Postnatal Apoptosis in Cerebellar Granule Cells of ...

1 downloads 0 Views 2MB Size Report
Keywords: P/Q-type calcium channels; Fluoro-Jade staining;. TUNEL staining ...... Westenbroek RE, T Sakurai, EM Elliott, JW Hell, TVB Starr, TP. Snutch and WA ...
Neurotoxicity Research, 2004, 2004, VOL. VOL.6(5). 6(4).pp. pp.001-014 267-280

F.P. Graham Publishing Co.

Postnatal Apoptosis in Cerebellar Granule Cells of Homozygous Leaner (tgla/tgla) Mice FRANCIS C. LAUa, TAMY C. FRANKb, SANG-SOEP NAHMd, GHEORGHE STOICAc and LOUISE C. ABBOTTb,* aUSDA, HNRCA, Tufts University, 711 Washington St., Boston, MA 02111, USA; bDepartment of Veterinary Anatomy and Public Health and cDepartment of Veterinary Pathobiology, College of Veterinary Medicine, Texas A & M University, 4458 TAMU, College Station, TX 77843-4458, USA; dDepartment of Human Anatomy and Medical Neurobiology, College of Medicine, Texas A&M University System Health Science Center, 1114 TAMU, College Station, TX 77843-1114, USA. [email protected] (Received 17 March 2004; Revised 17 July 2004; In final form 17 July 2004)

Leaner mice carry a homozygous, autosomal recessive mutation in the mouse CACNA1A gene encoding the α1A subunit of P/Q-type calcium channels, which results in an out-of-frame splicing event in the carboxy terminus of the α1A protein. Leaner mice exhibit severe ataxia, paroxysmal dyskinesia and absence seizures. Functional studies have revealed a marked decrease in calcium currents through leaner P/Q-type channels and altered neuronal calcium ion homeostasis in cerebellar Purkinje cells. Histopathological studies of leaner mice have revealed extensive postnatal cerebellar Purkinje and granule cell loss. We examined the temporospatial pattern of cerebellar granule cell death in the leaner mouse between postnatal days (P) 10 and 40. Our observations clearly indicate that leaner cerebellar granule cells die via an apoptotic process and that the peak time of neuronal death is P20. We did not observe a significant increase in microglial and astrocytic responses at P20, suggesting that glial responses are not a cause of neuronal cell death. We propose that the leaner cerebellar granule cell represents an in vivo animal model for low intracellular [Ca2+]-induced apoptosis. Since intracellular [Ca2+] is critical in the control of gene expression, it is quite likely that reduced intracellular [Ca2+] could activate a lethal cascade of altered gene expression leading to the apoptotic granule cell death in the leaner cerebellum. Keywords: P/Q-type calcium channels; Fluoro-Jade staining; TUNEL staining; Immunohistochemistry; Electron microscopy

INTRODUCTION Apoptosis or programmed cell death takes place in the normal developing central nervous system (CNS) to eliminate over-produced neurons (Oppenheim, 1991). Without appropriate amounts and timing of neuronal cell death, normal CNS formation cannot take place (Raff, 1992; Juurlink and Hertz, 1993; Sugimoto et al., 1994; Meier et al., 2000). It has been estimated that as many as 70% of some categories of neurons in the vertebrate nervous system die during the process of normal development (Raff et al., 1993; Vaudry et al., 2003). Failure of physiological neuronal cell death during development can result in severe CNS disorders. One example is medulloblastoma, which is a cerebellar tumor that arises through abnormal cerebellar granule cell proliferation and decreased apoptosis (Kim et al., 2003). On the other hand, increased apoptosis during CNS development is thought to occur in the disease process taking place in individuals with Down syndrome (Lubec and Engidawork, 2002). Equally important, however, is the occurrence of excessive neuronal cell death underlying a wide range of devastating adult onset neurodegenerative diseases such as Alzheimer's, Huntington's and Parkinson's diseases (Portera-Cailliau et al., 1995; Smale et al., 1995; Su et al., 1997; Marx, 2001; Fiskum et al., 2003; Gastard et al., 2003; Su et al., 2003). It has been shown that apoptosis is the common pathway of neuronal cell death in a number of cerebellar mutant mice, including weaver (wv), lurcher (lc) and Purkinje cell degeneration (pcd), making these mutant mice important models to study the processes by which neuronal cell death is accomplished (Rakic

*Corresponding author. Tel.: +1 (979) 845-2269; Fax: +1 (979) 847-8981; E-mail: [email protected] ISSN 1029 8428 print/ ISSN 1476-3524 online. © 2004 FP Graham Publishing Co., www.fpgrahamco.com

268

F.C. LAU et al.

and Sidman, 1973; Migheli et al., 1995; Norman et al., 1995; Wullner et al., 1995; Zhang et al., 1997). These mice all exhibit cerebellar ataxia to varying degrees. Postmortem studies in human patients with ataxia often reveal loss of cerebellar neurons. Recent studies have revealed at least two types of hereditary cerebellar ataxia in humans, including episodic ataxia type-2 and spinocerebellar ataxia type-6, which are caused by mutations in the gene encoding the α1A subunit of human calcium ion channels, designated CACNA1A. The α1A subunit is the pore forming subunit of P/Qtype voltage-gated calcium ion channels (Westenbroek et al., 1995; Gillard et al., 1997). Patients with these mutations exhibit progressive ataxia with signs of cerebellar atrophy (Ophoff et al., 1996; Zhuchenko et al., 1997). More recently, Jouvenceau et al. (2001) reported that a mutation in the coding region of the human CACNA1A gene resulted in early-onset absence epilepsy and cerebellar ataxia. The phenotype of this human disorder bears striking parallels to a mouse cerebellar ataxia model called the leaner mouse. The leaner mouse carries a homozygous, autosomal recessive mutation in the gene encoding the α1A subunit of mouse P/Q-type calcium channels. The mutation in the leaner mouse CACNA1A gene is located in a splice donor consensus sequence of the α1A subunit and results in an out-of-frame splicing event in the carboxy terminus of the α1A protein (Fletcher et al., 1996). Leaner mice exhibit severe ataxia, paroxysmal dyskinesia and absence seizures (Sidman et al., 1965; Meier and MacPike, 1971). While α1A mRNA and protein expression are not altered in leaner mouse cerebellar neurons (Lau et al., 1998), functional studies have revealed a marked decrease in calcium currents through leaner P/Q-type channels (Dove et al., 1998; Lorenzon et al., 1998) and altered neuronal calcium ion homeostasis (Dove et al., 2000; Murchison et al., 2002). Histopathological studies of leaner mice have revealed extensive postnatal cerebellar Purkinje and granule cell loss as well as atrophy of the inferior olive (Herrup and Wilczynski, 1982; Heckroth and Abbott, 1994). Thus, the genetic and pathophysiological characteristics of leaner mice provide unique opportunities to study cerebellar dysfunction and, specifically, cerebellar neurodegeneration related to calcium channel mutations. We examined the temporospatial pattern of cerebellar granule cell death in the leaner mouse between postnatal days (P) 10 and 40 and determined that the peak time of neuronal cell death occurred at P20. Based on information gained from this investigation we will be able to carry out a more detailed investigation of the molecular mechanisms that produce the

observed pattern of cerebellar granule cell death in the leaner cerebellum.

