Effects of Divalent Cations on Schaffer Collateral Axon Function

Abstract

Previously, we reported that distal Schaffer collaterals undergo biphasic changes in excitability during high-frequency stimulation (HFS), with an early hyper-excitability period followed by an excitability depression period. The extracellular divalent cations calcium and magnesium can regulate membrane excitability in neuronal tissue. Therefore, we hypothesized that altering the concentrations of extracellular calcium and magnesium would alter the biphasic excitability changes. We tested this hypothesis by recording distal Schaffer collateral fiber volleys in stratum radiatum of hippocampal area CA1 during 100 Hz HFS in artificial cerebral spinal fluid (ACSF) containing normal and altered concentrations of extracellular divalent cations. Our normal ACSF contained 2.0 mM calcium and 2.0 mM magnesium. We examined four solutions with altered divalent cation concentrations: 1) high-calcium/low-magnesium (3.8 mM/0.2 mM), 2) low-calcium/high-magnesium (0.2 mM/3.8 mM), 3) high-calcium/normal-magnesium (3.8 mM/2.0 mM), or 4) normal-calcium/high-magnesium (2.0 mM/10.0 mM), and assessed the effects on Schaffer collateral responses. Increasing or decreasing extracellular calcium enhanced or reduced (respectively) the early hyper-excitable period whereas increasing extracellular magnesium reduced the later excitability depression. Because these results might be explained by altered calcium influx through voltage-gated calcium (Ca_V) channels, we tested Ca_V blockers (ω–agatoxin IVA, ω-conotoxin-GVIA, cadmium), but observed no effects on responses during HFS. Some of the effects of altered divalent cation concentration may be explained by altered membrane surface charge. Although this mechanism does not completely explain our findings, calcium influx through Ca_V channels is not required.

INTRODUCTION

Extracellular calcium and magnesium ions are well known for their ability to regulate membrane excitability in both neuronal (Katz and Miledi 1967, Taylor and Dudek 1982, Mody et al. 1987) and non-neuronal (Hall and Fry 1992) tissues. Several mechanisms contribute to these effects on excitability. Extracellular divalent cations alter membrane surface charge by interacting with negatively-charged membrane proteins or lipids, thereby affecting voltage-dependent ions channel gating and kinetics (Bennet et al. 1997, Hanck and Sheets 1992, reviewed in Green and Andersen 1991). In addition, extracellular calcium ions may alter channel function by binding to membrane surface receptors or changing influx through calcium-permeable channels (reviewed in Jones and Smith 2016). Finally, altered synaptic function may also contribute via effects on neurotransmitter release (Dodge and Rahamimoff 1967) or postsynaptic receptors (e.g., magnesium block of NMDA glutamate receptors, Mayer et al. 1984, Nowak et al. 1984).

Because of this ability to regulate neuronal excitability and ion channel function, we examined the roles of calcium and magnesium on Schaffer collateral axon function during high-frequency stimulation (HFS). Previously, we described changes in hippocampal Schaffer collateral axon function during HFS (Kim et al. 2012, Owen and Grover 2015) in which distal segments of Schaffer collaterals undergo biphasic changes in excitability during HFS, with an initial, transient increase in excitability (hyper-excitability) followed by a sustained decrease in excitability (depression). Given the major effects of extracellular divalent cations on membrane excitability, we hypothesized that altering extracellular calcium or magnesium concentrations would affect the biphasic changes in Schaffer collateral excitability that occur during HFS. We tested this hypothesis by comparing extracellular recordings of Schaffer collateral fiber volleys in normal extracellular solution and in extracellular solutions with altered divalent cation concentrations. We found that changes in extracellular calcium, with or without changes in extracellular magnesium, altered the early hyper-excitable phase early during HFS, whereas changes in extracellular magnesium altered the later phase of depression.

EXPERIMENTAL PROCEDURES

Slice Preparation

All procedures were approved by the Institutional Animal Care and Use Committee at Marshall University. Hippocampal slices were prepared as previously described (Kim et al. 2012). Male Sprague-Dawley rats (60 days old; Hilltop Lab Animals, Scottdale, PA; RRID:RGD 25824850) were sedated by CO₂-air inhalation and decapitated. The brain was removed and placed into chilled artificial cerebrospinal fluid (ACSF) composed of the following (in mM): 124 NaCl, 26 NaHCO₃, 3.4 KCl, 1.2 NaH₂PO₄, 2 CaCl₂, 2 MgSO₄, and 10 glucose (pH 7.35, equilibrated with 95% O₂/5% CO₂). A block containing both hippocampi was glued to the stage of a vibratome (Campden Instruments, Lafayette, IN; RRID:SCR_018460), immersed in chilled ACSF, and sectioned into 400 – 500 μm-thick slices in the coronal or horizontal plane. Slices were dissected to remove the hippocampus from surrounding structures and stored at room temperature (20 – 22 °C) in an interface-holding chamber. For recordings, individual slices were transferred to a small-volume (~200 μL) interface-recording chamber heated to 34.5–35.5 °C and perfused with oxygenated ACSF at a rate of 1.0–1.5 ml/min.

