www.elsevier.com / locate / bres
Research report
Mechanisms underlying H O -mediated inhibition of synaptic
2 2transmission in rat hippocampal slices
*
Marat V. Avshalumov, Billy T. Chen, Margaret E. Rice
Departments of Physiology and Neuroscience and Neurosurgery, New York University School of Medicine, 550 First Avenue, New York, NY 10016,
USA
Accepted 8 August 2000
Abstract
Hydrogen peroxide (H O ) inhibits the population spike (PS) evoked by Schaffer collateral stimulation in hippocampal slices.2 2 21
Proposed mechanisms underlying this effect include generation of hydroxyl radicals (?OH) and inhibition of presynaptic Ca entry. We have examined these possible mechanisms in rat hippocampal slices. Inhibition of the evoked PS by H O was sharply concentration-2 2
dependent: 1.2 mM H O had no effect, whereas 1.5 and 2.0 mM H O reversibly depressed PS amplitude by roughly 80%. The iron2 2 2 2 chelator, deferoxamine (1 mM), and the endogenous?OH scavenger, ascorbate (400 mM), prevented PS inhibition, confirming ?OH involvement. Isoascorbate (400mM), which unlike ascorbate is not taken up by brain cells, also prevented PS inhibition, indicating an extracellular site of ?OH generation or action. We then investigated whether H O -induced PS depression could be overcome by2 2
21
prolonged stimulation, which enhances Ca entry. During 5-s, 10-Hz trains under control conditions, PS amplitude increased to over 200% during the first three–four pulses, then stabilized. In the presence of H O , PS amplitude was initially depressed, but began to2 2
recover after 2.5 s of stimulation, finally reaching 80% of the control maximum. In companion experiments, we assessed the effect of
21 21 21
H O on presynaptic Ca2 2 entry by monitoring extracellular Ca concentration ([Ca ] ) during train stimulation in the presence ofo
21 21
postsynaptic receptor blockers. Evoked [Ca ] shifts were apparently unaltered by H O , suggesting a lack of effect on Cao 2 2 entry. Taken together, these findings suggest new ways in which reactive oxygen species (ROS) might act as signaling agents, specifically as modulators of synaptic transmission. 2000 Elsevier Science B.V. All rights reserved.
Theme: Excitable membranes and synaptic transmission
Topic: Presynaptic mechanisms
21
Keywords: Hydrogen peroxide; Oxidative stress; Transmitter release; Ion-selective microelectrode; Ca ; Ascorbate
1. Introduction those involving kinase and phosphatase enzymes, which are sensitive to this ROS [10,23,30,51].
Reactive oxygen species (ROS) occur naturally during It has also been shown that H O can modulate neuronal2 2
cell metabolism and are strongly regulated by the intracel- activity. The effect of H O on hippocampus slice physi-2 2
lular antioxidant network. Under pathological conditions, ology has been particularly well-studied [13,21,23,29,35– including ischemia, hyperoxia, and exposure to certain 37,46]. Both the primary effect and the mechanism of toxins, ROS can cause damage to the central nervous action of H O2 2 in hippocampal slices, however, remain system [14,17]. Recent studies, however, have suggested uncertain. Several reports in the literature indicate that that ROS, particularly hydrogen peroxide (H O ), might2 2 H O2 2 causes a reversible depression of the population also play a role in cellular signaling [12]. Among the most spike (PS) recorded in CA1 stratum pyramidale during important signaling pathways modulated by H O2 2 are stimulation of the Schaffer collaterals [13,35–37], although an augmentation in hippocampal PS amplitude has also been reported [21]. In addition, it is not yet certain whether
*Corresponding author. Tel.: 11-212-263-5438; fax: 1
1-212-689-PS depression is mediated directly by H O or by
hy-0334. 2 2
E-mail address: margaret.rice@nyu.edu (M.E. Rice). droxyl radicals (?OH) that can be produced from the
interaction of H O with tissue iron or copper through the2 2 tobarbital sodium. After decapitation, the brain was rapidly
Fenton reaction [7,16,17]: removed and placed in ice-cold, oxygenated (95% O / 5%2
CO ) artificial cerebrospinal fluid (ACSF) for about 1 min,2
n1 (n11 )1 2
H O2 21Me →Me 1 ?OH1OH , then bisected, blocked and the tissue mounted on the stage
of a Vibratome (Ted Pella, St. Louis, MO, USA).
Trans-n1
where Me is the reduced form of the metal ion. Pellmar
verse hippocampal slices were cut in ice-cold ACSF that and colleagues concluded that PS depression was mediated
contained (in mM): 120 NaCl, 5 KCl, 35 NaHCO , 1.253
by?OH, because deferoxamine, an iron chelator, prevented
NaH PO , 1.5 CaCl , 1.3 MgCl , and 10 glucose. Slices2 4 2 2
H O -induced changes in PS amplitude [37]. In other2 2 were maintained in an incubation chamber at room
tem-experiments, however, although with lower concentrations
perature for at least 1 h before recording. of deferoxamine, Katsuki and colleagues [21] concluded
that ?OH was not involved in the effects of H O2 2 on
2.2. Extracellular recording hippocampal slice physiology.
The site of action of H O in mediating PS depression is2 2
For the measurement of evoked potentials, an individual also not yet clear, although there is strong evidence for a
slice was transferred to a submersion-recording chamber presynaptic location: H O had no effect on antidromically2 2
(Warner Instrument, Hamden, CT, USA), where it was evoked PS in CA1 stratum pyramidale [34]; nor did it alter
21
continuously superfused with ACSF at 1.5 ml min at EPSPs evoked by glutamate iontophoresis [35]. These data
318C. Extracellular PSs were elicited by stimulating the led Pellmar to propose that H O might inhibit transmitter2 2
21 Schaffer collaterals and recorded in the stratum pyramidale
release by decreasing presynaptic Ca entry [35].
