(Stroke. 1995;26:1259-1267.)
© 1995 American Heart Association, Inc.
Articles |
Delayed Antagonism of Calpain Reduces Excitotoxicity in Cultured Neurons
From the Departments of Neurology (J.R.B.) and Pharmacological and Physiological Sciences (C.J.M., R.J.M.), University of Chicago (Ill).
Correspondence to James R. Brorson, MD, Department of Neurology MC2030, University of Chicago, 5841 S Maryland Ave, Chicago, IL 60637. E-mail jbrorson@neurology.bsd.uchicago.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Background and Purpose Glutamate receptor antagonists can produce protection against the neurotoxicity of excessive glutamate stimulation. However, antagonism of the postreceptor processes that produce cell damage may provide a longer window of opportunity for protecting neurons after the initiation of excitotoxic injury. Among various processes that have been thought to mediate the toxic effects of glutamate are activation of the Ca2+-dependent proteases calpain I and II and the activation of nitric oxide synthase. We tested the potential for neuroprotection by delayed application of calpain antagonists after excitotoxic treatment.
Methods Primary cultures of cerebellar and hippocampal neurons were exposed to the glutamate receptor agonists kainate and N-methyl-D-aspartate (NMDA) for 20-minute periods, and survival was examined by fluorescent assay after 24 hours. Enzyme antagonists were applied at various time points during this interval.
Results The neurotoxic effects of NMDA in cultured hippocampal neurons and of kainate in cultured cerebellar neurons have been previously shown to be Ca2+ dependent. Here we show that in both of these examples of glutamate receptor–mediated toxicity, activation of a calpainlike proteolytic activity occurred, which was blocked by the calpain inhibitor MDL-28170. This inhibitor also limited the toxicity, even when applied at times up to 1 hour after the onset of the toxic exposure. Another protease inhibitor, E-64, also blocked the proteolysis and toxicity produced by kainate in cerebellar neurons. Blocking nitric oxide synthase activity after 1 hour with the antagonist NG-nitro-L-arginine was also protective of cerebellar and hippocampal neurons, as was the combination of MDL-28170 and NG-nitro-L-arginine.
Conclusions The activation of calpain is among several enzymatic processes that contribute to the toxicity of glutamate receptor stimulation, and blocking these postreceptor mechanisms can be effective in protecting neurons from excitotoxicity at delayed time points.
Key Words: calpain • excitotoxicity • glutamates • neuroprotection • nitric oxide
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Excessive action of glutamate at its excitatory receptors on central neurons is thought to play an important role in neuronal death in a number of brain diseases. Numerous in vitro and in vivo studies have shown that antagonists of glutamate receptors can decrease the neuronal damage induced by glutamate agonists or by experimentally induced stroke or ischemia.1 2 3 4 5 However, in several clinical situations in which neuronal cytoprotection would be of potential benefit, such as stroke, some delay necessarily elapses between the initiation of the insult and the application of therapy. Although some studies have found amelioration of damage with late application of glutamate receptor antagonists,6 7 8 it may be that blockade of the secondary processes, which occur as later steps in the pathways to structural damage in neurons, would produce better cytoprotection after a delay than would direct blockade of glutamate receptors.
Delayed glutamate receptor–mediated damage is largely the result
of excessive influx of Ca2+ into
neurons.9 A number of Ca2+-activated
processes are thought to mediate the toxic effects of elevated
cytoplasmic free Ca2+ concentrations
([Ca2+]i), including
Ca2+-activated lipases,10 the Ca2+-dependent
generation of free radicals,11 the generation of nitric
oxide (NO) by the
Ca2+/calmodulin-dependent enzyme
NO synthase,12 and the activation of proteases. The latter
category includes the Ca2+-activated proteases here
collectively referred to as calpain. The two related forms, calpain I,
activated at micromolar concentrations of
Ca2+, and calpain II, activated at millimolar
Ca2+ concentrations, have similar proteolytic
specificities and contain E-F hand structures typical for
Ca2+-binding proteins.13 14 Previous
authors have shown that glutamate agonists or hypoxia can
induce calpain activation and proteolysis in hippocampal
neurons15 16 and that inhibition of calpain can protect
hippocampal neurons against direct glutamate agonist–induced toxicity
and against ischemia-induced damage.17 18 19 We have
found that calpain inhibition can also block the
Ca2+-dependent death induced by the
non–N-methyl-D-aspartate (NMDA) agonist
kainate in cultured cerebellar Purkinje neurons,20 and
others have reported protection against
-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA)–induced
Purkinje cell damage in slices.21 Inhibitors
of calpain used include the nonspecific protease inhibitor
leupeptin and the more specific thiol protease inhibitors
E-6422 and MDL-28170 (or Cbz-Val-Phe-H); the latter agent
has the advantage of increased membrane permeability.23 No
studies have yet reported the effects of delayed application of calpain
inhibitors after glutamate receptor–induced damage in
vitro. We sought to assess the efficacy of delayed application of
calpain inhibitors in protecting against neuronal death
induced by NMDA in cultured hippocampal neurons and by kainate in
cultured cerebellar neurons, and we also studied the effects of
combining calpain inhibition with antagonism of NO synthase.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Neuronal Cultures
Cultures of hippocampal neurons taken from day 17 embryonic Holtzman rats and of cerebellar neurons from day 16 embryos were prepared largely as previously described in detail,24 25 except that 15 mmol/L HEPES was added to the culture medium for the cerebellar cultures. Procedures followed were in accordance with a protocol approved by the University of Chicago Institutional Animal Care and Use Committee. Trypsin-dissociated neurons were plated on 15-mm round glass coverslips and suspended over a feeding glial layer in a serum-free defined medium (N2.1). Hippocampal neurons for protein studies were cultured on poly-L-lysine–coated 35-mm culture dishes, with the feeding glial layer suspended over the neurons on perforated plastic disks. Neurons were used for toxicity experiments when of age 9 to 11 days in vitro and for protein blots when of age 10 to 14 days in vitro.
