B R A I N R E S E A R C H 1 1 9 6 ( 2 0 0 8 ) 1 2 1 – 1 3 0

ava i l ab l e a t www.sc i enced i rec t . com

www.e l sev i e r. com/ l oca te /b ra in res

Research Report

Contribution of calpain activation to early stages ofhippocampal damage during oxygen–glucose deprivation

Michael Grammer⁎, Dong Li, Nirubhana Arunthavasothy, Janusz LipskiDepartment of Physiology, Faculty of Medical and Health Sciences, University of Auckland, 85 Park Road, Grafton, Auckland 1023, New Zealand

A R T I C L E I N F O

⁎ Corresponding author. Fax: +64 9 373 7499.E-mail address: m.grammer@auckland.acAbbreviations: CsA, cyclosporin A; HSD,

oxygen–glucose deprivation; Rh 123, Rhodmembrane potential

0006-8993/$ – see front matter © 2007 Elsevidoi:10.1016/j.brainres.2007.12.006

A B S T R A C T

Article history:Accepted 1 December 2007Availabe online 8 December 2007

CalpainsareCa2+-activatedenzymeswhich cleavecytoskeletal andother proteins, contributingto neuronal damage in conditions of pathological intracellular Ca2+ elevation, including stroke.However, the consequences of calpain overactivation have typically been observed hours afterinsult. To identify the earliest events attributable to calpain activation, and thus potentiallyisolate calpain substrates involved in acute neuronal damage, we dynamically recorded theeffects of calpain inhibition inan invitromodelof stroke. ExtracellularDCpotentials and fEPSPswere monitored together with changes of light transmittance (as a measure of cell andmitochondrial swelling) and Rh 123 fluorescence (to monitor mitochondrial membranepotential; ΔΨm) in hippocampal slices obtained from P12–P17 rats. No differences wereobserved in the latencies of fEPSP disruption or onset of extracellular DC shifts associatedwith hypoxic spreading depression (HSD) evoked by oxygen–glucose deprivation (OGD) undercontrol conditions or in the presence of calpain inhibitor III (MDL 28170). However, a significantdifference was observed in transmitted light signals during OGDwith calpain inhibition. Giventhe potential contribution of mitochondrial swelling to changes in light transmittance, theseexperiments were also conducted in the presence of cyclosporin A to block opening ofthe mitochondrial permeability transition pore (MPTP). Our results indicate that differences inOGD-induced changes of light transmittance in the presence of MDL 28170 are not likelythe result of MPTP blockade or changes in dendritic beading. We propose that calpaininhibition may alter changes in light transmittance by limiting conformational changes ofmitochondria.

© 2007 Elsevier B.V. All rights reserved.

Keywords:Cell and mitochondrial swellingMitochondrial membrane potentialLight transmittanceIschemia and reperfusionBrain sliceCyclosporin A

1. Introduction

During ischemia/reperfusion, neurons undergo a series ofphysiological events leading to cell death or damage. Thisneurological injury is frequently modelled in vitro by oxygen–glucose deprivation (OGD) (Lipski et al., 2006; Somjen, 2001). Inthe absence of oxygen and glucose, ATP levels are quickly

.nz (M. Grammer).hypoxic spreading depr

amine 123; ROI, region

er B.V. All rights reserved

depleted and cellular ionic gradients are lost. The equilibra-tion of ions across the plasma membrane leads to neuronaldamage through elevated intracellular calcium concentrationas well as excessive cell and mitochondrial swelling. Para-doxically, restoring oxygen and glucose to hypoxic tissue (acondition mimicking reperfusion) initiates additional neuro-toxic responses. This ‘reperfusion injury’ is largely attributed

ession; MPTP, mitochondria permeability transition pore; OGD,of interest; ROS, reactive oxygen species; ΔΨm, mitochondrial

.

Fig. 1 – Electrophysiological responses to OGD and calpaininhibition. A: Examples of fEPSP recordings before (A1) andafter (A2) 3min of OGD. Each trace represents the average of 6sweeps. B: Extracellularly recorded DC shift observed 6 minafter the onset of OGD (double-headed arrow). C: Times(mean±S.E.M.) at which the loss of fEPSPs and extracellularDC shifts occurred during OGD under control conditions or inthe presence of MDL 28170 (40 μM).

122 B R A I N R E S E A R C H 1 1 9 6 ( 2 0 0 8 ) 1 2 1 – 1 3 0

to a burst of reactive oxygen species (ROS) generated bymitochondria and/or cytosolic oxidases that damages mem-brane lipids as well as cellular proteins and nucleic acids(Abramov et al., 2007; Nicholls and Budd, 2000). ROS produc-tion during reperfusion is further stimulated by elevatedintracellular calcium levels (Nicholls and Budd, 2000).

Excessive intracellular calcium during ischemia/reperfu-sion or OGD damages neurons through multiple pathways,among which activation of calpains appears to play a majorrole. Calpains comprise a family of 14 calcium-activatedcysteine proteases that, upon activation, trigger substrate-specific proteolysis resulting in degradation, activation ortranslocation of the substrate (Carragher, 2006; Goll et al.,2003). Following brief ischemic episodes, calpain activation(detected as the breakdown product of a preferred substrate,spectrin) was reported in CA1 pyramidal neurons (Seubertet al., 1989). Following longer ischemic events, calpainactivation was observed in other, progressively vulnerablebrain regions (Roberts-Lewis et al., 1994). Neuroprotectionfrom ischemia/reperfusion by calcium-limiting agents, suchas the NMDA channel blocker MK-801, correlates with de-creased calpain activation (Roberts-Lewis and Siman, 1993;Seubert et al., 1990; Zhou and Baudry, 2006). In addition,calpain inhibitors provide neuroprotection from ischemiaboth in vitro (Malagelada et al., 2005; Newcomb-Fernandezet al., 2001) and in vivo (Hong et al., 1994; Kawamura et al.,2005; Lee et al., 1991).