Recent-of CA1 with conventional glass electrodes (2–7 MV tip ly, our laboratory showed that H O can decrease stimu-2 2
21 resistance, backfilled with 1 M NaCl) connected to an
lated dopamine release in striatal slices in a Ca
-depen-Axoprobe 1A amplifier (Axon Instruments, Foster City, dent manner, consistent with a presynaptic site of action of
CA, USA). Pulse duration was 100ms, with the stimulus H O [6].2 2
intensity adjusted to the lowest level (0.2–1.3 V) required In the present studies, we tested the hypothesis that
to evoke a PS of maximal amplitude. The stimulating generation of ?OH is required for the action of H O and2 2
21 electrode was a twisted bipolar electrode, made from
investigated whether decreased presynaptic Ca entry
Teflon-insulated platinum–iridium wire. In some experi-might be the mode of action. To ascertain whether PS
ments, 10-Hz pulse trains of 5-s duration were used, with depression is mediated by ?OH, the effect of H O was2 2
the same pulse duration and amplitude as in single pulse assessed in the presence of deferoxamine [37], with
experiments. In all slices, the evoked PS was monitored for concomitant determination of H O2 2 concentration in the
25–30 min in normal ACSF to ascertain that the response presence of this chelator. In addition, we tested whether
was stable; only slices with stable PS responses during this the endogenous?OH scavenger, ascorbate [2,4,8,40], could
control period were tested further. For each experimental modulate the effect of H O . To determine whether the2 2
condition tested with single pulse stimulation, three evoked action of ascorbate was intracellular or extracellular, H O2 2
PS records were stored and averaged using locally written was applied in the presence of the stereoisomer of
ascor-software. The amplitude (in mV) of these averaged PS bate, isoascorbate (D-ascorbate), which has similar
electro-records was measured from the mean of the positive peak chemical properties to ascorbate, but is not a substrate for
preceding and the positive peak following the negative PS. the stereoselective ascorbate transporter [48].
Consequent-Acquisition of pulse train experiments data was controlled ly, isoascorbate is not taken up by cells and remains in the
by CLAMPEX 7.0 software (Axon Instruments) which
im-extracellular compartment [4]. In companion experiments,
21 ported PS records in the computer via a DigiData 1200B we evaluated the effect of H O on presynaptic Ca2 2 entry
21 21 D/A board (Axon Instruments).
by monitoring extracellular Ca concentration ([Ca ] )o
using ion-selective microelectrodes (ISMs) [31,42].
21
2.3. Ca ion-selective microelectrode recording
21
Stimulated [Ca ] shifts in hippocampal slices were
2. Materials and methods o
21
monitored using Ca -sensitive ion-selective
microelec-21 21
2.1. Hippocampal slice preparation trodes (Ca -ISMs). Ca -ISMs were positioned in
21
stratum radiatum of CA1; Ca entry was monitored Hippocampal slices (400-mm thickness) were prepared during 10-Hz pulse trains of 5 s duration in the presence of from young adult (50–60 days old), male Long–Evans postsynaptic receptor blockers: picrotoxin, 50 mM; AP5,
21 rats. All animal experimentation was conducted following 50mM; and CNQX, 25 mM. Double-barreled Ca -ISMs NIH guidelines and with approval by the NYU School of were prepared using theta glass (Warner Instrument) and Medicine Institutional Animal Care and Use Committee. calibrated as described previously [31,42]. The ion-sensing
21
Ronkon-koma, NY, USA) as the ion exchanger and was backfilled Students t-test or one-way ANOVA followed by Student– with 100 mM CaCl . The reference barrel was backfilled2 Newman–Keuls test was used as appropriate. The thres-with 150 mM NaCl. Extracellular potentials recorded thres-with hold of statistical significance was considered to be P,
the reference barrel were subtracted from the ion signals 0.05.
