Abstract

When deprived of spontaneous ongoing network activity by chronic exposure to tetrodotoxin (TTX), cultured cortical neurons retract their dendrites, lose dendritic spines, and degenerate over a period of 1–2 weeks. Electrophysiological properties of these slowly degenerating neurons prior to their death are normal, but they express very large miniature excitatory postsynaptic currents (mEPSCs). Chronic blockade of these mEPSCs by the alpha-amino-5-hydroxy-3-methyl-4-isoxazole propionic acid (AMPA) receptor antagonist 6,7-Dinitroquinoxaline-2,3-dione (DNQX) had no effect of its own on cell survival, yet, paradoxically, it protected the TTX-silenced neurons from degenerating. TTX-treated neurons also exhibited deficient Ca2+ clearance mechanisms. Thus, upscaled mEPSCs are sufficient to trigger apoptotic processes in otherwise chronically silenced neurons.

Introduction

Spontaneous ongoing neuronal activity is a common feature of the cerebral cortex during development and throughout life. In spite of its pivotal role in neuronal gene regulation (reviewed by West and others 2001) and in the formation of proper neuronal connectivity (reviewed by Katz and Shatz 1996; Zhang and Poo 2001), it is not completely understood if this bioelectrical activity has any long-term contribution to neuronal maintenance and viability. Evidence, however, is mounting to suggest that it does. For example, epidemiological studies in humans have shown that the risk for Alzheimer disease and age-related dementia is lower in more educated and more intellectually active individuals (Ott and others 1999; Wilson and others 2002). In rats, enriched environment, assumed to trigger elevated levels of brain activity, was found to decrease naturally occurring apoptosis by 45% in the hippocampus (Young and others 1999). It has also dramatically improved the prospects for survival of newly generated neurons (Kempermann and others 1997; Petreanu and Alvarez-Buylla 2002) usually suffering from high degeneration rate (Cameron and McKay 2001; Gould and others 2001; Dayer and others 2003). The need for ongoing neuronal activity is best manifested in culture preparations (both in slice and dissociated cultures) where neurons have been shown to gradually die following chronic “deprivation” of activity. The main agent used to block activity in such preparations is the voltage-gated sodium channel blocker tetrodotoxin (TTX) (Baker and Ruijter 1991; Ramakers and others 1991; Ruijter and others 1991). Nevertheless, not every brain region is affected by TTX in the same manner. For example, in the cat's retina, TTX was found to decrease naturally occurring cell death of ganglion cells rather than increase it (Scheetz and others 1995).

Apoptotic neuronal death, caused by intense network activity, has been studied extensively, mainly as a model for traumatic as well as neurodegenerative brain diseases (Dirnagl and others 1999; Hardingham and Bading 2003). Whereas it is intuitively clear that excessive accumulation of intracellular calcium ([Ca2+]i) in hyperactive neurons recruits apoptotic pathways and thus leads to neuronal death, why would neurons that do not fire action potentials, and thus do not influx calcium, degenerate, and what might be the molecular mechanisms responsible for this degeneration? It has been shown that moderate increases of intracellular calcium protect hippocampal and cortical neurons from oxygen and glucose as well as trophic deprivation (Bickler and Fahlman 2004; Papadia and others 2005). One could then speculate that lack of activity may harm cortical neurons by reducing calcium to below some minimal levels. Nevertheless, no direct evidence for this was ever provided.

The present study was aimed at the analysis of neurodegeneration caused by activity deprivation in cultured cortical neurons. The results of the present study indicate that contrary to common belief, chronic suppression of network activity via the blockade of action potential discharges leads to a rise in the postsynaptic response to glutamate. This, together with a reduced ability for calcium clearance in the affected cells, is suggested to trigger the molecular machinery that causes neuronal apoptosis. Thus, it is suggested that activity enables the neuron to better cope with transient rises in intracellular calcium concentrations.

Materials and Methods

Cultures

Cultures were prepared as detailed elsewhere (Papa and others 1995; Goldin and others 2001). Briefly, Wistar rat pups were decapitated on postnatal day 3 (P3) and their brains removed and placed in a chilled (4 °C), oxygenated Leibovitz L15 medium (Biological Industries, Beit Haemek, Israel) enriched with 0.6% glucose and Gentamicin (20 μg/mL; Sigma, St. Louis, MO). Bilateral cortical tissue was mechanically dissociated and plated on 12-mm glass coverslips at 3 × 105 to 4 × 105 cells per well in a 24-well plate. The plating medium consisted of 5% heat-inactivated horse serum (HS) and 5% fetal calf serum and was prepared in MEM (minimal essential medium)-Earl salts (Biological Industries, Beit Haemek, Israel) enriched with 0.6% glucose, Gentamicin, and 2 mM glutamax. Cells were left to grow in the incubator at 37 °C, 5% CO2 for 4 days, at which time the medium was changed to 10% HS in enriched MEM, plus a mixture of 5′-fluoro-2-deoxyuridine/uridine (Sigma, 20 μg and 50 μg/mL, respectively). The medium was changed 4 days later to 10% HS in enriched MEM. TTX (at 1 μM, Alomone labs, Jerusalem, Israel) was added to the growth medium at 4–5 days in culture for 10–13 days of incubation period (unless otherwise stated). Cells were recorded in their growth medium at several days after the drug application to verify that TTX is still effective in blocking action potential discharges.

Transfection

A lipofectamine 2000™ (Invitrogen, Carlsbad, CA) mix was prepared at 1 μL/well with 50 μL/well optimem™ (Invitrogen, Carlsbad, CA) and incubated for 5 min at room temperature. This was mixed with 1.5 μg/well total DNA in 50 μL/well optimem™ and incubated for 15 min at room temperature. The mix was then added on the transfected culture wells and allowed to sit for 4–6 h until change of medium. In most cases, at least several neurons were transfected. Cotransfection efficiency for several plasmids using this method was nearly 100% (Y Pilpel and I Fishbein, unpublished observation). Transfections were made on day 11 in vitro, (DIV, 6 days after onset of exposure to TTX—for the treated group) and visualized 2–3 days later.

Electrophysiology

The cultures were transferred to a recording chamber placed in a Nikon inverted microscope and washed with standard recording medium containing (in mM) NaCl 129, KCl 4, MgCl2 1, CaCl2 2, glucose 10, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) 10, pH was adjusted to 7.4 with NaOH, and osmolarity to 320 mOsm with sucrose. TTX (0.5 μM) and picrotoxin (20 μM) were also added to this medium for the recording of spontaneous miniature excitatory postsynaptic currents (mEPSCs). Neurons were recorded at room temperature with patch pipettes containing (in mM) K-gluconate 140, NaCl 2, HEPES 10, ethyleneglycol-bis(2-aminoethylether)-N,N,N′,N′-tetraacetic acid 0.2, Na-GTP 0.3, Mg-ATP 2, phosphocreatine 10, and pH 7.4 having a resistance in the range of 6–12 Mohm. Miniature “inhibitory” synaptic currents (mIPSCs) were recorded with a patch pipette where CsCl replaced K-gluconate and with the extracellular recording medium containing DNQX (20 μM) and DL-2-Amino-5-phosphonopentanoic acid (APV) (50 μM). Signals were amplified with Axopatch 200A (Axon Instruments Inc., Foster City, CA) and were stored on IBM PC. PClamp analysis software was used for the off-line analysis of voltage/current protocols. For the analysis of mEPSCs, a 600-Hz low-pass filter was first applied (in Clampfit analysis), and the events were then analyzed using Minianalysis software, with a threshold set at 9 pA currents.

Calcium Imaging

Cultures were incubated for 1 h at room temperature with the standard recording medium containing TTX and 2 μM Fura-2AM (or Fluo-4AM for recordings of spontaneous activity). Cells were imaged thereafter on the stage of an inverted Olympus microscope, equipped with a Till Photonics light source and an Andor Technology Ixon CCCD camera (Belfast, Northern Ireland). Basal fluorescence levels to illumination at 340/380 nm and responses to pulsed application of glutamate or AMPA were recorded in several fields of each cover glass. The results are presented as the ratio of fluorescence at 340/380 nm, which reflects the concentration of [Ca2+]i.

Immunocytochemistry

Cultures were fixed in 4% paraformaldehyde for 20 min, washed with phosphate-buffered saline, and incubated for 1 h in HS with 0.3% triton and overnight at 4 °C with the primary antibodies using Vectastain ABC kit. Neurons were visualized on a Nikon Eclipse E800 microscope, and digitized pictures were taken using Nikon Act-1 software. About 4 fields were taken per 12-mm coverslip for statistical analysis. For morphology, individual neurons were stained in fixed tissue using microdrops of the membrane bound dye DiI, solubilized in oil, as detailed elsewhere (Papa and others 1995). Cells were visualized and reconstructed on a Zeiss 510 confocal microscope.