While inhibition of calpains clearly provides neuroprotec-tion from ischemia/reperfusion, as well as other pathologicalconditions associated with calcium dysregulation, the specificsubstrates of calpain-mediated proteolysis responsible forneuronal damage under these conditions have not yet beenidentified. There are currently over 100 calpain substrates in-cluding ion channels/pumps, membrane anchoring proteins,enzymes and a large number of cytoskeletal/membraneassociated proteins. Given the involvement of cell and mito-chondrial swelling during the early stages of ischemia(Christophe and Nicolas, 2006; Kimelberg, 1995), and recentevidence that calpains may target mitochondria duringischemia (Chen et al., 2002), we hypothesized that calpaininhibition might decrease or slow the onset of cellular ormitochondrial swelling, thus implicating cytoskeletal and/ormitochondrial associated calpain substrates in early ischemicdamage. The identification of synaptic and membrane pro-teins that support synaptic transmission and cell membranepotential as calpain substrates (Khoutorsky and Spira, 2005;Vanderklish et al., 1995) also suggests that calpain activationmight influence electrophysiological responses during ische-mia and reperfusion.

To resolve the potential contribution of calpain activationduring early stages of cellular damage caused by ischemia/reperfusion, electrophysiological as well as intrinsic andfluorescent optical signals in the CA1 region of the hippo-campus were dynamically recorded during OGD and reperfu-sion. Cellular and mitochondrial swelling associated withOGD/reperfusion alters the optical properties of neural tissuewhich were recorded using conventional light microscopy. Inaddition, we simultaneously monitored changes in mitochon-drial membrane potential (ΔΨm) with the fluorescent indicatorRhodamine 123 (Rh 123).

2. Results

2.1. Effects of calpain inhibition duringOGDand reperfusion

One of the earliest events in ischemia, or during OGD regardedas a model of ischemic damage, is the disruption of synaptictransmission (Lipski et al., 2006; Schiff and Somjen, 1987). Inagreement with this finding, we observed rapid decay ofsynaptic transmission between Schaeffer collaterals andpyramidal CA1 neurons, measured as the loss of fEPSPs instratum (Str) radiatumwithin 2.5 min of OGD (Fig. 1A). Calpaininhibitor III (MDL 28170; 40 μM) did not alter the latency tosynaptic disruption (Fig. 1C; p=0.46), which correspondswith previous reports of adenosine-mediated inhibitionas the principle mechanism of synapse blockade during OGD(Abbracchio and Cattabeni, 1999).

Another event during ischemia or OGD is hypoxic spread-ing depression (HSD) associated with a large redistribution of

123B R A I N R E S E A R C H 1 1 9 6 ( 2 0 0 8 ) 1 2 1 – 1 3 0

ions between the extracellular and intracellular spaces and asudden loss of cell membrane potential which can be recordedextracellularly as a negative DC potential shift (Lipski et al.,2006; Somjen, 2001). Recently, the onset of HSD during hypoxiawas shown to be accelerated by mitochondrial uncouplingwith carbonyl cyanide p-trifluromethoxy-phenylhydrazone(Gerich et al., 2006). Conditions or agents that preserve ΔΨm

might therefore be expected to delay HSD onset. Under controlconditions, the DC shift occurred approximately 6 min afterOGD onset (Fig. 1B). Calpain inhibition with MDL 28170 had nosignificant effect on the latency to this event (Fig. 1C; p=0.99).

Ischemia and OGD are also associated with substantial celland mitochondrial swelling which alter optical properties ofbrain tissue and can be monitored by recording changes inlight transmittance (Aitken et al., 1999; Bahar et al., 2000;Kreisman and LaManna, 1999; Somjen, 2001). Fig. 2 shows atypical example of relative changes in light transmittancerecorded under our experimental conditions in Str. pyrami-dale and Str. radiatum of the CA1 hippocampal region. Soon

Fig. 2 – Changes in light transmittance in hippocampal CA1subfields during OGD and reperfusion. Panels A–C show ROIsin Str. pyramidale and Str. radiatum (black and grey ROIsrespectively; 5× objective). D: Time course of OGD-evokedchanged in light transmittance. Changes in light transmittanceduring OGD are attributed to two processes. Swelling of cellbodies and large processes during OGD results in increasedlight transmittance which peaks with spreading depression(B). Following the transmittance peak, and still under OGDconditions, tissue darkening occurs (oblique arrow) which hasbeen attributed to swelling of mitochondria and/or dendriticbeading. C: Upon reperfusion, hippocampal tissue undergoesfurther darkening which may represent additional,irreversible damage due to lipid peroxidation. Calibration barin A–C: 200 μm.

Fig. 3 – Effects of calpain inhibition on transmitted lightduring OGD and reperfusion. With calpain inhibitor III (MDL28170), peak light transmittance in response to OGD wasblunted compared to controls. Calpain inhibition eliminatedtissue darkening during OGD following spreading depressionin Str. pyramidale (A) and significantly decreased darkeningin Str. radiatum (B). Individual traces show mean±S.E.M.(*p<0.05).

after OGD onset, light transmittance in both regions increasedfirst slowly then rapidly, with the fast phase associated withthe extracellular DC shift (not shown) and onset of HSD.Increased light transmittance in this preparation is associatedwith cell swelling (Kreisman et al., 1995; Lipski et al., 2006). Therapid increase in light transmittance peaked and thendeclined gradually for the remainder of the OGD period. Thedecline in light transmittance observed during OGD (followingthe peak associated with HSD onset) is not fully understood,but has been attributed tomitochondrial swelling (Bahar et al.,2000; Muller and Somjen, 1999) and/or dendritic beading(Andrew et al., 1999; Fayuk et al., 2002). The light transmit-tance decrease duringOGDwas consistentlymore pronouncedin Str. radiatum. Upon reperfusion, light transmittance furtherdecreased in both examined regions. Light transmittancedecrease upon reperfusion may represent additional tissuedamage such as lipid peroxidation associated with rapid ROSproduction (Hochgraf et al., 1997; Imai et al., 2000) as is knownto occur during reperfusion (Abramov et al., 2007), however,additional experiments are necessary to explain this phenom-enon.We recently found that tissue swelling doesnot decreaseduring the early stages of reperfusion (Lipski et al., 2006),eliminating this as a possible explanation for the decreasedlight transmittance.