21 using the AxoProbe 1A amplifier. Stimulated [Ca ]o
shifts were recorded on a chart recorder and stored also on
21
computer. The maximum amplitude of [Ca ] shifts waso 3. Results
measured from chart records and calculated with respect to
21
the concentration of Ca in ACSF using the Nernst 3.1. Concentration dependence of H O -mediated PS2 2
equation [31]. depression
2.4. Hydrogen peroxide and ascorbate analyses Previously it was shown that H O (1.0–3.0 mM) can2 2
inhibit the Schaffer collateral-evoked PS in slices of guinea Actual media concentrations of H O were determined2 2 pig hippocampus [35–37]. We first determined whether in these experiments using a commercially available kit H O concentrations in this range were also effective in rat2 2
(Oxis International, Portland, OR, USA). This quantitative hippocampal slices. The effect of 15-min exposure to assay is a spectrophotometric method based on the oxida- H O on PS amplitude was sharply concentration depen-2 2
21 31
tion of ferrous ions (Fe ) to ferric ions (Fe ) by H O .2 2 dent: 1.2 mM H O2 2 had no significant effect (n56), The ferric ions bind to the indicator dye xylenol orange whereas 1.5 mM H O2 2 caused a significant (P,0.001), (3,39-bis[N,N -di(carboxymethyl)-aminomethyl]-o -cresol- reversible decrease in PS amplitude to 1862% (n519) of sulfone-phthalein, sodium salt) to form a stable colored control (Fig. 1). Similarly, 2.0 mM H O2 2 produced a complex, which was monitored at 560 nm using a Turner reversible depression of the PS, but only to the same extent 340 spectrophotometer (Barnstead / Thermolyne, Dubuque, as 1.5 mM H O (Fig. 1). In all subsequent experiments,2 2
IA, USA). therefore, 1.5 mM H O was considered to be the mini-2 2
Ascorbate and isoascorbate concentrations in ACSF in mum effective concentration. the presence and absence of H O were determined using2 2
a previously described HPLC method [4,41]. Media sam- 3.2. Influence of iron chelation ples collected from the holding flask or from the recording
chamber were immediately diluted 1:10 in ice-cold, de- It is known that deferoxamine, a metal ion chelator, oxygenated eluent, then kept on ice until analysis. Under prevents formation of ?OH from H O [16]. To determine2 2
these conditions, ascorbate samples were stable for at least the involvement of ?OH in mediating the effect of H O2 2
1 h. on PS amplitude in rat hippocampal slices, we investigated
whether the effect persisted in the presence of this chelator.
2.5. Chemicals Deferoxamine (1 mM) alone did not alter PS amplitude
(not illustrated); records obtained after 30 min of deferox-ACS-grade hydrogen peroxide solution, 8.8 M (30%), amine superfusion served as controls in these experiments. was obtained from Sigma (St. Louis, MO, USA). This Application of nominally 1.5 mM H O in the presence of2 2
stock solution was diluted in ACSF to the following 1 mM deferoxamine had no effect on PS amplitude (Fig. experimental concentrations in ACSF immediately before 2A). At face value, these data suggested that the PS slice application: 1.2 mM (0.004%); 1.5 mM (0.005%); decrease was mediated by ?OH.
2.0 mM (0.007%). Deferoxamine mesylate, ascorbic acid An important control, however, which had not been and isoascorbic acid, glutathione (GSH), picrotoxin, AP5 reported in previous studies of the effect of H O2 2 on (D,L-2-amino-5-phosphonopentanoic acid) were also pur- hippocampal slice physiology, was to determine actual
chased from Sigma; CNQX (6-cyano-7-nitroquioxaline- concentrations of H O under varying conditions. To do2 2
2,3-dione) was obtained from Tocris (Ballwin, MO, USA). so, we took samples of H O solutions immediately after2 2
Solutions containing these agents were also freshly pre- preparation and also from the recording chamber at the end
pared before application. of 15-min superfusion. Immediately after preparation,
actual initial concentrations of H O were the same as the2 2
2.6. Experimental design and statistical analysis nominal levels (Table 1). Concentrations of H O in the2 2
recording chamber after 15-min superfusion were slightly In all experiments, the amplitude of the stabilized PS lower than initial levels, with a decrease of 0.1 mM evoked by single pulse stimulation in ACSF alone was reaching significance (P,0.05; n55) for 1.5 mM H O in2 2
considered to be control (100%). Mean PS amplitude data ACSF alone. For nominally 1.5 mM H O2 21deferoxamine, are presented as percent of control6S.E.M.; concentrations initial H O2 2 concentrations were not altered by this
21
Fig. 2. Inhibition of H O -mediated PS depression by metal ion chela-2 2
tion. (A) Representative electrophysiological records of hippocampal PS in CA1 (each is the average of three records). After 30-min pretreatment Fig. 1. Reversible inhibition of the evoked population spike (PS) in rat
with 1 mM deferoxamine (Def), H O had no effect on evoked PS hippocampal slices by H O . (A) Electrophysiological records from CA12 2 2 2
amplitude (compare with Fig. 1A) during application of either 1.5 or 2.0 in stratum pyramidale evoked by stimulation of the Schaffer collaterals
mM H O . An initial concentration of 2.0 mM was required to maintain before H O exposure (control), after 15-min superfusion with 1.2, 1.5, or2 2 2 2
an H O concentration in the superfusion chamber of at least 1.5 mM (see 2.0 mM H O , and after 30-min washout of H O (each is the average of2 2 2 2 2 2
Section 3.2). (B) Average PS amplitude of the changes induced by H O three PS records). (B) PS amplitude changes (as % control) after 15-min 2 2
(2.0 mM initially) in the presence (treated) and absence (untreated) of exposure to H O and following 30-min washout (*** P2 2 ,0.001
com-deferoxamine (% control). The duration of com-deferoxamine application is pared with controls; n56 for 1.2 mM H O , n2 2 519 for 1.5 mM, and
indicated by the white bar, H O application is indicated by the black bar,
n515 for 2.0 mM). Data are mean6S.E.M.; dotted line indicates pre- 2 2
and washout by the striped bar (data are mean6S.E.M.; *** P,0.001 H O control amplitude (100%).2 2
indicates difference between PS amplitude in H O alone and in H O2 2 2 21
deferoxamine).