Excitotoxicity Assays
Cell death and survival were assayed with the fluorescent
markers fluorescein diacetate and propidium iodide.24 26
To reduce background staining of cellular debris, before agonist
exposure the coverslips of cultured neurons were placed for 24 hours in
medium to which 10% horse serum (GIBCO) had been added. The coverslips
were then washed and exposed at 37°C in room air to various agents in
saline buffers to which tetrodotoxin (0.5 µmol/L) and bicuculline (20
µmol/L) were added to eliminate indirect synaptic contributions to
toxicity. The buffer for kainate exposures contained the following
(mmol/L): NaCl 145, KCl 3, CaCl2 2, MgCl2 1,
HEPES 10, glucose 10, pH to 7.4 with NaOH. For NMDA exposures, a
similar buffer was used, except that the MgCl2 was omitted
and 10 µmol/L glycine added. After 20-minute exposures, the
coverslips were again washed in the same ionic solutions to remove the
agonists and returned to incubation in glia-conditioned, serum-free
medium. Where indicated, antagonists were added to the
medium bathing the cells at various times after agonist exposure.
After 24 hours, cell survival was assayed. The coverslips were washed in buffer and then exposed for 4 minutes to fluorescein diacetate (15 µg/mL) and propidium iodide (5 µg/mL). The stained cells were examined on an epifluorescence microscope at wavelengths appropriate for each fluorophore. Surviving neurons contained fluorescein, and the red fluorescence of propidium iodide marked the nuclei of dead neurons. These groups did not overlap. Living and dead neurons were counted on adjacent fields of each coverslip to totals of at least 100 by a blinded evaluator. The percentage of neurons surviving was determined on three to four coverslips for each condition in each experiment and normalized to parallel controls performed in the same ionic conditions. The average relative percent survival from at least three separate experiments for each condition is expressed in the text and figures as mean±SEM. The protective effects of various treatments in toxicity assays were compared with the effects of agonist alone in parallel trials by one-way ANOVA followed by Dunnett's test for multiple comparisons (SIGMASTAT, Jandel Scientific).
Protein Immunoblotting
Immunoblots of protein from cultured neurons were performed to
evaluate spectrin breakdown. Four to six 35-mm tissue culture dishes
containing hippocampal neurons or 18 to 25 coverslips of cultured
cerebellar neurons were taken to ensure enough protein for each
condition of each experiment. The cells were washed with buffer before
treatment and then treated with the specified agonist and
antagonist, dissolved from 1000-fold stock solutions into
saline buffers as described above. The neurons were incubated at 37°C
in the agonist-containing solution, washed again in the buffer,
returned to glia-conditioned medium (containing calpain or NO synthase
antagonists when indicated), and incubated at 37°C. At
the time of protein harvesting, the hippocampal neurons were washed
with cold PBS (0.9 g/L NaCl, 10 mmol/L phosphate buffer) and then
collected by mechanical scraping with 100 µL of PBS per tissue
culture dish. Phenylmethylsulfonylfluoride (0.5 mmol/L) was
added to halt further protease activity. The protein suspension was
centrifuged for 1 minute at 12 000 to 14 000 rpm. The
supernatant was discarded and the protein pellet brought up in 40 µL
of sample buffer. The cerebellar neurons were harvested by micropipette
trituration of 80 to 100 µL of sample buffer per condition on and off
the coverslips. Sample buffer contained 6.5% glycerol, 3.0% sodium
dodecyl sulfate, and 0.5% Tris in distilled water.
For in vitro digestion of spectrin by exogenous calpain, duck erythrocyte membranes (DEMs) were used as a plentiful source of spectrin. The digestion buffer contained the following (mmol/L): NaCl 25, CaCl2 1, HEPES 30, 2-mercaptoethanol 5, and 20 µg/mL of DEMs, with pH adjusted to 7.5. Calpain was obtained as rabbit skeletal muscle Ca2+-activated neutral protease from Sigma Chemical Co, dissolved at 2 U/mL in digestion buffer, and added at a final concentration of 0.2 U/mL to digestion buffer containing DEMs. This mixture was incubated at room temperature (except when indicated) alone or in the presence of MDL-28170 (10 µmol/L), E-64 (10 µmol/L), or EGTA (2.5 mmol/L). After 30 minutes the reactions were stopped with a stop solution and a 5-minute incubation at 55°C.
The protein from each condition was quantified spectrophotometrically (Micro BCA Protein Reagent Kit). Equal amounts of protein, as determined by the least amount found in any single condition, were loaded into each lane of the 12-cm 5% polyacrylamide gel, which was then run at 45 to 50 V for 12 to 15 hours, along with molecular weight markers and protein from calpain-treated purified DEMs, prepared as described above. The protein was then transferred from the gel to a nitrocellulose sheet at 1 A/h, run for 1 hour.
