The Journal of Neurophysiology Vol. 77 No. 5 May 1997, pp. 2349-2359
Copyright ©1997 by the American Physiological Society

Enhancement of Whole Cell Synaptic Currents by Low Osmolarity and by Low [NaCl] in Rat Hippocampal Slices

Rong Huang, Daniel F. Bossut, and George G. Somjen

Departments of Cell Biology and Neurobiology, Duke University Medical Center, Durham, North Caroline 27710

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Huang, Rong, Daniel F. Bossut, and George G. Somjen. Enhancement of whole cell synaptic currents by low osmolarity and by low [NaCl] in rat hippocampal slices. J. Neurophysiol. 77: 2349-2359, 1997. We recorded whole cell currents of patch-clamped neurons in stratum pyramidale of CA1 region of rat hippocampal tissue slices. Synaptic currents were evoked by orthodromic stimulation while holding potential of the neuron was varied from hyperpolarized to depolarized levels. Extracellular osmolarity (_o) was lowered by superfusion with artificial cerebrospinal fluid in which NaCl concentration ([NaCl]) was reduced. The effect of low extracellular NaCl was tested in additional trials in which NaCl was substituted by isosmolar fructose. Both lowering of _o and isosmotic lowering of extracellular [NaCl] ([NaCl]_o) caused reversible increase of excitatory postsynaptic currents. The effect of lowering _o was concentration dependent, and it was significantly stronger than the effect of equivalent isosmotic lowering of [NaCl]_o. Inhibitory postsynaptic currents also increased in many but not in all cases. Lowering of _o caused a prolongation of the time constant of relaxation of the capacitive charging current induced by small hyperpolarizing voltage steps. A virtual input capacitance, calculated by dividing this time constant by the input resistance, increased during hypotonic exposure. Isosmotic lowering of [NaCl]_o had no effect on time constant or input capacitance. Depolarizing voltage commands evoked spikelike inward currents presumably representing Na⁺-dependent action potentials generated outside the voltage-clamped region of the cell. These current spikes became smaller in low _o and in low [NaCl]_o. Broader, voltage-dependent, presumably Ca²⁺-mediated inward currents became more prominent during hypotonic exposure. We conclude that lowering of [NaCl]_o causes enhancement of excitatory synaptic transmission. Transmission may be facilitated by the uptake of Ca²⁺ into presynaptic terminals as well as into postsynaptic target neurons, induced by the low [NaCl]_o. Lowering of _o enhances synaptic transmission more than does a corresponding isosmotic lowering of [NaCl]. The excess increase recorded from the cell soma in low _o may in part be due to changing electrotonic length caused by the swelling of dendrites.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

The osmolarity of extracellular fluid (_o) can decrease because of water loading or salt loss, with serious clinical consequences, including coma and seizures (Avner 1995; Ellis 1995; Thorn et al. 1977). The clinical syndrome of low _o is complicated by transfer of water across the blood-brain barrier and elevation of intracranial pressure, because the barrier is more permeable to water than to most solutes (Bradbury 1979; Yamaguchi et al. 1994). From studies of cerebral tissue slices in vitro it is clear, however, that low _o also has direct effects on brain tissue itself. When such a slice is exposed to a hypotonic bathing solution, the amplitude of extracellularly recorded excitatory postsynaptic potentials (fEPSPS) is greatly enhanced (Andrew 1991; Ballyk et al. 1991; Chebabo et al. 1995a; Rosen and Andrew 1990). Electrographic seizures are facilitated (Andrew et al. 1989; Dudek et al. 1990), but, in the absence of another convulsant stimulus, low _o by itself rarely induces seizurelike discharges (Chebabo et al. 1995a; Huang et al. 1995). More commonly, in hippocampal tissue slices treated by strongly hypotonic bathing solution, recurrent episodes of spreading depression develop. If, however, calcium is withdrawn and magnesium concentration is raised, then similar hypotonic treatment induces recurrent seizures instead of spreading depression (Chebabo et al. 1995a; Huang et al. 1995).

Interpretation of extracellular recordings is complicated by the fact that lowering of _o causes concentration-dependent decrease of interstitial space, resulting in increase of tissue resistance (R_o), which alters the recorded signals. We found that the antidromic population spike amplitude increased in proportion to R_o but that fEPSP amplitude increased much more than did the antidromic spike and R_o (Chebabo et al. 1995a). Moreover, the duration and waveform of the fEPSPs were altered in ways that changes in the electrical properties of the tissue could not explain. We concluded that synaptic transmission was genuinely enhanced by low _o (Chebabo et al. 1995a). Earlier, Ballyk et al. (1991) found, however, that intracellularly recorded excitatory postsynaptic potentials of hippocampal pyramidal cells did not increase when _o was lowered, even though the same team reported that intracellular excitatory postsynaptic potentials of neocortical neurons were reliably increased by lowering of _o (Rosen and Andrew 1990).

To resolve the apparent discrepancy between extra- and intracellular recordings in hippocampus, we chose the whole cell patch-clamp method in voltage clamp configuration, because it enables reliable, stable recordings from single cells even while the tissue slice is subjected to insults that cause cell swelling (Czéh et al. 1993). We found a concentration-dependent reversible enhancement of excitatory postsynaptic currents (EPSCs) when bath osmolarity or NaCl concentration ([NaCl]) were lowered.

Abstracts of some of the results have already appeared (Bossut and Somjen 1995; Huang et al. 1996).

	METHODS

Abstract Introduction Methods Results Discussion References

Tissue slices were prepared from male Sprague-Dawley rats of 85-140 g body wt, as previously described (Dingledine 1984; Somjen et al. 1986). The rats were decapitated under ether anesthesia, and the brains were rapidly removed and dropped into chilled artificial cerebrospinal fluid (ACSF). Hippocampal 400 µm thick were prepared by a vibrating cutter (Vibroslice). In the more recent experiments the dissection was performed in a modified fluid, with NaCl and CaCl₂ replaced by equiosmolar (240 mM) sucrose.Vibrator-cut slices prepared in sucrose ACSF seemed to be in better functional state, inasmuch as strong orthodromic volleys evoked single population spikes, not double spikes or bursts. The slices were transferred into a dual-well Oslo-style interface chamber that was perfused with normal ACSF at a rate of ~1.5 ml/min at 36°C. In the initial experiments, slices in reserve were retained in a holding chamber in ACSF at room temperature saturated by 95% O₂-5% CO₂; later, slices in reserve were maintained at the gas interface in the unused well of the dual chamber at 36°C.

