Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Aug 1.
Published in final edited form as: Neuroscience. 2019 May 30;412:72–82. doi: 10.1016/j.neuroscience.2019.05.042

Oxygen-glucose deprivation differentially affects neocortical pyramidal neurons and parvalbumin-positive interneurons

Nadya Povysheva a, Aparna Nigam a, Alyssa K Brisbin a, Jon W Johnson a,b, Germán Barrionuevo a,b
PMCID: PMC6818263  NIHMSID: NIHMS1532362  PMID: 31152933

Abstract

Stroke is a devastating brain disorder. The pathophysiology of stroke is associated with an impaired excitation-inhibition balance in the area that surrounds the infarct core after the insult, the peri-infarct zone. Here we exposed slices from adult mouse prefrontal cortex to oxygen-glucose deprivation and reoxygenation (OGD-RO) to study ischemia-induced changes in the activity of excitatory pyramidal neurons and inhibitory parvalbumin (PV)-positive interneurons. We found that during current-clamp recordings, PV-positive interneurons were more vulnerable to OGD-RO than pyramidal neurons as indicated by the lower percentage of recovery of PV-positive interneurons. However, neither the amplitude of OGD-induced depolarization observed in current-clamp mode nor the OGD-associated current observed in voltage-clamp mode differed between the two cell types. Large amplitude, presumably action-potential dependent, spontaneous postsynaptic inhibitory currents recorded from pyramidal neurons were less frequent after OGD-RO than in control condition. Disynaptic inhibitory postsynaptic currents (dIPSCs) in pyramidal neurons produced predominantly by PV-positive interneurons were reduced by OGD-RO. Following OGD-RO, dendrites of PV-positive interneurons exhibited more pathological beading than those of pyramidal neurons. Our data support the hypothesis that the differential vulnerability to ischemia-like conditions of excitatory and inhibitory neurons leads to the altered excitation-inhibition balance associated with stroke pathophysiology.

Introduction

Ischemic stroke is a devastating brain disorder and one of the leading causes of disability worldwide. Interruption of blood flow to the affected brain area results in disrupted or insufficient supply of oxygen and glucose, two components essential for normal neuronal functioning. This condition is known as oxygen and glucose deprivation (OGD). OGD leads to massive neuronal cell death in the directly affected area, the ischemic core, and pathological changes in the surrounding area, the peri-infarct zone. The peri-infarct zone is a high-risk brain area following stroke since pathological changes can give rise to neurodegeneration in the brain (Dirnagl, Iadecola and Moskowitz 1999). There are no known therapeutics that can combat neurodegeneration in neurons after an ischemic insult. To rescue neurons in the peri-infarct zone, it is critical to understand the pathological changes that excitatory and inhibitory neurons undergo during OGD and subsequent reoxygenation (RO).

One of the pathological changes due to OGD-RO involves circuitry effects associated with differential vulnerability of excitatory and inhibitory neurons to ischemia, and, as a result, changes in the E/I balance that can lead to disease conditions (Stief et al. 2007, Haider and McCormick 2009). Indeed, previous studies showed that GABAergic cortical neurons are more vulnerable to OGD-RO than excitatory glutamatergic neurons (Wang 2003, Zhang et al. 2009). Several studies showed a decrease in the number of GABA receptors, and of inhibitory neurons in the neocortical peri-infarct zone (Neumann-Haefelin et al. 1998, Qu et al. 1998, Schiene et al. 1996), but see (Hiu et al. 2016). Interestingly, stroke had no effect on the density of the glutamatergic synapses assessed in various neocortical layers at one week or one month post-stroke (Hiu et al. 2016). Similarly, in a photothrombotic model of focal stroke, no changes were detected during phasic excitation of neurons over a 2-week period after stroke in the neocortical peri-infarct zone (Clarkson et al. 2010).

Inhibitory neocortical interneurons comprise a very diverse group of cells (Petilla Interneuron Nomenclature et al. 2008). PV-positive interneurons, a type of inhibitory neuron, are known to provide strong inhibitory control of pyramidal neuron activity (Wilson et al. 2012), and, thus, their vulnerability to ischemic insult can have a pronounced impact on the E/I balance in cortical circuitry. Although in general GABAergic neurons are vulnerable to ischemia, existing data about differential resistance of parvalbumin (PV)-positive interneurons to ischemic insults are somewhat contradictory. For example, PV-positive interneurons in the hippocampus are less vulnerable to ischemia than hippocampal CA1 pyramidal neurons according to morphological studies in gerbil (Nitsch et al. 1989) and rat (Papp et al. 2008). In contrast, in the neocortex, signaling through PV-positive interneurons was shown to be more susceptible to ischemia-associated changes than through glutamatergic neurons, since GABAergic synaptic network activity was substantially suppressed as a result of ischemic depolarization (Xie et al. 2014). A reduction in the number of PV-positive neurons as well as the number of their dendrites was reported in the neocortical peri-infarct zone (Neumann-Haefelin et al. 1998). Thus, the differential vulnerability of PV-positive interneurons and excitatory neurons to ischemia remains an unresolved issue.

In the present study we addressed the differential vulnerability of neocortical PV-positive interneurons and pyramidal neurons during OGD-RO. We report that inhibitory PV-positive interneurons are more prone to cell death than pyramidal neurons during OGD-RO. We found that the loss of PV-positive interneurons during OGD-RO led to a reduction of action potential dependent IPSCs in pyramidal neurons, as well as disynaptic inhibitory response.

Materials and Methods

Experiments were performed on prefrontal cortex (PFC) slices from 7–10-month-old transgenic mice (n=25) that express eGFP in PV-positive interneurons (CB6/Tg(Gad1-EGFP)G42jh/J mice (Chattopadhyaya et al. 2004)). All animal procedures were conducted in accordance with the guidelines outlined in the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and approved by the University of Pittsburgh Institutional Animal Care and Use Committee. Mice were deeply anesthetized with chloral hydrate and decapitated. The brain was quickly removed and immersed in ice-cold pre-oxygenated artificial cerebrospinal fluid (ACSF). A tissue block containing the prelimbic cortex was excised for slicing. Coronal slices (350–400 μm thick) were cut with a vibratome (Leica VT1000S, Leica, Germany). Slices were incubated at 37°C for 0.5–1 h and further stored at room temperature until they were transferred to a recording chamber perfused with ACSF at 31–32°C. In control condition, ACSF of the following composition was used (in mM): 126 NaCl, 2.5 KCl, 1.25 NaH2PO4, 1 MgSO4, 2 CaCl2, 24 NaHCO3, 10–20 glucose; pH 7.25–7.3. ACSF was bubbled with a 95% O2/5% CO2 gas mixture.

