- Email: [email protected]
Acute and chronic alterations in calcium homeostasis in 3-nitropropionic acid-treated human NT2-N neurons
Neuroscience Vol. 113, No. 3, pp. 699^708, 2002 C 2002 IBRO. Published by Elsevier Science Ltd All rights reserved. Printed in Great Britain 0306-4522 / 02 $22.00+0.00
PII: S 0 3 0 6 - 4 5 2 2 ( 0 2 ) 0 0 1 7 7 - X www.neuroscience-ibro.com
ACUTE AND CHRONIC ALTERATIONS IN CALCIUM HOMEOSTASIS IN 3-NITROPROPIONIC ACID-TREATED HUMAN NT2-N NEURONS W.-T. LEE,a;b T. ITOHb and D. PLEASUREb a b
Department of Pediatrics, National Taiwan University Hospital, Taipei, Taiwan
Division of Neurology Research, Room 516H, Abramson Research Building, The Children’s Hospital of Philadelphia, 34th Street and Civic Center Boulevard, Philadelphia, PA 19104, USA
Abstract33-Nitropropionic acid (3-NP), an irreversible inhibitor of succinate dehydrogenase, induced ATP depletion and both necrosis and apoptosis in human NT2-N neurons. Necrosis occurred predominantly within the ¢rst two days, and increased in a dose-dependent fashion with the concentration of 3-NP, whereas apoptosis was observed after 24 h or later at a similar rate in 0.1 mM and 5 mM 3-NP. We focused our e¡orts on intracellular calcium homeostasis during the ¢rst 48 h in 1 mM 3-NP, a period during which 10% of the neurons died by necrosis and 3% by apoptosis. All NT2-N neurons showed a stereotyped [Ca2þ ]i rise, from 48 ; 2 to 140 ; 12 nM (mean ; S.E.M.), during the ¢rst 2 h in 3-NP. Despite severe ATP depletion, however, [Ca2þ ]i remained above 100 nM in only 17% and 25% of the NT2-N neurons after 24 and 48 h in 3-NP, respectively, indicating that most neurons were able to recover from acute [Ca2þ ]i rise, and suggesting that chronic [Ca2þ ]i dysregulation is a better indicator of subsequent necrosis. Blockade of N-methyl-D-aspartate^glutamate receptor by MK-801 substantially ameliorated 3-NP-induced ATP depletion, subsequent chronic [Ca2þ ]i elevation, and survival. Moreover, xestospongin C, an inhibitor of endoplasmic reticulum Ca2þ release, enhanced the capacity of NT2-N neurons to maintain [Ca2þ ]i homeostasis and resist necrosis while subjected to sustained energy deprivation. As far as we know, this report is the ¢rst to employ human neurons to study the pathophysiology of 3-NP neurotoxicity. C 2002 IBRO. Published by Elsevier Science Ltd. All rights reserved. Key words: NT2-N neurons, 3-nitropropionic acid, necrosis, glutamate receptors, calcium, xestospongin C.
vival of these cultured neurons is enhanced by inhibiting NMDA^GluR (Fink et al., 1996; Pang and Geddes, 1997). These observations indicate that activation of NMDA^GluR plays an important role in 3-NP neurotoxicity. This activation is due, at least in part, to partial neuronal plasma membrane depolarization, which releases the voltage-dependent Mg2þ block of NMDA^ GluR (Greene et al., 1998). We previously reported that NT2-N neurons, prepared by retinoic acid-induced di¡erentiation of the human teratocarcinoma line, Ntera-2 (Pleasure et al., 1992), are susceptible to NMDA^GluR- and non-NMDA^ GluR-mediated excitotoxicity (Younkin et al., 1993; Hardy et al., 1994; Itoh et al., 1998), and that oxygen/ glucose deprivation causes NT2-N neuronal necrosis by an NMDA^GluR-mediated mechanism (Rootwelt et al., 1998; Almaas et al., 2000). In the present study, we examined the contributions of NMDA^GluR, nonNMDA^GluR, voltage-gated Ca2þ channels (VGCC), endoplasmic reticulum (ER), and mitochondria to acute and chronic alterations in [Ca2þ ]i homeostasis in these human neurons during a 3-NP treatment regimen chosen to elicit predominantly neuronal necrosis.
Systemic or intrahemispheric administration of 3-nitropropionic acid (3-NP), an irreversible inhibitor of succinate dehydrogenase, elicits striatal neuronal degeneration in rodents and non-human primates, and recapitulates many of the clinical and pathological features of Huntington’s disease (Beal et al., 1993; Wullner et al., 1994; Pal¢ et al., 1996). This neuronal degeneration is diminished by administration of MK-801, an N-methyl-Daspartate^glutamate receptor (NMDA^GluR) inhibitor (Kim et al., 2000; Lee et al., 2000). In primary cultures of rodent CNS, 3-NP elicits an early increase in neuronal [Ca2þ ]i (Greene et al., 1998; Fukuda et al., 1998; Keller et al., 1998; Olsen et al., 1999), and both apoptotic and necrotic neuronal death. As was observed in vivo, sur-
*Corresponding author. Tel. : +1-215-590-2090; fax: +1-215-5903709. E-mail address: [email protected] (D. Pleasure). Abbreviations : CNQX, 6-cyano-7-nitroquinoxaline-2,3-dione; ER, endoplasmic reticulum; FCCP, carbonyl cyanide p-(tri£uoromethoxy)phenylhydrazone ; HEPES, N-2-hydroxyethylpiperazineNP-2-ethane-sulfonic acid; IP3, inositol 1,4,5-trisphosphate ; LDH, lactate dehydrogenase ; MK-801, dizocilpine ; MPT, mitochondrial permeability transition ; MTT, 3-(4,5-dimethylthiazol2-yl)-2,5-diphenyltetrazolium bromide; 3-NP, 3-nitropropionic acid ; NADH, nicotinamide adenine dinucleotide; NMDA^ GluR, N-methyl-D-aspartate^glutamate receptor ; TUNEL, terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling; VGCC, voltage-gated Ca2þ channels.
EXPERIMENTAL PROCEDURES
Cell culture NT2-N neurons were prepared and maintained as previously 699
NSC 5622 23-7-02
700
W.-T. Lee et al.
described (Pleasure et al., 1992; Itoh et al., 1998; Rootwelt et al., 1998). The medium contained 2 mM glutamine (freshly prepared) and 10% (v/v) fetal bovine serum, without added glutamate. The neurons were plated in six-well, 12-well and 24-well plates or small chambers that had been coated with polylysine and Matrigel (Becton Dickinson, Bedford, MA, USA). Six-well plates were plated at 2U106 cells/well, 12-well plates at 1U106 cells/well, and 24-well plates at 0.5U106 cells/well. The cultures were studied between four and ¢ve weeks after termination of retinoic acid di¡erentiation. All experiments were done two to four days after the last feeding. Cell death assays To expose NT2-N neurons to 3-NP, the culture medium was totally replaced with fresh medium containing 3-NP at various concentrations and no exogenous glutamate. At the indicated time-points, the cells were treated with Triton X-100 to permeabilize plasma membranes, and double-stained with propidium iodide and for DNA nicking by the terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) procedure (Gavrieli et al., 1992; Pang and Geddes, 1997). Nuclei of dead cells were condensed and/or fragmented, and distinctly di¡erent from the oval or round nuclei of live cells. We counted nuclei of dead cells that were positive by the TUNEL reaction as apoptotic, and cells with condensed or fragmented nuclei that were TUNEL-negative as necrotic. A total of 300 to 500 neurons were counted for each experiment. In each experimental set, we simultaneously examined NT2-N neuronal cultures that had not been exposed to 3-NP in order to determine control incidences of apoptosis and necrosis. Plasma membrane integrity was also assessed by measuring lactate dehydrogenase (LDH) release from the cells to the medium (Younkin et al., 1993). LDH measurements were done with the Cytotox 96 non-radioactive cytotoxicity assay kit (Promega, Madison, WI, USA), based on the conversion of pyruvic acid to lactic acid by LDH in the presence of NADH, according to the manufacturer’s protocol. At indicated time-points after treatment with 3-NP, 50 Wl of culture medium was removed for LDH assay. The NT2-N cells were then lysed with Triton X-100 to a ¢nal concentration of 0.1% (w/v). The plates were stored at 37‡C for 45 min, and then 50 Wl of cell lysate was taken for the LDH measurement. The percentage of the cell death was expressed as the ratio of LDH content in the medium before and before plus after cell lysis U100. Reducing capacity of the neurons was assessed by the conversion of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) to formazan crystals as described previously (Itoh et al., 1998). Formazan production was measured 2 h after adding MTT (0.5 mg/ml) to the culture medium by spectroscopy at 560 nm, and was expressed as the percentage of simultaneously run control values.
neurons in randomly chosen ¢elds at room temperature (23^ 25‡C) as previously described (Itoh et al., 1998). In brief, the NT2-N neurons cultured on coverslips were loaded with 5 WM fura-2/AM with 0.02% (w/v) pluronic F-127 in the standard recording solution (NaCl 140, KCl 5.4, CaCl2 1.8, MgCl2 0.8, Na2 HPO4 0.8, HEPES 10, and D-glucose 25 (in mM), adjusted to pH 7.4 with NaOH) for 45 min. The coverslip was attached to a perfusion chamber (RC-21B; Warner Instrument, USA) on the stage of an upright epi£uorescence microscope (Optiphot, Nikon, Japan), and emission £uorescence images at 510 nm were alternatively taken at 340 and 380 nm excitation wavelengths. [Ca2þ ]i values were calculated from the emission ratio (R) at 340 and 380 nm excitation wavelengths, using the formula described by Grynkiewicz et al. (1985): [Ca2þ ]i = bKd (R3Rmin )/ (Rmax 3R). Kd is the e¡ective dissociation constant of fura-2, and b is the ratio of £uorescence intensity at 380 nm for Ca2þ -free and Ca2þ -saturated dye. Rmin and Rmax are the emission ratios obtained under Ca2þ -free and Ca2þ -saturated conditions, respectively. These parameters were obtained by in vitro calibration. We previously demonstrated that [Ca2þ ]i measurements could be done in NT2-N neurons up to 3 h, if emission images were taken at a 2.5 min interval to minimize damage to the cells by UV exposure as well as photobleaching of loaded fura-2 (Itoh et al., 1998). Even when the frequency of UV exposure was reduced, however, we still observed signi¢cant death of NT2N neurons at 24 h after UV exposure. Therefore, in order to evaluate chronic changes in [Ca2þ ]i after 24 or 48 h with 3-NP, we used sister cultures treated identically to those for acute experiments. Culture medium glutamate concentration Assays of glutamate in the medium were performed as previously described (Rootwelt et al., 1998). After centrifugation to remove cell debris, 0.5 ml of medium was mixed with 1 ml 100 mM Tris bu¡er (pH 7.0), and passed over a column containing AG-1 resin (chloride form; 100^200 mesh; 0.5 ml bed volume) (Bio-Rad, Hercules, CA, USA) to remove excessive glutamine. Glutamine was eluted with 4 ml of deionized water. Glutamate was then eluted with 4 ml of 2 M HCl and analyzed by high pressure liquid chromatography, using precolumn derivatization with o-phthalaldehyde and £uorescent detection (Jones and Gilligan, 1983) with an internal standard. Statistics In most instances (as indicated in the text and ¢gure legends), the results were expressed as means ; S.E.M., with statistical signi¢cances calculated by paired t-test or ANOVA. Chi-square testing was done for analysis of statistical signi¢cance of the data in Table 2.