Normal ACSF contained (in mmol/l) 130 NaCl, 3.5 KCl, 1.25 NaH₂PO₄, 24 NaHCO₃, 1.2 CaCl₂, 1.2 MgSO₄, and 10 dextrose, pH 7.4, saturated with 95% O₂-5% CO₂. Hypotonic solutions were prepared by reducing [NaCl] (Chebabo et al. 1995a,b). In the solution designated as H-20, NaCl was reduced from 130 to 110 mmol/l; in H-40 solution, [NaCl] was 90; in H-60, it was 70 mmol/l. In fructose-substituted low-NaCl solution, designatedF-40, 40 mmol/l NaCl were replaced by equiosmolar (~80 mmol) fructose. The osmolarities, measured with a freezing-point osmometer, were as follows: 301 ± 4 (SE) mosm/kg (normal ACSF); 261 ± 4 (H-20); 230 ± 5 (H-40); 195 (H-60); 305 ± 4 mosm/kg (F-40).

Patch-clamp pipettes were pulled from thin-walled borosilicate tubing (B159-10, Sutter Instrument) and were usually filled with a solution containing (in mM) 125 potassium gluconate, 10 KCl, 2 MgCl₂, 1 CaCl₂, 11 ethylene glycol-bis(-aminoethyl ether)-N,N,N,N-tetraacetic acid (EGTA), 10 N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid (HEPES), and 2 Na₂ATP, pH 7.25, osmolarity 285 mosM/kg, pipette resistance 3-6 M. In five trials the potassium gluconate was replaced by cesium gluconate, potassium acetate, or KF. Extracellular recording micropipettes were filled with 150 mmol/l NaCl, resistance 12-16 M.

Extracellular and patch-clamp recording electrodes were placed in stratum pyramidale of CA1 region, separated by ~0.5 mm. A tungsten microwire, insulated except at the tip, inserted in stratum radiatum, was used for orthodromic stimulation by photoisolated constant-current pulses from a Grass S88 stimulator. Cells were searched for with the patch pipette by "blind" approach (Blanton et al. 1989; Czéh et al. 1993). After a seal of 1-6 G was achieved, pipette capacitance was compensated and the whole cell configuration was established by gentle suction on the pipette. Series resistance was compensated 75-80%. Cell capacitance could not be fully compensated and, in the most recent trials, this was not attempted so as to be able to determine relative surface membrane capacitance (see RESULTS). Recordings were accepted if the membrane potential was greater than 50 mV and stable, and the holding current was stable. All recordings were made in voltage-clamp mode.

An Axopatch 1D amplifier controlled by the Axon Instruments Axoclamp suite of computer programs was used for recording. Current was sampled at 2,000 Hz. Extracellular responses were amplified through a FET-input stage and recorded on a second channel by the same program. Extracellular voltage was also monitored by chart recorder. The holding potential (V_h) of the patch-clamped cell was usually 65 mV, sometimes more negative. Postsynaptic currents evoked by orthodromic volleys were superimposed on a series of 6-10 consecutive 350- or 370-ms voltage steps, which took the membrane potential from hyperpolarized to depolarized levels in 15-mV increments (Fig. 1). Orthodromic stimuli were constant in each experiment but varied among different experiments from 80 to 150 µA, 0.05-0.1 ms. Individual sweeps were delivered at 15-s intervals; 5 or 6 min separated each set of recordings. In the majority of the experiments this protocol was alternated by ramp depolarizations to test the effective threshold of evoking inward current spikes (Czéh et al. 1993). The ramp took the potential from 120 to 0 mV in 95 ms.

View larger version (39K): [in this window] [in a new window]   FIG. 1. Whole cell currents in a hippocampal CA1 pyramidal neuron. The voltage was first stepped from 65 to 70 mV, then in 15-mV increments from 105 to 0 mV, and then back to 65 mV; for clarity, only 3 sweeps are superimposed here, stepping to 105, 45, and 0 mV (see labels next to traces in A). Voltage protocol is shown in the inset below C. Depolarizing voltages evoked trains of spikelike inward currents probably caused by Na+-dependent action potentials, as well as prolonged inward current surges that are probably Ca2+ currents. In A, st marks the stimulus pulse. The afferent volley evoked by the stimulus caused the synaptic currents, which were inward at hyperpolarizing and outward at depolarizing voltages; the outward synaptic current interrupted the train of current spikes. Currents are shown before leak subtraction. Synaptic currents recorded in another experiment at 6 voltages are shown on an expanded scale in Fig. 2. A: control recording, 20 min after whole cell condition was established. B: 15 min after switching to hypotonic (H-40) solution. C: 25 min of H-40 superfusion. D: after 30 min of wash with normal artificial cerebrospinal fluid (ACSF).

After having been placed in the recording chamber, slices were left undisturbed for 90 min. After whole cell recording conditions were established, control data were collected for 20 min, followed by perfusion of the chamber by hypoosmolar solution. After 20-30 min (usually 25 min) of hypotonic treatment, normal ACSF perfusion was resumed and recovery from hypotonia was observed for 20 min (usually >30 min). In some cases a second, and in rare instances a third hypotonic treatment was applied to the same cell, always allowing 55-65 min of washing with normal ACSF between treatments. After completion of the study of a cell, all slices that had been exposed to hypotonia were discarded and a fresh slice was taken from the holding chamber.

Data were analyzed off-line by ClampFit 6 software (Axon Instruments). Whole cell input resistance (R_N) was determined from the hyperpolarizing current required to raise the clamp potential from V_h by a steady 5 mV. Such hyperpolarizing steps evoked an initial capacitive charging current, the relaxation of which usually could be fitted by a second-order exponential with the use of either the simplex or the Chebyshev algorithm of the ClampFit software. A virtual input capacitance (C_N) was calculated as₁/R_N, where ₁ is the larger of the two time constants (Jackson 1992; Rall 1969). The shorter time constant (₂) was partly suppressed in the trials in which cell capacity compensation had been attempted after whole cell condition was established, and sometimes second-order exponential fitting failed. In these cases a first-order exponential was fitted to the segment of the capacitive current transient beginning 0.8-1.0 ms after the onset of the hyperpolarizing step and ending either 25 or 50 ms later. When both first- and second-order exponentials were fitted for the same cell, the ₁ values were very similar.

The amplitudes of synaptically transmitted currents were measured after correction for linear leak current as the difference between peak current and the average baseline (holding) current shortly before the orthodromic stimulus. EPSC peak was defined as the minimum (i.e., negative peak) and inhibitory postsynaptic current (IPSC) as the maximum (positive peak) current measured by the software between two cursors placed to delimit the relevant part of the current trace.

Significance of differences in mean values was calculated with the use of the t-test.