Field Potential Recordings

Extracellular field potentials were recorded through glass micropipettes that were pulled with a P97 Flaming/Brown micropipette puller (Sutter Instruments; RRID:SCR_016842) and filled with ACSF (3 – 5 MΩ); in some recordings, the tip was broken before placement in the slice to lower resistance (1 – 2 MΩ) and noise. Fiber volleys were recorded in CA1 stratum radiatum with a DAM50 amplifier (WPI; RRID:SCR_018415) providing a gain of 1,000. Signals were band-pass filtered (0.1 – 3000 Hz) and digitized at 100 kHz. Digitized signals were stored on a personal computer running Windows XP (Microsoft) using WinWCP and WinEDR software (Strathclyde Electrophysiology Suite, John Dempster, University of Strathclyde, http://spider.science.strath.ac.uk/sipbs/software_ses.htm, RRID:SCR_014270). Fiber volley amplitudes were measured as the difference between the maximum negativity and following positivity. Latencies were measured as the time difference between the beginning of the stimulus artifact and the response at 10% of peak amplitude. Half-widths were measured as response duration at 50% of amplitude.

Stimulation

A bipolar Teflon-insulated, stainless steel stimulating electrode was placed in stratum radiatum near the border of areas CA3 and CA1 in the approximate location of area CA2. Constant voltage biphasic stimuli (0.1 ms duration) were delivered using an A-M Systems model 2100 stimulator (RRID:SCR_016677). The stimulus intensity was adjusted to produce the largest response that could be obtained without contamination of the response by the stimulus artifact. Stimulus intensities were 4 – 10 V. Low-frequency stimulation (LFS) and HFS were performed in the same slice. For LFS, stimuli were delivered at 0.06 Hz. For HFS, trains of 160 stimuli were delivered at 100 Hz. Fiber volleys were isolated using an AMPA receptor blocker (30 μM 6,7-dinitroquinoxaline-2,3-dione, DNQX; RRID:CAS_2379–57-9). In some recordings an NMDA receptor blocker (CGP-37849, 5 μM; RRID:CAS_127910–31-0), alone or in combination with a GABA_A receptor blocker, (bicuculline methiodie, 10 μM; RRID:CAS_40709–69-1) was applied with DNQX.

To determine the effects of altering extracellular divalent cation concentrations, fiber volley responses to LFS and HFS were first recorded in normal divalent cation concentrations, followed by a solution of either high-calcium/low-magnesium (HiCal/LoMag; 3.8 mM CaCl₂/0.2 mM MgSO₄), low-calcium/high-magnesium (LoCal/HiMag; 0.2 mM/3.8 mM MgSO₄), high-calcium/normal magnesium (HiCal; 3.8 mM CaCl₂/2.0 mM MgSO₄), or normal-calcium/high-magnesium (HiMag; 2.0 mM CaCl₂/10.0 mM MgSO₄). Application times for solutions with altered ion concentrations (10 – 20 min) were determined in separate control experiments by the time required to produce a stable change in excitatory synaptic transmission.

To examine the contribution of Ca_V channels, fiber volley responses during HFS were recorded in normal ACSF before and after applying a non-specific Ca_V channel (600 μM cadmium, Cd²⁺), a specific Ca_V2.1 blocker (200 nM ω–agatoxin IVA; RRID:CAS_145017–83-0), or a specific Ca_V2.2 blcokers (1 μM ω-conotoxin-GVIA; RRID:CAS_106375–28-4). Application times and concentrations for Ca_V channel blockers (10 min for Cd²⁺, 30 min for ω–agatoxin IVA and ω-Ctx-GVIA) were taken from separate control experiments which assessed inhibition of evoked excitatory synaptic transmission.

Reagents

Drugs were prepared as concentrated stock solutions and stored at −20° C. DNQX (Sigma-Aldrich or Tocris), was dissolved in DMSO; bicuculline methiodide (Sigma-Aldrich or Tocris) and CGP-37849 (Sigma-Aldrich or Tocris) were dissolved in distilled water; ω-Agtx-IVA (Alomone) and ω-Ctx-GVIA (Alomone) were dissolved in sterile phosphate-buffered saline. Stock solutions were diluted to final concentrations by addition to ACSF perfusing the tissue. Salts and all other reagents were from either Fisher Scientific or Sigma-Aldrich.