Western Blot Analysis

After TTX treatment, cortical neurons were harvested into a TT buffer supplemented with a cocktail of protease inhibitors (Sigma) and a cocktail of phosphates inhibitors (80 mM β-Glycerophosphate, 2 mM Sodium Orthovanadate, 50 mM NaF) and lysed mechanically by vortex. Protein concentration in the cleared lysate was determined using the Bradford assay (Bio-Rad, Hercules, CA). Samples containing equal amounts of protein were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. The membrane was blocked in Tris-buffered saline containing 10% skim milk. The indicated antibodies were added and immunoreactive proteins were visualized with horseradish peroxidase–conjugated protein-A and enhanced chemiluminescence. For assessment of GluR1 levels in the neurons, protein samples were first normalized to obtain similar neuron-specific enolase (NSE) bands before the membrane was stripped and the anti glutamate receptor one (GluR1)/ post synaptic density 95 (PSD95) antibodies were added. Western blot quantification was done using the Image-Pro Plus software. Protein amounts were assessed using the following formula: Band size × (band density − background density). Each of the GluR1 and PSD95 bands were normalized to their relevant NSE band level (from the same column), and finally the mean of each control was normalized to 100% to which the 2 TTX groups were compared.

Statistical Analysis

The morphological and immunocytochemical data were summarized and analyzed automatically using Image-Pro Plus software. Every experiment was repeated at least twice, with several wells used for each treatment. The stained neurons were counted in several fields of view in each glass by an independent observer. The analysis was followed by t-tests, analysis of variance (ANOVA), or nonparametric tests, as the case required. Significance was set at P < 0.05. Further, Tukey's multiple comparisons were performed when needed.

Results

Morphological Analysis of Slow Neurodegeneration in Culture

We were able to demonstrate a time dependent, massive degeneration of cultured cortical neurons over 2 weeks of exposure to TTX (Fig. 1A–C). This neuronal death was also reflected in a massive reduction of western blotting for NSE. By comparison, glial cells, which support the growth of the cultured cortical neurons, were not affected by the TTX treatment, as indicated by western blotting for the glial marker glial fibrillary acidic protein (GFAP) (Fig. 1D). Cultured hippocampal neurons were also found to be vulnerable to activity deprivation by TTX, although to a lesser extent: After 12 days of exposure to TTX, hippocampal neurons numbers declined to 38% of their respective controls (n = 8 fields for each group), whereas cortical neurons grown on the same well plate were down to 24% (n = 8 fields for each group). To examine whether TTX produced neuronal degeneration through blockade of sodium currents or some other unknown toxic properties, cultures were also exposed to saxitoxin (STX), a sodium channel blocker that acts through a different molecular mechanism. STX at 100 nM also caused degeneration of cultured neurons much like TTX: under the same testing conditions, both TTX and STX caused a decrease in cell number down to about 33% of control from 373 ± 20 to 123 ± 18 and 129 ± 33 cells per field, respectively (9 fields examined for both TTX and STX groups and 16 fields for control).

Figure 1.

Blockade of electrical activity produces a time-dependent neuronal degeneration. (A, B) The 17-day-old cortical cultures stained for NeuN in control (A) and in cultures exposed to TTX for 11 days (B). (C) Time-dependent cell death: Presence of TTX for 5 days produced a 35% decline in cell density, whereas a 15-day exposure to TTX produced a nearly complete disappearance of the cultured neurons (cells were treated with TTX at day 4 in culture, n = 6 fields for each group). (D) Western blotting for NSE and GFAP, a glial marker, in TTX and control cultures. A clear disappearance of staining for NSE but no change in GFAP in TTX-treated culture is obvious.

Figure 1.

Blockade of electrical activity produces a time-dependent neuronal degeneration. (A, B) The 17-day-old cortical cultures stained for NeuN in control (A) and in cultures exposed to TTX for 11 days (B). (C) Time-dependent cell death: Presence of TTX for 5 days produced a 35% decline in cell density, whereas a 15-day exposure to TTX produced a nearly complete disappearance of the cultured neurons (cells were treated with TTX at day 4 in culture, n = 6 fields for each group). (D) Western blotting for NSE and GFAP, a glial marker, in TTX and control cultures. A clear disappearance of staining for NSE but no change in GFAP in TTX-treated culture is obvious.

To examine if the massive TTX-induced cell death is unique to developing neurons or will it also take place when the cells are mature and extensively interconnected, 10-day-old cultures were exposed to TTX for an additional 11 days. Here too there was a massive cell death following exposure to the drug (reducing numbers to 23% of control, n = 16 for each group). To examine if the age of culture preparation makes a difference for the TTX effects, we exposed cortical cultures to TTX using the same standard conditions, except that in this condition, cells were not taken from P3 pups (as previously described, Fig. 1C) but from P0 newborn rats. A decrease to 31% of control numbers was found in these cultures after 11 days of treatment, although absolute numbers were higher for both TTX and control groups (n = 17 and 15). In yet a different experimental condition, neurons were taken from P3 rats but treated immediately after plating rather than after 4 or 5 days. In this condition, death rate was higher, with surviving cells reaching 10% of control (n = 8 and 12, data summarized in Table 1).

Table 1

Summary of the different conditions in which the effect of chronic exposure to TTX on neuronal cells survival was tested.

 The age of the newborn rats from which the culture was taken Age of culture (DIV) upon TTX treatment onset TTX treatment duration (days) Control TTX Neuronal population reduced to (% control) P
Condition one Hippocampus P3 12 365 ± 333 138 ± 8 38 0.001 
Condition two Cortex P3 12 512 ± 34 125 ± 22 24 0.001 
Condition three Cortex P3 11 496 ± 67 51 ± 9 10 0.001 
Condition four Cortex P3 10 11 638 ± 31 145 ± 34 23 0.001 
Condition five Cortex P0 11 821 ± 66 256 ± 40 31 0.001 
 The age of the newborn rats from which the culture was taken Age of culture (DIV) upon TTX treatment onset TTX treatment duration (days) Control TTX Neuronal population reduced to (% control) P
Condition one Hippocampus P3 12 365 ± 333 138 ± 8 38 0.001 
Condition two Cortex P3 12 512 ± 34 125 ± 22 24 0.001 
Condition three Cortex P3 11 496 ± 67 51 ± 9 10 0.001 
Condition four Cortex P3 10 11 638 ± 31 145 ± 34 23 0.001 
Condition five Cortex P0 11 821 ± 66 256 ± 40 31 0.001 

To test if activity deprivation affects selective populations of neurons, cultures were stained for neuronal nuclei (NeuN) and glutamic acid decarboxylase (GAD). Although staining for the γ-aminobutyric acidergic (GABAergic) marker GAD might have been slightly lighter than in the control case, the same proportion of GABAergic neurons (about 25%) was seen in the control and the TTX-treated cultures, indicating that the cell loss is not caused by a selective elimination of a subpopulation of neurons in the culture (Fig. 2).

Figure 2.

The proportion of GAD-positive neurons is the same in control and TTX-treated cultures. Cells were stained for NeuN (red) and GAD (green). The total number of cells, as well as the yellow cells (neuN + GAD–positive cells), were counted in 5 different fields for each group, to examine if the proportion of GABAergic cells changes following a massive reduction in the total number of cells in the TTX-treated dishes. In total, 273 TTX-treated neurons and 628 control neurons were counted. (A) Control culture, yellow cells indicate GAD-positive neurons; (B) TTX-treated culture; (C) Magnification of field (A); (D) Percentage of GAD-positive neurons is the same in control and TTX-treated cultures.

Figure 2.

The proportion of GAD-positive neurons is the same in control and TTX-treated cultures. Cells were stained for NeuN (red) and GAD (green). The total number of cells, as well as the yellow cells (neuN + GAD–positive cells), were counted in 5 different fields for each group, to examine if the proportion of GABAergic cells changes following a massive reduction in the total number of cells in the TTX-treated dishes. In total, 273 TTX-treated neurons and 628 control neurons were counted. (A) Control culture, yellow cells indicate GAD-positive neurons; (B) TTX-treated culture; (C) Magnification of field (A); (D) Percentage of GAD-positive neurons is the same in control and TTX-treated cultures.