The blot was first stained with ponceau red (0.1%) for 30 minutes, revealing the success of protein transfer. Ponceau red was washed away with distilled water, and the blots were then blocked with 2.5% nonfat dry milk and 0.1% Tween 20 in PBS for 1 hour. After two 10-minute washes in PBS, the blot was stained for 3 hours with a polyclonal antispectrin antibody, the kind gift of Dr Ron Debreuil. This antibody was diluted to 1:1200 in PBS containing 5% fetal calf serum (FCS), 0.1% bovine serum albumin, and 0.1% Tween 20 (PBS/FCS) and was preserved with approximately 20 mg/L of sodium azide, which allowed it to be reused up to four times. The blot was washed in PBS/FCS three times for 10 minutes each. Then a horseradish peroxidase–conjugated anti-rabbit secondary antibody (Amersham) was applied, diluted to 1:5000 in PBS and 2.5% nonfat dry milk. After two 5-minute washes in PBS/FCS, subsequent washes with 0.3% and then 0.1% Tween 20 in PBS were performed (each three times for 5 minutes). Finally, an enhanced chemoluminescent detection solution (Amersham ECL RPN 2106 Kit) was applied for 1 minute, and the blot was exposed to Kodak 8x10 film for various times ranging from 15 seconds up to 8 hours. Quantification of immunolabeled protein bands was performed by scanning laser densitometry with the use of an UltroScan XL Laser Densitometer (Pharmacia) and subsequent analysis with GELSCAN XL 2.1 software (Pharmacia). Calculated ratios of the density of the two primary spectrin breakdown fragment bands divided by the sum of the densities of these bands and the two bands of intact spectrin were taken as the percent spectrin breakdown. The means of the percent spectrin breakdown from four separate experiments for controls and agonist plus calpain antagonists were compared with agonist alone with the use of repeated-measures ANOVA, followed by Dunnett's test for comparisons between groups (SIGMASTAT, Jandel Scientific).
Whole-Cell Patch Clamping
Whole-cell patch-clamp measurements of ligand-gated
Ca2+ currents were performed as previously
described.27 Cells were accepted for study if a stable
seal formed with a whole-cell resistance of at least 150 M
.
Intracellular solutions contained the following (mmol/L): CsF 145,
1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid
10, pH to 7.2 with CsOH (ATP was omitted to allow rundown of the
voltage-gated Ca2+ currents). The extracellular
saline buffer was as described above, with Mg2+
omitted and the additions of glycine (10 µmol/L), tetrodotoxin (0.5
µmol/L), and Cd2+ (100 µmol/L). Cells were held at a
membrane potential of -80 mV, and agonists were applied by a rapid
perfusion method as described by Tang et al.28 Briefly, a
tapered theta-tube applicator was brought near to the cell, and
parallel solenoid valves (Lee Co) were controlled so as to switch off
the control solution on one side of the theta-tubing
simultaneously with the opening of agonist flow on the
other side. All experiments were performed at room temperature. Data
were recorded on a PC-based system with the use of an Axopatch 1D
amplifier (Axon Instruments) and acquisition software based on
C-LAB II (Indec Systems Inc). The mean peak inward
currents were compared by means of paired two-tailed Student's
t tests.
Materials
The calpain antagonist MDL-28170 was the gift of Dr
Shujaath Mehdi (Marion Merrell Dow), and E-64 was obtained from
Calbiochem. Propidium iodide was obtained from Aldrich
Biochemicals, and 6-cyano-7-nitroquinolone-2,3-dione (CNQX) was
purchased from Research Biochemicals, Inc. The polyclonal anti-spectrin
antibody was the kind gift of Dr Ron Dubreuil, and DEMs were the gift
of Dr Clive Palfrey, both of the Department of Pharmacological and
Physiological Sciences, University of Chicago.
Other agents came from Sigma Chemical Co.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Toxicity of Glutamate Antagonists in Hippocampal and Cerebellar Neurons
The application of 100 µmol/L kainate for 20 minutes was substantially toxic to the cultured cerebellar neurons, resulting in a survival relative to controls of 42±4% (n=7) when assayed 24 hours later (Fig 1
). NMDA was a much less effective toxin to
these cells, even in conditions designed to maximize the activation of
NMDA receptors (Mg2+-free, glycine-supplemented
buffer), producing 78±5% relative survival. We have previously shown
that kainate produces toxicity in the cerebellar neurons by a
mechanism that is largely independent of activation of NMDA receptors
or of voltage-gated Ca2+ channels.20 In
contrast, in cultured hippocampal pyramidal neurons,
100 µmol/L NMDA, applied for 20 minutes, was an effective
excitotoxin, resulting in 49±4% relative survival, whereas, similar
to the results of others, 100 µmol/L kainate had very little
toxicity in hippocampal neurons, resulting in 83±3% relative survival
(Fig 1). The toxicity of kainate in cerebellar neurons was largely
blocked by the coapplication of the competitive non-NMDA
antagonist CNQX (10 µmol/L) and the toxicity of NMDA in
hippocampal neurons by the coapplication of the antagonist
D-2-amino-5-phosphonovaleric acid (D-AP5) (50 µmol/L)
(not shown). Thus, the hippocampal and the cerebellar neurons differ in
their sensitivity to toxicity due to activation of NMDA versus non-NMDA
receptors, with the cerebellar neurons vulnerable to
kainate-induced toxicity and the hippocampal neurons vulnerable to
NMDA-induced toxicity.
|
Glutamate Agonists and Calpain Activation
Both the toxicity of kainate in cultured cerebellar
neurons20 and the toxicity of NMDA in hippocampal
neurons29 can be shown to be largely dependent on
Ca2+ influx. One of the cytoplasmic processes
activated by elevation of [Ca2+]i is
protein degradation by the Ca2+-dependent proteases
or calpains. Using Western blots for spectrin, a structural protein
that is readily broken down by calpain,30 we investigated
whether the toxic agonist exposure in these cultured neurons was
accompanied by evidence of proteolytic activity suggestive of the
activation of calpain. We first treated purified protein from DEMs, a
plentiful source of the cytoskeletal protein spectrin, with exogenous
calpain in the presence of 1 mmol/L Ca2+. As shown
in Fig 2A, such treatment resulted in a decreased
density of the intact spectrin bands and a large increase in spectrin
breakdown products (BDPs). This breakdown of spectrin by calpain
could be inhibited by omitting or chelating the
Ca2+, by the calpain antagonist
MDL-28170 (10 µmol/L), and to a small extent by the
antagonist E-64 (10 µmol/L). The effect of E-64 was seen
as a preservation of the density of the intact spectrin band. The
locations of the BDP bands produced by exogenous calpain were used as
size markers in subsequent blots.
