By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Address correspondence and reprint requests to Dr David Schubert, The Salk Institute for Biological Studies, 10010 N. Torrey Pines Road, La Jolla, CA 92037. E-mail: email@example.com
Two clonal nerve-like cell lines derived from HT22 and PC12 have been selected for resistance to glutamate toxicity and amyloid toxicity, respectively. In the following experiments it was asked if these cell lines show cross-resistance toward amyloid beta peptide (Aβ) and glutamate as well as toward a variety of additional neurotoxins. Conversely, it was determined if inhibitors of oxytosis, a well-defined oxidative stress pathway, also protect cells from the neurotoxins. It is shown that both glutamate and amyloid resistant cells are cross resistant to most of the other toxins or toxic conditions, while inhibitors of oxytosis protect from glutathione and cystine depletion and H2O2 toxicity, but not from the toxic effects of nitric oxide, rotenone, arsenite or cisplatin. It is concluded that while there is a great deal of cross-resistance to neurotoxins, the components of the cell death pathway which has been defined for oxytosis are not used by many of the neurotoxins.
Oxidative stress, defined in terms of an imbalance between reactive oxygen species (ROS) production and destruction, is thought to be involved in nerve cell death associated with ischemia, trauma and neurodegenerative diseases such as Parkinson's and Alzheimer's. In all of these pathologies a fraction of the nerve cells survive the initial insult, suggesting that protective mechanisms are available to some cells. We have previously isolated a series of cell lines which are resistant to oxidative stress caused by amyloid beta peptide (Aβ) or oxidative glutamate toxicity (Sagara et al. 1996, 1998). Here we ask if there is any overlap in the resistance of the Aβ and glutamate resistant cell lines to the original toxic agents as well as to a variety of other neurotoxins. In addition, the commonalities of the cell death pathways used in response to several additional neurotoxins are examined.
There are several ways in which the oxidative burden of cells can be regulated. One of these is through extracellular glutamate. Although glutamate is generally thought of as both a neurotransmitter and an excitotoxin, extracellular glutamate can also kill neurons through a non-receptor mediated pathway which involves the glutamate-cystine antiporter, system Xc− (Bannai and Kitamura 1980; Murphy et al. 1989; Sato et al. 1999). Under normal circumstances the concentration of extracellular cystine is high relative to intracellular cystine, and cystine is imported via the Xc– antiporter in exchange for intracellular glutamate. Cystine is rapidly converted to cysteine and utilized for protein synthesis and to make the antioxidant glutathione (GSH). However, when there is a high concentration of extracellular glutamate, the exchange of glutamate for cystine is inhibited and the cell becomes depleted of cysteine and GSH, resulting in severe oxidative stress. The cell eventually dies via a series of well defined events which have the characteristics of both apoptosis and necrosis, a process called oxytosis (for a review see Tan et al. 2002). Oxytosis has been most extensively studied in HT22 cells. HT22 cells are immortalized mouse hippocampal neurons that lack ionotropic glutamate receptors but die within 24 h after exposure to 1–2 mm glutamate. Several HT22 subclones have been selected for growth in high exogenous glutamate and are over 10-fold more resistant to glutamate than the parental cells (Sagara et al. 1998). These cell lines have increased expression of catalase, but not glutathione peroxidase (GPx) or superoxide dismutase. In addition, the activities of three enzymes involved in GSH metabolism, γ-glutamylcysteine synthetase, GSH reductase, and GSH-S-transferase, are elevated. Since transfection of some of these activities into the wild-type cell line does not confer complete resistance, it is likely that additional molecules play a role in the resistance pathway.
As with oxidative glutamate toxicity, reactive oxygen species (ROS) are involved in Aβ-induced cell death (Behl et al. 1994). Clones of the rat nerve-like cell line PC12 have been selected that are resistant to exogenously applied Aβ (Sagara et al. 1996). In these cells there is increased expression of both catalase and GPx, and the transfection of these enzymes into the wild-type line partially increases Aβ resistance. As with resistance to glutamate toxicity, a number of additional undefined factors are responsible for their overall resistance to amyloid.