Neuronal ATP content ATP was quanti¢ed by a luciferin/luciferase-based assay. Twelve-well culture plates were rinsed with ice-cold phosphatebu¡ered saline (PBS), and the neurons were then lysed with 0.1 ml of ATP releasing bu¡er (Sigma). ATP concentrations were measured using the ATP Bioluminescence Assay kit CH II (Boehringer Mannheim) and a luminometer. Fifty microliters of sample was mixed with 50 Wl of luciferase reagent. After 1 s, the reaction was recorded for 10 s. A standard curve was constructed by assaying calibration samples of known ATP concentration. Samples were diluted so that the readings fell within the linear part of the standard curve. Protein content was determined using the BCA protein assay kit (Pierce, Rockford, IL, USA). The ATP levels at each time-point were compared with normal controls at the same time-points. ATP content was expressed as percentage of non-3-NP-treated control at each time-point. [Ca2+]i measurements [Ca2þ ]i measurements were performed on all isolated NT2-N
RESULTS
Acute necrosis is induced by 1 mM or higher 3-NP in NT2 neurons Treatment of NT2-N neurons with 3-NP induced two modes of neuronal death, apoptotic and non-apoptotic (necrotic). Apoptosis, demonstrated by TUNEL-positivity, increased linearly with incubation time and was independent of 3-NP concentration (Fig. 1B). In contrast, necrosis, estimated by subtracting the TUNEL-positive population from total death, increased signi¢cantly with concentrations of 3-NP of more than 1 mM. This increase was mainly observed within 48 h after 3-NP was added. Thereafter, the slopes of increases in total cell death were similar to those observed in TUNEL-
NSC 5622 23-7-02
3-NP excitotoxicity in human NT2-N neurons
701
Fig. 1. Time- and 3-NP-concentration-dependent apoptotic and total death of NT2-N neurons, assessed by counting nuclei of dead neurons and those positive to TUNEL staining (see Experimental procedures for detail). Open circles in each panel show background death in two independent control cultures. We always observed similar percentages of the background during the observation periods. Data are means ; S.D. of three independent experiments.
positive cells (Fig. 1A, B). These results indicate that, whereas delayed apoptosis predominates at 0.1 mM 3-NP, 1 mM or higher 3-NP induces NT2-N neuronal necrosis, mainly within 48 h. We focused on this necrotic 3-NP-induced death in this study. As shown in Fig. 1, control NT2-N cultures contained between 12% to 19% dead cells by morphological evaluation. We calculated the actual population acutely killed by 3-NP at each time-point by subtracting this control background level. Results of this calculation correlated well, during the ¢rst four days of 3-NT exposure, with results of LDH release assays (Table 1). MTT assays, which are frequently used to quantify cell survival by measuring cellular reducing capacity, yielded substantially higher estimates of the loss of viability caused by 48 h of incubation of the neurons with 1 mM 3-NP. MK-801 markedly diminished the 3-NP-induced loss of neuronal MTT reducing capacity in the neurons (Fig. 2). Table 1. Morphological and LDH estimates of the incidence of NT2-N neuronal necrosis as a function of 3-NP concentration and time [3-NP] (mM)
0.1 1 5
Two days with 3-NP
Four days with 3-NP
Morph LDH release
Morph LDH release
1% 10% 15%
1% 11% 15%
1% ; 1% 6% ; 2% 13% ; 2%
2% ; 1% 8% ; 2% 13% ; 3%
LDH release results are the means ; S.E.M. of at least three independent experiments. [3-NP] = concentration of 3-NP in the medium. Morph = Morphological estimates of necrosis (proportions of nuclei non-fragmented, and usually shrunken, but TUNEL-negative), and corresponds to total minus apoptotic neurons as shown in Fig. 1.
Neuroprotective e¡ects of MK-801 and other agents We con¢rmed by LDH release assay that neuronal necrosis that resulted from incubation with 1 mM 3-NP for 48 h was totally prevented by MK-801 (Fig. 3). Thus, activation of NMDA^GluR plays an initial and essential role in 3-NP-induced necrosis in NT2 neurons. The non-NMDA^GluR inhibitor, 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), and the L-type VGCC blocker, nifedipine, had signi¢cant, but relatively minor, protective e¡ects. A larger protective e¡ect against NT2N neuronal necrosis was obtained by treatment with xestospongin C, an inhibitor of inositol 1,4,5-trisphosphate (IP3)-sensitive Ca2þ release from ER. One micromolar of xestospongin C used in this study was three times as much as the reported IC50 , 358 nM, for inhibition against IP3-sensitive Ca2þ release (Gafni et al., 1997). Two agents that block the mitochondrial permeability transition (MPT), bongkrekic acid, and cyclosporin A (Li et al., 1997; Dubinsky and Levi, 1998; Friberg et al., 1998; Beutner et al., 1998; Budd et al., 2000), were modestly neuroprotective (Fig. 3). At a 10-fold higher concentration, however, cyclosporin A was no longer protective, and, indeed, was itself toxic to the NT2-N neurons even in the absence of 3-NP (data not shown). The antioxidant, propyl gallate (10 WM), previously reported to diminish 3-NP-induced neuronal apoptosis (Keller et al., 1998), also diminished 3-NP-induced NT2-N neuronal LDH release by 44% ; 4% (mean ; S.E.M., n = 8, P 6 0.01) at 48 h. 3-NP diminishes NT2-N neuronal ATP content. This fall is inhibited by MK-801 3-NP blocks ATP synthesis by inhibition of succinate
NSC 5622 23-7-02
702
W.-T. Lee et al.
Fig. 2. Concentration-dependent e¡ects of 3-NP on MTT reduction by NT2-N neurons in the absence and presence of 10 WM MK-801. Data are means ; S.E.M. of at least three independent experiments.
dehydrogenese, an essential enzyme in the TCA cycle. ATP content of control NT2-N neurons was 12.2 ; 1.5 nM/mg protein (mean ; S.E.M., n = 18). Neuronal ATP content did not fall during the ¢rst 6 h of incubation with 0.1 mM 3-NP, but then declined to 48 ; 9% and 47 ; 2% (mean ; S.E.M., n = 6) of the basal value at 24 and 48 h, respectively (Fig. 4). Higher concentrations of 3-NP caused an earlier and more profound decline. In 1 mM 3-NP, for example, ATP content varied between 67 and 86% of the basal value during the ¢rst 6 h, and had fallen to 22 ; 3% and 25 ; 4% of the control value after 24 and 48 h, respectively. Thus, high concentrations of 3-NP, which led predominantly to neuronal necrosis, caused a much earlier fall in neuronal ATP content. Addition of 10 WM MK-801 concurrently with the 1 mM 3-NP substantially diminished ATP depletion. For example, at 48 h, ATP content remained at 80 ; 2% (n = 10) of the control value, more than threefold higher than in parallel cultures in the absence of MK-801 (Fig. 4). This result demonstrates that NMDA^GluR activation contributed substantially to the NT2-N neuronal ATP depletion caused by 3-NP, likely by accelerating ATP consumption during the early phase of exposure to high concentrations of 3-NP.
tures was not a consequence of increased release of glutamate from NT2-N neurons by 3-NP. 3-NP elicits a consistent early rise in NT2-N neuronal [Ca2+]i Loss of calcium homeostasis is considered to be a key mechanism in NMDA^GluR-mediated neuronal excitotoxicity (Kiedrowski, 1999). During the ¢rst 2 h after addition of 1 mM 3-NP in a Ca2þ -containing medium,
3-NP does not increase extracellular glutamate concentration in NT2-N cultures We previously observed that NT2 neurons secrete glutamate into the culture medium (Rootwelt et al., 1998). In this study, we totally removed the old medium and added 3-NP dissolved in fresh culture medium which did not contain exogenous glutamate. After a 48 h incubation with or without 3-NP, the glutamate concentration of the medium was 98 ; 10 WM in cultures not treated with 3-NP, 58 ; 16 WM in cultures in 1 mM 3-NP, and 74 ; 20 WM in cultures in 5 mM 3-NP. Thus, activation of NMDA^GluR by 3-NP in the NT2-N neuronal cul-
Fig. 3. Protective e¡ects of various treatments against 3-NPinduced necrosis, estimated by LDH release. The neurons were incubated with 1 mM 3-NP (or without, in the control) for two days, with one or more of the following agents : MK-801 (10 WM), CNQX (100 WM), nifedipine (nifed, 10 WM), xestospongin C (xesto, 1 WM), bongkrekic acid (bong, 50 WM), or cyclosporin A (CSA, 1 WM). Data are means ; S.E.M. of at least three independent experiments. * and **, signi¢cantly lower than the % in 3-NP, P 6 0.05 and P 6 0.01, respectively.
NSC 5622 23-7-02
3-NP excitotoxicity in human NT2-N neurons
703
Fig. 4. Time-dependent e¡ects of 3-NP on NT2-N neuronal ATP content. Data are means ; S.E.M. of at least four independent experiments. For the data point shown in the presence of MK-801, S.E.M. bars are contained within the symbol.
there was a gradual increase in [Ca2þ ]i in all NT2-N neurons, from a basal [Ca2þ ]i level of 48 ; 2 nM to 140 ; 12 nM (mean ; S.E.M.). This [Ca2þ ]i increase was prevented by removal of Ca2þ from the medium (Fig. 5A). Thus, in£ux of extracellular Ca2þ is required for 3-NP to induce a sustained increase in NT2-N neuronal [Ca2þ ]i . The rate of the rise in neuronal [Ca2þ ]i for the ¢rst 2 h was markedly slower in the presence of MK801. The L-type VGCC inhibitor, nifedipine, did not further augment the inhibitory e¡ect of MK-801 on this rise in [Ca2þ ]i . We concluded, therefore, that NMDA^GluR played a predominant role in the early rise in [Ca2þ ]i in 3-NP-treated NT2-N neurons. Roles of ER and mitochondria in bu¡ering [Ca2+]i during the ¢rst 30 min of treatment of NT2-N neurons with 3-NP When NT2-N neurons were exposed to 1mM 3-NP in a Ca2þ -free medium, there was still an early rise in [Ca2þ ]i (Fig. 5B). During the ¢rst few minutes, the rate of this rise was nearly identical to that in a Ca2þ -containing medium (compare Fig. 5A, B, noting the changes in x and y scales). This demonstrates that, during this brief initial time period, the increase in [Ca2þ ]i caused by 3-NP was the result of release of Ca2þ from internal stores (ER and/or mitochondria). With longer periods of observation, the increase in [Ca2þ ]i decelerated, con¢rming the requirement for extracellular Ca2þ to sustain a 3-NPinduced [Ca2þ ]i elevation. The rise in [Ca2þ ]i in Ca2þ free medium was totally blocked by addition of xestospongin C plus dantrolene. These drugs inhibit IP3-sensitive ER Ca2þ release and Ca2þ -activated ER Ca2þ release, respectively (Wei and Perry, 1996; Gafni et al., 1997; Yu et al., 1999; Mattson et al., 2000). We concluded, therefore, that release of Ca2þ from ER was largely responsible for the 3-NP-induced initial rise in [Ca2þ ]i . Release of Ca2þ from mitochondria may substantially
increase the rise in [Ca2þ ]i caused by activation of neuronal NMDA^GluR (Mody and MacDonald, 1995; Stout et al., 1998). To evaluate the role of mitochondria in bu¡ering [Ca2þ ]i during the early phase of exposure of NT2-N neurons to 3-NP, carbonyl cyanide p-(tri£uoromethoxy)phenylhydrazone (FCCP), a proton ionophore which blocks mitochondrial retention of Ca2þ (Luo et al., 1997; Khodorov et al., 1999) was added to the Ca2þ -free medium. This elicited a rapid, approximately 20 nM, increase in [Ca2þ ]i . The slope of the rise in [Ca2þ ]i that followed during the next 30 min in these 3-NP-treated neurons was very similar to that in the absence of FCCP (Fig. 5B). We interpret these results to indicate that, in a Ca2þ -free medium, the NT2-N neuronal mitochondria contained a small amount of Ca2þ , su⁄cient, when released, to raise [Ca2þ ]i by 20 nM, and that, during the remainder of the observation period, mitochondrial Ca2þ uptake did not substantially contribute to [Ca2þ ]i homeostasis. Chronic e¡ects of 3-NP on NT2-N neuronal Ca2+ homeostasis We measured [Ca2þ ]i following 24 or 48 h of incubation with 1 mM 3-NP in a Ca2þ -containing medium (chronic studies in a Ca2þ -free medium were not performed, because chronic incubation in Ca2þ -free medium was toxic to these neurons even in the absence of 3-NP). There were two populations of neurons in these 3-NPtreated cultures: the majority (83% at one day, 75% at two days) had restored [Ca2þ ]i to below 100 nM, whereas in the remaining neurons (17% at 24 h, 25% at 48 h), [Ca2þ ]i remained above 100 nM (Table 2). The distributions of [Ca2þ ]i values in these neurons are shown in Fig. 6. Inspection of this ¢gure indicates that the increase in neurons with [Ca2þ ]i above 100 nM between 24 and 48 h of treatment with 3-NP without other additives (column 2 in the ¢gure) was entirely accounted for by
NSC 5622 23-7-02
704
W.-T. Lee et al.