To examine a possible contribution of calpain activity tocell or mitochondrial swelling, we measured light transmit-tance during OGD experiments with calpain inhibition. There

124 B R A I N R E S E A R C H 1 1 9 6 ( 2 0 0 8 ) 1 2 1 – 1 3 0

were two notable differences in light transmittance attribu-table to calpain activity. In the presence of MDL 28170 (40 μM),light transmittance increased and peaked, but, compared tocontrols, peak transmittance tended to be diminished in Str.pyramidale and Str. radiatum (Fig. 3). Throughout the re-mainder of OGD, light transmittance declined under controlconditions, but, in contrast to controls, light transmittanceremained stable in Str. pyramidale and decreased significantlyless than controls in Str. radiatum with calpain inhibition. Inboth conditions, light transmittance decreased sharply uponreperfusion. Analysis of a specific calpain effect upon reperfu-sion is obscured by calpain's affect on the preceding condition,however, equalizing light transmittance at the onset ofreperfusion reveals no difference in the magnitude or timecourse of change during reperfusion.

The relationship between mitochondrial swelling and mito-chondrial depolarization is not well established, although thetwo events are typically correlated in intact cells (Kaasik et al.,2007).Todetermine if thesmallerdecrease in light transmittanceobservedduringOGD in thepresenceof the calpain inhibitorwasassociated with a reduction inmitochondrial depolarization, weconducted OGD experiments in the presence of the mitochon-drial membrane potential (ΔΨm) indicator, Rh 123.

As an initial verification of Rh 123 fluorescent signal inour submerged preparation, we measured transmitted lightand Rh 123 fluorescence in slices that underwent OGD twice(Fig. 4). During the first OGD episode, mitochondrial depolar-

Fig. 4 – Light transmittance (A) and Rh 123 signals (B) in responA: During the first OGD episode, light transmittance followed therecording of Rh 123 signal revealed an initial signal increase witpeaked in concert with the peak in light transmittance. After peaUpon reperfusion, Rh 123 signal showed a rapid phase of recovethroughout reperfusion. At the second episode of OGD, light transame transmittance value observed near the end of the initial OGand plateaued, remaining steady throughout the episode of OGDIndividual traces show mean±S.E.M.

ization tracked closelywith changes in transmitted light. Uponreperfusion, Rh 123 signal decreased suddenly in parallel witha decrease in light transmittance. During the second OGDepisode, Rh 123 fluorescence again increased simultaneouslywith transmitted light as in the first OGD episode, however,both signals increased and plateaued rather than peaking. Theincrease of Rh 123 fluorescence during the secondOGDepisodeargues against loss of Rh 123 signal due to leakage from cellsduringmitochondrial depolarization or due to photobleaching.Thus, the decrease in Rh 123 fluorescence during reperfusion isbest explained by mitochondrial repolarization, subsequentre-sequestration and self-quenching of Rh 123 (Gerich et al.,2006). The relationship between mitochondrial depolarizationand Rh 123 fluorescence was further verified by our observa-tion that the addition of the mitochondrial uncoupler FCCP(1μM),whichabolishes themitochondrialmembranepotential(Gerich et al., 2006), strongly increased the Rh 123 signal undernormoxic conditions (n=3; data not shown).

In the presence of the calpain inhibitor MDL 28170, latencyto onset and rate of mitochondrial depolarization were notdifferent from controls in Str. pyramidale or Str. radiatum(Fig. 5). However, MDL 28170 tended to diminish the peakamplitude of the Rh 123 signal, suggesting reduced depolar-ization of mitochondrial membrane potential evoked by OGD.During the remainder of the OGD period and subsequentreperfusion, the fluorescent signal was not different betweenMDL 28170 and control conditions.

se to repeated episodes of OGD in Str. pyramidale (n=4).previously described response (see Fig. 3A). B: Simultaneoush the onset of OGD, followed by a more rapid increase thatking, the Rh 123 signal declined for the duration of OGD.ry to baseline followed by a slower, continuous declinesmittance increased slowly and eventually plateaued at theD episode. At the same time, Rh 123 signal increased quickly, then decayed during the second reperfusion episode.

Fig. 5 – The effect of calpain inhibition onRh123 fluorescenceduring OGD and reperfusion. Calpain inhibition (MDL 28170,40μM) showed a trend towards a smaller initial peak of the Rh123 signal during OGD in both Str. pyramidale (A; p=0.08)and Str. radiatum (B; p=0.06). Note the identical rise time of Rh123 in both conditions, in contrast to responses observed inthe presence of CsA (Fig. 6).

Fig. 7 – Effect of CsA on electrophysiological responses toOGD. Times (mean±S.E.M.) at which loss of fEPSPs andextracellular DC shifts occurred during OGD under controlconditions and in the presence of CsA (10 μM).

125B R A I N R E S E A R C H 1 1 9 6 ( 2 0 0 8 ) 1 2 1 – 1 3 0

2.2. Effects of cyclosporin A (CsA) during OGDand reperfusion

To examine whether the smaller reduction in light transmit-tance we observed during OGD in the presence of calpaininhibition (Fig. 3) might be due to reduced mitochondrialswelling, we examined changes of light transmittance in the

Fig. 6 – Effects of CsA on transmitted light and ΔΨm during OGDtransmittance (A, B) or significantly effect ΔΨm (C, D) during the c

presence of CsA. CsA reduces mitochondrial swelling follow-ing acute hypoglycaemia in vivo (Zieminska et al., 2006), andreduces swelling of isolated mitochondria in response toexcessive calcium load (Friberg et al., 1998). In the presence ofCsA (10 μM), the OGD-induced changes in light transmittancewere not different from those observed under control condi-tions in Str. pyramidale or Str. radiatum, as previouslyreported at lower concentration (5 μM) (Muller and Somjen,1999) (Figs. 6A, B).

We also tested CsA's ability to modify OGD-induced mito-chondrial depolarization and examined its effect on the laten-cies to synaptic disruption and extracellular DC shifts. In thepresence of CsA, no significant difference in Rh 123 fluorescencewas observed during OGD or reperfusion (Figs. 6C, D). Likewise,there were no differences in the latencies to onset of extra-cellular DC shifts (p=0.54) or disruption of synaptic transmission(p=0.75) between control and CsA conditions.

and reperfusion. CsA (10 μM) did not alter changes in lightourse of OGD and reperfusion.

126 B R A I N R E S E A R C H 1 1 9 6 ( 2 0 0 8 ) 1 2 1 – 1 3 0

2.3. Effect of CsA on hypo-osmotic swelling

To further investigate the relationship between light trans-mittance and mitochondrial swelling, we measured changesin light transmittance evoked by hypo-osmotic challenge inthe presence of CsA (10 μM). CsA significantly reduced theamplitude of light transmittance in response to moderate(230 mosM) hypo-osmotic challenge in Str. radiatum andtended to decrease the amplitude of light transmittance inresponse to hypo-osmotic challenge in Str. pyramidale(p=0.12; Fig. 8). Hypo-osmotic challenge did not decreaseΔΨm (as measured with Rh 123 fluorescence; data not shown),suggesting a dissociation betweenmitochondrial swelling andΔΨm under these conditions (c.f. Safiulina et al., 2006) (Fig. 7).