Data Analysis and Statistics

Fiber volleys were analyzed for amplitude, latency, and half-width using WinWCP and custom routines written in the Python programming language (https://www.python.org, RRID:SCR_008394). For comparison among slices, amplitudes were normalized relative to the first response recorded during HFS. To compare the effects of changing divalent cation concentrations or applying channel blockers, two-way (solution and time) repeated measures ANOVAs were performed on the means of the first 10 and last 10 responses during HFS as these reflect the hyper-excitable and depression periods, with post-hoc analysis using Student’s t-test with Bonferroni correction if a significant effect of solution and/or significant interaction between solution and time were detected. To assess effects on responses evoked by LFS, responses were averaged over 5 min periods immediately preceding the start of solution change and after solution change was completed compared by Student’s t-tests. Means, standard errors and t-tests were calculated using Gnumeric (RRID:SCR_018462). A p-value of 0.05 or less indicated significance. All data are reported as mean ± S.E.M.

RESULTS

Schaffer Collateral Hyper-excitability During HFS Is Modulated By Extracellular Divalent Cation Concentrations

We previously reported that distal Schaffer collateral excitability, measured as fiber volley amplitude and latency changes, is altered in a frequency-dependent manner during continuous HFS (Kim et al. 2012, Owen and Grover 2015, Owen et al. 2017). Early during HFS, Schaffer collateral excitability increases, reaching a peak at around stimulus 10. This hyper-excitable period is followed by a period of sustained excitability depression. Both the early hyper-excitability and late depression periods are frequency-dependent over the range of 10 – 100 Hz (Owen and Grover 2015). As changing the concentrations of divalent cations, such as calcium, can modulate excitability (Adelman and Moore 1961, Traub et al. 1994, Bikson et al. 1999), we tested if distal Schaffer collateral excitability during LFS and the biphasic excitability changes during 100 Hz HFS are modulated by altering extracellular divalent cation concentrations. In an initial set of experiments (n = 12), we recorded fiber volleys in three different solutions: control (normal calcium and magnesium concentrations), HiCal/LoMag, and LoCal/HiMag. In both the HiCal/LoMag and LoCal/HiMag solutions, the total divalent cation concentration (4.0 mM) was kept equal to that of the control solution.

The effects of manipulating extracellular divalent cation concentrations during LFS are summarized in Table 1. Fiber volleys evoked by LFS decreased in amplitude in HiCal/LoMag solution (control vs. HiCal/LoMag: 0.53 ± 0.08 mV vs. 0.45 ± 0.07 mV, p < 0.05; n = 12 for both) while latencies slightly, but significantly, increased (control vs. HiCal/LoMag: 1.85 ± 0.10 ms vs. 1.90 ± 0.09 ms, p < 0.01), indicating a reduction in Schaffer collateral excitability. Changes in amplitude or latency were not observed in the LoCal/HiMag solution (amplitude: p = 0.12; latency: p = 0.37). Fiber volley half-widths were unaltered by either the HiCal/LoMag (p = 0.36) or LoCal/HiMag (p = 0.14) solutions.

Table 1.

Effects of Extracellular Divalent Cations on Low-frequency Stimulation

Condition	[Ca2+]o	[Mg2+]o	Amplitude (mV)	Latency (ms)	Half-width (ms)
Control (n = 12)	2.0	2.0	0.62 ± 0.12	1.60 ± 0.10	0.67 ± 0.04
Low Ca2+/High Mg2+	0.2	3.8	0.67± 0.13	1.59 ± 0.10	0.65 ± 0.04
Control (n = 12)	2.0	2.0	0.53 ± 0.08	1.85 ± 0.10	0.68 ± 0.04
High Ca2+/Low Mg2+	3.8	0.2	0.45 ± 0.07 *	1.90 ± 0.09 **	0.70 ± 0.03
Control (n = 7)	2.0	2.0	0.44 ± 0.04	1.51 ± 0.04	0.71 ± 0.04
High Ca2+	3.8	2.0	0.42 ± 0.04 *	1.55 ± 0.04 *	0.75 ± 0.04 **
Control (n = 9)	2.0	2.0	0.62 ± 0.05	1.57 ± 0.12	0.73 ± 0.04
High Mg2+	2.0	10.0	0.50 ± 0.06 **	1.64 ± 0.10	0.86 ± 0.08

Bold/italics indicates significant difference compared to control:

^*p<0.05

^**p<0.01.

