Hypoxia affects the physiological behavior of rat cortical synaptosomes

Free Radical Biology & Medicine 42 (2007) 1749–1756www.elsevier.com/locate/freeradbiomed

Original Contribution

Hypoxia affects the physiological behavior of rat cortical synaptosomes

Carlo Aldinucci a, Alessandra Carretta a, Lucia Ciccoli b, Silvia Leoncini b, Cinzia Signorini b,Giuseppe Buonocore c, Gian Paolo Pessina a,⁎

a Department of Physiology, University of Siena, 53100 Siena, Italyb Department of Pathophysiology, Experimental Medicine, and Public Health, University of Siena, 53100 Siena, Italy

c Department of Pediatrics, Obstetrics, and Reproductive Medicine, University of Siena, 53100 Siena, Italy

Received 15 June 2006; revised 19 February 2007; accepted 6 March 2007Available online 13 March 2007

Abstract

Nerve cells, especially synaptosomes, are very susceptible to hypoxia and the subsequent oxidative stress. In this paper, we examined theeffects of hypoxia (93% N2:2% O2:5% CO2, v/v/v) on rat cortical synaptosomes by evaluating modifications of synaptosomal mitochondrialrespiration rate and ATP production, membrane potential, intrasynaptosomal mitochondrial Ca2+ concentration ([Ca2+]i), and desferoxamine-chelatable free iron and esterified F2-isoprostane levels after different periods of hypoxia and after 30 min of reoxygenation. Oxygen consumptiondecreased significantly during 120 min of hypoxia and was restored after reoxygenation. At the same time, ATP production decreased andremained significantly lower even after reoxygenation. This involved a depolarization of the synaptosomal mitochondrial membrane, although the[Ca2+]i remained practically unchanged. Indeed, iron and F2-isoprostane levels, representing useful prediction markers for neurodevelopmentaloutcome, increased significantly after hypoxia, and there was a strong correlation between the two variables. On the whole our results indicate thatsynaptosomal mitochondria undergo mitoptosis after 2 h of hypoxia.© 2007 Published by Elsevier Inc.

Keywords: Mitochondrial respiration; Mitochondrial membrane potential; Lipid peroxidation; Free iron release; F2-isoprostanes; Free radicals

Introduction

Normal brain function is highly dependent on adequatetissue oxygenation, and acute brain ischemia is a major cause ofmorbidity and mortality. Increasing evidence suggests thatmitochondrial dysfunctions are also caused by hypoxia [1].Mitochondria play a key role in the normal life of the cell andare involved in both apoptosis and necrosis. In fact, apoptosisinvolves defects in the maintenance of adenosine 5′-tripho-sphate (ATP) production and mitochondrial membrane potential(ΔΨm) [2], as well as in the opening of the mitochondrialpermeability transition pore (PTP), which is regulated by theΔΨm, by the pH of the mitochondrial matrix, and bymodifications in the intracellular Ca2+ concentration [3]. Duringhypoxia/anoxia electrons accumulate in respiratory carriers andpromote a reductive stress. After reperfusion, there is anapparent hyperoxidation of components of the mitochondrial

⁎ Corresponding author. Fax: +39 0577 234219.E-mail address: [email protected] (G.P. Pessina).

0891-5849/$ - see front matter © 2007 Published by Elsevier Inc.doi:10.1016/j.freeradbiomed.2007.03.010

respiratory chain and increased lipid peroxidation [4], accom-panied by the closure of the voltage-dependent anion channel[5]. During stress conditions mitochondria overproduce reactiveoxygen species which cause the PTP to open with a collapse ofthe ΔΨm and cessation of the import of mitochondrial proteinprecursors. So mitochondria undergo mitoptosis [6], whichmeans the suicide of the mitochondrion and can result in the celldeath. Moreover, the role of oxidative stress and mitochondrialdysfunction in neurodegenerative diseases is well demonstratedand suggests that strategic therapies targeting physiologicalmitochondrial mechanisms hold great promise [7].

Iron is an essential element for mammalian cells and abalance between its accumulation and its release is necessary forthe maintenance of cell viability [8]. An increase in intracellulardesferoxamine(DFO)-chelatable free iron has been demon-strated after cerebral ischemia [9,10]. Moreover, iron and lipidperoxidation play an important role in oxidant injury and in thepathogenesis of hypoxic/ischemic encephalopathy [11]. Mor-row et al. [12] demonstrated the formation of a series ofprostaglandin F2-like compounds named F2-isoprostanes,

1750 C. Aldinucci et al. / Free Radical Biology & Medicine 42 (2007) 1749–1756

which are formed by free radical-catalyzed peroxidation ofesterified arachidonic acid in phospholipids, a pathway in-dependent of cyclo-oxygenase activity. Therefore, the measure-ment of F2-isoprostanes is a reliable method to assess oxidativestate in vivo and in vitro [11].

