Potent inhibition of anoxic depolarization by the sodium channel blocker dibucaine

Abstract

Recurring waves of peri-infarct depolarizations (PIDs) propagate across gray matter in the hours and days following stroke, expanding the primary site of injury. Ischemic depolarization (termed anoxic depolarization or AD in live brain slices) is PID-like but immediately arises in the more metabolically compromised ischemic core. This causes dramatic neuronal and astrocyte swelling and dendritic beading with spine loss within minutes, resulting in acute cell death. AD is evoked in rodent neocortical slices by suppressing the Na⁺/K⁺-ATPase pump with either oxygen/glucose deprivation (OGD) or exposure to ouabain. The process driving AD and PIDs remains poorly understood. Here we show that dibucaine is a potent drug inhibiting AD because of its high binding affinity to the Na⁺ channel. Field recording reveals that, when superfused with ouabain (5 min), neocortical slices pretreated with 1 μM dibucaine for 45 min display either no AD or delayed AD onset compared with untreated controls. If ouabain exposure is extended to 10 min, 1 μM dibucaine is still able to delay AD onset by ∼60%. Likewise, it delays OGD-evoked AD onset by ∼54% but does not depress action potentials (APs) or evoked orthodromic field potentials. Increasing dibucaine to 10 μM inhibits AP firing, gradually putting the slice into a stasis that inhibits AD onset but also renders the slice functionally quiescent. Two-photon microscopy reveals that 10 μM dibucaine pretreatment prevents or helps reverse ouabain-induced structural neuronal damage. Although the therapeutic range of dibucaine is quite narrow, dibucaine-like drugs could prove therapeutically useful in inhibiting PIDs and their resultant neuronal damage.

within minutes of stroke onset in animal models, neurons and glia in the ischemic core undergo ischemic depolarization, an event where failure of the Na⁺/K⁺-ATPase pump coincides with loss of neuronal membrane selectivity to ion permeation. In live cortical slices, this process is usually termed anoxic depolarization (AD) and is induced by oxygen/glucose deprivation (OGD). In vivo and in slices, cytoskeletal disruption leads to immediate neuronal swelling and dendritic beading, but membrane repolarization and homeostasis are not reestablished because energy supply is inadequate. Over several minutes, permeability of neuronal membrane to smaller molecules (<1,000 Da) increases as Pannexin 1 channels open (Thompson et al. 2006; 2008). However, AD can still proceed with Pannexin 1 channels blocked (Thompson and MacVicar 2008). To date, no drug has been proven clinically useful for neuroprotection following stroke. No channel blocker of one particular ion, nor any single transmitter antagonist alone (including glutamate receptor antagonists), blocks AD outright in brain slices (see Anderson et al. 2005). Likewise, glutamate receptor antagonists are ineffective in vivo (Murphy et al. 2008). However, certain σ-receptor ligands (Anderson et al. 2005) and Na⁺ channel blockers (Weber and Taylor 1994; Taylor and Meldrum 1995) delay AD onset in cortical slices of rodents.

The initiating mechanism of AD is of major interest because animal studies reveal that AD generation is the major determinant of ischemic brain damage in the ischemic core (Kaminogo et al. 1998). More therapeutically relevant, during the hours and days following stroke onset (Hartings et al. 2003) recurring AD-like events termed peri-infarct depolarizations (PIDs) spread from the border of the ischemic core out into the penumbra (Nedergaard and Astrup 1986; Nedergaard and Hansen 1993; Oliveira-Ferreira et al. 2010), the surrounding brain area that is partially metabolically compromised. PIDs may propagate out further into healthy brain tissue, where they are recorded as spreading-depression (SD) events. Although SD is innocuous, AD results in irreversible damage because afflicted cells lack the energy required to repolarize and so can die within minutes. PIDs are also proposed to promote damage by depleting energy reserves of penumbral neurons. PID propagation promotes acute injury to synaptic circuitry in the ischemic penumbra (Risher et al. 2010) and increases both growth rate (Hartings et al. 2003) and volume (Busch et al. 1996; Takano et al. 1996; Back et al. 1996) of the ischemic territory. Lesion volume in human (Schiemanck et al. 2005a; 2005b) and rodent (Alexis et al. 1996) stroke subjects correlates with functional deficits. If indeed PIDs promote this expansion by increasing metabolic demand and inducing hypoperfusion as suggested by patient recordings (Dreier et al. 2006; Dohmen et al. 2008; Oliveira-Ferreira et al. 2010), then it opens a dramatic temporal window of opportunity for drug treatment to reduce the progressing injury. In addition, we suggest that blood-brain-barrier disruption caused by recurring depolarization (Gursoy-Ozdemir et al. 2004) could provide a direct route for drug delivery to inhibit PIDs and so limit infarct expansion.

Using brain slices to study these depolarizing events is useful for studying the effects of drugs on the depolarizations. SD can be repeatedly evoked in brain slices by exposing briefly elevating [K⁺]_o (Anderson and Andrew 2002). AD can be initiated by exposing brain slices to OGD or briefly to ouabain, which, like OGD, inhibits Na⁺/K⁺-ATPase. The AD induced by OGD or ouabain is identical as measured by changes in light transmittance (LT), electrophysiology, and drug effects (Jarvis et al. 2001; Anderson et al. 2005). Considering these similarities, we use the term AD to describe both the OGD-induced depolarization and the ouabain-induced depolarization (Joshi and Andrew 2001; Jarvis et al. 2001; Anderson et al. 2005).

Here we examine dibucaine, a local anesthetic and Na⁺ channel blocker of the caine family that delays AD in human cortical slices during simulated ischemia (Risher et al. 2007). We show that dibucaine pretreatment at 1 μM inhibits AD onset yet does not reduce evoked field potentials. Also, two-photon laser-scanning microscopy (2PLSM) reveals dramatic protection of pyramidal neurons from AD-induced injury in neocortical slices pretreated with 10 μM of dibucaine. We also compare binding studies and orders of potency used to compare dibucaine with other Na⁺ channel blockers to help determine how AD onset may be inhibited.

MATERIALS AND METHODS

Cortical brain-slice preparation.