During HFS in control solution, Schaffer collaterals underwent the same biphasic changes described previously (Kim et al. 2012, Owen and Grover 2015, Owen et al. 2017); namely, an early hyper-excitability, where fiber volley amplitudes increased and latencies decreased, followed by excitability depression, where amplitudes decreased and latencies increased (Figs. 1B – C and 2B – C). After changing from control to HiCal/LoMag solution, the normalized amplitudes of the first 10 (hyper-excitability period) and last 10 responses (depression period) were increased (Fig. 1B), suggesting an increase in excitability. Significant effects of solution and time were detected by ANOVA (solution- F ratio: 11.88, p < 0.01; time-F ratio: 199.5, p < 0.001) as well as a significant interaction between solution and time (F ratio: 6.871, p < 0.01). Post-hoc analysis detected significant increases in normalized fiber volley amplitudes during the first 10 (control vs. HiCal/LoMag: 125.24 ± 7.29 % vs. 156.02 ± 12.28 %, p < 0.01) and last 10 (control vs. HiCal/LoMag: 23.2 ± 2.5 % vs. 32.9 ± 4.6 %, p < 0.01) stimuli. No effect of HiCal/LoMag solution was detected for either fiber volley latency (Fig. 1C; p = 0.19) nor half-width changes (Fig. 1D; p = 0.98), nor was a significant interaction detected (latency change- p = 0.14; half-widths- p = 0.84), although a significant effect of time was detected for fiber volley latency change (F ratio: 108.1, p < 0.001) but not half-widths (p = 0.77). When control solution was replaced with LoCal/HiMag solution (n = 12), normalized amplitudes during the hyper-excitable period (first 10 responses) were reduced compared to control (Fig. 2B;),but were still larger than controls during the depression period (last 10 responses; Fig. 2B). An effect of time (F ratio: 1002, p < 0.001) and a significant interaction between solution and time (F ratio: 11.07, p < 0.01) with post-hoc analysis detecting a significant reduction in fiber volley amplitudes during the first 10 responses (control vs. LoCal/HiMag: 126.57 ± 7.34 % vs. 106.1 ± 2.45 %, p < 0.05) and a significant increase during the last 10 responses (control vs. LoCal/HiMag: 23.6 ± 2.5 % vs. 28.1 ± 3.3%, p < 0.05). Fiber volley latency changes during the first 10 stimuli were reduced in the LoCal/HiMag solution (Fig. 2C), but were not affected during the last 10 stimuli. Only a significant effect of time was detected for fiber volley latency changes (F ratio: 83.09, p < 0.001) with post-hoc analysis detecting a significant reduction during the first 10 stimuli (control vs. LoCal/HiMag: −0.033 ± 0.006 ms vs. −0.004 ± 0.007 ms, p < 0.01) but no effect was detected during the last 10 stimuli (p = 0.91). The effects of LoCal/HiMag solution on amplitude and latency changes during the first 10 stimuli indicate a reduction of the normal hyper-excitability during the initial portion of HFS. No effects of solution (p = 0.84), time (p = 0.21), or an interaction between solution and time (p = 0.93) were observed in fiber volley half-width changes after switching to LoCal/HiMag solution (Fig. 2D;).

Fig. 1. Increasing extracellular calcium and lowering extracellular magnesium enhanced Schaffer collateral hyper-excitability during HFS.

A) Example traces of fiber volleys recorded in control (top) and HiCal/LoMag (bottom) solutions. Fiber volley responses evoked by the first (1) and sixth (6) stimuli are shown on the left; responses evoked by the first (1) and last (160) stimuli are shown at right. B) Normalized Schaffer collateral fiber volley amplitudes recorded during 100 Hz HFS in HiCal/LoMag ACSF were significantly larger than those recorded in control ACSF during the hyper-excitability period (first 10 stimuli) and depression period (last 10 stimuli). C) Fiber volley latency changes did not differ between control or HiCal/LoMag ACSF during either the hyper-excitability or depression periods. D) Fiber volley half-width changes did not differ for either the hyper-excitability or depression periods. Shaded areas indicate ± 1 S.E.M.; **p < 0.01; n = 12 for control and HiCal/LoMag.

Fig. 2. Lowering extracellular calcium and raising extracellular magnesium reduced Schaffer collateral hyper-excitability during HFS.

A) Example traces of fiber volleys recorded in control (top) and LoCal/HiMag (bottom) solutions. Fiber volley responses evoked by the first (1) and sixth (6) stimuli are shown on the left; responses evoked by the first (1) and last (160) stimuli are shown at right. B) The normal fiber volley amplitude increase seen during the first 10 stimuli in control ACSF was abolished by application of LoCal/HiMag solution. LoCal/HiMag solution also slightly, but significantly, reduced fiber volley amplitude depression during the final 10 stimuli. C) The decrease in fiber volley latency normally seen during the first 10 stimuli in control ACSF was significantly reduced by LoCal/HiMag solution, but there was no effect during the depression period (last 10 stimuli). D) Half-width changes were not significantly different between control and LoCal/HiMag solution. Shaded areas indicate ± 1 S.E.M.; *p < 0.05, **p < 0.01; n = 12 for control and LoCal/HiMag.

In summary, increasing extracellular Ca²⁺ concentration while reducing Mg²⁺ concentration to maintain total divalent cation concentration reduced Schaffer collateral excitability during LFS and enhanced hyper-excitability during HFS. The opposite manipulation, decreasing extracellular Ca²⁺ concentration while increasing extracellular Mg²⁺, failed to affect excitability during LFS, but did reduce hyper-excitability during the early portion of HFS. Because both divalent cations were manipulated in these experiments (an increase in one and equal decrease in the other), our observations may be the result of opposing effects counteracting each other. In the next sets of experiments, we increased the concentration of either calcium or magnesium without a compensating decrease in the other divalent ion in order to elucidate the effect of each divalent cation on Schaffer collateral excitability; that is, we allowed total divalent cation concentration to increase.

Increasing Extracellular Calcium and Magnesium Have Different Effects On Schaffer Collateral Excitability During HFS

We first assessed the effect of increasing extracellular calcium from the control level of 2.0 mM to 3.8 mM without changing extracellular magnesium (HiCal solution). The effects on responses during LFS are summarized in Table 1. During LFS, application of HiCal solution significantly decreased fiber volley amplitudes (control vs. HiCal: 0.44 ± 0.04 mV vs. 0.42 ± 0.04 mV, p < 0.05; n = 7) and increased latencies (control vs. HiCal: 1.51 ± 0.04 ms vs. 1.55 ± 0.04 ms, p < 0.05), resembling the effects of HiCal/LoMag. Unlike HiCal/LoMag solution, application of HiCal solution also significantly increased fiber volley half-width (control vs. HiCal: 0.71 ± 0.04 ms vs. 0.75 ± 0.04 ms, p < 0.01). The HiCal solution also increased fiber volley amplitude changes during the hyper-excitability period of HFS (Fig. 3A – B; control vs. HiCal: 110.16 ± 2.07 % vs. 125.8 ± 6.02 %, p < 0.05), similar to what we observed in the HiCal/LoMag solution (compare Figs. 1B and 3B), but had no effect on fiber volley amplitudes during the depression period (Fig. 3A – B; p = 0.16). Significant effects of solution (F ratio: 8.11, p < 0.05) and time (F ratio: 553.8, p < 0.001) were detected. HiCal solution failed to affect latency (Fig. 3C) or half-width (Fig. 3D) changes during either the hyper-excitability period (latency changes, p = 0.71; half-width changes, p = 0.28) or the depression period (latency changes, p = 0.42; half-width changes, p = 0.94) of HFS; no effect of solution or a significant interaction between solution and time was detected for either latency changes (solution- p = 0.465, interaction- p = 0.42) or half-width changes (solution- p = 0.81, interaction- p = 0.72), only a significant effect of time for both measures (latency changes- F ratio: 71.61, p <0.001; half-widths- F ratio: 23.61, p < 0.05).

Fig. 3. Increasing extracellular calcium without changing extracellular magnesium enhanced Schaffer collateral hyper-excitability during HFS.

A) Example traces of fiber volleys recorded in control (top) and HiCal (bottom) solutions. Fiber volley responses evoked by the first (1) and sixth (6) stimuli are shown on the left; responses evoked by the first (1) and last (160) stimuli are shown at right. B) Fiber volley amplitude changes during the hyper-excitability period were significantly larger in HiCal solution compared to control, but were not different during the depression period. No differences in fiber volley latency (C) or half-width changes (D) were observed between control and HiCal solution either during the hyper-excitability or depression periods. Shaded areas indicate ± 1 S.E.M.; *p < 0.05; n = 7 for control and HiCal.

We next assessed the effect of increasing extracellular magnesium from the control level of 2.0 mM to 10 mM without changing extracellular calcium (HiMag solution; n = 9). Data collected during LFS are summarized in Table 1. Fiber volley amplitudes during LFS were significantly decreased after switching to HiMag solution (control vs. HiMag: 0.62 ± 0.05 mV vs. 0.50 ± 0.06 mV, p < 0.01), but no effect was observed on either fiber volley latency (p = 0.19) or half-width (p = 0.11). During HFS, fiber volley amplitudes during the hyper-excitability period remained unchanged whereas amplitude depression during the last 10 stimuli was significantly reduced (Fig. 4B) after applying the HiMag solution. Significant effects of solution (F ratio: 7.716, p <0.05) and time (F ratio: 143.5, p <0.001) were detected by ANOVA with post-hoc analysis confirming a significant reduction in amplitude depression (control vs. HiMag: 28.9 ± 2.5 % vs. 43.3 ± 3.8 %, p < 0.01) but no difference during the hyper-excitable period (p = 0.36). No significant interaction between solution and time was detected for fiber volley amplitudes (p = 0.09). The normal fiber volley latency decrease was attenuated in the presence of HiMag solution (Fig. 4C), with a significant effect of time (F ratio: 28.36, p < 0.01) and a significant interaction between solution and time (F ratio: 7.049, p < 0.05) detected by ANOVA. Post-hoc analysis confirmed the attenuation of fiber volley latency decrease during the early hyper-excitable period (control vs. HiMag: −0.069 ± 0.014 ms vs. 0.021 ± 0.011 ms, p < 0.01). No effect of HiMag solution (p = 0.847) or a significant interaction between solution and time (p = 0.843) was detected in half-width changes (Fig. 4D), only an effect of time (F ratio: 45.6, p < 0.001).