	RESULTS

Abstract Introduction Methods Results Discussion References

Whole cell currents of voltage-clamped CA1 pyramidal neurons in tissue slices in normal solution

GENERAL PROPERTIES. The current holding the membrane at 65 mV in the initial control state, 20 min after establishment of whole cell recording, was on average 35.7 ± 14.6 (SE) pA (range 136 to +59 pA; n = 15). The average control R_N was 153 ± 11 M (n = 39).

The standard test protocol consisted of an initial 5-mV hyperpolarizing step followed after 50 ms by a series of 6, 8, or 10 successive 350-ms voltage step commands taking the potential from hyperpolarized to depolarized levels in 15-mV increments. An orthodromic volley activated synaptic transmission 200 or 262 ms after the start of the voltage steps. Figure 1A shows uncorrected sample current recordings from a neuron in control solution, and the Fig. 1C, inset, illustrates part of the voltage protocol. Figure 2A shows a complete family of synaptic current recordings on an expanded scale, after linear leak subtraction and adjustment of "baseline" current.

View larger version (13K): [in this window] [in a new window]   FIG. 2. Families of synaptic currents of a CA1 neuron (A-C), and extracellular voltage responses (D-F) in CA1 stratum pyramidale. Six current traces are superimposed in each family of tracings (A-C) obtained at membrane voltages of 105, 90, 75, 60, 45, and 30 mV, preset by a voltage protocol similar to the one partially illustrated in Fig. 1C, inset. Currents were corrected for linear leak, and "baseline" was adjusted to 0 current. Each extracellular recording (D-F) is the average of 6 consecutive sweeps, recorded concurrently with the synaptic currents. A and D: control. B and E: during superfusion by hypotonic (H-40) solution. C and F: recovery, after 30 min of wash with normal ACSF. IPSC, inhibitory postsynaptic current; EPSC, excitatory postsynaptic current.

CAPACITIVE CURRENT. The 5-mV hyperpolarization evoked a brief capacitive transient current that could not be fully eliminated by the capacity compensation circuit. To explore the nature of this current, in nine trials the capacity compensation was not adjusted once whole cell recording was achieved. Figure 3 illustrates the relaxation of the capacitive transient current on an expanded scale in one such trial. The decay of the capacitive current usually appeared to be the composite of two parts showing a change of course 0.8-1.0 ms after its onset, and it could be fitted by a second-order exponential (Jackson 1992; Rall 1969). The average time constants of these nine cells in normal solution were ₁ = 11.59 ± 0.78 and ₂ = 0.39 ± 0.04 ms. In seven other trials in which the cell capacity compensation was attempted after whole cell recording conditions were established, the mean of ₁ was 11.62 ± 1.06 ms (n = 7), which is not significantly different. ₂ Was not measured in these seven cells because adjusting capacity compensation in whole cell conditions distorted or eliminated the fast component of the capacitive current.

View larger version (18K): [in this window] [in a new window]   FIG. 3. Relaxation of capacitive current transients of a CA1 neuron. Voltage was stepped from 65 to 70 mV (see Fig. 1); traces show parts of the currents on an expanded scale, immediately after the peak of the capacitive transient. Top trace: control. Bottom trace: at the end of 25 min of hypotonic exposure (H-40). Middle trace (labeled R): recovery after 30 min of wash with normal ACSF. Each trace is the average of 8 sweeps, with fitted 2nd-order exponentials shown as dashed lines. Ordinate scale represents pA for the control record, but the other 2 traces have been scaled to match maximal amplitudes to facilitate comparison of time courses. Time constants of the decay of the currents are as follows. Controllonger time constant (1): 10.4 ms, shorter time constant (2): 0.45 ms. Hypotonia1: 19.4 ms, 2: 0.84 ms. Recovery1: 12.5, 2: 0.67 ms. Capacity compensation was not applied to the whole cell currents (see METHODS). Data from the same cell as Fig. 1.

VOLTAGE-ACTIVATED CURRENTS. Prolonged hyperpolarizing voltage steps usually evoked a slowly activating inwardly rectifying current. With depolarizing voltage steps a small, persistent voltage-dependent inward current became manifest in many but not in all cells, probably a noninactivating voltage-dependent Na⁺ current (French et al. 1990; Hoehn et al. 1993). These low-amplitude slow currents were clearly revealed after subtraction of "leak" current.

Depolarizing ramp or step voltage commands in the medium range evoked series of brief, very large amplitude inward current spikes (Fig. 1A, trace labeled 45 mV; Fig. 4A). These spike currents probably represent Na⁺-dependent action potentials generated in the axon, outside the region controlled by the voltage clamp, because they are not seen when the patch pipette contains the local anesthetic drug N-(2,6-dimethylphenylcarbamoylmethyl)-triethylammonium bromide (QX-314) (Czéh et al. 1993). At strongly depolarized potentials, complex and more prolonged inward current transients appeared, with superimposed Na⁺ spikes (Figs. 1 and 4). These broad inward current surges are probably caused by Ca²⁺ currents, because they are not suppressed by cesium gluconate plus QX-314 (Czéh et al. 1993).

View larger version (15K): [in this window] [in a new window]   FIG. 4. Currents evoked by ramp depolarization that took the membrane from 120 to 0 mV. A: control. B: after 25 min of hypotonic treatment. C: recovery. Note reduced amplitude of inward "spike" currents, and slightly increased (more positive) threshold level triggering of spikes during hypotonic superfusion. Data from the same cell as Fig. 1.

SYNAPTICALLY EVOKED CURRENTS. With the potential held at a hyperpolarized level, afferent volleys evoked a synaptically transmitted transient inward current (Figs. 1A and 2A, traces labeled 105 mV). At depolarized potentials the same volley evoked an outward current (Fig. 1A, trace labeled 0 mV; Fig. 2A, at 30 mV). The outward current rose more slowly to a later summit and it decayed more slowly than did the inward current. At a V_h of 105 or 100 mV the synaptic inward current reached peak 7.8 ± 0.4 (SE) ms after the stimulus; at V_h between 30 and 0 mV the outward current was maximal at 20.0 ± 0.7 ms. The difference (12.2 ms) is highly significant (P < 0.001). With the membrane potential held at 90 or 105 mV, the mean maximal amplitude of the inward current in control solution was about 180 pA; at 15 or 0 mV the average amplitude of the outward current was ~290 pA.