|
When 100 µmol/L NMDA was applied to cultured hippocampal neurons for
20 minutes and spectrin immunoblotting was performed on equal amounts
of total protein from the cells harvested at several time points
relative to the start of NMDA treatment, it could be seen that the NMDA
exposure produced an elevation in BDPs and a decrease in intact
spectrin over time, increasing from the initiation of the agonist
exposure up to 4 hours later (Fig 2B
). By 24 hours, the increase in
BDPs as a fraction of total protein was no longer evident (not shown).
For this reason, for subsequent experiments examining the effects of
various antagonists on spectrin breakdown, protein was
harvested from the treated neurons at 4 hours after agonist exposure.
The BDPs produced by NMDA treatment of hippocampal neurons were of the
same size and pattern as those produced by spectrin hydrolysis by
exogenous calpain (estimated molecular weights, 169 and 173 kD), and
production of the BDPs was blocked by MDL-28170. E-64 did little to
diminish the amount of BDPs but resulted in a small increase in the
intact spectrin band over NMDA alone (Fig 2C). A substantial background
level of BDPs was present in the hippocampal neurons, and prolonged
treatment with MDL-28170 in NMDA-treated cells actually reduced the
levels of BDPs below those in controls.
A similar result was found in cerebellar neurons treated with
kainate. Again, with equal amounts of total protein loaded in each
lane, in the kainate-treated samples the intact spectrin bands were
decreased in intensity and spectrin-immunoreactive fragments appeared,
which suggested calpain activation (Fig 3A). It appeared
that kainate treatment was sometimes associated with more extensive
spectrin degradation, resulting in lesser signals in the primary BDP
bands and still smaller fragments appearing (Fig 3B). The breakdown of
spectrin could be blocked by MDL-28170 or partly by E-64. The effects
of the calpain inhibitors were again best seen in the
preservation of the intact spectrin bands rather than as a decrease in
the intensity of the kainate-induced spectrin BDPs. Quantification
of these immunoblots by scanning densitometry was performed, with the
integrated density of the two primary BDP bands relative to the total
amount of spectrin immunoreactivity in these bands and in the two
primary bands of intact spectrin taken as the percent spectrin
breakdown. Although such a method might underestimate the amount of
spectrin breakdown by ignoring any more extensive breakdown of spectrin
into smaller fragments, it gives a lower limit to the spectrin
breakdown. These results are presented in Figs 2D and 3C and
confirm the qualitative impressions that the spectrin breakdown
produced by NMDA or kainate was blocked extensively by MDL-28170
and partially by E-64. In summary, the exposure of either cultured
hippocampal neurons or cultured cerebellar neurons to excitotoxic
stimuli produced the appearance of fragmentation of the cytoskeletal
protein spectrin over the first several hours, consistent with
that expected to result from activation of calpain.
|
The effects of the calpain antagonists are unlikely to be a
result of direct effects on the glutamate receptors. We previously
showed that the enzyme antagonists MDL-28170, E-64, and
NG-nitro-L-arginine (N-Arg)
were found to have little direct effect on the size of membrane
currents evoked by application of kainate to cerebellar
neurons.20 Similarly, MDL-28170 and N-Arg were without
significant effect on NMDA-induced currents in hippocampal neurons,
with currents in 10 µmol/L MDL-28170 and in 100 µmol/L N-Arg found
to be 95±6% and 92±5%, respectively, of control currents (Fig 4). Thus, the inhibition of the production of BDPs
or of the toxicity by these antagonists must have been a
result of their effects at postreceptor mechanisms, most likely at
their target enzymes.
|
Neuroprotective Effects of Calpain Antagonists
What involvement does calpain activation have in the
excitotoxicity produced by NMDA or by kainate? If calpain
activation is essential to the toxicity, then application of calpain
antagonists should block the toxicity. We applied the
antagonists MDL-28170 (10 µmol/L) or E-64 (10 µmol/L)
to the cultured cerebellar neurons and MDL-28170 to the hippocampal
neurons at various times relative to the start of a 20-minute exposure
to kainate or NMDA and continued the application of the
antagonist for the duration of a 24-hour period, after
which assays of cell survival were performed. In the cerebellar
neurons, both MDL-28170 and E-64 effectively improved neuronal
survival, preventing most of the toxicity when applied before or
concurrent with the kainate exposure and remaining significantly
protective even when applied at times as late as 1 hour after
kainate application (Fig 5A). There was no
significant protective effect when the calpain antagonists
were applied 4 hours after kainate exposure, in which case the
survival fell nearly to that of kainate treatment without
antagonist. CNQX was also protective when applied
immediately after the kainate exposure or when applied after 1
hour, although to a somewhat lesser extent than the calpain
antagonists. For comparison, each of the
antagonists was individually applied to the neurons for 24
hours (without any prior agonist exposure), and they each seemed to
have a slight direct toxicity, although only the effect of MDL-28170
reached statistical significance when compared with controls, resulting
in 89±7% survival.
|
In the hippocampal neurons exposed to NMDA, MDL-28170 was again found
to be protective when applied at times up to 1 hour after NMDA
application, but not at 4 hours (Fig 5B). In contrast, the NMDA
antagonist D-AP5 (50 µmol/L) was not found to be
significantly protective when applied at delayed time points after the
NMDA exposure. In separate experiments, E-64 also had protective
effects in hippocampal neurons when applied concurrently with or
immediately after NMDA exposure, but not when applied after 1 hour (not
shown). D-AP5 alone, unlike MDL-28170 or E-64 alone, applied for 24
hours was itself significantly toxic to the cells, producing a relative
survival of 81±5%.