3. Discussion

Although the involvement of calpain activation in cell deathassociated with ischemia/reperfusion was first reported over15 years ago (Seubert et al., 1989), the specific downstreamtargets responsible for neuronal damage remain largelyunknown. Identifying the calpain substrates responsible forcell damage is necessary for the development of specificinhibitors of neural damage and to potentially extend thewindow of effective therapeutic treatments. Most previousstudies of calpain activation following ischemia/reperfusionor OGD examined cells or tissue at delayed time points beyondwhich many downstream substrates of calpain may havealready been cleaved. In this study physiological responses inhippocampal tissue were monitored in real time during andimmediately after OGD to investigate the earliest calpain-mediated events associated with neuronal damage.

3.1. Electrophysiology

Calpain inhibition had no effect on loss of synaptic functionduring OGD. Calpain activity has been implicated in normalsynaptic function both pre- and post-synaptically (Khoutorskyand Spira, 2005; Vanderklish et al., 1995). Due to the rapidonset of adenosine-mediated synaptic inhibition duringischemic conditions however (Abbracchio and Cattabeni,1999), any effect of pathological calpain activation related toOGD-induced intracellular calcium increases may likely beobscured. In submerged hippocampal slices loss of fEPSPs isirreversible after HSD, which does not permit examination ofcalpain's effect on recovery of synapse function.

The mechanism of HSD and the accompanying extracel-lular DC shift observed during OGD is not fully understood butATP depletion, inhibition of Na+/K+-ATPase activity andrelease of intracellular K+ are contributing factors (Allenet al., 2005; Somjen, 2001). The depolarization of cell mem-brane potential associated with HSD can cause release ofglutamate from neurons (Rossi et al., 2000) and glia (Seki et al.,1999) resulting in excitotoxicity (Choi and Rothman, 1990).Since intracellular calcium begins to increase early duringOGD, before HSD onset (Lipski et al., 2006), we hypothesizedthat calpain inhibitionmight stabilize the cellmembrane, thusdelaying the ionic redistribution responsible for cellulardepolarization associated with HSD. Contrary to this view,

there was no difference in latency to the OGD-inducedextracellular DC shifts with calpain inhibition, suggestingeither that calpains are not activated under these conditionsor that calpain activity does not contribute to the onset of HSD.

3.2. Calpains effects on cell swelling

Cell swelling is a major component of brain damage duringischemia, contributing to ion channel gating, gene transcrip-tion, and activation of signalling pathways (Hoffmann andSimonsen, 1989). Swelling is associated with the unfolding ofmembrane invaginations which are otherwise stabilized bycytoskeletal proteins such as spectrin II, a preferred calpainsubstrate, and F-actin (Okada, 1997; Sun and Levitan, 2003).Intracellular dispersion of spectrin II (fodrin) is one of theearliest events of cell swelling under hypo-osmotic stress,corresponding closely with increased cell volume (Sun andLevitan, 2003). With continued hypo-osmotic stress, regulateddecrease of cell volume correlates with the reorganization ofspectrin II into its original, pre-swelling submembrane dis-tribution (Sun and Levitan, 2003).

We hypothesized that inhibiting calpain-mediated proteo-lysis of spectrin II and related cytoskeletal proteins mightdelay or limit the early stages of cell swelling in response toOGD. While there was no difference in the latency to peakswelling during OGD as assessed by changes in transmittedlight, the light transmittance peak tended to be lower withcalpain inhibition. Since transmitted light signals have twoopposing components (Fayuk et al., 2002) (see also Fig. 2legend), it is not possible to unequivocally conclude thatcalpain inhibition limited cell swelling. However, it is unlikelythat calpain inhibition increased or accelerated the ‘darken-ing’ component of the transmitted light signal (typicallyattributed to mitochondrial swelling), which would alsodecrease the light transmittance peak.

The possible effect of calpain inhibition on dendriticbeading under these conditions is not clear. Reports haveshown that calpain activity is necessary for the recovery ofdendritic structure following insult resulting in beading(Adamec et al., 1998; Faddis et al., 1997), but have also shownthat calpain is involved in loss of MAP2, which is normallylocalized to dendrites, following OGD (Buddle et al., 2003).Recent evidence suggests that calpain activity is not asso-ciated with the formation of dendritic beads, rather that asbeading occurs MAP2 becomes redistributed, which is thenfollowed by calpain-mediated proteolysis and loss of MAP2(Hoskison and Shuttleworth, 2006). Based on these reports, itdoes not seem likely that the effects of calpain inhibition ontransmitted light during OGD are due to reduced dendriticbeading.

3.3. Calpain and mitochondrial swelling

In the presence of MDL 28170, OGD resulted in significantlyless darkening of hippocampal tissue observed after the rapidincrease in light transmittance associated with HSD. Tissuedarkening after HSD has been proposed to be partly the resultof mitochondrial swelling associated with high intracellularcalcium and strongly depolarized mitochondrial membranepotential (Aitken et al., 1999; Muller and Somjen, 1999). Until

127B R A I N R E S E A R C H 1 1 9 6 ( 2 0 0 8 ) 1 2 1 – 1 3 0

recently, this interpretation was problematic because swellingof isolated mitochondria results in increased, rather thandecreased, light transmission (Kristian et al., 2000; Raaflaub,1953). However, it has now been shown that under anoxicconditions brain mitochondria swell, but the overall result is adecrease in transmitted light due to shrinkage/condensationof the mitochondrial matrix (Fujii et al., 2004).

Although reversible after brief periods, matrix shrinkageimpairs mitochondrial ATP synthesis (Korge et al., 2005) andmay precede ischemia/reperfusion-induced mitochondrialdamage. In contrast, conditions that result in mitochondrialswelling without shrinkage of the matrix (e.g. hypo-osmolar-ity) decrease light scattering and thereby increase transmittedlight (see Fig. 8). The mechanism(s) of matrix condensationduring anoxia are not known, however subfractionation ofmitochondria revealed that calpain is localized to the mito-chondrial matrix (Arrington et al., 2006). Analysis of matrixcondensation during anoxia in calcium-free conditions mightthus help to elucidate calpain's role in this form of mitochon-drial damage.