Fig. 4. Increasing extracellular magnesium without changing extracellular calcium reduced fiber volley amplitude depression during HFS.

A) Example traces of fiber volleys recorded in control (top) and HiMag (bottom) solutions. Fiber volley responses evoked by the first (1) and sixth (6) stimuli are shown on the left; responses evoked by the first (1) and last (160) stimuli are shown at right. B) Fiber volley amplitude changes did not differ between control and HiMag solution during the early hyper-excitability period, but fiber volleys showed significantly less amplitude depression in HiMag ACSF compared to control during the last 10 stimuli. C) Fiber volley latency changes were significantly smaller in HiMag solution during the hyper-excitability period compared to control, but no difference was present during the depression period. D) No differences were observed in fiber volley half-width changes between responses recorded in control or HiMag solutions. Shaded areas indicate ± 1 S.E.M.; **p < 0.01. n = 9 for control and HiMag.

Collectively, the results shown above indicate that increasing extracellular calcium, with or without compensatory changes in extracellular magnesium, enhances Schaffer collateral hyper-excitability during early HFS. In contrast, increasing extracellular magnesium alone did not affect responses during the hyper-excitable period, but did reduce the later excitability depression. The effect of extracellular divalent cations might be mediated by calcium currents through voltage-gated calcium (Ca_V) channels, which are contained in the en passant synaptic boutons of Schaffer collaterals. To determine if calcium flux through Ca_V channels contributes to Schaffer collateral hyper-excitability during HFS, we tested for effects of selective Ca_V 2.1 and 2.2 (ω-Agtx-IVA and ω-Ctx-GVIA, respectively) and a non-selective (cadmium) channel blockers.

Ca_V Channels Do Not Contribute to Hyper-excitability During HFS

We first investigated a role for Ca_V 2.1 and 2.2 channels using selective channel blockers. In separate recordings, we applied either ω-Agtx-IVA (Agtx, n = 7) or ω-Ctx-GVIA (Ctx, n = 6). Applying 200 nM ω-Agtx-IVA during LFS (summarized in Table 2) had no effect on fiber volley amplitudes (p = 0.4), latencies (p = 0.68), or half-widths (p = 0.08). 1 μM ω-Ctx-GVIA application during LFS (summarized in Table 2), did reduce fiber volley amplitudes (control vs. Ctx: 0.53 ± 0.05 mV vs. 0.5 ± 0.04 mV, p < 0.05), but had no effects on either latencies (p = 0.175) or half-widths (p = 0.91). No effect of ω-Agtx-IVA application or a significant interaction between ω-Agtx-IVA and solution was observed on amplitudes, latency changes, or half-width changes (amplitudes: solution- p = 0.22, interaction- p = 0.30; latency changes: solution- p = 0.495, interaction- p = 0.465; half-width changes: solution- p = 0.495,) during HFS (Fig. 5), although a significant effect of time was detected (amplitudes: F ratio- 366.1, p < 0.001; latency changes: F ratio- 7.154, p < 0.05; half-widths: F ratio- 8.895, p < 0.05). Similarly, no effect of ω-Ctx-GVIA application (Fig. 6), or a significant interaction between ω-Ctx-GVIA and time was detected fiber volley amplitudes, latency changes, or half-width changes (amplitudes: solution- p = 0.32, interaction- p = 0.23; latency changes: solution- p = 1.0, interaction- p = 0.71; half-width changes: solutions- p = 0.39, interaction- 0.955) during HFS.

Table 2.

Effects of CaV Blockers on Low-frequency Stimulation

Condition	[Ca2+]o	[Mg2+]o	Amplitude (mV)	Latency (ms)	Half-width (ms)
Control (n = 6)	2.0	2.0	0.53 ± 0.05	1.35 ± 0.05	0.60 ± 0.04
1 μM ω-Ctx-GVIA	2.0	2.0	0.50± 0.04 *	1.34 ± 0.05	0.60 ± 0.04
Control (n = 7)	2.0	2.0	0.72±0.12	1.50±0.16	0.63±0.05
200 nM ω-Agtx-IVA	2.0	2.0	0.70±0.11	1.50±0.16	0.65±0.06
Control (n = 4)	2.0	2.0	0.64 ± 0.37	2.02 ± 0.18	0.84 ± 0.10
Cd2+ (600 μM)	2.0	2.0	0.66 ± 0.33	1.94 ± 0.18	0.83 ± 0.08

Bold/italics indicates significant difference compared to control:

^*p<0.05

^**p<0.01.

Fig. 5. Selective Ca_V2.1 channel blocker did not alter the biphasic changes in Schaffer collateral excitability during HFS.

