Introduction trauma, brain pH acidifies to 6.2 to 6.8 [32]. Even more
Brain Research 881 2000 77–87 www.elsevier.com locate bres
Interactive report
Synaptic depression and neuronal loss in transiently acidic
1
hippocampal slice cultures
2
Zhong-Min Xiang , Peter J. Bergold
Department of Physiology and Pharmacology , State University of New York-Downstate Medical Center, 450 Clarkson Avenue, Brooklyn,
NY 11203, USA
Accepted 11 August 2000
Abstract
Acidosis is a rapid and inevitable event accompanying cerebral ischemia or trauma. We used hippocampal slice cultures to examine an immediate effect of acidosis, synaptic depression; and a delayed effect, neuronal loss. Exposure to low bicarbonate artificial cerebral
spinal fluid aCSF, pH 6.70 for 30 min at 328C, acidified intracellular pH from 7.3160.12 to 6.5360.08. Accompanying intracellular acidosis was a depression of synaptic responses. Both effects rapidly reversed after treatment with normal aCSF pH 7.35. Death analysis
after acidosis treatment revealed no delayed neuronal loss. Increasing the duration of the acidosis to 60 min, however, induced irreversible synaptic depression and delayed neuronal loss. Increasing acidosis temperature to 378C acidified intracellular pH and depressed synaptic
responses. Delayed neuronal loss was also observed. Acidosis using lactate aCSF, pH 6.70 for 30 min at 328C acidified intracellular pH from 7.1960.13 to 6.4360.07 and depressed synaptic responses. After reperfusion with lactate containing aCSF pH 7.35, intracellular pH
recovered yet synaptic responses remained depressed and delayed neuronal loss was observed. This suggested that, for a 30-min treatment at 328C, lactate acidosis was neurotoxic while low bicarbonate acidosis was not. Increasing the duration or temperature of low bicarbonate
acidosis induced neuronal loss. These data provide additional evidence that acidosis contributes to the neurotoxicity during stroke and trauma.
2000 Elsevier Science B.V. All rights reserved.
Theme : Disorders of the nervous system
Topic : Ischemia
Keywords : Intracellular pH; Lactate; Bicarbonate; Hypothermia; Slice culture
1. Introduction trauma, brain pH acidifies to 6.2 to 6.8 [32]. Even more
severe brain acidification occurs if hyperglycemia precedes Intracellular and extracellular pH are highly regulated in
ischemia or if the ischemia is incomplete so that glucose the brain. Intracellular neuronal pH ranges from 6.9 to 7.2
continues to be delivered to hypoxic cells [32,18]. In these and intracellular glial pH ranges between 6.8 and 7.6
conditions, pH can acidify to as low as 6.0 [20,13,2]. [4,7,23]. Larger deviations of intracellular pH occur only
Acidosis of the brain is typically transient. In both during pathological conditions such as cerebral ischemia or
ischemia and trauma, pH recovers to physiological levels trauma [32]. During ischemia, intracellular pH of neurons
in minutes to hours. and glia typically acidify to 6.5 [32,20,13,2]. During
Both early and delayed effects of acidosis have been described. Early effects of acidosis include depression of
neuronal activity [1,44], cell swelling [34] and enhanced production of free radicals [32]. Unless acidosis is severe,
these effects are reversible. Acidosis has minor effect on
1
Published on the World Wide Web on 1 September 2000.
resting membrane potential and excitability of neurons [5].
Corresponding author. Tel.: 11-718-270-3927; fax: 11-718-270-
Synaptic transmission is strongly depressed [1,19,37,17].
2241.
The site of synaptic depression is thought to be post-
E-mail address : pbergoldnetmail.hscbklyn.edu P.J. Bergold.
2
synaptic, since presynaptic fiber volley is more resistant to
Present address: Department of Psychiatry, Mount Sinai School of Medicine, One Gustave Levy Place, New York, NY 10029, USA.
acidosis than excitatory post-synaptic potential EPSP
0006-8993 00 – see front matter
2000 Elsevier Science B.V. All rights reserved. P I I : S 0 0 0 6 - 8 9 9 3 0 0 0 2 7 9 5 - 5
78 Z
[44]. Post-synaptic effects of acidosis include suppression 2. Materials and methods
of ionotropic glutamate receptor function and enhanced function of GABA-A receptors [36,39,11].