As oxidative stress in the CNS caused by glutamate and Aβ may occur concomitantly or sequentially during both the aging process and AD pathology, it is important to determine if cells which survive one insult may be more resistant to the other. To this end the following experiments compare the cross-resistance of Aβ and glutamate resistant cell lines to their respective selective agents as well as to a number of additional toxins. We also examine the overlap in the cell death programs initiated by several additional neurotoxins. It is shown that while there is indeed a shared resistance to a wide spectrum of pro-oxidant conditions, the cells die by distinct programmed cell death pathways initiated by the individual neurotoxins.
Materials and methods
Cell culture and toxicity studies
The HT22 hippocampal nerve cell line is a subclone of HT4 (Morimoto and Koshland 1990), which was selected for its sensitivity to glutamate toxicity. The cells do not possess active ionotropic glutamate receptors and are not subject to excitotoxicity (Davis and Maher 1994). The glutamate resistant cell line (HT22r2) was selected for growth in 10 mm glutamate and is maintained in 2 mm glutamate (Sagara et al. 1998). HT22 cells are propagated in Dulbecco's modified Eagle's medium (DMEM) (Vogt and Dulbecco 1963) supplemented with 10% fetal bovine serum. Cell survival was determined by the MTT (3-(4,5-dimethyldiazol-2-yl)-2,5-diphenyl tetrazolium bromide) assay as described (Schubert et al. 1992). In HT22 cells the MTT assay correlates with cell death as determined by trypan blue exclusion and a colony forming assay (Davis and Maher 1994). Briefly, HT22 cells are dissociated with pancreatin (Life Technologies, Gaithersburg, MD) and seeded onto 96-well microtiter plates in 10% fetal bovine serum at a density of 2.5 × 103 cells per well in 100 µL medium. The next day cells are treated with various reagents according to the experimental design. Twenty hours after the addition of glutamate, 10 µL of the MTT solution (2.5 mg/mL) is added to each well and the cells are incubated for 4 h at 37°C. Solubilization solution (100 µL: 50% dimethylfomamide, 20% SDS, pH 4.8) is then added to the wells and the next day the absorption values at 570 nm are measured. The results are expressed relative to the controls specified in each experiment, and were subjected to statistical analysis (Student's t-test).
Rat pheochromocytoma (PC12) cells are a subclone of a high-passage cell line originally obtained from L. Greene (Greene and Tischler 1976). PC12 cells were grown on tissue culture dishes (Falcon, Indianapolis, IN) in DMEM supplemented with 10% fetal bovine serum and 5% horse serum. Aβ-resistant cells were selected for growth in high concentrations of Aβ and are maintained in the presence of 5 µm Aβ(25−35). The data presented here are from one Aβ-resistant clone, AβrCl 8 (Sagara et al. 1996). For amyloid toxicity the MTT assay does not necessarily reflect cell viability although it does measure an early necessary event in the cytotoxicity pathway (Liu and Schubert 1997; Liu et al. 1997). Therefore growth curves were also done to measure toxicity.
The caspase inhibitor, AcYVADfmk, was obtained from Bachem (Torrance, CA, USA) (R,S)-3,5-dihydroxyphenylglycine (DHPG) from Tocris-Cookston (England), and all other reagents were from Sigma (St. Louis, MO, USA).
As both glutamate and Aβ are toxic to nerve cells, it was asked if cells selected for resistance to one of these toxins are more or less resistant to the other. Two cell lines were used. The hippocampal nerve cell line HT22, which is sensitive to glutamate, and the PC12 sympathetic nerve-like line which has been extensively studied with respect to amyloid toxicity. Clonal cell lines that are very resistant to each toxin were selected by growth in the presence of high concentrations of Aβ or glutamate (Sagara et al. 1996; Sagara et al. 1998). For example, the EC50 for killing HT22 cells is 1.5 mm glutamate, while the resistant clone requires 25 mm glutamate (Table 1). Similarly, Aβ slows down the rate of growth of wild type PC12 cells but has no effect on the PC12 resistant clone 8 (Fig. 1a). Likewise, using the MTT assay, which measures the biological activity of the Aβ peptide (Liu and Schubert 1997), the EC50 is increased from 10 µm Aβ to 55 µm Aβ (Table 1). When the cells selected for glutamate resistance are screened for Aβ toxicity and vice versa, there is a surprising degree of cross resistance relative to their parental cell lines (Fig. 1b, Table 1). The Aβ resistant PC12 line is 8 times more resistant to glutamate than its parental cell line, while the glutamate resistant HT22 clone is about twice as resistant to Aβ as the wild type clone. These data show that selection for resistance to one neurotoxin can lead to resistance to other toxins.