Neurons and astrocytes are very susceptible to hypoxia andthe subsequent oxidative stress. In a previous study, we ob-served an increase in intracellular Ca2+ mobilization byhypoxia-induced NO generation in rat brain striatal slices andhuman astrocytoma U-373 MG cells [13]. Synaptosomes arepinched off of resealed isolated nerve terminals prepared fromcortical brain neurons, which contain essentially mitochondria,enzymes, and vesicles of neurotransmitters. Retaining most ofthe physiological behavior of normal nerve terminals, such asthe maintenance of ATP levels and the release of theneurotransmitters, these synaptosomes are widely used forneurochemical research [14]. Synaptosomes are particularlyaffected by oxidative stress, and mitochondria from differenttissues, including synaptic and nonsynaptic brain mitochondria,have different susceptibilities to oxidants [15–17].

Because the presence of both neuronal and glial mitochon-dria in nonsynaptic preparations could be a confounding factor,in this paper we examined the effects of hypoxia (2% O2) on ratcortical synaptosomes by evaluating modifications of thesynaptosomal mitochondrial respiration (JO2) rate and mem-brane potential, intrasynaptosomal mitochondrial Ca2+ concen-tration ([Ca2+]i), and free iron DFO-chelatable and esterifiedF2-isoprostane levels after different periods of hypoxia.Because reperfusion after brief periods of global cerebralischemia is accompanied by hyperoxidation of components ofthe mitochondrial respiratory chain due to decreased substrateavailability or loss of electron carriers from mitochondria [18],we evaluated the same parameters after 30 min of synaptosomereoxygenation.

Materials and methods

Preparation of synaptosomes

Synaptosomes were isolated from adult male Sprague–Dawley rats. We used the preparation from a single brain foreach experiment. The brain was homogenized (1/10 w/v) in0.32 M saccharose, 10 mM Tris–HCl, and 1 mM K+EGTA(STE) buffer using a tight Potter–Elvehjem glass–Teflonhomogenizer. The homogenate was centrifuged at 1200g for3 min at 4°C in a Beckman Allegra 21R centrifuge. After twowashes, both supernatants were centrifuged at 12,000g for10 min at 4°C. The sediments, containing synaptosomes andmyelin, were resuspended in 9 ml STE buffer and stratified on aconventional 11 and 8% Ficoll gradient. They were then centri-fuged in a Beckman Optima LE-70 ultracentrifuge (RotorSW28) at 28,000 rpm for 45 min at 4°C. The myelin layer wasremoved by aspiration and discarded. The synaptosomal layerwas resuspended in STE buffer and centrifuged at 12,000g for15 min at 4°C. The resulting pellet containing pure synapto-somes was resuspended in 1 ml of STE buffer and homogenizedusing a 2-ml capacity Potter–Elvehjem homogenizer. The pure

synaptosomal suspension was kept on ice until used for theexperiments. Proteins were evaluated by the method of Gornallet al. [19].

Hypoxia conditions

Synaptosomes were resuspended in Kreb's buffer at a finalconcentration of 2 mg/ml. The densities of synaptosomes in oursuspensions were far lower than in cell cultures or in intacttissues and the volume of the surrounding medium was muchhigher than the extracellular tissue volume. Therefore, to ensurethe established hypoxic conditions, we resuspended thesynaptosomes in hypoxic medium (obtained by bubbling theappropriate gas mixture for 5 min) in 25-ml flasks (one flask foreach hypoxic period) and gassed them for 5 min with 2%O2:93% N2:5% CO2. The flasks were then closed and incubatedfor 2 h at 37°C. After 30, 60, and 120 min of hypoxia, an aliquotof synaptosomal suspension was collected for the variousdeterminations. The remainder was incubated for 30 min in 95%air:5% CO2 (reoxygenation) and then subjected to the samedeterminations as above. Control flasks were gassed with 95%air:5% CO2 and treated as the hypoxic flasks. Because theincubation periods were very brief, we pregassed the mediumwith the hypoxic mixture and used a closed system for thehypoxic conditions in order to ensure better saturation of thesystem.

Determination of oxygen consumption and ATP production

Oxygen consumption (JO2) of synaptosomal mitochondriawas measured with a Clark-type oxygen electrode (RankBrothers Ltd; Dual Digital Model 20). Synaptosomes (2 mg/ml)in Kreb's buffer were incubated under hypoxic or normoxicconditions. After incubation (30, 60, 120 min) and after 30 minof reoxygenation, 400 μl of synaptosomal suspension wasplaced in a water-jacketed Perspex reaction chamber (1 mlcapacity). Temperature was maintained at 37°C and the currentflow was recorded for 500 s using Pico Technology Oxygraphsoftware. A polarizing voltage of 0.616 was applied across thecells to get the current flowing proportional to the dioxygenconcentration in the reaction chamber. Calibration wasperformed previously using sodium hydrosulfite to define“zero-oxygen.” An oxygen concentration of 313 nmol O2/ml inthe air-saturated reaction medium was assumed at 37°C. JO2

was calculated as nmol O2 min−1 mg−1.ATP production was determined with an ATP biolumines-

cence assay kit (Sigma Chemical, Dorset, England) followingthe manufacturer's instructions.