Mitochondrial swelling, as occurs during ischemic events inneural tissue, has been linked to opening of themitochondrialpermeability transition pore (MPTP) both in vivo (Friberg et al.,1998) and in vitro (Friberg et al., 1999; Zieminska et al., 2006).MPTP opening under conditions of ischemia/reperfusion(elevated intracellular Ca++, high levels of ROS, ATP depletion)initiates mitochondrial swelling and damage through loss ofmitochondrial ionic gradients and the release of moleculessuch as cytochrome c and apoptosis-inducing factor (AIF) (Liu

Fig. 8 – Changes in light transmittance during mild(260 mosM) and moderate (230 mosM) hypo-osmotic swellingin the presence of CsA. Swelling induced by hypo-osmoticstress corresponded with increased light transmittance. CsA(10 μM) limited the light transmittance increase, suggestingthatmitochondrial swellingandopeningof theMPTP increaseslight transmittance under these conditions (*p<0.05).

et al., 1996; Susin et al., 1996). Importantly, MPTP opening withsubsequent mitochondrial swelling and AIF release has alsobeen attributed to calpain activity (Arrington et al., 2006; Goreset al., 1998). If the decrease in tissue darkening observedduringcalpain inhibition was due to reducedmitochondrial swelling,wewould expect a similar result from other agents that inhibitmitochondrial swelling, such as CsA.

3.4. CsA effects on osmotically and OGD-induced cell andmitochondrial swelling

The ability of CsA to inhibit MPTP opening and limit mito-chondrial swelling in brain mitochondria during ischemic/hypoxic events in vitro depends on the degree of ischemicchallenge (Brustovetsky and Dubinsky, 2000; Hansson et al.,2003). In contrast to calpain inhibition, we observed no effectof CsA on OGD-induced changes in light transmittance.However, CsA reduced changes in light transmittance inresponse to hypo-osmotic stress, a condition known to inducemitochondrial swelling and increase light transmittance inintact cells (Devin et al., 1998), indicating effective drug de-livery in our preparation. While CsA appears to have limitedmitochondrial swelling during hypo-osmotic stress, it ispossible that, in spite of its delivery to mitochondria, themuch stronger insult associated with OGD overwhelmed CsA-mediated blockade of MPTP opening. This interpretation isconsistent with the findings of Brustovetsky and Dubinsky(2000) who reported that CsA blocked MPTP permeability ofbrain mitochondria under mildly stressful conditions, but wasineffective under conditions that lead to full, sustained mito-chondrial depolarization.

3.5. Mitochondrial depolarization

Under normal physiological conditionsmitochondria undergochanges in ΔΨm associated with the energetic demands of thecell (Hayakawa et al., 2005; Weeber et al., 2002). Such changesare believed to result from alteration of the mitochondrialproton gradient (Chen, 1988). However, during ischemia/reperfusion or OGD, mitochondrial depolarization (and swel-ling) is associated with opening of the MPTP. As mentionedabove, calpain activity has been linked to MPTP opening, andneuroprotection during ischemia/reperfusion by CsA is typi-cally attributed to prevention of MPTP opening.

Calpain inhibition resulted in a small, non-significantreduction of mitochondrial depolarization during OGD as as-sessed by Rh 123 fluorescence. A similar reduction in mito-chondrial membrane depolarization was observed with CsA.Given the typical correlation between mitochondrial swellingand ΔΨ (at least in intact cells, Devin et al., 1998; see alsoSafiulina et al., 2006), it seems therefore likely that mitochon-drial swelling was not significantly different between slicestreated with calpain inhibitor compared with slices treatedwith CsA. Yet, we observed significantly less darkening ofhippocampal tissue following HSD with calpain inhibition.

3.6. Conclusions

The results here demonstrate that calpain activation signifi-cantly effects the early stages of cellular response during OGD

128 B R A I N R E S E A R C H 1 1 9 6 ( 2 0 0 8 ) 1 2 1 – 1 3 0

observed here as light transmittance changes. Reduced tissuedarkening during OGD with calpain inhibition is not likely tobe the result of reduced dendritic beading nor does it appear tobe due to reducedmitochondrial swelling. A possible explana-tion of our results that remains to be tested is that calpaininhibition may limit condensation of the mitochondrialmatrix. While the neuroprotective benefits of calpain inhibi-tion are well-established, these findings provide new insightinto potential mechanisms by which calpain activation con-tributes to cell death.

4. Experimental procedures

4.1. Slice preparation

All experimental protocols were approved by the AnimalEthics Committee of the University of Auckland. Acute hip-pocampal slices were obtained from P12–P17 Wistar rats aspreviously described (Lipski et al., 2006). Briefly, brains wererapidly removed under CO2 anaesthesia and placed inoxygenated (95% O2/5% CO2), ice-cold, ‘standard’ ACSF withthe addition of 3.0 mMMgSO4. Slices (350 μm) were cut using avibratome (VT1000S, Leica) and transferred for 20–30 min toa holding chamber containing oxygenated, standard ACSF at34±0.5 °C. Subsequently, sliceswere kept at room temperaturewith continuous oxygenation for an additional minimum of1 h recovery time. Standard ACSF consisted of (mM): NaCl 120,NaHCO3 26, KCl 3.0, NaH2PO4 1.25, CaCl2 2.0, MgSO4 2.0, glucose20 and L-ascorbic acid, 0.4 (285–295 mosM; pH 7.4). Forrecordings, slices were transferred to a low-volume (1.5 ml),thermoregulated chamber (JG-23, Warner Instruments) andkept submerged at 36 °C (3.5 ml/min). ACSF used duringexperiments consisted of standard ACSF with lower glucoseconcentration (5 mM), higher NaCl concentration (126 mM)and contained no ascorbic acid. For OGD conditions, ACSFconsisted of (mM) NaCl 120, NaHCO3 26, KCl 5, NaH2PO4 1.25,CaCl2 2.0, MgSO4 2.0; osmotically balanced with sucrose(15 mM) and gassed with 95% N2/5% CO2. Before collectingbaseline data, slices were exposed for 20 min to MDL 28170 orcyclosporin A. Drug concentrations weremaintained through-out the entire experiment.