15-min superfusion decreased by 0.3 mM (n55; P,
0.001).
Because we had noted that 1.2 mM H O did not alter2 2 and its non-physiological stereoisomer, isoascorbate, could PS amplitude (Fig. 1), we repeated evaluation of the effect also afford protection from H O .2 2
of deferoxamine in the presence of initially 2.0 mM H O .2 2 Initial studies with ascorbate and isoascorbate indicated Although the concentration of H O2 2 again fell after 15- that both caused a decrease in H O2 2 concentration (P,
min superfusion in the presence of deferoxamine, an 0.05) (Table 1). To compensate for the decrease in H O2 2
effective level, greater than 1.5 mM, was maintained concentration and to be consistent with our deferoxamine (Table 1). Importantly, under these conditions, deferox- studies, we used initially 2.0 mM H O2 2 for further amine also prevented the H O -induced depression of the2 2 experiments with these antioxidants. Under these
con-PS (Fig. 2A, B). ditions, the concentration of H O2 2 remained above 1.5
mM (Table 1). Somewhat surprisingly, ascorbate was relatively stable in the presence of H O . Ascorbate2 2
3.3. Effect of?OH scavengers on the H O -induced2 2 concentration in the chamber after 15-min superfusion with
depression of the evoked PS H O2 2 was 360622 mM (n55), which was only 10% below the initial concentration of 400mM. The stability of Prevention of H O -induced PS suppression by deferox-2 2 isoascorbate was similar, with 354619 mM (n55) after amine confirmed the involvement of ?OH. We then as- 15-min superfusion with H O .2 2
Table 1
a
H O concentration in the absence and in the presence of 1 mM deferoxamine, 4002 2 mM ascorbate, or 400mM isoascorbate
Nominal H O concentration (mM)2 2 Initial [H O ]2 2 [H O ] after 15-min superfusion2 2
Data are mean6S.E.M.; n55. Samples were taken immediately after preparation of the H O solution at room temperature and from the recording2 2
chamber at 318C, after 15-min superfusion.
b
Def, deferoxamine; Asc, ascorbate; IsoAsc, isoacorbate.
c
P,0.05with respect to the initial concentration.
d
P,0.001 with respect to the initial concentration.
e
P,0.001 compared with H O concentration in the recording chamber in ACSF alone.2 2
3A). Indeed, the average PS amplitude in the presence of was intracellular or extracellular, H O2 2 was applied in ascorbate1H O did not differ significantly from control2 2 presence of isoascorbate. Like ascorbate, isoascorbate also (Fig. 3B). To determine whether the action of ascorbate prevented PS suppression by H O , indicating an extracel-2 2
lular component to the effect.
Endogenous GSH, another biologically relevant scavenger of ?OH [8], has been shown previously to enhance recovery in this H O model of oxidative stress2 2
[38]. When co-applied with H O , however, the con-2 2
centration of GSH in the presence of H O2 2 rapidly fell below detection limits (low micromolar levels). Unsurpris-ingly, therefore, GSH (initially 400 mM) did not protect against the effect of 2.0 H O on the PS amplitude (not2 2
illustrated).
21 3.4. Pulse train stimulation and presynaptic Ca entry
21
Decreased presynaptic Ca entry has been proposed as a possible mechanism for the inhibitory effect of H O on2 2
hippocampal PS amplitude [35]. To evaluate whether
21
increased Ca influx during prolonged stimulation could overcome the H O -induced PS depression, we examined2 2
evoked PS during pulse train stimulation (5 s, 10 Hz) in the presence of 1.5 mM H O . We also assessed the effect2 2
21
of H O on presynaptic Ca2 2 influx during train
stimula-21 21
tion by monitoring [Ca ] using Cao -ISMs. Prior to the delivery of a pulse train, single pulses were used to elicit a PS with stable amplitude. The average amplitude of three single PSs served as the control against which PS am-plitudes during pulse train stimulation were compared.
In ACSF alone, 10-Hz stimulation for 5 s caused a twofold increase in PS amplitude that was maximal after
Fig. 3. Inhibition of H O -induced PS depression by ascorbate and2 2
three to four pulses, then was constant until the end of the
isoascorbate. (A) Representative electrophysiological records of
hip-train (Fig. 4). The amplitude of the last pulse of the hip-train
pocampal PS in CA1 (each is the average of three records). The effect of
H O on PS amplitude in ACSF alone was prevented by co-application of2 2 was 213614% of single pulse control (n56; P,0.001). In
ascorbate (Asc, 400 mM) or isoascorbate (IsoAsc, 400 mM) with the presence of H O , PS amplitude was depressed
2 2
nominally 2.0 mM H O . (B) Average changes in PS amplitude (as %2 2 through much of the train (Fig. 4). The average PS
control). The amplitude of the PS did not differ significantly from control
amplitude evoked by the first pulse of the train in H O2 2
when 2.0 mM H O was applied together with ascorbate or isoascorbate2 2
was decreased to 1862% (n56; P,0.001) of the single
(n55), but did in ACSF alone (n515; *** P,0.001). Data are
recovered to 150613% (n56; P,0.05) of single pulse control, although this still differed significantly from the amplitude of the last pulse of the train in ACSF alone (P,0.001) (Fig. 4).