MATERIALS AND METHODS Animals Mice with the control genotype C57BL/6J:+/+ (wild type), or mutant genotype, C57BL/6J: tgla/tgla (homozygous leaner) or heterozygous leaner mice that were also heterozygous for the dominant mutation, Oligosyndactyly (C57BL/6J:tgla/+:+/Os) were housed in the Laboratory Animal Research and Resource building at Texas A&M University. Male and female heterozygous leaner (tgla/+:+/Os) or homozygous leaner mice (tgla/tgla) were bred to produce homozygous tgla/tgla offspring. The gene that carries the mutation resulting in oligosyndactyly is closely linked to the tg locus so very few crossing over events occur between the two genes (Isaacs and Abbott, 1992). Oligosyndactyly causes fusion of digits, and embryos that are homozygous for Oligosyndactyly (Os/Os) die in utero. Therefore, we were able to use the presence of fused digits on newborn pups to determine whether they are heterozygous leaners (they have fused digits) or homozygous leaner pups (they have the normal number of digits). This is necessary in order to determine the genotype of the youngest mice examined (P10), which is just before the time of onset of ataxia (P12-15) in homozygous leaner mice. All mice were housed and bred in a constant-temperature room (21-22°C) with 12-hour light/dark cycle and allowed access to food and water ad libitum. All mice were weaned between 30 to 40 days of age. Male and female wild type and homozygous tgla/tgla mice between the ages of P10 and P40 were used in this study. Homozygous leaner pups often die of hypothermia and dehydration at approximately three weeks of age due to their neurological disorders, which limits their ability to move about the cage and adequately access food and water. Consequently, homozygous leaner mice were kept alive by fostering newborn homozygous leaner pups to lactating Swiss White Webster female mice. In addition, moistened rodent chow was placed into the cages starting when leaner pups were approximately 18 days old and changed daily. Weaned leaner mice had their diet of dry rodent chow and water, given ad libitum, supplemented with moistened rodent chow, which was changed daily. All experimental procedures were carried out in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals

APOPTOSIS IN LEANER MOUSE CEREBELLAR GRANULE CELLS

(NIH Publication No.85-23, revised, 1996), and the minimum number of animals necessary for each experiment were used. Fluoro-Jade Staining Mice were anesthetized at the appropriate age (P10, 20, 30 and 40; n = 3 to 6 for each time point and genotype) with 3.0 mg/kg ketamine plus 0.5 mg/kg xylazine per 20 g body weight, given intraperitoneally (IP), then briefly perfused with Tyrode's saline followed by 4% paraformaldehyde in 0.1 M phosphate buffer (PB, pH 7.4). The brains were removed, and some were paraffin embedded and others were cryoprotected in 20% sucrose in PB, rapidly frozen on powdered dry ice and then stored at -70°C until sectioned. The frozen, cryoprotected brains were cut sagittally at 25 µm on a SLEE cryostat (Mainz, Germany) and stored at -70°C until used. The frozen, slide-mounted sections were thawed, dried in an oven at 54°C for 2-3 hours, and cooled to room temperature. The paraffin-embedded brains were cut sagittally through the region of the vermis and paravermis at 5.0 µm, placed on poly-L-lysine coated glass slides and deparaffinized in two changes of xylene followed by two changes of 100% ethanol. Following the protocol described by Schmued et al. (1997), slides with either frozen sections or paraffin sections were immersed in 100% ethanol, then 70% ethanol followed by deionized water. They were then incubated in 0.06% potassium permanganate for 15 minutes, rinsed in deionized water and incubated in 0.001% Fluoro-Jade in 0.1% acetic acid in deionized water for 30 minutes. After staining, slides were rinsed in three 1 minute washes of deionized water, thoroughly dried with a hot air gun, immersed in 2 changes of xylene, and coverslipped with DPX mounting media (Electron Microscopy Sciences, Ft. Washington, PA, USA). Since apoptosis occurs normally and continuously in spermatogonia throughout life (Lee et al., 1997), two slides with 25 µm testes sections were included with each staining group as a positive staining control. TUNEL Staining For in situ labeling of DNA strand breaks, mice were anesthetized as described above, and at the appropriate age (P10, 15, 20, 30 and 40; n = 4 or 5 for each time point and genotype) perfused as described above. The perfused bodies were stored at 4°C for 4 hours then the brains were removed and stored in fresh fixative at 4°C for 24 hours. The cerebella were embedded in paraffin and sectioned sagittally at 5.0 µm through the vermis. Representative sections were stained with hematoxylin and eosin for histological examination. Other paraffin

269 3

sections were processed for use with an in situ apoptosis detection kit (Oncor, Inc., Gaithersburg, MD, USA) according to the instructions provided by the manufacturer. Briefly, sections were deparaffinized, delipidated and then treated with 20 µg/ml proteinase K (Amersham Biosciences, Piscataway, NJ, USA). Endogenous peroxidase activity was quenched with 2% hydrogen peroxide (H2O2). The sections were rinsed with phosphate buffered saline (PBS; pH 7.4), then incubated with terminal deoxynucleotidyl transferase (Tdt) and digoxigenin-11-dUTP at 37°C in a humidified chamber for 1 hour. The reaction was stopped with washing buffer provided in the kit. The sections were rinsed with PBS and incubated with anti-digoxigeninperoxidase for 1 hour at room temperature. After washing in PBS, the color reaction was carried out using 0.05% 3,3-diaminobenzidine (DAB; Sigma, St. Louis, MO, USA) as the chromogen. The slides were counterstained with methyl green (0.5% in 0.1 M sodium acetate, pH 4.0). Sections from adult mouse testes were used as positive and negative controls. In each experiment both positive and negative control testes sections were processed under identical conditions as cerebellar sections except for the negative control sections H2O was used in place of Tdt enzyme. Cells were considered positive for apoptosis if they were isolated, the nuclei stained brown and showed signs of chromatin condensation. For quantification, the number of apoptotic granule cells per midsagittal cerebellar section was analyzed. To compare regional differences in the presence of apoptotic granule cells, the cerebellar sections were divided into the anterior lobe (lobules I through V) and posterior lobe (lobules VI through X) and the number of apoptotic cells per lobule recorded. Coronal sections through wild type and leaner cerebella also were processed to determine whether apoptotic granule cells were compartmentalized in either an anterior to posterior or medial to lateral manner. Electron Microscopy Male and female P20, 30 and 40 wild type and leaner (n = 3 for each genotype and age) were anesthetized with ketamine/xylazine (IP) as described above and perfused intracardially with 250 ml of 1% paraformaldehyde/1% glutaraldehyde (Electron Microscopy Sciences) in 0.12 M PB (pH 7.4). The protocol of Abbott and Sotelo (2000) was followed when preparing samples for electron microscopy. The fixed cerebellar samples were embedded in Araldite and ultrathin sections (70-90 nm) were mounted on uncoated copper grids and observed using a Zeiss 10C electron microscope. Ultrathin sec-

4270

F.C. LAU et al.