As the V_h was moved to less negative levels, the synaptically transmitted inward current amplitude decreased and the outward current amplitude increased. The voltage dependence of the maximum amplitudes of the currents is illustrated in Fig. 2 and in the current-voltage (I-V) plots of Fig. 5. A true reversal potential cannot be determined from the I-V plots, but the polarity, time course, and voltage dependence of these currents identify them as EPSCs and IPSCs, respectively (see DISCUSSION). When an EPSC was evoked at a slightly depolarized V_h, it frequently triggered an inward current spike, presumably an Na⁺ spike, especially when augmented by hypotonic treatment (Fig. 2B). By contrast, the IPSC interrupted and suppressed the train of inward spike currents that were evoked by depolarizing voltages (Fig. 1A, traces labeled 45 mV and 0 mV).

View larger version (13K): [in this window] [in a new window]   FIG. 5. Current-voltage (I-V) relation of EPSCs. Amplitudes of the inward synaptic currents (presumed EPSCs) at 6 holding potentials, in control conditions, after 25 min of hypotonic exposure (H-40), and after 30 min of wash with normal ACSF. Straight lines: regression of current on voltage for the control data and for hypotonia (regression for recovery deleted for clarity). Points: mean values of 12 trials in 5 cells (from 5 animals).

Hypotonic treatment causes reversible, concentration-dependent enhancement of EPSCs

Replacing the bath with hypotonic solution caused, as expected, an increase of the extracellularly recorded orthodromic population spike (Ballyk et al. 1991; Chebabo et al. 1995a). At the same time the EPSC recorded from the patch-clamped cell increased markedly in all but one case (Figs. 1, 2, 5, and 6). The current amplitudes in Table 1 show the means ± SE of the EPSC amplitudes measured at the most hyperpolarized potentials. The number of trials with hypoosmotic solutions other than H-40 was limited, but even so it is evident from Table 1 that the increase of EPSC amplitude was concentration dependent. Washing with normal solution resulted in every case in a return toward normal, often followed by posthypotonic depression. The data for recovery in Table 1 are the averages taken after 30 min of wash with normal solution. The mean fEPSC rose during H-40 treatment from the control amplitude of 189-359 pA (190%), and after 30 min of wash it returned to 194 pA, which is very near the normal value (104%). This nearly normal mean at 30-min recovery conceals the widely varying recovery rates: after 30 min of wash, the EPSC of some cells was still somewhat augmented whereas that of some returned to control and yet others were depressed below control level (undershooting).

View larger version (17K): [in this window] [in a new window]   FIG. 6. Changes with time in the maximal amplitudes of the EPSC, IPSC, and the extracellular population spike. Horizontal bar: duration of exposure to hypotonic (H-40) solution. Data from a single experiment.

The changes of the EPSC during exposure to hypotonic solution and during washing with normal solution coincided with changes in the amplitude of the extracellular population spike, which were similar in direction but greater (Figs. 2 and 6; Table 1).

EPSC amplitude increased in approximately equal proportion at all V_hs at which it was detectable. Plotting EPSC amplitude against holding voltage yields an I-V plot, which became much steeper during hypotonia than it was under control condition or after recovery. Figure 5 illustrates this change of slope in 12 trials on five cells where all conditions of the experiment were uniform. During hypotonic treatment, EPSCs tended to trigger inward current spikes at holding voltages at which they did not have this effect under control condition (Fig. 2). When such current spikes were present, EPSC amplitude could not be measured, and for this reason the N value of Fig. 5 at intermediate voltages is <12. At the strongly hyperpolarized as well as the most depolarized voltages, the Na⁺ current spikes were absent. The steepening of slope during hypotonia was evident in all individual I-V plots, even when intermediate points were missing.

IPSCs show variable response to hypoosmotic treatment

In many trials the increase of the EPSC was paralleled by an increase in IPSC amplitude (Fig. 6). In some cells, however, the IPSC reached a maximum after 10-15 min hypotonic exposure and then it started to decrease, and in some cases it was depressed at the end of 25 min of hypotonic treatment. Of 23 trials, at the end of 25 min of treatment with H-40 solution, the IPSC increased reversibly in 15 trials, decreased in 5 cases, and remained unchanged (within 4% of control) in 3 cases. Of the four trials with H-20 solution, the IPSC increased in two, decreased in one, and remained unchanged (98%) in one. In both trials with H-60 solution the IPSC remained unchanged. As a result of this variability, the increase of the mean IPSC amplitude is statistically not significant (Table 1). Yet in the cases when the IPSC did reversibly increase, its amplitude grew much above the range of variation seen in the initial control state (e.g., Fig. 6); in other words, the change was significant (i.e., "real") in that cell. Moreover, whenever the IPSC amplitude increased during hypotonia, the slope of its I-V plot steepened, similarly to that of the EPSC, indicating increased slope conductance. In the few cases when the IPSC was depressed at the end of hypotonia, it became even smaller during wash with normal solution without signs of recovery suggesting possible "rundown."

Isosmotic low-NaCl solution also induces enhancement of EPSCs

In our previous study, extracellular fEPSPs were reversibly increased during exposure to solutions in which NaCl was replaced by equiosmolar mannitol or fructose (Chebabo et al. 1995a). Similarly, in six patch-clamped cells, EPSC amplitudes increased during treatment with fructose-substituted low-NaCl (F-40) solution. EPSC consistently increased in all trials but, on average, not by as much as in comparably NaCl-deficient hypotonic (H-40) solution (Table 1). The mean EPSC increment (difference between treated and control amplitudes) amounted to 170 ± 25.7 pA during H-40 hypotonia and to 88 ± 40.1 pA during F-40 low-NaCl treatment. Despite the wide variability, the difference between the two conditions is significant (P < 0.005). IPSCs were usually also augmented by treatment with F-40 solution but, because of variability, the mean values were not significantly different (Table 1).

Cells maintained in normal solution

In control cells observed for extended periods without change of _o or [NaCl], EPSC and IPSC amplitudes fluctuated somewhat in random fashion, but they showed no consistent trend in any direction (Table 1).

C_N increases during hypotonic treatment but not during low [NaCl]

In the initial series of experiments the capacity compensation dials of the amplifier were repeatedly adjusted during recording, to obtain optimal capacity cancellation. In later experiments cell capacity compensation was adjusted only before recording was started. In the final experiments, cell capacity was not compensated at all, to measure C_N and to follow its changes during treatments.

The longer of the two time constants of the decay of the capacitive charge current, ₁, (Jackson 1992; Rall 1969) was consistently prolonged during hypotonic treatment (Fig. 3, Table 2). Wash with normal ACSF after hypotonia partially reversed this increase (Table 2). In a few instances in which recovery was observed for 55 min, if ₁ was still increased after 30 min of wash, it returned to near normal values at the end of 1 h.

