Peter G. Aitken

Preparative methods for brain slices: a discussion

Criteria for slice health and factors that affect slice health were discussed by many of the participants in the conference. In addition to the standard parameters of slice health (energy metabolism, morphology, electrophysiological responsiveness) more subtle but possibly equally important manifestations of slice health were discussed. These included protein synthesis, and more subtle changes, of which we are becoming increasingly aware. The latter include synthesis of stress-related proteins, altered levels of phosphorylation, altered levels of proteolysis. These last were only touched on, but it is becoming apparent they do in fact constitute important manifestations of differences between the slice preparation and the in vivo tissue. They may well lead to quite different responses in slices from those that occur in vivo. While many ways of optimizing slice wellness were discussed, there was a fair consensus that certain adjustments will optimize the most widely measured aspects of cell function. These include the following, wherever possible. Use of young animals, use of the interface chamber, preparing slices with the vibratome, pre-treating animals with ice-cold cardiac perfusion before sacrificing, using pre-incubation media which reduce NMDA receptor activation, free radical formation and cell swelling. When possible these treatments should perhaps be continued into the normal incubation. This being said, many viewpoints were actually expressed in the discussion, and it should be read to get a feel for the usefulness of the different approaches.

Block of (Na+,K+)ATPase with ouabain induces spreading depression-like depolarization in hippocampal slices

We used ouabain (100 microM) to block Na+,K(+)ATPase of in vitro rat hippocampal slices. This treatment was sufficient to cause the sudden depolarization that is the hallmark of both spreading depression (SD) and of the SD-like anoxic depolarization (AD). This depolarization was accompanied by a large and sudden increase in [K](o), also reminiscent of that observed during both SD and AD. Ouabain-induced SD did not require a complete inactivation of Na+,K(+)ATPase, as it occurred when the enzyme was still capable of providing recovery of both V(o) and [K](o). The data indicate that functional inactivation of Na+,K(+)ATPase per se initiates events that lead to an SD-like AD. This ouabain-induced depolarization was not affected by block of synaptic transmission, instead it was abolished by hyperosmolarity of the extracellular space. The possible relevance of these findings to the pathophysiology of AD is discussed.

Hypotonic exposure enhances synaptic transmission and triggers spreading depression in rat hippocampal tissue slices

Low extracellular osmotic pressure (pi o) is known to enhance CNS responsiveness and the chance of seizures, but the mechanism of the hyperexcitability is not clear. We recorded evoked potentials in st. radiatum and st. pyramidale of CA1. Tissue electrical resistance (Ro) was determined from the voltage drop (VRo) evoked by constant current pulses. Lowering of pi o by reducing [NaCl] caused a concentration-dependent increase of amplitude and duration of extracellular excitatory postsynaptic potentials (fEPSPs). fEPSPs increased much more than did VRo, but antidromic population spikes increased in proportion to VRo. fEPSP increased also in isosmotic low NaCl (fructose or mannitol substituted) solutions, but not as much as in low pi o. In moderately hypotonic solutions orthodromic population spikes increased as expected from the augmented fEPSP, but in strong hypotonia input-output curves shifted to the left and single stimuli evoked multiple population spikes, indicating lowering of threshold of postsynaptic neurons. Blocking N-methyl-D-aspartate (NMDA) receptors did not diminish the enhancement of fEPSP amplitude. Spreading depression (SD) erupted in most slices in very low pi o, but not in isoosmotic low [NaCl] solutions. We conclude that the hypotonic enhancement of EPSPs depends, in part, on the lowering of [Na+]o and/or of [Cl-]o, and it may be augmented by dendritic swelling favoring electrotonic spread of EPSPs from dendrites to somata, and buildup of transmitter concentration due to swelling of perisynaptic glia. SD can be initiated by cell swelling, but the depolarization associated with SD is probably not caused by the opening of stretch-gated ion channels.

