Brain Research 911 (2001) 37–42 www.elsevier.com / locate / bres
Research report
Calbindin overexpression buffers hippocampal cultures from the energetic impairments caused by glutamate Michelle L. Monje, Russell Phillips, Robert Sapolsky* Department of Biological Sciences, Stanford University MC5020, Stanford, CA 94305 -5020, USA Accepted 10 April 2001
Abstract A dramatic rise in free cytosolic calcium concentration is thought to be a central event in the pathogenesis of glutamate excitotoxicity in neurons. We have previously demonstrated that gene transfer of the calcium-binding protein calbindin D28k via a Herpes simplex amplicon vector decreases the rise in intracellular calcium and promotes cell survival following glutamatergic challenge. This study explores the effect of calbindin transgene expression on cellular metabolism following glutamate excitotoxicity. Because excitotoxic insults are often energetic in nature, and because calcium sequestering and extrusion place heavy energy demands on a cell, we hypothesized that calbindin overexpression may help preserve cellular energy levels during an insult. We overexpressed calbindin in primary hippocampal cultures, using a Herpes simplex amplicon vector system. We found that calbindin overexpression protected neurons from the decline in ATP levels, mitochondrial potential and metabolic rate following a glutamatergic insult. These results indicate that calbindin expression helps preserve cellular energy state following glutamate excitotoxicity. This illustrates the energetic load placed on neurons by increased free cytosolic calcium and may help explain the neuroprotective effects of calbindin. 2001 Elsevier Science B.V. All rights reserved. Theme: Disorders of the nervous system Topic: Neurotoxicity Keywords: Excitotoxicity; Hippocampus; Neuron death; Calbindin; Energetics; Gene therapy
1. Introduction A dramatic rise in free cytosolic calcium is thought to be a central event in the pathogenesis of excitotoxin-mediated neurodegeneration [2,21,24]. Effects of cytosolic calcium accumulation include activation of proteases, phospholipases and endonucleases, generation of reactive oxygen species, inhibition of protein synthesis and derrangement of the cytoskeleton. A thorough understanding of the role calcium plays in neuron death may help guide the development of effective new therapies for neurologic diseases such as ischemic stroke, epilepsy and cranial trauma. An important aspect of excitotoxic calcium influx is the heavy energetic demand which cytosolic calcium accumu-
*Corresponding author. Tel.: 11-650-723-2649; fax: 11-650-7255356. E-mail address:
[email protected] (R. Sapolsky).
lation places upon the injured neuron. Cellular calcium regulation is energetically expensive, in that it relies upon sequestration and extrusion. Calcium is extruded from the cell and sequestered in the endoplasmic reticulum by active transport using the calcium ATPase, and is extruded from the cell by secondary active transport using the sodium–potassium ATPase and the sodium–calcium exchanger. The mitochondrion also plays a critical role in regulating cytosolic calcium concentration. Mitochondria sequester calcium via an inner mitochondrial membrane uniporter when the rise in cytosolic calcium exceeds the capacity of active transport mechanisms [20]. Excessive intramitochondrial calcium sequestration compromises mitochondrial function, thereby impairing the neuron’s ATP-generating capability [5,9]. Thus, the rise in free cytosolic calcium which occurs in the setting of an excitotoxic insult is theoretically costly both in terms of ATP expenditure and decreased ATP production. In this study, we sought to investigate the importance of the
0006-8993 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 01 )02568-9
38
M.L. Monje et al. / Brain Research 911 (2001) 37 – 42
energetic load of cytosolic calcium accumulation following glutamate excitotoxicty. We have previously demonstrated that gene transfer of the calcium-binding protein calbindin D28k via a Herpes simplex amplicon vector decreases the rise in intracellular calcium and promotes cell survival following hypoglycemic and glutamatergic challenge [17,18]. In this study, we have investigated the impact of overexpressing this calcium binding protein on cellular metabolic parameters following glutamatergic challenge. Specifically, we examined the effects of calbindin D28k overexpression on mitochondrial membrane potential, cellular ATP levels and overall metabolic activity as measured by silicon microphysiometry in primary rat mixed hippocampal cultures.
2. Methods
2.1. Generation of plasmids (amplicons) and of vectors A rat calbindin cDNA clone (pBC40), first isolated by Hunziker and Schrinkel [12] was obtained from S. Christakos (University of Medicine and Dentistry, Newark, NJ, USA). The calbindin cDNA was removed from pBC40 and was cloned into the bicistronic expression plasmid pa22bgal [15], yielding construct pa4CBa22bgal [17]. A negative control plasmid, pa4cbsa22bgal was generated by insertion of stop codons in all three reading frames into the deletion resulting from removal of codons 6–43 near the 59 end of the CaBP cDNA [11,17]. The plasmid constructs pa4CBa22bgal and pa4cbsa22bgal were transfected into E5 cells using Lipofectamine cationic liposome. At 24 h after transfection, cells were superinfected with HSV mutant d120 [4], at an MOI (multiplicity of infection) of 0.03. After complete or nearly complete cytopathic effect was reached, cells were frozen at 2808C, thawed quickly, and then sonicated. Viral stocks va4CBa22bgal (CB vector), va4cbsa22bgal (stop codon control vector) and uninfected cell lysate were partially purified by centrifugation through a sucrose gradient [11]. About 0.5–3.0310 7 vectors and 1.0 –10310 7 helper virus particles were produced per milliliter by these methods.
2.2. Cell culture Mixed hippocampal cultures were grown from Day 18 fetal rats by standard, published methods [10]. Briefly, the procedure was as follows: after dissection the tissue was treated with papain (Worthington Biochemical, Freehold, NJ, USA) according to the procedure recommended by the company. The cells were dissociated, filtered through an 80um cell strainer and resuspended in a modified MEM media (UCSF Tissue Culture Facility, San Francisco, CA, USA) and supplemented with 10% horse serum (Hyclone, Logan, UT, USA).
For the mitochondrial membrane potential experiments, cells were plated at a density of 20 000 / cm 2 on no. 0 coverslips from Carolina Biological (North Carolina, USA), cleaned and coated with poly-D-lysine. For the ATP study cells were plated at a similar density on poly-Dlysine coated 48-well plates. For the microphysiometry studies, cells were plated at a density of 50 000 cells / cm 2 on transwell inserts (Falcon, New Jersey, USA). In all cases the cells were used on days 11–13. These cultures are typically 20–30% neurons, the rest of the cells being glia of various types.
2.3. Rhodamine fluoroscopy Mixed primary rat hippocampal culture cells (day 12–14) were transfected 12 h prior to the beginning of experiment with 3 either calbindin vector (va4CBa22bgal), control vector (va4cbsa22bgal) or no vector (‘mock infection’ with sterile media). A 3.8-ml volume was used / well; the titers of the stock HSV vector and replication-deficient helper virus averaged 2.5310 7 infectious particles / ml and 2.9310 7 plaque-forming units / ml, respectively. Experiments were conducted at room temperature in Hank’s balanced salt solution–Hepes buffer (HBSSh, Gibco). Immediately prior to the beginning of the experiment, cells were incubated for 10 min with trimethyl-ester rhodamine (non-quenching; Molecular Probes, Eugene, OR, USA) in HBSSh buffer. Coverslips were then placed on a custom-designed microscope stage in 500 ml of HBSSh. Readings were taken every 30 s for 40 min. Stable baseline fluorescence was measured for 5 min, and then 7.5 ml of 5 mM glutamate in HBSSh buffer solution was added (exposing cells to 75 mM glutamate.) Data were collected and analyzed using METAMORPH software (Universal Imaging).
2.4. ATP analysis Mixed primary rat hippocampal culture cells (day 12– 14) were transfected 12 h prior to the beginning of experiment with 3.8 ml of either calbindin vector, control vector, or were mock infected. The titers of the stock HSV vector and replication-deficient helper virus were, on average, 2.5310 7 infectious particles / ml and 2.9310 7 plaque-forming units / ml, respectively. Experiments were conducted at 378C in a humidified 5% CO 2 incubator, and in Earl’s balanced salt solution with bicarbonate, pH 7.4, 5 mM glucose. Cells were exposed to glutamate (50 or 75 mM, 30 min) allowed to recover for 12 h, then frozen and stored at 2808C. Pilot data examining ATP levels 4, 8 and 12 h post-glutamate exposure indicated that 12 h postinsult was the earliest time at which we could detect a decline in ATP levels (data not shown). ATP was extracted and assayed using the luciferin–luciferase technique with a kit from Sigma (St. Louis, MO, USA). ATP levels were quantified using a standard curve. Protein concentration was determined using the Pierce assay.