Table 1. Glutamate and Aβ Resistant Lines (EC50)
Exponentially dividing cultures of HT22 and PC12 cells as well as their clones resistant to glutamate and Aβ, respectively, were plated in 96 well plates at 2.5 ×103 per well and 18 h later various concentrations of the indicated toxins were added and the cells incubated for an additional 20 h before the MTT viability assay. In all cases, cells were also visually scored for toxicity and both sets of data were in agreement. The toxins were added in triplicate at 3-fold serial dilutions and the EC50 for toxicity is given as the mean of 3 determinations plus or minus the standard error of the mean. ΔCys means cystine depletion. The normal concentration of cystine in the culture medium is 260 µm. *The cells are resistant to BSO up to a concentration of 2 mm.
To determine if the cross-resistance can be generalized to other forms of neurotoxicity, the toxicity of a variety of additional neurotoxins or toxic conditions were examined on all four cell lines. Three conditions, cystine depletion, GSH depletion, and H2O2 toxicity were chosen because of their relationship to the oxytosis pathway (Tan et al. 2002), while five additional toxins (arsenite, cisplatin, MPP +, rotenone, and nitric oxide) were selected because their toxicities are thought to involve reactive oxygen species (ROS) (see later). Although the depletion of intracellular glutathione by buthionine-[S,R]-sulfoximine (BSO) and cystine depletion (ΔCys) are related to cell death caused by exogenous glutamate due to the fact that they both deplete GSH, there are mechanistic differences (Tan et al. 2002). Both glutamate and Aβ resistant cell lines are, however, much more resistant than their parental cells to both cystine starvation and BSO toxicity (Table 1). In the case of HT22, 50% of the wild-type cells are killed by the reduction of cystine in the culture medium from 260 µm (normal) to 129 µm, while 50% of the resistant cells will survive in 20 µm cystine (Fig. 2). A similar increase in resistance to cystine starvation occurs with PC12r8 (Fig. 2). It should be noted that there are also very large baseline differences between the sensitivities of the wild type HT22 and PC12 cell lines to the different toxic conditions. There could be many reasons for these differences but they are not known at this time.
Hydrogen peroxide is frequently used as a standard for the induction of oxidative stress in cultured cells. Since both oxytosis and Aβ toxicity have peroxide intermediates in their cell death pathways (Behl et al. 1994; Tan et al. 2001), it would be predicted that both resistant cell lines would also be more resistant to H2O2. The data in Table 1 show that this is indeed the case. As with glutamate and BSO, it takes a much higher concentration of H2O2 to kill PC12 cells than HT22 cells, suggesting that the antioxidant defense system in PC12 is much more effective than that of HT22.
Finally, five other toxins with proposed ROS intermediates in their toxicity pathways were examined: arsenite, cisplatin, nitric oxide (NO), MPP+ and rotenone. Cisplatin and arsenite damage cells at least in part via oxidative damage to DNA (Hannemann and Baumann 1988). Both resistant cell lines are much more resistant to arsenite, while there is no significant difference with respect to cisplatin. Rotenone and MPP+ are mitochondrial poisons and can cause Parkinson's disease-like symptoms when administered to animals (for review, see Andersen 2001). Both resistant cell lines are more resistant to rotenone, but only PC12r8 is slightly more resistant to MPP+. Finally, NO has been implicated in many forms of cell death associated with CNS trauma and neurological disease (for review, see Bredt 1999). Cells were exposed to the NO donor diethylenetriamine/nitric oxide adduct, using diethylenetriamine as a control. There is no difference however, between cells which were selected for resistance to glutamate and Aβ and their parental cell lines regarding NO toxicity (Table 1). It follows that the regulation of oxidative damage caused by NO is very distinct from those of the other pro-oxidants.