Determination of ΔΨm

The membrane potential-sensitive probe 5,5′,6,6′-tetra-chloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide(JC-1) was used to measure the relative change in mitochondrialmembrane potential. JC-1 is a cationic dye that exhibitspotential-dependent accumulation in mitochondria, indicatedby a fluorescence emission shift from green (∼535 nm) to red

1751C. Aldinucci et al. / Free Radical Biology & Medicine 42 (2007) 1749–1756

(∼590 nm). Consequently, mitochondrial depolarization isindicated by an increase in the green/red fluorescence intensityratio. The ratio of green to red fluorescence depends solely onthe membrane potential and not on other factors such asmitochondrial size, shape, and density that could influencesingle-component fluorescence signals.

After hypoxia and after reoxygenation, synaptosomes wereloaded with JC-1 (6 μM) for 15 min at 37°C, then washed threetimes, resuspended in Kreb's buffer, and kept on ice. Twohundred microliters of synaptosomes (50 μg/ml in Kreb'sbuffer) were placed in the wells of a microtiter plate and mixed.A Fluoroskan Ascent fluorimeter (ThermoLabsystems) wasused. During the measurement, synaptosomes were maintainedat 37°C and protected from light. The Ψm was estimated as theratio of JC-1 monomers (green, λem=538 nm) to JC-1 aggre-gates (red, λem=590 nm) at λex=485 nm. All fluorescencemeasurements were corrected for autofluorescence of synapto-somes not loaded with JC-1, which was constant throughout theexperiments. In control experiments, no photobleaching wasobserved during fluorescence measurements.

Measurement and calculation of [Ca2+]i

The intrasynaptosomal calcium concentration was measuredimmediately after 120 min of hypoxia and after 30 min ofreoxygenation. Synaptosomes were distributed in 96-wellmicrotiter plates (200 μl/well) and then washed two timeswith Hepes-buffered saline (HBS) containing 145 mM NaCl,5 mM KCl, 1 mM MgCl2, 1.3 mM CaCl2, 10 mM Hepes, and10 mM glucose, pH 7.4, supplemented with 0.03% Pluronic F-127. The synaptosomes were then incubated for 75 min at 25°Cin the dark with 10 μM fura-2 acetoxymethyl ester (Fura-2) indimethyl sulfoxide, after which they were washed three timeswith HBS.

[Ca2+]i was monitored with a Fluoroskan Ascent fluorimeter(ThermoLabsystems) and fluorescence (F) was recorded at340 nm excitation and 505 nm emission. The calibration wasthen performed on each sample to evaluate [Ca2+]i. Fura-2leakage was estimated by quenching the extracellular dyefluorescence with 0.2 mM MnCl2, which was then chelated bythe addition of 0.5 mM CaDTPA.

The maximal fluorescence (Fmax) was obtained by addingsequentially 10 nMCaCl2, 2.3 μMCa2+ ionophore A 23187, and100 μM digitonin. Finally, 20 mM MnCl2 was added to recordthe autofluorescence, from which the minimum fluorescence(Fmin) was calculated according to Hesketh et al. [20]. The F,Fmin, and Fmax values were corrected for autofluorescence andthe [Ca2+]i was calculated from F according to Tsien et al. [21]and Grynkiewicz et al. [22].

Measurement of DFO-chelatable iron

For these experiments we used synaptosomes that were moreconcentrated (10 mg proteins/ml), and DFO-chelatable free ironwas measured according to a previous study [23]. Briefly, afterhypoxia, DFO was added to 1 ml of synaptosomal suspension toobtain a final concentration of 25 μM DFO, and the synap-

ztosomes were ruptured by the addition of 1 ml water, freeze–thawing, and sonication. The samples were then ultrafiltered incentrifuge filter devices (Centriplus YM30; Amicon) with mo-lecular exclusion of 30,000 Da. The DFO excess was removedby silica column chromatography and the DFO–iron complexwas determined by HPLC. Control samples were treated asabove.

Determination of esterified F2-isoprostanes

The levels of esterified F2-isoprostanes in the synaptosomalsuspensions (10 mg proteins/ml) were determined by gaschromatography/negative ion chemical ionization tandemmass spectrometry (GC/NICI-MS/MS) analysis, a very sensitive(picograms) and specific method [24]. After incubation,butylated hydroxytoluene was added to 1 ml of synaptosomalsuspension as an antioxidant to obtain a final concentration of90 μMand the samples were frozen at−70°C until use. Later, thesamples were exposed to basic hydrolysis (1 M KOH) at 45°Cfor 45min. After acidification to pH 3with 1MHCl, the sampleswere spiked with tetradeuterated PGF2α (500 pg in 50 μlethanol), as internal standard. After extraction with ethyl acetatethe total lipid extract was applied directly onto an aminopropyl(NH2) cartridge and the final eluate was derivatized topentafluorobenzyl ester and trimethylsilyl ether and examinedby GC/NICI-MS/MS. Control samples were treated as above.