21
To observe presynaptic Ca entry during 5-s, 10-Hz
21
stimulation, Ca -ISM measurements were made in the presence of a cocktail of postsynaptic receptor blockers (see Materials and methods). Under these conditions, the
21
mean evoked decrease in [Ca ] was 52o 64 mM (n56)
21
(Fig. 5). These [Ca ] shifts were unaffected by 1.5 mMo
H O , with an mean fall of 502 2 64mM (n56). The lack of
21
difference between the size of the [Ca ] decrease in theo
absence and presence of H O , as well as the apparently2 2 21
similar time course of the [Ca ] shifts (Fig. 5) arguedo 21
against an effect of H O on presynaptic Ca2 2 entry.
4. Discussion
Previous studies of the effect of H O2 2 on evoked PS amplitude in guinea pig hippocampal slices raised several unanswered questions. Here, we have addressed these questions in slices of rat hippocampus and confirm that H O -induced depression of the evoked PS is mediated by2 2
?OH, but that the mechanism does not appear to involve
21
presynaptic Ca entry per se.
Fig. 4. PS amplitude changes during pulse train stimulation in the
Consistent with previous reports [13,37], we found that
presence and absence of H O . (A) Average PS records for first, 25th and2 2
H O2 2 decreased the amplitude of the PS evoked by
last pulse during 10-Hz stimulation for 5 s (each record is the average
Schaffer collateral stimulation, with a minimum effective
from six slices). (B) Time course of PS amplitude changes during train
stimulation (5 s, 10 Hz) in ACSF alone, then after 15-min exposure to 1.5 concentration of 1.5 mM H O2 2 (Fig. 1). In contrast to
mM H O . Data are mean2 2 6S.E.M. (n56) and are given as a percentage these results, however, Katsuki and colleagues [21] re-of the PS amplitude evoked by single pulse stimulation. Throughout the
ported that 0.3–3.0 mM H O caused an augmentation of2 2
pulse train, PS amplitudes in the presence of H O2 2 were significantly
the PS amplitude in rat hippocampal slices. In the present
lower than those recorded during train stimulation in ACSF alone (P,
studies, no enhancement was seen in any experiment (n5 0.001).
19). This difference in response to H O2 2 might be explained by several differences in experimental conditions s of stimulation), however, PS amplitude had begun to between the previous and present studies. The most recover and was 5265% (n56; P,0.001) of control. By important of these is likely to be that in the earlier study, the end of the train in the presence of H O , PS amplitude2 2 slice incubation temperature during recovery after slicing
21
Fig. 5. Evoked [Ca ] decreases during pulse train stimulation in the presence and absence of H O . Measurements were made in the presence ofo 2 2 21
postsynaptic receptor antagonists (50mM AP5, 25mM CNQX, and 50mM picrotoxin) so that the recorded [Ca ] shifts during train stimulation reflectedo
21 21
primarily presynaptic Ca entry (see Materials and methods). The control record is the average of three [Ca ] shifts recorded during 5-s, 10-Hzo
21 21
stimulation in ACSF1receptor blockers; the [Ca ] shift in the presence of H O is the average of [Cao 2 2 ] records taken at the same site after 10-, 15-,o 21
and 20-min superfusion with 1.5 mM H O in the continued presence of the receptor blockers. Maximum Ca2 2 influx was not affected by H O (see2 2 21
and in the incubation chamber was 378C, whereas in our 4.2. Dissociation of the H O effect on PS amplitude2 2 21
studies, recovery was at room temperature and recording and Ca entry during pulse train stimulation
was at 318C. The lower temperatures of recovery and
21
incubation presumably kept slices in more viable condition Presynaptic Ca influx through voltage-sensitive cal-[11]. The fact that H O -induced PS depression was2 2 cium channels is essential for the coupling of excitation reversible upon washout of H O is further consistent with2 2 and neurotransmitter release [15]. There is a strong,
21
maintained slice viability following H O exposure.2 2 positive relationship between the extent of Ca entry and the amount of transmitter released, with enhancement of both during prolonged stimulation [20,39,47,53]. The 4.1. The effect of H O is mediated by2 2 ?OH present results are consistent with this basic relationship. Under control conditions during 10-Hz train stimulation, Many studies of the effects H O on a variety of in vivo2 2 PS amplitude doubled within the first few pulses of the and in vitro systems have concluded that the effective train, then remained constant for the rest of the stimulation
21 agent is ?OH, because iron chelators, like deferoxamine, (Fig. 4). With this same stimulus paradigm, [Ca ]o
could prevent the H O -dependent effect [24,33,37,44,54].2 2 showed an initial rapid decrease, then continued to fall In the present studies, H O had no effect on PS amplitude2 2 gradually through the stimulus train (Fig. 5).