All procedures follow National Institutes of Health guidelines for the humane care and use of laboratory animals and undergo yearly review by the Animal Care and Use Committees at the Queen's University and the Medical College of Georgia. All efforts were made to minimize animal discomfort and reduce the number of animals used. Male Sprague-Dawley rats, 21–29 days old (Charles River, St. Constant, Quebec, Canada), and 12 male and female transgenic mice of the B6.Cg-Tg(Thy1-EGFP)MJrs/J strain [GFP-M; where GFP is green fluorescent protein] at an average age of 4 mo were used in this study. The founding mice of [GFP-M] colony were kindly provided by Dr. J. Sanes (Harvard University, Cambridge, MA). Mice of this strain display bright fluorescence in a small fraction of pyramidal neurons of the neocortex and hippocampus, providing high contrast, thus facilitating 2PLSM imaging. Rodents were housed with food and water supplied ad libitum in approved controlled environments at Queen's University and the Medical College of Georgia. Rats were either guillotined in a rodent restrainer or intracardially perfused with a high-sucrose solution while anesthetized by an intraperitoneal injection of pentobarbital sodium (Somnotol; MTC Pharmaceuticals, Joliet, IL) and then guillotined. For 2PLSM, mice were deeply anesthetized with halothane and decapitated. The rodent brain was quickly excised and placed in ice-cold oxygenated (95% O₂-5% CO₂) high-sucrose-based fluid (below). Coronal slices (400 μm) were taken from neocortical frontal and parietal regions using a vibrating blade microtome (Leica VT1000S; Leica, Wetzlar, Germany). More caudal slices included hippocampus. Slices were transferred to a net submerged in a beaker of artificial cerebrospinal fluid (aCSF) gassed with 95% O₂-5% CO₂ at 22°C. Slices were incubated at 29°C for at least 1 h before experimentation at 33–35°C.

Experimental solutions.

The aCSF contained (in mM): 120 NaCl, 3.3 KCl, 26 NaHCO₃, 1.3 MgSO₄, 1.2 NaH₂PO₄, 11 d-glucose, 1.8 and CaCl₂ (pH 7.3–7.4). The high-sucrose-based fluid contained the same constituents except that NaCl was removed and replaced with equimolar sucrose. Sucrose is impermeable to cells, and excising and slicing the brain in high-sucrose solution reduces damage to the slices (Kirov et al. 2004).

Ischemia was simulated by either OGD or by addition of the Na⁺/K⁺-ATPase inhibitor ouabain (100 μM, Sigma-Aldrich, St. Louis, MO). For OGD, aCSF glucose was reduced from 11 to 1 mM (replaced with equimolar NaCl), and the 95% O₂-5% CO₂ mixture gassing the aCSF was replaced with 95% N₂-5% CO₂. Our laboratory has previously shown that 100 μM ouabain induces LT changes during and after AD indistinguishable from OGD (Jarvis et al. 2001). Several laboratories, including ours, have shown the same with the AD waveform during intracellular recording of pyramidal neurons. The one difference we found was that OGD eliminated synaptic input pre-AD, whereas ouabain did not (Anderson et al. 2005).

Drugs tested (Sigma-Aldrich) were all voltage-gated sodium channel blockers including the following: dibucaine hydrochloride (0.1, 1.0, and 10 μM), tetrodotoxin (1, 10 μM), procaine hydrochloride (10, 100 μM), lidocaine hydrochloride (10, 100 μM), tetracaine hydrochloride (1, 10 μM) and bupivacaine hydrochloride (1, 10 μM). All drugs were dissolved in aCSF and added to aCSF to the desired concentration. To prepare some of the low-concentration drug solutions, stock solutions were made by dissolving the drug in double-distilled H₂O, and aliquots were added to aCSF.

Imaging LT.

Changes in near-infrared LT (Fig. 1, A and B) were monitored in real time. A front of cell swelling is imaged as an increase in LT during AD initiation and propagation in cortical brain slices (Obeidat and Andrew 1998; Basarsky et al. 1998). Dendritic damage is imaged as decreased LT in the wake of the AD front that follows the LT increase. This LT reduction is caused by the formation of dendritic beads (Polischuk et al. 1998; Obeidat et al. 2000), which form within minutes of AD onset in brain slices (Tanaka et al. 1999; Andrew et al. 2007) and effectively scatter light (Jarvis et al. 1999). Beading is also observed in vivo following focal ischemia (Hori and Carpenter 1994; Murphy et al. 2008; Risher et al. 2010).

Fig. 1.

A: equipment for imaging light transmittance (LT) consists of a broadband halogen light source filtered by a near-infrared pass filter. The light is scattered, absorbed, or transmitted by the brain slice. Transmitted light is collected with a charge-coupled device (CCD) and digitized using a frame grabber controlled by the imaging software. Each pixel of the image is pseudocolored according to the change in LT (ΔLT) through that area of brain slice. B: neocortical slices pretreated for 45 min with 1 μM dibucaine display delayed or blocked anoxic depolarization (AD). Left: digitized pseudocolored images show propagation of AD in rat neocortex in an untreated control slice, as well as the dendritic damage that follows in its wake. At time 0:00 oxygen/glucose deprivation (OGD) begins. Time course of ΔLT in zone of interest is plotted in neocortical layers II/III (white box) before, during, and after 10 min of OGD. AD appears within 5 min as a bilateral wave of increased LT in gray matter moving medially (arrows). Magenta pseudocoloring indicates dendritic damage in the wake of AD and develops post-OGD throughout reintroduction of normal artificial cerebrospinal fluid (aCSF) (10:00–20:00). Right: slice pretreated for 45 min with 1 μM dibucaine. At 0:00, cosuperfusion of OGD and 1 μM dibucaine begins. Compared with the untreated slice in A, the dibucaine-treated slice undergoes AD ∼2.5 min later (arrows). However, damage is still extensive after the slice has been superfused with normal aCSF for 10 min (at 20:00). C: evoked orthodromic population spike (PS) (average of 5 sweeps) recorded in layers II/III of the neocortical slice upon stimulation of axons in layer VI or in white matter (wm). Asterisk shows stimulus artifact.

A rat brain slice was transferred to a chamber for imaging. The slice was weighted at its edges with silver wire and submerged in flowing, oxygenated aCSF (3–4 ml/min). The temperature was raised from 29 to 35°C over several minutes prior to the start of the experiment. The slice was illuminated using a broadband, voltage-regulated halogen light source (Fig. 1A) on an inverted light microscope. The light traversed a band-pass filter that transmitted red and near-infrared light (690–1,000 nm). Video frames were acquired using a COHU (San Diego, CA) charge-coupled device (CCD) that was set at maximum gain and medium black level. The γ-level was set to 1.0 so that the CCD output was linear with respect to changes in light intensity. Frames acquired at 30 Hz were averaged and digitized using a frame grabber (DT 3155; Data Translation, Marlboro, MA) and controlled by Axon Imaging Workbench (AIW 2.2) software (Axon Instruments, Foster City, CA).