Statistical analysis

Twelve Sprague–Dawley rats were used for this study. Foreach experiment, the synaptosomes were prepared from a singlerat brain and the results were reported as means±SD. The datawere analyzed by two-tailed Student t test for paired sampleswith p<0.05 as the minimum level of significance.

Results

JO2 of synaptosomal mitochondria

The JO2 of synaptosomal mitochondria was measured with aClark-type oxygen electrode in synaptosomal suspensions(2mg/ml) in Kreb's buffer incubated under hypoxic or normoxic(controls) conditions for 30, 60, or 120 min and after 30 min ofreoxygenation. In control samples, JO2 increased slowly duringthe 120-min incubation period (Fig. 1, bottom). In hypoxicsynaptosomes, JO2 decreased significantly after 30 minhypoxia, from 2.3±0.45 to 1.4±0.5 nM O2/min/mg proteinand the decrease was maintained after the subsequent incubationperiod (Fig. 1, bottom). When the hypoxic synaptosomalsuspensions were maintained in normoxia for 30 min (reox-ygenation), normal oxygen consumption was restored in allsamples (Fig. 1, top).

ATP production

ATP production was determined with an ATP biolumines-cence assay kit on experimental samples during and 30 min after

Fig. 2. ATP production, determined with an ATP bioluminescence assay kit onexperimental samples, during 120 min hypoxia (bottom) and after 30 min ofreoxygenation (top). Controls were incubated under normoxia. At each timepoint, the results of controls were compared with those of the various hypoxicsamples by two-tailed t test for paired samples. Each point represents the mean±SD of 12 independent experiments. ⁎p<0.05.

Fig. 1. Time course of O2 consumption (JO2) by synaptosomal mitochondriameasured with a Clark-type oxygen electrode in synaptosomal suspensions(2 mg/ml in Kreb's buffer) incubated under hypoxic or normoxic (control)conditions for 30, 60, or 120 min (bottom) and after 30 min of reoxygenation(top). Each point represents the means±SD of 12 independent experiments. Ateach time point, results of controls were statistically compared with those of thevarious hypoxia times by two-tailed t test for paired samples. ⁎p<0.05,⁎⁎p<0.01.

1752 C. Aldinucci et al. / Free Radical Biology & Medicine 42 (2007) 1749–1756

hypoxia. Fig. 2 (bottom) shows that hypoxia affected the ATPproduction by synaptosomal mitochondria: it decreased sig-nificantly to 80±2% during the first 60 min and remainedunchanged for the following 60 min. Interestingly the ATPproduction remained significantly lower than that of controlseven after the reoxygenation period (Fig. 2, top), indicatingdamage to the mitochondria during hypoxia.

Modifications of mitochondrial membrane potential

After hypoxia and after a 30-min reoxygenation period,synaptosomes were loaded with 6 μM JC-1. The ΔΨm wasestimated as the ratio of JC-1 monomers (green, λem=538 nm)to JC-1 aggregates (red, λem=590 nm) at λex=485 nm. The ratioof green to red in control synaptosomes incubated for 30, 60, and120 min in 95% air:5% CO2 was 0.123±0.003. During 2 h ofhypoxia the ratio increased progressively and significantly to0.149±0.018 (Fig. 3, bottom), indicating depolarization of thesynaptosomal mitochondrial membrane. Also in this case, themembrane potential in synaptosomes incubated in hypoxia for2 h did not return to the initial values after 30 min of reoxy-

genation, indicating that the mitochondrial membrane remaineddepolarized (Fig. 3, top).

Modifications of [Ca2+]i

The decreased ATP production and mitochondrial membranepotential after hypoxia prompted us to evaluate eventualmodifications of the [Ca2+]i. In controls, the calcium concentra-tion, which is strictly regulated in synaptosomes, remainedpractically unvaried at about 160 nM during the entireincubation period (Table 1). Likewise, [Ca2+]i remainedunchanged after 1 and 2 h of hypoxia and after the subsequentreoxygenation period.

Variations of DFO-chelatable iron

Iron is normally transported and stored in proteins, but it canpromote oxidative damage when released from its complexes.Hence, we measured the iron released in a free form(desferoxamine-chelatable) after 1 and 2 h of hypoxia. Incontrols, after 60 and 120 min aerobic incubation, the DFO-chelatable iron was practically unvaried at about 0.23 nM/mgproteins. The value increased slowly, but significantly(p<0.001), after the first and second hour of hypoxia, reaching

Fig. 4. Desferoxamine-chelatable free iron measured as described underMaterials and methods in synaptosomal suspensions (10 mg/ml) immediatelyafter the various hypoxic periods. Controls were incubated under normoxia. Ateach time point, results of controls were compared with those of the varioushypoxic samples by two-tailed t test for paired samples. Each point representsthe mean±SD of 12 independent experiments. ⁎p<0.001.