2.1. Hippocampal slice cultures A potential delayed effect of acidosis is neurotoxicity.
Hyperglycemia or hypercapnea preceding ischemia results Hippocampal
slice cultures
were prepared
from in more severe acidosis and neuronal loss than ischemia
Sprague–Dawley rats aged 20–30 days [45]. Rats were alone [13,18,14]. These data suggest that acidosis is toxic.
treated with the anesthetics ketamine and halothane to In contrast, hypercapnea preceding ischemia has been
ensure that any pain or discomfort were minimized during reported to reduce brain injury [33]. Direct application of
brain removal. One to three slices were plated on a lactic acid in vivo suggested that acidosis may not be
Millipore-CM filter Millipore, Woburn, MA. The filters neurotoxic since very acidic lactic acid pH
5.3 was were placed above 1 ml of slice culture media SCM, 50
o
needed to induce neuronal damage [15]. Such pH extremes Basal Medium Eagle’s, 25 Earles Balanced Salt solution,
do not occur in vivo. Intracellular pH was not measured in 25 horse serum, 25 mM N-2-Hydroxyethyl piperazine-
this study so the extent and duration of the acidosis is not N9-2-ethanesulfonic acid HEPES, 1 mM glutamine, 28
known. mM glucose, pH57.2 and incubated at 328C in a 5 CO
2
In vitro studies of acidosis toxicity have not been atmosphere. After 1 week, slices were cultured in slice
conclusive. Depending upon the culture system, acidosis culture medium containing 5 horse serum. Cultures were
has been shown to be either neuroprotective or neurotoxic. fed weekly if the insert contained one slice, every 3 days
The evidence that acidosis is neuroprotective largely for two slices and every 2 days for three slices. Cultures
comes from embryonic primary neuronal cultures. Inhibi- were maintained for 2 weeks before use.
tion of NMDA receptors at acidic pH suggests a mecha- nism how acidosis may be neuroprotective. The neuro-
2.2. Electrophysiological recording toxicity of hypoxia or excitotoxicity toward primary
embryonic neurons was strongly attenuated by acidosis Slice cultures attached to the Millicell-CM filter were
[37,38,12]. These cultures are highly resistant to acidosis transferred to an air-interface slice recording chamber that
[34,38,10,21]. The resistance of young neurons to acidosis had been modified to fit the filter insert Fine Science
toxicity may underestimate acidosis toxicity factor toward Tools, Foster City, CA. Cultures were perfused at 1
older neurons. ml min with aCSF that was aerated with 95 O and 5
2
Direct evidence for acidosis toxicity comes from studies CO , and maintained at 328C or 378C aCSF: in mM
2
using hippocampal slice cultures. Moderate and transient NaCl, 124; KCl, 3; MgCl , 1.6; CaCl , 1.7; NaH PO , 1.2;
2 2
2 4
acidosis of pH 6.6 induced widespread, delayed neuronal NaHCO
25; glucose, 11; pH57.35. Schaffer collaterals
3,
necrosis and apoptosis. Increasing the duration of acidosis of slice cultures were stimulated every 30 s with bipolar
increased the amount of neuronal loss [8]. These data tungsten electrodes. Glass recording electrodes were filled
support a neurotoxic role of acidosis [8,31]. In these with aCSF 4–8 MV and placed in the CA1 pyramidal
studies, intracellular acidosis was induced by perfusing cell layer. Amplified signals were digitized with a
slice cultures with reduced bicarbonate. To further study McAdios digitizer GW Instruments, Somerville, MA and
the potential
neurotoxicity of
acidosis, intracellular
processed using Superscope 2.0 software. Field excitatory acidosis was induced by either low bicarbonate or lactic
post-synaptic potentials fEPSPs were recorded in CA1 acid perfusion. The mechanism of induction of intracellu-
pyramidal cell layer by maximally stimulating Schaffer lar acidosis differs using low bicarbonate or lactic acid.