NO Synthase Inhibition
Another Ca2+-activated enzymatic process that
has been found by some investigators12 to mediate
glutamate receptor–induced toxicity in central neurons is the
production of NO by NO synthase, which can be competitively
inhibited by N-Arg. We also tested the protective effects of N-Arg (100
µmol/L) in these systems. When N-Arg, or N-Arg combined with
MDL-28170, was applied alone for 24 hours, it produced slight
reductions in survival compared with controls, with only the effect of
N-Arg alone in hippocampal cultures reaching statistical significance,
at 85±7% relative survival. When used after glutamate agonists, N-Arg
applied 1 hour after excitotoxic exposure was found to be protective of
either the cerebellar neurons or the hippocampal neurons, to a degree
at least as great as the protease inhibitors (Fig 6). In the cerebellar neurons, the combined application
of the protease inhibitor MDL-28170 with the NO synthase
inhibitor N-Arg at 1 hour afforded additional protection
over that of MDL-28170 alone at 1 hour, resulting in 92±6% relative
survival (P<.05). However, this combination of
antagonists was not significantly more protective than
N-Arg alone, and in the hippocampal neurons the neuroprotective effect
of combined MDL-28170 and N-Arg was similar to that of either agent
alone.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The toxicity induced by kainate in cerebellar neurons and by NMDA in hippocampal neurons is accompanied by the activation of a proteolytic activity consistent with that of calpain, as evidenced by a pattern of proteolytic fragments of spectrin produced similar to that produced by exogenous calpain and by the similar antagonist effects of the protease inhibitors MDL-28170 and E-64. Either form of calpain (calpain I, activated by micromolar [Ca2+]i, or calpain II, activated at millimolar [Ca2+]i) could be involved, since the inhibitors do not discriminate between these two forms.31 32 33 Although the [Ca2+]i as measured in single neurons exposed to NMDA or kainate, which usually only reaches the low micromolar range, might be thought insufficient to activate calpain I, evidence is mounting that the submembrane domain undergoes localized [Ca2+]i increases far exceeding those measured in the bulk cytoplasm and contains localized configurations of enzymes with relatively low affinity for Ca2+ positioned to respond to these microdomains of Ca2+ influx.34 35 Although most of the calpain is distributed in the cytoplasm, a fraction seems to be associated with membranes.13 32 Furthermore, the [Ca2+]i requirement of calpain II is decreased by autolytic activation.36 37 Immunocytochemical studies of the distribution of calpain I suggested that cerebellar Purkinje cells and deep nuclear neurons are richly supplied with calpain I, whereas the levels are low in hippocampal pyramidal cells.38 Nevertheless, Siman and Noszek15 argued that calpain I but not calpain II was specifically involved in the NMDA- or kainate-induced damage in hippocampal neurons, since only calpain I underwent autoproteolysis in their model. In sum, although the possible involvement of calpain II cannot be discounted, calpain I may be the most likely isoform involved in the toxic processes.
The protective effects of these antagonists are not the result of blocking effects at the glutamate receptors themselves since the effects on agonist-evoked currents were insignificant, while the neuroprotective effects were substantial. Also, the antagonists were nearly as effective when applied immediately after the agonist exposure as when applied during the agonist exposure.
It is possible that other proteolytic activities, which may make important contributions to toxicity, are also activated by the Ca2+ influx produced during excitotoxin exposures. It is highly likely that calpain is at least one of the important proteolytic activities leading to cell death after excitotoxin exposure, since the pattern of spectrin breakdown that accompanies the toxicity suggests widespread calpain activation and substantial breakdown of this important structural protein. Furthermore, the time course of the appearance of BDPs and the declining opportunity to protect the neurons with a calpain inhibitor coincide, as expected if this proteolytic process is causative and not merely a side effect of the toxic process. Others have argued for the specific involvement of calpain I in excitotoxic damage on a similar basis.30 However, although E-64 appears to be substantially less effective than MDL-28170 in preventing spectrin breakdown by calpain, its neuroprotective effect in cerebellar neurons is equivalent. The protective effects of E-64 and MDL-28170 in hippocampal neurons were not directly compared, but only MDL-28170 was significantly protective at 1 hour after NMDA exposure. E-64 has inhibitory activity at other cysteine proteases such as the lysosomal proteases cathepsin B and L,23 and perhaps these may also contribute to toxicity. On the other hand, it may be that the difference between the substantial neuroprotective effects of E-64 and its weak effects in blocking calpain-mediated spectrin breakdown relates to the high sensitivity of spectrin to proteolysis by calpain, so that incomplete block of calpain activity by E-64 may allow spectrin hydrolysis but still suppress the proteolysis of other proteins enough to prevent neuronal death. However, other processes, including activation of other proteases, may also be important to neuronal excitotoxicity. Until highly specific inhibitors of calpain are available, we cannot completely exclude the possibility that the neuroprotective effects of calpain inhibitors are mediated by effects on these other processes.