Electrophysiological recordings

Whole-cell recordings were performed from layer 2–3 neurons visualized by IR-DIC video microscopy using a Zeiss Axioskop microscope (Carl Zeiss, Inc., Thornwood, NY) with a 40x water immersion objective and a digital video camera (CoolSnap, Photometrics, Tucson, AZ). Pyramidal neurons were identified by their apical dendrites and triangular somata. PV-positive interneurons were recognized based on their green color in epifluorescent light. For the current-clamp experiments, patch electrodes were filled with an internal solution containing (in mM): 120 K-gluconate, 2 MgCl2, 10 NaCl, 10 HEPES, 0.2 EGTA, 14 phosphocreatine, 4 ATP-Mg, 0.3 GTP-Na; pH 7.25. For the voltage-clamp experiments, patch electrodes were filled with an internal solution containing (in mM): 105 Cs-gluconate, 2 MgCl2, 10 NaCl, 10 HEPES, 10 phosphocreatine, 4 ATP-Mg, 0.3 GTP, and 10 BAPTA; pH 7.25. Alexa hydrazide 568 (0.075%; Molecular Probes, Eugene, OR) was added to the intracellular solution during current-clamp recordings for later morphological identification of the recorded neurons. Electrodes had 5–10 MΩ open-tip resistance. Voltage and current-clamp recordings were performed with a MultiClamp 700A amplifier (Axon Instruments, Union City, CA). Current-clamp recordings were performed in bridge-balance mode. Signals were filtered at 2 kHz and acquired at a sampling rate of 10 kHz using a Digidata 1440 digitizer and Clampex l0.2 software (Molecular Devices Corporation, Sunnyvale, CA). Access resistance and capacitance were compensated on-line. Access resistance typically was 10–20 MΩ and remained relatively stable during experiments (≤ 30% increase) for the cells included in the analysis. Membrane potential was corrected for the liquid junction potential of −13 mV. The following pharmacological agents were used in this study: gabazine (10 μM; Ascent Scientific LTD, Bristol, UK) to inhibit GABAA receptors; 2,3-dihydroxy-6-nitro-7-sulfamoylbenzo(F)quinoxaline (NBQX; 20 μM; Ascent Scientific) to inhibit α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors (AMPARs) and kainate receptors; tetrodotoxin (TTX; 0.5 µM; Sigma) to inhibit voltage-gated Na+ channels.

To mimic ischemic conditions in the brain, PV-positive and pyramidal neurons were subjected to OGD-RO. First, the slices were perfused with an extracellular solution containing sucrose (10 mM) substituted for glucose and bubbled with a mixture of 95% N2/5% CO2 (OGD solution) for 10 min. It was critical to keep OGD exposure under 15 min time window was identified as the maximum survival window of pyramidal neurons (Zhang et al. 2008). After that, the extracellular solution was switched to oxygenated ACSF for 15 min. Recording were made either in current-clamp or in voltage-clamp mode.

Spontaneous inhibitory post-synaptic currents (sIPSCs) were recorded for 3 min in voltage-clamp mode at a holding potential of +12 mV from two separate groups of pyramidal neurons: 1) in control conditions (ACSF with oxygen and glucose); and 2) after exposure to the OGD for 10 min and RO for 15 min. Neurons in the second group were patched after the OGD-RO exposure. Spontaneous excitatory post-synaptic currents (sEPSCs) were recorded for 3 min in voltage-clamp mode at a holding potential of −70 mV from two experimental groups of pyramidal neurons similar to those described for sIPSCs.

To evoke disynaptic inhibitory postsynaptic currents (dIPSC), a bipolar stimulating electrode was placed on the border of white matter and layer 6 (Povysheva et al. 2006). dIPSCs were evoked at 0.1 Hz. A pyramidal neuron was voltage clamped at a holding potential of +12 mV and dIPSCs were recorded for 3 min in control conditions. Slices then were exposed to OGD-RO, and dIPSCs were recorded from the same pyramidal neuron for an additional 3 min.

Electrophysiological data analysis

Membrane properties of neurons were analyzed using the Clampfit 10.2 software package (Molecular Devices Corporation, Sunnyvale, CA). To characterize the membrane properties of neurons, hyper- and depolarizing current steps were applied for 500 ms in 5–10 pA increments at 0.5 Hz while recording membrane potential. Input resistance was measured from the slope of a linear regression fit to the voltage-current relation in a voltage range hyperpolarized from resting potential. The membrane time constant was determined by single-exponential fits to the average voltage relaxation activated by the hyperpolarizing current steps (5–15 pA). A series of depolarizing current steps of gradually increasing amplitude were used to evoke action potentials. Action potential properties were quantified using the first evoked action potential, defined as a regenerative depolarization with maximum slope of at least 10 mV/ms. Peak amplitudes of the action potential and the afterhyperpolarization were measured relative to the action potential threshold. Duration of the action potential was measured at its half amplitude. Firing frequency was calculated in Hz as the ratio between number of action potentials and current step duration.

The OGD-induced depolarization was measured as the difference between the resting membrane potential before OGD and the most positive membrane potential observed during OGD-RO. Latency of the OGD-RO depolarization was measured from the beginning of application of OGD solution until membrane potential depolarized more quickly than 1 mV/s. The OGD-RO current amplitude was measured as the difference between the baseline current (holding current before exposure to OGD-RO) and the maximum negative peak current during OGD-RO for the cells that recovered from the OGD-RO exposure. Latency of the OGD-RO current was measured from the beginning of application of OGD solution until membrane current changed more quickly than −1 pA/s. Spontaneous events were analyzed using the MiniAnalysis Program (Synaptosoft, Decatur, GA). Peak events were first detected automatically using an amplitude threshold of 1.5 times the average RMS noise. RMS noise was around 3 pA for recordings at −70 mV holding potential and 7 pA for those at +12 mV. More than 500 events in each cell were included in the analysis.

Amplitude of evoked dIPSCs was measured on averaged traces as the most positive current value compared to baseline current using Clampfit. Latency was measured between the stimulation artifact and the beginning of the outward synaptic current.

Morphological data analysis

During current-clamp recordings, neurons were filled with the fluorescent dye Alexa 568, which was added to the recording pipette solution as previously described (Povysheva et al. 2006). Whole-cell recordings were maintained for at least 30 min to ensure extensive cell labeling by the dye. Slices were fixed in ice-cold 4% paraformaldehyde for at least 72 h, then transferred into an anti-freeze solution (ethylene glycol and glycerol in 0.1 M phosphate buffer) and stored in the freezer. Neurons were reconstructed three-dimensionally using an Olympus Fluoview BX61 confocal microscope (Olympus America Inc, Melville, NY) with FITC and CY3 filters. Images were acquired with Fluoview software (Olympus America Inc, Melville, NY).