Each image of a series was an average of 128 or 256 frames. Averaged images were saved to the hard drive. The first averaged image in a series served as a control (T_cont), which was subtracted from each subsequent experimental image of that series (T_exp). The resulting series of subtracted images revealed changes in LT over time. The change (ΔT) was expressed as the digital intensity of the subtracted images (T_exp − T_cont), but the software does not divide ΔT by T_cont, meaning the images themselves are not normalized with respect to their regional variations in LT. The gain was set using the AIW software. The change in LT was displayed using a pseudocolor intensity scale. Zones of interest were selected to quantify and graphically display the experimental data offline.

Graphing of data was carried out using SigmaPlot for Windows (Jandel Scientific, San Rafael, CA). The data from the imaging experiments were analyzed such that changes in LT of a given zone of interest were expressed as percent change in T_cont for that region, taken from the control image. That is: ΔLT = [(T_exp − T_cont)/gain]/T_cont × 100 = ΔT%/T. This normalized the graphical data across adjacent regions of gray matter, which was necessary because of the variation in opacity that caused different initial LT values (T_cont). The means of maximum and minimum changes in LT among slices were compared for significance. AD onset time in each drug-treated slice was expressed as a percentage of the AD onset time of a subsequent untreated (control) slice. This accommodated for hourly and daily variation in AD onset time. In general, latency to AD onset increased throughout each day.

2PLSM.

For 2PLSM imaging a mouse slice was transferred into the imaging/recording chamber (RC-29, Warner Instruments, Hamden, CT) mounted on the Luigs and Neumann microscope stage. The slice was held down with an anchor (SHD-27LP/2, Warner) while perfused with oxygenated 34°C aCSF. Images of CA1 neurons in hippocampus were collected with infrared-optimized ×40/0.8 NA water-immersion objective (Carl Zeiss, Jena, Germany) using the Zeiss LSM 510 NLO META multiphoton system mounted on the motorized upright Axioscope 2FS microscope (Zeiss). These pyramidal neurons display responses to simulated ischemia identical to those in neocortex (Andrew et al. 2007). The scan module was directly coupled with Ti:sapphire broad-band, mode-locked laser (Mai-Tai; Spectra-Physics, Mountain View, CA) tuned to 910 nm or 960 nm for two-photon excitation. To monitor structural changes, three-dimensional (3D) time-lapse images were taken at 1- or 2-μm increments using a ×1.1 or ×2 optical zoom, yielding a nominal spatial resolution of 5.03 or 9.14 pixels/μm (12 bits/pixel, 2.24 μs pixel time) across a 204 × 204 μm or 112 × 112 μm imaging fields respectively. Emitted light was detected by internal photomultiplier tubes (PMTs) of the scan module with pinhole entirely opened or in a whole-field detection mode by external nondescanned detectors (Zeiss). Data acquisition was controlled by Zeiss LSM 510 software. If shifting of the focal plane occurred, the field of focus was adjusted and recentered before acquiring image stacks (Andrew et al. 2007; Risher et al. 2009). An LSM 510 Image Examiner (Zeiss) and Bitplane Imaris software were used for image analysis. A median filter (radius = 2) was applied to images in Fig. 9 to reduce photon and PMT noise. Dendritic beading was identified by rounded regions extending beyond the diameter of the parent dendrite separated by “interbead” segments. Dendritic recovery was defined as the reversal of beading (Risher et al. 2010). Because axial resolution of 2PLSM is relatively poor (∼2 μm), we used two-dimensional maximum-intensity projections (MIPs) of image stacks to assess relative changes in the area of neuronal soma as was described by us before (Risher et al. 2009). MIP images were digitally traced by hand to measure area of soma profiles at different experimental points and then were compared for significance. Although the neuronal injury was induced with ouabain, we previously published OGD effects on pyramidal cell dendrites using two-photon microscopy (Andrew et al. 2007). The two-photon imaging is identical whether OGD or ouabain is used to induce AD.

Fig. 9.

Dibucaine (Dib) protects pyramidal neurons from irreversible injury induced by ouabain-evoked AD by preventing soma swelling and dendritic beading as well as by promoting structural recovery. A1–A3: two-photon laser-scanning microscopy (2PLSM) sequence of enhanced green fluorescent protein (eGFP)-expressing pyramidal neuron. The soma swells and dendrites bead (A1, control) during exposure to 100 μM ouabain for 10 min (A2). Neuronal morphology does not recover during next 40 min in aCSF (A3). The bar at the bottom shows the time course of the experiment and indicates when images A1–A3 were acquired. A4: summary from 11 neurons in 9 slices from 7 animals showing irreversible soma swelling induced by 10 min of superfusion with 100 μM ouabain. Values are shown as percent of control. Shading of each histogram bar corresponds to the same shading in the time-line bar above. Asterisks indicate significant difference from control (**P < 0.005). B1–B3: representative 2PLSM image sequence showing that pretreatment for 45 min with 10 μM dibucaine (B1) protects neurons and dendrites from injury during 10 min of coexposure to 100 μM ouabain (B2). Then ouabain applied to the same slice following dibucaine washout caused damage (B3). The bar at the bottom shows the time course of the entire experiment, including when B1–B3 were sampled. B4: summary from 8 neurons in 6 slices from 5 animals showing protective effect of pretreatment with 10 μM dibucaine. There was no detectable neuronal soma swelling during 10 min of coexposure to 100 μM ouabain/10 μM dibucaine. Values are shown as percent of measurements at the end of pretreatment with 10 μM dibucaine for 55.2 ± 17.5 min. Asterisks indicate significant increase in soma size in ouabain compared with aCSF/Dib, Ouabain/Dib, and aCSF/Dib experimental points (***P < 0.001). C1–C3: another representative 2PLSM image sequence illustrating that a healthy pyramidal cell (C1) became swollen and dendrites beaded during cosuperfusion with 100 μM ouabain/10 μM dibucaine (C2), but complete recovery of morphological damage was possible (C3) because a slice was pretreated with 10 μM dibucaine for 45 min. The bar at the bottom shows when experimental points C1–C3 were acquired. C4: summary from 9 neurons in 6 slices from 5 animals in experiments when pretreatment with dibucaine promoted structural recovery from ouabain-induced injury. Slices were pretreated with 10 μM dibucaine for 55.5 ± 17.9 min. Values are shown as percent of soma area at the end of pretreatment with dibucaine. Cosuperfusion with 100 μM ouabain/10 μM dibucaine resulted in soma swelling, but neurons recovered during subsequent washout in 10 μM dibucaine. Asterisk indicates significant change in soma size from the values at the end of pretreatment with dibucaine (*P < 0.05).

Evoked field potentials.