in the presence of deferoxamine, confirming that ?OH, One mechanism by which H O2 2 might inhibit
neuro-21 rather than H O itself, was responsible for the observed2 2 transmission has been proposed to be decreased Ca entry PS depression. In these experiments, unlike previous at presynaptic terminals [35]. Consistent with this
hypoth-21 1
studies, we monitored and maintained a normally effective esis, Ca influx evoked by K depolarization was found concentration of H O2 2 (Table 1). These findings again to be depressed in retina cells under conditions of oxida-differed from those of Katsuki et al. [21], who found no tive stress [1]. In the present studies, however, there was effect of deferoxamine on the H O -mediated enhance-2 2 no obvious change in either the time-course or maximum
21
ment of the hippocampal PS. In that study, however, the amplitude of decreases in [Ca ] during 10-Hz stimula-o
concentration of deferoxamine tested was low (20 mM) tion in the presence of H O2 2 (Fig. 5), even though PS and thus probably below an effective level. amplitude was depressed for over half of the stimulation Further evidence for ?OH involvement came from our period (Fig. 4). This suggests that H O2 2 did not affect
21
finding that ascorbate also prevented H O -induced PS2 2 presynaptic Ca -channels, but rather acted downstream
21
depression. This antioxidant is an effective scavenger of from Ca entry. One possible mechanism might be
?OH, as well as of peroxyl radical and superoxide [2,8,40]. oxidation of redox sensitive sites in vesicle fusion proteins, Importantly, ascorbate was effective within its normal which could disrupt the interaction of components of
21
range of extracellular concentration, which is 200–400mM Ca -dependent vesicle fusion machinery [15,43]. The
[28,40,45]. observation that PS depression begins to recover after a
The site of ?OH generation in the present studies is not certain point during train stimulation might reflect
con-21
yet entirely clear, although data suggest an extracellular ditions under which sufficient Ca entry has occurred to location. Because H O is highly permeable through the2 2 overcome such disruption, thus permitting transmitter membrane [25], an intra- or extracellular site would be release [43].
possible. The concentration and duration of deferoxamine exposure in the present experiments, however, may not
have been sufficient for this somewhat membrane-perme- 4.3. Physiological relevance of H O -dependent2 2
able iron chelator to enter cells [5,22,54], such that modulation of neurotransmission
deferoxamine could have been acting in either
compart-ment. An extracellular component of the H O2 2 effect, The present results provide new information about the however, was suggested by the results with isoascorbate. possible role of ROS in modulating synaptic transmission. While ascorbate and isoascorbate have identical redox Endogenously, H O is produced during normal oxidative2 2
properties [19,32], only ascorbate is a substrate for the metabolism in the mitochondria, with levels that can stereoselective ascorbate transporter [50] and can enter the exceed 2% of total O2 consumption [3]. In presynaptic intracellular compartment [4,40,48]; isoascorbate, there- terminals, mitochondria are often found within 250 nm of fore, remains extracellular. Because both stereoisomers the presynaptic membrane [9], where they provide a local prevented PS depression by H O , this suggests that either2 2 supply of ATP needed for transmitter release, signal
21
the generation of?OH or its site of action is extracellular. transduction, and regulation of [Ca ]i [18,26,49,52]. Because ?OH is a neutral molecule, however, it would be Consequently, the mitochondria are also well positioned to expected to be membrane-permeable and thus not con- provide a source of endogenously produced H O2 2 to strained to remain in the compartment where it was modulate neurotransmission. Additionally,?OH, which can generated. In addition, it is relevant to consider that the be generated from H O , also has modulatory properties,2 2
phos-[13] J.C. Fowler, Hydrogen peroxide opposes the hypoxic depression of
phorylation reactions [27], which might further alter
evoked synaptic transmission in rat hippocampal slices, Brain Res.
synaptic processes.
766 (1997) 255–258.
In light of these recent findings, understanding processes [14] I. Fridovich, The biology of oxygen radicals, Science 201 (1978) that regulate both ROS generation and endogenous ROS- 875–880.
scavenging capacity takes on new importance. It has long [15] Y. Goda, T.C. Sudhof, Calcium regulation of neurotransmitter¨ release: reliably unreliable?, Curr. Opin. Cell. Biol. 9 (1997) 513–
been recognized that redox balance is critical for the
518.
prevention of oxidative damage. The present data show
[16] E. Graf, J.R. Mahoney, R.G. Bryant, J.W. Eaton, Iron-catalyzed
that regulation of ROS, for example by antioxidants like hydroxyl radical formation. Stringent requirement for free iron ascorbate, may also be critical for fine-tuning ROS- coordination site, J. Biol. Chem. 259 (1984) 3620–3624. mediated modulation of normal physiological processes, [17] B. Halliwell, Reactive oxygen species and the central nervous
system, J. Neurochem. 59 (1992) 1609–1623.
including synaptic transmission.
[18] R. Heidelberger, Adenosine triphosphate and the late steps in calcium-dependent exocytosis at a ribbon synapse, J. Gen. Physiol. 111 (1998) 225–241.
Acknowledgements [19] E.N. Iheanacho, N.H. Hunt, R. Stocker, Vitamin C redox reactions in blood of normal and malaria-infected mice studied with isoascorbate as a nonisotopic marker, Free Radic. Biol. Med. 18 (1995) 543–552.