A) Example traces of fiber volleys recorded before (top) and after (bottom) application of 200 nM ω-Agtx-IVA (Agtx). Fiber volley responses evoked by the first (1) and sixth (6) stimuli are shown on the left; responses evoked by the first (1) and last (160) stimuli are shown at right. B-D) Application of ω-Agtx-IVA had no effect on fiber volley amplitude (B), latency (C), or half-width (D) changes during HFS. Shaded areas indicate ± 1 S.E.M.; n = 7 for control and ω-Agtx-GVIA.

Fig. 6. Selective Ca_V2.2 channel blocker did not alter the biphasic changes in Schaffer collateral excitability during HFS.

A) Example traces of fiber volleys recorded before (top) and after (bottom) application of 1 μM ω-Ctx-GVIA (Ctx). Fiber volley responses evoked by the first (1) and sixth (6) stimuli are shown on the left; responses evoked by the first (1) and last (160) stimuli are shown at right. B-D) Application of ω-Ctx-GVIA had no effect on fiber volley amplitude (B), latency (C), or half-width (D) changes during HFS. Shaded areas indicate ± 1 S.E.M.; n = 6 for control and ω-Ctx-GVIA.

Because the previous experiments investigated the roles of specific Ca_V channels, it is possible that one channel type could have compensated for loss of the other. To test this possibility, we investigated the effects of blocking Ca_V channels non-selectively with Cd²⁺ (600 μM; n = 4). During LFS, summarized in Table 2, Cd²⁺ application had no effect on Schaffer collateral fiber volley amplitudes (p = 0.83), latencies (p = 0.21) or half-widths (p = 0.79). Likewise, during HFS, Cd²⁺ application (Fig. 7) had no effect on fiber volley amplitudes (solution p = 0.90, interaction p = 0.41), latency changes (solution p = 0.61, interaction p = 0.62), or half-width changes (solution p = 0.97, interaction p = 0.25), although there were significant effects of time for amplitudes (p < 0.005) and latency changes (p < 0.05), but not half-width changes (p = 0.09). Together, our data suggests that the effects of altering extracellular divalent cation concentrations on Schaffer collateral excitability are not mediated by a Ca_V channels.

Fig. 7. Non-specific Ca_V channel blocker did not alter the biphasic changes in Schaffer collateral excitability during HFS.

A) Example traces of fiber volleys recorded before (top) and after (bottom) application of 600 μM Cd²⁺. Fiber volley responses evoked by the first (1) and sixth (6) stimuli are shown on the left; responses evoked by the first (1) and last (160) stimuli are shown at right. No effect of Cd²⁺ application was observed on fiber volley amplitude (B), latency (C), or half-width (D) changes. Shaded areas indicate ± 1 S.E.M.; n = 4 for control and Cd²⁺.

DISCUSSION

Manipulating extracellular divalent cation concentration had varying effects on Schaffer collateral excitability, depending on the frequency of stimulation (LFS or HFS) and specific cation species manipulated (calcium or magnesium), which are summarized in Table 3. We chose to alter divalent cation concentration beyond what would be expected under physiological conditions in order to maximize effects and make it easier to observe any changes that resulted from our manipulations. Supra-physiological manipulations are commonly done in electrophysiological experiments for this reason; for example, using non-physiologically elevated chloride concentration in internal pipette solution for whole-cell recordings in order to better observe inhibitory postsynaptic currents. As discussed in more detail below, effects of extracellular divalent cations on responses evoked during LFS can be explained by alterations in membrane surface charge. In contrast, effects on Schaffer collateral responses during HFS are not easily explained by a membrane surface charge mechanism.

Table 3:

Summary of Findings

Experiment	LFS			HFS (first 10 stimuli)			HFS (last 10 stimuli)
	Amplitude	Latency	Half-width	Amplitude	Δ Latency	Δ Half-width	Amplitude	Δ Latency	Δ Half-width
HiCal/LoMag	−	+	n.s.	+	n.s.	n.s.	+	n.s.	n.s.
LoCal/HiMag	n.s.	n.s.	n.s.	−	−	n.s.	+	n.s.	n.s.
HiCal	−	+	+	+	n.s.	n.s.	n.s.	n.s.	n.s.
HiMag	−	n.s.	n.s.	n.s.	−	n.s.	+	n.s.	n.s.
ω-Agtx-IVA	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.
ω-Ctx-GVIA	−	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.
Cadmium	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.