Fig. 3. Modifications of mitochondrial membrane potential measured byloading synaptosomal suspensions with 6 μM JC-1 immediately after thevarious hypoxia periods and after 30 min reoxygenation. The ΔΨm wasestimated as the ratio of JC-1 monomers (green, λem=538 nm) to JC-1aggregates (red, λem=590 nm) at λex=485 nm. Controls were incubated undernormoxia. At each time point, results of controls were compared with those ofthe various hypoxic samples by two-tailed t test for paired samples. Each pointrepresents the mean±SD of 12 independent experiments. ⁎p<0.05.

1753C. Aldinucci et al. / Free Radical Biology & Medicine 42 (2007) 1749–1756

0.29±0.09 nM/mg proteins (Fig. 4). This value remainedunvaried during the following 30 min of reoxygenation (data notshown).

Variations of concentration of esterified F2-isoprostanes

Free iron has been linked to the lipid peroxidation of thecell membrane, a major consequence of oxidative damage in avariety of tissues in vivo. Therefore, we evaluated themodifications of F2-isoprostane production in synaptosomal

Table 1Concentrations of intrasynaptosomal Ca2+ determined by Fura-2 fluorimetricanalysis before (control) and during 120 min hypoxia and after 30 minreoxygenation

Time Ca2+ measured immediatelyafter hypoxia (nM)

Ca2+ measured after 30 minof reoxygenation (nM)

0 (control) 160±22 145±9060 min 180±65 154±58120 min 195±77 182±49

Values represent the means±SD of 12 independent experiments.

suspensions before and after hypoxia by gas chromatography/negative ion chemical ionization tandem mass spectrometry(Fig. 5). The concentration of esterified F2-isoprostanesremained unvaried in controls after 1 and 2 h of normoxicincubation. The value increased significantly (p<0.001) after1 and 2 h of hypoxia. Once again, the level was maintainedafter 30 min of reoxygenation (data not shown). Fig. 6 showsthat there was a strong correlation (r=0.873, p<0.001)between free iron and esterified F2-isoprostanes after 2 h ofhypoxia.

Discussion

The brain is highly dependent on a continuous blood flow forthe supply of oxygen and glucose; indeed, it has one of the

Fig. 5. Esterified F2-isoprostanes in synaptosomal suspensions (10 mg/ml) weremeasured by gas chromatography/negative ion chemical ionization tandemmassspectrometry immediately after hypoxia and reported as pg/mg protein. Controlswere incubated under normoxia. At each time point, results of controls werecompared with those of the various hypoxic samples by two-tailed t test forpaired samples. Each point represents the mean±SD of 12 independent ex-periments. ⁎p<0.001.

Fig. 6. Correlation between DFO-chelatable free iron and esterified F2-isoprostanes evaluated in synaptosomal suspension (10 mg/ml) after 2 h ofhypoxia. n=12, p<0.001.

1754 C. Aldinucci et al. / Free Radical Biology & Medicine 42 (2007) 1749–1756

highest metabolic rates in the body. Therefore, it is veryvulnerable to ischemic conditions. It is generally accepted thatone of the main factors involved in cerebral ischemia andreperfusion injury is the generation of oxygen-derived freeradicals after hypoxia and postischemic reperfusion. In fact,neurons in the ischemic brain undergo ATP depletion, adecrease in pH, an increase in [Ca2+]i, and membranedepolarization [1]. Moreover, mitochondria are closely relatedto cell physiology and pathology in terms of ATP and Ca2+

homeostasis, reactive oxygen species generation, and apoptosis.Especially, synaptic mitochondria are implicated in neuralplasticity, which allows people to interact with the environmentand to struggle with neural injuries, and alteration of synapticplasticity could be implicated in some neurodegenerativediseases [25]. Mitochondrial functioning depends on the abilityof this organelle to maintain a constant ΔΨm across the innermitochondrial membrane [26]. Therefore, the aim of our studywas to evaluate the effects of hypoxia on mitochondria ofcortical synaptosomes and to correlate eventual differences inoxygen consumption and ATP production with free iron releaseand isoprostane production. We used synaptosome preparationsas it has been demonstrated that they retain the morphological,biochemical, and electrophysiological properties of the synapsewithout any contamination with nonsynaptic material [27]. Wedemonstrated that the oxygen consumption and consequentATP production by the synaptosomal mitochondria signifi-cantly decreased after 120 min of hypoxia. These results areconsistent with the loss of the mitochondrial electron transportchain at the level of complexes I, II–III, and V (ATP synthetase)during hypoxia [28]. Moreover, synaptosomal mitochondriawere unable to restore normal ATP production after 30 min ofreoxygenation. Because the maintenance of ATP production isassociated with apoptosis, whereas a rapid decline in mitochon-

drial function and ATP is associated with necrosis [29], we canassume that necrosis prevailed.