These studies were supported by NIH / NINDS grants
[20] M. Kano, R. Schneggenburger, A. Verkhratsky, A. Konnerth,
NS 34115 and NS 36362. We appreciate the expert
Depolarization-induced calcium signals in the somata of cerebellar
technical assistance Mei Lan Chao in the HPLC analysis of Purkinje neurons, Neurosci. Res. 24 (1995) 87–95.
ascorbate and isoascorbate. [21] H. Katsuki, C. Nakanishi, H. Saito, N. Matsuki, Biphasic effect of hydrogen peroxide on field potentials in rat hippocampal slices, Eur. J. Pharmacol. 337 (1997) 213–218.
[22] H. Keberle, The biochemistry of desferrioxamine and its relation to
References iron metabolism, Ann. NY Acad. Sci. 119 (1964) 758–768.
[23] E. Klann, E.D. Roberson, L.T. Knapp, J.D. Sweatt, A role for [1] P. Agostinho, C.B. Duarte, A.P. Carvalho, C.R. Oliveira, Oxidative superoxide in protein kinase C activation and induction of long-term
21
stress affects the selective ion permeability of voltage-sensitive Ca potentiation, J. Biol. Chem. 273 (1998) 4516–4522.
channels in cultured retinal cells, Neurosci. Res. 27 (1997) 323– [24] D.A. Long, K. Ghosh, A.N. Moore, C.E. Dixon, P.K. Dash,
334. Deferoxamine improves spatial memory performance following
[2] R.E. Beyer, The role of ascorbate in anti-oxidant protection of experimental brain injury in rats, Brain Res. 717 (1996) 109–117. biomembranes: interaction with vitamin E and coenzyme Q, J. [25] N. Makino, Y. Mochizuki, S. Bannai, Y. Sugita, Kinetic studies on Bioenerg. Biomembr. 26 (1994) 349–358. the removal of extracellular hydrogen peroxide by cultured fi-[3] A. Boveris, B. Chance, The mitochondrial generation of hydrogen broblasts, J. Biol. Chem. 269 (1994) 1020–1025.
peroxide. General properties and effect of hyperbaric oxygen, [26] M.P. Mattson, A.J. Bruce-Keller, Compartmentalization of signaling Biochem. J. 134 (1973) 707–716. in neurons: evolution and deployment, J. Neurosci. Res. 58 (1999) [4] B. Brahma, R.E. Forman, E. E Stewart, C. Nicholson, M.E. Rice, 2–9.
Ascorbate inhibits edema in brain slices, J. Neurochem. 74 (2000) [27] N. Maulik, R.M. Engelman, J.A. Rousou Jr., J.E. Flack, D. Deaton, 1263–1270. D.K. Das, Ischemic preconditioning reduces apoptosis By upregulat-[5] K.R. Bridges, A. Cudkowicz, Effect of iron chelators on the ing anti-death gene bcl-2, Circulation 100 (1999) II369–II375.
transferrin receptor in K562 cells, J. Biol. Chem. 259 (1984) [28] M. Miele, M. Fillenz, In vivo determination of extracellular brain
12970–12977. ascorbate, J. Neurosci. Methods 70 (1996) 15–19.
¨
[6] B.T. Chen, M.V. Avshalumov, M.E. Rice, Calcium modulates the [29] M. Muller, A. Fontana, G. Zbinden, B.H. Gahwiler, Effects of inhibition of striatal dopamine release by H O , Soc. Neurosci.2 2 interferons and hydrogen peroxide on CA3 pyramidal cells in rat Abstr. 25 (1999) 1501. hippocampal slice cultures, Brain Res. 619 (1993) 157–162. [7] C.C. Chiueh, D.L. Murphy, H. Miyake, K. Lang, P.K. Tulsi, S.J. [30] K. Nakamura, T. Hori, N. Sato, K. Sugie, T. Kawakami, J. Yodoi,
Huang, Hydroxyl free radical (?OH) formation reflected by salicyl- Redox regulation of a src family protein tyrosine kinase p56lck in T ate hydroxylation and neuromelanin. In vivo markers for oxidant cells, Oncogene 8 (1993) 3133–3139.
injury of nigral neurons, Ann. NY Acad. Sci. 679 (1993) 370–375. [31] C. Nicholson, M.E. Rice, Use of ion-selective microelectrodes and [8] G. Cohen, Enyzmatic / nonenzymatic sources of oxyradicals and voltammetric microsensors to study brain cell microenvironment, in: regulation of antioxidant defenses, Ann. NY Acad. Sci. 738 (1994) A.A. Boulton, G.B. Baker, W. Walz (Eds.), The Neuronal
Mi-8–14. croenvironment, Neuromethods, Vol. 9, Humana Press, Clifton, NJ,
[9] K.T. Delle Donne, S.R. Sesack, V.M. Pickel, Ultrastructural im- 1988, pp. 247–361.
munocytochemical localization of the dopamine D receptor within2 [32] E. Niki, Action of ascorbic acid as a scavenger of active and stable GABAergic neurons of the rat striatum, Brain Res. 746 (1997) oxygen radicals, Am. J. Clin. Nutr. 54 (1991) 1119S–1124S. 239–255. [33] C. Palmer, R.L. Roberts, C. Bero, Deferoxamine post-treatment [10] J.M. Denu, K.G. Tanner, Specific and reversible inactivation of reduces ischemic brain injury in neonatal rats, Stroke 25 (1994)
protein tyrosine phosphatases by hydrogen peroxide: evidence for a 1039–1045.
sulfenic acid intermediate and implications for redox regulation, [34] T. Pellmar, Electrophysiological correlates of peroxide damage in Biochemistry 37 (1998) 5633–5642. guinea pig hippocampus in vitro, Brain Res. 364 (1986) 377–381. [11] R. Dingledine, J. Dodd, J.S. Kelly, The in vitro brain slice as a [35] T.C. Pellmar, Peroxide alters neuronal excitability in the CA1 region useful neurophysiological preparation for intracellular recording, J. of guinea-pig hippocampus in vitro, Neuroscience 23 (1987) 447–
Neurosci. Methods 2 (1980) 323–362. 456.
[12] T. Finkel, Oxygen radicals and signaling, Curr. Opin. Cell. Biol. 10 [36] T.C. Pellmar, Use of brain slices in the study of free-radical actions,
[37] T.C. Pellmar, K.L. Neel, K.H. Lee, Free radicals mediate peroxida- [46] V. Seutin, J. Scuvee-Moreau, L. Massotte, A. Dresse, Hydrogen tive damage in guinea pig hippocampus in vitro, J. Neurosci. Res. peroxide hyperpolarizes rat CA1 pyramidal neurons by inducing an 24 (1989) 437–444. increase in potassium conductance, Brain Res. 683 (1995) 275–278. [38] T.C. Pellmar, D. Roney, D.L. Lepinski, Role of glutathione in repair [47] C.W. Shuttleworth, T.K. Smith, Action potential-dependent calcium of free radical damage in hippocampus in vitro, Brain Res. 583 transients in myenteric S neurons of the guinea-pig ileum,
Neuro-(1992) 194–200. science 92 (1999) 751–762.
[39] J. Qian, P. Saggau, Modulation of transmitter release by action [48] R. Spector, A.V. Lorenzo, Ascorbic acid homeostasis in the central potential duration at the hippocampal CA3–CA1 synapse, J. Neuro- nervous system, Am. J. Physiol. 225 (1973) 757–763.
physiol. 81 (1999) 288–298. [49] Y. Tang, R.S. Zucker, Mitochondrial involvement in post-tetanic [40] M.E. Rice, Ascorbate regulation and its neuroprotective role in the potentiation of synaptic transmission, Neuron 18 (1997) 483–491.
brain, Trends. Neurosci. 23 (2000) 209–216. [50] H. Tsukaguchi, T. Tokui, B. Mackenzie, U.V. Berger, X.Z. Chen, Y.
1
[41] M.E. Rice, E.J. Lee, Y. Choy, High levels of ascorbic acid, not Wang et al., A family of mammalian Na -dependentL-ascorbic acid glutathione, in the CNS of anoxia-tolerant reptiles contrasted with transporters, Nature 399 (1999) 70–75.
levels in anoxia-intolerant species, J. Neurochem. 64 (1995) 1790– [51] R.L. Whisler, M.A. Goyette, I.S. Grants, Y.G. Newhouse, Sublethal 1799. levels of oxidant stress stimulate multiple serine / threonine kinases [42] M.E. Rice, C. Nicholson, Glutamate- and aspartate-induced extracel- and suppress protein phosphatases in Jurkat T cells, Arch. Biochem.
lular potassium and calcium shifts and their relation to those of Biophys. 319 (1995) 23–35.
kainate, quisqualate and N-methyl-D-aspartate in the isolated turtle [52] C. Wiedemann, T. Schafer, M.M. Burger, T.S. Sihra, An essential cerebellum, Neuroscience 38 (1990) 295–310. role for a small synaptic vesicle-associated phosphatidylinositol [43] J. Rettig, C. Heinemann, U.H. Ashery, Z. Sheng, C.T. Yokoyama, 4-kinase in neurotransmitter release, J. Neurosci. 18 (1998) 5594–
21
W.A. Catterall, E. Neher, Alteration of Ca dependence of neuro- 5602.
21
transmitter release by disruption of Ca channel / syntaxin inter- [53] Y.C. Wu, T. Tucker, R. Fettiplace, A theoretical study of calcium action, J. Neurosci. 17 (1997) 6647–6656. microdomains in turtle hair cells, Biophys. J. 71 (1996) 2256–2275. [44] R. Sah, R.D. Schwartz-Bloom, Optical imaging reveals elevated [54] W. Ying, S.K. Han, J.W. Miller, R.A. Swanson, Acidosis potentiates intracellular chloride in hippocampal pyramidal neurons after oxida- oxidative neuronal death by multiple mechanisms, J. Neurochem. 73 tive stress, J. Neurosci. 19 (1999) 9209–9217. (1999) 1549–1556.
![Fig. 5. Evoked [Ca21]decreases during pulse train stimulation in the presence and absence of H O](https://thumb-ap.123doks.com/thumbv2/123dok/3139235.1382737/6.612.50.284.62.373/fig-evoked-decreases-pulse-train-stimulation-presence-absence.webp)