We then examined whether neuronal degeneration induced by activity deprivation is reversible; in other words, can the TTX-treated neurons that are destined to die be rescued? TTX was removed from the growth medium at 10 days in culture, 5 days after their initial exposure to TTX, and 2 days before termination of the experiment. At this time, washout of TTX could restore action potential discharges virtually instantaneously in spite of 5 days of continuous exposure to the drug (Fig. 3B). Whereas a large portion of the neurons was still alive when TTX was removed, most of the cells continued to degenerate within 2 days of TTX removal, indicating that they were unable to recover from the extended period of activity deprivation and apoptotic mechanisms were apparently already initialized (cell density was reduced from 221 ± 76 to 45 ± 9 cells per field, n = 12 and 28, respectively, 1-way ANOVA followed by Tukey comparisons, P < 0.005, Fig. 3B). To ensure that spontaneous activity was restored on TTX washout, calcium imaging was used to examine spontaneous activity of control neurons and of neurons treated with TTX for 5 days (after TTX washout) (Fig. 3B). Spontaneous activity was not only restored but it actually exceeded that of controls. This hyperactivity level gradually declined, though after 70 min, it was still higher than control. (Immediately after TTX washout, the hyperactive culture exhibited about 14 bursts/min with the mean amplitude of 167 ± 10% of control amplitude and a baseline level that was 370 ± 10% of control. Seventy minutes later, activity level was down to 5 bursts/min with the mean amplitude of 124.5 ± 12% of control and baseline 155 ± 11% of control. Control culture exhibited about 2 bursts/min.) Because Galvan and others (2000) reported that chronic in vivo exposure of the hippocampus to TTX will produce focal epilepsy, this hyperactivation of the neuronal network is not surprising. At any rate, all 14 monitored cells remained alive and active during the whole experiment so that simple excitotoxicity is not likely to be the cause for their death after TTX washout.

Figure 3.

Removal of TTX 2 days before termination of the experiment did not reverse the death process of the remaining neurons. (A) Left column, control cells at day 10 in culture. Although on day of removal of TTX (after 5 days of treatment), about a third of the cells were still alive (second column from the left), removal of TTX did not enhance survival, and 2 days later almost all of them died (third column from the left). Continuous presence of TTX caused a nearly complete disappearance of cells in culture (4th column). Right column, control cells that underwent medium replacement at the same time as the TTX-treated/washed cultures and counted on day 12 in culture. Number of fields for analysis in each group are, from left to right, 7, 12, 28, 24, and 24. (In this and the following figures * indicates 0.01 < P < 0.05; ** indicates 0.001 < P < 0.01; ***P < 0.001.) (B) Calcium imaging of spontaneous activity at 10 DIV, each line representing one cell in a field containing 10–15 neurons. Each field was imaged at 5 frames/s for 60 s and spontaneous synchronized activity can be easily detected. Top, activity in control neurons right after removal of TTX present during the incubation period (1 h). Middle, activity in the TTX group (treated for 5 days) immediately after TTX washout, Note that activity is not only regained but largely exceeds that of control. Bottom, activity in the TTX group 70 min after washout, note that neurons are still active, viable, and activity level is becoming similar to controls.

Figure 3.

Removal of TTX 2 days before termination of the experiment did not reverse the death process of the remaining neurons. (A) Left column, control cells at day 10 in culture. Although on day of removal of TTX (after 5 days of treatment), about a third of the cells were still alive (second column from the left), removal of TTX did not enhance survival, and 2 days later almost all of them died (third column from the left). Continuous presence of TTX caused a nearly complete disappearance of cells in culture (4th column). Right column, control cells that underwent medium replacement at the same time as the TTX-treated/washed cultures and counted on day 12 in culture. Number of fields for analysis in each group are, from left to right, 7, 12, 28, 24, and 24. (In this and the following figures * indicates 0.01 < P < 0.05; ** indicates 0.001 < P < 0.01; ***P < 0.001.) (B) Calcium imaging of spontaneous activity at 10 DIV, each line representing one cell in a field containing 10–15 neurons. Each field was imaged at 5 frames/s for 60 s and spontaneous synchronized activity can be easily detected. Top, activity in control neurons right after removal of TTX present during the incubation period (1 h). Middle, activity in the TTX group (treated for 5 days) immediately after TTX washout, Note that activity is not only regained but largely exceeds that of control. Bottom, activity in the TTX group 70 min after washout, note that neurons are still active, viable, and activity level is becoming similar to controls.

The TTX-induced degeneration was accompanied by a gradual shrinkage of neuronal soma size, down to 57% of control (Fig. 4B), as well as by a dramatic reduction in the number and extent of their dendrites (Fig. 4A–B). Sholl analysis of DiI-stained neurons indicated a significant decrease in the number of dendrites at medium distances (50–150 μm) from the soma (Fig. 4C). Further analysis of cells transfected with green fluorescent protein (GFP) for detailed morphological visualization (Fig. 5) revealed a 34% decrease in dendritic protrusions (spines and filopodia), from 0.38 ± 0.03 protrusions/μm (control) to 0.25 ± 0.018 in TTX (n = 8 cells for each group, P < 0.01, Fig. 6A). In addition, the remaining dendritic spines in these neurons were strikingly thin, lacking the distinct mushroom shape and were barely distinguishable from sheer filopodia (Fig. 6B). These headless spines were therefore grouped together with filopodia for further analysis. Whereas in control neurons, filopodia and headless spines constituted 17.5 ± 2% of the total protrusion, in the TTX group, they amounted to 48 ± 3% (P < 0.001) of the total protrusions. (For technical reasons, neurons were transfected at 11 DIV, [already treated for 6 days] and visualized 2 or 3 days later. Thus, transfected neurons were exposed to TTX for a shorter period of time [8–9 days instead of 11 days] than DiI-stained cells, which may explain the less dramatic morphological effect in the transfected cells as opposed to DiI-stained cells.)

Figure 4.

Dendritic tree morphology is drastically reduced in neurons that were exposed to TTX for 12 days (B) compared with control (A). (C) Sholl analysis demonstrates a significant difference in dendritic arborization between control and TTX-treated neurons (n = 7 cells for control and 8 cells for TTX-treated group). (D) Neurons exposed to TTX for 11 days exhibited significantly smaller somata when compared with control. (neurons were stained for NeuN, and their soma size automatically analyzed with Image-Pro Plus software, n = 161 neurons for experimental and control groups).

Figure 4.

Dendritic tree morphology is drastically reduced in neurons that were exposed to TTX for 12 days (B) compared with control (A). (C) Sholl analysis demonstrates a significant difference in dendritic arborization between control and TTX-treated neurons (n = 7 cells for control and 8 cells for TTX-treated group). (D) Neurons exposed to TTX for 11 days exhibited significantly smaller somata when compared with control. (neurons were stained for NeuN, and their soma size automatically analyzed with Image-Pro Plus software, n = 161 neurons for experimental and control groups).

Figure 5.

Spine density and morphology are different in control and TTX neurons. (A) Control neuron transfected with GFP (14 DIV), shown in 3 different magnifications (low, top to high, bottom, note scale bar). (B) TTX neuron transfected with GFP (14 DIV treated for 9 days). Protrusions, especially spines, are less abundant in TTX than in control cells. Existing protrusions in the TTX-treated cell are thin and lack the distinct mushroom shape of spines.

Figure 5.

Spine density and morphology are different in control and TTX neurons. (A) Control neuron transfected with GFP (14 DIV), shown in 3 different magnifications (low, top to high, bottom, note scale bar). (B) TTX neuron transfected with GFP (14 DIV treated for 9 days). Protrusions, especially spines, are less abundant in TTX than in control cells. Existing protrusions in the TTX-treated cell are thin and lack the distinct mushroom shape of spines.

Figure 6.

Morphometric analysis reveals a lower protrusion density in TTX-treated neurons and higher percentage of filopodia and filopodia-like (i.e., headless spines) structures in TTX-treated cells. (n = 8 neurons for each group, between 100 and 190 μm of dendrites from 2 different 40× fields, were analyzed for each cell, 731 protrusions from 2337 μm of dendrites on total.)

Figure 6.

Morphometric analysis reveals a lower protrusion density in TTX-treated neurons and higher percentage of filopodia and filopodia-like (i.e., headless spines) structures in TTX-treated cells. (n = 8 neurons for each group, between 100 and 190 μm of dendrites from 2 different 40× fields, were analyzed for each cell, 731 protrusions from 2337 μm of dendrites on total.)