We also observed depolarization of the mitochondrialmembrane, which was maintained after the reoxygenationperiod. It is well known that both ΔΨm and acidification of thematrix are responsible for opening of the mitochondrial PTP,consisting of the mitochondrial outer membrane voltage-dependent anion channel (VDAC), the inner membrane adeninenucleotide translocase, and the mitochondrial benzodiazepinereceptor [5,30]. In previous studies, we observed that tyrosinephosphorylation in the VDAC was increased after hypoxia[31,32]. The inability to restore the normal ΔΨm after thereoxygenation period was probably due to the classical high-conductance state of the pore, which is responsible for the earlysteps in the apoptotic process [30]. The intracellular calciumconcentration markedly increases in neuronal cells during andafter hypoxia due to inflow from interstitial spaces and releasefrom intracellular stores [33,34]. In synaptosomes, we observeda slight, nonsignificant increase in [Ca2+]i after hypoxia andreoxygenation. This agrees with Keelan et al. [35], whoexplained the maintenance of normal [Ca2+]i with the fact thatATP remained above a critical threshold.

Iron homeostasis is closely regulated by iron-regulatoryproteins (IRP) 1 and 2, which operate at high (IRP1) and low(IRP2) oxygen concentrations [36]. An imbalance between theaccumulation and the release of iron in free form is responsiblefor damage to nerve cells [8]. After 2 h hypoxia, we observed ahighly significant increase in desferoxamine-chelatable freeiron, which represents a potential liability to synaptosomesbecause of the participation of iron in the metabolism ofreactive oxygen species. This agrees with our previous resultson hypoxic erythrocytes [37]. Moreover the role of free iron inbrain damage is confirmed by a recent work of Freret et al. [38]demonstrating a reduction in brain damage after desferoxaminetreatments in rats subjected to transient focal cerebral ischemia.Cells possess a variety of antioxidant species involved in themaintenance of redox status [16]. However, the production ofsuperoxide ion radicals in synaptosomes, where mitochondriaare more concentrated, could improve the functional behaviorof the synaptic termination and render the neuron morevulnerable.

Together with the increase in free iron, we observed asignificant increase in F2-isoprostanes. These substances arehighly accurate quantitative markers of oxidative stress, as theyare derived from free radical-initiated peroxidation of arachi-donic acid [39]. Reactive products of lipid peroxidation areserious contributors to neurodegenerative diseases, but their useas in vivo biomarkers is limited because of their chemicalinstability. Thus, isoprostanes, which are more chemically andmetabolically stable, represent good in vivo biomarkers ofoxidative damage [40]. Oxidative stress plays a key role duringand after the early developmental phases as demonstrated in ourprevious work [41]. Moreover, the change in the mitochondrialpotential leads to overproduction of superoxide ion radicals and,as the synaptic region of the neuron uses more energy andmitochondria are more concentrated in this region, synapto-somes become more vulnerable to oxidative stress. On the

1755C. Aldinucci et al. / Free Radical Biology & Medicine 42 (2007) 1749–1756

whole, our results demonstrated that the incubation of synap-tosomes in 2%O2 caused a decrease in oxygen consumption andATP production by the synaptosomal mitochondria, followed byan increase in free iron and F2-isoprostanes due to the increase insuperoxide ion radicals, which are the most potent inducers ofPTP opening. PTP opening inevitably entails dissipation ofΔΨm with the cessation of ΔΨm-dependent importation ofmitochondrial protein precursors. In our experimental system,because the reparation pathways were reduced, the mitochondriaunderwent mitoptosis.

Acknowledgments

This work was supported in part by PAR Progetti 2004 and inpart by MIUR, PRIN 2005 PROT.2005 06 9071-005.

References

[1] Banasiak, K. J.; Xia, Y.; Haddad, G. G. Mechanisms underlying hypoxia-induced neuronal apoptosis. Prog. Neurobiol. 62:215–249; 2000.

[2] Kroemer, G.; Reed, J. C. Mitochondrial control of cell death. Nat. Med.6:513–519; 2000.

[3] Desagher, S.; Martinou, J.-C. Mitochondria as the central control point ofapoptosis. Trends Cell Biol. 10:369–377; 2004.

[4] Maia, F. D.; Pitombeira, B. S.; Araujo, D. T.; Cunha, G. M.; Viana, G. S.L-Deprenyl prevents lipid peroxidation and memory deficits produced bycerebral ischemia in rats. Cell. Mol. Neurobiol. 24:87–100; 2004.

[5] Lemasters, J. J.; Holmuhamedov, E. Voltage-dependent anion channel(VDAC) as mitochondrial governator—Thinking outside the box. Bio-chim. Biophys. Acta 1762:181–190; 2006.