FIGURE 1 Photomicrographs of cerebellar sections stained with Fluoro-Jade. Sections were taken from P20 wild type (A) and leaner (B) mice. M = molecular layer; PC = Purkinje cell layer; g = granule cell layer; w = white matter layer. Note the absence of any obvious FluoroJade-positive cells in the granule cell layer of the wild type cerebellar section (A). In contrast, there are many Fluoro-Jade positive cells in the granule cell layer of the leaner cerebellar section (B). The inset in B shows a higher magnification of the leaner cerebellar granule cell layer. Scale bar in B = 300 µm is the same for both A and B. Scale bar in the inset in B = 60 µm.

tions from the rostral and caudal vermis of each animal were observed and photographed. Lectin Histochemistry Male and female P20 and P30 mice (n = 3 for each group at each age) were anesthetized and perfused as described for Fluoro-Jade staining. For labeling of microglial cells the protocol of Stoica et al. (1993) was used. The paraformaldehyde fixed brains were processed for paraffin embedding. Sections were cut at 5 µm and placed on poly-L-lysine coated microscope slides. The sections were deparaffinized and incubated for 1 hour at room temperature with 15 µg/ml of biotinylated lectin, RCA-I (Vector Laboratories, Burlingame, CA, USA) in Tris buffer with Tween-20 (TTB, 0.05 M Tris base and 0.1% Tween-20 at pH 7.5). Then the sections were washed three times in TTB plus 0.9% NaCl and the red chromogen color was developed using an alkaline phosphatase kit (Vector Laboratories), then dehydrated and coverslipped as described previously. For quantification the RCA-1 stained sections were scored by an investigator blinded to the age and genotype of the sections being examined. For each section the number of RCA-I-positive, non-endothelial cells were counted in the granule cell layer of each lobule. Four to six non-adjacent sections were counted for each animal and then averaged. Immunohistochemistry The standard immunohistochemistry protocol of Abbott and Jacobowitz (1995) was followed for glial fibrillary acidic protein (GFAP) and activated caspase3 staining. Frozen sections from sets of P20 and P30 male and female mice (n = 3 to 4 for each genotype and

age) were used. Animals were anesthetized and perfused as described previously in the Fluoro-Jade staining section. The perfused brains were cryoprotected with 20% sucrose in PB, rapidly frozen using powdered dry ice and 20 µm sagittal frozen sections were cut in a SLEE cryostat. The sections were treated with 1% H2O2 to quench endogenous peroxidase activity, then blocked with 5% normal goat serum in PBS (Sigma). The sections were incubated with anti-GFAP serum (1:1,000 dilution, Dako, Glostrup, Denmark) or antiactivated caspase-3 serum (1:25,000 dilution, R&D Systems, Minneapolis, MN, USA), followed by biotinylated secondary antibody and peroxidaselabeled streptavidin (Vector Laboratories). Tissuebound peroxidase activity was visualized with 0.024% DAB (Sigma) and 0.006% H2O2 in 0.05 M Tris-HCl buffer (pH 7.6). The DAB reaction was terminated by transferring the sections to 0.05 M Tris-HCl buffer. Following termination of the DAB reaction, the sections were rinsed twice in PBS and dehydrated through an ascending series of ethyl alcohol, then xylene and coverslipped using Permount (Fisher, Pittsburgh, PA, USA). Sections from both genotypes were processed simultaneously in order to ensure equal exposure to all solutions. For quantification for both GFAP immunoreactivity and activated caspase-3 immunopositive cells the sections were scored by an investigator blinded to the age and/or genotype of the sections being examined. To assess the number of activated caspase-3-positive cells present in the granule cell layer of each section, the number of positive cells were counted in each lobule. Four to six non-adjacent sections were counted for each animal and then averaged. GFAP-immunoreactivity was quantified using a densitometry program.

APOPTOSIS IN LEANER MOUSE CEREBELLAR GRANULE CELLS

2715

FIGURE 2 Photomicrographs of TUNEL-positive cerebellar granule cells in wild type (A) and homozygous leaner (B-D) mice at P20. Black arrows point to TUNEL-positive nuclei. Red arrows point to small, pyknotic nuclei stained with methyl green that are visible in Figure C. Note the few TUNEL-positive nuclei in the cerebellar section from a wild type mouse (A). Figure D shows a single TUNEL-positive leaner cerebellar granule cell nucleus at higher magnification. w = white matter; PC = Purkinje cell layer; g = granule cell layer; c = capillary. Scale bar in B = 150 µm (same scale for A and C); scale bar in D = 20 µm.

Images from anterior and posterior lobules from each section were captured as tiff files using Adobe Photoshop 7.0. Uniform lighting and magnification parameters were used for all sections imaged. The spot densitometry program, AlphaEaseFC, (Alpha Innotech, San Leandro, CA, USA) was used to quantify the intensity of GFAP immunopositive staining in the granule cell layer in each section examined. A standardized rectangle (45,000 µm2) was applied to the granule cell layer and the optical density was recorded. Four areas in the anterior cerebellum and 4 areas in the posterior cerebellum were measured on each section. Background staining was determined by measuring the optical density over five Purkinje cell somata (400 µm2) observed in each section because Purkinje cells do not contain any GFAP. The average background value was subtracted from the measurements made from the granule cell layer. Four to six non-adjacent sections were counted for each animal and then averaged.

Statistics Data are presented as means ± SEM. For two-sample comparisons, the Students t-test was used. For multiple-sample comparisons, a two-way ANOVA was used followed by a Tukey's post hoc test. α was set at 0.05. RESULTS Temporal Pattern of Cell Death We first examined the temporal pattern of cell death occurring in the postnatal leaner cerebellum. We stained both wild type and leaner cerebellar sections taken from mice at postnatal days 10, 20, 30 and 40 with Fluoro-Jade staining, which is not specific for any one type of cell death; cells dying via either necrosis or apoptosis will stain positive with Fluoro-Jade (Schmued et al., 1997). We observed a high proportion of leaner granule cells but not wild type granule cells staining with Fluoro-Jade at P20 (FIG. 1). We did not observe very many Fluoro-Jade positive leaner granule

F.C. LAU et al.

272 6

cells at P10 (data not shown). We also observed that the number of Fluoro-Jade-positive leaner granule cells decreased at the older ages examined (P30 and P40; data not shown). Very few Fluoro-Jade-positive granule cells were observed in sections from wild type cerebella at all ages examined (see FIG. 1 for P20 wild type cerebellar section). To further investigate the mode of leaner granule cell death, we used TUNEL staining, which identifies cells with numerous breaks in the nuclear DNA, and is thought to be a marker of apoptosis (Enari et al., 1998; Lubec and Engidawork, 2002).