To record evoked field potentials (Fig. 1C) or the spontaneous negative shift representing the AD (Fig. 2A), a micropipette was pulled from thin-walled capillary glass, filled with 2 M NaCl (5–10 MΩ), and mounted on a 3-D micromanipulator. It was connected by a chlorided silver wire to an amplifier probe, and output was monitored on an online oscilloscope. Neocortical (Fig. 1C) slices were placed in the experimental chamber, weighted by silver wires, and the solution superfusing the slice was warmed from 29°C to 33–35°C. The micropipette tip was placed in layers II/III of the neocortex, and a concentric bipolar electrode (Rhodes Electronics, Houma, LA) was placed in white matter or layer VI to stimulate the immediately overlying layers. A current pulse (0.1 ms duration; 0.5 Hz) was applied to produce a population spike (PS) at just-maximal amplitude (Fig. 1C). The amplified signals were digitized, signal-averaged (6–15 sweeps/trace), displayed, and plotted using pCLAMP software (Axon Instruments). The effect of a specific drug on the averaged PS amplitude was measured and normalized as a percent of the control amplitude ± SE. The ability of a drug to maintain or help recover PS amplitude compared with initial values indicated neuroprotection.

Fig. 2.

Dibucaine can promote neuronal recovery. A: negative deflection in extracellular voltage in response to 5 min of slice superfusion with 100 μM ouabain indicates AD. B: mean (+SE) PS response to 5 min of 100 μM ouabain exposure to slices. Upon ouabain exposure, neocortical PS amplitude (○) is preserved until the moment of AD onset. In contrast, PS amplitude in CA1 pyramidale (●) decreases slightly upon ouabain exposure. *Show stimulus artifact. C: 5 of 10 neocortical slices pretreated in 1 μM dibucaine for 45 min do not undergo AD, and their mean (+SE) PS amplitude is regained to ∼90% of preouabain values. The other 5 slices undergo AD with an increased latency to onset compared with untreated controls and regain their PSs to a mean of ∼40% of preouabain amplitudes. Such recovery does not statistically differ from untreated slices. D: 45-min pretreatment with 0.1 μM dibucaine increases mean (±SE) latency to AD onset in 6 of 6 neocortical slices tested. Post-AD recovery of PS amplitude is not statistically improved compared with untreated controls.

σ-Receptor binding studies.

Radioligand binding was conducted using [³H]pentazocine under conditions favoring binding to σ-1 receptors (Hashimoto et al. 1997). Concentrations ranging from (0.01 to 100.00 μM) of the voltage-gated, sodium-channel-blocking drugs dibucaine, procaine, lidocaine, tetracaine, bupivacaine. and tetrodotoxin (TTX) were tested for their ability to inhibit [³H]pentazocine binding in rat cortical homogenates. Male Hooded Wistar rats (400–450 g) were killed by decapitation, and the brains were rapidly removed. Forebrain membranes were prepared according to the method of Bowen et al. (1993). Forebrains were homogenized in 10 volumes of ice-cold 10 mM Tris-sucrose buffer (pH 7.4) using a Potter-Elvehjem homogenizer. The homogenate was then centrifuged at 3,000 revolution/min for 10 min (4°C). The homogenate was retained, and the resulting pellet was resuspended in buffer (2 ml/g) and was recentrifuged. The supernatants from these two centrifugations were combined and incubated for 30 min at 25°C and then centrifuged at 16,000 revolution/min for 15 min (4°C). The pellet was resuspended in a final volume of 1.52 ml/g tissue in 10 mM Tris·HCl (pH 7.4) and stored at −80°C until required.

For binding assays the pellet was thawed and resuspended in 10 volumes of 10 mM Tris-sucrose buffer (pH 7.4). Aliquots of crude membranes containing 500 μg of protein were incubated with 3 nM [³H](+)-pentazocine (35.3 Ci/mmol; DuPont/NEN, Boston, MA), 50 μl cold test drug, and 50 mM Tris·HCl buffer (pH 8.0 at 25°C) in a final volume of 0.5 ml for 2 h at 25°C. Binding reactions were terminated with ice ice-cold 10 mM Tris·HCl (pH 8.0), and the membranes were rapidly filtered using a Brandell 48-channel cell harvester (Biochemical Research Laboratory, Gaithersburg, MD) through Whatman GF/B filters pretreated with 0.5% polyethyleneimine for at least 2 h. The radioactivity trapped by the filters was determined by liquid scintillation counting (Beckman Coulter, Fullerton, CA). Nonspecific binding was estimated in the presence of 10 μM haloperidol or (+) pentazocine. EC₅₀ values were determined using Prism GraphPad (GraphPad Software, San Diego, CA).

Statistics.

SigmaStat (Systat) and SigmaPlot were used to compute unpaired and paired Student's t-test and one-way repeated-measures ANOVA followed by Tukey's post hoc method. The significance criterion was set at P < 0.05. Data are presented as the means ± SE. Time values are means ± SD.

RESULTS

Ouabain exposure or OGD induces AD in cortical slices.

Imaging LT and electrophysiological monitoring of the negative extracellular voltage shift reveal that superfusion of neocortical or hippocampal slices with either 100 μM ouabain or OGD induces AD. AD is imaged in neocortex as a front of increased LT (Fig. 1B, untreated) that arises focally in each hemisphere in 3.5–6 min after onset of ouabain exposure (n = 120 slices, mean = 4:52 ± 59) (min:s ± SD in seconds) or in 4–7 min (n = 258 slices, mean = 5:32 ± 60.5) after OGD exposure. Depolarization and subsequent swelling propagates across neocortex, leaving dendritic damage in its wake, which is visualized as reduced LT that results partly from dendrites forming beads that scatter near-infrared light. Despite the replacement of ouabain or OGD aCSF with normal aCSF following 10 min, LT continues to decrease for several minutes (Fig. 1B).

The electrophysiological hallmark of AD is a negative shift in extracellular voltage as neurons suddenly depolarize (Fig. 2A). In this study evoked field potentials reveal that electrical silence is recorded within 5 min of ouabain exposure in both neocortex and hippocampus (Fig. 2B) and within 6 min of OGD in neocortex (Fig. 6A). Maintained neuronal injury following AD is evident from this persistent loss of the orthodromic PS in untreated neocortical slices undergoing AD induced by ouabain (Fig. 2B) or OGD (Fig. 6B).

Fig. 6.

Blocking AD promotes recovery. Mean (±SE) evoked PS amplitudes in neocortex during and after 10 min of superfusion with OGD aCSF. A: untreated slices all generated AD and lost their evoked PS. Slice pretreatment with 1 μM dibucaine for 45 min prevents AD during OGD and preserves PS amplitude post-OGD in slices. B: in untreated slices, expanding the time course (area “B” in A) reveals that OGD induces synaptic failure before AD onset, unlike ouabain-evoked AD.