To examine whether this reduction in spines and the dendritic morphological change are reflected also in the number of actual synaptic density, DiI-labeled neurons were immunostained for the presynaptic marker synaptophysin. Spines in close proximity (0.3 μm or less) to synaptophysin puncta were regarded as innervated spines and were counted (Fig. 7A–B). A marked decrease in innervated spine density to about 18% of control levels was found (from 0.23 ± 0.06 spine/μm to 0.04 ± 0.008; n = 7 cells for each group, P < 0.001, Fig. 7C). In addition, the percentage of innervated spines of the total population was smaller in TTX than in controls (36 ± 10% in TTX and 80 ± 3% in control, P < 0.01, Fig. 7D).

Figure 7.

TTX-treated neurons express fewer innervated spines. (AB) DiI-stained dendrites immunostained for the presynaptic marker synaptophysin in control (A) and TTX (B) groups. (C) Density of synaptophysin immunopositive spines was significantly lower in TTX neurons (n = 7 neurons for each group). (D) Among the existing spines, the percentage of synaptophysin immunopositive spines was also significantly lower in the TTX group.

Figure 7.

TTX-treated neurons express fewer innervated spines. (AB) DiI-stained dendrites immunostained for the presynaptic marker synaptophysin in control (A) and TTX (B) groups. (C) Density of synaptophysin immunopositive spines was significantly lower in TTX neurons (n = 7 neurons for each group). (D) Among the existing spines, the percentage of synaptophysin immunopositive spines was also significantly lower in the TTX group.

Moreover, when cotransfected with the GFP-tagged AMPA receptor subunit GluR1 and DsRed, TTX-treated neurons tended to form fluorescent receptor clusters mainly on the dendritic shaft, whereas in control cells, clusters were formed primarily on dendritic spines (Fig. 8A–C, n = 8 neurons for each group P < 0.001). Because overexpression of glutamate receptors by itself may have some effect on dendritic morphology, we repeated the experiments in neurons cotransfected with GFP-tagged PSD95 and DsRed for imaging of cell morphology. Similar results were obtained (Fig. 8D–F, n = 8 neurons for control and n = 7 for TTX, P < 0.001). These results indicate that neurons deprived of action potentials have their synapses located primarily on shafts of dendrites that are shorter, and thus, synapses are electrically and physically closer to the soma than those of controls.

Figure 8.

TTX-treated neurons tend to form synapses on the dendritic shaft rather than on dendritic spines. (AC) Dendrites from neurons trasfected for both pDsRed and GFP-tagged GluR1, control (A) and TTX (B). (C) Whereas in control dendrites, GluR1 clusters were formed primarily on spines, in TTX-treated cells, they were formed primarily on the shaft (n = 8 neurons for each group, P < 0.001). (DF) Dendrites from neurons trasfected for both pDsRed and GFP-tagged PSD95, control (D) and TTX (E). (F) As in the GluR1 case, also PSD95 clusters were formed primarily on dendritic shafts in TTX-treated neurons.

Figure 8.

TTX-treated neurons tend to form synapses on the dendritic shaft rather than on dendritic spines. (AC) Dendrites from neurons trasfected for both pDsRed and GFP-tagged GluR1, control (A) and TTX (B). (C) Whereas in control dendrites, GluR1 clusters were formed primarily on spines, in TTX-treated cells, they were formed primarily on the shaft (n = 8 neurons for each group, P < 0.001). (DF) Dendrites from neurons trasfected for both pDsRed and GFP-tagged PSD95, control (D) and TTX (E). (F) As in the GluR1 case, also PSD95 clusters were formed primarily on dendritic shafts in TTX-treated neurons.

To further assess possible effects of activity deprivation on synapse formation, we examined the total level of native GluR1. Cultures were treated with TTX for 7 days, when a substantial number of neurons were still alive. Cultures were then homogenized and prepared for western blotting. Protein quantity was normalized to reach similar density of NSE bands and the gels blotted for the GluR1 AMPA receptor subunits. The total amount of GluR1 immunoreactivity was 4-fold higher in TTX than in control, indicating up regulation of the receptor (397 ± 81% change, P < 0.01, n = 3 for each group, Fig. 9A–B). To compare this large increase with that of a genuine synaptic marker (after all, GluR1 might not be located at the postsynaptic site), we exposed the gels to a second antibody against the postsynaptic marker PSD95. Surprisingly, this protein exhibited no difference between the TTX and control groups (105 ± 20% change, P > 0.5, Fig. 9A–B). This observation suggests that whereas both the distribution along the dendrite and the total quantity of glutamate receptors are affected by chronic silencing of cortical neurons, the number of postsynaptic sites is probably not affected. To further elucidate this possibility, we resorted to electrophysiological analysis of the chronically silenced neurons.

Figure 9.

Western blot from 11 DIV cultures treated with TTX for 7 days. (A) Levels of protein loading were normalized to produce similar bands of anti NSE antibody (lower band). Once this was achieved, the membrane was stripped and stained first for the GluR1 AMPA receptor subunit (middle band) and then for the postsynaptic marker PSD95 (top band). (B) Western blot semi quantification, n = 3 for every condition. Whereas GluR1 was markedly upregulated in TTX-treated group, PSD95 protein levels did not seem to differ from control.

Figure 9.

Western blot from 11 DIV cultures treated with TTX for 7 days. (A) Levels of protein loading were normalized to produce similar bands of anti NSE antibody (lower band). Once this was achieved, the membrane was stripped and stained first for the GluR1 AMPA receptor subunit (middle band) and then for the postsynaptic marker PSD95 (top band). (B) Western blot semi quantification, n = 3 for every condition. Whereas GluR1 was markedly upregulated in TTX-treated group, PSD95 protein levels did not seem to differ from control.

Electrophysiological Analysis of Degenerating Neurons

Electrophysiological properties of the slowly degenerating neurons were examined in whole-cell patch clamp recordings from control and long-term (11–14 days) TTX-treated neurons. Resting membrane potentials were the same for TTX and the control neurons (−57 ± 2 vs. −58 ± 2 mV, P > 0.8), yet the treated cells displayed significantly lower capacitance (48 ± 5 vs. 115 ± 3.13 pF, P < 0.0001), indicating that their surface area is smaller than controls, as expected by their minimized dendritic tree. In response to a series of voltage commands in the presence of 0.5 μM TTX, the TTX-treated neurons expressed higher input resistance (558 ± 72 vs. 168 ± 23 MΩ in the linear range around resting membrane potentials [−70 to −40 mV]). TTX-treated cells therefore exhibited slightly different voltage–current relations (F1,253 = 1.22, P > 0.27 for groups and F8,253 = 2.18, P < 0.029 for interaction, n = 14 and 16 cells for control and TTX, respectively) (Fig. 10A). We also explored possible differences in voltage-gated Ca2+ currents, studied in the presence of 0.5 μM TTX, 2 mM barium in the external solution (replacing Ca2+), and CsCl in the recording pipette, and found no differences in voltage–current relations, (2-way ANOVA for repeated measures: F1,17 < 1 for both peak and steady state fractions, n = 10 cells for TTX and n = 9 cells for control, Fig. 10B–C).

Figure 10.

Electrophysiological comparisons between control and TTX-treated cells. (A) Little differences in potassium currents between TTX-treated neurons (for 11 days), and controls were observed. Left: Hyperpolarizing and depolarizing voltage steps in the presence of TTX, and a sample of the potassium currents induced in TTX and control cells. Right: potassium currents IV curves, n = 14 and 16 cells for control and TTX, respectively. (BC) Lack of difference in calcium currents between control and TTX. Left: Hyperpolarizing and depolarizing voltage steps, and a sample of the calcium (barium) currents induced in TTX and control cells. Right: Current to voltage relations in both control and TTX, immediately (40 ms) after the step (B) to activate T-type and L-type currents, and in steady state, 500 ms after step beginning, (C) primarily L-type currents. Recordings were made in the presence of 2 mM barium (replacing calcium) in the external medium, and 120 mM cesium chloride (replacing potassium gluconate) in the pipette solution. Currents were corrected for the cells individual capacitance, n = 9 and 10 cells for control and TTX, respectively.

Figure 10.

Electrophysiological comparisons between control and TTX-treated cells. (A) Little differences in potassium currents between TTX-treated neurons (for 11 days), and controls were observed. Left: Hyperpolarizing and depolarizing voltage steps in the presence of TTX, and a sample of the potassium currents induced in TTX and control cells. Right: potassium currents IV curves, n = 14 and 16 cells for control and TTX, respectively. (BC) Lack of difference in calcium currents between control and TTX. Left: Hyperpolarizing and depolarizing voltage steps, and a sample of the calcium (barium) currents induced in TTX and control cells. Right: Current to voltage relations in both control and TTX, immediately (40 ms) after the step (B) to activate T-type and L-type currents, and in steady state, 500 ms after step beginning, (C) primarily L-type currents. Recordings were made in the presence of 2 mM barium (replacing calcium) in the external medium, and 120 mM cesium chloride (replacing potassium gluconate) in the pipette solution. Currents were corrected for the cells individual capacitance, n = 9 and 10 cells for control and TTX, respectively.