Evidence for Apoptotic Granule Cell Death in the Leaner Cerebellum TUNEL-positive cerebellar granule cells were observed during postnatal cerebellar development in both wild type and leaner mice at all ages examined (FIG. 2A-2B and 2D). However, no TUNEL-positive Purkinje cells were observed in either wild type or leaner mice at any age examined. Very few TUNELpositive granule cells were found in the external granule cell layer (EGL) in either genotype at P10 and P15. By P20 the external granule layer had essentially disappeared in both wild type and leaner mice.

80 70

*,#

60

Mean # of 50 TUNEL-positive 40 nuclei per cerebellar 30 section

wild type leaner

*

20

*

10 0 P10

P15

P20

P30

P40

Postnatal Age FIGURE 3 Graph showing the mean number of TUNEL-positive granule cell nuclei per cerebellar section for both wild type (black bars) and homozygous leaner (light bars) mice at postnatal days 10 (P10) through P40. Data are presented as means ± SEM; n = 4 or 5 for each genotype and age. The * indicates a significant difference between genotypes at the same age, while the # indicates that the number of TUNEL-positive cerebellar granule cells in leaner mice is significantly greater at P20 than at all other ages of leaner mice that were examined. A two way ANOVA was significant at α = 0.05, which was followed by a post hoc Tukey's test (p = 0.05). 60 50

*

40

Mean # of TUNELpositive nuclei per 30 leaner mouse cerebellar section

anterior lobe rostral lobe lobe posterior caudal lobe

20

*

10 0 P10

P15

P20

P30

P40

Postnatal Age FIGURE 4 Graph showing the mean number of TUNEL-positive granule cell nuclei per cerebellar section for different regions of the homozygous leaner cerebellum at postnatal days 10 (P10) through P40. Data are presented as means ± SEM; n = 4 or 5 for each time point examined. The dark bars indicate cell counts taken in the anterior lobe of the leaner cerebellum (black bars), while the gray bars indicate cell counts taken in the posterior lobe of the leaner cerebellum. A two way ANOVA was significant at α = 0.05, which was followed by a posthoc Tukey's test (p = 0.05). The * indicates a significant difference between the rostral and caudal cerebellar lobes at the same age.

APOPTOSIS IN LEANER MOUSE CEREBELLAR GRANULE CELLS

A

B

2737

C

FIGURE 5 Transmission electron photomicrographs of wild type (A) and leaner (B & C) cerebellar granule cells from mice at P20. The wild type granule cell nuclei in A are normal. Note the large amount of euchromatin and relatively few peripherally located clumps of heterochromatin within the nuclei. Several of the leaner granule cell nuclei in B (denoted by an *) show increased condensation of the nuclear chromatin. Figure C shows a leaner granule cell with a completely condensed or pyknotic nucleus. Note that the cytoplasm of this granule cell is still relatively normal. Scale bar in C = 10 µm; the scale is the same for A-C.

There was no difference in the number of TUNELpositive granule cells observed between wild type and leaner mice at P10 FIG. 3). However, the number of TUNEL-positive leaner granule cells was significantly higher than observed in wild type mice at P20, P30 and P40. In leaner cerebella, the number of TUNEL-positive granule cells was significantly higher at P20 when all the ages of leaner mice were compared and the number of TUNEL-positive granule cells decreased dramatically at P30 and P40. The number of TUNEL-positive granule cells was significantly higher in the anterior lobe compared to the posterior lobe of leaner cerebella at P20 and P30 (FIG. 4). This regional difference was not observed in wild type cerebella at any age nor was it present in leaner cerebella at P10, P15 or P40 (FIG. 4). Methyl green counterstaining revealed condensed (pyknotic) nuclei in the inner cerebellar granule of leaner mice at all ages. The pyknotic nuclei were numerous in leaner cerebella starting at P20, then decreased at the older ages (FIG. 2C). However, condensed, methyl green-stained granule cell nuclei were seldom observed in wild type cerebella at any age examined (FIG. 2A). We also examined the ultrastructural morphology of cerebellar granule cells in P20 wild type and leaner mice using electron microscopy (FIG. 5A-5C). The ultrastructural morphology of dying leaner granule cells showed several hallmarks of apoptosis (Kerr, 1972) including the appearance of dense chromatin clumps in the nuclei (FIG. 5B), which became more prominent until the nucleus appeared as completely condensed chromatin or pyknotic (FIG. 5C). The plasma membranes of granule cells with clumped and/or condensed chromatin were frequently convoluted, but there was no evidence of swelling or loss of integrity of the cellular organelles (FIG. 5B and 5C).

Examination for the Presence of Microglia and Astrocytes in the Leaner Cerebellum It is possible that increased numbers of glial cells could account for some of the small, dark nuclei stained with methyl green that were observed in leaner cerebella. To address this question we first qualitatively examined sections for the presence of microglia in leaner and wild type cerebella at P20 and P30, using histochemistry, specifically the biotinylated lectin, RCA-I, which binds to microglia and endothelial cells in the CNS (Stoica et al., 1993). There was no obvious difference in the number of microglia that stained positive for RCA-I between wild type and leaner mice at P20; very few microglia were positively stained in either genotype. The microglia also did not appear to be activated or reactive in either genotype at P20. However, RCA-I staining at P30 appeared to stain more RCA-1-positive microglia in the leaner cerebellar granule cell layer than observed in age-matched wild type cerebella. In addition, some of the microglia that stained positive for RCA-1 in the P30 leaner cerebella appeared to have larger, more obvious cytoplasmic projections than those observed in the age-matched wild type cerebella (FIG. 6A-6D). However, when the actual number of microglia were counted in sections from the two different ages and two genotypes the ANOVA was not significant (FIG. 6E, p=0.15). The same general observations also were true for GFAP immunohistochemical staining, which specifically stains astrocytes. There were no obvious differences in the number of GFAP-positive cells or the intensity of staining observed at P20 when leaner and wild type cerebella were compared. When we qualitatively assessed the sections, there appeared to be increases in the intensity GFAP-positive staining in the leaner cerebellum at P30 when compared to age-

F.C. LAU et al.

8 274

30.0

E 25.0

20.0

average # of cells 15.0 per slice 10.0

5.0

0.0 WT 20

LA 20

WT 30

LA 30

genotype and age

FIGURE 6 Light microscopic photomicrographs of microglia that are positive for RCA-1 histochemical staining in wild type (WT) (A and C) and homozygous leaner (LA) (B and D) mice at P30. The arrows indicate RCA-1-positive microglial cells. c = capillaries whose endothelial cells also stain positive for RCA-1; w = white matter; g = granule cell layer; PC = Purkinje cell layer. Scale bar for photomicrographs A and B = 100 µm; scale bar for C and D = 25 µm. Graph in E shows the quantification of the number of RCA-1-positive microglial cells in the cerebellar granule cell layer in wild type and leaner mice at P20 and P30. No significant differences were observed (p = 0.15), although there was a trend towards increased numbers of microglia present in the leaner cerebellar granule cell layer, especially at P30. Data are presented as means ± SEM; n = 3 for each group examined.

