Materials and methods membrane integrity were visualized by excitation at 488

100 A necrosis in the rat hippocampus [25]. With concurrent Val-Ala-Asp-CH F Z-VAD was purchased from Enzyme 2 21 depolarization, the toxicity of extracellular Zn is en- Systems Products. PC12 cells were transfected using hanced such that short exposure to low micromolar con- FuGene 6 Boehringer Mannheim, and stable transfectants 21 centrations of Zn becomes neurotoxic [54]. This de- were selected in medium containing 500 mg ml G418. polarization-induced enhancement likely reflects preferen- 21 tial entry of zinc through L-type voltage-gated Ca 2.2. Constructs channels, leading to toxic elevations of intracellular free 21 21 Zn [Zn ] in the range of 300–500 nM [4,40]. Wild type rZnT-1 cDNA kindly provided by Dr. R.D. i 21 Palmiter and dominant negative rZnT-1 cDNA an 83 bp Likely opposing such toxic elevations of [Zn ] are i EagI fragment deletion; [34] were subcloned into a CMV several mechanisms responsible for maintaining intracellu- 21 promoter-driven mammalian expression vector. After lar Zn homeostasis, including binding to metallothio- mutating the terminal stop codon in both rZnT-1 forms, a neins [1,10,50,53], and export across the plasma membrane fragment encoding a hexameric myc epitope tag gener- mediated by the ubiquitously expressed zinc transporter, ously provided by Dr. R. Kopan was inserted in-frame ZnT-1 [34]. Related zinc transporters appear to bear 21 distal to the altered stop codon. responsibility for transporting Zn into endosomes ZnT- 2 [32] or synaptic vesicles ZnT-3 [33]. Baby hamster 2.3. Toxicity experiments kidney BHK cells [34] and N2A cells [49] overexpres- sing ZnT-1 exhibit resistance to death induced by pro- 21 Immediately before toxic exposures, cells were washed longed Zn exposure. Following transient global ischemia twice using a HEPES-buffered salt solution HSS with the in gerbils, ZnT-1 transcription is upregulated in CA1 1 1 following composition: 130 mM Na , 5.4 mM K , 0.8 pyramidal neurons, consistent with the possibility that 21 21 2 mM Mg , 1.8 mM Ca , 131 mM Cl , 20 mM HEPES ZnT-1 induction may be a cellular strategy to counter 21 pH 7.4 at 258C, 15 mM glucose. Exposures of 15 min to ischemia-induced toxic Zn influx [49]. 21 toxic solutions were conducted at room temperature in We set out to develop a model of Zn influx-induced 1 1 HSS with equimolar substitution of K for Na in cell death in a neuronal cell line. We chose PC12 cells 1 solutions with elevated [K ]. To terminate toxic expo- because they have been widely used for investigating sures, solutions were washed twice with media stock MS, multiple aspects of neurobiology, including neuronal dif- which consists of Eagle’s minimal essential medium plus ferentiation, intracellular signaling pathways, and cell 21 21 mM glucose, and then replaced with MS containing 1 survival [14,56], and specifically express L-type Ca 21 serum 2:1, horse serum:fetal bovine serum before being channels [18,47], the primary route of toxic Zn entry 21 returned to a 378C incubator. A23187 exposures were into neurons. There is one previous report of Zn -induced conducted at 378C in MS after sham washes. Z-VAD was death of PC12 cells, utilizing prolonged 24 h exposure to 21 added to the final replacement solution MS and 1 Zn in the absence of serum, thought to be due to direct 21 serum following exposures and washes. inactivation of NGF by Zn [39]. To eliminate this death 21 mechanism and focus on toxic Zn entry, we utilized 21 2.4. Cell death assays brief 15 min exposure to Zn , followed by return to serum-containing medium, as the presence of serum abro- 21 Cell death was assessed morphologically by phase-con- gated Zn -mediated neurotrophin deprivation-induced trast microscopy and propidium iodide fluorescence. For death [39]. propidium iodide staining, cells were incubated for 30 min at 378C with propidium iodide solution 5 mg ml Molec- ular Probes, and dead cells cells with compromised

2. Materials and methods membrane integrity were visualized by excitation at 488

nm. Cell death was quantified by measuring lactate dehy- 2.1. Cell culture, drugs, and stable transfection drogenase LDH released by injured cells into the cell medium [21]. LDH values were normalized by subtracting PC12 cells were grown in Dulbecco’s modified Eagle’s the background LDH released by sham-washed cells from medium DMEM supplemented with 10 heat-inacti- treated cells and scaling to the signal associated with vated horse serum, 5 fetal bovine serum, and 2 mM complete cell death 5100, induced by 24 h exposure to L -glutamine GIBCO BRL in a humidified incubator at 30 mM A23187. 378C and 5 CO . Twenty-four hours prior to experi- 2 ments, cells were plated onto poly- L -lysine-coated 24-well 2.5. Electron microscopy 5 16 mm diameter plates at a density of 1.25310 cells well. ZnCl , GdCl , and A23187 were purchased from Cells were harvested 12 h following toxic exposures and 2 3 Sigma, and S2-Bay K 8644 and nimodipine from fixed with 1 paraformaldehyde 2 glutaraldehyde in 0.1 Research Biochemicals International. Benzyloxycarbonyl- M phosphate buffer pH 7.4 overnight at 48C. Cultures A .H. Kim et al. Brain Research 886 2000 99 –107 101 were then fixed and embedded. After cutting and staining 100 mM GdCl , and the cultures were placed on ice. After 3 sections with lead citrate and uranyl acetate, cells were 10 min, the quench solution was aspirated, and the cells examined and photographed with a JEOL 100 CS electron were lysed by addition of 100 ml hot 0.2 SDS. The lysate microscope. was then dotted onto filter paper Whatman 3MM, What- man, Inc., dried by heating lamp, and counted in a gamma 2.6. Immunofluorescence counter Beckman 4000, Beckman Instruments, Inc.. Other cultures from the same plating were used to de- Plated cells were fixed with cold methanol and immuno- termine protein concentrations for each sample. stained with a monoclonal antibody against the myc epitope Calbiochem. After incubating cells with a fluorescence-conjugated anti-mouse Alexa 488 secondary

3. Results