To quantitatively evaluate dendritic beading, which is considered a hallmark of the OGD pathology (Kislin et al. 2017, Brisson and Andrew 2012), the dendritic width was measured using the ImageJ 1.x software (Schneider, Rasband and Eliceiri 2012) similar to the approach in a previous study (Weilinger et al. 2016). To sample segments of the dendritic tree in both cell types similarly, a grid composed of 80 × 80 μm squares was surrounding the cell body were analyzed. Dendritic widths were estimated by placing a line perpendicular to the dendritic axis in three equally spaced locations on each dendritic segment within each square. 14–16 dendritic segments per neurons were analyzed. To obtain distribution frequencies of the dendritic width, histograms were plotted by binning the individual data into intervals of ~0.3 µm. The histograms were then fit with either a Gaussian or the sum of two Gaussian equations with the least-squares fitting method. A single Gaussian function was used to fit a histogram if the R2 value (a measure of goodness of fit) was >0.95; otherwise, the sum of two Gaussians was used for the fit.

Statistical analysis

For electrophysiological experiments, sample sizes were determined based on our previous studies (Povysheva and Johnson 2016). Electrophysiological data were collected from one neuron per brain slice. The Shapiro-Wilk test was used to check for normal distribution of data sets and variances were compared using the F-test. Normally distributed experimental groups were compared using two-tailed t-tests (assuming unequal variances) or paired t-test for means. The Fisher exact test was used to compare percentage of cells that recovered from OGD-RO.

For morphological experiments, a fixed-effect model was used to analyze data. Data were collected from at least 3 animals/group, 2–3 slices per animal. Sixty-nine dendrites from 4 pyramidal neurons (3 animals) in control condition, and 70 dendrites from 5 pyramidal neurons (4 animals) exposed to OGD-RO. Fifty-six dendrites (3 animals) from 4 neurons in control condition, and fifty-eight dendrites (4 animals) from 5 neurons in OGD-RO condition. Morphological changes were assessed with regression analysis and two-way analysis of variance using Tukey’s post-hoc test. Values are reported as mean ± SEM. For parametric and non-parametric tests, p-values <0.05 (t values >1.96 and <−1.96) and 95% confidence intervals excluding zero were considered statistically significant. Statistical tests were performed using Excel (Microsoft Corp., Redmond, WA), Origin (Origin Lab Corporation, Northampton, MA, USA) and GraphPad Prism 7 (GraphPad Software, Inc., San Diego, CA).

Results

Identification of pyramidal neurons and PV-positive interneurons

All experiments for this study were performed in adult animals (7–10 months) to model pathological changes that take place during stroke in human adults. Pyramidal neurons in PFC slices were identified under IR-DIC based on their distinct morphological features: large triangular cell body, partially visible, pronounced apical dendrite. In addition, their cellular identity was confirmed based on their intrinsic membrane properties. All pyramidal neurons studied exhibited typical physiological phenotypes including large action potential amplitudes and strong adaptation (Fig. 1A, left). The identity of some pyramidal neurons was further verified by confocal reconstruction (Fig. 1A, right). PV-positive interneurons were identified based on expression of eGFP in CB6/Tg(Gad1-EGFP)G42jh/J mice (Fig. 1B, right). All cells that expressed eGFP demonstrated a typical fast-spiking phenotype characterized by short duration of action potential, firing without adaptation, high firing frequency, and fast time constant (Fig. 1B, left) (Povysheva et al. 2008).

Figure 1. Identification of pyramidal neurons and PV-positive interneurons.

Figure 1

Open in a new tab

A. Identification of pyramidal neurons. Left, responses produced in pyramidal neurons by depolarizing and hyperpolarizing current pulses in ACSF; right, pyramidal neuron filled with fluorescent dye during recording and later reconstructed with a confocal microscope (scale bar 50 μm). B. Left, responses produced in PV-positive interneurons by depolarizing and hyperpolarizing current pulses recorded in ACSF. Right, image of PV-positive interneurons expressing eGFP (scale bar 50 μm).

The effects of OGD-RO on pyramidal neurons and PV-positive interneurons

To mimic ischemic conditions in the brain, PFC slices were bathed in extracellular solution that lacked oxygen and glucose, a condition known as OGD, followed by reoxygenation (OGD-RO). In current-clamp experiments, the resting membrane potential first was recorded for 5–10 min while the PFC slice was perfused with ACSF. The slice then was perfused with an extracellular solution that contained sucrose (10 mM) substituted for glucose and was bubbled with a mixture of 95% N2/5% CO2 (OGD solution). The OGD solution was applied for 10 min (time interval know to produce substantial physiological and morphological neuronal responses (Liu et al. 2005)), and then the extracellular solution was switched back to ACSF for 15 min to explore the tolerance of cells to OGD and subsequent RO. During both OGD and RO the resting membrane potential was continuously recorded. As a result of the OGD-RO exposure, the membrane potential exhibited a very strong depolarizing shift. Resting potential recovered to its pre OGD-RO value for some cells (Fig. 2A, left), but not for others (Fig. 2A, right). Cells that recovered from the OGD-RO sequence exhibited the following characteristics within 15 min of the start of reoxygenation: 1) return of membrane potential to a value similar to that before OGD-RO solution exposure (± 10 mV); 2) return of spontaneous postsynaptic potentials; and 3) generation of trains of action potentials in response to depolarizing current pulses (Fig. 2B). Resistance of pyramidal neurons and PV-positive interneurons to OGD-RO was quantified as the percent of cells that recovered following exposure to OGD-RO. Only 14% of PV-positive interneurons (1 out of 7) recovered following OGD-RO, whereas 80% of pyramidal neurons recovered (6 out of 8) (Fisher exact test (7, 8), p=0.04; odds ratio=0.056, 95% confidence interval=0.004 to 0.760) (Fig. 2C).

Figure 2. OGD-RO effects in current-clamp mode.

Figure 2

Open in a new tab

A. OGD-RO sequence produced strong depolarization of pyramidal neurons and PV-positive interneurons. B. Recordings following OGD-RO demonstrating two of the criteria used to assess cell recovery. Above, ability to produce action potential trains in response to depolarizing current pulses; below, observation of spontaneous postsynaptic currents. C. Fisher exact test revealed that cell recovery was significantly higher for pyramidal neurons (6 out of 8) than for PV-positive interneurons (1 out of 7). D. Latency of the OGD-induced depolarization was similar in pyramidal neurons (n=8) and PV-positive interneurons (n=7). E. Depolarization amplitude produced by OGD was similar in pyramidal neurons (n=8) and PV-positive interneurons (n=7).