matched wild type cerebella (FIG. 7A-7D). However, when we quantified the density of GFAP immunopositive staining in the granule cell layer of P20 and P30 wild type and leaner mice, the difference in staining intensity was not significant (FIG. 7E, p= 0.53). Expression of Activated Caspase-3 in Leaner Cerebellar Granule Cells We have presented evidence that leaner cerebellar granule cells die via apoptosis, which is a concerted, energy-dependent process requiring discrete signaling cascades, depending on the nature of the pro-apoptotic stimuli (Hengartner, 2000). Disruption of intracellular [Ca2+] homeostasis is known to lead to apoptotic cell death (Haughey et al., 2002; Li et al., 2003; Yoon et

al., 2003). One common cell death signaling pathway that is triggered by altered intracellular [Ca2+] homeostasis involves caspases, with activation of caspase-3 as a primary "effector" protease during apoptosis (Alavez et al., 2003). We used immunohistochemistry to examine cerebellar granule cells for the presence of activated caspase-3 and observed a large number of activated caspase-3-positive granule cells in the cerebella of P20 leaner mice when compared to agematched wild type mice (FIG. 8). We quantified the number of activated caspase-3-positive cells in the granule cell layer from wild type and leaner mice at P20, since this was the peak time of granule cell death identified using TUNEL staining (FIG. 3). We observed a significant effect with respect to genotype

APOPTOSIS IN LEANER MOUSE CEREBELLAR GRANULE CELLS

A

B

C

D

2759

PC

w g g

w PC 70

E

60 50

optical density

40 30 20 10 0 WT20

LA20

WT30

LA30

genotype and age

FIGURE 7 Light microscopic photomicrographs of microglia that are positive for GFAP immunohistochemical staining in wild type (WT) (A and C) and homozygous leaner (LA) (B and D) mice at P30. Arrows indicate several GFAP-positive astrocytes. w = white matter; g = granule cell layer; PC = Purkinje cell layer; M = molecular layer. Scale bar for photomicrographs A and B = 100 µm; scale bar for C and D = 25 µm. Graph in E shows the quantification of intensity of GFAP immunohistochemical staining for wild type and leaner mice at P20 and P30. No significant differences were observed (p = 0.53). Data are presented as means ± SEM; n = 3 for each genotype and age examined.

(p=0.005) but not with respect to anterior versus posterior cerebellum (p= 0.21), with many more activated caspase-3-positive cells observed in the leaner mouse cerebellar granule cell layer when compared to wild type mice. While the numbers of activated caspase 3positive cerebellar granule cells observed in leaner cerebella at later ages were still noticeable, the differences between leaner mice and age-matched wild type mice were not as dramatic as observed at P20 (data not shown). DISCUSSION The observations reported here clearly indicate that

leaner cerebellar granule cells die via an apoptotic process and that the peak time of apoptosis is P20. We also observed that in early postnatal development, P10P15, apoptotic cell death occurred in the normal developing mouse cerebellum and that there was no difference in the occurrence of cerebellar granule cell apoptosis between leaner and wild type mice during these early days of postnatal development. This is in agreement with other previously published reports of postnatal cerebellar development in general (Lossi et al., 2002) and in normal and leaner mouse cerebellar postnatal development in particular (Fletcher et al., 1996). In the wild type mouse, the occurrence of cerebellar granule cell apoptosis started to decrease at P20 and by

276 10

F.C. LAU et al.

C

50

*

40

*

average 30 # of cells per section 20

*

10 0

WT-ant

WT-post

WT-total

LA-ant

LA-post

LA-total

genotype and region FIGURE 8 Photomicrographs of activated caspase-3 stained sections from wild type (A) and homozygous leaner (B) mice at P20. Arrows indicate caspase-3-positive cerebellar granule cells. Note the large number of caspase-3-positive granule cells in the leaner cerebellar section (B) while no positive neurons are visible in the wild type cerebellar section (B). The inset in Figure B shows a higher magnification of the leaner granule cell layer. Three activated caspase-3-positive granule cells are seen in the inset. M = molecular layer; PC = Purkinje cell layer; g = granule cell layer; w = white matter layer. Scale bar for photomicrographs A and B = 200 µm; scale bar in the inset in B = 50 µm. There were significantly more activated caspase-3-positive cells observed in the leaner cerebellar granule cell layer when compared to the wild type cerebellum (*, p = 0.005). The graph shown in C reveals a significant difference between the average number of activated caspase-3-positive cells in P20 leaner cerebella when compared to age-matched wild type mice. Data are presented as means ± SEM; n = 4 for each age group. There was no significant difference between the number of activated caspase-3-positive cells observed in the anterior leaner cerebellum when compared to the posterior leaner cerebellum (p=0.21). Data are presented as means ± SEM.

P40 TUNEL-positive granule cells were barely detectable. The increased fragmentation DNA indicated by TUNEL-positive cerebellar granule cells in leaner mice, which peaked at P20, correlates with the presence of a large number of pyknotic nuclei and FluoroJade-positive granule cells at this same time point. Apoptosis as the mode of leaner cerebellar granule cell death was confirmed ultrastructurally. Using electron microscopy, we observed highly condensed granule cell nuclei without significant swelling of cytoplasmic organelles in the leaner mouse cerebellum (FIG. 5). The excessive apoptotic cell death observed in the leaner mouse was not due to altered expression and/or localization of the mutated α1A mRNA or protein based on observations reported previously (Lau et al., 1998). It was noted that the greatest number of pyknotic nuclei

were observed at the same age as the peak number of TUNEL-positive nuclei, which was P20 (data not shown). It is possible that the increased number of dark-staining or pyknotic nuclei observed in the leaner cerebellum could be due to increased numbers of microglia and/or astrocytes and not represent only neurons dying via apoptosis. Both microglia and astrocytes will respond to many different stimuli by increasing in numbers and/or in their activation state. Using histochemistry we specifically stained for microglia using the biotinylated lectin, RCA-1. When we visualized microglia at P20 and at P30 we did not find any significant difference in staining characteristics between leaner mouse cerebella and cerebella from agematched wild type mice. These observations suggested that the increased numbers of pyknotic nuclei observed in the leaner mouse cerebellum at P20 were not prima-