The above data show that resistance to glutamate toxicity confers resistance to Aβ toxicity and vice versa, as well as resistance to several other neurotoxins. A related question is whether or not components of a defined programmed cell death pathway are shared in the toxicity of other toxins toward a single nerve cell type. To answer this question we killed HT22 hippocampal nerve cells with several neurotoxins and conditions which induce oxidative stress, and asked if these cell death pathways are blocked by reagents which inhibit a form of programmed cell death called oxytosis (Tan et al. 2002). Oxytosis is quite distinct from apoptosis in that the Bcl-2/Bax system is not involved (Dargusch et al. 2000), there is no DNA laddering and there are a number of morphological differences from apoptosis (Tan et al. 1998). A group of reagents have been identified which inhibit this pathway. These include the caspase inhibitor tyrosine-valine-alanine-aspartate-fluoromethyketone (YVADfmk), nordihydroguaiaretic acid (NDGA), a lipoxygenase inhibitor (R,S)-3,5-dihydroxyphenylglycine (DHPG), a group I glutamate metabotropic receptor agonist, PD168,077, a dopamine D4 receptor agonist, the dietary antioxidant curcumin, diphenyleneiodium (DPI), a monoamine oxidase and mitochondrial ROS inhibitor, and finally, LY83583, a competitive inhibitor of soluble guanylate cyclase (sGC) (Tan et al. 2002). Table 2 shows the results when HT22 cells were challenged with glutamate, BSO, H2O2, the deprivation of exogenous cystine and glucose, NO, rotenone, sodium arsenite or cisplatin in the presence of these oxytosis inhibitors. Since cells can be overwhelmed by excess toxin, frequently masking some aspects of the cell death pathway, concentrations of toxins were chosen which only kill between 60 and 80% of the cells. Cell death caused by glutamate toxicity and cystine deprivation were both blocked by all of the inhibitors of the oxytosis pathway. This is to be expected since glutamate is toxic by virtue of its ability to inhibit cystine uptake (Bannai and Kitamura 1980; Murphy et al. 1989).
Table 2. Protection of HT22 Cells from Various Toxins (EC50)
Exponentially dividing HT22 cells were plated in microtiter plates at 2.5 ×103 cells per well and one day later challenged with the various toxins or toxic conditions at concentrations which killed between 60 and 80% of the cells as defined by the MTT assay 20 h later (see Table 1 for approximate concentrations in each toxic condition). Thirty minutes before adding the toxins the various inhibitors were added in triplicate at 6 different concentrations around their estimated EC50 as determined previously from glutamate toxicity. If no inhibition was observed, 6 higher concentrations were tested, or until the inhibitor itself became toxic. The data are presented as the concentration of the inhibitor that protected 50% of the cells from cell death. ‘No’ indicates that there was no protection at concentrations of up to 100 fold greater than those required to protect 50% of the cells from glutamate toxicity. *Δcys indicates the medium contains 65 µm cystine, 25% of the normal amount. ‡DHPG is a metabotropic glutamate receptor agonist whose biologic effect (protection) cannot be seen in the presence of high extracellular glutamate (see Sagara and Schubert 1998; for details).
The depletion of intracellular cystine uptake leads to the loss of intracellular GSH, the cell's major antioxidant. Since the loss of GSH is thought to initiate oxytosis, it would be expected that all reagents that inhibit glutamate toxicity would also block BSO-induced cell death. This is, however, not the case, for DPI and the metabotropic glutamate receptor agonist DHPG do not prevent BSO induced cell death (Table 2). DPI and DHPG are also ineffective against H2O2 toxicity. Since DPI blocks ROS production from mitochondria, which appears to be a necessary intermediate in oxytosis, it is not clear why BSO toxicity is unaffected. Since oxytosis involves the production of peroxides up to 300-fold above basal level, it would be expected that blocking mandatory cell death enzymes downstream of peroxide production would inhibit cell death caused by the addition of exogenous H2O2. This appears to be in part the case, for both PD168, 077 and LY83583, two reagents that inhibit Ca2+ influx, also block H2O2 toxicity (Li et al. 1997; Ishige et al. 2001). The activation of metabotropic glutamate receptors by DHPG does not block H2O2 toxicity, but the pathway used by DHPG to inhibit cell death is not well defined (Sagara and Schubert 1998).