The protective efficacy of late applications of calpain inhibitors after excitotoxic exposures suggests that resulting Ca2+-dependent proteolysis may continue for some time after the agonist exposure. Reasons for the extended activity of calpain probably include (1) prolonged [Ca2+]i elevations after glutamate agonist exposures while cytoplasmic buffering mechanisms slowly bring the [Ca2+]i down to normal levels39 and (2) late elevations of [Ca2+]i that can occur and are linked to toxicity.29 40 Furthermore, the autolysis of calpain, lowering its [Ca2+]i requirement, may result in repeated activation of calpain during subsequent [Ca2+]i fluctuations. In the cerebellar neurons, the non-NMDA antagonist CNQX was also somewhat effective in preventing toxicity when administered at times after the agonist exposure. This indicates that ongoing activity at non-NMDA receptors continued at a level sufficient to contribute to cell death for some time after the kainate was removed. Previous studies have shown that excessive synaptic activity alone is sufficient to cause toxicity,24 41 and CNQX may be protective by blocking such continuing synaptic activity. Although the NMDA receptor antagonist D-AP5 was not found to be significantly protective with late applications, the experimental variability was too great to rule out a small protective effect of delayed NMDA antagonism such as others have found4 or to statistically verify the apparent superiority of the calpain inhibitor MDL-28170 in protecting hippocampal neurons at delayed time points.
We also used the NO synthase inhibitor N-Arg in excitotoxicity assays. We have previously shown that the neuroprotective effect of N-Arg in cerebellar neurons can be reversed by a high concentration of the natural NO synthase substrate L-arginine.20 We have also found that the effect of N-Arg is relatively stereoselective for the L form, although the D enantiomer also has some protective effects at 100 µmol/L (J.R.B. et al, unpublished data, 1995), similar to its partial inhibitory effects reported in endothelium-dependent relaxation of rat aorta.42 Thus, the effects of N-Arg in these neurons are consistent with a competitive inhibitory action at NO synthase. However, as in the case of the calpain inhibitors, we cannot exclude contributions to neuroprotection from other effects of N-Arg. N-Arg was also quite protective at 1 hour after excitotoxin exposure, and the combination of the calpain inhibitor MDL-28170 and the NO synthase inhibitor N-Arg nearly completely rescued the cerebellar neurons at 1 hour. The combination of MDL-28170 and N-Arg, however, had no additive neuroprotective effects in hippocampal neurons. The reason for this possible difference between cell types may relate to the heterogeneous makeup of the cerebellar cultures that consist of Purkinje cells, deep nuclear neurons, and interneurons, whereas the hippocampal cultures are relatively pure populations of pyramidal cells.24 If toxicity is primarily mediated by calpain in some neurons and primarily by NO in other neurons, additive protective effects of antagonist may result in heterogeneous cultures. In the hippocampal cells, synergism of NO production and calpain activation may produce the death of individual neurons, so that blocking either process alone is protective, with little additivity. It is clear that the relevance of calpain activation and of NO production to toxicity is specific to the neuronal cell type, since unlike the neurons studied here, cerebellar granule cells are protected by neither the inhibition of calpain43 nor the inhibition of NO synthase.44
All of these results underscore the fact that the processes leading to the demise of a cell after excitotoxin exposure are not irreversible but are subject to delayed intervention. The specific time course found here in cultured neurons, with effective protection at 1 hour but little or no effect at 4 hours, cannot be expected to correspond directly to a similar duration of reversibility of toxicity in situations in vivo where the influences of local tissues hold sway. Only in vivo studies can determine the parameters of protective effects of these agents in such settings. Recent studies showed that concurrent treatment with MDL-28170 reduced proteolysis and infarct size in a model of stroke in rats with vascular occlusion45 and that another novel calpain inhibitor was effective in reducing infarct size even when administered 1.25 hours after middle cerebral artery occlusion.46
Multiple cellular processes are known to accompany excessive glutamate receptor activation and [Ca2+]i elevation. Besides the two studied here, other processes thought important in mediating neurotoxicity in glutamate exposure or ischemic damage include the activation of lipases,10 lipid peroxidation,47 and enzymatic production of oxidative free radicals.11 Blockade of each of these processes has been shown to be protective of neurons to various degrees. It appears that it is the combination of several processes, acting in synergy, that results in the demise of an individual cell rather than any single process alone, so that the interruption of any one of the toxic processes may often be sufficient to prevent the death of individual cells. Blockade of more than one of the processes may afford more complete protection in certain neurons, especially when treatment is delayed after the initiation of cellular injury.
|
Acknowledgments |
|---|
This study was supported by National Institutes of Health grant NS-01630 and a Grant-in-Aid from the American Heart Association of Metropolitan Chicago (Dr Brorson) and by grants DA-02121 and MH-40165 (Dr Miller). Mr Marcuccilli was supported by MD-PhD training grant HD-07009. We would like to thank Reginaldo Sulit and Patricia Manzolillo for their excellent technical assistance in performing the protein immunoblots and cell toxicity assays.
Received August 31, 1994; revision received February 3, 1995; accepted March 30, 1995.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1. Simon RP, Swan JH, Meldrum BS. Blockade of N-methyl-D-aspartate receptors may protect against ischemic damage in the brain. Science. 1984;226:850-852.
2.
Goldberg MP, Weiss JW, Pham PC, Choi DW.
N-Methyl-D-aspartate receptors mediate hypoxic neuronal injury
in cortical cultures. J Pharmacol Exp
Ther. 1987;243:784-791.