APOPTOSIS IN LEANER MOUSE CEREBELLAR GRANULE CELLS

rily due to increased numbers of microglial cells. There was a tendency towards increased size of microglia and towards increased numbers in the granule cell layer in leaner mice at P30, but the trend did not reach statistical significance. Microglia are the immune cell population of the CNS (Polazzi et al., 2001) and their role in responding to CNS trauma or neuropathology in order to remove dead and dying cells or foreign agents has been well documented (Liu et al., 2001; Penkowa et al., 2001). It is quite likely that any increase in microglia after the peak time of leaner cerebellar neuronal cell death could be such a response. However, the range of reciprocal neuron-microglial interactions is not well understood, especially in the context of neuronal death and damage. It has been demonstrated that microglial cells can release molecules that are able to rescue neurons from apoptotic cell death (Polazzi et al., 2001). Therefore, it would be consistent that activation of microglia after the observed peak of apoptosis in the leaner cerebellum could help to limit cell death in the leaner cerebellum after P20. Activation of astrocytes in the CNS often has been associated with promotion of neuronal apoptosis (Hu et al., 1997; Cassia et al., 2002). On the other hand, specific astrocyte-derived trophic factors have been shown to increase neuronal survival in culture through inhibition of apoptotic cell death (Sortwell et al., 2000). Thus, it is tempting to speculate that activation of both microglia and astrocytes after P20 may participate in limiting the amount of apoptotic neuronal death taking place in the leaner cerebellum. The relationship between neuronal apoptosis and glial cell involvement is clearly complex. Since microglial and astrocyte changes appear to take place after peak leaner granule cell death, the numerous pyknotic nuclei observed in the leaner cerebellum at P20 were most likely nuclei of dying granule cells. The number of pyknotic nuclei and the number of FluoroJade positive cells observed in the leaner cerebellum were both greater than the number of TUNEL-positive granule cell nuclei observed in all of the cerebellar sections examined. It also is clear that TUNEL labeling only occurs in nuclei during a specific segment of the apoptotic process (Frank et al., 2003). What is likely taking place in the leaner cerebellum is that a relatively small number of granule cells are TUNEL-positive at any one time, but a larger number of nuclei are pyknotic or positive for Fluoro-Jade and may remain visible for a relatively longer period of time before they are destroyed. When cerebellar granule cells are instructed to die they will activate several final common elements of

27711

their death-inducing machinery (Vaudry et al., 2003). While neuronal cell death may occur through diverse mechanisms, apoptosis is one of the three main types of cell death, the other two being necrosis and autophagy (Yuan et al., 2003). The key pathway involved in apoptotic cell death involves three categories of molecules: 1) the family of cysteine proteases called caspases; 2) adaptor proteins that act as activators of caspases; and 3) the Bcl-2 family of mitochondrial-associated proteins (Yuan and Yanker, 2000). One of the core components of the caspase cascade is caspase-3, which, when activated, carries out the execution steps of apoptosis by cleaving a number of downstream substrates. We observed that caspase-3 was activated in leaner cerebellar granule cells at P20 much more frequently than in age-matched wild type granule cells (FIG. 8). Activated caspase-3 is capable of cleaving ICAD (inhibitor of caspase-activated DNAase) and activating CAD (caspase-activated DNAase), which results in DNA fragmentation (Enari et al., 1998). Activated caspase-3 causes the breakdown of additional cellular components such as Bcl-2, actin, and PARP (polyADP-ribose polymerase), which bring about at least some of the morphological and biochemical alterations associated with apoptosis (Vaudry et al., 2003). The observation of increased numbers of leaner cerebellar granule cells that are positive for activated caspase-3 supports the conclusion that leaner cerebellar cells die via the process of apoptosis. Of course there are many possible "upstream" molecules that are either regulators of apoptosis or transducers of apoptosis that are involved in the entire cell death cascade taking place in leaner cerebellar granule cells. We have yet to elucidate these additional molecules. It seems likely that the observed leaner cerebellar granule cell death is a result of altered α1A function due to the truncation of the carboxyl terminus of the α1A protein (Fletcher et al., 1996). Recent studies have shown that the β subunit of voltage-gated calcium channels and G-protein modulation of P/Q-type channel function takes place through interaction with the carboxyl terminus of the α1A subunit (Mark et al., 2000; Zhou et al., 2003). The leaner mutation could very likely result in abnormal modulation of P/Q-type channel function via the β subunits and/or G-proteins because of the abnormal carboxy terminal structure that may interfere with subunit interactions. Several interesting questions that are not yet answered concern the timing of granule cell death and the number of dying granule cells in the leaner cerebellum. Why is postnatal day 20 the time of peak granule cell death and why do some granule cells survive and

12 278

F.C. LAU et al.

other granule cells die when all granule cells apparently express the mutated form of the α1A protein? It is possible that cerebellar granule cells do not start dying until after P/Q-type channels are expressed. Meacham et al. (2003) reported that in rat cerebellar Purkinje cells and granule cells, α1A subunit protein expression is low in the first postnatal week and then expression increases until maximal expression is observed at P21. This would suggest that the effects of the altered α1A subunit protein in the leaner cerebellum may not be observed until the channels utilizing the protein (P/Qtype calcium channels) are functioning. The time of onset of neuronal function also may be important. The rodent cerebellum is fully formed at approximately 2021 days and it is known that normal synaptic function and calcium ion channel activation are important for neuronal survival (Toescu, 1999; Verhage et al., 2000). Because P/Q-type calcium ion channels are very important for neurotransmission, we predict that neuronal signaling is not normal in the leaner cerebellum where P/Q-type channels are highly expressed by granule cells and Purkinje cells. It is quite possible that inadequate or inappropriate neurotransmission is causing approximately half of the cerebellar granule cells to die while the remaining granule cells perhaps receive enough input to survive. Finally, it is possible that lack of trophic factor production may result in significant granule cell death, although levels of neuronal trophic factors such as BDNF or NGF have not been assessed in the developing leaner cerebellum. Calcium ions are involved in many diverse aspects of cell functions so it is not surprising that disruption of neuronal calcium homeostasis could lead to neuronal cell death. Many in vitro experiments suggest that intracellular [Ca2+] overload will lead to cell death via apoptosis (reviewed by Mattson, 2000; Sastry and Rau, 2000; Annunziato et al., 2003). However, too little intracellular [Ca2+] also can be detrimental to neuronal survival and specifically cerebellar granule cell survival, at least in vitro (Ichikawa et al., 1998; Tabuchi et al., 2003). We propose that the leaner cerebellar granule cell represents an in vivo animal model for low intracellular [Ca2+]-induced apoptosis. Since intracellular [Ca2+] is critical in the control of gene expression, it is quite likely that reduced intracellular [Ca2+] could activate a lethal cascade of altered gene expression leading to the apoptotic granule cell death taking place in the leaner cerebellum. Identification of genes and ultimately the proteins involved in this cascade will provide valuable information to further our understanding of the role of intracellular [Ca2+] in neuronal survival and death in the CNS.