Spreading depression-like hypoxic depolarization in CA1 and fascia dentata of hippocampal slices: relationship to selective vulnerability

Hippocampal tissue slices were made hypoxic for 4-10 min and then reoxygenated for 60-120 min. Postsynaptic evoked potentials were recorded and extracellular DC potential was monitored continuously in stratum (st.) pyramidale of CA1 and st. granulosum of fascia dentata (FD). In some preparations extracellular potassium ([K+]o) and calcium ([Ca2+]o) were also recorded in both regions. Postsynaptic responses disappeared sooner during hypoxia and were less likely to recover upon reoxygenation in CA1 than in FD. The CA1 region exhibited a spreading depression (SD)-like response to hypoxia more often than did FD. When both regions showed SD-like depolarization, voltage shift and elevation of [K+]o were of greater magnitude and shorter latency in CA1. The probability of posthypoxic recovery of synaptic transmission was inversely related to the time spent in the SD-like state in both CA1 and FD. We conclude that the selective vulnerability of CA1 neurons to hypoxic and ischemic damage may be due, at least in part, to the regions propensity to undergo prolonged and severe SD-like depolarization.

Membrane currents in CA1 pyramidal cells during spreading depression (SD) and SD-like hypoxic depolarization.

We used the patch clamp technique in whole-cell configuration to investigate the membrane current and membrane resistance of neurons in rat hippocampal tissue slices during spreading depression (SD) induced by high K+ solution or electrical stimulation and during SD-like depolarization caused by hypoxia. The potential of the patch pipette was referred to an extracellular micropipette electrode to ensure control of the true membrane potential during large shifts of extracellular potential, delta Vo. During both hypoxic and normoxic SD, increase of holding current indicated a large inward current which reached a mean maximum of about 1.75 nA. This virtual inward current started and ended at the same time as the extracellularly recorded negative delta Vo shift, but the trajectories of the two differed. When the membrane was clamped at strongly positive potential, the current during SD was outward. The average apparent reversal potential of the current during SD was near zero but in individual cases varied from -26 mV to + 12 mV. During SD the input resistance decreased on the average to 43% of the resting control value. The decrease of the input resistance was not voltage dependent. The increase of holding current and decrease of resistance occurred with both Cs- and K-gluconate recording pipettes and was not suppressed by 2 mM intracellular QX-314. Voltage-gated currents disappeared during SD; a small, Cs(+)-resistant outward rectifying current was the last to be lost. During recovery, reversal potential and input resistant overshot the control level and then returned to normal within about 5 min. The data are consistent with change of both driving potential and conductance for several ions, but the decrease of overall membrane resistance was less than earlier estimates with other methods had suggested. Normoxic SD and hypoxic SD-like depolarization could not be distinguished by these tests.

The effects of moderate changes of extracellular K+ and Ca2+ on synaptic and neural function in the CA1 region of the hippocampal slice

The effects of moderate changes of the concentration of ions on the function of mammalian central nervous tissue have not exactly been determined. We placed tissue slices from rat hippocampal formation in an interface chamber for study in vitro. Extracellular potentials were recorded in stratum radiatum and stratum pyramidale in response to stimuli of varying intensity applied to the Schaffer collateral bundle. The overall input-output relationship of excitatory synaptic transmission was gauged by expressing postsynaptic population spike amplitude as a function of presynaptic volley amplitude. The components of the transmission process were also examined by plotting the maximal rate of rise (slope) of the focally recorded synaptic potential (fEPSP) as a function of presynaptic volley amplitude, and the population spike amplitude as a function of the fEPSP slope. Raising the concentration of K+ from the normal level of 3.5 mM to 5 mM caused an average increase of 48% in the population spike evoked by a given presynaptic volley. This was due to an increased electrical excitability of pyramidal cells, as indicated by an increase of the population spike evoked by a given magnitude of fEPSP. Conversely, lowering [K+]o from 3.5 to 2 mM caused a decrease of the population spike relative to a given magnitude of either the presynaptic volley or the fEPSP. Changing [K+]o within these limits caused no significant change of the fEPSP evoked by a given presynaptic volley. Raising [Ca2+]o from 1.2 to 1.8 mM caused a 35% increase in both the fEPSP and the population spike evoked by a given presynaptic volley, and lowering [Ca2+]o to 0.8 mM caused a decrease of both these functions. The amplitude of the population spikes evoked by given fEPSPs changed surprisingly little (but consistently) when [Ca2+]o was varied within these limits. We conclude that moderate changes of [K+]o influence mainly the electric excitability of hippocampal pyramidal cells, with little effect on transmitter release or on the response of the postsynaptic membrane to transmitter, while moderate changes of [Ca2+]o affect the release of excitatory synaptic transmitter more than they affect postsynaptic membrane function.