In contrast to glutamate, BSO, H2O2 and cystine deprivation, the toxic insults of arsenite, cisplatin, rotenone, and NO were not inhibited by most of the reagents that block oxytosis. The major exception was the sGC inhibitor LY83583, which showed the widest spectrum of inhibition, blocking cell death induced by all of the toxic agents tested except arsenite. In addition, both PD168, 077 and the antioxidant, curcumin, blocked rotenone toxicity. DHPG was the only other inhibitor of NO toxicity besides the guanylate cyclase inhibitor. Therefore the specific cell death pathways which are activated by NO, rotenone, arsenite and cisplatin are distinct from the oxytosis pathway despite the fact that cells selected for resistance to oxytosis are generally more resistant to these toxins (Table 1).
The above data show that nerve cells which are selected for resistance to Aβ or high extracellular glutamic acid are also more resistant to a variety of neurotoxins, including arsenite, rotenone, MPP+ and the strong oxidizing agent, H2O2. While glutamate, BSO, cystine depletion, and H2O2 toxicities all share most of the cell death components required for oxytosis, cell death caused by arsenite, cisplatin, MPP+, NO and rotenone all seem to occur by different mechanisms because most of the inhibitors that block oxytosis fail to inhibit the cell death caused by these toxins. These observations are somewhat surprising since the latter group of toxins are all thought to cause cell death via pathways which are ROS dependent, and the Aβ and glutamate resistant cells all have elevated antioxidant enzyme levels and are much more resistant to peroxides (Sagara et al. 1996; Sagara et al. 1998; Table 1). Therefore there must be multiple cell death pathways for these neurotoxins, and simple ROS elevation is apparently not always sufficient to complete the death program. The following paragraphs briefly discuss the cross-resistance and the putative role of ROS in cell death caused by each toxin.
There is rather extensive cross-resistance between cells selected in the presence of Aβ or glutamate and a variety of additional neurotoxins (Table 1). The shared resistance to cystine deprivation and BSO is not surprising since the loss of cystine and GSH are central to the oxytosis pathway (Tan et al. 2002). The pleotropic effects of cells selected for resistance to H2O2 has also been noted. For example, fibroblasts selected for resistance to H2O2 become resistant to cadmium(II) and mercury(II) (Sugiyama et al. 1993) as well as C2 ceramide (Kim et al. 2001) and exposure to pure oxygen (Spitz et al. 1995). Conversely, cells selected for resistance to NiCl2 or cadmium also become resistant to H2O2 and menadione (Mello-Filho et al. 1988; Salnikow et al. 1994). Therefore cross-tolerance to a variety of toxic agents is not unusual and the common denominator appears to be resistance to oxidative stress that can be assayed by the cellular response to H2O2. It does not necessarily follow, however, that changes in antioxidant metabolism, although frequently observed, are the sole mechanism that leads to resistance (see above references and those that follow).
Table 1 shows that both the Aβ and glutamate resistant cells are not significantly more resistant to cisplatin than their parental cell lines. Cisplatin is generally considered a DNA damaging agent, causing DNA-DNA cross-linking as well as single strand breaks. There is evidence, however, that oxidative stress and resultant lipid peroxidation also contribute to the cytotoxicity caused by cisplatin (Hannemann and Baumann 1988; Spitz et al. 1993). H2O2-resistant HA1 fibroblasts are 1.5–3 fold more resistant to cisplatin than the parental lines (Spitz et al. 1993). It was shown, however, that the elevated catalase activity in the H2O2 and cisplatin resistant lines was not responsible for conferring resistance, and that elevated GSH was only in part responsible for cisplatin resistance (Spitz et al. 1993). These data are consistent with the fact that both Aβ-resistant PC12 cells (Sagara et al. 1996) and glutamate resistant HT22 cells (Sagara et al. 1998) have highly elevated antioxidant enzymes but are not resistant to cisplatin. It is therefore likely that although cisplatin can cause oxidative damage, its major mechanism of toxicity is independent of ROS metabolism.