We measured the OGD-induced depolarization as the difference between the resting potential before OGD and the most positive membrane potential during OGD-RO. The OGD-induced depolarization occurred at comparable latencies in the both cell types (t (12) = 2.18, p=0.9) (Fig. 2D), and its amplitude did not differ between pyramidal neurons and PV-positive interneurons depolarization (63 ± 9 mV in pyramidal neurons and 63 ± 11 mV in PV-positive interneurons; t (12) = 2.18, p=0.9) (Fig. 2E). Interestingly, despite the similar magnitude of depolarization produced by the OGD-RO sequence in the two cell types, PV neurons were more vulnerable to OGD-RO as indicated by their percentage recovery (Fig. 2C).

Next we assessed in voltage-clamp mode the effects of OGD-RO on membrane currents of pyramidal neurons and PV-positive interneurons (Lauderdale et al. 2015, Thompson 2014). Neurons were held at −70 mV before, during OGD and during reoxygenation. All neurons exhibited an inward holding current during OGD-RO. The current often peaked during exposure to the OGD solution and again during RO (Fig. 3A). Both pyramidal neurons and PV-positive interneurons developed a substantial inward current during OGD with comparable latencies (pyramidal neurons: 6.5 ± 1.8 min; PV-positive interneurons: 7.3 ± 1.8 min, t (14) = 2.18, p=0.42; Fig. 3B). Recovery from OGD-RO (defined as return of holding current to within 10 pA of control holding current) was similar for pyramidal neurons (6 out of 7) and PV-positive neurons (7 out of 11; (Fisher exact test (7, 11), p=0.6; odds ratio=3.43, 95% confidence interval=0.31 to 48); Fisher exact test). The OGD-RO current amplitude of cells that recovered from OGD-RO did not differ between pyramidal neurons and PV-positive interneurons (t (14) = 2.14, p=0.94) (Fig. 3C). Thus, differences in the OGD-RO current amplitude in pyramidal neurons and PV-positive interneurons cannot explain their differential vulnerability to OGD-RO exposure.

Figure 3. OGD-RO effects in voltage-clamp mode.

Figure 3

Open in a new tab

A. Changes in holding current produced by the OGD-RO sequence in pyramidal neurons and PV-positive interneurons. Neurons were held at −70 mV. B. Latency of the OGD-RO-induced inward current was similar in pyramidal neurons (n=7) and PV-positive interneurons (n=11). C. OGD current of cells that recovered from OGD-RO was not different between pyramidal neurons (n=6) and PV-positive interneurons (n=7).

The effects of OGD-RO on inhibitory inputs to pyramidal neurons

Differential vulnerability of pyramidal neurons and PV-positive interneurons to OGD-RO would be expected to result in a shift of the E/I balance in cortical circuitry. The E/I balance can be assessed through measurement of excitatory and inhibitory inputs to pyramidal neurons (Povysheva and Johnson 2016). Greater vulnerability of PV-positive interneurons to OGD-RO would result in decreased inhibitory drive to pyramidal neurons. Thus, we next assessed the OGD-RO effects on inhibitory input to pyramidal neurons (see Methods). To record sIPSCs in isolation, pyramidal neurons were held at +12 mV (Cossart et al. 2001), a membrane potential close to the reversal potential of glutamate receptor-mediated current. Addition of gabazine abolished all visible postsynaptic currents (Fig. 4A), suggesting that glutamatergic responses were undetectable at +12 mV. Since OGD-RO resulted in the death of PV-positive interneurons in PFC slices, inhibitory inputs to the pyramidal neurons are expected to be reduced. Thus, we assessed the amplitude and frequency of the sIPSCs in pyramidal neurons under control conditions, and after the OGD-RO insult. Surprisingly, the amplitude (t (15) = 2.1, p=0.44) and frequency (t (16) = 2.1, p=0.5) of the sIPSCs in pyramidal neurons were comparable in both conditions (Fig. 4B,C).

Figure 4. Effects of OGD-RO on inhibitory inputs to pyramidal neurons.

Figure 4

Open in a new tab

A. sIPSCs recorded from pyramidal neurons at a holding potential (Vhold) of +12 mV. 10 μM gabazine eliminated spontaneous events, indicating that sEPSCs were not detectable at +12 mV. B,C. Frequency (B) and amplitude (C) of sIPSCs were not different between pyramidal neurons exposed to OGD-RO (n=10) and those in control conditions (n=9) (solid circles, control conditions; empty circles, OGD-RO). D,E. Large amplitude sIPSCs were abolished by TTX (1 μM) (representative examples; black, control; gray, TTX). F. Large amplitude sIPSC frequency was significantly lower in the pyramidal neurons that were exposed to OGD-RO than in those recorded under control conditions. G. A series of 8 dIPSCs recorded in a pyramidal neuron are superimposed. H. dIPSCs (average of 15 recordings) in a pyramidal neuron in control conditions (upper trace), a pyramidal neuron after exposure to OGD-RO (middle trace), and a pyramidal neuron exposed to OGD-RO conditions after application of 20 μM NBQX + 50 μM AP-5 (lower trace). I. dIPSC amplitude was significantly lower in pyramidal neurons exposed to OGD-RO than in those recorded under control conditions (n=5). Arrows in G and H indicate time of stimulation; stimulus artifacts were blanked.

Importantly, sIPSCs are not a homogeneous population, but consist of larger action potential-dependent IPSCs, and smaller action potential-independent (miniature) IPSCs. Since OGD-RO resulted in the death of 86% of PV-positive interneurons in PFC slices, large-amplitude sIPSCs associated with firing of PV-positive neurons are expected to be reduced. We thus analyzed in isolation large, action-potential dependent sIPSCs. We defined as “large” all sIPSCs with amplitude at least two standard deviations above the average sIPSC amplitude. Large sIPSCs were reliably abolished by TTX in 7 pyramidal neurons (Fig. 4D, E), supporting their dependence on presynaptic action potentials. We found that the frequency of the large-amplitude, presumably action potential dependent sIPSCs, was decreased after OGD-RO (t (10) = 2.2, p=0.03) (Fig. 4F), indicating that OGD-RO reduced inhibitory drive to pyramidal neurons.

PV-positive interneurons produce feed-forward disynaptic inhibition of pyramidal neurons providing temporal fidelity of the outputs of the latter (Cruikshank et al. 2010, Povysheva et al. 2006, Marek et al. 2018). The effect of OGD-RO on PV-positive interneurons should consequently reduce disynaptic inhibitory responses of PFC pyramidal neurons. Disynaptic inhibitory postsynaptic currents (dIPSCs) were evoked in pyramidal neurons by extracellular stimulation at a holding voltage of +12 mV. In accord with the disynaptic nature of the response, its latency was 4.0 ± 0.2 ms (Fig.4G) (Povysheva and Johnson 2016, Povysheva et al. 2006). Importantly, dIPSCs were abolished by addition of AMPAR and NMDAR antagonists, excluding the possibility of monosynaptic IPSCs being present (Fig. 4H). Consistent with the deleterious effect of the OGD on PV-positive interneurons, OGD-RO decreased the amplitude of dIPSCs (control, 79 ± 17 pA; OGD-RO, 45 ± 13 pA; n=5, t (4) = 2.8, p<0.05) (Fig. 4H, I), thus providing further evidence of differential vulnerability of PV-positive interneurons.