View this table: [in this window] [in a new window]   TABLE 2. 1, RN, and CN

If resistance stays constant, prolongation of ₁ indicates increase of C_N. R_N changes were inconsistent during hypotonia but, on average, R_N did increase. The virtual capacitance obtained from ₁/R_N also increased during H-40 treatment and it recovered partially after 30 min of wash with normal ACSF (Table 2).

There was no increase of ₁ or of C_N during exposure to isosmotic low [NaCl] (F-40) solution (Table 2). Time-dependent changes of ₁ and of C_N during prolonged observation in normal ACSF were small and inconsistent.

In the trials in which whole cell capacity compensation was not applied, ₂ also increased during H-40 hypotonic treatment, from 0.39 ± 0.04 ms to 0.53 ± 0.08 ms (n = 8, P < 0.01).

Other changes caused by hypotonic treatment

The amplitude of the brief inward current spikes decreased when either _o or [NaCl] was lowered (Figs. 1, B and C, and 4B). This is as expected if these current spikes represent Na⁺ dependent action potentials.

Although the putative Na⁺ spike currents decreased, the broader, presumably Ca²⁺-mediated inward current transients became more prominent (Fig. 1, B and C).

The apparent threshold potential at which ramp depolarization triggered putative Na⁺ current spikes was 36 ± 1.3 mV (n = 15). During exposure to H-40 hypotonic solution, the apparent trigger potential shifted inconsistently, with the mean change being 1.6 ± 1.4 mV (not significant). The current corresponding to the trigger was normally 161 ± 17.7 pA (n = 15). In all but three cases of H-40 exposure, the ramp-current triggering inward spikes shifted to more positive levels, on average by 38 ± 11.4 pA (Fig. 4B). Changes of threshold current and of threshold voltage were not always correlated.

Electrochemical potential shifts

Electrochemical junctions were present between the bath and the ground connection, between the extracellular electrode and interstitial fluid, and between the patch pipette solution and the cytosol.

The junction potential change in the extracellular pipette is the sum of the junctions at the pipette tip and at the ground connection. The change of the Cl potential of the Ag/AgCl ground wire can be calculated to be 9.2 mV (wire more positive, bath more negative) (Kruyt and Overbeek 1969). The calculated shift of the junction between the 150 mM NaCl of the extracellular micropipette and interstitial fluid is 2.05 mV (pipette more positive). During perfusion with H-40 solution, the actual extracellular baseline potential change was 3.9 ± 0.5 mV (n = 15); during F-40 perfusion, which is electrochemically equivalent, change was 10.5 ± 0.42 mV (n = 5). It seems that voltages of unidentified source were added to the junction potentials. The potential of the ground electrode does not influence the recording when the pipette voltage is referred to the extracellular tissue potential instead of "ground." EPSC amplitude increased during H-40 perfusion to 203 ± 16% of control amplitude (n = 16) when the voltage was clamped with reference to bath ground, and to 191 ± 23% (n = 7) with reference in the tissue. The IPSC grew to 127 ± 10% of control when referred to bath and to 138 ± 15% with reference in the tissue. The differences are small, amounting to less thanthe SE.

If the neuronal cytosol was diluted during hypotonic swelling, then the cytosol/patch pipette junction would also be altered. The holding current, which was usually but not always negative (mean 35.7 pA, range 136 to +59 pA), consistently became more positive, on average by 33.5 ± 5.5 pA (n = 15), during H-40 hypotonia. Some of the change of holding current can be attributed to the shift of the summed junction potentials and perhaps to other uncontrolled factors. Because linear leak was subtracted and baselines were adjusted, changes of holding current did not influence measured EPSC and IPSC amplitudes.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Identification of the synaptic currents

Timing, polarity, and voltage dependence identified the two synaptic whole cell currents. The inward current was maximal at hyperpolarized V_hs, where the driving potential of the glutamatergic EPSCs is known to be strongest, wheras the outward current grew as the cell was depolarized, as expected for -aminobutyric acid (GABA)-controlled IPSCs, whether mediated by Cl or by K⁺ (Kaila 1994; Otis et al. 1993). Moreover, in all cases the outward current reached maximum later than did the inward current. The EPSC evoked by Schaffer collateral-commissural volleys is known to be transmitted monosynaptically, whereas IPSCs evoked by the same volley require at least one intermediate synapse if the stimulating electrode is placed >0.5 mm from the recording point (Andersen et al. 1964a,b; Schwartzkroin and Mueller 1987; Turner 1990; Zhu and Krnjevi 1994).

Even though the IPSC reached maximum later than did the EPSC, the two overlapped in time. While both were activated, the two conductances must have interfered with each other. Even though at hyperpolarized potentials there was no detectable outward current, the GABA-controlled Cl and/or K⁺ conductance presumably curtailed the glutamate-activated inward current. At depolarized potentials the glutamate-controlled conductance must have limited the GABA-induced outward current. From the reported time courses of the two currents (Andersen et al. 1964a,b; Schwartzkroin and Mueller 1987; Turner 1990) it seems that the lingering EPSC could subtract more from the amplitude of the IPSC than the IPSC could from the EPSC.

True reversal (0 current) potentials could not be determined for the synaptic currents, in part because of the overlapping time courses and in part because of the location of the synapses. Synapses on the dendritic tree lie outside the effective range of the voltage clamp (Rall 1969).

Interpretation hypotonia-induced synaptic current changes. Possible sources of error

The increase of EPSC amplitude during exposure to low _o was robust, reproducible, concentration-dependent, and reversible. Isosmotic low [NaCl] treatment caused a similar if on the average less powerful effect. These results are especially striking, because the reduced extracellular [Na⁺] ([Na⁺]_o) would be expected to lessen both the driving force and the conductance of glutamate-induced inward current. The observed increases of whole cell current must therefore underestimate the actual change in channel function.

The measured change of the EPSC could be subject to some error due to the change in junction potential. A change in junction potential shifts the I-V plot along the abscissa but does not alter its slope. The slope of the I-V plots of EPSC amplitude became much steeper under the influence of hypotonic exposure, indicating a change in slope conductance. Moreover, the change of EPSC amplitudes measured during hypotonia either with reference to ground or to tissue interstitium were not significantly different.

Changes in IPSC amplitude were inconsistent. Inasmuch as the peak of the IPSC was dominated by GABA_A-induced current (Kaila 1994; Schwartzkroin and Mueller 1987), the reduced [Cl]_o is expected to lessen the driving force and conductance of the IPSC, similarly to the effect of low [Na⁺]_o on the EPSC. Nevertheless, more often than not, the IPSC also increased (Fig. 6), in a few cases even more than the EPSC. A problem with evaluating the data is the overlap of the inward and outward currents. Although each must have influenced the other, the EPSC probably spilled over into the IPSC more than the reverse. A powerful increase of the EPSC could therefore obscure an increase of the IPSC. This reasoning is reinforced by the observation in a few cells in which the IPSC was considerably increased after 15 min of hypotonic treatment, but barely or not at all different from control at the end of 25 min. It seems likely that hypotonia augments the IPSC, but to definitely resolve this question, monosynaptically elicited IPSCs will have to be recorded in isolation after pharmacological suppression of EPSCs (Zhu and Krnjevi 1994).