Dibucaine and AD evoked by ouabain exposure (5 min).

Neocortical slices were pretreated in 1 μM dibucaine aCSF for 45 min, and then AD induction was attempted with 100 μM ouabain exposure for 5 min. The evoked field potential was unaffected in five slices because, in five of the ten slices, AD onset was completely blocked (Fig. 2C, ▴). The full amplitude of the PS 30 min after removal of ouabain was preserved, similar to untreated slices. The other five pretreated slices generated AD but with a delayed onset of 6.78 min ± 1.07 (Fig. 2C, ○) compared with untreated slices of 4.34 min ± 0.57 (●). AD onset was significantly delayed, and PS recovery was initially faster; however, then recovery measured 30 min after AD was not significantly improved (P = 0.49) compared with no pretreatment (○ vs. ●).

Slices pretreated with 0.1 μM dibucaine for 45 min (Fig. 2D) showed AD onset 5.63 min ± 0.48 vs. untreated at 4.14 min ± 0.27. Again, AD was significantly delayed. and recovery of the evoked response was better compared with untreated slices but not significantly.

Dibucaine and AD evoked by ouabain exposure (10 min).

Next, to determine whether dibucaine could inhibit ouabain-induced AD during a more severe metabolic challenge, the chemical ischemia was extended from 5 to 10 min (Figs. 3 and 4). Preincubation in 1 μM dibucaine was effective (Fig. 3A), yielding AD block following 10 min of ouabain exposure in 20 of 34 trials and significantly delaying AD in the remainder by an average of 59 ± 7.1% (Figs. 3B and 4A). LT remained unchanged throughout the experiment when AD was blocked (Fig. 3A), indicating prevention of AD and of the damage that ensues. In those trials of 1 μM dibucaine where AD was only delayed (Fig. 3B), slice swelling (i.e., increased LT) during AD was significantly decreased with 1 μM dibucaine pretreatment (Fig. 4B), but subsequent LT reduction was not different than untreated slices.

Fig. 3.

Dibucaine blocks or delays AD. In untreated slices, AD is imaged as a sharp increase in LT followed by a reduction in LT that signifies dendritic damage. A: in some slices from the same animal pretreated with 1 μM dibucaine, AD is prevented, so subsequent damage is minimal. B: in other slices 1 μM dibucaine delays AD onset, but subsequent AD-induced injury remains substantial.

Fig. 4.

Potencies of Na⁺ channel blockers vary in inhibiting neocortical AD. Imaging is carried out during 10-min application of 100 μM ouabain to slices pretreated with sodium channel blockers for 45 min. A: AD onset times. The x-axis represents the percentage of trials for each drug treatment in which AD was blocked. The y-axis represents the mean (±SE) AD onset time in the remainder of trials as a percentage of corresponding controls. Drugs at the top right corner inhibit AD onset better than drugs at the bottom left. B: mean (±SE) maximum (cell swelling) and minimum (dendritic damage) LT values in slices undergoing AD. Except for the reduction in maximum LT seen with 1 μM dibucaine, slice pretreatment with sodium channel blockers does not affect maximum and minimum LT compared with untreated slices (*P < 0.05). Dibucaine at 1 or 10 μM does not affect post-AD dendritic damage compared with untreated slices. Note that LT change in drug-treated slices generating AD was compared with untreated slices tested on the same experimental day. TTX, tetrodotoxin.

Slice preincubation in 10 μM dibucaine blocked AD during 10 min of ouabain exposure in 10 of 14 trials and significantly delayed AD onset time in the remaining four trials by 28 ± 14.7% (mean ± SE) (Fig. 4A). As noted below, however, this may not be surprising given that preincubation in 10 μM dibucaine for +40 min electrically silences the slices.

Because the best-documented action of dibucaine is blockade of voltage-gated Na⁺ channels, the efficacy of other voltage-gated Na⁺ channel blockers to inhibit AD was also investigated. In imaging experiments in which 100 μM ouabain superfused each neocortical slice for 10 min, pretreatment for 45 min in the local anesthetic procaine at (100 μM) or the potent and selective Na⁺ channel blocker TTX (10 μM) prevented AD from initiating in some slices, as did dibucaine at 1 and 10 μM (Fig. 4A). However, incubation in 10 μM procaine, 1 μM TTX, or 10 μM lidocaine (also a local anesthetic) did not block AD in any slice and only slightly delayed its onset (Fig. 4A). Although 10 μM TTX was more effective than the same concentration of dibucaine, dibucaine at 1 μM displayed higher potency (Fig. 4A). Thus dibucaine was the most potent caine in delaying or blocking AD in our study. In addition, ΔLT was compared in drug-treated slices that had generated AD vs. control slices performed on the same experimental days. Drug incubation usually did not affect peak LT (slice swelling) during AD and maximum damage afterward (Fig. 4B).

Does dibucaine incubation affect normal health of slices?

Clearly, dibucaine is effective at 1–10 μM in inhibiting AD onset during 10 min of simulated ischemia. However, to be an effective neuroprotective drug at these concentrations, it should not impair normal electrophysiological function. To investigate this, evoked field potentials in neocortex were recorded throughout a 45-min preincubation time at 30°C in dibucaine, then during increased warming to 35°C, and finally during return to aCSF at 35°C. Incubation in 10 μM dibucaine reduced the PS amplitude by ∼80%, whereas 1 μM dibucaine incubation reduced PS amplitude by only ∼10% (Fig. 5). Dibucaine at 1 μM largely preserved normal electrophysiological function, whereas 10 μM dibucaine essentially shut down evoked orthodromic activity irreversibly at either 35°C or 30°C.

Fig. 5.

Mean (+SE) evoked responses in neocortical layers II/III to orthodromic stimulation of layer VI throughout the 45-min slice pretreatment period in either 1 μM or 10 μM dibucaine. From time 40–45 min the superfusate was heated from 30–35°C, and at 45 min switched to normal aCSF. Dibucaine at 10 μM almost abolishes the PS (stasis), whereas 1 μM dibucaine preserves the spike amplitude at ∼90% of its initial value. *Show stimulus artifact.

Dibucaine and AD evoked by 10 min of OGD exposure.