The effects of OGD-RO on excitatory inputs to pyramidal neurons

To complete analysis of the effects of OGD-RO on the E/I balance, we assessed spontaneous excitatory postsynaptic currents (sEPSCs). sEPSCs were isolated by holding pyramidal neurons at −70 mV, a membrane potential at which sEPSCs were clearly observable, and GABAA receptor-mediated currents were undetectable because −70 mV is near ECl (−69 mV). Indeed, inhibition of AMPA receptors with NBQX eliminated all observable postsynaptic currents at −70 mV (Fig. 5A). OGD-RO did not affect either amplitude (t (14) = 2.2, p=0.36) or frequency (t (17) = 2.1, p=0.6) of sEPSCs (Fig. 5B,C). OGD-RO also did not affect the frequency of large amplitude, presumably action potential dependent sEPSCs (t (10) = 2.1, p=0.9) (Fig. 5F). As for large sIPSCs, large sEPSCs were defined as sEPSCs with amplitudes of at least 2 standard deviations above the average sEPSC amplitude. Large sEPSCs were reliably abolished by TTX in 6 pyramidal neurons (Fig. 5D,E). The lack of an effect of OGD-RO on sEPSCs is consistent with the high percentage of pyramidal neurons that recovered from exposure to OGD-RO (Fig. 2C).

Figure 5. OGD-RO does not affect excitatory inputs to pyramidal neurons.

Figure 5

Open in a new tab

A. sEPSCs recorded in pyramidal neurons at −70 mV. 20 μM NBQX eliminated spontaneous events, indicating that sIPSCs were not visible at −70 mV. Frequency (B) and amplitude (C) of sEPSCs were not different between the pyramidal neurons that were exposed to OGD-RO (n=9) and those recorded in control conditions (n=10; solid circles, control conditions; empty circles, OGD-RO). D,E. Large amplitude sIPSCs were abolished by TTX (1 μM) (n=6; representative examples; black, control; gray, TTX). F. The frequency of large sEPSCs was not different in pyramidal neurons in control conditions and after OGD-RO.

Effects of OGD-RO on the morphology of pyramidal neurons and PV-positive interneurons

OGD-RO conditions can affect neuronal morphology (Brown et al. 2009, Xie et al. 2014). One of the early effects of the lack of oxygen and glucose is dendritic beading (formation of local swellings along dendrites). To quantify dendritic beading in control conditions and after OGD-RO exposure, we measured dendritic width of pyramidal neurons and PV-positive interneurons (see Methods) using an approach similar to one used in a previous study (Weilinger et al. 2016). In the control and OGD groups, neurons were labeled with fluorescent dye during whole-cell recording in ACSF. Neurons in the OGD group were exposed to the OGD-RO sequence during whole-cell recording (Fig. 6A,B).

Figure 6. OGD-RO effects on the morphology of pyramidal neurons and PV-positive interneurons.

Figure 6

Open in a new tab

A. Morphology of pyramidal neurons in control conditions (a) and after exposure to OGD-RO (b) (scale bar 20 μm). Dendrites of pyramidal neurons in control conditions (c) and after exposure to OGD-RO (d) (scale bar 4 μm). B. Morphology of PV-positive interneurons in control conditions (a) and after exposure to OGD-RO (b) (scale bar 20 μm). Dendrites of PV-positive interneurons in control conditions (c) and after exposure to OGD-RO (d) (scale bar 4 μm). C. Distributions of dendritic width of pyramidal neurons in control conditions (a) and after the OGD-RO exposure (b). D. Distributions of dendritic width of PV-positive interneurons in control conditions (a) and after the OGD-RO exposure (b).

Sixty-nine dendrites from 4 pyramidal neurons in control condition and 70 dendrites from 5 pyramidal neurons exposed to OGD-RO were included in the analysis. In PV-positive interneurons, we measured 56 dendrites from 4 neurons in control conditions and 58 dendrites from 5 neurons exposed to OGD-RO.

Non-linear regression curve fitting of histograms revealed that only PV-positive interneurons exposed to OGD-RO showed two populations of dendritic widths centered at 1.67 ± 0.34 μm and 2.63 ± 0.41 μm (Fig. 6Db). This effect is starkly different from control PV-positive interneurons, which only have one population of dendritic width centered at 1.71 ± 0.41 μm (Fig. 6Da). The emergence of a rightward shifted dendritic width population indicates that OGD-RO exposed PV-positive neurons are prone to dendritic beading. There was no difference between the dendritic widths of control pyramidal neurons (2.07 ± 0.56 μm) and OGD-RO exposed pyramidal neurons (2.31 ± 0.63 μm; p>0.1), both of which showed only one population of dendritic widths (Fig.6Ca,b). We found that dendritic width of PV-positive interneurons differed significantly in OGD-RO versus control condition (mean difference = −0.44; 95% confidence interval = −0.69 to −0.19; p <0.0001). In contrast, dendritic width of pyramidal neurons did not differ significantly in OGD-RO versus control condition (mean difference = −0.22; 95% confidence interval = −0.45 to 0.008; p=0.06).

The observed alterations in dendritic morphology provide further evidence of the higher vulnerability of PV-positive interneurons to OGD-RO (Fig. 2C).

Discussion

Acute and long-term changes in excitatory and inhibitory neurons produced by ischemia

Ischemic insults in multiple in vitro and in vivo animal models result in severe cellular changes, including anoxic depolarization, and, ultimately, cell death (Allen, Karadottir and Attwell 2005, Thompson 2014). In this study we assessed both depolarization and cell death produced by OGD-RO in excitatory pyramidal neurons and inhibitory PV-positive interneurons. We showed that PV-positive interneurons are more vulnerable to OGD-RO than pyramidal neurons based on the lower survival ratio of PV-positive interneurons. As a result, inhibition produced by PV-positive interneurons is reduced in pyramidal neurons following OGD-RO, leading to changed excitation/inhibition ratio in PFC circuitry. Interestingly, amplitude and latency of depolarizing responses to OGD-RO were not different in the two cell types. This finding, however, cannot be extrapolated to intact neurons because intracellular ion concentrations were kept constant during our recordings at the concentrations in the recording pipette solution (Du et al. 2018).