[6] Skulachev, V. P. Bioenergetic aspects of apoptosis, necrosis and mitoptosis.Apoptosis 11:473–485; 2006.

[7] Lin, M. T.; Flint Beal, M. Mitochondrial dysfunction and oxidative stressin neurodegenerative diseases. Nature 443:787–795; 2006.

[8] Kress, G. J.; Dineley, K. E.; Reynolds, I. J. The relationship betweenintracellular free iron and cell injury in cultured neurons, astrocytes, andoligodendrocytes. J. Neurosci. 22:5848–5855; 2002.

[9] Lipscomb, D. C.; Gorman, L. G.; Traystman, R. J.; Hurn, P. D. Lowmolecular weight iron in cerebral ischemic acidosis in vivo. Stroke29:487–492; 1998.

[10] Allen, D. R.; Wallis, G. L.; McCay, P. B. Catechol adrenergic agentsenhance hydroxyl radical generation in xanthine oxidase systems con-taining ferritin: implications for ischemia/reperfusion. Arch. Biochem.Biophys. 315:235–243; 1994.

[11] Montuschi, P.; Barnes, P. J.; Roberts, L. J. Isoprostanes: markers andmediators of oxidative stress. FASEB J. 18:1791–1800; 2004.

[12] Morrow, J. D.; Awad, J. A.; Boss, H. J.; Blair, I. A.; Roberts, L. J. Non-cyclooxygenase-derived prostanoids (F2-isoprostanes) are formed in situon phospholipids. Proc. Natl. Acad. Sci. USA 89:10721–10725; 1992.

[13] Meini, A.; Benocci, A.; Frosini, M.; Sgaragli, G. P.; Blanco Garcia, J.;Pessina, G. P.; Aldinucci, C.; Palmi, M. Potentiation of intracellular Ca2+

mobilization by hypoxia-induced NO generation in rat brain striatal slicesand human astrocytoma U-373 MG cells and its involvement in tissuedamage. Eur. J. Neurosci. 17:692–700; 2003.

[14] Almeida, A.; Brooks, K. G.; Sammut, I.; Keelan, J.; Davey, G. P.; Clark,J. B.; Bates, T. E. Postnatal development of the complexes of the electrontransport chain in synaptic mitochondria from rat brain. Dev. Neurosci.17:212–218; 1995.

[15] Gutierrez-Martin, Y.; Martin-Romero, F. J.; Henao, F.; Gutierrez-Merino,C. Synaptosomal plasma membrane Ca(2+) pump activity inhibition byrepetitive micromolar ONOO(−) pulses. Free Radic. Biol. Med. 32:46–55;2002.

[16] Joshi, G.; Sultana, R.; Perluigi, M.; Allan, B. D. In vivo protection ofsynaptosomes from oxidative stress mediated by Fe2+/H2O2 or 2,2-

azobis-(2-amidinopropane) dihydrochloride by the glutathione mimetictricyclodecan-9-yl-xanthogenate. Free Radic. Biol. Med. 38:1023–1031;2005.

[17] Kristian, T.; Gertsch, J.; Bates, T. E.; Siesjo, B. K. Characteristics ofthe calcium-triggered mitochondrial permeability transition in nonsy-naptic brain mitochondria: effect of cyclosporin A and ubiquinone O.J. Neurochem. 74:1999–2009; 2000.

[18] Perez-Pinzon, M. A.; Sick, T. J.; Rosenthal, M. Mechanism(s) ofmitochondrial hyperoxidation after global cerebral ischemia. Adv. Exp.Med. Biol. 471:175–180; 1999.

[19] Gornall, A. G.; Bardawil, C. J.; David, M. M. Determination of serumprotein by means of the biuret reaction. J. Biol. Chem. 177:751–766;1949.

[20] Hesketh, T. R.; Smith, G. A.; Moore, J. P.; Taylor, M. V.; Metcalfe, J. C.Free cytoplasmic calcium concentration and the mitogenic stimulation oflymphocytes. J. Biol. Chem. 258:4876–4882; 1983.

[21] Tsien, R. Y.; Pozzan, T.; Rink, T. J. Calcium homeostasis in intactlymphocytes: cytoplasmic free calcium monitored with a new, intracellu-larly trapped fluorescent indicator. J. Cell Biol. 94:325–338; 1982.

[22] Grynkiewicz, G.; Poenie, M.; Tsien, R. Y. A new generation of Ca2+

indicators with greatly improved fluorescence properties. J. Biol. Chem.260:3440–3450; 1985.

[23] Ferrali, M.; Ciccoli, L.; Signorini, C.; Comporti, M. Iron release anderythrocyte damage in allyl alcohol intoxication in mice. Biochem.Pharmacol. 40:1485–1490; 1990.

[24] Comporti, M.; Signorini, C.; Leoncini, S.; Buonocore, G.; Rossi, V.;Ciccoli, L. Plasma F2-isoprostanes are elevated in newborns andinversely correlated to gestational age. Free Radic. Biol. Med. 37:724–732; 2004.