Acknowledgements Supported in part by CERH NIEHS grant P30EF09106 to LCA. The authors would like to thank Ms. Marla Jurek and Dr. Amy Brandl for their help with staining and quantification of TUNEL-positive cerebellar granule cells. The authors also gratefully acknowledge use of the facilities maintained by the Image Analysis Laboratory of the College of Veterinary Medicine. References Abbott LC and DM Jacobowitz (1995) Development of calretininimmunoreactive unipolar brush-like cells and an afferent pathway to the embryonic and early postnatal mouse cerebellum. Anat. Embryol. 191, 541-559. Abbott LC and C Sotelo (2000) Ultrastructural analysis of catecholaminergic innervation in weaver and normal mouse cerebellar cortices. J. Comp. Neurol. 426, 316-329. Alavez S, D Pedroza and J Moran (2003) Mechanisms of cell death by deprivation of depolarizing conditions during cerebellar granule neurons maturation. Neurochemistry 43, 581-590. Annunziato L, S Amoroso, A Pannaccione, M Cataldi, G Pignataro, A D'Alessio, R Sirabella, A Secondo, L Sibaud and GF Di Renzo (2003) Apoptosis induced in neuronal cells by oxidative stress: role played by caspases and intracellular calcium ions. Toxicol. Lett. 139, 125-133. Cassia P, H Peluffo, M Pehar, L Martinez-Palma, A. Ressia, JS Beckman, AG Estevez and L Barbeito (2002) Peroxynitrite triggers a phenotypic transformation in spinal cord astrocytes that induces motor neuron apoptosis. J. Neurosci. Res. 67, 21-29. Dove LS, LC Abbott and WH Griffith (1998) Whole-cell and single-channel analysis of P-type calcium currents in cerebellar Purkinje cells of leaner mutant mice. J. Neurosci. 18, 7687-7699. Dove LS, SS Nahm, D Murchison, LC Abbott and WH Griffith (2000) Altered calcium homeostasis in cerebellar Purkinje cells of leaner mutant mice. J. Neurophysiol. 84, 513-524. Enari M, H Sakahira, H Yokoyama, K Okawa, A Iwamatsu and S Nagata (1998) A caspase-activated DNase that degrades DNA during apoptosis, and its inhibitor ICAD. Nature 391, 43-50. Fiskum G, A Starkov, BM Polster and C Chinopoulos (2003) Mitochondrial mechanisms of neural cell death and neuroprotective interventions in Parkinson's disease. Ann. NY Acad. Sci. 991, 111-119. Fletcher CF, CM Lutz, TN O'Sullivan, JD Shaughnessy Jr, R Hawkes, WN Frankel, NG Copeland and NA Jenkins (1996) Absence epilepsy in tottering mutant mice is associated with calcium channel defects. Cell 87, 607-617. Frank TC, M Nunley, R Ramon and LC Abbott (2003) Fluoro-Jade identification of cerebellar granule cell and Purkinje cell death in the α1A calcium ion channel mutant mouse, leaner. Neuroscience 118, 667-680. Gastard MC, JC Troncoso and VE Koliatsos (2003) Caspase activation in the limbic cortex of subjects with early Alzheimer's disease. Ann. Neurol. 54, 393-398. Gillard SE, SG Volsen, W Smith, RE Beattie, D Bleakman and D Lodge (1997) Identification of the pore-forming subunit of calcium channels: an antisense study on rat cerebellar Purkinje cells in culture. Neuropharmacology 36, 405-409. Haughey NJ, A Nath, SL Chan, AC Borchard, MS Rao and MP

APOPTOSIS IN LEANER MOUSE CEREBELLAR GRANULE CELLS Mattson (2002) Disruption of neurogenesis by amyloid beta-peptide, and perturbed neural progenitor cell homeostasis, in models of Alzheimer's disease. J. Neurochem. 83, 1509-1524. Heckroth JA and LC Abbott (1994) Purkinje cell loss from alternating sagittal zones in the cerebellum of leaner mutant mice. Brain Res. 658, 93-104. Hengartner MO (2000) The biochemistry of apoptosis. Nature 407, 770-776. Herrup K and SL Wilczynski (1982) Cerebellar cell degeneration in the leaner mouse. Neuroscience 7, 2185-2196. Hu J, A Ferreira and LJ Van Eldik (1997) S100beta induces neuronal cell death through nitric oxide release from astrocytes. J. Neurochem. 69, 2294-2301. Ichikawa D, A Tabuchi, A Taoka, T Tsuchiya and M Tsuda (1998) Attenuation of cell death mediated by membrane depolarization different from that by exogenous BDNF in cultured cerebellar granule cells. Mol. Brain Res. 56, 218-226. Isaacs KR and LC Abbott (1992) Development of the paramedian lobule of the cerebellum in wild type and tottering mice. Dev. Neurosci. 14, 386-393. Jouvenceau A, LH Eunson, A Spauschus, V Ramesh, SM Zuberi, DM Kullmann and MG Hanna (2001) Human epilepsy associated with dysfunction of the brain P/Q-type calcium channel. Lancet 358, 801-807. Juurlink B and L Hertz (1993) Ischemia-induced death of astrocytes and neurons in primary culture: pitfalls in quantifying neuronal cell death. Dev. Brain Res. 71, 239-246. Kerr JF, AH Wyllie and AR Currie (1972) Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br. J. Cancer 26, 239-257. Kim JY, AL Nelson, SA Algon, O Graves, LM Sturla, LC Goumnerova, DH Rowitch, RA Segal and SL Pomeroy (2003) Medulloblastoma tumorigenesis diverges from cerebellar granule cell differentiation in patched heterozygous mice. Dev. Biol. 263, 50-66. Lau FC, LC Abbott, IJ Rhyu, DS Kim and HM Chin (1998) Expression of calcium channel α1A mRNA and protein in the leaner mouse (tgla/tgla) cerebellum. Mol. Brain Res. 59, 93-99. Lee JW, JH Richburg, SC Younkin and K Boekelheide (1997) The Fas system is a key regulator of germ cell apoptosis in the testis. Endocrinology 138, 2081-2988. Li SY, JH Ni, DS Xu and HT Jia (2003) Down-regulation of GluR2 is associated with Ca2+-dependent protease activities in kainateinduced apoptotic cell death in cultured rat hippocampal neurons. Neurosci. Lett. 352, 105-108. Liu B, K Wang, HM Gao, B Mandavilli, JY Wang and JS Hong (2001) Molecular consequences of activated microglia in the brain: overactivation induces apoptosis. J. Neurochem. 77, 182189. Lorenzon NM, CM Lutz, WN Frankel and KG Beam (1998) Altered calcium channel currents in Purkinje cells of the neurological mutant mouse leaner. J. Neurosci. 18, 4482-4489. Lossi L, S Mioletti and A Merighi (2002) Synapse-independent and synapse-dependent apoptosis of cerebellar granule cells in postnatal rabbits occur at two subsequent but partly overlapping developmental stages. Neuroscience 112, 509-523. Lubec G and E Engidawork (2002) The brain in Down syndrome (TRISOMY 21). J. Neurol. 249, 1347-1356. Mark MD, S Whittemann and S Herlitze (2000) G protein modulation of recombinant P/Q-type calcium channels by regulators of G protein signaling proteins. J. Physiol. 528, 65-77. Marx J (2001) Neuroscience. New leads on the 'how' of Alzheimer's. Science 293, 2192-2194.