Multiple studies indicate that the E/I balance is affected by ischemic insults. For example, changes in excitatory and inhibitory circuits of the rat hippocampus 12–14 months after complete forebrain ischemia resulted in hyperexcitability (Arabadzisz and Freund 1999). Similarly, cerebral phototrombosis and resulting ischemia led to hyperexcitability in the somatosensory cortex (Schiene et al. 1996). It was shown that acute ischemic brain injury initiates early hyperexcitability and hypersynchronous discharges of large populations of excitatory neurons that could spread through the brain (Lei et al. 2016). Many studies indicate that ischemia appears to modify the E/I balance by producing a deficit in inhibitory, but not in excitatory circuitry. Thus, in the photothrombotic model of focal stroke, the mean phasic excitation in the neocortex remained unchanged over the 2-week period after stroke (Clarkson et al. 2010). Another study also showed no observable effects on the density of glutamatergic synapse subclasses investigated in layers 5 and 2/3 at 1 week or 1 month post-stroke (Hiu et al. 2016). The study utilizing the OGD model of ischemic insult showed that GABAergic cortical neurons are more vulnerable to OGD-RO than excitatory glutamatergic neurons (Wang 2003). OGD decreased sIPSC frequency in pyramidal neurons from rat cortical slices, indicating that short-term cerebral ischemia causes dysfunction of interneurons resulting in less inhibition of pyramidal neurons from rat cortical slices (Wang 2003). In rat hippocampus, interneurons in CA1 stratum radiatum appeared to be vulnerable to the OGD insult than pyramidal neurons (Zhang et al. 2009). Substantial loss of inhibitory neurons in the hippocampus was detected after forebrain ischemia in rats (Arabadzisz and Freund 1999). In GABAergic medium spiny neurons in the striatum, ischemia diminished the frequency and amplitude of sIPSCs, and enhanced the frequency, decay time and half-width of sEPSCs compared to control animals (Bacigaluppi et al. 2016). Several studies showed decreases in the number of GABA receptors in the neocortical peri-infarct zone (Neumann-Haefelin et al. 1998, Qu et al. 1998, Schiene et al. 1996). Increase in BDNF-TrkB signaling produced by ischemia can lead to degradation of the K+-Cl− co-transporter, and as a result, a change in GABAergic signaling from hyperpolarizing to depolarizing (Kang, Johnston and Kadam 2015).

In accord with those studies, here we show that OGD-RO induces a decrease in large action potential-dependent sIPSCs recorded in pyramidal neurons. Consistent with the decrease in large sIPSCs, we found that PV-positive interneurons are more vulnerable to OGD-RO than excitatory pyramidal neurons as assessed by their cell death and morphological changes. In contrast, a recent study (Hiu et al. 2016) demonstrated an increase in the number of GABA receptors as well as sIPSCs one week after stroke lesion in the mouse neocortex.

Role of PV-positive interneurons in pathophysiology of ischemia

Although existing evidence points to alterations in inhibitory neuron function, the type of inhibitory neurons affected by ischemia is not clear. Here we demonstrate that PV-positive interneurons are more vulnerable to OGD-RO than pyramidal neurons.

Since PV-positive interneurons provide perisomatic inhibition to pyramidal neurons, their differential vulnerability to ischemic insult could have a pronounced impact on the E/I balance in cortical circuitry. In the somatosensory cortex, optogenetic stimulation of PV-positive terminals resulted in GABAergic synaptic network activity that was substantially suppressed as a result of ischemic depolarization during global ischemia in mice (Xie et al. 2014). In the neocortical peri-infarct zone a reduction of the number of PV-positive neurons was reported as well as a reduction of the number of their dendrites (Neumann-Haefelin et al. 1998). However, it was shown that PV-immunoreactive interneurons in gerbil hippocampus are resistant to the effects of ischemia, as they survive up to 28 days after the insult. It was concluded that the presence of the Ca2+-binding protein parvalbumin protects the GABAergic neurons from the deleterious consequences of ischemia-induced Ca2+-influx (Nitsch et al. 1989). Consistent with this finding, parvalbumin was shown to have neuroprotective properties in ischemia (Van Den Bosch et al. 2002). However, our data demonstrate higher vulnerability of PV-positive interneurons to OGD-RO than pyramidal neurons. Higher vulnerability of PV-positive interneurons than pyramidal neurons to OGD-RO may be explained by their differential properties. Specifically, PV-positive interneurons, unlike pyramidal neurons, express Ca2+-permeable AMPA receptors (McBain and Fisahn 2001) that, when activated during OGD-RO, could be responsible for death-inducing Ca2+ elevations. Another potential explanation for the high vulnerability of cortical PV-positive interneurons is the large number of strong excitatory inputs they receive (Povysheva et al. 2006, Povysheva et al. 2008). These inputs can be overactivated by elevation of glutamate during OGD-RO, leading to cell death. In addition, failure of the Na+-K+ pump is an important contributor to the development of anoxic depolarization in neurons and eventually to cell death (Thompson 2014). The different cell size and surface to volume ratios of pyramidal neurons and PV-positive interneurons may lead to distinct responses to pump failure, and thus to the differential OGD-RO sensitivity.

Morphological changes produced by OGD-RO conditions

Our study indicates that OGD-RO-associated morphological alterations are cell-type specific. PV-positive interneurons, which are more prone to cell death in OGD-RO conditions, also undergo more substantial morphological changes. Specifically, we showed that PV-positive interneuron morphology was more affected by OGD-RO than pyramidal neuron morphology. PV-positive interneurons exhibited substantial beading in virtually all dendritic branches after OGD-RO that is reflected by variations of dendritic thickness (Fig. 6). In contrast, pyramidal neurons exhibited a substantial number of non-beaded dendrites after OGD-RO. Alterations in morphology are consistent with the higher vulnerability of PV-positive interneurons to the OGD-RO insult.

Dendritic beading and swellings in neurons were reported in multiple studies (Risher et al. 2010, Brisson and Andrew 2012). For example, very pronounced dendritic beading was reported in the proximal dendrites of medium spiny neurons in brain slices exposed to OGD (Brisson and Andrew 2012). Severe dendritic beading also was reported in the somatosensory cortex after ischemic insult in vivo (Risher et al. 2010). Other morphological changes reported after ischemia in the striatum and motor cortex of mice include decreased dendrite spine density and reduced dendritic growth (Luo et al. 2014). Several studies suggest that morphological changes produced by OGD-RO are transient, and normal morphology could be restored during RO. Rapidly reversible damage to dendrites and spines were reported in vivo, where the majority of spines and dendrites regained their original structure during RO (Murphy et al. 2008, Xie et al. 2014). Although swellings occur during the acute phase of stroke and can be transient, according to our data (Fig. 6) they can persist for longer periods after OGD. The observed differences in dendritic swelling between neuronal subtypes may be attributable to cell swelling during the acute phase. Interestingly, dendrite swellings were also produced by an excitotoxicity insult in vitro through NMDA application, pointing to potential common mechanisms (Weilinger et al. 2016).