[25] Palop, J. J.; Chin, J.; Mucke, L. A network dysfunction perspective onneurodegenerative diseases. Nature 443:768–773; 2006.

[26] Gorman, A. M.; Ceccatelli, S.; Orrenius, S. Role of mitochondria inneuronal apoptosis. Dev. Neurosci. 22:348; 2000.

[27] Whittaker, V. P. Thirty years of synaptosome research. J. Neurocytol.22:735–742; 1993.

[28] Allen, K. L.; Almeida, A.; Bates, T. E.; Clark, J. B. Changes of respiratorychain activity in mitochondrial and synaptosomal fractions isolated fromthe gerbil brain after graded ischaemia. J. Neurochem. 64:2222–2229;1995.

[29] Budd Haeberlein, S. L. Mitochondrial function in apoptotic neuronal celldeath. Neurochem. Res. 29:521–530; 2004.

[30] Ly, J. D.; Grubb, D. R.; Lawen, A. The mitochondrial membrane potential(deltapsi(m)) in apoptosis: an update. Apoptosis 8:115–128; 2003.

[31] Buonocore, G.; Liberatori, S.; Bini, L.; Mishra, O. P.; Delivoria-Papadopoulos, M.; Pallini, V.; Bracci, R. Hypoxic response of synapto-somal proteins in term guinea pig fetuses. J. Neurochem. 73:2139–2148;1999.

[32] Liberatori, S.; Canas, B.; Tani, C.; Bini, L.; Buonocore, G.; Godovac-Zimmermann, J.; Mishra, O. P.; Delivoria-Papadopoulos, M.; Bracci, R.;Pallini, V. Proteomic approach to the identification of voltage-dependentanion channel protein isoforms in guinea pig brain synaptosomes. Pro-teomics 4:1335–1340; 2004.

[33] Mishra, O. P.; Delivoria-Papadopoulos, M. Inositol tetrakisphosphate(IP4)- and inositol triphosphate (IP3)-dependent Ca2+ influx in corticalneuronal nuclei of newborn piglets following graded hypoxia. Neurochem.Res. 29:391–396; 2004.

[34] Jeong, J. I.; Lee, Y. W.; Kim, Y. K. Chemical hypoxia-induced cell death inhuman glioma cells: role of reactive oxygen species, ATP depletion,mitochondrial damage and Ca2+. Neurochem. Res. 28:1201–1211; 2003.

[35] Keelan, J.; Bates, T. E.; Clark, J. B. Intrasynaptosomal free calciumconcentration during rat brain development: effects of hypoxia, aglycae-mia, and ischaemia. J. Neurochem. 66:2460–2467; 1996.

[36] Meyron-Holtz, E. G.; Ghosh, M. C.; Rouault, T. A. Mammalian tissueoxygen levels modulate iron-regulatory protein activities in vivo. Science306:2087–2090; 2004.

[37] Ciccoli, L.; Rossi, V.; Leoncini, S.; Signorini, C.; Blanco-Garcia, J.;Aldinucci, C.; Buonocore, G.; Comporti, M. Iron release, superoxideproduction and binding of autologous IgG to band 3 dimers in newborn

1756 C. Aldinucci et al. / Free Radical Biology & Medicine 42 (2007) 1749–1756

and adult erythrocytes exposed to hypoxia and hypoxia–reoxygenation.Biochim. Biophys. Acta 1672:203–213; 2004.

[38] Freret, T.; Valable, S.; Chazalviel, L.; Saulnier, R.; Mackenzie, E. T.; Petit,E.; Bernaudin, M.; Boulouard, M.; Schumann-Bard, P. Delayed admin-istration of deferoxamine reduces brain damage and promotes functionalrecovery after transient focal cerebral ischemia in the rat. Eur. J. Neurosci.23:1757–1765; 2006.

[39] Montine, T. J.; Montine, K. S.; Reich, E. E.; Terry, E. S.; Porter, N. A.;Morrow, J. D. Antioxidants significantly affect the formation of different

classes of isoprostanes and neuroprostanes in rat cerebral synaptosomes.Biochem. Pharmacol. 65:611–617; 2003.

[40] Montine, K. S.; Quinn, J. F.; Zhang, J.; Fessel, J. P.; Roberts, L. J.; Morrow,J. D.; Montine, T. J. Isoprostanes and related products of lipid peroxidationin neurodegenerative diseases. Chem. Phys. Lipids 128:117–124; 2004.

[41] Longini, M.; Perrone, S.; Kenanidis, A.; Vezzosi, P.; Marzocchi, B.;Petraglia, F.; Centini, G.; Buonocore, G. Isoprostanes in amniotic fluid: apredictive marker for fetal growth restriction in pregnancy. Free Radic.Biol. Med. 38:1537–1541; 2005.