Strikingly, neurons treated with TTX for 11–14 days expressed an 82% “increase” in mEPSCs amplitudes over control neurons (25.6 ± 2.4 vs. 14 ± 1 pA at the rate of -60 mV, n = 9 and 13 cells, respectively, P < 0.001, Fig. 11A1–A4). The increase in amplitude was accompanied by a dramatic increase in mEPSCs frequency as well, from 120 ± 65 to 552 ± 146 events/min (P < 0.01, Fig. 11A5). The rise time of the mEPSCs, on the other hand, did not significantly differ between the 2 groups (4.72 ± 0.28 ms for control and 4.15 ± 0.32 ms for TTX, P > 0.2, Fig. 11A6). Furthermore, the correlation coefficient between event amplitude and rise time was small and nonsignificant (R = 0.354, P > 0.1, n = 22, Fig. 11A7), indicating that despite the reduced dendritic arborization, cable filtering did not have a major contribution to mEPSCs higher amplitude. This striking difference was not found when cells were exposed to TTX for only 2–3 days prior to recording, unlike earlier reports (Turrigiano and others 1998; Turrigiano and Nelson 2004) but similar to those of Nakayama and others (2005). Such a short exposure to TTX caused a nonsignificant 19% increase in mEPSCs amplitude (15.6 ± 3 pA [n = 11] compared with 13.1 ± 1.8 pA [n = 8]) (Fig. 11B1–B3). Likewise, for the short exposure group, the frequency of events was not significantly different between control and experimental groups (Fig. 11B4).

Figure 11.

Prolonged exposure to TTX produces a marked and specific increase in size and frequency of mEPSCs. (A1) Sample records of mEPSCs in control (top) and TTX-treated cells (bottom) to illustrate the increase in both the frequency and amplitudes of the mEPSCs. (A2) Average amplitudes of mEPSCs are significantly different between the 2 groups (P < 0.0001), with little difference in decay kinetics. (A3) Cumulative histograms of the amplitudes of the mEPSCs showing that the TTX-treated cells have much larger events than the controls. (A4,5) The mean amplitudes (A4) and frequencies (A5) of mEPSCs are significantly larger for the TTX-treated compared with control cells (P < 0.01), and the mean event rise time (A6), on the other hand, was not significantly different between the 2 groups nor was a significant correlation between events amplitude and rise time found (A7). (B) Short, 2- to 3-day exposure to TTX produced only a small and nonsignificant increase in mEPSC amplitudes. (B1) Cumulative histogram of the 2 groups of mEPSCs. (B2) Average size of the 2 groups is nearly identical, (B3) The mEPSC amplitudes are not different from each other. (B4) Frequencies of mEPSCs are not changed by a 2-day presence of TTX.

Figure 11.

Prolonged exposure to TTX produces a marked and specific increase in size and frequency of mEPSCs. (A1) Sample records of mEPSCs in control (top) and TTX-treated cells (bottom) to illustrate the increase in both the frequency and amplitudes of the mEPSCs. (A2) Average amplitudes of mEPSCs are significantly different between the 2 groups (P < 0.0001), with little difference in decay kinetics. (A3) Cumulative histograms of the amplitudes of the mEPSCs showing that the TTX-treated cells have much larger events than the controls. (A4,5) The mean amplitudes (A4) and frequencies (A5) of mEPSCs are significantly larger for the TTX-treated compared with control cells (P < 0.01), and the mean event rise time (A6), on the other hand, was not significantly different between the 2 groups nor was a significant correlation between events amplitude and rise time found (A7). (B) Short, 2- to 3-day exposure to TTX produced only a small and nonsignificant increase in mEPSC amplitudes. (B1) Cumulative histogram of the 2 groups of mEPSCs. (B2) Average size of the 2 groups is nearly identical, (B3) The mEPSC amplitudes are not different from each other. (B4) Frequencies of mEPSCs are not changed by a 2-day presence of TTX.

Furthermore, the increase in frequency and amplitude of the mEPSCs in TTX-treated cells was probably not due to some morphological amplification of synaptic currents: another group of TTX-treated neurons was compared with controls for the presence of spontaneous mIPSCs. There was a marked decrease in both amplitude and frequency of mIPSCs in the TTX-treated cells recorded with patch pipettes where CsCl replaced K-gluconate and where excitatory postsynaptic currents were blocked by DNQX (amplitudes in control cells = 27 ± 3.5 pA, at the rate of −60 mV, n = 10 cells, and in TTX-treated cells 10.7 ± 1.0 pA, n = 12, P < 0.01, frequencies in control = 12 ± 2.8 and in TTX cells 2.5 ± 0.8 mIPSCs/min, P < 0.01, Fig. 12). These results indicate that the increase in mEPSCs is selective and is due to excitatory deprivation of the TTX-exposed neurons and that the inhibitory tone is governed by different rules in the activity-deprived cells.

Figure 12.

Spontaneous mIPSCs are reduced by prolonged exposure to TTX. (A) Sample illustration of mIPSCs recorded with a CsCl containing patch pipette, in the presence of TTX, APV, and 20 μM DNQX. Large and slow decaying events are obvious in the control but not in the TTX-treated cultures. (B) Frequency of the mIPSCs and (C) their mean amplitudes are significantly reduced by exposure to TTX for 12 days.

Figure 12.

Spontaneous mIPSCs are reduced by prolonged exposure to TTX. (A) Sample illustration of mIPSCs recorded with a CsCl containing patch pipette, in the presence of TTX, APV, and 20 μM DNQX. Large and slow decaying events are obvious in the control but not in the TTX-treated cultures. (B) Frequency of the mIPSCs and (C) their mean amplitudes are significantly reduced by exposure to TTX for 12 days.

Protecting Neurons from Degeneration

The upscaling of mEPSCs is proposed to serve as a “homeostatic” mechanism counteracting the reduced level of activity (Turrigiano and others 1998). It was therefore assumed that blocking the enhanced mEPSCs will speed up the neuronal death rate over TTX alone. Cultures were chronically exposed to postsynaptic glutamatergic and GABAergic receptor antagonists (using 2-APV, DNQX, and bicuculline), thus depriving the neurons of any excitatory or inhibitory receptor activity. Strikingly, long-term deprivation of postsynaptic glutamatergic and GABAergic receptors in and of themselves did not trigger any neuronal death (Fig. 13A) (as was also recently shown by De Lima and others [2004]). Because blockade of synaptic currents should also bring about the elimination of action potentials and therefore should enhance cell death, it was suspected that some tonic activity might still be present and that this activity might be sufficient to keep the neurons alive. The blockade of AMPA and N-methyl-D-aspartate (NMDA) receptors was then combined with the elimination of action potentials by TTX. Paradoxically, when the AMPA receptor antagonist DNQX was applied together with TTX, it enhanced survival of neurons dramatically compared with TTX alone (from 6% of TTX to 79%, ANOVA followed by Tukey comparison, TTX group was significantly different from the TTX + DNQX groups [P < 0.001], whereas DNQX was insignificantly different from the control group, Fig. 13B). This indicates that the enhanced miniature AMPA-mediated current initiates TTX-induced neuronal cell death. On the other hand, the NMDA receptor antagonist MK-801 had only a marginally significant protective effect on the viability of TTX-treated neurons (from 24% of control to 51% [ANOVA followed by Tukey comparison, P > 0.063]). Interestingly, MK-801 by itself had a neuroprotective effect on neurons, increasing their numbers to 167% of control (ANOVA followed by Tukey comparisons, P < 0.0001) (Fig. 13C). These results indicate that NMDA receptors have only a marginal role, if any, in neuronal cell death induced by activity deprivation.

Figure 13.

Glutamate antagonists and TTX toxicity. (A) Silencing of cultured neurons with a combination of drugs that block synaptic activity did not mimic the effects of TTX. (N = 9 fields for each groups.) (B) Exposure to DNQX in addition to TTX almost completely reversed the neurotoxic effects of TTX. (TTX group is significantly different from the TTX + DNQX groups, n = 16 fields for control, TTX and TTX + DNQX groups, and 14 fields for the DNQX group (P < 0.0001). (C) Although NMDA blocker MK-801 by itself had a strong effect on neuronal survival (P < 0.0001), it produced only a mild, marginally significant improvement in the TTX/MK-801–treated groups (n = 11, 12, 9, 12 for control, MK-801, TTX, and TTX-MK-801, respectively).