NMDA antagonists: Lack of protective effect against hypoxic damage in CA1 region of hippocampal slices

Rat hippocampal slices were exposed to a hypoxic insult in control medium or while exposed to the N-methyl-D-aspartate (NMDA) receptor antagonists DL-2-amino-7-phosphonoheptanoic acid (100 microM) or DL-2-amino-5-phosphonovaleric acid (25 or 100 microM). Synaptic transmission between Schaffer collaterals and CA 1 pyramidal cells was evaluated before and after the hypoxic period, and the DC potential in the CA 1 pyramidal cell layer was monitored during the hypoxic period. Neither antagonist significantly increased the proportion of slices in which synaptic transmission recovered following hypoxia, nor did they increase the latency of the spreading depression-like anoxic depolarization. We suggest that NMDA receptors are not involved in neural damage caused by severe hypoxia with sudden onset, while they may play a role in the effects of more moderate, gradual onset hypoxia and in post-ischemic reperfusion effects.

Role of calcium channels in spreading depression in rat hippocampal slices.

In the CA1 region of rat hippocampal slices, spreading depression (SD) was provoked by a brief period of hypoxia or by localized application of high potassium solution. We measured extracellular DC voltage (Vo), extracellular potassium concentration ([K+]o) and/or extracellular Ca2+ concentration ([Ca2+]o). SD was provoked under control conditions and also when voltage-gated Ca2+ channels were blocked by application of 2 mM Ni2+ or Co2+. In some experiments, CPP, DNQX, or the two together were also applied to block glutamate receptor-coupled channels. When SD was provoked by hypoxia, these treatments significantly increased the latency of SD onset and decreased the amplitudes of the accompanying delta Vo, delta [Ca2+]o and delta [K+]o. Hypoxia-induced SD was never blocked completely, however and delta [Ca2+]o was reduced at most by 50%. When SD was provoked by application of high K+ solution near the recording site, Ni2+ or Co2+ partially suppressed the Vo and [Ca2+]o shifts but did not block SD altogether. When high K+ solution was applied at a distance, Ni2+ or Co2+ blocked the propagation of SD to the recording site. We conclude that during SD, a significant proportion of the calcium ions flowing into neurons does not pass through voltage-gated or glutamate receptor-linked channels.

Ion channel involvement in hypoxia-induced spreading depression in hippocampal slices

Rat hippocampal tissue slices were made hypoxic in control medium and in medium containing the ion channel blockers tetraethylammonium (TEA), 4-aminopyridine (4-AP), or tetrodotoxin (TTX). Postsynaptic evoked potentials, extracellular DC potential Vec, and in some experiments extracellular potassium concentration [K+]o were monitored in stratum pyramidale of the CA1 region. TEA (10 mM) decreased the latency of hypoxia-induced spreading depression (SD), and reduced the amplitudes of the changes in Vec and [K+]o. 4-AP (50 microM) also decreased the latency of SD but had no effect on the Vec shift. In most slices, TTX (1 microM) increased SD latency but had no effect on the Vec shift. In some slices, TTX blocked the occurrence of SD.

Kainic acid and penicillin: Differential effects on excitatory and inhibitory interactions in the CA1 region of the hippocampal slice

The effects of kainic acid (KA, 0.05-1.0 microM), and penicillin (PN, 3.4 mM) were studied in the CA1 region of rat hippocampal slices. Three components of the overall input/output function were taken: (1) the amplitude of the presynaptic compound action potential (prevolley) vs stimulation current applied to Schaffer collaterals, (2) the magnitude of the focally recorded synaptic potential (population EPSP) vs prevolley amplitude; and (3) the amplitude of the focally recorded population spike vs population EPSP magnitude. Recurrent inhibition was measured using the antidromic-orthodromic paired pulse method. KA caused a significant and reversible enhancement of all 3 component input/output functions while having no effect on paired pulse inhibition. PN caused a left shift in the EPSP-population spike relationship and decreased or abolished paired pulse inhibition; the other two measures of excitability were not changed. These results suggest that PN and KA differ fundamentally in the mechanisms by which they produce seizures: PN by removing inhibition while not affecting neuronal excitability per se; KA by exerting a generalized excitatory effect on neural membranes and on synaptic function while leaving recurrent inhibition unchanged.