3. Germano IM, Lawrence HP, Meldrum BS, Bartkowski HM, Simon RP. Kynurenate inhibition of cell excitation decreases stroke size and deficits. Ann Neurol. 1987;22:730-734. [Medline] [Order article via Infotrieve]
4. Choi DW, Koh J-y, Peters S. Pharmacology of glutamate neurotoxicity in cortical cell culture: attenuation by NMDA antagonists. J Neurosci. 1988;8:185-196. [Abstract]
5. Kaku DA, Goldberg MP, Choi DW. Antagonism of non-NMDA receptors augments the neuroprotective effect of NMDA receptor blockage in cortical cultures subjected to prolonged deprivation of oxygen and ischemia. Brain Res. 1991;554:344-347. [Medline] [Order article via Infotrieve]
6. Park CK, Nehls DG, Graham DI, Teasdale GM, McCulloch J. The glutamate antagonist MK-801 reduces focal ischemic brain damage in the rat. Ann Neurol. 1988;24:543-551. [Medline] [Order article via Infotrieve]
7.
Sheardown MJ, Nielsen EØ, Hansen AJ, Jacobsen P,
Honore T. 2,3-Dihydroxy-6-nitro-7-sulfamoyl-benzo(F)quinoxaline:
a neuroprotectant for cerebral ischemia.
Science. 1990;247:571-574.
8.
Smith SE, Meldrum BS. Cerebroprotective effect
of a non–N-methyl-D-aspartate
antagonist, GYKI 52466, after focal ischemia in the
rat. Stroke. 1992;23:861-864.
9. Choi DW. Ionic dependence of glutamate neurotoxicity. J Neurosci. 1987;7:369-379.[Abstract]
10.
Umemura A, Mabe H, Nagai H. A phospholipase C
inhibitor ameliorates postischemic neuronal
damage in rats. Stroke. 1992;23:1163-1166.
11.
Coyle JT, Puttfarcken P. Oxidative stress,
glutamate, and neurodegenerative disorders. Science. 1993;262:689-695.
12.
Dawson VL, Dawson TM, London ED, Bredt DS, Snyder SH.
Nitric oxide mediates glutamate neurotoxicity in primary
cortical cultures. Proc Natl Acad Sci U S A. 1991;88:6368-6371.
13. Johnson P. Calpains (intracellular calcium-activated cysteine proteinases): structure-activity relationships and involvement in normal and abnormal cellular metabolism. Int J Biochem. 1990;22:811-822. [Medline] [Order article via Infotrieve]
14.
Emori Y, Kawasaki H, Sugihara H, Imajoh S, Kawashima S,
Suzuki K. Isolation and sequence analyses of cDNA clones
for the large subunits of two isozymes of rabbit calcium-dependent
protease. J Biol Chem. 1986;261:9465-9471.
15. Siman R, Noszek JC. Excitatory amino acids activate calpain I and induce structural protein breakdown in vivo. Neuron. 1988;1:279-287. [Medline] [Order article via Infotrieve]
16. Arai A, Vanderklish P, Kessler M, Lee K, Lynch G. A brief period of hypoxia causes proteolysis of cytoskeletal proteins in hippocampal slices. Brain Res. 1991;555:276-280. [Medline] [Order article via Infotrieve]
17. Arlinghaus L, Mehdi S, Lee KS. Improved posthypoxic recovery with a membrane-permeable calpain inhibitor. Eur J Pharmacol. 1991;209:123-125. [Medline] [Order article via Infotrieve]
18.
Lee KS, Frank S, Vanderklish P, Arai A, Lynch G.
Inhibition of proteolysis protects hippocampal neurons from
ischemia. Proc Natl Acad Sci U S A. 1991;88:7233-7237.
19. Rami A, Krieglstein J. Protective effects of calpain inhibitors against neuronal damage caused by cytotoxic hypoxia in vitro and ischemia in vivo. Brain Res. 1993;609:67-70. [Medline] [Order article via Infotrieve]
20. Brorson JR, Manzolillo PA, Miller RJ. Ca2+ entry via AMPA/KA receptors and excitotoxicity in cultured cerebellar Purkinje cells. J Neurosci. 1994;14:187-197. [Abstract]
21. Caner H, Collins JL, Harris SM, Kassell NF, Lee KS. Attenuation of AMPA-induced neurotoxicity by a calpain inhibitor. Brain Res. 1993;607:354-356. [Medline] [Order article via Infotrieve]
22. Komatsu K, Inazuki K, Hosoya J, Satoh S. Beneficial effect of new thiol protease inhibitors, epoxide derivatives, on dystrophic mice. Exp Neurol. 1986;91:23-29. [Medline] [Order article via Infotrieve]
23. Mehdi S. Cell-penetrating inhibitors of calpain. Trends Biochem Sci. 1991;16:150-153. [Medline] [Order article via Infotrieve]
24. Abele AE, Scholz KP, Scholz WK, Miller RJ. Excitotoxicity induced by enhanced excitatory neurotransmission in cultured hippocampal pyramidal neurons. Neuron. 1990;4:413-419. [Medline] [Order article via Infotrieve]
25. Brorson JR, Bleakman D, Gibbons SJ, Miller RJ. The properties of intracellular calcium stores in cultured rat cerebellar neurons. J Neurosci. 1991;11:4024-4043. [Abstract]
26. Jones KH, Senft JA. An improved method to determine cell viability by simultaneous staining with fluorescein diacetate-propidium iodide. J Histochem Cytochem. 1985;33:77-79. [Abstract]
27.
Brorson JR, Bleakman D, Chard PS, Miller RJ.
Calcium directly permeates
kainate/-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid
receptors in cultured cerebellar Purkinje neurons. Mol
Pharmacol. 1992;41:603-608. [Abstract]
28.
Tang C-M, Dichter M, Morad M. Quisqualate
activates a rapidly inactivating high conductance ionic channel in
hippocampal neurons. Science. 1989;243:1474-1476.
29. Randall RD, Thayer SA. Glutamate-induced calcium transient triggers delayed calcium overload and neurotoxicity in rat hippocampal neurons. J Neurosci. 1992;12:1882-1895. [Abstract]
30. Siman R, Noszek JC, Kegerise C. Calpain I activation is specifically related to excitatory amino acid induction of hippocampal damage. J Neurosci. 1989;9:1579-1590. [Abstract]
31. Mellgren RL. Canine cardiac calcium-dependent proteases: resolution of two forms with different requirements for calcium. FEBS Lett. 1980;109:129-133. [Medline] [Order article via Infotrieve]
32. Siman R, Baudry M, Lynch G. Purification from synaptosomal plasma membranes of calpain I, a thiol protease activated by micromolar calcium concentrations. J Neurochem. 1983;41:950-956. [Medline] [Order article via Infotrieve]
33.
Saito K-I, Nixon RA. Specificity of
calcium-activated neutral proteinase (CANP) inhibitors for
human 
CANP and mCANP. Neurochem Res. 1993;18:231-233. [Medline]
[Order article via Infotrieve]
34.
Llinás R, Sugimori M, Silver RB.
Microdomains of high calcium concentration in a presynaptic
terminal. Science. 1992;256:677-679.
35. Stanley EF. Single calcium channels and acetylcholine release at a presynaptic nerve terminal. Neuron. 1993;11:1007-1011. [Medline] [Order article via Infotrieve]
36.
Coolican SA, Haiech J, Hathaway DR. The role of
subunit autolysis in activation of smooth muscle
Ca2+-dependent proteases. J
Biol Chem. 1986;261:4170-4176.
37.
DeMartino GN, Huff CA, Croall DE.
Autoproteolysis of the small subunit of calcium-dependent
protease II activates and regulates protease activity.
J Biol Chem. 1986;261:12047-12052.
38. Siman R, Gall C, Perlmutter LA, Christian C, Baudry M, Lynch G. Distribution of calpain I, an enzyme associated with degenerative activity, in rat brain. Brain Res. 1985;347:399-403. [Medline] [Order article via Infotrieve]
39.
Thayer SA, Miller RJ. Regulation of the
intracellular free calcium concentration in single rat dorsal root
ganglion neurones in vitro. J
Physiol (Lond). 1990;425:85-115.
40. Tymianski M, Wallace MC, Spigelman I, Uno M, Carlen PL, Tator CH, Charlton MP. Cell-permeant Ca2+ chelators reduce early excitotoxic and ischemic neuronal injury in vitro and in vivo. Neuron. 1993;11:221-235. [Medline] [Order article via Infotrieve]
41.
Rothman SM. Synaptic activity mediates death of
hypoxic neurons. Science. 1983;220:536-537.
42.
Wang Y-X, Poon CI, Pang CCY. In vitro
and ex vivo inhibitory effects of L- and
D-enantiomers of NG-nitro-arginine on
endothelium-dependent relaxation of rat aorta.
J Pharmacol Exp Ther. 1993;265:112-119.
43. Manev H, Favaron M, Siman R, Guidotti A, Costa E. Glutamate neurotoxicity is independent of calpain I inhibition in primary cultures of cerebellar granule cells. J Neurochem. 1991;57:1288-1295. [Medline] [Order article via Infotrieve]
44. Puttfarcken PS, Lyons WE, Coyle JT. Dissociation of nitric oxide generation and kainate-mediated neuronal degeneration in primary cultures of rat cerebellar granule cells. Neuropharmacology. 1992;31:565-575. [Medline] [Order article via Infotrieve]
45. Hong S-C, Goto Y, Lanzino G, Soleau S, Kassell NF, Lee KS. Neuroprotection with a calpain inhibitor in a model of focal cerebral ischemia. Stroke. 1994;25:663-669. [Abstract]
46. Bartus RT, Hayward NJ, Elliott PJ, Sawyer SD, Baker KL, Dean RL, Akiyama A, Straub JA, Harbeson SL, Li Z, Powers J. Calpain inhibitor AK295 protects neurons from focal brain ischemia: effects of postocclusion intra-arterial administration. Stroke. 1994;25:2265-2270. [Abstract]
47. Hall ED. Novel inhibitors of iron-dependent lipid peroxidation for neurodegenerative disorders. Ann Neurol. 1992;32:S137-S142.
This article has been cited by other articles:
 |
|
 |
|
  M. Zhou and M. Baudry Developmental changes in NMDA neurotoxicity reflect developmental changes in subunit composition of NMDA receptors. J. Neurosci., March 15, 2006; 26(11): 2956 - 2963. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. R. Lynch and R. P. Guttmann Excitotoxicity: Perspectives Based on N-Methyl-D-Aspartate Receptor Subtypes J. Pharmacol. Exp. Ther., March 1, 2002; 300(3): 717 - 723. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. W. R. Shuttleworth and J. A. Connor Strain-Dependent Differences in Calcium Signaling Predict Excitotoxicity in Murine Hippocampal Neurons J. Neurosci., June 15, 2001; 21(12): 4225 - 4236. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Tremblay, B. Chakravarthy, K. Hewitt, J. Tauskela, P. Morley, T. Atkinson, and J. P. Durkin Transient NMDA Receptor Inactivation Provides Long-Term Protection to Cultured Cortical Neurons from a Variety of Death Signals J. Neurosci., October 1, 2000; 20(19): 7183 - 7192. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Lipton Ischemic Cell Death in Brain Neurons Physiol Rev, October 1, 1999; 79(4): 1431 - 1568. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. T. Faddis, M. J. Hasbani, and M. P. Goldberg Calpain Activation Contributes to Dendritic Remodeling after Brief Excitotoxic Injury In Vitro J. Neurosci., February 1, 1997; 17(3): 951 - 959. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||