Figure 13.

Glutamate antagonists and TTX toxicity. (A) Silencing of cultured neurons with a combination of drugs that block synaptic activity did not mimic the effects of TTX. (N = 9 fields for each groups.) (B) Exposure to DNQX in addition to TTX almost completely reversed the neurotoxic effects of TTX. (TTX group is significantly different from the TTX + DNQX groups, n = 16 fields for control, TTX and TTX + DNQX groups, and 14 fields for the DNQX group (P < 0.0001). (C) Although NMDA blocker MK-801 by itself had a strong effect on neuronal survival (P < 0.0001), it produced only a mild, marginally significant improvement in the TTX/MK-801–treated groups (n = 11, 12, 9, 12 for control, MK-801, TTX, and TTX-MK-801, respectively).

Calcium Dynamics

AMPA receptor-dependent synaptic currents may affect neuronal viability in several ways, involving Ca2+ influx and a subsequent rise of intracellular Ca2+ concentrations (Jensen and others 2001). This can be achieved either directly, through Ca2+ permeable AMPA receptors lacking the GluR2 subunits (Lu and others 1996; Liu and others 2004) or indirectly through an activation of voltage-gated Ca2+ channels and the NMDA subtype of the glutamate receptor. Furthermore, morphological differences in dendritic arborization and spine density/size may also have an important role in calcium kinetics.

There was no distinct enhancement of GluR1-lacking GluR2-mediated responses to glutamate, the hallmark of which is a rectification of the responses to AMPA at positive membrane potentials (data not shown, see also Pellegrini-Giampietro and others 1997), indicating that a differential expression of the 2 receptor subtypes is not likely to underlie the enhanced toxicity/mEPSCs. An alternative cause for the TTX-induced toxicity is that while the synaptic stimulation activates the same subsets of receptors, it produces a larger, more persistent rise of intracellular calcium concentration in the TTX-treated cells. Calcium transients in response to pulsed application of AMPA were then studied using the ratiometric Ca2+ indicator dye Fura-2. The experiments were conducted both with young cultures (10 DIV treated with TTX for 5 days, Fig. 14D–E) when neurons were still abundant and with older neurons (17 DIV treated with TTX for 12 days, Fig. 14A–C) when their numbers were drastically reduced. In both young and mature cells, TTX-treated neurons responded to the application of AMPA (0.5 mM in a pressure pipette in the presence of TTX), with higher [Ca2+]i responses and slower Ca2+ clearance (F1,64 = 28.89, P < 0.0001) for the mature neurons and (F1,78 = 28.92, P < 0.001) for the younger ones (Fig. 14). Nevertheless, higher [Ca2+]i responses were more prominent in the older cells: 170% versus 120% of respective controls and so was the increase in [Ca2+]i decay time. Mature, longer treated cells exhibited a 70% increase in the time to 50% recovery when compared with controls, whereas younger cells exhibited a 30% increase relative to their respective controls. This time-dependent enhancement of the TTX effect resembles the time-dependent increase in the mEPSC amplitudes described above.

Figure 14.

Ca2+ response to pulse application of AMPA was larger in TTX-treated neurons than in controls. (A) Sample illustrations of the Fura-2–loaded neurons' responses to the application of AMPA (50-ms pulse of 500 μM applied above the field of view from a patch pipette) control, top image, and TTX-treated, bottom image. (B) Average response of both control (n = 39) and TTX (n = 27) groups to the AMPA application in mature cells (17 DIV), treated for 12 days. (C) Same averaged response as in (B), normalized to the peak response. (D) Averaged response for younger neurons (10 DIV) treated for only 5 days (n = 47 for control, n = 49 for TTX). (E) Same response as in (D), normalized to the peak.

Figure 14.

Ca2+ response to pulse application of AMPA was larger in TTX-treated neurons than in controls. (A) Sample illustrations of the Fura-2–loaded neurons' responses to the application of AMPA (50-ms pulse of 500 μM applied above the field of view from a patch pipette) control, top image, and TTX-treated, bottom image. (B) Average response of both control (n = 39) and TTX (n = 27) groups to the AMPA application in mature cells (17 DIV), treated for 12 days. (C) Same averaged response as in (B), normalized to the peak response. (D) Averaged response for younger neurons (10 DIV) treated for only 5 days (n = 47 for control, n = 49 for TTX). (E) Same response as in (D), normalized to the peak.

For comparison with the effects of AMPA on [Ca2+]i, the responses to pressure application of NMDA were also examined in control and TTX-treated cells. NMDA (10 μM) was applied in the presence of 0 mM magnesium and 10 μM glycine. Neurons treated with TTX for 10 days had a 2.3-fold larger response than controls (F1,104 = 102.97, P < 0.0001) (Fig. 15A–B). This increased response was larger than the difference in the calcium response to AMPA, indicating that the increase in calcium responses to AMPA is not a selective effect and that TTX may upregulate the NMDA receptor as well (as previously reported by Nakayama and others 2005). Nevertheless, because the NMDA receptor is not likely to be activated in the TTX conditions for lack of sufficient depolarization needed to remove the Mg2+ block, it came as no surprise that its blockade with MK-801 did not have a major impact on neuronal survival (see above). Finally, as expected, Ca2+ decay was found to be slower in TTX-treated neurons than in controls: There was an 88% increase in the time to 50% recovery when compared with control cells (Fig. 15C).

Figure 15.

Ca2+ responses to pulse application of NMDA were larger in TTX-treated neurons than in controls. (A) Sample illustrations of the Fura-2–loaded neurons before (left) and after (right) responses to the application of NMDA (100-ms pulse of 10 μM NMDA applied above the field of view from a patch pipette) control (top) and TTX treated (bottom). (B) Averaged response of both control (n = 68 cells) and TTX (n = 38 cells) groups to the NMDA application in mature cells (15 DIV), treated for 10 days. (C) Same average response as in (D), normalized to the peak response.

Figure 15.

Ca2+ responses to pulse application of NMDA were larger in TTX-treated neurons than in controls. (A) Sample illustrations of the Fura-2–loaded neurons before (left) and after (right) responses to the application of NMDA (100-ms pulse of 10 μM NMDA applied above the field of view from a patch pipette) control (top) and TTX treated (bottom). (B) Averaged response of both control (n = 68 cells) and TTX (n = 38 cells) groups to the NMDA application in mature cells (15 DIV), treated for 10 days. (C) Same average response as in (D), normalized to the peak response.

Can the differences in responses to topical application of glutamate ligands reflect a more basic difference in the resting levels of [Ca2+]i? Comparisons among basal levels of 131 control cells, 93 TTX-treated, 47 TTX and DNQX–treated, and 86 DNQX-treated cells indicated that, indeed, [Ca2+]i levels were consistently higher in TTX-treated neurons, whereas exposure of DNQX together with TTX greatly prevented this rise in [Ca2+]i (Fig. 16).

Figure 16.

Basal [Ca2+]i is significantly higher in TTX-treated cells compared with controls. The effect of TTX was markedly reduced in the presence of DNQX, which by itself did not cause a change in basal calcium concentration. N indicates the number of cells measured.

Figure 16.

Basal [Ca2+]i is significantly higher in TTX-treated cells compared with controls. The effect of TTX was markedly reduced in the presence of DNQX, which by itself did not cause a change in basal calcium concentration. N indicates the number of cells measured.

Finally, it is not likely that the observed difference in response to AMPA and NMDA is due to a change in activation of voltage-gated calcium channels, as no differences between control- and TTX-treated cells were seen in calcium currents evoked by depolarizing voltage steps (see above). Thus, TTX is likely to cause an increase in reactivity of the AMPA receptors to glutamate, which produce larger synaptic currents, causing a larger influx of sodium and perhaps also calcium, eventually leading to activity deprivation–induced cell death.