Thus, electrophysiological and morphological data indicate that ischemia has a more profound effect on PFC inhibitory PV-positive interneurons than on excitatory pyramidal neurons. As a result, excitation-inhibition balance in PFC circuitry shifts away from inhibition, a mechanism that may be critical to the pathophysiology of stroke.

Highlights.

PV-positive interneurons are more vulnerable to oxygen-glucose deprivation-reoxygenation (OGD-RO) than pyramidal neurons.

Large amplitude spontaneous inhibitory but not excitatory responses in pyramidal neurons are reduced by OGD-RO.

Evoked disynaptic inhibitory responses in pyramidal neurons produced by PV-positive interneurons are reduced by OGD-RO.

OGD-RO changes the excitation-inhibition balance in the neocortex resulting in insufficient inhibition of pyramidal neurons.

Dendrites of PV-positive interneurons exhibited more pathological beading than those of pyramidal neurons following OGD-RO.

Acknowledgments

This work was supported by the National Institutes of Health Grants R01 MH045817 and R01 GM128195 (JWJ). The authors thank Lihua Ming for excellent technical assistance.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

The authors declare no competing financial interests.

References

  1. Allen NJ, Karadottir R & Attwell D (2005) A preferential role for glycolysis in preventing the anoxic depolarization of rat hippocampal area CA1 pyramidal cells. J Neurosci, 25, 848–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Arabadzisz D & Freund TF (1999) Changes in excitatory and inhibitory circuits of the rat hippocampus 12–14 months after complete forebrain ischemia. Neuroscience, 92, 27–45. [DOI] [PubMed] [Google Scholar]
  3. Bacigaluppi M, Russo GL, Peruzzotti-Jametti L, Rossi S, Sandrone S, Butti E, De Ceglia R, Bergamaschi A, Motta C, Gallizioli M, Studer V, Colombo E, Farina C, Comi G, Politi LS, Muzio L, Villani C, Invernizzi RW, Hermann DM, Centonze D & Martino G (2016) Neural Stem Cell Transplantation Induces Stroke Recovery by Upregulating Glutamate Transporter GLT-1 in Astrocytes. J Neurosci, 36, 10529–10544. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brisson CD & Andrew RD (2012) A neuronal population in hypothalamus that dramatically resists acute ischemic injury compared to neocortex. J Neurophysiol, 108, 419–30. [DOI] [PubMed] [Google Scholar]
  5. Brown CE, Aminoltejari K, Erb H, Winship IR & Murphy TH (2009) In vivo voltage-sensitive dye imaging in adult mice reveals that somatosensory maps lost to stroke are replaced over weeks by new structural and functional circuits with prolonged modes of activation within both the periinfarct zone and distant sites. J Neurosci, 29, 1719–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chattopadhyaya B, Di Cristo G, Higashiyama H, Knott GW, Kuhlman SJ, Welker E & Huang ZJ (2004) Experience and activity-dependent maturation of perisomatic GABAergic innervation in primary visual cortex during a postnatal critical period. J Neurosci, 24, 9598–611. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Clarkson AN, Huang BS, Macisaac SE, Mody I & Carmichael ST (2010) Reducing excessive GABA-mediated tonic inhibition promotes functional recovery after stroke. Nature, 468, 305–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cossart R, Dinocourt C, Hirsch JC, Merchan-Perez A, De Felipe J, Ben-Ari Y, Esclapez M & Bernard C (2001) Dendritic but not somatic GABAergic inhibition is decreased in experimental epilepsy. Nat Neurosci, 4, 52–62. [DOI] [PubMed] [Google Scholar]
  9. Cruikshank SJ, Urabe H, Nurmikko AV & Connors BW (2010) Pathway-specific feedforward circuits between thalamus and neocortex revealed by selective optical stimulation of axons. Neuron, 65, 230–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dirnagl U, Iadecola C & Moskowitz MA (1999) Pathobiology of ischaemic stroke: an integrated view. Trends Neurosci, 22, 391–7. [DOI] [PubMed] [Google Scholar]
  11. Du Y, Wang W, Lutton AD, Kiyoshi CM, Ma B, Taylor AT, Olesik JW, McTigue DM, Askwith CC & Zhou M (2018) Dissipation of transmembrane potassium gradient is the main cause of cerebral ischemia-induced depolarization in astrocytes and neurons. Exp Neurol, 303, 1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Haider B & McCormick DA (2009) Rapid neocortical dynamics: cellular and network mechanisms. Neuron, 62, 171–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hiu T, Farzampour Z, Paz JT, Wang EH, Badgely C, Olson A, Micheva KD, Wang G, Lemmens R, Tran KV, Nishiyama Y, Liang X, Hamilton SA, O’Rourke N, Smith SJ, Huguenard JR, Bliss TM & Steinberg GK (2016) Enhanced phasic GABA inhibition during the repair phase of stroke: a novel therapeutic target. Brain, 139, 468–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kang SK, Johnston MV & Kadam SD (2015) Acute TrkB inhibition rescues phenobarbital-resistant seizures in a mouse model of neonatal ischemia. Eur J Neurosci, 42, 2792–804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kislin M, Sword J, Fomitcheva IV, Croom D, Pryazhnikov E, Lihavainen E, Toptunov D, Rauvala H, Ribeiro AS, Khiroug L & Kirov SA (2017) Reversible Disruption of Neuronal Mitochondria by Ischemic and Traumatic Injury Revealed by Quantitative Two-Photon Imaging in the Neocortex of Anesthetized Mice. J Neurosci, 37, 333–348. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Lauderdale K, Murphy T, Tung T, Davila D, Binder DK & Fiacco TA (2015) Osmotic Edema Rapidly Increases Neuronal Excitability Through Activation of NMDA Receptor-Dependent Slow Inward Currents in Juvenile and Adult Hippocampus. ASN Neuro, 7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Lei Z, Zhang H, Liang Y & Xu ZC (2016) Reduced expression of IA channels is associated with post-ischemic seizures. Epilepsy Res, 124, 40–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Liu LY, Wei EQ, Zhao YM, Chen FX, Wang ML, Zhang WP & Chen Z (2005) Protective effects of baicalin on oxygen/glucose deprivation- and NMDA-induced injuries in rat hippocampal slices. J Pharm Pharmacol, 57, 1019–26. [DOI] [PubMed] [Google Scholar]