Effects on Low-frequency Stimulation

Schaffer collateral excitability during LFS was reduced by three of our four manipulations of extracellular divalent cations: (1) increasing extracellular calcium with a compensating reduction in extracellular magnesium, (2) increasing extracellular calcium while maintaining normal extracellular magnesium, and (3) increasing extracellular magnesium while maintaining normal extracellular calcium. Schaffer collateral excitability during LFS was not affected when extracellular calcium was reduced with a compensating increase in extracellular magnesium. We suggest that these effects during LFS can be explained by altered membrane surface charge. Increasing either calcium or magnesium alone, would have increased total extracellular divalent cation concentration, which should have reduced and slowed voltage-dependent sodium currents (Frankenhaeuser and Hodgkin 1957; Hanck and Sheets 1992). This increase in total divalent cations can explain the reduced excitability. Negative charges associated with membrane proteins or lipids can influence can influence ion channel kinetics and conductance by generating an electrostatic force nearby (Bennet et al. 1997, reviewed in Green and Andersen 1991). By increasing extracellular divalent cations, such as calcium and magnesium, the influence that negative electrostatic force would exert on ion channels would lessen, effectively screening that force and reducing ion conductance and shifting voltage-gated sodium (Na_V) channel kinetics to more depolarized membrane potentials (Bennet et al. 1997, Hanck and Sheets 1992, reviewed in Green and Andersen 1991). This combination of reduced ion conductance and a depolarized shift in Na_V channel activation would lead to the reduced excitability we observed. Assuming extracellular calcium is more effective at altering surface charge than magnesium (Hanck and Sheets 1992), the first manipulation, increasing calcium and decreasing magnesium concentrations in equal amounts, could have had a similar effect. For the same reasons, our fourth manipulation, reducing extracellular calcium with a compensating increase in extracellular magnesium, should have had the opposite effect (an increase) on Schaffer collateral excitability during LFS. While this manipulation had no significant effects on Schaffer collateral excitability, the trend was in the appropriate direction. A larger reduction in extracellular calcium or a decrease in total extracellular divalent cation concentration might have yielded a significant increase in excitability.

Effects on High-frequency Stimulation

Unlike during LFS, where increased extracellular calcium reduced excitability but decreased extracellular calcium did not change excitability, during HFS clearly opposite effects on hyper-excitability (increased fiber volley amplitude and decreased fiber volley latency) were observed by increasing or decreasing extracellular calcium. Moreover, during HFS, increasing extracellular calcium promoted hyper-excitability, whereas during LFS calcium reduced excitability. In addition, calcium and magnesium had different effects on the biphasic excitability change during HFS, with calcium – but not magnesium – affecting hyper-excitability. Increasing extracellular magnesium, in comparison, substantially reduced the depression of Schaffer collateral responses that occurs after the hyper-excitability period, similar to when extracellular calcium was increased and extracellular magnesium was decreased. This, however, occurred at a large concentration of magnesium (10 mM) as compared to calcium (3.8 mM), possibly because calcium has greater effect on surface charge than magnesium. In support of this possibility, increasing extracellular calcium, with magnesium reduced to maintain total divalent cation concentration, also significantly reduced depression, with the magnitude of effect being smaller than that of magnesium. While some of our data is consistent with a surface charge mechanism, other results are not. For instance, increasing extracellular calcium without a compensating decrease in extracellular magnesium, which should have increased surface charge screening, failed to significantly affect depression. Thus, effects of altering extracellular divalent cations on our LFS results can be explained by a surface change mechanism, but our HFS results cannot be explained completely by this mechanism.

Voltage-gated calcium channels are present in the en passant synaptic boutons that periodically interrupt the long Schaffer collateral fibers (Shepherd and Harris 1998, Shepherd et al. 2002). To test whether calcium flux through these channels contributes to excitability changes during HFS, we examined effects of two selective Ca_V blockers, ω-Agtx-IVA and ω-Ctx-GVIA, and a non-selective blocker, cadmium. None of the blockers altered the biphasic changes in Schaffer collateral responses during HFS, ruling out a contribution of calcium influx through Ca_V channels. It is possible that calcium influx through another method might contribute to the early hyper-excitability phase. Changes in extracellular calcium may have altered membrane excitability during HFS by affecting the sodium-calcium exchanger, calcium-dependent potassium channels, and/or calcium-dependent signaling (e.g. calcium-sensing receptor). Further experimentation will be needed to determine which of these potential mechanisms may contribute to the calcium sensitivity of excitability changes during HFS.

Potential Physiological Relevance

Extracellular calcium is known to decrease during periods of intense neuronal activity (Nicholson et al. 1978, Benninger et al. 1980, for review Heinemann et al. 1990), including during high frequency afferent stimulation. Because the early hyper-excitability during HFS is directly dependent on extracellular calcium, any change in extracellular calcium triggered by HFS might alter the hyper-excitability process. This would constitute a form of negative feedback, with HFS initiating hyper-excitability and also triggering a decrease in extracellular calcium concentration, which would then feedback to reduce hyper-excitability. With regards to long-term synaptic potentiation (LTP), the excitability and extracellular calcium changes that occur during HFS would tend to reduce the magnitude of LTP, which requires postsynaptic depolarization and calcium influx and is triggered by HFS at many central nervous system synapses. This would constitute a longer-term form of feedback acting to potentially reduce future long-lasting increases in excitation.