Neuron dendrites swell in hypotonic solution

While studying neurons freshly dissociated from rat hippocampal CA1 region, we were surprised that brief exposure to very strongly hypotonic solution did not cause the cells to swell (Somjen et al. 1993). A more detailed recent series of experiments revealed that the volume of most isolated hippocampal neurons does increase when they are placed in hypotonic solution, but the swelling usually begins only after a latent period that can sometimes last several minutes (Somjen et al. 1996).

Earlier we also estimated the swelling of cells in hippocampal tissue slices, with the use of the shrinkage of interstitial volume as an index of cell volume expansion (Chebabo et al. 1995b). Graded lowering of _o appeared to cause concentration-dependent cell swelling in hippocampal slices, and there were also indications of regulatory volume decrease during prolonged hypotonia. Because of the slow exchange of the interstitial fluid of the slice with the bath, the cells in the slice experience a more gradual change of _o than do isolated cells. Both the washout of Na⁺ and the shrinkage of interstitial volume take 20-25 min to complete (Chebabo et al. 1995b), similarly to the changes in extracellular electrical signals (Chebabo et al. 1995a) and synaptic currents (Fig. 6).

Measurements of interstitial volume (Chebabo et al. 1995b) did not distinguish, however, between the swelling of neurons and of glial elements. In the present experiments, the calculated C_N of the patch-clamped neurons increased reversibly during exposure to low _o, but not during exposure to isosmotic low-[NaCl] solutions (Table 2). The question is, can we infer hypotonic swelling?

Rall (1969; also Jackson 1992) gave the theoretical treatment of current or voltage steps applied to the soma ofa neuron with an extended dendritic tree. In our trials,5-mV step hyperpolarization evoked a transient current the relaxation of which usually could be fitted by a second-order exponential. This capacitive current could not be eliminated by the compensation circuit of the amplifier. The amplitude of the capacitive current was smaller with cell capacity compensation than without, but ₁ was not affected by cell capacity compensation. This longer time constant is probably related to Rall's ₁, representing an aggregate time constant of the dendritic tree (Rall 1969), mostly outside the reach of the clamp circuit. Hypotonic treatment caused both ₂ and ₁ to increase, but ₂ could not reliably be measured after cell capacity compensation.

To derive C_N, ₁ had to be corrected by R_N. The measured R_N includes the uncompensated fraction of the access resistance, plus leak resistance (which is the voltage-independent resistance of the plasma membrane shunted somewhat through the pipette-to-membrane seal), plus the resistance of the extracellular tissue (and bath) to ground. Of these components, resistance of the cell membrane is by far the greatest. The leak current flows not only through the soma membrane but also, accessed by way of dendritic cytoplasm, through the wide expanse of the dendritic membranes. Although the calculated C_N cannot be equated with the absolute value of cell membrane capacitance, the changes in C_N can be taken to reflect changes in membrane capacitance, very probably due mainly to changes in dendritic surface area.

We conclude that the increased C_N indicates a hypotonic expansion of the surface area of the neuronal dendritic tree, perhaps also of the neuronal soma. It is generally held that biological membranes can be stretched only slightly (Evans and Hochmut 1978). Cell swelling without stressing the membrane is attributed either to the unfolding of previously bent or pleated membrane, or to the addition ("traffic") of "spare" membrane from cell organelles to the plasma membrane. Smoothing of a wrinkled membrane would add to the capacitance only if the pleating was so tight that fluid was excluded between apposed membrane surfaces.

Possible mechanisms of the EPSC increase

Swelling of dendritic shafts (reduced core resistance) and increased "resting" membrane resistance (reduced leak) would expand the dendritic space constant and thereby increase the impact of distant synapses at the soma (Turner 1984). There was evidence for dendritic swelling during hypotonia, but without more detailed data it is difficult to model the magnitude of such electrotonic effects. In addition, because cell swelling reduces interstitial space and perhaps also the width of the synaptic cleft, it could bring presynaptic terminals closer to postsynaptic membranes. Moreover, swelling of perisynaptic glia could hinder the diffusion of transmitter away from the synaptic cleft. These latter two effects could combine to raise transmitter concentration at the receptive surface.

None of these geometric and physical factors could, however, play a part in the enhancement seen with isosmotic low [NaCl], which causes no increase of C_N (Table 2) and little or no change in interstitial volume (Chebabo et al. 1995b). The data in Table 1 as well as our previous measurements (Chebabo et al. 1995a) show the enhancement of the EPSC caused by F-40 solution to be about half of that caused by H-40 solution. It is tempting to suggest that the physical effects of swelling account for half the hypotonic enhancement, and the reduced [NaCl] accounts for the remainder. We must then seek an explanation for the part of the effect that is due to the low [NaCl]_o.

Increased CNS excitability induced by low [Cl]_o has been attributed to reduced synaptic inhibition (Andersen et al. 1964a,b; Chamberlin and Dingledine 1988; Kaila 1994). In our trials in which both [Na⁺]_o and [Cl]_o were reduced, IPSCs were, however, more often increased than decreased or unchanged, and therefore reduced inhibition could not contribute to the increased EPSC.

We must conclude that there was a genuine enhancement of synaptic efficacy, well demonstrated by the increased whole cell conductance associated with the peak EPSC (Fig. 5) and in many cases also with the IPSC. This may seem in conflict with findings in isolated CA1 neurons, where muscimol-induced currents were depressed when the cell was exposed to low _o or low [NaCl]_o (Vreugdenhil et al. 1995). Enhancement of IPSCs could be due to presynaptic facilitation, and competition between postsynaptic depression and presynaptic enhancement could contribute to the variability of IPSC changes. We have no comparable data for glutamate-induced currents, but Paoletti and Ascher (1994) reported that NMDA-induced currents in cultured mouse neurons were enhanced by low osmolarity.