Discussion

The present results indicate that the increase in the size of glutamatergic mEPSCs, triggered by the long-term deprivation of action potentials, may be instrumental in causing neuronal death, which can be prevented by blockade of these receptors. If this is indeed the case, why then do spontaneously active neurons, which activate more glutamate receptors and influx higher amounts of calcium ions, not die? Several possibilities should be considered; first, the dendritic morphology in TTX-treated neurons is strikingly different from that of control neurons; they have simpler and shorter arborizations, and their synaptic terminals are localized primarily on the dendritic shaft and not on dendritic spines. This may lead to poor partition and removal of calcium. A second issue is the slower [Ca2+]i clearance apparent in TTX-treated neurons. This feature is typical of higher [Ca2+] intake by mitochondria, known to capture free calcium when it reaches high concentrations, thus competing with other mechanisms of calcium clearance and release it once its concentration starts to fall (reviewed by Thayer and others 2002). Furthermore, because [Ca2+] release from the mitochondria is primarily via Na+/Ca2+ exchange processes (Zhang and Lipton 1999), it is highly likely that the lack of action potentials affects its ability to unload Ca2+. Chronically high concentration of calcium in the mitochondria might be sufficient to trigger apoptosis (Pivovarova and others 2004). Finally, action potential discharges may cause release of some survival/protective factors without which neurons may be supersensitive to even a relatively small increase in glutamate receptor activation. The nature of this survival factor is yet unknown, but several factors are potential candidates, with brain-derived neurotrophic factor (BDNF) being an obvious one. BDNF is at least partially neuroprotective to activity-deprived neurons (Ghosh and others 1994; and our own preliminary results), and its production and release were shown to be coupled to neuronal activity (Gorba and others 1999; Kohara and others 2001; Balkowiec and Katz 2002; Ichisaka and others 2003). Furthermore, blocking action potentials without blocking miniature synaptic currents was shown to dramatically decrease protein synthesis in the dendrites (Sutton and others 2004). We would therefore propose that synaptic calcium currents, when not accompanied by periodic depolarization and by the consequent growth factor release, might gradually divert protein synthesis from the dendrites (where it is synaptic in nature) to the cell body where it might be related to apoptotic pathways. Thus, the lack of action potentials impairs neurons ability to cope with even small rises of calcium, unless the neuron is supplied with the growth factor exogenously. This might also explain why blocking the NMDA receptor dramatically diminishes Ca influx but does not impair network activity (with the possible trophic release coupled to it) and has a positive effect on neuronal viability, while blocking both NMDA and AMPA or only AMPA has no such positive effect.

One major issue relates to the sequence of cellular events leading to the apoptotic cell death. We observed 3 categories of changes: “morphological” changes including the shrinkage of dendrites and disappearance of spines; “electrophysiological”, involving the increase in mEPSCs; and “biochemical”, including the changes in glutamate receptors/responses as well as changes in calcium handling machinery. Obviously, the blockade of AMPA receptors salvages the neurons and prevents their morphological deterioration, but does the morphological change precede or follow the electrophysiological change? In our hands and others (Papa and Segal 1995; Collin and others 1997; Tyler and Pozzo-Miller 2003), even a 2-day exposure to TTX causes marked changes in dendritic spine shape, but it is not clear that there is a significant change in dendritic morphology. A more detailed time-course analysis is required to address this issue.

The biological relevance of TTX-induced cell death may be expressed in the improved chances for survival of cortical neurons found in active circuits in the brain. A condition in which neuronal network activity is completely blocked (as in TTX treatment) is not likely to be found in the mature functioning cortex. On the other hand, deafferentation, caused by removal of sensory input or through the loss of other network components, is a common situation in cortical structures and is accompanied by morphological changes in the affected neurons (Harmon and Wellman 2003; Flores and others 2005). Furthermore, during development, neurons that fail to integrate into an active neuronal network are likely to be eliminated in this manner. As previously mentioned, newly generated neurons are more susceptible to degenerate (Gould and others 2001) and some of the effects of TTX such as the induction of hyperexcitability in the hippocampus are age dependent (Galvan and others 2000). Indeed, Sun and others (2005) have recently tied the phosphorylation of c-jun (triggered by a rise in intracellular calcium concentration) with programmed cell death, providing a possible molecular death trigger. Thus, the processes of deafferentation, linked to marked dendritic shrinkage and eventual cell death, are common in the cerebral cortex (Capurso and others 1997; Johnson and others 1997) but less common in other structures (e.g., cerebellum (Bravin and others 1999) or retina (scheetz and others 1995), where dendrite/spine changes are noticed but not neuronal cell death.

In conclusion, the blockade of action potential discharges triggers a slowly developing sequence of cellular events, including shrinkage of dendrites and disappearance of synaptic dendritic spines, sensitization to the postsynaptic action of glutamate, and a reduced ability to clear calcium, all of which eventually causes neuronal death.

We wish to thank V. Greenberger for the preparation of the cultures, Dr N. Maggio for participation in preliminary studies, Dr A. Avital and M. Brodt for comments on the manuscript and help with the statistical analysis, and A. Meidanik for help with the data analysis. This work was supported by a grant from The Nash Family Foundation, The Weizmann Institute. Conflict of Interest: None declared.