4.2. Electrophysiology

Schaeffer collateral-evoked field potentials (fEPSPs) and OGD-induced extracellular DC negative shifts (Lipski et al., 2006,2007) were recorded in Str. radiatum of the CA1 region using aDC amplifier (NL102, Digitimer; 3 kHz low-pass) and glassmicroelectrodes (2–6 MΩ) filled with standard ACSF. fEPSPswere evoked by current pulses delivered with a twistedtungsten wire (0.1 mm, Teflon coated) at 0.2 Hz. Data wereacquired using Clampex (v.9.2, Molecular Devices) and wereanalyzed with Clampfit (v.9.2, Molecular Devices). The latencyto the DC shift was measured from the onset of OGD to thetime at which the DC potential reached 50% of peak deflection.The latency to disappearance of the fEPSP wasmeasured fromthe onset of OGD to the timewhen the fEPSP decayed to 10% ofbaseline amplitude. Slices with unstable fEPSP baseline wereconsidered unhealthy and not included in analysis.

4.3. Light transmittance

Light transmittance through selected regions of the slice wasmeasured using an approach similar to that described before(Andrew et al., 1999; Davidson and Lipski, 2005; Hoffman et al.,2000; Kreisman et al., 1995; Muller and Somjen, 1999).Measurements were made in Str. radiatum and Str. pyrami-dale of CA1 in submerged slices with a long working distance(10.8 mm) 5× objective (numerical aperture 0.12) using anupright microscope (Axioskop, Zeiss). Transmitted light froma 50-W halogen lamp passed through a 0.2% neutral densityfilter and long pass filter (590 nm) and was detected using acooled CCD camera (Orca-ER, Hamamatsu) with 100 msexposure time (binning 2×2; gain 50; (Davidson and Lipski,2005). Light transmittance values were collected every 10 s in 6identical, non-overlapping circular regions of interest (ROIs,100 μm diameter, 3 in each subregion of CA1), using ImageWorkbench (v.4.0, Molecular Devices). For analysis, the 3values from the individual ROIs in Str. pyramidale and Str.radiatumof each slice were averaged at every time point usingImage Workbench and Excel (Microsoft).

4.4. Mitochondrial membrane potential

OGD-induced mitochondrial depolarization (ΔΨm) was evalu-ated using the fluorescent indicator, Rhodamine 123 (Rh 123,Sigma) under similar conditions as described by Bahar et al.(2000). Rh 123 (20 mg/ml) was dissolved in 95% EtOH as a stocksolution and stored at −20 °C. Slices were loaded with Rh 123under low light conditions at a final concentration of 5 μg/ml(13.3 μM) for 30 min in a holding chamber filled with standardACSF (gassed with 95% O2/5% CO2 at room temperature.Illumination was provided by a 175 W xenon lamp (SutterInstruments) and fluorescent signals were recorded in thesame ROIs as transmitted light (5× objective) every 10 s with arhodamine filter set (BP 510–560, dichroic 580, LP 590) and200 ms exposure time. A neutral density filter (80% reduction)was used in the light excitation pathway. Data were collectedand analyzed off-line with Image Workbench and Excel. Themean values recorded during the control period were used tonormalize responses.

4.5. Hypo-osmotic swelling

Tissue swelling in Str. pyramidale and Str. radiatum of CA1was inducedwith rapid exchange of standard ACSF containingno ascorbic acid (285–295 mosM) for hypo-osmotic ACSF withdecreased NaCl concentration (105 and 90 mM) givingapproximately 260 and 230 mosM, respectively. Light trans-mittance was measured as described above (3 ROIs in Str.pyramidale and 3 in Str. radiatum).

4.6. Drugs

MDL 28170 (Sigma) was dissolved in DMSO (40 mM stock) andstored at −20 °C. For experiments, MDL 28170 (or DMSO con-trol) was delivered by syringe pump (SP100IZ, WPI) at 3.5 μl/min to the inflow of the recording chamber, providing a 1000-fold dilution (final concentration, 40 μM). Cyclosporin A (CsA,Sigma) was dissolved in DMSO: 95% EtOH; 1:1 (10 mM stock)

129B R A I N R E S E A R C H 1 1 9 6 ( 2 0 0 8 ) 1 2 1 – 1 3 0

and stored at −20 °C. Delivery of CsA to the inflow of therecording chamber via syringe pump resulted in drug pre-cipitation. Therefore, CsA was slowly dissolved in ACSF for afinal concentration of 10 μM. Due to this constraint, arecirculating system of solution delivery was used in this setof experiments.

4.7. Statistics

Data are presented asmeans+S.E.M. Statistical difference wasdetermined by Student's t tests with significance levels <0.05.

Acknowledgments

This study was supported by the New Zealand NeurologicalFoundation and the University of Auckland Research Com-mittee. The authors kindly thank Drs. Alistair Stewart andDenis Loiselle for helpful discussions.

R E F E R E N C E S

Abbracchio, M.P., Cattabeni, F., 1999. Brain adenosine receptors astargets for therapeutic intervention in neurodegenerativediseases. Ann. N.Y. Acad. Sci. 890, 79–92.

Abramov, A.Y., Scorziello, A., Duchen, M.R., 2007. Three distinctmechanisms generate oxygen free radicals in neurons andcontribute to cell death during anoxia and reoxygenation.J. Neurosci. 27, 1129–1138.

Aitken, P.G., Fayuk, D., Somjen, G.G., Turner, D.A., 1999. Use ofintrinsic optical signals to monitor physiological changes inbrain tissue slices. Methods 18, 91–103.

Allen, N.J., Karadottir, R., Attwell, D., 2005. A preferential role forglycolysis in preventing the anoxic depolarization of rathippocampal area CA1 pyramidal cells. J. Neurosci. 25, 848–859.

Andrew, R.D., Jarvis, C.R., Obeidat, A.S., 1999. Potential sources ofintrinsic optical signals imaged in live brain slices. Methods 18,185–196 (179).

Arrington, D.D., Van Vleet, T.R., Schnellmann, R.G., 2006. Calpain10: a mitochondrial calpain and its role in calcium-inducedmitochondrial dysfunction. Am. J. Physiol., Cell Physiol. 291,C1159–C1171.

Bahar, S., Fayuk, D., Somjen, G.G., Aitken, P.G., Turner, D.A., 2000.Mitochondrial and intrinsic optical signals imaged duringhypoxia and spreading depression in rat hippocampal slices.J. Neurophysiol. 84, 311–324.