The findings above using ouabain indicated that 1 μM dibucaine was effective at inhibiting AD without impairing slice health, so we examined the potency of the drug during AD induced by OGD (which better simulates stroke). Supporting a previous study (Anderson et al. 2005), we found that AD induced by 10 min of OGD was more difficult to block than by 10 min of 100 μM ouabain. This probably reflects the fact that all ATP-requiring processes are blocked by OGD, not just the Na⁺/K⁺-ATPase pump. Pretreatment in 1 μM dibucaine did not block OGD-induced AD but significantly delayed AD onset by 54.2 ± 8.2% (mean ± SE) (Fig. 7A). Peak LT was significantly reduced in slices treated with 1 μM dibucaine compared with untreated slices (Fig. 7B). LT reduction in dibucaine-treated neocortical slices 5 and 10 min after AD onset and later following 10 min of recovery from OGD was not reduced compared with untreated slices studied on the same experimental days (Fig. 7B). Thus post-AD damage was the same. In neocortex studied electrophysiologically, 1 μM dibucaine pretreatment blocked AD and subsequent PS loss in all five slices, whereas, in every untreated slice (n = 6), AD initiated within the 10 min of OGD with almost complete loss of PS amplitude (Fig. 6A). The evoked field potential was lost in the early minutes of OGD, prior to AD onset (Fig. 6B), apparently the result of synaptic failure. Incubation in 0.1 μM dibucaine did not significantly alter AD onset time (P = 0.44) (Fig. 7A).

Fig. 7.

Imaging shows that AD induced by 10 min of OGD in neocortical slices is robustly delayed when slices are pretreated for 45 min with 1 μM dibucaine. A: pretreatment with dibucaine at 1 μM, but not 0.1 μM, significantly increases mean (±SE) latency to AD. B: analysis of LT during and after OGD. Data from 1 zone of interest per neocortical hemisphere are pooled. Zones were selected in layer II/III areas that produced the least decrease in LT after AD in untreated slices. Slice pretreatment with 1 μM dibucaine does not affect LT at the post-AD time points tested but slightly reduces mean swelling recorded as AD sweeps through the zones of interest (*P < 0.05). Numbers by each bar represent slices tested.

The effect of other Na⁺ channel blockers on AD onset induced by 10 min of OGD was also studied. The AD delay by 1 μM dibucaine was greater (but not significantly) than the same concentration of tetracaine, bupivacaine, or TTX (Fig. 8). Only at 100 μM did procaine and lidocaine delay AD onset more than 1 μM dibucaine (Fig. 8). Together, OGD and ouabain experiments reveal that dibucaine was the most effective at inhibiting AD. All of the Na⁺ channel blockers studied did delay AD at a concentration of 10 μM or more.

Fig. 8.

Dibucaine is highly effective at inhibiting AD. Comparison of the effects of sodium channel blockers on onset time of AD in neocortex induced by 10-min OGD. All drug treatments delay mean AD (±SE) onset time significantly except for 0.1 μM dibucaine. Longest AD delay is by 1 μM dibucaine unless other drugs were applied at 10 or 100× concentration. Numbers by each bar represent slices tested.

Do Na⁺ channel blockers inhibit AD through their action at σ-1 receptors?

Previous studies showed that certain ligands that bind σ-1 receptors can delay AD, especially carbetapentane (Anderson et al. 2005). We tested whether the caines showed significant binding in this respect. There was no apparent relationship between caine binding to these receptors and AD inhibition (Table 1). However, there is a strong correlation between the strength of Na⁺ channel binding and AD onset inhibition. Recovery from AD was harder to grade across studies because caine concentrations varied considerably.

Table 1.

Comparison of drugs that inhibit AD onset

Drug	Na+ Channel Binding IC50, μM	σ-Receptor Binding IC50, μM	Relative Effectiveness Blocking AD (ouabain)	Relative Effectiveness Blocking AD (OGD)	Relative Effectiveness for Electrophys. Recovery (OGD)
Dibucaine	1.4	59	+++++	++++	+++
Tetracaine	3.4	>100	++	++	+ (Niiyama et al. 2002)
Bupivacaine	5.4	52	++	++	+++ (Niiyama et al. 2002)
Procaine	110	>100	+	+	+ (Niiyama et al. 2005)
Lidocaine	240	3.8	+	+	+++ (Niiyama et al. 2005)
Carbetapentane	78* (Trube and Netzer 1994)	3.6	+++ (Anderson et al. 2005)	++ (Anderson et al. 2005)	+++ (Anderson et al. 2005)
4-PPBP	—	<1	+	+	—

Na⁺ channel binding IC₅₀ concentrations for displacement of [³H]batrachotoxinin-a 20 α-benzoate binding in a vesicular preparation from guinea pig cerebral cortex (Creveling et al. 1983). σ-Receptor binding IC₅₀ concentrations for displacement of [³H]pentazocine binding to rat forebrain P2 fraction homogenates (this study).

^*Inferred from Trube and Netzer (1994). Na⁺ current was inhibited by dextromethorphan (DM) (IC₅₀ = 78 ± 8 μM) although intracellular recording shows that 100 μM of the σ-1 receptor ligand DM is not potent enough to alter the action potential of CA1 pyramidal neurons in slices (Wong et al. 1988). Carbetapentane displays similar binding and a common anticonvulsant action of DM but independent of the N-methyl-d-aspartate antagonist action by DM. AD, anoxic depolarization; OGD, oxygen/glucose deprivation.

Neuroprotective effects of dibucaine evaluated at the cellular level.

We next used live imaging with 2PLSM of enhanced GFP-expressing pyramidal neurons to assess the dynamic process of ouabain-induced injury and extent of neuroprotection by dibucaine while resolving single neurons and dendrites in real time. We chose ouabain rather than OGD because experiments using ouabain were logistically easier to set up.

We exposed neocortical and CA1 hippocampal pyramidal neurons in nine slices from seven GFP-M mice to 100 μM ouabain for 10 min. Pyramidal neurons were imaged first in control aCSF, then aCSF with 100 μM ouabain, and finally during return to control aCSF (Fig. 9, A1–A3). By 10 min of ouabain exposure dendrites were beaded and the neuronal somata were swollen by 37.5 ± 8.7% (P < 0.005, Fig. 9A4). As predicted by our LT imaging and PS recording, there was no recovery during a 40-min wash in control aCSF.

Because the success rate in blocking ouabain-induced AD was slightly higher when slices were preincubated in 10 μM dibucaine (10 of 14 trials or 71%) vs. 1 μM dibucaine (20 of 34 trials or 59%), we selected 10 μM for evaluating acute neuroprotection at the single-neuron level. Fourteen slices from nine animals were pretreated with 10 μM dibucaine for 54.1 ± 15.8 min. This pretreatment had no effect on dendritic morphology or the area of neuronal soma as derived from MIP images (P = 0.5, Fig. 9, B1 and C1).