Fig. 5. Acute e¡ects of 1 mM 3-NP on NT2-N neuronal [Ca2þ ]i . After recording basal [Ca2þ ]i for one minute, the perfusion solution was changed to a solution containing the indicated drugs. (A) In a Ca2þ -containing medium, 3-NP induced a linear increase in [Ca2þ ]i , which was attenuated by MK-801. Nifedipine did not further increase the attenuation elicited with MK-801. Changing to a Ca2þ -free medium at 120 min (indicated by the bar) caused the rise in [Ca2þ ]i to cease. (B) In a Ca2þ -free medium, the initial rise in [Ca2þ ]i in the 3-NP-treated neurons was similar to that in a Ca2þ -containing medium, but then slowed. The rise in [Ca2þ ]i was abolished by treatment with xestospongin C and dantrolene. Treatment of the neurons with 1 WM FCCP and 1 mM 3-NP caused an initial, very brief, rise in [Ca2þ ]i that was more rapid than in the absence of FCCP. After the ¢rst 5 min, the slope of the rise was similar to that in the absence of FCCP. Numbers of neurons examined (in two independent experiments) are indicated in parentheses, and are expressed as the means ; S.E.M. Abbreviations and concentrations used: MK, MK-801 10 WM; Nif, nifedipine 10 WM; Xesto, xestospongin C 1 WM; Dant, dantrolene 10 WM; FCCP 1 WM.
neurons in which [Ca2þ ]i was barely above 100 nM. The proportion of neurons in which [Ca2þ ]i was above 100 nM at 24 or 48 h was markedly reduced by treatment with MK-801, and was also signi¢cantly reduced by
treatment with xestospongin C, but not by treatment with CNQX or nifedipine (Table 2). Combination treatments with MK-801 and xestospongin C or MK-801 plus CNQX were more e¡ective in diminishing the incidence
NSC 5622 23-7-02
3-NP excitotoxicity in human NT2-N neurons
705
Table 2. Incidence of NT2-N neurons with [Ca2þ ]i s 100 nM after one or two days in 1 mM 3-NP
Control 3-NP 3-NP+MK-801 3-NP+nifedipine 3-NP+CNQX 3-NP+xestospongin C 3-NP+MK-801+nifedipine 3-NP+MK-801+CNQX 3-NP+MK-801+xestospongin C
One day
Two days
ND 17% (48/280) 6% (17/300)** 20% (13/66) 21% (15/72) 9% (10/106)* 5% (6/126)** 3% (3/115)** 0% (0/105)**
2% (5/244) 25% (44/179) 6% (9/158)** 19% (29/156) 18% (32/171) 14% (26/182) 7% (5/74)** 1% (1/104)** 1% (2/135)**
ND = not done; * and **, signi¢cantly lower than the % in 3-NP, P 6 0.05 and P 6 0.01, respectively. All drugs were added at time 0, and [Ca2þ ]i was measured at 24 and 48 h, using fura-2/AM. In parentheses, we indicate the number of neurons assayed in which [Ca2þ ]i was greater than 100 nM, and the total number of neurons assayed for each condition at each time-point. Concentrations used: 3-NP 1 mM ; MK-801 10 WM; nifedipine 10 WM; CNQX 100 WM; xestospongin C 1 WM.
of neurons with [Ca2þ ]i greater than 100 nM than treatment with MK-801 alone.
DISCUSSION
Our aim in this study has been to characterize the changes in Ca2þ homeostasis that accompany neuronal energy deprivation and precede neuronal death. We chose conditions (1 mM 3-NP) and a time-frame (¢rst 48 h of 3-NP exposure) in which most NT2-N neurons died by necrosis rather than apoptosis. [Ca2þ ]i increased in a linear fashion in the neurons during the ¢rst few hours after addition of 3-NP, to almost three-fold higher than the resting level by 2 h. Several prior reports of the acute e¡ects of 3-NP, as well as the present study, have documented that activation of NMDA^GluR is one of responsible mechanisms for this early rise in neuronal [Ca2þ ]i (Greene et al., 1998; Olsen et al., 1999). Addition of MK-801 diminished this early rise by 70%, with no additional inhibitory e¡ect by nifedipine, thus indicating the relative importance of activation of NMDA^GluR, and the relative unimportance of activation of L-type VGCC in mediating this increase in [Ca2þ ]i . Our shortterm observations in a Ca2þ -free medium demonstrated that release of internal stores of Ca2þ contributed substantially to the ¢rst few minutes of 3-NP-induced rise in [Ca2þ ]i , but were quantitatively less signi¢cant at later time-points. Because the early rise in [Ca2þ ]i was observed in all of the 3-NP-treated NT2-N neurons, but most of these neurons survived prolonged 3-NP exposure, it is clear that an early [Ca2þ ]i rise of the magnitude we observed was not su⁄cient to elicit necrosis. While little has previously been known about the chronic e¡ects of 3-NP on neuronal [Ca2þ ]i homeostasis, prior studies have emphasized the signi¢cance of prolonged [Ca2þ ]i dysregulation and consequent mitochondrial calcium overload in the pathophysiology of neuronal excitotoxicity (Stout et al., 1998; Nicholls et al., 1999; Reynolds, 1999). We observed that, at 24 and 48 h after addition of 1 mM 3-NP to the medium, [Ca2þ ]i remained above 100 nM in 17% and 25% of the neurons, respectively. The incidence of neurons with [Ca2þ ]i above 100 nM at 24 and 48 h was signi¢cantly decreased by MK-801. MK-801 was also by far the most
e¡ective of the drugs we tested in protecting the NT2-N neurons against necrosis. CNQX or nifedipine alone did not signi¢cantly diminish the proportion of neurons maintaining their [Ca2þ ]i above 100 nM at 24 and 48 h, but the combinations of MK-801 and CNQX or MK-801 and nifedipine were signi¢cantly more e¡ective in this regard than MK-801 alone, and CNQX or nifedipine alone did marginally diminish the incidence of neuronal necrosis. These data indicate that non-NMDA^GluR and VGCC do play roles in NT2-N neuronal 3-NP toxicity, though these are quantitatively minor. Since [Ca2þ ]i was 140 ; 12 nM (mean ; S.E.M.) at 2 h, and had fallen to below 100 nM in all but 17% of the neurons by 24 h, most of the NT2-N neurons must have lowered [Ca2þ ]i toward normal in the interval between 2 and 24 h, despite marked ATP depletion. Bcl-2 was a candidate responsible for the capacity of these neurons to restore normal or near-normal [Ca2þ ]i , because Bcl-2 expression in NT2-N neurons is signi¢cantly increased after terminal di¡erentiation (Singer et al., 1998; Guillemain et al., 2000). However, no further induction of Bcl-2 protein was observed after the exposure to 3-NP (our preliminary observation), although this might explain why NT2-N neurons are more resistant to energy deprivation compared to primary neuronal cultures. For the following reasons, it is likely those neurons with chronically elevated [Ca2þ ]i had not yet undergone necrosis but were mostly destined not to survive. First, at 24 h only 1% of the neurons in the cultures treated with 1 mM 3-NP showed morphological features of necrosis. Second, successful fura-2 imaging requires a cell to maintain its plasma membrane intact, in order to retain loaded fura-2; hence, neurons already undergoing necrosis are probably not represented in our [Ca2þ ]i data. Third, the incidence of neurons with [Ca2þ ]i above 100 nM at 24 and 48 h correlated well with actual neuronal death evaluated by LDH assay. A concentration of 3-NP (0.1 mM) su⁄cient to induce a gradually increasing incidence of NT2-N neuronal apoptosis, but not signi¢cant necrosis, caused a delayed fall in neuronal ATP content, whereas concentrations of 3-NP su⁄cient to induce substantial NT2-N neuronal necrosis (1 or 5 mM) caused an early and more profound fall in neuronal ATP content. With 1 or 5 mM 3-NP, ATP content reached a near-steady-state level less than a
NSC 5622 23-7-02
706
W.-T. Lee et al.
Fig. 6. Scattergram of [Ca2þ ]i values at 24 and 48 h in 3-NP-treated NT2-N neurons. The concentration of 3-NP in the medium was 1 mM. Each dot represents a single neuron. Data pooled from three independent experiments are shown. Abbreviations and concentrations used: MK, MK-801 10 WM; Nif, nifedipine 10 WM; Xes, xestospongin C 1 WM; CNQX 100 WM.
third that of control neurons within 24 h. The reduced ATP level was presumably maintained by continued ATP generation by glycolysis. In parallel experiments in which NMDA^GluR activation was blocked with MK-801, the neurons were able to maintain ATP at 80% of the control level, indicating that, with the strong inhibition of aerobic ATP production, ATP consumption induced by NMDA^GluR is a critical factor to determine the severity of ATP depletion in NT2-N neurons. One factor likely to have contributed to the ATP-sparing e¡ect of inhibiting NMDA^GluR was a diminution in ATP consumption by plasma membrane Naþ ,Kþ -ATPase in response to NMDA^GluR-mediated neuronal Naþ loading (Chinopoulos et al., 2000). Collapse of mitochondrial membrane potential induced by activation of NMDA^GluR is also known to induce reverse operation of the ATP synthase and reduce ATP/ADP ratio
(Budd and Nicholls, 1996). Our ATP data also suggest that early and severe ATP depletion is likely to be an essential step leading to chronic Ca2þ dysregulation. The early increase in [Ca2þ ]i elicited by 3-NP in a Ca2þ -free medium in the NT2-N neurons was entirely prevented by addition of xestospongin C and dantrolene, indicating that Ca2þ mobilization from intracellular Ca2þ storage into cytosol occurred over the initial phase of 3-NP administration. Longer term, xestospongin C diminished the proportion of NT2-N neurons in which [Ca2þ ]i was chronically elevated, and, in conjunction with MK-801, almost totally prevented chronic elevation in [Ca2þ ]i . Finally, xestospongin C signi¢cantly diminished the proportion of NT2-N neurons undergoing necrosis at 48 h. These results raise the idea that initial Ca2þ mobilization from intracellular Ca2þ storage as an acute e¡ect of 3-NP enhances NMDA^GluR-medi-
NSC 5622 23-7-02
3-NP excitotoxicity in human NT2-N neurons
ated neurotoxicity. It has been reported recently that, as a model of Huntington’s disease, a 20-min treatment with 3-NP produced a long-term potentiation of the NMDA^GluR-mediated synaptic excitation in striatal spiny neurons by the mechanism requiring D2 dopamine receptor activation (Calabresi et al., 2001). Although the exact mechanism by which 3-NP induces Ca2þ mobilization from intracellular Ca2þ storage in NT2-N neurons remains to be clari¢ed, Ca2þ release particularly from IP3-sensitive Ca2þ storage might also alter the physiological characteristics of NMDA^GluR. In addition, our data strongly suggest that drugs that block ER Ca2þ release deserve consideration as therapeutic agents in stroke and other disorders of brain energy metabolism (Mody and MacDonald, 1995; Wei and Perry, 1996; Yu et al., 1999). We observed that two MPT inhibitors and an antioxidant diminished 3-NP-induced NT2-N neuronal necrosis. The MPT results con¢rm a prior study in which one of these inhibitors, cyclosporin A, diminished the extent of neuronal necrosis elicited by systemic administration of 3-NP (Leventhal et al., 2000). Antioxidants and free radical spin traps have also been reported to minimize 3-NP toxicity in vivo (Schulz et al., 1996; Fontaine et al., 2000), and glutathione peroxidase knockout mice are hypersusceptible to 3-NP-induced striatal neuronal loss (Klivenyi et al., 2000). In our own study, NT2-N neuronal necrosis caused by 3-NP was diminished by addition of an antioxidant, propyl gallate, to the medium.
707
Together, the protective e¡ects of MPT inhibitors and antioxidants suggest that reactive oxygen species produced in excess by mitochondria play a role in 3-NPmediated NT2-N neuronal necrosis, as they do in ischemic necrosis in these neurons (Rootwelt et al., 1998; Kim et al., 2000; Almaas et al., 2000).
CONCLUSION
In conclusion, our study is the ¢rst to employ human neurons to study the pathophysiology of 3-NP neurotoxicity, and yielded three novel ¢ndings. First, we demonstrated the preponderant role that activation of NMDA^ GluR plays in mediating 3-NP-induced depletion of neuronal ATP. Second, by assaying neuronal [Ca2þ ]i at 24 and 48 h after application of 3-NP, we demonstrated chronic [Ca2þ ]i dysregulation similar to that previously reported in neuronal excitotoxicity (Itoh et al., 1998; Nicholls et al., 1999). Third, we demonstrated that inhibiting ER Ca2þ release substantially enhanced the capacity of the 3-NP-treated neurons to maintain [Ca2þ ]i homeostasis and resist necrosis.