Brustovetsky, N., Dubinsky, J.M., 2000. Limitations of cyclosporin Ainhibition of the permeability transition in CNS mitochondria.J. Neurosci. 20, 8229–8237.

Carragher, N.O., 2006. Calpain inhibition: a therapeutic strategytargeting multiple disease states. Curr. Pharm. Des. 12,615–638.

Chen, L.B., 1988. Mitochondrial membrane potential in living cells.Annu. Rev. Cell Biol. 4, 155–181.

Chen, M., Won, D.J., Krajewski, S., Gottlieb, R.A., 2002. Calpain andmitochondria in ischemia/reperfusion injury. J. Biol. Chem.277, 29181–29186.

Choi, D.W., Rothman, S.M., 1990. The role of glutamateneurotoxicity in hypoxic–ischemic neuronal death. Annu. Rev.Neurosci. 13, 171–182.

Christophe, M., Nicolas, S., 2006. Mitochondria: a target forneuroprotective interventions in cerebral ischemia–reperfusion.Curr. Pharm. Des. 12, 739–757.

Davidson, J., Lipski, J., 2005. Factors Contributing to Cell Swelling inan In Vitro Model of Stroke. University of Virginia. Oculus U-21.

Devin, A., Espie, P., Guerin, B., Rigoulet, M., 1998. Energetics ofswelling in isolated hepatocytes: a comprehensive study. Mol.Cell. Biochem. 184, 107–121.

Fayuk, D., Aitken, P.G., Somjen, G.G., Turner, D.A., 2002. Twodifferent mechanisms underlie reversible, intrinsic opticalsignals in rat hippocampal slices. J. Neurophysiol. 87,1924–1937.

Friberg, H., Ferrand-Drake, M., Bengtsson, F., Halestrap, A.P.,Wieloch, T., 1998. Cyclosporin A, but not FK 506, protectsmitochondria and neurons against hypoglycemic damage andimplicates the mitochondrial permeability transition in celldeath. J. Neurosci. 18, 5151–5159.

Friberg, H., Connern, C., Halestrap, A.P., Wieloch, T., 1999.Differences in the activation of the mitochondrial permeabilitytransition among brain regions in the rat correlate withselective vulnerability. J. Neurochem. 72, 2488–2497.

Fujii, F., Nodasaka, Y., Nishimura, G., Tamura, M., 2004. Anoxiainduces matrix shrinkage accompanied by an increase in lightscattering in isolated brainmitochondria. Brain Res. 999, 29–39.

Gerich, F.J., Hepp, S., Probst, I., Muller, M., 2006. Mitochondrialinhibition prior to oxygen-withdrawal facilitates the occurrenceof hypoxia-induced spreading depression in rat hippocampalslices. J. Neurophysiol. 96, 492–504.

Goll, D.E., Thompson, V.F., Li, H., Wei, W., Cong, J., 2003. Thecalpain system. Physiol. Rev. 83, 731–801.

Gores, G.J., Miyoshi, H., Botla, R., Aguilar, H.I., Bronk, S.F., 1998.Induction of the mitochondrial permeability transition as amechanism of liver injury during cholestasis: a potential role formitochondrial proteases. Biochim. Biophys. Acta 1366, 167–175.

Hansson, M.J., Persson, T., Friberg, H., Keep, M.F., Rees, A.,Wieloch,T., Elmer, E., 2003. Powerful cyclosporin inhibition ofcalcium-induced permeability transition in brain mitochondria.Brain Res. 960, 99–111.

Hayakawa, Y., Nemoto, T., Iino, M., Kasai, H., 2005. RapidCa2+-dependent increase in oxygen consumption bymitochondria in single mammalian central neurons. CellCalcium 37, 359–370.

Hochgraf, E., Mokady, S., Cogan, U., 1997. Dietary oxidized linoleicacid modifies lipid composition of rat liver microsomes andincreases their fluidity. J. Nutr. 127, 681–686.

Hoffmann, E.K., Simonsen, L.O., 1989. Membrane mechanisms involume and pH regulation in vertebrate cells. Physiol. Rev. 69,315–382.

Hoffman, B., Olson, J.E., Kreisman, N., 2000. The role of taurine forcellular volume regulation of the hippocampus. Acad. Emerg.Med. 7, 1168.

Hong, S.C., Goto, Y., Lanzino, G., Soleau, S., Kassell, N.F., Lee, K.S.,1994. Neuroprotection with a calpain inhibitor in a model offocal cerebral ischemia. Stroke 25, 663–669.

Imai, K., Aimoto, T., Shima, T., Nakashima, T., Sato, M., Kimura, R.,2000. Alteration in membrane fluidity of rat liver microsomesand of liposomes by protoporphyrin and itsanti-lipidperoxidative effect. Biol. Pharm. Bull. 23, 415–419.

Kaasik, A., Safiulina, D., Zharkovsky, A., Veksler, V., 2007.Regulation of mitochondrial matrix volume. Am. J. Physiol.,Cell Physiol. 292, C157–C163.

Kawamura, M., Nakajima, W., Ishida, A., Ohmura, A., Miura, S.,Takada, G., 2005. Calpain inhibitor MDL 28170 protectshypoxic–ischemic brain injury in neonatal rats by inhibition ofboth apoptosis and necrosis. Brain Res. 1037, 59–69.

Khoutorsky, A., Spira, M.E., 2005. Calcium-activated proteases arecritical for refilling depleted vesicle stores in culturedsensory-motor synapses of Aplysia. Learn. Mem. 12, 414–422.

Kimelberg, H.K., 1995. Current concepts of brain edema. Review oflaboratory investigations. J. Neurosurg. 83, 1051–1059.

Korge, P., Honda, H.M., Weiss, J.N., 2005. K+-dependent regulationof matrix volume improves mitochondrial function under

130 B R A I N R E S E A R C H 1 1 9 6 ( 2 0 0 8 ) 1 2 1 – 1 3 0

conditions mimicking ischemia–reperfusion. Am. J. Physiol,Heart Circ. Physiol. 289, H66–H77.

Kreisman, N.R., LaManna, J.C., 1999. Rapid and slow swellingduring hypoxia in the CA1 region of rat hippocampal slices.J. Neurophysiol. 82, 320–329.