Dibucaine provided complete morphological protection to dendrites and neuronal soma during cosuperfusion with 100 μM ouabain for 10 min in 6 of 14 slices (Fig. 9, B1 and B2). There was no change in dendritic structure (Fig. 9B2) or mean area of neuronal soma during application of ouabain compared with dibucaine aCSF alone (P = 0.6, Fig. 9B4). No additional changes in dendritic morphology or mean soma area were observed following 15-min wash in 10 μM dibucaine containing aCSF (P = 0.67, Fig. 9B4). However, when 100 μM of ouabain was added after 20 min of dibucaine washout, dendrites beaded and pyramidal somata swelled by 81.8 ± 11.8% (P < 0.001, Fig. 9, B3 and B4). This is expected if 10 μM dibucaine pretreatment blocked ouabain-induced AD in these slices, consistent with our ΔLT experiments. In the remaining 8 of 14 slices pretreated with 10 μM dibucaine, neurons swelled by 20.4 ± 4.7% (P < 0.05) and dendrites beaded after 10 min of cosuperfusion with 10 μM ouabain (Fig. 9, C1, C2, and C4). However, during 40-min wash in 10 μM dibucaine containing aCSF, neurons dramatically recovered from soma swelling and dendritic beading in six of eight slices (Fig. 9, C3 and C4). The two different neuroprotective responses to 10 μM dibucaine are illustrated in Fig. 9, B and C. In Fig. 9B, the dibucaine inhibits substantial soma swelling and beading compared with Fig. 9A. In Fig. 9C, recovery from soma swelling and beading is enhanced, so it appears that dibucaine is remarkably effective in reversing the neuronal swelling and dendritic beading by ouabain-induced AD. In some cases, 10 μM dibucaine completely protects neurons from injury but at the cost of inhibiting synaptic communication (Fig. 5).

DISCUSSION

The present study investigated AD, a propagating depolarization that induces acute neuron damage in the ischemic core. Cultured or dissociated neurons are often used to gauge ischemic damage, but these cells do not undergo an immediate and rapid depolarization as is observed in cortical slices (Obeidat and Andrew 1998) and in vivo (Murphy et al. 2008), and therefore they do not undergo the key event upstream to subsequent neuronal injury. Living brain slices preserve local chemical and electrical synaptic connections and the physiological relationship between neurons, astrocytes, and extracellular space. Importantly, sliced neocortex and hippocampus undergo propagating AD during stroke simulation. In the present study, electrophysiological recording and imaging of LT in these slice preparations were used to study the effects of selected caines upon AD onset and propagation as well as upon the resultant injury. In support, 2PLSM imaging was used to determine at the subcellular level both the extent of neuronal injury arising from ouabain-induced AD and whether neuroprotection was conferred by dibucaine.

The AD generated in cortical slices is similar to the initial depolarization of the ischemic core in that it is a one-time event leading to acute neuronal damage within minutes. It also resembles propagating PIDs recorded in the penumbra that can recur over many hours after a stroke, adding to the metabolic stress of the neurons and thereby extending the damage. However, PIDs are difficult to simulate in brain slices because a gradient of metabolic stress over a slice (akin to that emanating from the ischemic core) is difficult to replicate. We have proposed that PIDs are similar to AD but can recur in less metabolically compromised brain areas (Andrew et al. 2002). Therefore, a drug that inhibits the AD should also inhibit the less severe (but progressively damaging) PID waves.

Dibucaine inhibits AD onset.

In the present study, slice exposure to ouabain or OGD both reliably evoked AD. The one difference was that, upon ouabain exposure, the orthodromic field potential was preserved until the onset of AD. In contrast, synaptic failure preceded AD in those slices exposed to OGD as also reported by Anderson et al. (2005) and Kim et al. (2006). This initial loss of synaptic function before AD is also considered to occur in vivo.

Although there was no effect of dibucaine on AD propagation speed (not shown), a 45-min slice incubation in 1 μM dibucaine inhibited AD onset induced by either ouabain or OGD. This was 10 times the potency of the σ-1 receptor ligand carbetapentane, the drug that we previously found best inhibited AD in slices (Anderson et al. 2005). There are other reports of drugs at concentrations of 1 μM or less that also affect ischemia-induced events. Rat neocortical slices exposed to OGD displayed reduced lactate dehydrogenase leakage, a marker of tissue damage, when treated with 0.1 or 1.0 μM TTX (Tatsumi et al. 1998; Oka et al. 2002). Also, slice treatment with TTX as low as 0.5 μM significantly delayed OGD-induced AD onset in one study (Fung et al. 1999). Another group (Weber and Taylor 1994; Taylor and Meldrum 1995) observed that TTX at 300 nM blocked AD and TTX at 100 nM or lidocaine at 0.5 μM delayed AD. Thus Na⁺ channel blockers in concentrations that are apparently too low to affect AP discharge can nevertheless inhibit AD onset.

A role for dibucaine in reducing ischemic damage has been examined in only a few studies. Dibucaine at 10–30 μM decreased OGD-induced tissue damage as determined by lactate dehydrogenase levels (Tatsumi et al. 1998; Oka et al. 2002). It has been reported to inhibit the mitochondrial permeability transition pore (Kowaltowski and Castilho 1997; Hoyt et al. 1997), the megachannel that opens after stressful conditions like ischemia to help mediate apoptosis. Panov et al. (2004) suggested that dibucaine stabilizes the mitochondrial membrane. Recent in vivo experiments indicate that the mitochondrial permeability transition pore coactivates with ischemic depolarization (Liu and Murphy 2009).

One earlier study using intracellular electrophysiology directly linked dibucaine with AD inhibition (Yamada et al. 2004). They found that 10 μM dibucaine treatment of hippocampal slices for 20 min before OGD prolonged the latency to the rapid depolarization (i.e., AD), similar to the present study. Also pretreatment decreased the maximal slope of AD and caused significant recovery of neuronal membrane potential after reperfusion with normal aCSF. This fits with a present study where we see PS amplitude decreasing by 20-min pretreatment, whereas membrane potential remains stable (White and Andrew 2008).

Resolution of neuronal damage by dibucaine.

Dendritic beading during ischemia is an early sign of acute injury leading to neuronal death (Hsu and Buzsaki 1993; Hori and Carpenter 1994; Enright et al. 2007). It is a direct consequence of AD in cortical slices (Obeidat and Andrew 1998). Recovery from neuronal swelling and dendritic beading signifies the structural resiliency of mature pyramidal neurons (Kirov et al. 2004). Here we show that dibucaine can protect cortical slices from swelling and beading by inhibiting AD onset using ouabain. Furthermore, if AD does initiate, injury can be reversed if the slice is quickly returned to dibucaine-aCSF. Similar recovery from dendritic beading has been observed with 2PLSM in vivo during reperfusion after stroke (Zhang et al. 2005), during recovery from brief global ischemia (Murphy et al. 2008) and during rounds of PIDs in induced penumbra-like regions (Risher et al. 2010). Reversal of structural dendritic damage observed in the present study demonstrates that a drug can promote this reparative effect after stroke simulation.