27913

Mattson MP (2000) Apoptosis in neurodegenerative disorders. Nat. Rev. Mol. Cell Biol. 1, 120-129. Meacham, CA, LD White, S Barone Jr and TJ Shafer (2003) Ontogeny of voltage-sensitive calcium channel 1A and 1E subunit expression and synaptic function in rat central nervous system. Brain Res. Dev. Brain Res. 142, 47-65. Meier H and AD McPike (1971) Three syndromes produced by two mutant genes in the mouse. Clinical, pathological, and ultrastructural bases of tottering, leaner, and heterozygous mice. J. Heredity 62, 297-302. Meier P, A Finch and G Evan (2000) Apoptosis in development. Nature 407, 796-801. Migheli A, A Attanasio, WH See, SA Bayer and B Ghetti (1995) Detection of apoptosis in weaver cerebellum by electron microscopic in situ end-labeling of fragmented DNA. Neurosci. Lett. 199, 53-56. Murchison D, LS Dove, LC Abbott and WH Griffith (2002) Homeostatic compensation maintains Ca2+ signaling functions in Purkinje neurons in the leaner mutant mouse. Cerebellum 1, 119127. Norman DJ, L Feng, SS Cheng, J Gubbay, E Chan and N Heintz (1995) The lurcher gene induces apoptotic death in cerebellar Purkinje cells. Development 121, 1183-1193. Ophoff RA, GM Terwindt, MN Vergouwe, R van Eijk, PJ Oefner, SM Hoffman, JE Lamerdin, HW Mohrenweiser, DE Bulman, M Ferrari, J Haan, D Lindout, G-JB van Ommen, MH Hofker, MD Ferarri and RR Frants (1996) Familial hemiplegic migraine and episodic ataxia type-2 are caused by mutations in the Ca2+ channel gene CACNL1A4. Cell 87, 543-552. Oppenheim RW (1991) Cell death during development of the nervous system. Ann. Rev. Neurosci. 14, 453-501. Penkowa M, A Molinero, J Carrasco and J Hidalgo (2001) Interleukin-6 deficiency reduces the brain inflammatory response and increases oxidative stress and neuroprotection after kainic acid-induced seizures. Neuroscience 102, 805-818. Polazzi E, ET Gianni and A Contestablile (2001) Microglial cells protect cerebellar granule neurons from apoptosis: evidence for reciprocal signaling. Glia 36, 271-280. Portera-Caillaiu C, JC Hedreen, DL Price and VE Koliatsos (1995) Evidence for apoptotic cell death in Huntington disease and excitotoxic animal models. J. Neurosci. 15, 3775-3787. Raff MC (1992) Social controls on cell survival and cell death. Nature 356, 397-400. Raff MC, BA Barres, JF Burn, HS Coles, Y Ishizaki and MD Jacobson (1993) Programmed cell death and the control of cell survival: lessons from the nervous system. Science 262, 695-700. Rakic P and RL Sidman (1973) Weaver mutant mouse cerebellum: defective neuronal migration secondary to abnormality of Bergmann glia. Proc. Natl. Acad. Sci. USA 70, 240-244. Sastry PS and KS Rau (2000) Apoptosis and the nervous system. J. Neurochem. 74, 1-20. Schmued LC, C Albertson and W Slikker Jr (1997) Fluoro-Jade: a novel fluorochrome for the sensitive and reliable histochemical localization of neuronal degeneration. Brain Res. 751, 37-46. Sidman RL, MC Green and SH Appel (Eds.) (1965) Catalogue of the Neurological Mutants of the Mouse (Harvard University Press: Cambridge). Smale G, NR Nichols, DR Brady, CE Finch and WE Horton Jr (1995) Evidence for apoptotic cell death in Alzheimer's disease. Exp. Neurol. 133, 225-230. Sortwell CE, BF Daley, MR Pitzer, SO McGuire, JR Sladek Jr and TJ Collier (2000) Oligodendrocyte-type 2 astrocyte-derived trophic factors increase survival of developing dopamine neurons

14 280

F.C. LAU et al.

through the inhibition of apoptotic cell death. J. Comp. Neurol. 426, 143-153. Stoica G, O Illanes, SI Tasca and PKY Wong (1993) Temporal central and peripheral nervous system changes induced by a paralytogenic mutant of moloney murine leukemia virus TB. Lab. Invest. 69, 724-735. Su JH, G Deng and CW Cotman (1997) Bax protein expression is increased in Alzheimer's brain: correlations with DNA damage, bcl-2 expression, and brain pathology. J. Neuropathol. Exp. Neurol. 56, 86-93. Su JH, AJ Anderson, DH Cribbs, C Tu, L Tong, P Kesslack and CW Cotman (2003) Fas and Fas ligand are associated with neuritic degeneration in the AD brain and participate in beta-amyloidinduced neuronal death. Neurobiol. Dis. 12, 182-193. Sugimoto A, PA Friesen and JH Rothman (1994) Baculovirous p35 prevents developmental programmed cell death and rescues a ced-9 mutant in the nematode Caenorhabditis elegans. EMBO J. 13, 2023-2028. Tabuchi A, E Funaji, J Nakatsubo, M Fukuchi, T Tsuchiya and M Tsuda (2003) Inactivation of aconitase during the apoptosis of mouse cerebellar granule neurons induced by a deprivation of membrane depolarization. J. Neurosci. Res. 71, 504-515. Toescu EC (1999) Activity of voltage-opened calcium channels in rat cerebellar granule neurons and neuronal survival. Neuroscience 94, 561-570. Vaudry D, A Falluel-More, S Leuillet, H Vaudry and BJ Gonzalez (2003) Regulators of cerebellar granule cell development act through specific signaling pathways. Science 300, 1532-1534. Verhage M, AS Maia, JJ Plomb, AB Brussaard, JH Heeroma, H Vermeer, F Toonen Ruud, RE Hammer, TK van den Berg, M

Missler, HJ Geuze and TC Sudhof (2000) Synaptic assembly of the brain I the absence of neurotransmitter secretion. Science 287, 864-869. Westenbroek RE, T Sakurai, EM Elliott, JW Hell, TVB Starr, TP Snutch and WA Catterall (1995) Immunochemical identification and subcellular distribution of the α1A subunits of brain calcium channels. J. Neurosci. 15, 6403-6418. Wullner U, PA Loschmann, M Weller and T Klockgether (1995) Apoptotic cell death in the cerebellum of mutant weaver and lurcher mice. Neurosci. Lett. 200, 109-112. Yoon WJ, SJ Won, RB Ryu and BJ Gwag (2003) Blockade of ionotropic glutamate receptors produces neuronal apoptosis through the Bax-cytochrome C-caspase pathway: the causative role of Ca2+ deficiency. J. Neurochem. 85, 525-533. Yuan J and BA Yanker (2000) Apoptosis in the nervous system. Nature 407, 802-809. Yuan J, M Lipinski and A Degterev (2003) Diversity in the mechanisms of neuronal cell death. Neuron 40, 401-413. Zhang W, B Ghetti and WH Lee (1997) Decreased IGF-I gene expression during the apoptosis of Purkinje cells in pcd mice. Dev. Brain Res. 98, 164-176. Zhou YD, TJ Turner and K Dunlap (2003) Enhanced G proteindependent modulation of excitatory synaptic transmission in the cerebellum of the Ca2+ channel-mutant mouse, tottering. J. Physiol. 547, 497-507. Zhuchenko O, J Bailey, P Bonnen, T Ashizawa, DW Stockton, C Amos, WB Dobyns, SH Subramony, HY Zoghbi and CC Lee (1997) Autosomal dominant cerebellar ataxia (SCA6) associated with small polyglutamine expansions in the α1A-voltage-dependent calcium channel. Nat. Genet. 15, 62-69.