A significant increase in both basal and maximal calcineurin activity in the rat pilocarpine model of status epilepticus

Authors


Address correspondence and reprint requests to Dr Severn B. Churn, Department of Neurology, Medical College of Virginia, Box 980599, Richmond, VA 23298, USA. E-mail: schurn@hsc.vcu.edu

Abstract

This study focused on the effects of status epilepticus on the activity of calcineurin, a neuronally enriched, calcium-dependent phosphatase. Calcineurin is an important modulator of many neuronal processes, including learning and memory, induction of apoptosis, receptor function and neuronal excitability. Therefore, a status epilepticus-induced alteration of the activity of this important phosphatase would have significant physiological implications. Status epilepticus was induced by pilocarpine injection and allowed to continue for 60 min. Brain region homogenates were then assayed for calcineurin activity by dephosphorylation of p-nitrophenol phosphate. A significant status epilepticus-dependent increase in both basal and Mn2+-dependent calcineurin activity was observed in homogenates isolated from the cortex and hippocampus, but not the cerebellum. This increase was resistant to 150 nm okadaic acid, but sensitive to 50 µm okadaic acid. The increase in basal activity was also resistant to 100 µm sodium orthovanadate. Both maximal dephosphorylation rate and substrate affinity were increased following status epilepticus. However, the increase in calcineurin activity was not found to be due to an increase in calcineurin enzyme levels. Finally, increase in calcineurin activity was found to be NMDA-receptor activation dependent. The data demonstrate that status epilepticus resulted in a significant increase in both basal and maximal calcineurin activity.

Abbreviations used
CaN

calcineurin

SE

status epilepticus

MOPS

3-(N-morpholino)propane sulfonic acid

DTT

d,l-dithiothreitol

LTD

long-term depression

LTP

long-term potentiation

pNPP

p-nitrophenol phosphate

pNP

p-nitrophenol

Tween-20

polyoxyethylene sorbitan monolaurate

PAGE

polyacrylamide gel electrophoresis

PBS

phosphate-buffered saline

PMSF

phenylmethylsulfonyl fluoride

SDS

sodium dodecyl sulfate.

Status epilepticus (SE) is a severe medical condition, affecting 102 000–152 000 people per year in the USA (DeLorenzo et al. 1995, 1996). Defined as continuous seizure activity lasting 30 minutes or greater, SE is a life-threatening medical emergency, associated with significant mortality rates (Bleck 1991; DeLorenzo et al. 1996; Fountain 2000). According to recent studies, status epilepticus results in between 22 000 and 42 000 fatalities every year in the USA (DeLorenzo et al. 1995, 1996). In addition, SE is also associated with physiological and neurochemical changes which take place during the seizure activity (Wasterlain et al. 1993; Rice et al. 1996; Fountain 2000). These cellular changes may later manifest themselves as spontaneous recurrent seizure activity (i.e. epilepsy), through a process known as epileptogenesis (Lothman and Betram 1993; Fountain 2000). Therefore, victims of status epilepticus incur both an immediate, serious risk of death, and another, long-term risk of developing recurrent seizure activity.

Modulation of intracellular calcium levels is important in a number of normal neuronal processes including regulation of gene expression, development of learning and memory, and neurotransmitter synthesis and release (Churn 1995; Chittajallu et al. 1998). However, alterations in calcium-regulated systems and loss of calcium homeostasis have also been implicated in many pathological conditions, such as ischemia (Choi 1987; Parsons et al. 1997, 1999), traumatic brain injury (Tymianski and Tator 1996; Rzigalinski et al. 1998), and SE (Pal et al. 1999; Parsons et al. 2000). Specifically, influx of calcium through the NMDA subtype of glutamate receptor is suspected to be important for physiological changes occurring after SE (Wasterlain et al. 1993; Rice and DeLorenzo 1998; Pal et al. 1999). NMDA-linked increases in intracellular calcium affect a number of calcium-controlled cellular mechanisms and enzymes including calcium/calmodulin-dependent kinase II (EC 2.7.1.123, CaM kinase II) (Churn et al. 1995; Kochan et al. 2000), calpain I (EC 3.4.22.17 del Cerro et al. 1994), and calcineurin (EC 3.1.3.16 CaN) (Montoro et al. 1993).

Calcineurin is a calcium/calmodulin-stimulated phosphatase enriched in neural tissue (Pallen and Wang 1985). CaN-mediated dephosphorylation is an important modulatory factor in many cellular processes, including development of learning and memory (Riedel 1999), regulation of long-term potentiation (LTP) and long-term depression (LTD) (Wang and Stelzer 1994; Thiels et al. 2000), and induction of apoptosis (Springer et al. 2000). Additionally, CaN may depress the activity of the GABA receptor, an important inhibitory receptor in the brain (Huang and Dillon 1998). These characteristics make CaN an intriguing enzyme for study in SE. In spite of this, no studies have yet focused on SE-dependent changes in CaN activity.

This study was undertaken to characterize changes in calcineurin activity in specific brain regions following status epilepticus. Basal and maximal CaN activities were investigated, with an NMDA-receptor activation-dependent increase present in both. Since CaN regulates many important neuronal processes (Wang and Stelzer 1994; Riedel 1999; Springer et al. 2000; Thiels et al. 2000), an increase in the activity of this important enzyme may be involved in SE-induced alterations of neuronal function.

Materials and methods

All materials were reagent grade and were purchased from Sigma Chemical Co. (St Louis, MO, USA) unless otherwise noted. Bio-Rad protein assay concentrate, sodium dodecyl sulfate (SDS), 2-mercaptoethanol, blotting grade dry milk, and Mini-Protean II SDS-PAGE apparatus and power supplies were purchased from Bio-Rad Laboratories (Richmond, CA, USA). Okadaic acid was purchased from BioMol (Plymouth Meeting, PA, USA). ABC reagents were obtained from Vector Laboratories (Burlingame, CA, USA). Absolute ethanol was purchased from AAPER Alcohol and Chemical Co. (Shelbyville, KY, USA). Adult male Sprague–Dawley rats were obtained from Harlan Laboratories (Indianapolis, IN, USA).

Pilocarpine model of status epilepticus

All animal use procedures were in strict accordance with the Guide for the Care and Use of Laboratory Animals described by the National Institutes of Health and were approved by the Virginia Commonwealth University Animal Use Committee. SE was induced in adult male Sprague–Dawley rats by the method described in Mello et al. (1993). Rats were handled regularly for approximately 1 week to acclimate them to handling prior to drug administration. Adult rats weighing 175–230 g were injected intraperitoneally (i.p.) with 350 mg/kg pilocarpine HCl, a potent muscarinic agonist. Thirty minutes prior to the pilocarpine injection, all rats were treated i.p. with 1 mg/kg methylscopolamine. Methylscopolamine is a peripheral muscarinic antagonist; administering it prior to pilocarpine injection minimizes adverse peripheral effects. Control rats received all injections, except that pilocarpine was administered at 1/10th its seizure-inducing dose. Control animals did not exhibit any behavioral seizure activity. When appropriate, rats were injected with the noncompetitive NMDA antagonist, MK-801 (4 mg/kg, i.p.), 20 min before pilocarpine injection. All drugs were dissolved in 0.9% saline.

Behavioral seizures typically began 20 min following pilocarpine injection. Seizures were rated according to the scale described by Racine (1972). Only animals reaching stage-5 seizures were included in this study. Stage 5 was classified as the most severe level of seizure activity, characterized by rearing and falling. Rats not responding to the initial pilocarpine injection received a second injection, often of a lower dose. Rats that still did not display stage-5 seizure activity following a second pilocarpine injection were excluded from the SE group (Kochan et al. 2000).

Isolation and homogenization of brain regions

Rats were rapidly killed by decapitation after 1 h of SE. Brains were dissected on a Petri dish on ice to preserve post mortem enzyme activity. Cortical, hippocampal, and cerebellar brain regions were quickly isolated and immediately homogenized with 10 strokes (up and down) at 12 000 r.p.m., using a motorized homogenizer (TRI-R Instruments, Inc. Rockville Center, NY, USA). Brain regions were homogenized into ice-cold homogenization buffer containing 5 mm HEPES (pH 7.0), 7 mm EGTA, 5 mm EDTA, 1 mm dithiothreitol (DTT), 0.3 mm phenylmethylsulfonyl fluoride (PMSF), and 300 mm sucrose. Cortical regions were homogenized into 7 mL of buffer, hippocampal regions into 3 mL of buffer, and cerebellar regions into 5 mL. All brain homogenates were normalized for protein, separated into aliquots and stored at −80°C until used.

PNPP assay of phosphatase activity

Calcineurin activity was assayed using a modification of the procedure detailed by Pallen and Wang (1983a). All reaction tubes were prepared on ice and contained the following: 25 mm MOPS (pH 7.0), 1 mm DTT, 2 mmp-nitrophenol phosphate (pNPP). Some basal tubes also contained 2 mm EGTA and 2 mm EDTA. Maximal tubes contained the same reagents as basal reactions, with the addition of 2 mm MnCl2. Final reaction volumes were 1 mL. Both Mn2+ and Ca2+/calmodulin activate calcineurin in this assay; however, Mn2+ activates calcineurin more strongly than calcium (Pallen and Wang 1983b). Therefore, Mn2+ was used instead of calcium in maximal reactions to better visualize the cation-stimulated activity of the enzyme. When called for, 150 nm okadaic acid, 50 µm okadaic acid (sodium salt), or 100 µm NaVO4 was added to both basal and maximal tubes. Prior to use, the protein concentration of all homogenates was determined using the method of Bradford (1976). Reactions were initiated by the addition of 200 µg/mL brain region homogenate. Unless otherwise specified, standard reactions were incubated at 37°C for 15 min in a shaking water bath. Tubes were then removed from the water bath and placed in ice to stop the reaction. Absorbance of the reaction mixture was immediately measured at 405 nm in a spectrophotometer (UV-2101, Shimadzu Scientific Instruments, Inc. Columbia, MD, USA). Absorbance units were converted to nmol of pNP produced by comparison with a pNP concentration standard absorption curve.

PNPP substrate kinetics

Basal reaction tubes were prepared containing 25 mm MOPS, 1 mm DTT, 2 mm EGTA, 2 mm EDTA, and 150 nm okadaic acid to minimize the influence of other phosphatases on the results of the assay. Maximal reaction tubes contained 25 mm MOPS, 1 mm DTT, 2 mm MnCl2, and 150 nm okadaic acid. All reaction tubes were prepared on ice. pNPP was added to the tubes for final concentrations of 0, 0.1, 0.2, 0.5, 1, 2, 4, 8, 16 and 24 mm. CaN reactions were initiated by the addition of homogenate and assayed as described above. Values for Km and Vmax were obtained via a nonlinear regression analysis of the substrate isotherm.

CaN protein level analysis

Western analysis was performed essentially as described previously (Churn et al. 1992; Churn et al. 1995). Briefly, homogenates were resolved on SDS–PAGE and transferred to a nitrocellulose membrane using the Geni blot system (Idea Scientific, Minneapolis, MN, USA). Nitrocellulose was then immersed for 1 h in blocking solution containing phosphate-buffered saline (PBS, pH 7.4), 0.05% (v/v) polyoxyethylene sorbitan monolaurate (Tween-20), 50 g/L Bio-Rad blotting grade dry milk, and horse serum (Vector Laboratories Inc., diluted 50 µL/10 mL). The nitrocellulose membrane was then incubated with primary antibody in blocking solution for 1 h. Anti-CaN A antibody (clone CN-A1, mouse monoclonal IgG; Sigma Chemical Co.) was diluted 1 : 10 000 for western analysis. Membranes were then washed three times in a wash solution containing PBS, Tween-20, and dry milk. Nitrocellulose was then reacted with a secondary antibody in blocking solution for 45 min. The membrane was then washed three times and incubated with an avidin-biotinylated horseradish peroxidase complex for 30 min. Nitrocellulose was then washed three times in PBS for 10 min per wash. Blots were developed with a solution containing 10 mL PBS, 0.025% (v/v) H2O2, and 8 mg 4-chloro-1-napthol dissolved in 2 mL of methanol. Specific immunoreactive bands were quantified by computer-assisted densitometry (Inquiry, Loats Associates, Westminster, MD, USA) as described previously (Churn et al. 1992).

Statistical analysis

All statistical analysis of data was performed using GraphPad Prism 3.0 (GraphPad Software, San Diego, CA, USA), or Sigma Stat 2.03 (SPSS Inc, Richmond, CA, USA). Levels of phosphatase activity were compared with control using the Student's t-test (two-tailed distribution, unpaired or paired where appropriate) or one-way anova with Tukey post analysis to control for type 1 errors in multiple comparisons. Values for Vmax and Km were determined by a nonlinear regression curve for single-site substrate binding, included with the Prism software. Computer assisted densitometric analysis was performed using Inquiry (Loats Associates Inc, Westminster, MD, USA).

Results

Status epilepticus results in an increase in calcineurin activity in specific brain regions

Calcineurin dephosphorylates the low molecular weight compound p-nitrophenolphosphate (pNPP) with high affinity, producing p-nitrophenol (pNP) (Pallen and Wang 1983a). It is possible to measure this dephosphorylation by spectrophotometry, providing a simple and accurate method for quantifying calcineurin activity. In the present study, CaN activity was measured via the pNPP assay in homogenates obtained from the cortex, hippocampus, and cerebellum of SE and control rats.

Both basal and maximal CaN activity were measured in SE and control animals (Fig. 1). In every sample studied, SE caused an increase in phosphatase activity above control levels in both the cortex and hippocampus, but not the cerebellum. Average basal activity in control samples was 0.79 ± 0.10 µg pNP/min in cortical homogenates, and 0.86 ± 0.03 µg pNP/min in hippocampal homogenates. Following SE, basal CaN activity was elevated 48.6 ± 14.5% above control in the cortex (Fig. 1a) (1.18 ± 0.04 µg pNP/min, p < 0.01, n = 6, two-tailed unpaired Student's t-test) and 20.8 ± 4.9% in the hippocampus (Fig. 1b) (1.04 ± 0.03 µg pNP/min, p < 0.001, n = 9). The elevated activity was biological, and could be eliminated by boiling the homogenate prior to beginning the reaction. The increase was present in spite of the addition of ion chelators to the reaction mix (see methods).

Figure 1.

SE resulted in a significant increase in basal and maximal CaN. Basal (white) and maximal (black) calcineurin activity was assayed in cortical, hippocampal, and cerebellar homogenates isolated immediately after 1 h of SE. (a) Cortical homogenates isolated from SE animals displayed a significant increase in both basal and maximal CaN activity. (**p < 0.01, n = 6, two-tailed unpaired Student's t-test). (b) Basal and maximal CaN activity was also elevated in the hippocampal homogenates isolated from SE animals when compared with control animals (*p < 0.05, n = 9; **p < 0.01, n = 9, two-tailed unpaired Student's t-test). (c) No significant differences were observed in either basal or maximal calcineurin activity in cerebellar homogenates (p > 0.1, n = 7).

Unlike homogenates from the cortex and hippocampus, basal activity in cerebellar homogenate was not significantly increased after SE. Basal activity in cerebellar homogenates isolated from control animals was 0.92 ± 0.08 µg PNP/min. After SE, activity in the cerebellum was elevated to 1.00 ± 0.05 µg PNP/min, an increase of 8.0 ± 10.1% (Fig. 1c). This was not a significant increase (p > 0.05, n = 7). The difference in the response of the cerebellum from that of the forebrain suggests that different mechanisms are involved in the expression of SE in the cerebellum from that of the forebrain.

Maximal activity, like basal activity, was increased after SE in both the cortex and hippocampus, but not the cerebellum. In cortical homogenates isolated from control animals, pNPP dephosphorylation was found to be 1.34 ± 0.18 µg pNP/min. Following SE, maximal cortical dephosphorylation activity was elevated to 1.80 ± 0.05 µg pNP/min, an increase of 34.2 ± 10.2% (Fig. 1a, p < 0.01, n = 6). An increase in maximal activity was also observed in hippocampal homogenates. Control maximal activity was 1.36 ± 0.05 µg pNP/min. SE resulted in a 13.7 ± 5.1% increase in activity, to a post-SE value of 1.55 ± 0.02 µg pNP/min. (Fig. 1b, p < 0.05, n = 9). No significant change in maximal activity was observed in the cerebellum. Control maximal activity was 1.54 ± 0.12 µg pNP/min in cerebellar homogenates. After SE, maximal activity increased to 1.60 ± 0.08 µg pNP/min. This was an increase of 4.0 ± 3.4% over control, which was not statistically significant (Fig. 1c, p > 0.05, n = 7).

The SE-dependent increase in CaN activity is due to an alteration in enzyme kinetics

To further characterize the SE-dependent effect on calcineurin activity, several experiments were conducted in cortical homogenate to examine the effects of SE on CaN-mediated pNPP dephosphorylation kinetics. Figure 2 shows the linear regression analysis of the time course for presteady state pNPP dephosphorylation under basal and maximal CaN conditions. The slope of the linear regression for each condition was used to estimate the rate of pNPP dephosphorylation. The rate of CaN-dependent dephosphorylation of pNPP in cortical homogenates from control animals was estimated to be 0.26 µg/min under basal conditions. SE resulted in a 36.7 ± 7.6% increase in the rate of pNPP dephosphorylation, to 0.36 µg/min (Fig. 2a, p < 0.001, n = 3, unpaired Student's t-test). The increase in pNPP dephosphorylation rate was also observed in maximal CaN reactions (Fig. 2b). The rate of control pNPP dephosphorylation was 0.68 µg/min under maximal conditions. SE resulted in a significant increase in pNPP dephosphorylation, to a value of 0.80 µg/min (17.6 ± 4.7%, p < 0.05, n = 3, unpaired Student's t-test). The data demonstrate a significant increase in both basal and maximal pNPP dephosphorylation rates in cortical homogenate

Figure 2.

SE modulates the rate of CaN-mediated dephosphorylation in rat cortical homogenates. CaN dephosphorylation reactions were performed in cortical homogenates isolated from control (white circles) and SE (black circles) animals. Basal (a) and maximal (b) CaN reactions were allowed to continue for specific time intervals, and the corresponding activities plotted to determine reaction rate. The portion of the reaction shown is during the linear (pre-steady state) portion of the reaction (a). Under basal conditions, the reaction rate (as indicated by linear regression analysis of the data with time as the independent variable and µg pNP produced as the dependent variable. r2 = 0.992) was 0.257 µg/min in control homogenates (white circles). CaN-mediated dephosphorylation in SE cortical homogenates (black circles) occurred at a rate of 0.355 µg/min. This represents an increase of 36.72 ± 7.6% when compared with the control rate. (b) Maximal dephosphorylation occurred at a rate of 0.684 µg/min in control homogenates, as indicated by linear regression analysis (r2 = 0.991). The maximal reaction rate in SE homogenates was 0.804 µg/min, an increase of 17.55 ± 4.74% over the control rate. Data represents an average of three experiments.

To determine the effect of SE on CaN affinity for pNPP, cortical homogenates from SE and control animals were tested in the presence of specific pNPP concentrations. (Fig. 3). Through nonlinear regression analysis of the data, maximal pNPP dephosphorylation (Vmax) and substrate affinity (Km) for the dephosphorylation of pNPP were calculated. The control value for Vmax was 1.21 ± 0.06 µg pNP/min under basal reaction conditions. Following SE, Vmax increased to 1.46 ± 0.03 µg pNP/min, an increase of 20.5 ± 5.3% above the control value (p < 0.05, n = 3, two-tailed unpaired Student's t-test). Substrate affinity was also increased under basal conditions. In homogenates isolated from control animals, Km was found to have a value of 1.25 ± 0.22 mm. Post-SE, Km decreased to 0.54 ± 0.06 mm. This decrease in Km corresponded to a 56.8 ± 14.5% increase in substrate affinity in SE homogenates when compared with control (p < 0.05, n = 3). Under maximal conditions, Vmax was also increased after SE. In control homogenates, Vmax was determined to be 1.73 ± 0.05 µg pNP/min. Following SE, Vmax was elevated to 1.93 ± 0.04 µg pNP/min, an increase of 12.8 ± 3.5% over control (p < 0.05, n = 3). Substrate affinity appeared to increase in SE homogenates under maximum conditions; however, this difference was not statistically significant (p > 0.05, n = 3).

Figure 3.

SE alters CaN substrate affinity in rat cortical homogenates. Substrate affinity (Km) and maximal pNPP dephosphorylation (Vmax) for CaN were determined under basal and maximal conditions for both SE and control cortical homogenates. (a) SE increased both Km and Vmax for pNPP dephosphorylation by CaN under basal conditions. A substrate concentration isotherm was generated under basal dephosphorylation conditions in cortical homogenates isolated from control animals (white circles) and SE animals (black circles). Basal CaN reactions were performed at specific concentrations of pNPP (see methods). Both Vmax and substrate affinity were increased in SE homogenates when compared with control. Inset: Double reciprocal (Lineweaver–Burk) plots of data from control and SE homogenates. SE resulted in an increase in maximal dephosphorylation (Vmax) under basal conditions. Vmax for SE homogenates (black circles) was 20.5 ± 5.29% higher than that of control homogenates (white circles) (p < 0.05, n = 3, two-tailed unpaired Student's t-test). Substrate affinity, Km, was increased by 56.75 ± 14.49% after SE (p < 0.05, n = 3, two-tailed unpaired Student's t-test). (b) SE increased Vmax for pNPP dephosphorylation under conditions for maximal CaN activity. Substrate concentration isotherms under maximal dephosphorylation conditions for pNPP dephosphorylation were generated in cortical homogenates isolated from control (white circles) and SE (black circles) animals. Maximal CaN reactions were performed at specific concentrations of pNPP (see Materials and methods). Maximal dephosphorylation was increased in SE homogenates when compared with control. However, substrate affinity in SE and control homogenates was not significantly different under maximal conditions. Inset: Double reciprocal plot of substrate isotherms from control and SE homogenates. Vmax was increased in SE homogenates by 12.80 ± 3.47% (p < 0.05, n = 3, two-tailed unpaired Student's t-test). Substrate affinity appeared to increase, however, the change was small and not statistically significant (p > 0.05, n = 3, two-tailed unpaired Student's t-test).

The increase in both basal and maximal pNPP dephosphorylation is due to CaN

The enzyme activity observed in the various brain regions is consistent with CaN activity: calcineurin dephosphorylates pNPP readily, is stimulated by Mn2+, and is most active at pH 6.8–7.0 (Pallen and Wang 1983a,b). However, pNPP is a versatile substrate, and is also dephosphorylated by some tyrosine and alkaline phosphatases (Thompson et al. 1991). All experiments were conducted at a neutral, rather than alkaline, pH (see Materials and methods). Therefore, it would be expected that alkaline phosphatases have only a minimal contribution to the observed dephosphorylation of pNPP. To rule out tyrosine phosphatases and other serine/threonine phosphatases, the phosphatase inhibitors sodium orthovanadate and okadaic acid were tested.

Sodium orthovanadate (NaVO4) is a potent inhibitor of tyrosine phosphatases (IC50 = 10 µm) (Swarup et al. 1982) Additionally, orthovanadate is a strong inhibitor of the Mn2+-stimulated activity of calcineurin (IC50 = 1 µm, Morioka et al. 1998). It does not, however, inhibit any non-Mn2+-stimulated calcineurin activity (Morioka et al. 1998). Dephosphorylation reactions were carried out in cortical homogenates in the presence of 100 µm NaVO4 under both basal and maximal (Mn2+-stimulated) conditions (Fig. 4). The addition of orthovanadate did not cause any significant change in basal pNPP dephosphorylation in either control or SE homogenates (p > 0.05, n = 4, one-way anova). Therefore, even in the presence of NaVO4, basal pNPP dephosphorylation in homogenates isolated from SE animals was still significantly greater when compared with control homogenates (p < 0.001, n = 4, unpaired Student's t-test). Thus, SE induced a NaVO4-resistant increase in pNPP dephosphorylation. The data support the hypothesis that the SE induced dephosphorylation was not due to tyrosine phosphatases.

Figure 4.

The SE-induced increase in pNPP dephosphorylation was not due to tyrosine phosphatase activity. Cortical homogenates were isolated from control and SE rats and subjected to basal (white bars) and maximal (black bars) reaction conditions. The reactions were performed in the absence (–) or presence (+) or 100 µm NaVO4. This level of NaVO4 is sufficient to inhibit tyrosine phosphatases and Mn2+-stimulated CaN activity. The inclusion of NaVO4 did not significantly affect basal pNPP dephosphorylation. However, inclusion of 100 µm NaVO4 resulted in complete inhibition of Mn2+-stimulated pNPP dephosphorylation (***p < 0.001, n = 4, two-tailed unpaired Student's t-test)

As expected, the inclusion of orthovanadate did have a significant effect on maximal pNPP dephosphorylation. Orthovanadate reduced maximal enzyme activity to basal levels in both SE and control homogenates. The addition of NaVO4 reduced control maximal activity from 1.19 µg pNP/min to 0.62 µg pNP/min, a decrease of 48.1 ± 4.6% (p < 0.001, n = 4, two-tailed paired Student's t-test). In SE homogenates, maximal activity was reduced by 40.7 ± 8.9%, from 1.67 µg pNP/min to 0.99 µg pNP/min (p < 0.001, n = 4). The NaVO4-inhibited maximal activity was significantly greater in SE homogenates when compared with control homogenates (p < 0.001, n = 4, two-tailed unpaired Student's t-test). Since orthovanadate specifically inhibits only Mn2+-stimulated CaN activity, the observed inhibition of maximal activity, but not basal activity, is consistent with CaN-dependent dephosphorylation.

To further characterize the CaN-dependent dephosphorylation of pNPP, okadaic acid was utilized. Okadaic acid (OA) is a potent inhibitor of serine/threonine phosphatases 1 and 2 A (Bialojan and Takai 1988), with an IC50 of 10–15 nm for PP1, and an IC50 of 100 pm for PP2A. Okadaic acid is also an inhibitor of calcineurin, although with a much lower affinity (IC50 = 5 µm) (Bialojan and Takai 1988). This difference in sensitivity makes OA a valuable tool in differentiating between the activities of these three phosphatases. CaN activity was first assayed in the presence of 150 nm OA, a concentration sufficient to inhibit both PP1 and 2 A, but not high enough to significantly affect CaN activity. The inclusion of 150 nm OA did not significantly affect pNPP dephosphorylation in cortical homogenates (Fig. 5a). The lack of OA-dependent inhibition was observed under both basal and maximal conditions for CaN activation. SE basal activity in the presence of 150 nm OA was 1.03 µg pNP/min, only 1.7% lower than the 1.05 µg pNP/min that was measured in the absence of inhibitor. This difference was not significant (p > 0.05, n = 4, two-tailed paired Student's t-test). SE maximal activity was 2.14 µg pNP/min in the absence of inhibitor, and 2.00 µg pNP/min in the presence of 150 nm OA, a decrease of 8.3% that was not statistically significant (p > 0.05, n = 4). Similarly, control basal activity was not significantly modified by low concentrations of OA (p > 0.05, n = 4). Control basal activity in the absence of OA was 0.62 µg pNP/min, and in the presence of OA activity was 0.69 µg pNP/min. Control maximal activity was 1.61 µg pNP/min without OA and 1.68 µg pNP/min with OA, a difference of 4.7% (p > 0.05, n = 4). These results indicated that PP1 and 2 A are not responsible for the SE-dependent increase in phosphatase activity observed under either basal or maximal conditions.

Figure 5.

The SE-induced increase CaN activity was inhibited by okadaic acid in a concentration-dependent manner. (a) Low concentration OA has no significant effect on pNPP dephosphorylation in cortical homogenate. The addition of 150 nm okadaic acid did not affect pNPP dephosphorylation under either basal or maximal activity in homogenates isolated from either SE or control animals (p > 0.05, n = 4). This suggests that neither PP1 (IC50 for OA: 10–15 nm) nor PP2A (IC50: 100 pm) is responsible for the measured activity. (b) High concentration of OA results in significant inhibition of pNPP dephosphorylation. Okadaic acid (50 µm) inhibited maximal activity in both control and SE cortical homogenates (**p < 0.01, n = 6, two-tailed paired Student's t-test). Maximal activity was inhibited by 10.65% in SE homogenates, and by 6.61% in homogenate isolated from control animals. High concentration OA also inhibited basal activity by 11.17% in SE homogenates (**p < 0.01, n = 6, two-tailed paired Student's t-test), but did not inhibit pNPP dephosphorylation in control homogenates (p > 0.05, n = 6, two-tailed paired Student's t-test). The data suggests that SE results in an OA-sensitive increase in pNPP dephosphorylation, measured under both basal and maximal conditions for CaN activity.

Dephosphorylation of pNPP was also carried out at a higher concentration (50 µm) of OA. This concentration of OA was high enough to block the activity of CaN, and was found to significantly decrease pNPP dephosphorylation in both control and SE cortical homogenates (Fig. 5b). Maximal activity in control homogenates was reduced from 1.29 µg pNP/min to 1.20 µg pNP/min by the addition of inhibitor (a decrease of 8.8 ± 2.8%, p < 0.01, n = 6). Similarly, maximal activity was reduced in SE homogenates by 10.7 ± 4.2%, from a value of 1.57 µg pNP/min without inhibitor to a value of 1.40 µg pNP/min in the presence of OA (p < 0.01, n = 6, two-tailed paired Student's t-test). However, basal activity was not inhibited by 50 µm OA in control homogenates (p > 0.05, n = 6). Under basal reaction conditions, most of the required ions for CaN activation are either not present or bound to chelators, and CaN activity in control basal reactions is probably very limited. Therefore, it is not surprising that 50 µm OA does not inhibit the control basal pNPP dephosphorylation. However, it is interesting to note that the SE basal activity, unlike the control basal activity, was inhibited by high-concentration OA. Basal activity in the SE homogenate was reduced from 1.06 µg pNP/min to 0.94 µg pNP/min by the addition of 50 µm OA, a decrease of 11.2 ± 4.4% (p < 0.01, n = 6). The data suggest that SE results in the production of a Ca2+-independent form of CaN that is sensitive to inhibition by high concentrations of OA. However, the activity in SE homogenates was not entirely reduced to control levels by the addition of 50 µm OA. This may be due to the solubility or efficacy limits of OA as a CaN inhibitor under these conditions, or it may suggest that another, OA-insensitive phosphatase is also contributing to the observed SE-dependent increase in phosphatase activity.

The lack of response to low concentrations of OA, the significant inhibition by high concentration OA, and the distinctive reaction of the homogenate to orthovanadate, all strongly suggest that CaN is responsible for the observed increase in activity. In addition, the response of the SE homogenate to high concentration OA under basal conditions suggests that the observed increase in basal activity is due to the presence of a calcium independent form of CaN.

The increase in enzyme activity is not due to an increase in protein

A potential explanation for the observed SE-dependent changes in CaN activity and kinetics is an increase in the amount of enzyme present in the cell. To investigate this possibility, western blot analysis was performed on homogenate isolated from the cortex and hippocampus (Fig. 6). Cerebellar homogenate was not tested since there was no SE-dependent change in CaN activity in the cerebellum. Immunoreactivity of the catalytic A subunit was determined using a monoclonal antibody to CaN A (CN-1; Sigma). In both the cortex (Fig. 6a) and hippocampus (Fig. 6b), a single band with molecular weight of 61 kDa was resolved. This size corresponds with the size of the CaN A subunit (Klee et al. 1998). Computer-assisted densitometric analysis was performed on SE and control fractions (see Methods and materials). Optical density of the immunoreactive band in SE homogenate was found to be 93.62 ± 3.01% of control in the cortex and 104.17 ± 14.57% of control in the hippocampus. No significant difference in the density of the 61 kDa band was found in either brain region, indicating equal total amounts of the enzyme in both SE and control animals in both regions (p > 0.05, n = 4, two-tailed unpaired Student's t-test). The data suggest that the increase in enzyme activity was not due to increased protein synthesis, but due to another mechanism, such as post-translational modification of the enzyme, or increased Ca/CaM binding to the enzyme following SE.

Figure 6.

The post-SE increase in CaN activity was not due to a SE-induced increase in CaN protein expression. To determine if SE results in increased CaN enzyme levels, homogenates from control and SE rats were subjected to western analysis. Cortical (a) and hippocampal (b) homogenates were resolved on SDS–PAGE and blotted to nitrocellulose (see Materials and methods). The resulting blots were reacted with monoclonal antibodies against the A subunit of CaN (Clone CN-1; Sigma Chemical Co.). SE did not result in a significant alteration of CaN immunoreactivity in any brain region studied. (a) Immunoreactivity was 93.62 ± 3.01% of control enzyme levels in cortical homogenates isolated from SE animals. (b) CaN immunoreactivity in hippocampal homogenates isolated from SE animals was 104.17 ± 14.57% of the control enzyme level.

The increase in calcineurin activity occurs through an NMDA-dependent mechanism

Calcium influx through the NMDA subtype of glutamate receptor is essential for the development of a number of SE-dependent alterations in neurochemical and physiological processes (Churn et al. 1995; Rice and DeLorenzo 1998; Kochan et al. 2000; Parsons et al. 2000). To investigate the involvement of NMDA-regulated calcium influx on the SE-dependent increase in calcineurin activity, MK-801 was used to block the NMDA receptor-associated calcium channel during seizure activity. Pre-injection of MK-801 has been shown to prevent many of the NMDA-receptor dependent consequences of SE, but does not prevent or inhibit the pilocarpine-induced seizures in this model (Rice and DeLorenzo 1998).

Cortical (Fig. 7a) and hippocampal (Fig. 7b) homogenates from MK-801 treated SE animals were assayed for phosphatase activity via the pNPP assay as described. Injection of MK-801 alone (no SE) did not induce a significant change in calcineurin activity. However, MK-801 injection did block the SE-induced increase in calcineurin activity. In the cortex, basal calcineurin activity in MK-801 animals was 0.88 µg pNP/min, only slightly higher than the control activity of 0.82 µg pNP/min. The observed increase was not statistically significant (7.2 ± 16.8% increase over control, p > 0.05, n = 4, one-way anova with Tukey's post hoc analysis). Maximal activity of MK-801 and control animals was also not significantly different: 1.38 µg pNP/min for control homogenates and 1.48 µg pNP/min for MK-801 homogenates (an increase of 7.1 ± 11.4%, p > 0.05, n = 4). Cortical calcineurin activity was significantly higher in pilo-SE animals than in either control or MK-801 treated animals. SE basal CaN activity was 1.21 µg pNP/min, an increase of 37.8 ± 11.0% over the MK-801 group (p < 0.05, n = 4). Maximal activity was 1.86 µg pNP/min, 26.0 ± 6.8% higher than that of the MK-801 homogenates (p < 0.05, n = 4). This data suggests that NMDA-dependent calcium influx during SE is essential for the observed increase in calcineurin activity in the cortex.

Figure 7.

MK-801 pretreatment prevented the SE-dependent increase in CaN activity in cortical homogenates. (a) Pre-injection of animals with MK-801 (a noncompetitive NMDA antagonist) completely prevented the SE-dependent increase in CaN activity in cortical homogenates. CaN activity in MK-801 cortical homogenates was not significantly different from that of control homogenates for both basal (white) and maximal (black) activities (p > 0.05, n = 4, one-way anova with Tukey post hoc analysis). SE homogenates had significantly higher activity than both MK-801 and control homogenates under both basal and maximal conditions (*p < 0.05, n = 4). (b) MK-801 pretreatment reduced, but did not completely prevent, the SE-dependent increase in CaN activity in hippocampal homogenates. MK-801 treated animals had significantly lower CaN activity than SE animals under both basal (white) and maximal (black) conditions (p < 0.05, n = 4). However, both basal and maximal hippocampal CaN activity in MK-801 animals was significantly higher than that of control animals (p < 0.05).

Hippocampal homogenates from MK-801-treated animals had significantly lower basal and maximal CaN activity than hippocampal SE homogenates (p < 0.05, n = 4, one-way anova with Tukey's post hoc analysis). However, unlike the cortex, CaN activity was not entirely reduced to control levels by MK-801 pre-injection. A significant increase in CaN activity was still found in MK-801 homogenates (Fig. 7b). The MK-801 basal value of 1.23 µg pNP/min was elevated 19.1% above the control value of 1.03 µg pNP/min, but still significantly lower than the activity in SE homogenates, which was 1.35 µg pNP/min, or a 31.1% increase over control (p < 0.05, n = 4). Likewise, maximal activity was elevated above control in the MK-801 group, with a value of 1.67 µg pNP/min, a 24.6% increase over the control value of 1.34 µg pNP/min. However, this increase was not as large as the increase in activity found in SE homogenates. SE maximal activity was 31.3% above control, with a value of 1.76 µg pNP/min. This suggests that either the MK-801 did not entirely block the NMDA receptor in hippocampal neurons, or that the SE-dependent increase in calcineurin activity is controlled by additional mechanisms in the hippocampus. Overall, NMDA receptor activation was found to play a significant role in the SE-induced increase in CaN activity in both brain regions.

Discussion

Understanding the cellular and molecular changes that take place during SE is essential to understanding the physiological mechanisms that cause this severe seizure activity. This study characterizes the changes in both basal and Mn2+-stimulated CaN activity in specific brain regions following SE. The data demonstrate a significant increase in CaN activity in cortical and hippocampal homogenates. The increase in phosphatase activity in cortical homogenates was found to be OA-sensitive. SE did not significantly affect CaN activity in homogenates isolated from the cerebellum. In addition, the SE-induced increase in CaN activity was NMDA-receptor activation-dependent. The present study is the first definitive investigation of SE-dependent modulation of calcineurin activity. The data supports the hypothesis that SE results in an NMDA-dependent increase in both basal and maximal calcineurin activity in the rat forebrain, but not in the cerebellum.

There are several possible mechanisms that could account for the increase in CaN-mediated dephosphorylation. Calcineurin is a calcium/calmodulin-stimulated enzyme (Klee et al. 1998) and would be stimulated by increased intracellular calcium concentrations. SE induces a loss of function of the endoplasmic reticulum Mg2+/Ca2+ ATPase (Parsons et al. 2000). This enzyme sequesters calcium ions into the microsomes of the smooth ER, providing a high-affinity mechanism for regulating intracellular calcium concentration (Carafoli 1987; Miller 1991). After SE, ATPase-mediated uptake of calcium into the microsomes is less efficient (Parsons et al. 2000), which could potentially result in higher than normal resting calcium concentrations inside the cell. In fact, an increase in intracellular free calcium has been found during and after SE (Pal et al. 1999). The increased intracellular calcium associated with SE could be responsible for activating CaN above its normal physiological level. However, it is unlikely that increased calcium observed in vivo is directly responsible for the increase in CaN activity following SE observed in this study. Increasing chelator concentrations did not effect the SE-induced increase in CaN dephosphorylation (data not shown). Additionally, no difference in free calcium between SE and control homogenates was present, as determined by calcium-specific electrode measurements as described by Parsons et al. (1997) (93–20 calcium electrode, Orion, Boston, MA, USA). Thus, the effects of any alterations in intracellular free calcium would be minimal under the conditions used in the present study.

Another possible explanation for the observed increase in CaN activity is an increase in the amount of enzyme present in the cell. This is unlikely, since it would require translation of a significant amount of enzyme in a relatively short amount of time (animals actively seized for slightly less than 1 h). Western analysis of SE and control homogenates further eliminated this possibility, as no significant increase in total enzyme level was observed in any brain region studied.

Another possible mechanism is post-translational modification of CaN, producing a more active form of the enzyme. One such modification could be limited proteolysis by the calcium-stimulated neutral protease, calpain I. Calpain may be activated by the increased intracellular calcium concentrations present during and after SE. Supporting this hypothesis, calpain is activated in other neurotraumatic events that involve glutamate excitotoxicity, such as ischemia (White et al. 2000) and traumatic brain injury (Buki et al. 1999). Calpain proteolysis could account for the increased Vmax, decreased Km, and increased enzyme activity observed post-SE in this study. Calpain cleaves the catalytic CaN A subunit, removing the calmodulin binding domain and creating a highly active, 43 kDa, calcium-independent form of the enzyme (Tallant et al. 1988; Wang et al. 1989). Some researchers also surmise that calpain-mediated proteolysis removes the autoinhibitory domain present in the A subunit (Wang et al. 1989), although other studies suggest that this may not be the case (Hashimoto et al. 1990).

Significant proteolysis of CaN was not responsible for the observed alteration in enzyme function, since no decrease in total enzyme was observed in western analysis of the SE homogenate. Additionally, no degradation products reacted with the anti-CaN A antibody. However, it is possible that the total amount of enzyme broken down by calpain was too small to detect in the western analysis that was used. Calpain-degraded CaN is reported to be 1100% more active than normal enzyme, while SE produced only a 20–50% increase in activity. The magnitude of this difference suggests that, if the calcium-independent 43 kDa fragment is present, it constitutes only a small amount (3–5%) of total enzyme. A difference in protein levels of this magnitude would be difficult to detect through western blot analysis. It is also possible that the epitope recognized by the antibody was no longer present following proteolysis, explaining why no small fragments appeared on the western blot.

Finally, post-translational modification other than calpain proteolysis may be taking place, activating CaN through another, unidentified mechanism. For example, calcineurin is phosphorylated at the same location by both protein kinase C and CaM kinase II, inhibiting the enzyme through a decrease in substrate affinity (Hashimoto and Soderling 1989; Martensen et al. 1989). Calcineurin and CaM kinase II share a similar distribution in the brain (Sola et al. 1999), and regulation of CaN by CaM kinase II is not unlikely. However, long-lasting inhibition of CaM kinase II has been observed following SE in multiple models (Bronstein et al. 1988; Churn et al. 2000; Kochan et al. 2000). It is possible that a decrease in the CaM kinase II-mediated phosphorylation of CaN produces a net increase in the amount of active calcineurin. While this mechanism would account for the observed increase in maximal calcineurin activity, it does not explain the SE-dependent development of calcium-independent CaN activity.

Calcium influx through the NMDA receptor has been implicated in many of the physiological changes that take place following SE (Rice and DeLorenzo 1998; Kochan et al. 2000). For this reason, it was of interest to determine if the observed SE-dependent increase in CaN activity was also NMDA-dependent. Blocking the NMDA receptor-associated Ca2+ channel with MK-801 completely prevented the SE-dependent increase in CaN activity in the cortex, and significantly reduced the magnitude of the increase in the hippocampus. This data suggests that calcium influx through the NMDA-receptor associated channel is involved in the SE-induced increase in calcineurin activity.

The increase in CaN activity described in this study was not observed in the cerebellum. Although pilocarpine-induced seizure activity generalizes into the cerebellum, many SE-induced physiological changes do not develop in this region. For example, CaM kinase II activity is altered in the cortex and hippocampus following SE, but not in the cerebellum (Kochan et al. 2000). Similarly, the spontaneous recurrent seizures that develop following pilocarpine treatment do not generalize into the cerebellum (Mello et al. 1993). Considering this, it is not surprising that SE-induced increase in calcineurin activity was not noted in the cerebellum. It is possible that SE activates different mechanisms in the cerebellum than in the forebrain, or that an enzyme or receptor necessary for the SE-dependent activation of CaN is not present in the cerebellum.

The observed SE-dependent increase in CaN activity has broad physiological implications. Neuronal mechanisms which are regulated by CaN dephosphorylation under normal conditions could be highly altered as a result of elevated enzyme activity. One important CaN-mediated mechanism is modulation of the GABAA receptor. GABA receptors are the primary receptor responsible for the fast inhibitory response in neuronal tissue (Macdonald and Olsen 1994), and play a major role in preventing the neuronal hyperexcitability associated with epilepsy and SE (Bleck 1991). Several recent studies have demonstrated an inhibitory modulation of GABAA receptor function by CaN (Chen et al. 1990; Chen and Wong 1995; Amico et al. 1998; Huang and Dillon 1998; Lu et al. 2000). A post-SE elevation in CaN activity may result in increased dephosphorylation of GABAA receptors, producing a net disinhibition of cellular excitability. This hypothesis is strengthened by the fact that recent studies have described a post-SE decrease in GABAA activity (Kapur and Coulter 1995). Chronic increases in CaN-mediated dephosphorylation could lead to a long-term loss of GABA receptor function, which may play a role in epileptogenesis. The results of one kindling study already suggest that calcineurin may have an epileptogenic role (Moia et al. 1994), and disinhibitory mechanisms have been found to be epileptogenic in several studies.

This study is the first to detail an SE-dependent effect on calcineurin activity. The results presented above describe an SE-induced increase in CaN activity which has far-reaching implications, from possible avenues for SE treatment to a potential role in development of epilepsy following SE. While this study focused on the total enzyme pool contained in the homogenate, future studies will investigate the effect of SE on CaN activity in specific cellular fractions, as well as on purified enzyme. A great deal of further research on this topic is possible, offering exciting insight into the mechanisms underlying status epilepticus.

Acknowledgements

The authors would like to thank Matthew Joose for his assistance and helpful suggestions. This work was supported by National Institute of Neurological Disorders and Stroke awards PO1-NS25630 (SBC and RJD) and RO1-NS23350 (RJD), and by a VCU Undergraduate Research Grant (JEK).

Ancillary