Protection of dendritic circuitry by dibucaine observed in some of our 2PLSM experiments supports the preservation of field potentials recorded in several of the dibucaine-pretreated slices during exposure to ouabain. Previous electrophysiological findings have also found that drug treatment can reduce post-AD damage. For example, certain drugs are reported to improve membrane-potential recovery (Yamada et al. 2004) or PS amplitude recovery (Anderson et al. 2005) compared with untreated control slices.

However, there is a discrepancy between dendritic recoveries from ouabain-induced injury seen in 2PLSM experiments and lack of recovery of LT post-AD. Because the light scatter by dendritic beading is assumed to underlie reduction of LT after AD-induced injury (Jarvis et al. 2001; Anderson et al. 2005), it was anticipated that LT should return toward baseline levels upon dendritic recovery, but this was not observed, possibly because of significant light scatter generated by nonrecovering dendrites close to both cut surfaces of the brain slice. Direct imaging of neurons away from cut surfaces using 2PLSM is a powerful approach to follow recovery after AD.

Na⁺ channel blockers do not inhibit AD by binding σ-1 receptors.

Certain voltage-gated Na⁺ channel blockers and σ-1 receptor ligands are reliable and reasonably potent AD inhibitors. To examine a possible common action by both groups, dibucaine and other Na⁺ channel blockers were assayed for σ-1 receptor ligand affinity. Our binding studies showed that the order of affinity of these compounds for σ-1 receptors (lidocaine >> bupivacaine ≥ dibucaine >> tetracaine = procaine) did not resemble our order for AD inhibition (dibucaine ≥ tetracaine ≥ bupivacaine >> lidocaine ≥ procaine). Thus caines appear to inhibit AD through an action other than σ-1 receptor binding. A recent study of CA1 neurons using whole cell current clamp confirms that, although both drugs inhibit depolarization, dibucaine and carbetapentane have distinctly different actions (White and Andrew 2008). Moreover at 10 μM, the order of potency for prolonging AD onset in rat CA1 hippocampal slices in another study (Yamada et al. 2004) was bupivacaine ≥ dibucaine > procaine ≥ tetracaine > mepivacaine ≥ lidocaine. This corresponds to our study (except that we found procaine to be less potent) and confirms that dibucaine and bupivacaine, the most potent caines in blocking Na⁺ channels, are also the most effective inhibitors of AD.

Insight into AD initiation.

The action of dibucaine is not mediated through σ-1 receptors but clearly involves an interaction with Na⁺ channels. Urenjak and Obrenovitch (1996) proposed that the sodium channel blockers TTX and lidocaine delay AD onset principally by blocking Na⁺ channel-dependent action potential firing, thereby decreasing the metabolic load of maintaining membrane ion gradients. However, in the present study dibucaine at 1 μM inhibits AD without altering AP discharge as judged by the evoked field potential. As well, if drugs delayed AD principally by saving energy, then bypassing the need for ATP by directly inhibiting the Na⁺/K⁺-ATPase pump with ouabain should remove the AD-delaying effect of the drug. However, dibucaine and TTX blocked or delayed ouabain-induced AD, showing that a general preservation of energy is not the neuroprotective mechanism of these drugs. At higher concentration (10 μM), dibucaine blocks enough Na⁺ channels to inhibit action potential discharge, thereby inducing a stasis whereby neurons resist firing and so also cannot function via synaptic transmission. We have carried out a detailed study comparing effects of dibucaine and carbetapentane on whole cell patched pyramidal neurons (White and Andrew 2008). The two drugs inhibit AD onset by markedly different mechanisms. However, a common mechanism is probably inhibition of a persistent Na⁺ current, which will decrease excitability (Cheng et al. 2008; Taylor and Meldrum 1995).

Our work here and that of others indicates that Na⁺ channel opening is the major component of the depolarizing event although other channels are involved. This explains why Na⁺ channel blockers can delay AD but do not consistently reduce the magnitude of depolarization once AD initiates. The “leaky” neuronal membrane that results from AD may involve other presently unidentified channels that allow molecules as large as 2,000 Da across (Tanaka et al. 1999). Pannexin 1 (Px1) may provide a further conduit for the flux of ions and, in particular, for larger molecules across neuronal membrane (Thompson and MacVicar 2008). However, the opening of Px1 is responsible for neither the initiation of AD nor the major depolarization moments after AD initiation, as AD can still proceed and cause damage in the presence of the Px1 antagonist carbenoxolone (Douglas et al. 2006) and Px1 channels appear to open several minutes after AD onset (Thompson and MacVicar 2008).

Potential of dibucaine-like drugs as stroke therapeutics.

We are presently investigating whether dibucaine-like drugs are candidates for further testing and development as anti-PID therapeutics. At 1 μM, dibucaine can inhibit AD onset while also preserving synaptic function. As well, we have performed preliminary experiments that suggest that dibucaine also inhibits the onset of SD in rodent brain slices (unpublished observations, L. Carr and R. D. Andrew). Because we believe that PIDs are attributable to the same mechanism as AD and SD, in tissue at an intermediate level of metabolic compromise, we expect that dibucaine would also inhibit PIDs.

Dibucaine has never been tested clinically in stroke, probably because it has a narrow concentration range where it is both clinically useful and safe (Ogawa et al. 1998; Noda et al. 2001). Presently its therapeutic use is limited to topical anesthesia (Nupercainal), usually for relief from minor burns, insect bites, or hemorrhoid pain.

In our study, dibucaine inhibits the damage arising from AD by stopping or delaying AD onset in rat neocortical and hippocampal brain slices, and in this respect it is the most potent of the caine family of Na⁺ channel blockers. Dibucaine-like drugs could prove to be potent in reducing acute stroke damage. If given prophylactically or in the first hours or days after stroke onset, inhibition of recurring PIDs should reduce infarct size and therefore improve functional outcome.

GRANTS

This work was supported by the Heart and Stroke Foundation of Ontario grant T-4478 (R. Andrew), a HSFO studentship (H. Douglas), the Canadian Institutes of Health Research grant MOP69044 (R. Andrew), and the National Institutes of Health grants NS057113 (S. Kirov) and NS062154 (S. Kirov).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.