Acknowledgements/Supported by the Muscular Dystrophy Association, the Wolfson Fund, and by NIH Grants NS25044, NS08075, and HD26979. We appreciate the assistance of Marc Yudko¡ in the glutamate assays.
REFERENCES
Almaas, R., Saugstad, O.D., Pleasure, D., Rootwelt, T., 2000. E¡ect of anesthetics on hydroxyl radicals, lipid peroxidation and hypoxic neuronal cell death in human NT2-N neurons. Anesthesiology 92, 764^774. Beal, M.F., Brouillet, E., Jenkins, B.G., Ferrante, R.J., Kowall, N.W., Miller, J.M., Storey, E., Srivastava, R., Rosen, B.R., Hyman, B.T., 1993. Neurochemical and histologic characterization of striatal excitotoxic lesions produced by the mitochondrial toxin 3-nitropropionic acid. J. Neurosci. 13, 4181^4192. Beutner, G., Ruck, A., Riede, B., Brdiczka, D., 1998. Complexes between porin, hexokinase, mitochondrial creatine kinase and adenylate translocator display properties of the permeability transition pore. Implication for regulation of permeability transition by the kinases. Biochim. Biophys. Acta 1368, 7^18. Budd, S.L., Nicholls, D.G., 1996. A reevaluation of the role of mitochondria in neuronal Ca2þ homeostasis. J. Neurochem. 66, 403^411. Budd, S.L., Tenneti, L., Lishnak, T., Lipton, S.A., 2000. Mitochondrial and extramitochondrial apoptotic signaling pathways in cerebrocortical neurons. Proc. Natl. Acad. Sci. USA 97, 6161^6166. Calabresi, P., Gubellini, P., Picconi, B., Centonze, D., Pisani, A., Bonsi, P., Greengard, P., Hipskind, R.A., Borrelli, E., Bernardi, G., 2001. Inhibition of mitochondrial complex II induces a long-term potentiation of NMDA-mediated synaptic excitation in the striatum requiring endogenous dopamine. J. Neurosci. 21, 5110^5120. Chinopoulos, C., Tretter, L., Rozsa, A., Adam-Vizi, V., 2000. Exacerbated response to oxidative stress by an Naþ load in isolated nerve terminals : the role of ATP depletion and rise of [Ca2þ ]i . J. Neurosci. 20, 2094^2103. Dubinsky, J.M., Levi, Y., 1998. Calcium-induced activation of the mitochondrial permeability transition in hippocampal neurons. J. Neurosci. Res. 53, 728^741. Fink, S.L., Ho, D.Y., Sapolsky, R.M., 1996. Energy and glutamate dependency of 3-nitropropionic acid neurotoxicity in culture. Exp. Neurol. 138, 298^304. Fontaine, M.A., Geddes, J.W., Banks, A., Butter¢eld, D.A., 2000. E¡ect of exogenous and endogenous antioxidants on 3-nitropropionic acidinduced in vivo oxidative stress and striatal lesions. Insights into Huntington’s disease. J. Neurochem. 75, 1709^1715. Friberg, H., Ferrand-Drake, M., Bengtsson, F., Halestrap, A.P., Wieloch, T., 1998. Cyclosporin A, but not FK 506, protects mitochondria and neurons against hypoglycemia damage and implicates the mitochondrial permeability transition in cell death. J. Neurosci. 18, 5151^5159. Fukuda, A., Deshpande, S.B., Shimano, Y., Nishino, H., 1998. Astrocytes are more vulnerable than neurons to cellular Ca2þ overload induced by a mitochondrial toxin, 3-nitropropionic acid. Neuroscience 87, 497^507. Gafni, J., Munsch, J.A., Lam, T.H., Catlin, M.C., Costa, L.G., Molinski, T.F., Pessah, I.N., 1997. Xestospongins : potent membrane permeable blockers of the inositol 1,4,5-trisphosphate receptor. Neuron 19, 723^733. Gavrieli, Y., Sherman, Y., Ben-Sasson, S., 1992. Identi¢cation of programmed cell death in situ via speci¢c labeling of nuclear DNA fragmentation. J. Cell Biol. 119, 493^501. Greene, J.G., Sheu, S.-S., Gross, R.A., Greenamyre, J.T., 1998. 3-Nitropropionic acid exacerbates N-methyl-D-aspartate toxicity in striatal culture by multiple mechanisms. Neuroscience 84, 503^510. Grynkiewicz, G., Poenie, M., Tsien, R.Y., 1985. A new generation of Ca indicators with greatly improved £uorescence properties. J. Biol. Chem. 260, 3440^3450.
NSC 5622 23-7-02
708
W.-T. Lee et al.
Guillemain, I., Gaboyard, S., Fontes, G., Saunier, M., Privat, A., Patey, G., 2000. Di¡erential expression of Bcl-2-related proteins in di¡erentiating NT2 cells. NeuroReport 11, 1421^1425. Hardy, M., Younkin, D., Tang, C.-M., Pleasure, J., Williams, M., Pleasure, D., 1994. Expression of non-NMDA glutamate receptor channels by clonal NT2-N human neurons. J. Neurochem. 63, 482^489. Itoh, T., Itoh, A., Horiuchi, K., Pleasure, D., 1998. AMPA receptor-mediated excitotoxicity in human NT2-N neurons results from loss of intracellular Ca2þ homeostasis following marked elevation of intracellular Naþ . J. Neurochem. 71, 112^124. Jones, B.N., Gilligan, J.P., 1983. o-Phthaldialdehyde precolumn derivatization and reversed-phase high-performance liquid chromatography of polypeptide hydrolysates and physiological £uids. J. Chromatogr. 266, 471^482. Keller, J.N., Guo, Q., Holtsberg, F.W., Bruce-Keller, A.J., Mattson, M.P., 1998. Increased sensitivity to mitochondrial toxin-induced apoptosis in neural cells expressing mutant presenilin-1 is linked to perturbed calcium homeostasis and enhanced oxyradical production. J. Neurosci. 18, 4439^4450. Khodorov, B., Pinelis, V., Storozhevykh, T., Yuravichus, A., Khaspekhov, L., 1999. Blockade of mitochondrial Ca2þ uptake by mitochondrial inhibitors ampli¢es the glutamate-induced calcium response in cultured cerebellar granule cells. FEBS Lett. 458, 162^166. Kiedrowski, L., 1999. N-Methyl-D-aspartate excitotoxicity: relationships among plasma membrane potential, Naþ /Ca2þ exchange, mitochondrial Ca2þ overload, and cytoplasmic concentrations of Ca2þ , Hþ , and Kþ . Mol. Pharmacol. 56, 619^632. Kim, G.W., Copin, J.C., Kawase, M., Chen, S.F., Sato, S., Gobbel, G.T., Chan, P.H., 2000. Excitotoxicity is required for induction of oxidative stress and apoptosis in mouse striatum by the mitochondrial toxin, 3-nitropropionic acid. J. Cereb. Blood Flow Metab. 20, 119^129. Klivenyi, P., Andreassen, O.A., Ferrante, R.J., Dedeoglu, A., Mueller, G., Lancelot, E., Bogdanov, M., Andersen, J.K., Jiang, D., Beal, M.F., 2000. Mice de¢cient in cellular glutathione peroxidase show increased vulnerability to malonate, 3-nitropropionic acid, and 1-methyl-4-phenyl1,2,5,6-tetrahydropyridine. J. Neurosci. 20, 1^7. Lee, W.T., Shen, Y.Z., Chang, C., 2000. Neuroprotective e¡ect of lamotrigine and MK-801 on rat brain lesions Induced by 3-nitropropionic acid : evaluation by magnetic resonance imaging and in vivo proton magnetic resonance spectroscopy. Neuroscience 95, 89^95. Leventhal, L., Sortwell, C.E., Hanbury, R., Collier, T.J., Kordower, J.H., Pal¢, S., 2000. Cyclosporin A protects striatal neurons in vitro and in vivo from 3-nitropropionic acid toxicity. J. Comp. Neurol. 425, 471^478. Li, P.A., Uchino, H., Elmer, E., Siesjo, B.K., 1997. Amelioration by cyclosporin A of brain damage following 5 or 10 min of ischemia in rats subjected to preischemic hyperglycemia. Brain Res. 753, 133^140. Luo, Y., Bond, J.D., Ingram, V.M., 1997. Compromised mitochondrial function leads to increased cytosolic calcium and to activation of MAP kinases. Proc. Natl. Acad. Sci. USA 94, 9705^9710. Mattson, M.P., Zhu, H., Yu, J., Kindy, M.S., 2000. Presenilin-1 mutation increases neuronal vulnerability to focal ischemia in vivo and to hypoxia and glucose deprivation in cell culture : involvement of perturbed calcium homeostasis. J. Neurosci. 20, 1358^1364. Mody, I., MacDonald, J.F., 1995. NMDA receptor-dependent excitotoxicity: the role of intracellular Ca2þ release. Trends Pharmacol. Sci. 16, 356^359. Nicholls, D.G., Budd, S.L., Castilho, R.F., Ward, M.W., 1999. Glutamate excitotoxicity and neuronal energy metabolism. Ann. NY Acad. Sci. 893, 1^12. Olsen, C., Rustad, A., Fonnum, F., Paulsen, R.E., Hassel, B., 1999. 3-Nitropropionic acid: an astrocyte-sparing neurotoxin in vitro. Brain Res. 850, 144^149. Pal¢, S., Ferrante, R.J., Brouillet, E., Beal, M.F., Dolan, R., Guyot, M.C., Peschanski, M., Hantraye, P., 1996. Chronic 3-nitropropionic acid treatment in baboons replicates the cognitive and motor de¢cits of Huntington’s disease. J. Neurosci. 16, 3019^3025. Pang, Z., Geddes, J.W., 1997. Mechanisms of cell death induced by mitochondrial toxin 3-nitropropionic acid : acute excitotoxic necrosis and delayed apoptosis. J. Neurosci. 17, 3064^3073. Pleasure, S.J., Page, C., Lee, V.M.-Y., 1992. Pure, postmitotic, polarized human neurons derived from Ntera 2 cells provide a system for expressing exogenous proteins in terminally di¡erentiated neurons. J. Neurosci. 12, 1802^1815. Reynolds, I.J., 1999. Mitochondrial membrane potential and the permeability transition in excitotoxicity. Ann. NY Acad. Sci. 893, 33^41. Rootwelt, T., Dunn, M., Yudko¡, M., Itoh, T., Almaas, R., Pleasure, D., 1998. Hypoxic cell death in human NT2-N neurons : involvement of NMDA- and non-NMDA-glutamate receptors. J. Neurochem. 71, 1544^1553. Schulz, J.B., Henshaw, D.R., MacGarvey, U., Beal, M.F., 1996. Involvement of oxidative stress in 3-nitropropionic acid neurotoxicity. Neurochem. Int. 29, 167^171. Singer, C.A., Rogers, K.L., Dorsa, D.M., 1998. Modulation of Bcl-2 expression : a potential component of estrogen protection in NT2 neurons. NeurouReport 9, 2565^2568. Stout, A.K., Raphael, H.M., Kanterewicz, B.I., Klann, E., Reynolds, I.J., 1998. Glutamate-induced neuron death requires mitochondrial calcium uptake. Nature Neurosci. 1, 366^373. Wei, H., Perry, D.C., 1996. Dantrolene is cytoprotective in two models of neuronal cell death. J. Neurochem. 67, 2390^2398. Wullner, U., Young, A.B., Penney, J.B., Beal, M.F., 1994. 3-Nitropropionic acid toxicity in the striatum. J. Neurochem. 63, 1772^1781. Younkin, D., Tang, C.-M., Hardy, M., Reddy, U.R., Pleasure, S., Lee, V., Pleasure, D., 1993. Induction of NMDA receptor expression and glutamate excitotoxicity in the clonal human cell line NTera 2. Proc. Natl. Acad. Sci. USA 90, 2174^2178. Yu, Z., Luo, H., Fu, W., Mattson, M.P., 1999. The endoplasmic reticulum stress-responsive protein GRP78 protects neurons against excitotoxicity and apoptosis: suppression of oxidative stress and stabilization of calcium homeostasis. Exp. Neurol. 155, 302^314. (Accepted 4 February 2002)
NSC 5622 23-7-02
PII: S 0 3 0 6 - 4 5 2 2 ( 0 2 ) 0 0 1 7 7 - X www.neuroscience-ibro.com
ACUTE AND CHRONIC ALTERATIONS IN CALCIUM HOMEOSTASIS IN 3-NITROPROPIONIC ACID-TREATED HUMAN NT2-N NEURONS W.-T. LEE,a;b T. ITOHb and D. PLEASUREb a b
Department of Pediatrics, National Taiwan University Hospital, Taipei, Taiwan
Division of Neurology Research, Room 516H, Abramson Research Building, The Children’s Hospital of Philadelphia, 34th Street and Civic Center Boulevard, Philadelphia, PA 19104, USA
Abstract33-Nitropropionic acid (3-NP), an irreversible inhibitor of succinate dehydrogenase, induced ATP depletion and both necrosis and apoptosis in human NT2-N neurons. Necrosis occurred predominantly within the ¢rst two days, and increased in a dose-dependent fashion with the concentration of 3-NP, whereas apoptosis was observed after 24 h or later at a similar rate in 0.1 mM and 5 mM 3-NP. We focused our e¡orts on intracellular calcium homeostasis during the ¢rst 48 h in 1 mM 3-NP, a period during which 10% of the neurons died by necrosis and 3% by apoptosis. All NT2-N neurons showed a stereotyped [Ca2þ ]i rise, from 48 ; 2 to 140 ; 12 nM (mean ; S.E.M.), during the ¢rst 2 h in 3-NP. Despite severe ATP depletion, however, [Ca2þ ]i remained above 100 nM in only 17% and 25% of the NT2-N neurons after 24 and 48 h in 3-NP, respectively, indicating that most neurons were able to recover from acute [Ca2þ ]i rise, and suggesting that chronic [Ca2þ ]i dysregulation is a better indicator of subsequent necrosis. Blockade of N-methyl-D-aspartate^glutamate receptor by MK-801 substantially ameliorated 3-NP-induced ATP depletion, subsequent chronic [Ca2þ ]i elevation, and survival. Moreover, xestospongin C, an inhibitor of endoplasmic reticulum Ca2þ release, enhanced the capacity of NT2-N neurons to maintain [Ca2þ ]i homeostasis and resist necrosis while subjected to sustained energy deprivation. As far as we know, this report is the ¢rst to employ human neurons to study the pathophysiology of 3-NP neurotoxicity. C 2002 IBRO. Published by Elsevier Science Ltd. All rights reserved. Key words: NT2-N neurons, 3-nitropropionic acid, necrosis, glutamate receptors, calcium, xestospongin C.
vival of these cultured neurons is enhanced by inhibiting NMDA^GluR (Fink et al., 1996; Pang and Geddes, 1997). These observations indicate that activation of NMDA^GluR plays an important role in 3-NP neurotoxicity. This activation is due, at least in part, to partial neuronal plasma membrane depolarization, which releases the voltage-dependent Mg2þ block of NMDA^ GluR (Greene et al., 1998). We previously reported that NT2-N neurons, prepared by retinoic acid-induced di¡erentiation of the human teratocarcinoma line, Ntera-2 (Pleasure et al., 1992), are susceptible to NMDA^GluR- and non-NMDA^ GluR-mediated excitotoxicity (Younkin et al., 1993; Hardy et al., 1994; Itoh et al., 1998), and that oxygen/ glucose deprivation causes NT2-N neuronal necrosis by an NMDA^GluR-mediated mechanism (Rootwelt et al., 1998; Almaas et al., 2000). In the present study, we examined the contributions of NMDA^GluR, nonNMDA^GluR, voltage-gated Ca2þ channels (VGCC), endoplasmic reticulum (ER), and mitochondria to acute and chronic alterations in [Ca2þ ]i homeostasis in these human neurons during a 3-NP treatment regimen chosen to elicit predominantly neuronal necrosis.
Systemic or intrahemispheric administration of 3-nitropropionic acid (3-NP), an irreversible inhibitor of succinate dehydrogenase, elicits striatal neuronal degeneration in rodents and non-human primates, and recapitulates many of the clinical and pathological features of Huntington’s disease (Beal et al., 1993; Wullner et al., 1994; Pal¢ et al., 1996). This neuronal degeneration is diminished by administration of MK-801, an N-methyl-Daspartate^glutamate receptor (NMDA^GluR) inhibitor (Kim et al., 2000; Lee et al., 2000). In primary cultures of rodent CNS, 3-NP elicits an early increase in neuronal [Ca2þ ]i (Greene et al., 1998; Fukuda et al., 1998; Keller et al., 1998; Olsen et al., 1999), and both apoptotic and necrotic neuronal death. As was observed in vivo, sur-
*Corresponding author. Tel. : +1-215-590-2090; fax: +1-215-5903709. E-mail address: [email protected] (D. Pleasure). Abbreviations : CNQX, 6-cyano-7-nitroquinoxaline-2,3-dione; ER, endoplasmic reticulum; FCCP, carbonyl cyanide p-(tri£uoromethoxy)phenylhydrazone ; HEPES, N-2-hydroxyethylpiperazineNP-2-ethane-sulfonic acid; IP3, inositol 1,4,5-trisphosphate ; LDH, lactate dehydrogenase ; MK-801, dizocilpine ; MPT, mitochondrial permeability transition ; MTT, 3-(4,5-dimethylthiazol2-yl)-2,5-diphenyltetrazolium bromide; 3-NP, 3-nitropropionic acid ; NADH, nicotinamide adenine dinucleotide; NMDA^ GluR, N-methyl-D-aspartate^glutamate receptor ; TUNEL, terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling; VGCC, voltage-gated Ca2þ channels.
EXPERIMENTAL PROCEDURES
Cell culture NT2-N neurons were prepared and maintained as previously 699
NSC 5622 23-7-02
700
W.-T. Lee et al.
described (Pleasure et al., 1992; Itoh et al., 1998; Rootwelt et al., 1998). The medium contained 2 mM glutamine (freshly prepared) and 10% (v/v) fetal bovine serum, without added glutamate. The neurons were plated in six-well, 12-well and 24-well plates or small chambers that had been coated with polylysine and Matrigel (Becton Dickinson, Bedford, MA, USA). Six-well plates were plated at 2U106 cells/well, 12-well plates at 1U106 cells/well, and 24-well plates at 0.5U106 cells/well. The cultures were studied between four and ¢ve weeks after termination of retinoic acid di¡erentiation. All experiments were done two to four days after the last feeding. Cell death assays To expose NT2-N neurons to 3-NP, the culture medium was totally replaced with fresh medium containing 3-NP at various concentrations and no exogenous glutamate. At the indicated time-points, the cells were treated with Triton X-100 to permeabilize plasma membranes, and double-stained with propidium iodide and for DNA nicking by the terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) procedure (Gavrieli et al., 1992; Pang and Geddes, 1997). Nuclei of dead cells were condensed and/or fragmented, and distinctly di¡erent from the oval or round nuclei of live cells. We counted nuclei of dead cells that were positive by the TUNEL reaction as apoptotic, and cells with condensed or fragmented nuclei that were TUNEL-negative as necrotic. A total of 300 to 500 neurons were counted for each experiment. In each experimental set, we simultaneously examined NT2-N neuronal cultures that had not been exposed to 3-NP in order to determine control incidences of apoptosis and necrosis. Plasma membrane integrity was also assessed by measuring lactate dehydrogenase (LDH) release from the cells to the medium (Younkin et al., 1993). LDH measurements were done with the Cytotox 96 non-radioactive cytotoxicity assay kit (Promega, Madison, WI, USA), based on the conversion of pyruvic acid to lactic acid by LDH in the presence of NADH, according to the manufacturer’s protocol. At indicated time-points after treatment with 3-NP, 50 Wl of culture medium was removed for LDH assay. The NT2-N cells were then lysed with Triton X-100 to a ¢nal concentration of 0.1% (w/v). The plates were stored at 37‡C for 45 min, and then 50 Wl of cell lysate was taken for the LDH measurement. The percentage of the cell death was expressed as the ratio of LDH content in the medium before and before plus after cell lysis U100. Reducing capacity of the neurons was assessed by the conversion of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) to formazan crystals as described previously (Itoh et al., 1998). Formazan production was measured 2 h after adding MTT (0.5 mg/ml) to the culture medium by spectroscopy at 560 nm, and was expressed as the percentage of simultaneously run control values.
neurons in randomly chosen ¢elds at room temperature (23^ 25‡C) as previously described (Itoh et al., 1998). In brief, the NT2-N neurons cultured on coverslips were loaded with 5 WM fura-2/AM with 0.02% (w/v) pluronic F-127 in the standard recording solution (NaCl 140, KCl 5.4, CaCl2 1.8, MgCl2 0.8, Na2 HPO4 0.8, HEPES 10, and D-glucose 25 (in mM), adjusted to pH 7.4 with NaOH) for 45 min. The coverslip was attached to a perfusion chamber (RC-21B; Warner Instrument, USA) on the stage of an upright epi£uorescence microscope (Optiphot, Nikon, Japan), and emission £uorescence images at 510 nm were alternatively taken at 340 and 380 nm excitation wavelengths. [Ca2þ ]i values were calculated from the emission ratio (R) at 340 and 380 nm excitation wavelengths, using the formula described by Grynkiewicz et al. (1985): [Ca2þ ]i = bKd (R3Rmin )/ (Rmax 3R). Kd is the e¡ective dissociation constant of fura-2, and b is the ratio of £uorescence intensity at 380 nm for Ca2þ -free and Ca2þ -saturated dye. Rmin and Rmax are the emission ratios obtained under Ca2þ -free and Ca2þ -saturated conditions, respectively. These parameters were obtained by in vitro calibration. We previously demonstrated that [Ca2þ ]i measurements could be done in NT2-N neurons up to 3 h, if emission images were taken at a 2.5 min interval to minimize damage to the cells by UV exposure as well as photobleaching of loaded fura-2 (Itoh et al., 1998). Even when the frequency of UV exposure was reduced, however, we still observed signi¢cant death of NT2N neurons at 24 h after UV exposure. Therefore, in order to evaluate chronic changes in [Ca2þ ]i after 24 or 48 h with 3-NP, we used sister cultures treated identically to those for acute experiments. Culture medium glutamate concentration Assays of glutamate in the medium were performed as previously described (Rootwelt et al., 1998). After centrifugation to remove cell debris, 0.5 ml of medium was mixed with 1 ml 100 mM Tris bu¡er (pH 7.0), and passed over a column containing AG-1 resin (chloride form; 100^200 mesh; 0.5 ml bed volume) (Bio-Rad, Hercules, CA, USA) to remove excessive glutamine. Glutamine was eluted with 4 ml of deionized water. Glutamate was then eluted with 4 ml of 2 M HCl and analyzed by high pressure liquid chromatography, using precolumn derivatization with o-phthalaldehyde and £uorescent detection (Jones and Gilligan, 1983) with an internal standard. Statistics In most instances (as indicated in the text and ¢gure legends), the results were expressed as means ; S.E.M., with statistical signi¢cances calculated by paired t-test or ANOVA. Chi-square testing was done for analysis of statistical signi¢cance of the data in Table 2.
Neuronal ATP content ATP was quanti¢ed by a luciferin/luciferase-based assay. Twelve-well culture plates were rinsed with ice-cold phosphatebu¡ered saline (PBS), and the neurons were then lysed with 0.1 ml of ATP releasing bu¡er (Sigma). ATP concentrations were measured using the ATP Bioluminescence Assay kit CH II (Boehringer Mannheim) and a luminometer. Fifty microliters of sample was mixed with 50 Wl of luciferase reagent. After 1 s, the reaction was recorded for 10 s. A standard curve was constructed by assaying calibration samples of known ATP concentration. Samples were diluted so that the readings fell within the linear part of the standard curve. Protein content was determined using the BCA protein assay kit (Pierce, Rockford, IL, USA). The ATP levels at each time-point were compared with normal controls at the same time-points. ATP content was expressed as percentage of non-3-NP-treated control at each time-point. [Ca2+]i measurements [Ca2þ ]i measurements were performed on all isolated NT2-N
RESULTS
Acute necrosis is induced by 1 mM or higher 3-NP in NT2 neurons Treatment of NT2-N neurons with 3-NP induced two modes of neuronal death, apoptotic and non-apoptotic (necrotic). Apoptosis, demonstrated by TUNEL-positivity, increased linearly with incubation time and was independent of 3-NP concentration (Fig. 1B). In contrast, necrosis, estimated by subtracting the TUNEL-positive population from total death, increased signi¢cantly with concentrations of 3-NP of more than 1 mM. This increase was mainly observed within 48 h after 3-NP was added. Thereafter, the slopes of increases in total cell death were similar to those observed in TUNEL-
NSC 5622 23-7-02
3-NP excitotoxicity in human NT2-N neurons
701
Fig. 1. Time- and 3-NP-concentration-dependent apoptotic and total death of NT2-N neurons, assessed by counting nuclei of dead neurons and those positive to TUNEL staining (see Experimental procedures for detail). Open circles in each panel show background death in two independent control cultures. We always observed similar percentages of the background during the observation periods. Data are means ; S.D. of three independent experiments.
positive cells (Fig. 1A, B). These results indicate that, whereas delayed apoptosis predominates at 0.1 mM 3-NP, 1 mM or higher 3-NP induces NT2-N neuronal necrosis, mainly within 48 h. We focused on this necrotic 3-NP-induced death in this study. As shown in Fig. 1, control NT2-N cultures contained between 12% to 19% dead cells by morphological evaluation. We calculated the actual population acutely killed by 3-NP at each time-point by subtracting this control background level. Results of this calculation correlated well, during the ¢rst four days of 3-NT exposure, with results of LDH release assays (Table 1). MTT assays, which are frequently used to quantify cell survival by measuring cellular reducing capacity, yielded substantially higher estimates of the loss of viability caused by 48 h of incubation of the neurons with 1 mM 3-NP. MK-801 markedly diminished the 3-NP-induced loss of neuronal MTT reducing capacity in the neurons (Fig. 2). Table 1. Morphological and LDH estimates of the incidence of NT2-N neuronal necrosis as a function of 3-NP concentration and time [3-NP] (mM)
0.1 1 5
Two days with 3-NP
Four days with 3-NP
Morph LDH release
Morph LDH release
1% 10% 15%
1% 11% 15%
1% ; 1% 6% ; 2% 13% ; 2%
2% ; 1% 8% ; 2% 13% ; 3%
LDH release results are the means ; S.E.M. of at least three independent experiments. [3-NP] = concentration of 3-NP in the medium. Morph = Morphological estimates of necrosis (proportions of nuclei non-fragmented, and usually shrunken, but TUNEL-negative), and corresponds to total minus apoptotic neurons as shown in Fig. 1.
Neuroprotective e¡ects of MK-801 and other agents We con¢rmed by LDH release assay that neuronal necrosis that resulted from incubation with 1 mM 3-NP for 48 h was totally prevented by MK-801 (Fig. 3). Thus, activation of NMDA^GluR plays an initial and essential role in 3-NP-induced necrosis in NT2 neurons. The non-NMDA^GluR inhibitor, 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), and the L-type VGCC blocker, nifedipine, had signi¢cant, but relatively minor, protective e¡ects. A larger protective e¡ect against NT2N neuronal necrosis was obtained by treatment with xestospongin C, an inhibitor of inositol 1,4,5-trisphosphate (IP3)-sensitive Ca2þ release from ER. One micromolar of xestospongin C used in this study was three times as much as the reported IC50 , 358 nM, for inhibition against IP3-sensitive Ca2þ release (Gafni et al., 1997). Two agents that block the mitochondrial permeability transition (MPT), bongkrekic acid, and cyclosporin A (Li et al., 1997; Dubinsky and Levi, 1998; Friberg et al., 1998; Beutner et al., 1998; Budd et al., 2000), were modestly neuroprotective (Fig. 3). At a 10-fold higher concentration, however, cyclosporin A was no longer protective, and, indeed, was itself toxic to the NT2-N neurons even in the absence of 3-NP (data not shown). The antioxidant, propyl gallate (10 WM), previously reported to diminish 3-NP-induced neuronal apoptosis (Keller et al., 1998), also diminished 3-NP-induced NT2-N neuronal LDH release by 44% ; 4% (mean ; S.E.M., n = 8, P 6 0.01) at 48 h. 3-NP diminishes NT2-N neuronal ATP content. This fall is inhibited by MK-801 3-NP blocks ATP synthesis by inhibition of succinate
NSC 5622 23-7-02
702
W.-T. Lee et al.
Fig. 2. Concentration-dependent e¡ects of 3-NP on MTT reduction by NT2-N neurons in the absence and presence of 10 WM MK-801. Data are means ; S.E.M. of at least three independent experiments.
dehydrogenese, an essential enzyme in the TCA cycle. ATP content of control NT2-N neurons was 12.2 ; 1.5 nM/mg protein (mean ; S.E.M., n = 18). Neuronal ATP content did not fall during the ¢rst 6 h of incubation with 0.1 mM 3-NP, but then declined to 48 ; 9% and 47 ; 2% (mean ; S.E.M., n = 6) of the basal value at 24 and 48 h, respectively (Fig. 4). Higher concentrations of 3-NP caused an earlier and more profound decline. In 1 mM 3-NP, for example, ATP content varied between 67 and 86% of the basal value during the ¢rst 6 h, and had fallen to 22 ; 3% and 25 ; 4% of the control value after 24 and 48 h, respectively. Thus, high concentrations of 3-NP, which led predominantly to neuronal necrosis, caused a much earlier fall in neuronal ATP content. Addition of 10 WM MK-801 concurrently with the 1 mM 3-NP substantially diminished ATP depletion. For example, at 48 h, ATP content remained at 80 ; 2% (n = 10) of the control value, more than threefold higher than in parallel cultures in the absence of MK-801 (Fig. 4). This result demonstrates that NMDA^GluR activation contributed substantially to the NT2-N neuronal ATP depletion caused by 3-NP, likely by accelerating ATP consumption during the early phase of exposure to high concentrations of 3-NP.
tures was not a consequence of increased release of glutamate from NT2-N neurons by 3-NP. 3-NP elicits a consistent early rise in NT2-N neuronal [Ca2+]i Loss of calcium homeostasis is considered to be a key mechanism in NMDA^GluR-mediated neuronal excitotoxicity (Kiedrowski, 1999). During the ¢rst 2 h after addition of 1 mM 3-NP in a Ca2þ -containing medium,
3-NP does not increase extracellular glutamate concentration in NT2-N cultures We previously observed that NT2 neurons secrete glutamate into the culture medium (Rootwelt et al., 1998). In this study, we totally removed the old medium and added 3-NP dissolved in fresh culture medium which did not contain exogenous glutamate. After a 48 h incubation with or without 3-NP, the glutamate concentration of the medium was 98 ; 10 WM in cultures not treated with 3-NP, 58 ; 16 WM in cultures in 1 mM 3-NP, and 74 ; 20 WM in cultures in 5 mM 3-NP. Thus, activation of NMDA^GluR by 3-NP in the NT2-N neuronal cul-
Fig. 3. Protective e¡ects of various treatments against 3-NPinduced necrosis, estimated by LDH release. The neurons were incubated with 1 mM 3-NP (or without, in the control) for two days, with one or more of the following agents : MK-801 (10 WM), CNQX (100 WM), nifedipine (nifed, 10 WM), xestospongin C (xesto, 1 WM), bongkrekic acid (bong, 50 WM), or cyclosporin A (CSA, 1 WM). Data are means ; S.E.M. of at least three independent experiments. * and **, signi¢cantly lower than the % in 3-NP, P 6 0.05 and P 6 0.01, respectively.
NSC 5622 23-7-02
3-NP excitotoxicity in human NT2-N neurons
703
Fig. 4. Time-dependent e¡ects of 3-NP on NT2-N neuronal ATP content. Data are means ; S.E.M. of at least four independent experiments. For the data point shown in the presence of MK-801, S.E.M. bars are contained within the symbol.
there was a gradual increase in [Ca2þ ]i in all NT2-N neurons, from a basal [Ca2þ ]i level of 48 ; 2 nM to 140 ; 12 nM (mean ; S.E.M.). This [Ca2þ ]i increase was prevented by removal of Ca2þ from the medium (Fig. 5A). Thus, in£ux of extracellular Ca2þ is required for 3-NP to induce a sustained increase in NT2-N neuronal [Ca2þ ]i . The rate of the rise in neuronal [Ca2þ ]i for the ¢rst 2 h was markedly slower in the presence of MK801. The L-type VGCC inhibitor, nifedipine, did not further augment the inhibitory e¡ect of MK-801 on this rise in [Ca2þ ]i . We concluded, therefore, that NMDA^GluR played a predominant role in the early rise in [Ca2þ ]i in 3-NP-treated NT2-N neurons. Roles of ER and mitochondria in bu¡ering [Ca2+]i during the ¢rst 30 min of treatment of NT2-N neurons with 3-NP When NT2-N neurons were exposed to 1mM 3-NP in a Ca2þ -free medium, there was still an early rise in [Ca2þ ]i (Fig. 5B). During the ¢rst few minutes, the rate of this rise was nearly identical to that in a Ca2þ -containing medium (compare Fig. 5A, B, noting the changes in x and y scales). This demonstrates that, during this brief initial time period, the increase in [Ca2þ ]i caused by 3-NP was the result of release of Ca2þ from internal stores (ER and/or mitochondria). With longer periods of observation, the increase in [Ca2þ ]i decelerated, con¢rming the requirement for extracellular Ca2þ to sustain a 3-NPinduced [Ca2þ ]i elevation. The rise in [Ca2þ ]i in Ca2þ free medium was totally blocked by addition of xestospongin C plus dantrolene. These drugs inhibit IP3-sensitive ER Ca2þ release and Ca2þ -activated ER Ca2þ release, respectively (Wei and Perry, 1996; Gafni et al., 1997; Yu et al., 1999; Mattson et al., 2000). We concluded, therefore, that release of Ca2þ from ER was largely responsible for the 3-NP-induced initial rise in [Ca2þ ]i . Release of Ca2þ from mitochondria may substantially
increase the rise in [Ca2þ ]i caused by activation of neuronal NMDA^GluR (Mody and MacDonald, 1995; Stout et al., 1998). To evaluate the role of mitochondria in bu¡ering [Ca2þ ]i during the early phase of exposure of NT2-N neurons to 3-NP, carbonyl cyanide p-(tri£uoromethoxy)phenylhydrazone (FCCP), a proton ionophore which blocks mitochondrial retention of Ca2þ (Luo et al., 1997; Khodorov et al., 1999) was added to the Ca2þ -free medium. This elicited a rapid, approximately 20 nM, increase in [Ca2þ ]i . The slope of the rise in [Ca2þ ]i that followed during the next 30 min in these 3-NP-treated neurons was very similar to that in the absence of FCCP (Fig. 5B). We interpret these results to indicate that, in a Ca2þ -free medium, the NT2-N neuronal mitochondria contained a small amount of Ca2þ , su⁄cient, when released, to raise [Ca2þ ]i by 20 nM, and that, during the remainder of the observation period, mitochondrial Ca2þ uptake did not substantially contribute to [Ca2þ ]i homeostasis. Chronic e¡ects of 3-NP on NT2-N neuronal Ca2+ homeostasis We measured [Ca2þ ]i following 24 or 48 h of incubation with 1 mM 3-NP in a Ca2þ -containing medium (chronic studies in a Ca2þ -free medium were not performed, because chronic incubation in Ca2þ -free medium was toxic to these neurons even in the absence of 3-NP). There were two populations of neurons in these 3-NPtreated cultures: the majority (83% at one day, 75% at two days) had restored [Ca2þ ]i to below 100 nM, whereas in the remaining neurons (17% at 24 h, 25% at 48 h), [Ca2þ ]i remained above 100 nM (Table 2). The distributions of [Ca2þ ]i values in these neurons are shown in Fig. 6. Inspection of this ¢gure indicates that the increase in neurons with [Ca2þ ]i above 100 nM between 24 and 48 h of treatment with 3-NP without other additives (column 2 in the ¢gure) was entirely accounted for by
NSC 5622 23-7-02
704
W.-T. Lee et al.
Fig. 5. Acute e¡ects of 1 mM 3-NP on NT2-N neuronal [Ca2þ ]i . After recording basal [Ca2þ ]i for one minute, the perfusion solution was changed to a solution containing the indicated drugs. (A) In a Ca2þ -containing medium, 3-NP induced a linear increase in [Ca2þ ]i , which was attenuated by MK-801. Nifedipine did not further increase the attenuation elicited with MK-801. Changing to a Ca2þ -free medium at 120 min (indicated by the bar) caused the rise in [Ca2þ ]i to cease. (B) In a Ca2þ -free medium, the initial rise in [Ca2þ ]i in the 3-NP-treated neurons was similar to that in a Ca2þ -containing medium, but then slowed. The rise in [Ca2þ ]i was abolished by treatment with xestospongin C and dantrolene. Treatment of the neurons with 1 WM FCCP and 1 mM 3-NP caused an initial, very brief, rise in [Ca2þ ]i that was more rapid than in the absence of FCCP. After the ¢rst 5 min, the slope of the rise was similar to that in the absence of FCCP. Numbers of neurons examined (in two independent experiments) are indicated in parentheses, and are expressed as the means ; S.E.M. Abbreviations and concentrations used: MK, MK-801 10 WM; Nif, nifedipine 10 WM; Xesto, xestospongin C 1 WM; Dant, dantrolene 10 WM; FCCP 1 WM.
neurons in which [Ca2þ ]i was barely above 100 nM. The proportion of neurons in which [Ca2þ ]i was above 100 nM at 24 or 48 h was markedly reduced by treatment with MK-801, and was also signi¢cantly reduced by
treatment with xestospongin C, but not by treatment with CNQX or nifedipine (Table 2). Combination treatments with MK-801 and xestospongin C or MK-801 plus CNQX were more e¡ective in diminishing the incidence
NSC 5622 23-7-02
3-NP excitotoxicity in human NT2-N neurons
705
Table 2. Incidence of NT2-N neurons with [Ca2þ ]i s 100 nM after one or two days in 1 mM 3-NP
Control 3-NP 3-NP+MK-801 3-NP+nifedipine 3-NP+CNQX 3-NP+xestospongin C 3-NP+MK-801+nifedipine 3-NP+MK-801+CNQX 3-NP+MK-801+xestospongin C
One day
Two days
ND 17% (48/280) 6% (17/300)** 20% (13/66) 21% (15/72) 9% (10/106)* 5% (6/126)** 3% (3/115)** 0% (0/105)**
2% (5/244) 25% (44/179) 6% (9/158)** 19% (29/156) 18% (32/171) 14% (26/182) 7% (5/74)** 1% (1/104)** 1% (2/135)**
ND = not done; * and **, signi¢cantly lower than the % in 3-NP, P 6 0.05 and P 6 0.01, respectively. All drugs were added at time 0, and [Ca2þ ]i was measured at 24 and 48 h, using fura-2/AM. In parentheses, we indicate the number of neurons assayed in which [Ca2þ ]i was greater than 100 nM, and the total number of neurons assayed for each condition at each time-point. Concentrations used: 3-NP 1 mM ; MK-801 10 WM; nifedipine 10 WM; CNQX 100 WM; xestospongin C 1 WM.
of neurons with [Ca2þ ]i greater than 100 nM than treatment with MK-801 alone.
DISCUSSION
Our aim in this study has been to characterize the changes in Ca2þ homeostasis that accompany neuronal energy deprivation and precede neuronal death. We chose conditions (1 mM 3-NP) and a time-frame (¢rst 48 h of 3-NP exposure) in which most NT2-N neurons died by necrosis rather than apoptosis. [Ca2þ ]i increased in a linear fashion in the neurons during the ¢rst few hours after addition of 3-NP, to almost three-fold higher than the resting level by 2 h. Several prior reports of the acute e¡ects of 3-NP, as well as the present study, have documented that activation of NMDA^GluR is one of responsible mechanisms for this early rise in neuronal [Ca2þ ]i (Greene et al., 1998; Olsen et al., 1999). Addition of MK-801 diminished this early rise by 70%, with no additional inhibitory e¡ect by nifedipine, thus indicating the relative importance of activation of NMDA^GluR, and the relative unimportance of activation of L-type VGCC in mediating this increase in [Ca2þ ]i . Our shortterm observations in a Ca2þ -free medium demonstrated that release of internal stores of Ca2þ contributed substantially to the ¢rst few minutes of 3-NP-induced rise in [Ca2þ ]i , but were quantitatively less signi¢cant at later time-points. Because the early rise in [Ca2þ ]i was observed in all of the 3-NP-treated NT2-N neurons, but most of these neurons survived prolonged 3-NP exposure, it is clear that an early [Ca2þ ]i rise of the magnitude we observed was not su⁄cient to elicit necrosis. While little has previously been known about the chronic e¡ects of 3-NP on neuronal [Ca2þ ]i homeostasis, prior studies have emphasized the signi¢cance of prolonged [Ca2þ ]i dysregulation and consequent mitochondrial calcium overload in the pathophysiology of neuronal excitotoxicity (Stout et al., 1998; Nicholls et al., 1999; Reynolds, 1999). We observed that, at 24 and 48 h after addition of 1 mM 3-NP to the medium, [Ca2þ ]i remained above 100 nM in 17% and 25% of the neurons, respectively. The incidence of neurons with [Ca2þ ]i above 100 nM at 24 and 48 h was signi¢cantly decreased by MK-801. MK-801 was also by far the most
e¡ective of the drugs we tested in protecting the NT2-N neurons against necrosis. CNQX or nifedipine alone did not signi¢cantly diminish the proportion of neurons maintaining their [Ca2þ ]i above 100 nM at 24 and 48 h, but the combinations of MK-801 and CNQX or MK-801 and nifedipine were signi¢cantly more e¡ective in this regard than MK-801 alone, and CNQX or nifedipine alone did marginally diminish the incidence of neuronal necrosis. These data indicate that non-NMDA^GluR and VGCC do play roles in NT2-N neuronal 3-NP toxicity, though these are quantitatively minor. Since [Ca2þ ]i was 140 ; 12 nM (mean ; S.E.M.) at 2 h, and had fallen to below 100 nM in all but 17% of the neurons by 24 h, most of the NT2-N neurons must have lowered [Ca2þ ]i toward normal in the interval between 2 and 24 h, despite marked ATP depletion. Bcl-2 was a candidate responsible for the capacity of these neurons to restore normal or near-normal [Ca2þ ]i , because Bcl-2 expression in NT2-N neurons is signi¢cantly increased after terminal di¡erentiation (Singer et al., 1998; Guillemain et al., 2000). However, no further induction of Bcl-2 protein was observed after the exposure to 3-NP (our preliminary observation), although this might explain why NT2-N neurons are more resistant to energy deprivation compared to primary neuronal cultures. For the following reasons, it is likely those neurons with chronically elevated [Ca2þ ]i had not yet undergone necrosis but were mostly destined not to survive. First, at 24 h only 1% of the neurons in the cultures treated with 1 mM 3-NP showed morphological features of necrosis. Second, successful fura-2 imaging requires a cell to maintain its plasma membrane intact, in order to retain loaded fura-2; hence, neurons already undergoing necrosis are probably not represented in our [Ca2þ ]i data. Third, the incidence of neurons with [Ca2þ ]i above 100 nM at 24 and 48 h correlated well with actual neuronal death evaluated by LDH assay. A concentration of 3-NP (0.1 mM) su⁄cient to induce a gradually increasing incidence of NT2-N neuronal apoptosis, but not signi¢cant necrosis, caused a delayed fall in neuronal ATP content, whereas concentrations of 3-NP su⁄cient to induce substantial NT2-N neuronal necrosis (1 or 5 mM) caused an early and more profound fall in neuronal ATP content. With 1 or 5 mM 3-NP, ATP content reached a near-steady-state level less than a
NSC 5622 23-7-02
706
W.-T. Lee et al.
Fig. 6. Scattergram of [Ca2þ ]i values at 24 and 48 h in 3-NP-treated NT2-N neurons. The concentration of 3-NP in the medium was 1 mM. Each dot represents a single neuron. Data pooled from three independent experiments are shown. Abbreviations and concentrations used: MK, MK-801 10 WM; Nif, nifedipine 10 WM; Xes, xestospongin C 1 WM; CNQX 100 WM.
third that of control neurons within 24 h. The reduced ATP level was presumably maintained by continued ATP generation by glycolysis. In parallel experiments in which NMDA^GluR activation was blocked with MK-801, the neurons were able to maintain ATP at 80% of the control level, indicating that, with the strong inhibition of aerobic ATP production, ATP consumption induced by NMDA^GluR is a critical factor to determine the severity of ATP depletion in NT2-N neurons. One factor likely to have contributed to the ATP-sparing e¡ect of inhibiting NMDA^GluR was a diminution in ATP consumption by plasma membrane Naþ ,Kþ -ATPase in response to NMDA^GluR-mediated neuronal Naþ loading (Chinopoulos et al., 2000). Collapse of mitochondrial membrane potential induced by activation of NMDA^GluR is also known to induce reverse operation of the ATP synthase and reduce ATP/ADP ratio
(Budd and Nicholls, 1996). Our ATP data also suggest that early and severe ATP depletion is likely to be an essential step leading to chronic Ca2þ dysregulation. The early increase in [Ca2þ ]i elicited by 3-NP in a Ca2þ -free medium in the NT2-N neurons was entirely prevented by addition of xestospongin C and dantrolene, indicating that Ca2þ mobilization from intracellular Ca2þ storage into cytosol occurred over the initial phase of 3-NP administration. Longer term, xestospongin C diminished the proportion of NT2-N neurons in which [Ca2þ ]i was chronically elevated, and, in conjunction with MK-801, almost totally prevented chronic elevation in [Ca2þ ]i . Finally, xestospongin C signi¢cantly diminished the proportion of NT2-N neurons undergoing necrosis at 48 h. These results raise the idea that initial Ca2þ mobilization from intracellular Ca2þ storage as an acute e¡ect of 3-NP enhances NMDA^GluR-medi-
NSC 5622 23-7-02
3-NP excitotoxicity in human NT2-N neurons
ated neurotoxicity. It has been reported recently that, as a model of Huntington’s disease, a 20-min treatment with 3-NP produced a long-term potentiation of the NMDA^GluR-mediated synaptic excitation in striatal spiny neurons by the mechanism requiring D2 dopamine receptor activation (Calabresi et al., 2001). Although the exact mechanism by which 3-NP induces Ca2þ mobilization from intracellular Ca2þ storage in NT2-N neurons remains to be clari¢ed, Ca2þ release particularly from IP3-sensitive Ca2þ storage might also alter the physiological characteristics of NMDA^GluR. In addition, our data strongly suggest that drugs that block ER Ca2þ release deserve consideration as therapeutic agents in stroke and other disorders of brain energy metabolism (Mody and MacDonald, 1995; Wei and Perry, 1996; Yu et al., 1999). We observed that two MPT inhibitors and an antioxidant diminished 3-NP-induced NT2-N neuronal necrosis. The MPT results con¢rm a prior study in which one of these inhibitors, cyclosporin A, diminished the extent of neuronal necrosis elicited by systemic administration of 3-NP (Leventhal et al., 2000). Antioxidants and free radical spin traps have also been reported to minimize 3-NP toxicity in vivo (Schulz et al., 1996; Fontaine et al., 2000), and glutathione peroxidase knockout mice are hypersusceptible to 3-NP-induced striatal neuronal loss (Klivenyi et al., 2000). In our own study, NT2-N neuronal necrosis caused by 3-NP was diminished by addition of an antioxidant, propyl gallate, to the medium.
707
Together, the protective e¡ects of MPT inhibitors and antioxidants suggest that reactive oxygen species produced in excess by mitochondria play a role in 3-NPmediated NT2-N neuronal necrosis, as they do in ischemic necrosis in these neurons (Rootwelt et al., 1998; Kim et al., 2000; Almaas et al., 2000).
CONCLUSION
In conclusion, our study is the ¢rst to employ human neurons to study the pathophysiology of 3-NP neurotoxicity, and yielded three novel ¢ndings. First, we demonstrated the preponderant role that activation of NMDA^ GluR plays in mediating 3-NP-induced depletion of neuronal ATP. Second, by assaying neuronal [Ca2þ ]i at 24 and 48 h after application of 3-NP, we demonstrated chronic [Ca2þ ]i dysregulation similar to that previously reported in neuronal excitotoxicity (Itoh et al., 1998; Nicholls et al., 1999). Third, we demonstrated that inhibiting ER Ca2þ release substantially enhanced the capacity of the 3-NP-treated neurons to maintain [Ca2þ ]i homeostasis and resist necrosis.
Acknowledgements/Supported by the Muscular Dystrophy Association, the Wolfson Fund, and by NIH Grants NS25044, NS08075, and HD26979. We appreciate the assistance of Marc Yudko¡ in the glutamate assays.
REFERENCES
Almaas, R., Saugstad, O.D., Pleasure, D., Rootwelt, T., 2000. E¡ect of anesthetics on hydroxyl radicals, lipid peroxidation and hypoxic neuronal cell death in human NT2-N neurons. Anesthesiology 92, 764^774. Beal, M.F., Brouillet, E., Jenkins, B.G., Ferrante, R.J., Kowall, N.W., Miller, J.M., Storey, E., Srivastava, R., Rosen, B.R., Hyman, B.T., 1993. Neurochemical and histologic characterization of striatal excitotoxic lesions produced by the mitochondrial toxin 3-nitropropionic acid. J. Neurosci. 13, 4181^4192. Beutner, G., Ruck, A., Riede, B., Brdiczka, D., 1998. Complexes between porin, hexokinase, mitochondrial creatine kinase and adenylate translocator display properties of the permeability transition pore. Implication for regulation of permeability transition by the kinases. Biochim. Biophys. Acta 1368, 7^18. Budd, S.L., Nicholls, D.G., 1996. A reevaluation of the role of mitochondria in neuronal Ca2þ homeostasis. J. Neurochem. 66, 403^411. Budd, S.L., Tenneti, L., Lishnak, T., Lipton, S.A., 2000. Mitochondrial and extramitochondrial apoptotic signaling pathways in cerebrocortical neurons. Proc. Natl. Acad. Sci. USA 97, 6161^6166. Calabresi, P., Gubellini, P., Picconi, B., Centonze, D., Pisani, A., Bonsi, P., Greengard, P., Hipskind, R.A., Borrelli, E., Bernardi, G., 2001. Inhibition of mitochondrial complex II induces a long-term potentiation of NMDA-mediated synaptic excitation in the striatum requiring endogenous dopamine. J. Neurosci. 21, 5110^5120. Chinopoulos, C., Tretter, L., Rozsa, A., Adam-Vizi, V., 2000. Exacerbated response to oxidative stress by an Naþ load in isolated nerve terminals : the role of ATP depletion and rise of [Ca2þ ]i . J. Neurosci. 20, 2094^2103. Dubinsky, J.M., Levi, Y., 1998. Calcium-induced activation of the mitochondrial permeability transition in hippocampal neurons. J. Neurosci. Res. 53, 728^741. Fink, S.L., Ho, D.Y., Sapolsky, R.M., 1996. Energy and glutamate dependency of 3-nitropropionic acid neurotoxicity in culture. Exp. Neurol. 138, 298^304. Fontaine, M.A., Geddes, J.W., Banks, A., Butter¢eld, D.A., 2000. E¡ect of exogenous and endogenous antioxidants on 3-nitropropionic acidinduced in vivo oxidative stress and striatal lesions. Insights into Huntington’s disease. J. Neurochem. 75, 1709^1715. Friberg, H., Ferrand-Drake, M., Bengtsson, F., Halestrap, A.P., Wieloch, T., 1998. Cyclosporin A, but not FK 506, protects mitochondria and neurons against hypoglycemia damage and implicates the mitochondrial permeability transition in cell death. J. Neurosci. 18, 5151^5159. Fukuda, A., Deshpande, S.B., Shimano, Y., Nishino, H., 1998. Astrocytes are more vulnerable than neurons to cellular Ca2þ overload induced by a mitochondrial toxin, 3-nitropropionic acid. Neuroscience 87, 497^507. Gafni, J., Munsch, J.A., Lam, T.H., Catlin, M.C., Costa, L.G., Molinski, T.F., Pessah, I.N., 1997. Xestospongins : potent membrane permeable blockers of the inositol 1,4,5-trisphosphate receptor. Neuron 19, 723^733. Gavrieli, Y., Sherman, Y., Ben-Sasson, S., 1992. Identi¢cation of programmed cell death in situ via speci¢c labeling of nuclear DNA fragmentation. J. Cell Biol. 119, 493^501. Greene, J.G., Sheu, S.-S., Gross, R.A., Greenamyre, J.T., 1998. 3-Nitropropionic acid exacerbates N-methyl-D-aspartate toxicity in striatal culture by multiple mechanisms. Neuroscience 84, 503^510. Grynkiewicz, G., Poenie, M., Tsien, R.Y., 1985. A new generation of Ca indicators with greatly improved £uorescence properties. J. Biol. Chem. 260, 3440^3450.
NSC 5622 23-7-02
708
W.-T. Lee et al.
Guillemain, I., Gaboyard, S., Fontes, G., Saunier, M., Privat, A., Patey, G., 2000. Di¡erential expression of Bcl-2-related proteins in di¡erentiating NT2 cells. NeuroReport 11, 1421^1425. Hardy, M., Younkin, D., Tang, C.-M., Pleasure, J., Williams, M., Pleasure, D., 1994. Expression of non-NMDA glutamate receptor channels by clonal NT2-N human neurons. J. Neurochem. 63, 482^489. Itoh, T., Itoh, A., Horiuchi, K., Pleasure, D., 1998. AMPA receptor-mediated excitotoxicity in human NT2-N neurons results from loss of intracellular Ca2þ homeostasis following marked elevation of intracellular Naþ . J. Neurochem. 71, 112^124. Jones, B.N., Gilligan, J.P., 1983. o-Phthaldialdehyde precolumn derivatization and reversed-phase high-performance liquid chromatography of polypeptide hydrolysates and physiological £uids. J. Chromatogr. 266, 471^482. Keller, J.N., Guo, Q., Holtsberg, F.W., Bruce-Keller, A.J., Mattson, M.P., 1998. Increased sensitivity to mitochondrial toxin-induced apoptosis in neural cells expressing mutant presenilin-1 is linked to perturbed calcium homeostasis and enhanced oxyradical production. J. Neurosci. 18, 4439^4450. Khodorov, B., Pinelis, V., Storozhevykh, T., Yuravichus, A., Khaspekhov, L., 1999. Blockade of mitochondrial Ca2þ uptake by mitochondrial inhibitors ampli¢es the glutamate-induced calcium response in cultured cerebellar granule cells. FEBS Lett. 458, 162^166. Kiedrowski, L., 1999. N-Methyl-D-aspartate excitotoxicity: relationships among plasma membrane potential, Naþ /Ca2þ exchange, mitochondrial Ca2þ overload, and cytoplasmic concentrations of Ca2þ , Hþ , and Kþ . Mol. Pharmacol. 56, 619^632. Kim, G.W., Copin, J.C., Kawase, M., Chen, S.F., Sato, S., Gobbel, G.T., Chan, P.H., 2000. Excitotoxicity is required for induction of oxidative stress and apoptosis in mouse striatum by the mitochondrial toxin, 3-nitropropionic acid. J. Cereb. Blood Flow Metab. 20, 119^129. Klivenyi, P., Andreassen, O.A., Ferrante, R.J., Dedeoglu, A., Mueller, G., Lancelot, E., Bogdanov, M., Andersen, J.K., Jiang, D., Beal, M.F., 2000. Mice de¢cient in cellular glutathione peroxidase show increased vulnerability to malonate, 3-nitropropionic acid, and 1-methyl-4-phenyl1,2,5,6-tetrahydropyridine. J. Neurosci. 20, 1^7. Lee, W.T., Shen, Y.Z., Chang, C., 2000. Neuroprotective e¡ect of lamotrigine and MK-801 on rat brain lesions Induced by 3-nitropropionic acid : evaluation by magnetic resonance imaging and in vivo proton magnetic resonance spectroscopy. Neuroscience 95, 89^95. Leventhal, L., Sortwell, C.E., Hanbury, R., Collier, T.J., Kordower, J.H., Pal¢, S., 2000. Cyclosporin A protects striatal neurons in vitro and in vivo from 3-nitropropionic acid toxicity. J. Comp. Neurol. 425, 471^478. Li, P.A., Uchino, H., Elmer, E., Siesjo, B.K., 1997. Amelioration by cyclosporin A of brain damage following 5 or 10 min of ischemia in rats subjected to preischemic hyperglycemia. Brain Res. 753, 133^140. Luo, Y., Bond, J.D., Ingram, V.M., 1997. Compromised mitochondrial function leads to increased cytosolic calcium and to activation of MAP kinases. Proc. Natl. Acad. Sci. USA 94, 9705^9710. Mattson, M.P., Zhu, H., Yu, J., Kindy, M.S., 2000. Presenilin-1 mutation increases neuronal vulnerability to focal ischemia in vivo and to hypoxia and glucose deprivation in cell culture : involvement of perturbed calcium homeostasis. J. Neurosci. 20, 1358^1364. Mody, I., MacDonald, J.F., 1995. NMDA receptor-dependent excitotoxicity: the role of intracellular Ca2þ release. Trends Pharmacol. Sci. 16, 356^359. Nicholls, D.G., Budd, S.L., Castilho, R.F., Ward, M.W., 1999. Glutamate excitotoxicity and neuronal energy metabolism. Ann. NY Acad. Sci. 893, 1^12. Olsen, C., Rustad, A., Fonnum, F., Paulsen, R.E., Hassel, B., 1999. 3-Nitropropionic acid: an astrocyte-sparing neurotoxin in vitro. Brain Res. 850, 144^149. Pal¢, S., Ferrante, R.J., Brouillet, E., Beal, M.F., Dolan, R., Guyot, M.C., Peschanski, M., Hantraye, P., 1996. Chronic 3-nitropropionic acid treatment in baboons replicates the cognitive and motor de¢cits of Huntington’s disease. J. Neurosci. 16, 3019^3025. Pang, Z., Geddes, J.W., 1997. Mechanisms of cell death induced by mitochondrial toxin 3-nitropropionic acid : acute excitotoxic necrosis and delayed apoptosis. J. Neurosci. 17, 3064^3073. Pleasure, S.J., Page, C., Lee, V.M.-Y., 1992. Pure, postmitotic, polarized human neurons derived from Ntera 2 cells provide a system for expressing exogenous proteins in terminally di¡erentiated neurons. J. Neurosci. 12, 1802^1815. Reynolds, I.J., 1999. Mitochondrial membrane potential and the permeability transition in excitotoxicity. Ann. NY Acad. Sci. 893, 33^41. Rootwelt, T., Dunn, M., Yudko¡, M., Itoh, T., Almaas, R., Pleasure, D., 1998. Hypoxic cell death in human NT2-N neurons : involvement of NMDA- and non-NMDA-glutamate receptors. J. Neurochem. 71, 1544^1553. Schulz, J.B., Henshaw, D.R., MacGarvey, U., Beal, M.F., 1996. Involvement of oxidative stress in 3-nitropropionic acid neurotoxicity. Neurochem. Int. 29, 167^171. Singer, C.A., Rogers, K.L., Dorsa, D.M., 1998. Modulation of Bcl-2 expression : a potential component of estrogen protection in NT2 neurons. NeurouReport 9, 2565^2568. Stout, A.K., Raphael, H.M., Kanterewicz, B.I., Klann, E., Reynolds, I.J., 1998. Glutamate-induced neuron death requires mitochondrial calcium uptake. Nature Neurosci. 1, 366^373. Wei, H., Perry, D.C., 1996. Dantrolene is cytoprotective in two models of neuronal cell death. J. Neurochem. 67, 2390^2398. Wullner, U., Young, A.B., Penney, J.B., Beal, M.F., 1994. 3-Nitropropionic acid toxicity in the striatum. J. Neurochem. 63, 1772^1781. Younkin, D., Tang, C.-M., Hardy, M., Reddy, U.R., Pleasure, S., Lee, V., Pleasure, D., 1993. Induction of NMDA receptor expression and glutamate excitotoxicity in the clonal human cell line NTera 2. Proc. Natl. Acad. Sci. USA 90, 2174^2178. Yu, Z., Luo, H., Fu, W., Mattson, M.P., 1999. The endoplasmic reticulum stress-responsive protein GRP78 protects neurons against excitotoxicity and apoptosis: suppression of oxidative stress and stabilization of calcium homeostasis. Exp. Neurol. 155, 302^314. (Accepted 4 February 2002)
NSC 5622 23-7-02