A possible clue for the mechanism of enhanced synaptic efficacy could be found in the lowering of [Ca²⁺]_o caused by both low _o and low [NaCl] (Chebabo et al. 1995b). Lowering of [Ca²⁺]_o indicates uptake of Ca²⁺ ions by cells, and recently we confirmed that intracellular [Ca²⁺] ([Ca²⁺]_i) measured with the fluorescent indicator FURA-2 is initially depressed but then raised well above control level during hypotonia in neurons in hippocampal slices (Somjen et al. 1997). The uptake could in part or whole be due to suppression or reversal of the Ca/Na exchange transport. Elevated [Ca²⁺]_i could enhance synaptic efficacy both pre- and postsynaptically. Presynaptic uptake of Ca²⁺ could enhance synaptic transmission because an increase of baseline [Ca²⁺]_i in the terminals could facilitate impulse-evoked transmitter release (Katz and Miledi 1968; Moore and Hines 1986). Contrary to this idea, in isolated cells hypotonia suppresses whole cell calcium currents (Somjen et al. 1993), but these neurons lose their dendritic tree during cell preparation and calcium channels in dendrites and in presynaptic terminals may behave differently from those in the soma membrane. Besides, we have recently seen that not only baseline [Ca²⁺]_i, but also [Ca²⁺]_i responses evoked by depolarizing pulses in neurons in hippocampal slices, are initially depressed but later enhanced during hypotonia (Somjen et al. 1997). In the present experiments there was also evidence of enhancement of Ca²⁺-mediated voltage-dependent inward currents, illustrated in Fig. 1, B and C. If Ca²⁺ currents are similarly augmented in presynaptic terminals as well, this would lead to an increased release of transmitter substances. In an earlier study we found indications of enhanced presynaptic Ca²⁺ uptake during hypotonia, but the data were not entirely conclusive because of technical problems (Chebabo et al. 1995a).

In addition to a presynaptic effect, it has been reported that a rise in [Ca²⁺]_i potentiates the postsynaptic response of pyramidal neurons (Kullman et al. 1992). Responses of [Ca²⁺]_i would be dampened by the EGTA in the pipette solution in patch-clamped cell somata, but much less so in the dendritic branches where the potentiating effect would take place.

Clinical significance

Clinically, acute hypotonia of extracellular fluid is by necessity always associated with hyponatremia. The normal range of Na⁺ levels in the blood of human subjects is 136-145 meq/l, but overt signs of hyponatremia start only when [Na⁺]_o drops below 125 meq/l (Thorn et al. 1977). In the H-20 solution used in these experiments, [Na⁺]_o (in the form of NaCl, NaHCO₃ and NaH₂PO₄) was 134 mmol/l, in H-40 solution it was 115 mmol/l, and in H-60 solution it was 95 mmol/l. In human patients the mildest level of hypotonia corresponding to the H-20 solution would, therefore, probably be asymptomatic, the middle level (H-40) corresponds to overt clinical hyponatremia-hypotonia, and the severest level (H-60) represents danger of death (Avner 1995; Ellis 1995; Katzman and Pappius 1973; Thorn et al. 1977; Zelingher et al. 1996). As already mentioned in the INTRODUCTION, slice experiments do not faithfully represent all the clinical features of hyponatremia, which include changes in systemic circulation and intracranial pressure. Nevertheless, the enhanced synaptic transmission documented in this study almost certainly does play a part in the clinical picture of acute hyponatremia-hypotonia.

	ACKNOWLEDGEMENTS

  We thank Dr. A. M. van Dongen for critical reading of the manuscript and helpful discussions.

  This work was supported by National Institute of Neurological Disorders and Stroke Grant NS-18670.

	FOOTNOTES

  Address for reprint requests: G. G. Somjen, Box 3709, Duke University Medical Center, Durham NC 27710.