Kreisman, N.R., LaManna, J.C., Liao, S.C., Yeh, E.R., Alcala, J.R.,1995. Light transmittance as an index of cell volume inhippocampal slices: optical differences of interfaced andsubmerged positions. Brain Res. 693, 179–186.

Kristian, T., Gertsch, J., Bates, T.E., Siesjo, B.K., 2000. Characteristicsof the calcium-triggeredmitochondrial permeability transitionin nonsynaptic brain mitochondria: effect of cyclosporin A andubiquinone O. J. Neurochem. 74, 1999–2009.

Lee, K.S., Frank, S., Vanderklish, P., Arai, A., Lynch, G., 1991.Inhibition of proteolysis protects hippocampal neurons fromischemia. Proc. Natl. Acad. Sci. U. S. A. 88, 7233–7237.

Lipski, J., Park, T.I., Li, D., Lee, S.C., Trevarton, A.J., Chung, K.K.,Freestone, P.S., Bai, J.Z., 2006. Involvement of TRP-like channelsin the acute ischemic response of hippocampal CA1 neurons inbrain slices. Brain Res. 1077, 187–199.

Lipski, J., Wan, C.K., Bai, J.Z., Pi, R., Li, D., Donnelly, D., 2007.Neuroprotective potential of ceftriaxone in in vitro models ofstroke. Neuroscience 146, 617–629.

Liu, X., Kim, C.N., Yang, J., Jemmerson, R.,Wang, X., 1996. Inductionof apoptotic program in cell-free extracts: requirement for dATPand cytochrome c. Cell 86, 147–157.

Malagelada, C., Xifro, X., Minano, A., Sabria, J., Rodriguez-Alvarez,J., 2005. Contribution of caspase-mediated apoptosis to the celldeath caused by oxygen–glucose deprivation in cortical cellcultures. Neurobiol. Dis. 20, 27–37.

Muller, M., Somjen, G.G., 1999. Intrinsic optical signals in rathippocampal slices during hypoxia-induced spreadingdepression-like depolarization. J. Neurophysiol. 82, 1818–1831.

Newcomb-Fernandez, J.K., Zhao, X., Pike, B.R., Wang, K.K., Kampfl,A., Beer, R., DeFord, S.M., Hayes, R.L., 2001. Concurrentassessment of calpain and caspase-3 activation afteroxygen–glucose deprivation in primary septo-hippocampalcultures. J. Cereb. Blood Flow Metab. 21, 1281–1294.

Nicholls, D.G., Budd, S.L., 2000. Mitochondria and neuronalsurvival. Physiol. Rev. 80, 315–360.

Okada, Y., 1997. Volume expansion-sensing outward-rectifierCl− channel: fresh start to the molecular identity and volumesensor. Am. J. Physiol. 273, C755–C789.

Raaflaub, J., 1953. Swelling of isolated mitochondria of the liverand their susceptibility to physicochemical influences. Helv.Physiol. Pharmacol. Acta 11, 142–156.

Roberts-Lewis, J.M., Siman, R., 1993. Spectrin proteolysis in thehippocampus: a biochemical marker for neuronal injury andneuroprotection. Ann. N.Y. Acad. Sci. 679, 78–86.

Roberts-Lewis, J.M., Savage, M.J., Marcy, V.R., Pinsker, L.R., Siman,R., 1994. Immunolocalization of calpain I-mediated spectrindegradation to vulnerable neurons in the ischemic gerbil brain.J. Neurosci. 14, 3934–3944.

Rossi, D.J., Oshima, T., Attwell, D., 2000. Glutamate release insevere brain ischaemia is mainly by reversed uptake. Nature403, 316–321.

Safiulina, D., Veksler, V., Zharkovsky, A., Kaasik, A., 2006. Loss ofmitochondrial membrane potential is associated with increaseinmitochondrial volume: physiological role in neurones. J. Cell.Physiol. 206, 347–353.

Schiff, S.J., Somjen, G.G., 1987. The effect of graded hypoxia on thehippocampal slice: an in vitromodel of the ischemic penumbra.Stroke 18, 30–37.

Seki, Y., Feustel, P.J., Keller Jr., R.W., Tranmer, B.I., Kimelberg, H.K.,1999. Inhibition of ischemia-induced glutamate release in ratstriatum by dihydrokinate and an anion channel blocker.Stroke 30, 433–440.

Seubert, P., Lee, K., Lynch, G., 1989. Ischemia triggers NMDAreceptor-linked cytoskeletal proteolysis in hippocampus. BrainRes. 492, 366–370.

Seubert, P., Peterson, C., Vanderklish, P., Cotman, C., Lynch, G.,1990. Increased spectrin proteolysis in the brindled mousebrain. Neurosci. Lett. 108, 303–308.

Somjen, G.G., 2001. Mechanisms of spreading depression andhypoxic spreading depression-like depolarization. Physiol.Rev. 81, 1065–1096.

Sun, A.C., Levitan, I., 2003. Osmotic stress alters the intracellulardistribution of non-erythroidal spectrin (fodrin) in bovineaortic endothelial cells. J. Membr. Biol. 192, 9–17.

Susin, S.A., Zamzami, N., Castedo, M., Hirsch, T., Marchetti, P.,Macho, A., Daugas, E., Geuskens, M., Kroemer, G., 1996. Bcl-2inhibits the mitochondrial release of an apoptogenic protease.J. Exp. Med. 184, 1331–1341.

Vanderklish, P., Saido, T.C., Gall, C., Arai, A., Lynch, G., 1995.Proteolysis of spectrin by calpain accompanies theta-burststimulation in cultured hippocampal slices. Brain Res. Mol.Brain Res. 32, 25–35.

Weeber, E.J., Levy, M., Sampson, M.J., Anflous, K., Armstrong, D.L.,Brown, S.E., Sweatt, J.D., Craigen, W.J., 2002. The role ofmitochondrial porins and the permeability transition pore inlearning and synaptic plasticity. J. Biol. Chem. 277, 18891–18897.

Zhou, M., Baudry, M., 2006. Developmental changes in NMDAneurotoxicity reflect developmental changes in subunit com-position of NMDA receptors. J. Neurosci. 26, 2956–2963.

Zieminska, E., Matyja, E., Kozlowska, H., Stafiej, A., Lazarewicz, J.W.,2006. Excitotoxic neuronal injury in acute homocysteineneurotoxicity: role of calcium and mitochondrial alterations.Neurochem. Int. 48, 491–497.