References

Baker
RE
Ruijter
JM
Chronic blockade of bioelectric activity in neonatal rat neocortex in vitro: physiological effects
Int J Dev Neurosci
 , 
1991
, vol. 
9
 (pg. 
321
-
329
)
Balkowiec
A
Katz
DM
Cellular mechanisms regulating activity-dependent release of native brain-derived neurotrophic factor from hippocampal neurons
J Neurosci
 , 
2002
, vol. 
22
 (pg. 
10399
-
10407
)
Bickler
PE
Fahlman
CS
Moderate increases in intracellular calcium activate neuroprotective signals in hippocampal neurons
Neuroscience
 , 
2004
, vol. 
127
 (pg. 
673
-
683
)
Bravin
M
Morando
L
Vercelli
A
Rossi
F
Strata
P
Control of spine formation by electrical activity in the adult rat cerebellum
Proc Natl Acad Sci USA
 , 
1999
, vol. 
96
 (pg. 
1704
-
1709
)
Cameron
HA
McKay
RD
Adult neurogenesis produces a large pool of new granule cells in the dentate gyrus
J Comp Neurol
 , 
2001
, vol. 
435
 (pg. 
406
-
417
)
Capurso
SA
Calhoun
ME
Sukhov
RR
Mouton
PR
Price
DL
Koliatsos
VE
Deafferentation causes apoptosis in cortical sensory neurons in the adult rat
J Neurosci
 , 
1997
, vol. 
17
 (pg. 
7372
-
7384
)
Collin
C
Miyaguchi
K
Segal
M
Dendritic spine density and LTP induction in cultured hippocampal slices
J Neurophysiol
 , 
1997
, vol. 
77
 (pg. 
1614
-
1623
)
Dayer
AG
Ford
AA
Cleaver
KM
Yassaee
M
Cameron
HA
Short-term and long-term survival of new neurons in the rat dentate gyrus
J Comp Neurol
 , 
2003
, vol. 
460
 (pg. 
563
-
572
)
De Lima
AD
Opitz
T
Voigt
T
Irreversible loss of a subpopulation of cortical interneurons in the absence of glutamatergic network activity
Eur J Neurosci
 , 
2004
, vol. 
19
 (pg. 
2931
-
2943
)
Dirnagl
U
Iadecola
C
Moskowitz Michael
A
Pathobiology of ischaemic stroke: an integrated view
Trends Neurosci
 , 
1999
, vol. 
22
 (pg. 
391
-
397
)
Flores
G
Alquicer
G
Silva-Gomez
AB
Zaldivar
G
Stewart
J
Quirion
R
Srivastava
LK
Alterations in dendritic morphology of prefrontal cortical and nucleus accumbens neurons in post-pubertal rats after neonatal excitotoxic lesions of the ventral hippocampus
Neurosci
 , 
2005
, vol. 
133
 (pg. 
463
-
470
)
Galvan
CD
Hrachovy
RA
Smith
KL
Swann
JW
Blockade of neuronal activity during hippocampal development produces a chronic focal epilepsy in the rat
J Neurosci
 , 
2000
, vol. 
20
 (pg. 
2904
-
2916
)
Ghosh
A
Carnahan
J
Greenberg
ME
Requirement for BDNF in activity-dependent survival of cortical neurons
Science
 , 
1994
, vol. 
263
 (pg. 
1618
-
1623
)
Goldin
M
Segal
M
Avignone
E
Functional plasticity triggers formation and pruning of dendritic spines in cultured hippocampal networks
J Neurosci
 , 
2001
, vol. 
21
 (pg. 
186
-
193
)
Gorba
T
Klostermann
O
Wahle
P
Development of neuronal activity and activity-dependent expression of brain-derived neurotrophic factor mRNA in organotypic cultures of rat visual cortex
Cereb Cortex
 , 
1999
, vol. 
9
 (pg. 
864
-
877
)
Gould
E
Vail
N
Wagers
M
Gross
CG
Adult-generated hippocampal and neocortical neurons in macaques have a transient existence
Proc Natl Acad Sci USA
 , 
2001
, vol. 
98
 (pg. 
10910
-
10917
)
Hardingham
GE
Bading
HU
The yin and yang of NMDA receptor signalling
Trends Neurosci
 , 
2003
, vol. 
26
 (pg. 
81
-
89
)
Harmon
KM
Wellman
CL
Differential effects of cholinergic lesions on dendritic spines in frontal cortex of young adult and aging rats
Brain Res
 , 
2003
, vol. 
992
 (pg. 
60
-
68
)
Ichisaka
S
Katoh_Semba
R
Hata
Y
Ohshima
M
Kameyama
K
Tsumoto
T
Activity-dependent change in the protein level of brain-derived neurotrophic factor but no change in other neurotrophins in the visual cortex of young and adult ferrets
Neurosci
 , 
2003
, vol. 
117
 (pg. 
361
-
371
)
Jensen
JB
Lund
TM
Timmermann
DB
Schousboe
A
Pickering
DS
Role of GluR2 expression in AMPA-induced toxicity in cultured murine cerebral cortical neurons
J Neurosci Res
 , 
2001
, vol. 
65
 (pg. 
267
-
277
)
Johnson
F
Hohmann
SE
DiStefano
PS
Bottjer
SW
Neurotrophins suppress apoptosis induced by deafferentation of an avian motor-cortical region
J Neurosci
 , 
1997
, vol. 
17
 (pg. 
2101
-
2111
)
Katz
LC
Shatz
CJ
Synaptic activity and the construction of cortical circuits
Science
 , 
1996
, vol. 
274
 (pg. 
1133
-
1138
)
Kempermann
G
Kuhn
HG
Gage
FH
More hippocampal neurons in adult mice living in an enriched environment
Nature
 , 
1997
, vol. 
386
 (pg. 
493
-
495
)
Kohara
K
Kitamura
A
Morishima
M
Tsumoto
T
Activity-dependent transfer of brain-derived neurotrophic factor to postsynaptic neurons
Science
 , 
2001
, vol. 
291
 (pg. 
2419
-
2423
)
Liu
S
Lau
L
Wei
J
Zhu
D
Zou
S
Sun
HS
Fu
Y
Liu
F
Lu
Y
Expression of Ca(2+)-permeable AMPA receptor channels primes cell death in transient forebrain ischemia
Neuron
 , 
2004
, vol. 
43
 (pg. 
43
-
55
)
Lu
YM
Yin
HZ
Chiang
J
Weiss
JH
Ca2+-permeable AMPA/Kainate and NMDA channels: high rate of Ca2+ influx underlies potent induction of injury
J Neurosci
 , 
1996
, vol. 
16
 (pg. 
5457
-
5465
)
Nakayama
K
Kiyosue
K
Taguchi
T
Diminished neuronal activity increases neuron-neuron connectivity underlying silent synapse formation and the rapid conversion of silent to functional synapses
J Neurosci
 , 
2005
, vol. 
25
 (pg. 
4040
-
4051
)
Ott
A
van Rossum
CT
van Harskamp
F
van de Mheen
H
Hofman
A
Breteler
MM
Education and the incidence of dementia in a large population-based study: the Rotterdam study
Neurology
 , 
1999
, vol. 
52
 (pg. 
663
-
666
)
Papa
M
Bundman
M
Greenberger
V
Segal
M
Morphological analysis of dendritic spine development in primary cultures of hippocampal neurons
J Neurosci
 , 
1995
, vol. 
15
 (pg. 
1
-
11
)
Papadia
S
Stevenson
P
Hardingham
NR
Bading
H
Hardingham
GE
Nuclear Ca2+ and the cAMP response element-binding protein family mediate a late phase of activity-dependent neuroprotection
J Neurosci
 , 
2005
, vol. 
25
 (pg. 
4279
-
4287
)
Pellegrini-Giampietro
DE
Gorter
JA
Bennett
MV
Zukin
RS
The GluR2 (GluR-B) hypothesis: Ca(2+)-permeable AMPA receptors in neurological disorders
Trends Neurosci
 , 
1997
, vol. 
20
 (pg. 
464
-
470
)
Petreanu
L
Alvarez-Buylla
A
Maturation and death of adult-born olfactory bulb granule neurons: role of olfaction
J Neurosci
 , 
2002
, vol. 
22
 (pg. 
6106
-
6113
)
Pivovarova
NB
Nguyen
HV
Winters
CA
Brantner
CA
Smith
CL
Andrews
SB
Excitotoxic calcium overload in a subpopulation of mitochondria triggers delayed death in hippocampal neurons
J Neurosci
 , 
2004
, vol. 
24
 (pg. 
5611
-
5622
)
Ramakers
GJ
Raadsheer
FC
Corner
MA
Ramaekers
FC
Van Leeuwen
FW
Development of neurons and glial cells in cerebral cortex, cultured in the presence or absence of bioelectric activity: morphological observations
Eur J Neurosci
 , 
1991
, vol. 
3
 (pg. 
140
-
153
)
Ruijter
JM
Baker
RE
De Jong
BM
Romijn
HJ
Chronic blockade of bioelectric activity in neonatal rat cortex grown in vitro: morphological effects
Int J Dev Neurosci
 , 
1991
, vol. 
9
 (pg. 
331
-
338
)
Scheetz
AJ
Williams
RW
Dubin
MW
Severity of ganglion cell death during early postnatal development is modulated by both neuronal activity and binocular competition
Vis Neurosci
 , 
1995
, vol. 
12
 (pg. 
605
-
610
)
Sun
W
Gould
TW
Newbern
J
Milligan
C
Choi
SY
Kim
H
Oppenheim
RW
Phosphorylation of c-Jun in avian and mammalian motoneurons in vivo during programmed cell death: an early reversible event in the apoptotic cascade
J Neurosci
 , 
2005
, vol. 
25
 (pg. 
5595
-
5603
)
Sutton
MA
Wall
NR
Aakalu
GN
Schuman
EM
Regulation of dendritic protein synthesis by miniature synaptic events
Science
 , 
2004
, vol. 
304
 (pg. 
1979
-
1983
)
Thayer
SA
Usachev
YM
Pottorf
WJ
Modulating Ca2+ clearance from neurons
Front Biosci
 , 
2002
, vol. 
7
 (pg. 
d1255
-
d1279
)
Turrigiano
GG
Leslie
KR
Desai
NS
Rutherford
LC
Nelson
SB
Activity-dependent scaling of quantal amplitude in neocortical neurons
Nature
 , 
1998
, vol. 
391
 (pg. 
892
-
896
)
Turrigiano
GG
Nelson
SB
Homeostatic plasticity in the developing nervous system
Nat Rev Neurosci
 , 
2004
, vol. 
5
 (pg. 
97
-
107
)
Tyler
WJ
Pozzo-Miller
L
Miniature synaptic transmission and BDNF modulate dendritic spine growth and form in rat CA1 neurons
J Physiol
 , 
2003
, vol. 
553
 (pg. 
497
-
509
)
West
AE
Chen
WG
Dalva
MB
Dolmetsch
RE
Kornhauser
JM
Shaywitz
AJ
Takasu
MA
Tao
X
Greenberg
ME
Calcium regulation of neuronal gene expression
Proc Natl Acad Sci USA
 , 
2001
, vol. 
98
 (pg. 
11024
-
11031
)
Wilson
RS
Bennett
DA
Bienias
JL
Aggarwal
NT
Mendes De Leon
CF
Morris
MC
Schneider
JA
Evans
DA
Cognitive activity and incident AD in a population-based sample of older persons
Neurology
 , 
2002
, vol. 
59
 (pg. 
1910
-
1914
)
Young
D
Lawlor
PA
Leone
P
Dragunow
M
During
MJ
Environmental enrichment inhibits spontaneous apoptosis, prevents seizures and is neuroprotective
Nat Med
 , 
1999
, vol. 
5
 (pg. 
448
-
453
)
Zhang
LI
Poo
MM
Electrical activity and development of neural circuits
Nat Neurosci
 , 
2001
, vol. 
4
 
Suppl
(pg. 
1207
-
1214
)
Zhang
Y
Lipton
P
Cytosolic Ca2+ changes during in vitro ischemia in rat hippocampal slices: major roles for glutamate and Na+-dependent Ca2+ release from mitochondria
J Neurosci
 , 
1999
, vol. 
19
 (pg. 
3307
-
3315
)