  Received 21 August 1996; accepted in final form 6 January 1997.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• ANDERSEN, P., ECCLES, J. C., LØYNING, Y. Location of postsynaptic inhibitory synapses on hippocampal pyramids. J. Neurophysiol. 27: 592-607, 1964a.[Free Full Text]
• ANDERSEN, P., ECCLES, J. C., LØYNING, Y. Pathway of postsynaptic inhibition in the hippocampus. J. Neurophysiol. 27: 608-619, 1964b.[Free Full Text]
• ANDREW, R. D. Seizure and acute osmotic change: clinical and neurophysiological aspects. J. Neurol. Sci. 101: 7-18, 1991.[Medline]
• ANDREW, R. D., FAGAN, M., BALLYK, B. A., ROSEN, A. S. Seizure susceptibility and the osmotic state. Brain Res. 498: 175-180, 1989.[Medline]
• AVNER, E. D. Clinical disorders of water metabolism: hyponatremia and hypernatremia. Pediatr. Ann. 24: 23-30, 1995.[Medline]
• BALLYK, B. A., QUACKENBUSH, S. J., ANDREW, R. D. Osmotic effects on the CA1 neuronal population in hippocampal slices with special reference to glucose. J. Neurophysiol. 65: 1055-1066, 1991.[Abstract/Free Full Text]
• BLANTON, M. G., TURCO, J.J.L., KRIEGSTEIN, A. R. Whole cell recording from neurons in slices of reptilian and mammalian cerebral cortex. J. Neurosci. Methods 30: 203-210, 1989.[Medline]
• BOSSUT, D. F., SOMJEN, G. G. Postsynaptic currents of hippocampal neurons are enhanced in low osmotic pressure (Abstract). Eur. J. Neurosci. Suppl. 8: 147, 1995.
• BRADBURY, M. In: The Concept Of The Blood-Brain Barrier., . Chichester, UK: Wiley, 1979
• CHAMBERLIN, N. L., DINGLEDINE, R. GABAergic inhibition and the induction of spontaneous epileptiform activity by low chloride and high potassium in the hippocampal slice. Brain Res. 445: 12-18, 1988.[Medline]
• CHEBABO, S. R., HESTER, M. A., AITKEN, P. G., SOMJEN, G. G. Hypotonic exposure enhances synaptic transmission and triggers spreading depression in hippocampal tissue slices. Brain Res. 695: 203-216, 1995a.[Medline]
• CHEBABO, S. R., HESTER, M. A., JING, J., AITKEN, P. G., SOMJEN, G. G. Interstitial space, electrical resistance and ion concentrations during hypotonia of hippocampal slices of rats. J. Physiol. Lond. 487: 685-697, 1995b.[Abstract/Free Full Text]
• CZÉH, G., AITKEN, P. G., SOMJEN, G. G. Membrane currents in CA1 hippocampal cells during spreading depression (SD) and SD-like hypoxic depolarization. Brain Res. 632: 195-208, 1993.[Medline]
• DINGLEDINE, R. In: Brain Slices., . New York: Plenum, 1984
• DUDEK, F. E., OBENAUS, A., TASKER, J. G. Osmolality-induced changes in extracellular volume alter epileptiform bursts independently of chemical synapses in the rat: importance of non-synaptic mechanisms in hippocampal epileptogenesis. Neurosci. Lett. 120: 267-270, 1990.[Medline]
• ELLIS, S. J. Severe hyponatremia: complications and treatment. Q. J. Med. 88: 905-909, 1995.
• EVANS, E. A., HOCHMUT, R. M. Mechanochemical properties of membranes. In: Current Topics in Membrane Transport, edited by F. Bronner, and A. Kleinzeller . New York: Academic, 1978, vol. 10, p. 1-64
• FRENCH, C. R., SAH, P., BUCKETT, K. J., GAGE, P. W. A voltage-dependent persistent sodium current in mammalian hippocampal neurons. J. Gen. Physiol. 95: 1139-1157, 1990.[Abstract/Free Full Text]
• HOEHN, K., WATSON, T.W.J., MACVICAR, B. A. A novel tetrodotoxin-insensitive slow sodium current in striatal and hippocampal neurons. Neuron 10: 543-552, 1993.[Medline]
• HUANG, R., AITKEN, P. G., SOMJEN, G. G. The extent and mechanism of the loss of function caused by strongly hypotonic solutions in rat hippocampal tissue slices. Brain Res. 695: 195-202, 1995.[Medline]
• HUANG, R., BOSSUT, D. F., LEWIS, D. V., SOMJEN, G. G. Enhancement of whole-cell synaptic currents by low osmolarity and low [NaCl]_o in rat hippocampal neurons. Soc. Neurosci. Abstr. 22: 1741, 1996.
• JACKSON, M. B. Cable analysis with whole-cell patch clamp. Theory and experiment. Biophys. J. 61: 756-766, 1992.[Medline]
• KAILA, K. Ionic basis of GABA-A receptor channel function in the nervous system. Prog. Neurobiol. 42: 489-537, 1994.[Medline]
• KATZ, B., MILEDI, R. The role of calcium in neuromuscular facilitation. J. Physiol. Lond. 195: 481-492, 1968.[Abstract/Free Full Text]
• KATZMAN, R., PAPPIUS, H. M. In: Brain Electrolytes and Fluid Metabolism., . Baltimore, MD: Williams & Wilkins, 1973
• KRUYT, H. R. AND OVERBEEK, J.T.G. Inleiding Tot De Fysische Chemie. Amsterdam: Paris, 1969.
• KULLMANN, D. M., PERKEL, D. J., MANABE, T., NICOLL, R. Ca²⁺ entry via postsynaptic voltage-sensitive Ca²⁺ channels transiently potentiates excitatory synaptic transmission in the hippocampus. Neuron 9: 1175-1183, 1992.[Medline]
• MOORE, J. W., HINES, M. Some consequences of intracellular calcium binding on phasic transmitter release. In: Calcium, Neuronal Function and Transmitter Release, edited by R. Rahamimoff, and B. Katz . Boston, MA: Nijhoff, 1986, p. 115-137
• OTIS, T. S., DEKONINCK, Y., MODY, I. Characterization of synaptically elicited GABA-B responses using patch-clamp recordings in rat hippocampal slices. J. Physiol. Lond. 463: 391-407, 1993.[Abstract/Free Full Text]
• PAOLETTI, P., ASCHER, P. Mechanosensitivity of NMDA receptors in cultured mouse central neurons. Neuron 13: 645-655, 1994.[Medline]
• RALL, W. Time constants and electrotonic length of membrane cylinders and neurons. Biophys. J. 9: 1483-1508, 1969.
• ROSEN, A. S., ANDREW, R. D. Osmotic effects upon excitability in rat neocortical slices. Neuroscience 38: 579-590, 1990.[Medline]
• SCHWARTZKROIN, P. A., MUELLER, A. L. Electrophysiology of hippocampal neurons. In: Cerebral Cortex, edited by E. G. Jones, and A. Peters . New York: Plenum, 1987, vol. 6, p. 295-343
• SOMJEN, G. G., AITKEN, P. G., BALESTRINO, M., SCHIFF, S. J. Uses and abuses of in vitro systems in the study of the pathophysiology of the central nervous system. In: Brain Slices: Fundamentals, Applications and Implications, edited by A. Schurr, T. Teyler, and M. Rigor . Basel: Karger, 1986, p. 89-104
• SOMJEN, G. G., BORGDORFF, A. J., AND WADMAN, W. J. Free intracellular calcium in neurons in rat hippocampal slices during hypo-osmotic stress. 33rd Int. Physiol. Congr. Abstr. In press.
• SOMJEN, G. G., FAAS, G. C., VREUGDENHIL, M., WADMAN, W. J. Channel shutdown: a response of hippocampal neurons to adverse environments. Brain Res. 632: 180-194, 1993.[Medline]
• SOMJEN, G. G., WADMAN, W. J., JUTA, A., BORGDORF, A., AITKEN, P. G., KIEHART, D. P. Osmotically induced volume changes of freshly isolated hippocampal neurons (Abstract). FASEB J. 10: 86, 1996.
• THORN, G. W., ADAMS, R. D., BRAUNWALD, E., ISSELBACHER, K. J., AND PETERSDORF, R. G. (Editors). Harrison's Principles of Internal Medicine. New York: McGraw-Hill, 1977.
• TURNER, D. A. Conductance transients onto dendritic spines in a segmental cable model of hippocampal neurons. Biophys. J. 46: 85-96, 1984.[Medline]
• TURNER, D. A. Feed-forward inhibitory potentials and excitatory interactions in guinea pig hippocampal pyramidal cells. J. Physiol. Lond. 422: 333-350, 1990.[Abstract/Free Full Text]
• VREUGDENHIL, M., SOMJEN, G. G., WADMAN, W. J. Effects of strongly anisosmotic and NaCl deficient solutions on muscimol- and glutamate evoked whole-cell currents in freshly dissociated hippocampal neurons. Brain Res. 670: 89-96, 1995.[Medline]
• YAMAGUCHI, M., WU, S., EHARA, K., NAGASHIMA, T., TAMAKI, N. Cerebral blood flow in rats with water-intoxicated brain edema. Acta Neurochir. 60, Suppl.: 190-192, 1994.
• ZELINGHER, J., PUTTERMAN, C., ILAN, Y., DANN, E. J., ZVEIBIL, F., SHVIL, Y., GALUN, E. Case series: hyponatremia associated with moderate exercise. Am. J. Med. Sci. 311: 86-91, 1996.[Medline]
• ZHU, P. J., KRNJEVI, K. Anoxia selectively depresses excitatory synaptic transmission in hippocampal slices. Neurosci. Lett. 166: 27-30, 1994.[Medline]

0022-3077/97 $5.00 Copyright ©1997 The American Physiological Society