M.L. Monje et al. / Brain Research 911 (2001) 37 – 42
39
2.5. Microphysiometry Silicon microphysiometry is a technique which allows sensitive, real time measurement of metabolic rate in cultured cells as a function of extracellular proton extrusion [23]. Microphysiometry was employed to measure overall metabolic rate of mixed primary rat hippocampal cells (day 12–14) before, during and after glutamate exposure (300 mM, 30 min). Cells were transfected 12 h prior to the beginning of experiment with 10 ml of either the calbindin vector, or control vector. The titers of the stock HSV vector and replication-deficient helper virus were, on average, 2.5310 7 infectious particles / ml and 2.9310 7 plaque-forming units / ml, respectively. Experiments were conducted at room temperature and ambient conditions in Earl’s balanced salt solution (EBSS) buffer lacking bicarbonate, pH 7.4, 5 mM glucose (Gibco). Metabolic rate measurements were taken every minute for the 2 h preinsult, 30 min glutamate exposure, and 4 h recovery periods.
2.6. Data analysis Data are presented as means6S.E.M. Results are analyzed by t-test, one- or two-way ANOVA, as indicated, using SIGMASTAT (SPSS).
3. Results
3.1. ATP levels ATP levels were analyzed and compared between cell culture wells treated with no vector (mock), control vector (b-gal) and calbindin vector (CB.) There was no significant difference between mock-infected cells and control vector-treated cells in either ATP levels either basally or 12 h following 75 mM glutamate (N.S. by t-test in both cases; n510–11 wells / group). Calbindin overexpression also had no effect on basal ATP levels (N.S. by t-test, as compared with control vector; n526 wells / group). In mock-infected and control vector-treated cultures, glutamate caused a significant, dose-dependent decline in ATP levels (P,0.02 by one-way ANOVA). In contrast, a significant decline in ATP levels was not observed in calbindin vector-treated cultures following glutamate exposure (N.S. by one-way ANOVA). Two way ANOVA yielded a significant difference in ATP levels as a function of vector treatment (P,0.001) (Fig. 1).
Fig. 1. Effects of overexpression of Calbindin on ATP levels in primary hippocampal cultures challenged with glutamate. Data are expressed as a fraction of ATP levels in control cultures (control vector with 0 glutamate). Glutamate caused a decrease in ATP levels in control-treated vectors (one-way ANOVA, P,0.019, F54.237, df52), whereas no such decline was seen in calbindin-overexpressing wells (one-way ANOVA, P,0.564, F50.580, df52.) * and *** indicate P50.05 and 0.001, in comparing the two treatments at the same glutamate dose by a posthoc test following two-way ANOVA (n59–26 / group). Absolute levels of ATP averaged 1.5 mmol of ATP per mg of protein.
strated a significant decline in fluorescence following glutamate exposure relative to cells not exposed to glutamate (Fig. 2; P,0.0001 by t-test). Cultures overexpressing calbindin demonstrated significantly less mitochondrial membrane depolarization relative to control vector-treated cells following 75 mM glutamate exposure.
3.2. Mitochondrial potential The change in rhodamine fluorescence over the course of the 40-min experiment did not differ in cultures treated with control vector compared with cultures treated with calbindin vector in the absence of insult (data not shown; N.S. by t-test). Cells treated with control vector demon-
Fig. 2. Effects of overexpression of calbindin on mitochondrial potential in cultures challenged with 75 mM glutamate. *** indicates P,0.001 comparing control with calbindin-treated cultures by t-test; n540 cells / group. Data are expressed as fraction of baseline fluorescence, which was defined as the average during the 5 min prior to glutamate exposure. The decline in fluorescence was calculated as the average of the last 5 min of the post-glutamate period. Fluorescence units are arbitrary.
40
M.L. Monje et al. / Brain Research 911 (2001) 37 – 42
Fig. 3. Effects of overexpression of calbindin on metabolic rate in primary hippocampal cultures (measured indirectly with silicon microphysiometery). The postglutamate response differed between the two groups (P,0.0001 by two-way ANOVA, F54384.2, df523). There was not a significant difference in the increase in metabolic rate observed during glutamate exposure in the two groups (N.S. by two-way ANOVA).
3.3. Microphysiometric measure of metabolism The metabolic rate of cells not exposed to glutamate declined only slightly over the duration of the 6.5-h experiment. Glutamate treatment caused a significant decline in metabolism in both calbindin vector-treated and control vector-treated cultures (Fig. 3; P,0.001, in both cases, by one-way ANOVA). However, the overall metabolic rate of cultures overexpressing calbindin declined significantly less than that of cells treated with control vector (P,0.0001, by two-way ANOVA).
4. Discussion In healthy neurons, calcium concentrations are under tight control, and are maintained at a concentration 10-fold less than that of the extracellular milieu; such a discrepancy in calcium concentrations allows small increases in cytosolic calcium levels to effect powerful changes in cellular signal transduction. Mechanisms to extrude and sequester excess cytosolic calcium are energetically expensive. Active transport mechanisms include the calcium ATPase and the sodium–potassium ATPase coupled to the sodium–calcium exchanger. Decreasing glutamate-induced calcium mobilization would be likely to decrease ATP expenditure by these mechanisms. In the glutamate-insulted neuron, sparing of cellular ATP could be critical to
cell recovery, particularly because excitotoxic insults are often themselves energetic in nature. Mitochondria, with their critical role in cellular energetics, can also act as a ‘calcium sink’ when the free cytosolic calcium concentration overwhelms active transport mechanism [20]. Such intramitochondrial calcium sequestration can uncouple the electron transport chain [9,25] and inhibit the TCA cycle enzymes [5], impairing the cell’s ATP-generating capacity. Intramitochondrial calcium sequestration can also result in the formation of the mitochondrial permeability transition (MPT), collapsing the proton gradient crucial to aerobic ATP generation [3,6,7,27]. Decreasing glutamate-induced intramitochondrial calcium sequestration would help prevent MPT pore formation, as well as uncoupling of the electron transport chain. Excitotoxic increases in free cytosolic calcium can activate proteases, endonucleases and phospholipases, producing cytoskeletal derrangement, membrane and oxidative damage. Compounding these consequences of calcium influx may be the energy expenditure required to repair such cellular damage. This represents another potential component of the energetic load of cytosolic calcium accumulation. Thus reducing cytosolic calcium accumulation should be energetically sparing. Overexpression of calbindin decreases the rise in intracellular calcium following hypoglycemia or exposure to excitotoxins in vitro [17,18].
M.L. Monje et al. / Brain Research 911 (2001) 37 – 42
Calbindin overexpression also decreases neurons death under those circumstances [17,18], as well as following necrotic insults in vivo [14,22,26]. We now demonstrate that calbindin overexpression preserves a number of indices of energetics. We observed that calbindin overexpression spared cultures completely from the glutamate-induced declines in ATP levels. We are most interested in the energetic changes preceding cell death, so a potential confound exists. We measured ATP at 12 h postinsult. As noted, our pilot data indicate that this was the earliest time at which we could detect a decline in ATP levels in control vectoror mock-infected cultures. There is likely to be some neuron loss at that time [24], raising the possibility that measures of ATP in living cells could be contaminated with ATP released from dead neurons. However, our ATP measurements likely reflected levels in living cells. This is because experiments were ended by aspirating media from the cultures immediately prior to freezing, thereby removing necrotic cell debris. Using rhodamine fluoroscopy, we also found that calbindin overexpression buffered mitochondrial potential from glutamate-induced declines. That finding, along with the sparing of ATP levels, correlate well with our finding that calbindin overexpression buffered overall metabolic rate from a glutamate-induced decline. Silicon microphysiometry was used to obtain real-time measurements of metabolic rate as a function of extracellular proton extrusion [23]. Such extrusion reflects the acidic metabolic products of glycolysis, respiration, and ATP hydrolysis, including lactic acid, CO 2 and protons. Because microphysiometric units are arbitrary, we expressed data as absolute change from baseline, rather than as percent change of baseline; results were identical when analyzed either way. The dose of glutamate used in the microphysiometry experiments was higher than that in the other experiments. We have consistently found that cultures are less sensitive to excitotoxins in microphysiometric studies, perhaps reflecting the necessity of carrying out studies at room temperature. Our interest in calbindin D28k stems from evidence that it plays an important role in neuronal health. Calbindin is abundant in the CNS, but is notably absent from the pyramidal cells of the cortex and hippocampus, which are selectively vulnerable to excitotoxic insults [1,19]. Cells expressing endogenous calbindin have been shown to be relatively resistant to glutamate or calcium inophore exposure [16]. Furthermore, calbindin gene expression decreases in regions affected by neurodegenerative disease such as in the substantia nigra in Parkinson’s disease, the corpus striatum in Huntington’s disease, and in the nucleus basalis in Alzheimer’s disease [13]. With respect to Alzheimer’s disease, calbindin overexpression has been shown to protect from A-b fragment-induced elevations in free intracellular calcium and reactive oxygen species, and to preserve mitochondrial function in cultured PC12 cells
41
[8]. Our findings illustrate the positive impact of decreasing free cytosolic calcium on cellular energy metabolism in the setting of excitotoxicity. Preservation of cellular energetic parameters using calcium reduction strategies may prove to be clinically valuable in the treatment of some neurologic disease.
Acknowledgements Support was provided by NIH grant NS37520 and a TDRDP grant from the State of California to RS. The silicon microphysiometer was made available through the generosity of the Molecular Devices Corporation, Sunnyvale, CA, USA. Manuscript assistance was provided by Sheila Brooke and Theodore Dumas.
References [1] K.G. Baimbridge, M.R. Celio, J.H. Rogers, Calcium-binding proteins in the nervous system, Trends. Neurosci. 15 (1992) 303–308. [2] D.W. Choi, Glutamate neurotoxicity in cortical cell culture is calcium dependent, Neurosci. Lett. 58 (1985) 293–297. [3] M. Crompton, H. Ellinger, A. Costi, Inhibition by cyclosporin A of a Ca 21 -dependent pore in heart mitochondria activated by inorganic phosphate and oxidative stress, Biochem. J. 255 (1988) 357–360. [4] N.A. DeLuca, A.M. McCarthy, P.A. Schaffer, Isolation and characterization of deletion mutants of Herpes simplex virus type 1 in the gene encoding immediate-early regulatory protein ICP4, J. Virol. 56 (1985) 558–570. [5] R.M. Denton, J.G. McCormack, Ca 21 as a second messenger within mitochondria of the heart and other tissues, Ann. Rev. Phys. 52 (1990) 451–466. [6] K.K. Gunter, T.E. Gunter, Transport of calcium by mitochondria, J. Bioenerg. Biomemb. 26 (1994) 471–485. [7] T.E. Gunter, D.R. Pfeiffer, Mechanisms by which mitochondria transport calcium, Am. J. Physiol. 258 (1990) C755–786. [8] Q. Guo, S. Christakos, N. Robinson, M. Mattson, Calbindin D28k blocks the proapoptotic actions of mutant presenilin. 1: Reduced oxidative stress and preserved mitochondrial function, Proc. Natl. Acad. Sci. USA 95 (1998) 3227–3232. [9] L. Hillered, P. Chan, Role of arachadonic acid and other free fatty acids in mitochondrial dysfunction in brain ischemia, J. Neurosci. Res. 20 (1988) 451–456. [10] D.Y. Ho, E.S. Mocarski, R.M. Sapolsky, Altering central nervous system physiology with a defective Herpes simplex virus vector expressing the glucose transporter gene, Proc. Natl. Acad. Sci. USA 90 (1993) 3655–3659. [11] D.Y. Ho, S. Fink, M. Lawrence, T. Meier, T. Saydam, R. Dash, R. Sapolsky, Herpes simplex virus vector system: analysis of its in vivo and in vitro cytopathic effects, J. Neurosci. Meth. 57 (1995) 205– 215. [12] W. Hunzinger, S. Schrinkel, Rat brain calbindin D28: six domain structure and extensive amino acid homology with chicken calbindin D28, Mol. Endo. 2 (1988) 465–473. [13] A. Iacopino, S. Christakos, Specific reduction of calcium-binding protein (28-kilodalton calbindin-D) gene expression in aging and neurodegenerative diseases, Proc. Natl. Acad. Sci. USA 87 (1990) 4078–4082. [14] M. Kindy, J. Yu, R. Miller, Adenoviral vectors in ischemic injury, in: J. Kriegelstein (Ed.), Pharmacology of Cerebral Ischemia, Medpharm Science, Stuttgart, 1996.
42
M.L. Monje et al. / Brain Research 911 (2001) 37 – 42
[15] M.S. Lawrence, D. Ho, R. Dash, R. Sapolsky, Herpes simplex virus vectors overexpressing the glucose transporter gene protect against seizure-induced neuron loss, Proc. Natal. Acad. Sci. USA 92 (1995) 7247–7251. [16] M. Mattson, B. Rychlik, C. Chu7, S. Christakos, Evidence for calcium-reducing and excito-protective roles for the calcium-binding protein calbindin-D28K in cultured hippocampal neurons, Neuron 6 (1991) 41–51. [17] T.J. Meier, D.Y. Ho, R.M. Sapolsky, Increased expression of calbindin D28k via Herpes simplex virus amplicon vector decreases calcium ion mobilization and enhances neuronal survival after hypoglycemic challenge, J. Neurochem. 69 (1997) 1039–1047. [18] T.J. Meier, D. Ho, T. Park, R. Sapolsky, Gene transfer of calbindin D28k cDNA via Herpes simplex virus amplicon vector decreases cytoplasmic calcium ion response and enhances neuronal survival following glutamatergic challenge but not following cyanide, J. Neurochem. 71 (1998) 1013–1023. [19] I. Mody, K.G. Baimbridge, J.J. Miller, Distribution of calbindinD28K 1 (CaBP) in the cerebral cortex and hippocampus of the epileptic (El) mouse, Epilep. Res. 1 (1987) 46–52. [20] D.G. Nicholls, A role for the mitochondrion in the protection of cells against calcium overload?, Prog. Brain. Res. 63 (1985) 97– 106.
[21] S. Orrenius, P. Nicotera, On the role of calcium in chemical toxicity, Arch. Tox. Suppl. 11 (1987) 11–19. [22] R.G. Phillips, T. Meier, L. Giuli, J. McLaughlin, D. Ho, R. Sapolsky, Calbindin D28K gene transfer via herpes simplex virus amplicon vector decreases hippocampal damage in vivo following neurotoxic insults, J. Neurochem. 73 (1999) 1200–1205. [23] K.M. Raley-Susman, K. Kersco, J. Owicki, R. Sapolsky, Effects of excitotoxin exposure on metabolic rate of primary hippocampal cultures: application of silicon microphysiometry to neurobiology, J. Neurosci. 12 (1992) 773–780. [24] S.M. Rothman, J.H. Thurston, R.E. Hauhart, Delayed neurotoxicity of excitatory amino acids in vitro, Neuroscience 22 (1987) 471–480. [25] Y. Takeuchi et al., A possible mechanism of mitochondrial dysfunction during cerebral ischemia: inhibition of mitochondrial respiration activity by arachidonic acid, Arch. Biochem. Biophys. 289 (1991) 33–38. [26] M.A. Yenari, et al., Calbindin D28k overexpression attenuates striatal neuron death following transient focal cerebral ischemia, Soc. Neurosci. Abstr. 841 (1999). [27] M. Zoratti, I. Szabo, The mitochondrial permeability transition, Biochim. Biophys. Acta 1241 (1995) 139–176.