Arsenic occurs naturally in soil, water, and air, and also as a by-product in the production of other metals. It has been used for centuries as both a therapeutic and as an intentional poison; it is also a potent carcinogen (Abernathy et al. 1999). Rodent cells exposed to arsenite (As+3), become cross resistant to arsenate (As+5) (Gurr et al. 1999), as well as other metals such as cadmium and nickel (Romach et al. 2000), at least in part through the increase of multidrug resistant transporters which reduce intracellular metalloid concentration (Romach et al. 2000; Liu et al. 2001). Table 1 shows that cells selected for resistance to two agents which induce oxidative stress in cells, Aβ and glutamate, are also much more resistant to arsenite. These data suggest that at least some of the toxic effects of arsenite may be due to ROS metabolism. The major metabolite of inorganic arsenic, dimethylarsinic acid, causes the oxidation of deoxyguanosine to 8-oxo-2′-deoxyguanosine, probably through a dimethylarsenic peroxy radical (Yamanaka et al. 2001). Arsenite also directly causes the production of hydroxyl radicals and other ROS (Wang et al. 2001). Since both Aβ and glutamate resistant cells have elevated levels of enzymes which are able to reduce ROS toxicity, our data support the previous work on the involvement of ROS in arsenic induced mutagenicity and toxicity (Gurr et al. 1999).
Exposure of neurons to the Parkinson's inducing drugs 1-methyl-4-phenyl 1,2,3,6-tetrahydropyridine (MPTP), its metabolite MPP+, or rotenone causes selective degeneration of dopaminergic neurons of the substantia nigra in mice and the death of HT22 and PC12 nerve-like cells (Table 1 and Andersen 2001; for recent review). Both rotenone and MPP+ interfere with complex I mitochondrial electron transport, decrease ATP production, and increase mitochondrial generated ROS. Oxytosis also results in a 100–300-fold increase in mitochondria ROS production from complex I, which is required for the activation of down stream enzymes such as soluble guanylate cyclase (Tan et al. 2002). Surprisingly, the Aβ resistant cell lines were more resistant to MPP+ and rotenone, but the glutamate resistant cells were only more resistant to rotenone.
Glutamate and cystine depletion kill cells by a well studied programmed cell death pathway called oxytosis (Tan et al. 2002). The death of HT22 cells and primary neurons caused by high extracellular glutamate is clearly inhibited by a number of reagents that interfere with defined steps in the pathway. Each step is necessary for cell death to occur. To determine if there are any shared steps in the pathways used by the various toxins, it was asked if the inhibitors of the oxytosis pathway also inhibit cell death caused by the other toxins. With the exception of cystine deprivation, BSO, and H2O2, all potent inducers of oxidative stress, the answer is generally no (Table 2). For example, NO, rotenone, cisplatin and arsenite toxicity are, with a few exceptions, not blocked by any inhibitor of oxytosis. The major exception is the ability of the sGC inhibitor, LY83583, to inhibit all toxicities except arsenite. sGC generates cGMP which, in turn, opens cGMP gated calcium channels, allowing the influx of extracellular calcium (Li et al. 1997; Ishige et al. 2001). If cyclase inhibitors such as LY83583 prevent this calcium flux, then the calcium does not enter and the cells do not die. The antioxidant curcumin and the dopamine D4 receptor agonist PD168, 077 also inhibit rotenone toxicity. It can be concluded from the combined data that although there is extensive overlap with respect to resistance to multiple neurotoxins, the cell death pathways employed by arsenic and cisplatin must be distinct from that causing oxytosis. Whether each of these neurotoxins employs a unique cell death pathway remains to be determined.
This work was supported by the Department of Defense grant #DAMD17-99-1-9562, and the National Institutes of Health. We thank Drs P. Maher, Y. Liu and R. Cumming for reading the manuscript.