collaterals. In all experiments input–output data from Perfusion with low bicarbonate induces acidosis by dimin-
stimulation intensities of 50 to 200 mA, fEPSP amplitude
2
ishing the strength of intracellular buffering by HCO was roughly proportional. The maximum stimulation was
3
CO . Perfusion with acidic solutions of lactic acid acidifies near 300 mA in all groups. At each stimulation intensity,
2
intracellular pH by the free diffusion of lactic acid and there were no statistical differences among groups
active transport of lactate across the cell membrane in ANOVA, P.0.05, data not shown. Similar results were
addition to diminishing the strength of intracellular buffer- obtained if half-maximal stimulation was used data not
2
ing by HCO CO [26,43]. In this study, two parameters shown. Evoked synaptic responses were also tested at 32,
3 2
were measured during the acidosis treatment, acidification 34 and 378C. Increased temperature resulted in larger
of intracellular pH and depression of synaptic responses. fEPSP regardless of the stimulation intensity. Prior to
Assay of these immediate effects ensure that slice cultures sterile recordings, the recording chamber and electrodes
display well-known immediate effects of extracellular were washed with 70 ethanol and perfusion solutions
acidosis. Following the end of the acidosis treatment were filtered using 0.2 mm filter Corning, Bedford, MA.
cultures were returned to the incubator and delayed After recording, the slices were cultured in SCM con-
neuronal loss assayed. These studies provide additional taining streptomycin 50 unit ml, Sigma. Streptomycin is
evidence that moderate acidosis seen in in vivo ischemic a cell impermeable antibiotic that did not effect the results
conditions is toxic enough to kill neurons. of this study.
Z .-M. Xiang, P.J. Bergold Brain Research 881 2000 77 –87
79
2.3. Assay of cell death slice cultures were perfused for 30 min with aCSF
containing 25 mM bicarbonate. The solution pH was 7.35 Cell death was measured using the cell impermeable
when gassed with 95 O and 5 CO . Acidosis was
2 2
fluorescent dye propidium iodide PI. When the cell induced for 30 or 60 min by perfusing the slice cultures
membrane is damaged, PI enters the cell, binds DNA, and with low bicarbonate aCSF, pH 6.70. A 30-min recovery
emits fluorescence when excited. Fluorescence intensity is period of perfusion with aCSF followed the low bicarbon-
proportional to the amount of DNA binding as well as cell ate acidosis treatment. In some experiments, cultures were
death [8,45,22]. Unlike earlier studies using slice cultures perfused with lactate aCSF. Lactate aCSF was prepared by
from rats younger than 10 days, this study uses slice replacing 20 mM NaCl with 20 mM lactate in aCSF
cultures from 20-day-old rats. These cultures remain buffered with 20 mM HEPES instead of 25 mM bicarbon-
substantially thicker than cultures from 10-day-old rats ate. The pH of lactate aCSF was adjusted to pH 7.35 or pH
[45]. Cell counting is difficult since the cell density in the 6.70 using HCl. Acidosis was induced using a 30-min
pyramidal cell layer is high and PI-positive cells are in perfusion of normal lactate aCSF pH 7.35, followed by
multiple focal planes. Studies from the authors [8] and 30 min of acidic lactate aCSF pH 6.70, followed by a
from other investigators [22,40] have demonstrated that recovery period of 30 min of normal lactate aCSF pH
propidium iodide labeling quantitatively measures cell 7.35.
loss. All images were normalized using an image of a calibration slide containing InSpeck red 100 fluorescent
2.5. Measurement of intracellular pH beads Molecular Probes, Eugene, OR. The use of the
InSpeck fluorescent beads ensures that the excitation Slice cultures were loaded for 60 min at 358C in 2 ml of
intensity and camera sensitivity are uniform and all modified Gey’s Balanced Salt solution in mM, NaCl, 120;
readings are in the linear range of the camera. Slice KCl, 5; KH PO , 0.2; Na HPO , 0.8; NaHCO
27;
2 4
2 4
3,
cultures were transferred to serum-free SCM with 4 mg ml MgSO , 0.3; MgCl , 10; CaCl , 1.5; glucose, 5.6; pH5
4 2
2
propidium iodide Sigma, St. Louis, MO, incubated in 5 7.2 containing 10 mM BCECF-AM and 0.001 v v
CO incubator at 328C for 30 min. Fluorescence was
pluronic acid Molecular Probes, Eugene, OR. Cultures
2
observed using a Zeiss Axiovert 100 microscope with a were excised from the filter insert, and transferred to a
rhodamine filter set. After each image session, slices were perfusion chamber assembled on the stage of a Nikon
rinsed with serum-free SCM, transferred to SCM con- Diaphot inverted fluorescence microscope. The slice cul-
taining 5 horse serum and returned to the incubator. The tures were slightly submerged and perfused 1 ml min
fluorescent images were taken using a PTI intensified CCD with aCSF or modified aCSF aerated with 95O ,
2
camera and analyzed using NIH Image v.1.59. Mean pixel 5CO . Depending upon the experiment, chamber tem-
2
values were assayed the CA1 pyramidal layer. The CA1 perature was maintained either at 32 or 378C. Slice cultures
pyramidal cell layer in organotypic hippocampal slice were illuminated with alternating excitation at 440 and 490
cultures is known to contain predominantly CA1 pyramidal nm and emissions at 510 nm were recorded Photon
cells [9]. Interneurons and glia are also present in the CA1 Technology International, Princeton, NJ. Ratio images of
pyramidal cell layer [9]. When PI florescence was observed 490 nm 440 nm were obtained. Calibration curve were
in the CA1 pyramidal cell layer, PI florescence was generated by exposing cultures to the solutions with 6 pH
minimal in the adjacent regions of stratum radiatum and values ranging from 5.8 to 7.5: in mM KCl 70, HEPES
stratum oriens data not shown. These regions contain 20, sucrose 90, Na HPO
2.5, glucose 10, MgSO 1,
2 4
4
predominantly neuropil, glia, and interneurons [9]. The CaCl 1, nigericin 10 mM. The acquired equation was used
2
presence of PI staining in CA1 stratum pyramidale and the to convert the ratio 490 nm 440 nm to pHi values: pH 5
i 2
absence of staining in adjacent regions suggests that the 4.0810.93Ratio, r 50.89. n59. The high K nigericin
staining in the CA1 pyramidal cell layer is predominantly method used to calibrate BCECF fluorescence potentially
neuronal. In all experiments, propidium iodide assay at day induces a pH-dependent error [3]. To address this issue, a
zero of staining corresponds to 1 h following the acidosis second calibration curve was generated using a buffer that
or mock acidosis treatment. The manipulations of the lacked nigericin in mM; NaCl, 10; KCl, 70; MgCl , 1;
2
experiment, including temperature changes, changes in HEPES, 20, plus BCECF, 2 mM at solution pH of pH 5.0,
buffers, are thought to induce propidium iodide staining at 5.5, 6.0, 6.5, 7.0, 7.5 [46]. A comparison of the pH values
this time. generated from these two calibration curves yielded similar
pH values except for a higher intracellular pH value 2.4. Acidosis treatments
7.4660.15 using aCSF, solution pH 7.35. The conclu- sions of this study are equivalent using either calibration
In low bicarbonate acidosis, the 25 mM bicarbonate in curve.
aCSF was reduced to 5 mM, and NaCl was corresponding- In this study, slice cultures were continuously perfused
ly increased by 20 mM. When gassed with 95 O and 5 during the pH measurements. In a previous paper, slice
2
CO , solution pH was 6.70. Before acidosis induction, the cultures were imaged under a coverslip without perfusion
2
80 Z
[8]. This was possible since cultures from post-natal day session, the cultures were placed in slice culture media
10 rats resist the hypoxia that occurs without perfusion containing the antibiotic streptomycin and returned to a
P.J.B., unpublished data. We believe that perfusing the 328C incubator. CA1 neuronal loss, a delayed response to
cultures resulted in better retention of the dye. The acidosis in slice cultures, was analyzed 1 and 2 days after
continuous perfusion of the cultures in this study rapidly the recording session Fig. 1C. Propidium iodide epi-
removes dye that leaked to the extracellular space. For this fluorescence in the cultures following low bicarbonate
reason, the extracellular contamination of pH values is acidosis treatment did not differ significantly from controls
likely minimal. suggesting that the 30-min acidosis treatment did not
induce neuronal loss. These results are consistent with the 2.6. Statistics
earlier findings of Ding et al. [6], that demonstrated that low bicarbonate acidosis was nontoxic when performed at
Statistical comparisons were performed using one-way 328C.
ANOVA or Student’s t-test using a significant level of Acidosis-induced neuronal loss was suppressed at 328C
0.05. All values are presented as mean6S.E.M. Fig. 1C; [8]. Intracellular pH, synaptic depression and
neuronal loss were examined at 37 or 328C. A 20-min low bicarbonate acidosis at 378C induced acidification of