  19. Luo CX, Lin YH, Qian XD, Tang Y, Zhou HH, Jin X, Ni HY, Zhang FY, Qin C, Li F, Zhang Y, Wu HY, Chang L & Zhu DY (2014) Interaction of nNOS with PSD-95 negatively controls regenerative repair after stroke. J Neurosci, 34, 13535–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Marek R, Jin J, Goode TD, Giustino TF, Wang Q, Acca GM, Holehonnur R, Ploski JE, Fitzgerald PJ, Lynagh T, Lynch JW, Maren S & Sah P (2018) Hippocampus-driven feed-forward inhibition of the prefrontal cortex mediates relapse of extinguished fear. Nat Neurosci, 21, 384–392. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. McBain CJ & Fisahn A (2001) Interneurons unbound. Nat Rev Neurosci, 2, 11–23. [DOI] [PubMed] [Google Scholar]
  22. Murphy TH, Li P, Betts K & Liu R (2008) Two-photon imaging of stroke onset in vivo reveals that NMDA-receptor independent ischemic depolarization is the major cause of rapid reversible damage to dendrites and spines. J Neurosci, 28, 1756–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Neumann-Haefelin T, Staiger JF, Redecker C, Zilles K, Fritschy JM, Mohler H & Witte OW (1998) Immunohistochemical evidence for dysregulation of the GABAergic system ipsilateral to photochemically induced cortical infarcts in rats. Neuroscience, 87, 871–9. [DOI] [PubMed] [Google Scholar]
  24. Nitsch C, Scotti A, Sommacal A & Kalt G (1989) GABAergic hippocampal neurons resistant to ischemia-induced neuronal death contain the Ca2(+)-binding protein parvalbumin. Neurosci Lett, 105, 263–8. [DOI] [PubMed] [Google Scholar]
  25. Papp E, Rivera C, Kaila K & Freund TF (2008) Relationship between neuronal vulnerability and potassium-chloride cotransporter 2 immunoreactivity in hippocampus following transient forebrain ischemia. Neuroscience, 154, 677–89. [DOI] [PubMed] [Google Scholar]
  26. Petilla Interneuron Nomenclature, G., Ascoli GA, Alonso-Nanclares L, Anderson SA, Barrionuevo G, Benavides-Piccione R, Burkhalter A, Buzsaki G, Cauli B, Defelipe J, Fairen A, Feldmeyer D, Fishell G, Fregnac Y, Freund TF, Gardner D, Gardner EP, Goldberg JH, Helmstaedter M, Hestrin S, Karube F, Kisvarday ZF, Lambolez B, Lewis DA, Marin O, Markram H, Munoz A, Packer A, Petersen CC, Rockland KS, Rossier J, Rudy B, Somogyi P, Staiger JF, Tamas G, Thomson AM, Toledo-Rodriguez M, Wang Y, West DC & Yuste R (2008) Petilla terminology: nomenclature of features of GABAergic interneurons of the cerebral cortex. Nat Rev Neurosci, 9, 557–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Povysheva NV, Gonzalez-Burgos G, Zaitsev AV, Kroner S, Barrionuevo G, Lewis DA & Krimer LS (2006) Properties of excitatory synaptic responses in fast-spiking interneurons and pyramidal cells from monkey and rat prefrontal cortex. Cereb Cortex, 16, 541–52. [DOI] [PubMed] [Google Scholar]
  28. Povysheva NV & Johnson JW (2016) Effects of memantine on the excitation-inhibition balance in prefrontal cortex. Neurobiol Dis, 96, 75–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Povysheva NV, Zaitsev AV, Rotaru DC, Gonzalez-Burgos G, Lewis DA & Krimer LS (2008) Parvalbumin-positive basket interneurons in monkey and rat prefrontal cortex. J Neurophysiol, 100, 2348–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Qu M, Buchkremer-Ratzmann I, Schiene K, Schroeter M, Witte OW & Zilles K (1998) Bihemispheric reduction of GABAA receptor binding following focal cortical photothrombotic lesions in the rat brain. Brain Res, 813, 374–80. [DOI] [PubMed] [Google Scholar]
  31. Risher WC, Ard D, Yuan J & Kirov SA (2010) Recurrent spontaneous spreading depolarizations facilitate acute dendritic injury in the ischemic penumbra. J Neurosci, 30, 9859–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Schiene K, Bruehl C, Zilles K, Qu M, Hagemann G, Kraemer M & Witte OW (1996) Neuronal hyperexcitability and reduction of GABAA-receptor expression in the surround of cerebral photothrombosis. J Cereb Blood Flow Metab, 16, 906–14. [DOI] [PubMed] [Google Scholar]
  33. Schneider CA, Rasband WS & Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. Nat Methods, 9, 671–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Stief F, Zuschratter W, Hartmann K, Schmitz D & Draguhn A (2007) Enhanced synaptic excitation-inhibition ratio in hippocampal interneurons of rats with temporal lobe epilepsy. Eur J Neurosci, 25, 519–28. [DOI] [PubMed] [Google Scholar]
  35. Thompson RJ (2014) Pannexin channels and ischaemia. J Physiol [DOI] [PMC free article] [PubMed]
  36. Van Den Bosch L, Schwaller B, Vleminckx V, Meijers B, Stork S, Ruehlicke T, Van Houtte E, Klaassen H, Celio MR, Missiaen L, Robberecht W & Berchtold MW (2002) Protective effect of parvalbumin on excitotoxic motor neuron death. Exp Neurol, 174, 150–61. [DOI] [PubMed] [Google Scholar]
  37. Wang JH (2003) Short-term cerebral ischemia causes the dysfunction of interneurons and more excitation of pyramidal neurons in rats. Brain Res Bull, 60, 53–8. [DOI] [PubMed] [Google Scholar]
  38. Weilinger NL, Lohman AW, Rakai BD, Ma EM, Bialecki J, Maslieieva V, Rilea T, Bandet MV, Ikuta NT, Scott L, Colicos MA, Teskey GC, Winship IR & Thompson RJ (2016) Metabotropic NMDA receptor signaling couples Src family kinases to pannexin-1 during excitotoxicity. Nat Neurosci, 19, 432–42. [DOI] [PubMed] [Google Scholar]
  39. Wilson NR, Runyan CA, Wang FL & Sur M (2012) Division and subtraction by distinct cortical inhibitory networks in vivo. Nature, 488, 343–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Xie Y, Chen S, Wu Y & Murphy TH (2014) Prolonged deficits in parvalbumin neuron stimulation-evoked network activity despite recovery of dendritic structure and excitability in the somatosensory cortex following global ischemia in mice. J Neurosci, 34, 14890–900. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Zhang H, Schools GP, Lei T, Wang W, Kimelberg HK & Zhou M (2008) Resveratrol attenuates early pyramidal neuron excitability impairment and death in acute rat hippocampal slices caused by oxygen-glucose deprivation. Exp Neurol, 212, 44–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Zhang H, Xie M, Schools GP, Feustel PF, Wang W, Lei T, Kimelberg HK & Zhou M (2009) Tamoxifen mediated estrogen receptor activation protects against early impairment of hippocampal neuron excitability in an oxygen/glucose deprivation brain slice ischemia model. Brain Res, 1247, 196–211. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES