|Home | About | Journals | Submit | Contact Us | Français|
The endoplasmic reticulum (ER) is an essential cellular compartment responsible for Ca2+ sequestration, signaling, protein translation, folding as well as transport. Several acute and chronic disease conditions impair ER function leading to ER stress. To study the impact of ER stress on synaptic transmission we applied tunicamycin (TM) or thapsigargin (TG) to hippocampal neurons, which triggered sustained elevation of key ER stress markers. We monitored evoked and spontaneous neurotransmission during 4 days of TM or TG treatment and detected only a 20% increase in paired pulse facilitation suggesting an increase in neurotransmitter release probability. However, the treatments did not significantly affect the number of active synapses or the size of the total recycling vesicle pool as measured by uptake and release of styryl dye FM1-43. In contrast, under the same conditions, we observed a dramatic 4-fold increase in spontaneous excitatory transmission, which could be reversed by chronic treatment with the NMDA receptor blocker AP-5 or by treatment with salubrinal, a selective inhibitor of eukaryotic translation initiation factor 2 (eIF2α) dephosphorylation. Furthermore, ER stress caused NMDA receptor dependent suppression of eukaryotic elongation factor-2 (eEF2) phosphorylation thus reversing downstream signaling mediated by spontaneous release. Taken together, these findings suggest that chronic ER stress augments spontaneous excitatory neurotransmission and reverses its downstream signaling in a NMDA receptor dependent manner, which may contribute to neuronal circuitry abnormalities that precede synapse degeneration in several neurological disorders.
The endoplasmic reticulum (ER) is a continuous system of membranous cisterns and tubules present in all eukaryotic cells including neuronal processes and dendritic spines (Martone et al., 1993; Spacek and Harris 1997; Terasaki et al., 1994). The ER is an essential cellular compartment responsible for calcium sequestration and signaling as well as protein translation, folding and transport (Paschen and Fradsen, 2001). The ER lumen contains the highest Ca2+ concentration within the cell, which is maintained by Ca2+ ATPases of the SERCA (Sarco/Endoplasmic Reticulum Ca2+ -ATPase) family. ER possesses several Ca2+ dependent molecular chaperons such as BiP, Grp94, calnexin and calreticulin (Lee, 1992; Little et al., 1994) and numerous protein disulfide isomerases (Braakman et al., 1991) providing critical environment for the formation of disulfide bonds and correct protein folding. Improperly folded proteins undergo ER-associated protein degradation through the proteosomal pathway (Friedlander et al., 2000; Werner et al., 1996).
Conditions such as ischemia, disruption of Ca2+ homeostasis, nitrosative and oxidative stress, glucose or nutrient deprivation, viral infections can interfere with ER function and lead to ER stress (Kaufman et al., 2002; Doroudgar et al., 2009; Dimcheff et al., 2004; He et al., 1997; Lipton, 2007). Under ER stress unfolded proteins are accumulated within the ER lumen and trigger an adaptive response known as the ER stress response or unfolded protein response which acts to restore normal ER function by bringing back the protein-folding capacity of the ER. However, if the ER stress persists and cellular homeostasis cannot be restored, the ER stress response can lead to apoptosis. ER stress-mediated apoptosis is triggered by up-regulation of glucose-regulated protein 78 (GRP78) or CHOP (C/EBP homologous protein)/GADD153 (growth arrest- and DNA damage-inducible gene 153) expression leading to caspase-12 activation (Barone et al., 1994; Lin et al., 2007; Marciniak et al., 2004; Matsumoto et al., 1996; Nakagawa et al., 2000).
Several neurological and neuropsychiatric disorders that include Alzheimer’s disease, Parkinson’s disease, polyglutamine expansion diseases, Amyotrophic Lateral Sclerosis as well as Tuberous Sclerosis involve ER stress and unfolded protein response which may underlie their pathophysiology (Scheper and Hoozemans, 2009; Di Nardo et al., 2009). These disease states have also been associated with synaptic deficits, which precede synapse degeneration and neuronal loss (Hartley et al., 1999; Kamenetz et al., 2003; Nikolaus et al., 2009; Tang et al., 2003; Ting et al., 2007). ER stress response has also been linked to potential alterations in synaptic function, because modulation of ATF4/CREB transcriptional pathway, which plays a key role in ER stress response, regulates synaptic plasticity as well as learning and memory (Costa-Mattioli et al., 2005).
Despite the evidence linking several brain disorders to ER stress and synaptic deficits, very little is known about the direct impact of stress conditions on synaptic transmission. Here, to study the impact of ER stress on synaptic transmission in hippocampal neurons, we employed electrophysiological and optical imaging techniques after ER stress induction by tunicamycin (TM) or thapsigargin (TG) application (Chang and Korolev, 1996; Gething and Sambrook, 1999; Kaneko and Tsukamoto, 1994; Perez-Sala and Mollinedo, 1995).
Dissociated hippocampal cultures were prepared as previously described (Kavalali et al., 1999). Briefly, whole hippocampi were dissected from postnatal day 0–3 (P0–3) Sprague-Dawley rats. Tissue was trypsinized (10 mg/ml trypsin) for 10 min at 37°C mechanically dissociated by pipetting and plated on Matrigel or Poly-lysin coated coverslips. Cytosine arabinoside (4 μM ARAC, Sigma, St. Louis, MO) was added at day 1 in vitro (DIV), at 4 DIV ARAC concentration was reduced to 2 μM. All experiments were performed on 14–21 DIV cultures. All experiments were performed on at least three independent cultures.
DL-2-amino-5-phosphonopentanoate (AP5, Sigma), 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, Sigma), tetrodotoxin (TTX, Calbiochem), picrotoxin (PTX, Sigma) were added to solutions as indicated.
To induce ER stress neurons were treated in conditioned media with 5 μM of tunicamycin (TM, Sigma) or 300 nM thapsigargin (TG, Sigma). The drugs were applied either alone or with 50 μM AP5 or 1 μM TTX. TTX was applied 12 h and AP5 at least 2 h prior to treatment with TM or TG.
Neurons were incubated with salubrinal (5 μM, Tocris) 72 h prior to application of tunicamycin or thapsigargin.
Neurons were incubated in Ca2+ -free Tyrode solution either with 100 μM EGTA-AM for 15 min, or with 10 μM BAPTA-AM for 20 min, then washed with Tyrode solution containing 1 μM TTX, 50 μM AP5 and 50 μM of PTX before mEPSC recordings.
Cultured hippocampal neurons were treated with 50 μM NMDA for 30 min, washed with PBS once and placed back in conditioned growth medium. mEPSCs were recorded 2 h after treatment. Samples for western blot analysis were collected at 2h and 24 h after treatment.
The number of apoptotic cells was assessed with Tunel staining (Promega) according to the manufacturer’s protocol (Supplemental Figure 1).
Hippocampal neurons were grown on poly-lysine covered coverslips for all immunocytochemistry experiments. Cells were washed once with phosphate buffered saline (PBS) and fixed in 4% (w/v) paraformaldehyde (Sigma) in PBS at RT for 15 min. After 4x rinse with PBS, cells were incubated for 10 min at RT in 1% (w/v) glycine in PBS to reduce fixative-induced autofluorescence. Next, cells were permeabilized with 0.2% (v/v) Triton X-100 (Roche, Roche Diagnostics, Indianapolis, IN) in PBS for 10 min at RT. After incubation for 1 hour at RT in NGS blocking solution (10% (v/v) normal goat serum (Sigma), 5% (w/v) bovine serum albumin (Sigma) and 0.025% (v/v) Triton X-100 in PBS) cells were incubated for 1 hour at RT with primary antibody rabbit anti-CHOP (1:200) and mouse anti-MAP2 (1:200). After incubation with primary antibody cells were washed 3 times with PBS-T for 10 min each at RT. Cells were then incubated with goat anti-mouse secondary antibodies conjugated to Alexa Fluor 488 (1:1000, Invitrogen) and goat anti-rabbit secondary antibodies conjugated to Alexa Fluor 594 (1:1000, Invitrogen) for 1 hour at RT in the dark. Following this incubation cells were washed 3 times with PBS-T for 10 min each then coverslips were mounted using Aqua Poly/Mount (Polysciences, Warrington, PA) and allowed to cure overnight at RT in the dark before imaging. Images were obtained by a Zeiss LSM 510 META laser-scanning microscope equipped with LSM 510 Laser module (Carl Zeiss, Jena, Germany) and analyzed with ImageJ software.
Samples of treated neurons were collected with 1x loading buffer containing: 62.5 mM Tris-HCl (pH 6.8 at 25°C), 2% w/v SDS, 10% glycerol, 50 mM DTT, 0.01% w/v bromophenol blue. Samples were sonicated for 30 sec, and boiled for 5 min at 95°C, 20 μl of each sample were then loaded and separated on 8% SDS PAGE, then transferred to nitrocellulose membranes. After blocking for 60 min at room temperature in blocking solution (Tris-buffered saline (TBS), 0.1% Tween-20, 5% w/v nonfat dry milk) blots were washed tree times with TBS-T and incubated overnight at 4°C with the following primary antibodies: anti-phospho eIF2a (1:1000), anti-eIF2α (1:1000), anti-phospho eEF2 (1:1000), anti-eEF2 (1:1000) (all rabbit polyclonal antibodies from Cell Signaling Technology), mouse monoclonal anti–CHOP (1:1000, Affinity Bioreagents, Golden, CO), mouse monoclonal anti-GDI (1:15000) (gift of Dr. T. Südhof). Primary antibody dilution buffer contained: TBS, 0.1% Tween-20 with 5% BSA. After washing three times with TBS-T, blots were incubated with HRP-conjugated anti-rabbit (1:2000, Cell Signaling Technology) or anti-mouse (1:2500, Chemicon) secondary antibodies. Immunoreactive bands were visualized by enhanced chemiluminescence, captured on autoradiography film (Eastman Kodak, Rochester, NY). Digital images were produced by densitometric scans of autoradiographs on ScanJet 4300C (Hewlett Packard, Palo Alto, CA) and quantified using ImageJ software. The amount of total eEF2 and eIF2α proteins were adjusted to GDI loading control. The phospho/total and CHOP/GDI protein ratios were calculated for each condition and presented as a percentage of condition control.
Whole-cell patch-clamp recordings were performed on hippocampal pyramidal neurons. Data were acquired using a MultiClamp 700B amplifier and Clampex 9.0 software (Molecular Devices, Sunnyvale, CA). Recordings were filtered at 2 kHz and sampled at 200 μs. A modified Tyrode’s solution containing (in mM): 150 NaCl, 4 KCl, 2 MgCl2, 2CaCl2, 10 glucose, 10 Hepes, pH 7.4, was used as external bath solution. For isolation of miniature EPSCs (mEPSCs) 1 μM of TTX), 50 μM of PTX and 50 μM AP5 were added. For isolation of miniature IPSCs (mIPSCs) 1 μM of TTX, 50 μM AP5 and 10 μM CNQX were added. Evoked EPSCs were recorded in presence of 50 μM of PTX. The pipette internal solution contained (in mM): 115 Cs-MeSO3, 10 CsCl, 5 NaCl, 10 HEPES, 0.6 EGTA, 20 Tetraethylammonium-Cl, 4 Mg-ATP, 0.3 Na3GTP, pH 7.35, and 10 QX-314 [N-(2,6-dimethylphenylcarbamoylmethyl)-triethylammonium bromide], 300 mOsm. Field stimulation was applied through parallel bipolar electrodes (FHC, St Bowdoin, ME) immersed in the perfusion chamber, delivering 15–20 mA pulses. Series resistance ranged between 10–30 MΩ.
Synaptic boutons were loaded with FM1-43 during 90 s incubation in Tyrode solution containing 47 mM K+. After washing with dye-free Tyrode solution for 10 min, synaptic terminals were destained using three 90 s applications of a 90 mM K+ Tyrode solution with each separated by 60 s intervals. All staining and washing solutions contained 10 μM CNQX and 50 μM AP5 to prevent recurrent activity. Isolated boutons were selected during the wash and fluorescence changes were measured during destaining. Images were obtained by a cooled, intensified CCD camera (Roper Scientific, Duluth, GA) during illumination at 480 nm via an optical switch (Sutter Instruments, Novato, CA). Images were acquired and analyzed using imaging software (Molecular Devices).
For each neuron FM1-43 puncta were counted at least on three dendrites. These numbers were averaged to represent the synapse density for a given neuron. Measurements were obtained from at least 10 neurons from three independent cultures for a given condition.
Fura-2 Ca2+ imaging experiments with neuronal cultures were performed as previously described (Tang et al., 2003) using a DeltaRAM illuminator, an IC-300camera, and ImageMaster Pro software (all from PTI). Briefly, the cultured neurons were maintained in artificial CSF (ACSF) (140 mM NaCl, 5mM KCl, 1 mM MgCl2, 2 mM CaCl2, and 10 mM HEPES, pH 7.3) at 37°C during measurements (PH1 heater, Warner Instruments). Fura-2 340/380 ratio images for baseline (2 min) measurements were collected every 6 s for the duration of the experiment.
Paired t- test or one-way ANOVA with Tukey post-hoc test were used to determine statistical significance. The mean difference was taken as significant at p < 0.05.
To study the impact of ER stress on synaptic transmission, we applied tunicamycin (TM) and thapsigargin (TG) to cultured hippocampal neurons. Here, dissociated primary cultures provide a distinct advantage by allowing examination of synaptic function independent of potential general alterations in brain homeostasis, thereby enabling a distinction between cell-autonomous defects and global systemic as well as metabolic dysfunction. In order to monitor the progression of ER stress we examined alterations in the levels of phosphorylated eukaryotic translation initiation factor 2 α (p-eIF2α) and CHOP proteins. Alpha subunit of the translation initiation factor 2 (eIF2α) is phosphorylated by PKR-like kinase (PERK) in response to ER stress (Harding et al., 1999; Hinnebusch, 2000). CHOP protein triggers ER stress-mediated apoptosis. Thus, both, p-eIF2α and CHOP protein indicate the presence of ER stress. Treatment of hippocampal cultures with 5 μM TM or 300 nM TG caused a rapid elevation in the levels of these markers within 4 hours (Fig 1A–F). Continued TM or TG treatment sustained this increase up to 96 hours. In order to visualize the distribution of ER stress across neurons in the culture, we co-immunostained hippocampal cultures with an antibody against the neuronal marker MAP-2 and CHOP after 24 hour of TM or TG application (Fig. 1G). Quantitative analysis of immunofluorescence in neurons identified with their MAP-2 staining revealed a nearly homogeneous increase in the levels of CHOP staining across the culture (Fig. 1H), indicating that under these conditions nearly all neurons were experiencing ER stress. As a consequence of the wide spread ER stress, neurons gradually transitioned into apoptosis. However, between 48 to 72 hours after ER stress induction, only a small fraction of cells displayed apoptotic markers. When we quantified apoptotic neurons using Tunel staining, we found that the number of apoptotic cells increased with the time of drug application and comprised on average 10–20% of control neurons at 72 hours after application of TM or TG respectively (Supplemental Fig. 1).
Next, we assessed the impact of this global ER stress induction paradigm on the functional integrity of individual presynaptic nerve terminals by monitoring activity-dependent uptake and release of the styryl dye FM1-43 (Betz et al., 1996). Elevated potassium stimulation (47 mM K+ for 90 sec) triggers near maximal uptake of FM1-43 into the total recycling synaptic vesicle pool (Harata et al., 2001) and subsequent application of 90 mM K+ stimulation causes swift dye loss due to rapid mobilization and fusion of these dye labeled vesicles. The amount of dye uptake and the kinetics of dye release monitored during this protocol are extremely susceptible to differences in maturational states of presynaptic terminals (Mozhayeva et al., 2002) or potential defects in their neurotransmitter release machinery (Bronk et al., 2007; Schoch et al., 2001).
Surprisingly, sustained treatment with TM or TG up to 72 hours did not induce a significant change in the number of functional presynaptic terminals detected using this protocol along dendritic processes (Fig. 2A–D). Moreover, 72 hours of TM or TG treatment failed to trigger a significant alteration in the kinetics of dye release during elevated potassium depolarization (Fig. 2E–H) despite its ability to induce maximal wide spread ER stress across neurons (Fig. 1).
Our findings so far suggest that under these conditions ER stress induction does not compromise gross functional integrity of neurons or their presynaptic nerve terminals. In the next set of experiments, we assessed whether the same ER stress conditions elicited alterations in the properties of excitatory synaptic transmission using whole-cell recordings of evoked synaptic responses. Excitatory synaptic responses triggered during sustained 10 Hz stimulation showed robust depression and swift recovery under control as well as TG or TM treatment conditions up to 72 hours with similar kinetics under all conditions (Fig. 3A–B). In contrast, 20 Hz stimulation induced more rapid depression in TM or TG treated neurons compared to controls, although under both conditions subsequent low frequency stimulation produced a swift response recovery back to baseline levels with similar kinetics (Fig. 3C–D). This increased depression could be attributed to a small but significant elevation in neurotransmitter release probability, as incubation with either TM or TG induced a decrease in paired pulse facilitation of excitatory postsynaptic currents within 48 hours of treatment (Fig. 3E, F).
We also evaluated whether the ER stress paradigm altered the balance between AMPA receptor-mediated and NMDA receptor-mediated evoked excitatory neurotransmission, a known hallmark of postsynaptic maturation in addition to being a key substrate for long-term synaptic plasticity (Poncer and Malinow, 2001; Chubykin et al., 2007). The ratios of evoked AMPA and NMDA-receptor mediated excitatory postsynaptic currents detected in individual neurons were not significantly different during the course of 72 hours of TM or TG treatment (Fig. 3G, H). Taken together, these results suggest that chronic ER stress induction causes a mild but significant elevation in neurotransmitter release probability coupled with increased synaptic depression during high frequency stimulation while it leaves several other key markers of evoked excitatory neurotransmission intact.
In the next set of experiments we investigated whether chronic ER stress alters the properties of spontaneous excitatory neurotransmission. Changes in spontaneous neurotransmission could underlie functional neuronal deficits associated with several neurological as well as neuropsychiatric disorders that are coupled to ER stress (Lipton, 2007; Rossi et al., 2000; Zhang et al., 2009). Interestingly, treatment with TM or TG resulted in a robust two- to four-fold increase in the frequency of miniature excitatory postsynaptic currents (mEPSCs) within 48 hours (Fig. 4A–D), and in the case of TG this increase in mEPSC frequency was maintained up to 72 hours (Fig. 4B, D). The elevation in spontaneous synaptic activity was not mediated by increases in overall network activity as the miniature events were recorded in the presence of tetrodoxin (TTX) and AP-5 to block action potential generation and NMDA receptor activation. In contrast, inclusion of TTX or AP-5 together with TM or TG during ER stress induction abolished this increase in mEPSC frequency (Fig. 4C–D) without significantly changing the amplitudes of mEPSCs (Fig. 4E, F).
In the next set of experiments we tested whether NMDA receptor activation alone was sufficient to trigger ER stress and augment spontaneous neurotransmission in the absence of TM or TG treatment. Indeed, 30 min long treatment with 50 μM NMDA triggered robust ER stress as early as 2 hours after treatment as indicated by the elevation in p-eIF2α and CHOP protein levels. This elevation in key ER stress markers was sustained up to 24 hours following brief NMDA application (Fig. 5A). Unlike TM or TG treatment, however, acute NMDA mediated excitotoxicity triggered rapid neuronal death after 24 hours. Therefore, we restricted our electrophysiological analysis to the 2-hour time point after initial NMDA treatment. Under these conditions, we observed approximately two-fold increase in mEPSC frequency (Control: 2.16±0.26 Hz; NMDA: 4.46±1.09 Hz; p<0.05) (Fig. 5B) without a significant change in the amplitudes of individual events (Control: 23.74± 1.26 pA; NMDA: 30.0±3.6 pA; p>0.05). This increase in mEPSC frequency after brief NMDA exposure agrees well with earlier results (e.g. Malgaroli and Tsien; 1992) and strongly suggests that pathophysiological events that produce ER stress such as excitotoxicity (Sokka et al., 2007) or ischemia (Tajiri et al., 2004; Hayashi et al., 2003) could mimic the effect of TM or TG treatment and lead to a similar increase in spontaneous excitatory neurotransmission.
Earlier studies have shown that acute treatment of neurons with agents such as TG could significantly elevate the rate of spontaneous neurotransmission by increasing cytosolic Ca2+ levels (Xu et al., 2009). To address whether the increased mEPSC frequency after chronic treatment with TM or TG was due to elevated cytosolic Ca2+, or an increase in Ca2+ sensitivity of the neurotransmitter release machinery, we recorded mEPSCs after acute application of the cell permeable Ca2+ buffer EGTA-AM or BAPTA-AM (Fig. 6A–D). Under control conditions EGTA-AM or BAPTA-AM incubation resulted in a two-fold decrease in mEPSC frequency suggesting that baseline mEPSC frequency is facilitated by resting cytosolic Ca2+ levels (Fig. 6E–H). Likewise, the increased mEPSC frequency detected after 48-hour incubation with TM or TG was also extremely susceptible to EGTA-AM or BAPTA-AM treatment, producing a two- to four-fold decrease in mEPSC frequency in the case of EGTA-AM, ten- to twenty five-fold decrease in the case of BAPTA-AM (Fig. 6E–H) under both conditions.
Next, we asked whether this ER stress induced increase in mEPSC frequency is due to a sustained increase in cytosolic Ca2+. For this purpose, we used the cell permeable ratiometric Ca2+ indicator dye FURA-2AM to estimate differences in basal Ca2+ levels. We loaded control as well as TM or TG neurons treated with FURA2-AM and monitored FURA-2 fluorescence emission in response to ultraviolet excitation at 340 or 380 nm. The ratio of fluorescence emission after excitation at these two wavelengths is an indicator of free Ca2+ levels (Grynkiewicz et al., 1985). These experiments were conducted 48 hours after TM or TG treatments under conditions similar to our mEPSC recordings. In this setting, we did not detect a significant change in resting cytosolic Ca2+ levels (Fig. 6I). This finding suggests that the augmentation of mEPSC frequency during ER stress is mainly due to increased sensitivity of spontaneous release machinery to Ca2+ rather than a significant increase in basal Ca2+ levels. This premise is also consistent with the observation that the increase in mEPSC frequency takes 48 hours to develop after induction of ER stress indicating a long-term modification of the release machinery rather than a sustained increase in Ca2+. A sustained increase in Ca2+ would have been expected to have dramatic impact on several other synaptic parameters including evoked neurotransmission as well as miniature inhibitory postsynaptic currents, which were all largely unaltered (Supplemental Fig. 2).
To test whether there is a causal link between activation of ER stress response and augmentation of excitatory spontaneous neurotransmission, we tested the impact of salubrinal on ER stress induced increase in neurotransmission. Salubrinal inhibits ER stress activated phosphatase complexes and stabilizes phosphorylated eIF2α. Recent work had identified salubrinal in a molecule screen against ER stress induced apoptosis following tunicamycin treatment in a rat pheochromocytoma (PC12) cell line (Boyce et al., 2005). In these experiments we incubated neurons with salubrinal (5 μM) 72 hours prior to application of TM or TG. mEPSC recordings performed 48 hours after TM (Fig. 7A) or TG (Fig. 7B) exposure revealed significant block of the augmentation of mEPSC frequency induced by ER stress in response to salubrinal treatment (Fig. 7C). The same conditions did not alter the amplitudes of individual mEPSCs (Fig. 7D). This result indicates that decrease in eIF2α phosphorylation and possibly ER-stress dependent apoptosis initiate the observed increases in spontaneous neurotransmission. This premise is in line with earlier work which had suggested a role for apoptotic signaling cascade in regulation of neurotransmission (Reimertz et al., 2003; Li et al., 2008). However, it is important to note that in the absence of salubrinal, the fraction of apoptotic cells did not exceed 20% even after 48 hours of TM or TG treatment (Supplementary Figure 1) although nearly all cells experienced ER stress (Fig. 1). Therefore, the augmentation of mEPSC is more likely to be a direct consequence of eIF2α dephosphorylation or possibly ER stress induced phosphatase activities rather than concomitant apoptosis.
Recent studies suggest that spontaneous neurotransmitter release, rather than evoked release, specifically suppresses dendritic protein translation machinery by promoting phosphorylation and inactivation of eukaryotic elongation factor-2 (eEF2), a critical catalytic factor for ribosomal translocation during protein synthesis (Sutton et al., 2007). This chronic suppression of protein translation, in turn, stabilizes postsynaptic sensitivity to released neurotransmitters by maintaining subunit composition glutamate receptors (Sutton et al., 2004; Sutton et al., 2006).
In the next set of experiments, we examined whether the sustained increase in spontaneous release seen after chronic ER stress could augment eEF2 phosphorylation. To monitor alterations in eEF2 phosphorylation we used antibodies specific to phosphorylated (Thr 56) and total eEF2 and performed Western blot analysis after blockade of action potentials with TTX as well as after blockade of spontaneous NMDA mEPSCs (TTX+AP-5) (Fig. 8A). Under control conditions, action potential blockade caused a slight but statistically insignificant increase in phospho-eEF2 (p-eEF2). In contrast, in agreement with earlier observations, blockade of NMDA mEPSCs significantly decreased levels of p-eEF2 (Sutton et al., 2007). Surprisingly, under ER stress (48 hours long TM or TG treatment) p-eEF2 levels were significantly lower compared to controls. Application of TTX with TM or TG maintained the low levels of p-eEF2, whereas application of TM (but not TG) together with TTX and AP5 (to suppress NMDA-mEPSCs) resulted in a significant increase in p-eEF2 towards control levels (Fig. 8B). These results suggest that under ER stress regulation of eEF2 phosphorylation by spontaneous release events operates in reverse aiming to maintain a relatively dephosphorylated eEF2 thus potentially leading to facilitation of protein translation machinery. Given the overall attenuation of protein translation during ER stress (Harding et al., 1999), the positive regulation of translational machinery by spontaneous release events may ensure maintenance of homeostatic regulation and synapse stability under the stress conditions.
In this study, we chronically treated cultured hippocampal neurons with tunicamycin (TM) or thapsigargin (TG), two agents that compromise ER homeostasis via different mechanisms. TM blocks the synthesis of all N-linked glycoproteins causing unfolded protein response leading to ER stress, whereas, TG is an inhibitor of ER Ca2+-ATPase that blocks Ca2+ sequestration within the ER lumen leading to depletion of ER Ca2+ stores. Irrespective of their mechanisms of action, chronic treatment with these compounds caused a sustained increase in ER stress markers and spontaneous excitatory neurotransmission without substantial changes in resting Ca2+ levels suggesting a causal link between ER stress response and regulation of neurotransmission. This link is further supported by the finding that the augmentation of excitatory spontaneous neurotransmission during ER stress was susceptible to treatment with salubrinal, a selective inhibitor of eukaryotic translation initiation factor 2 (eIF2α) dephosphorylation, a key component of the ER stress response.
Surprisingly, these chronic ER stress conditions could be maintained up to 72 hours without significant apoptosis as indicated by Tunel staining or in the absence of degeneration in the overall integrity of presynaptic function, synapse numbers or postsynaptic responsiveness when assessed with a combination of morphological and functional markers. More detailed examination of the properties of synaptic transmission revealed a mild increase (~20%) in paired pulse depression of evoked excitatory postsynaptic currents coupled with facilitation of synaptic depression detected during sustained stimulation at 20 Hz. In contrast to this modest alteration in evoked excitatory transmission, chronic ER stress conditions caused a robust increase in mEPSC frequency, which could be countered by increased Ca2+ buffering by acute EGTA-AM or BAPTA-AM incubation. In the absence of significant changes in baseline Ca2+ levels, this result suggests an increase in Ca2+ sensitivity of spontaneous neurotransmitter release machinery, akin to observations after knock out of synaptic vesicle associated Ca2+ sensor synaptotagmin-1 (Xu et al., 2009). However, under ER stress the increase in spontaneous release was specific to excitatory synapses as the frequency of mIPSCs were largely unaltered under the same conditions.
Substantial increases in the rate of spontaneous release can have deleterious consequences on neuronal function. Increased release under resting conditions can augment glutamatergic tone and render neurons more vulnerable to excitotoxicity (Nakanishi et al., 2009; Cavelier and Attwell, 2005, Wasser and Kavalali, 2009). In addition, given spontaneous release events’ impact on neuronal activity (Otmakov et al., 1993; Carter and Regehr, 2002), uncontrolled augmentation of this form of release may adversely impact neuronal information processing by altering spike timing and fidelity. Recent studies also suggest that in contrast to evoked release spontaneous neurotransmitter release events tonically suppress protein translation machinery by maintaining the key regulator eEF2 in a phosphorylated state (Sutton et al., 2007). These observations are supported by recent findings from our group which suggest that spontaneous release events activate a population of NMDA receptors relatively isolated from receptors activated in response to evoked release (Atasoy et al., 2008). Furthermore, under physiological resting membrane potentials and Mg2+ levels spontaneous release events can trigger significant NMDA receptor dependent activation and signaling due to incomplete Mg2+ block (Espinosa and Kavalali, 2009). In this study, we also evaluated downstream signaling mediated by spontaneous release events during ER stress and found that the regulation of eEF2 phosphorylation by mEPSCs was altered. In agreement with earlier work, under control conditions blockade of resting NMDA receptor activity caused a significant decrease in eEF2 phosphorylation (Sutton et al., 2007). In contrast, ER stress conditions caused a significant decrease in baseline levels of eEF2 phosphorylation despite a concomitant increase in the rate of mEPSCs. Interestingly, blockade of resting NMDA receptor activation during ER stress resulted in restoration of p-eEF2 levels towards the baseline. However, this reversal was specific to TM induced ER stress, whereas blocking NMDA receptor activity did not alter the decrease in p-eEF2 levels seen after TG treatment. This discrepancy may suggest that the ability of NMDA-mEPSC suppression to prevent the effects of ER stress on eEF2 phosphorylation may require Ca2+ release from ER stores.
These results suggest that in central neurons ER stress reverses the direction of downstream signaling mediated by spontaneous release events. This finding is in contrast to a recent study which had shown that in non-neuronal cell lines ER stress itself induces eEF2 kinase-dependent eEF2 phosphorylation implicating a direct impact of ER stress on eEF2 function in the absence of glutamatergic signaling (Boyce et al., 2008). However, it remains to be determined whether this reversal in mEPSC-mediated signaling (i.e. the decrease in p-eEF2 levels) serves to preserve a significant dynamic range for homeostatic plasticity during stress, or alternatively it represents a defect in synaptic signaling potentially contributing to neurodegenerative effects of long-term stress. It is important to note that in the ER stress model we used here, long term blockade of NMDA receptors did not significantly alter expression of ER stress markers or rescue neuronal loss seen after 96 hours of continued stress. This observation favors the former premise that resting NMDA receptor activity and its downstream signaling acts to maintain homeostasis rather than facilitating degenerative decline.
Taken together our findings suggest that chronic treatment of hippocampal neurons with TM or TG, unlike acute NMDA mediated excitotoxicity, creates an in vitro condition that mimics chronic neuronal stress without substantial neuronal death or synaptic disintegration. This condition resembles several clinical observations that accompany neurological disorders, which suggest synaptic transmission and signaling abnormalities precede synapse degeneration and neuron loss (Hartley et al., 1999; Kamenetz et al., 2003). Strikingly, a major consequence of chronic stress on synaptic transmission is slow elevation of spontaneous excitatory transmission in 48 hours. The facilitation of spontaneous release we observed here may be triggered by the ER stress induced transcriptional program altering Ca2+ sensitivity of release. Interestingly, this effect of ER stress on excitatory neurotransmission is not shared by inhibitory spontaneous neurotransmission leading to an imbalance between inhibitory and excitatory input at rest. Such an imbalance between excitation and inhibition has been suggested as possible underlying cellular factor for certain neurodevelopmental disorders (Rubenstein and Merzenich, 2003; Dani et al., 2005). Resilience of spontaneous inhibitory neurotransmission to ER stress may arise from inherently increased Ca2+ buffering capacity of inhibitory interneurons (Lee et al., 2000).
Importantly, the impact on ER stress on spontaneous glutamatergic signaling could be alleviated by NMDA receptor blockade. This finding may explain therapeutic effectiveness of NMDA receptor blockers in several neurological and neuropsychiatric conditions that are accompanied with chronic neuronal stress and imply NMDA receptor signaling mediated by spontaneous glutamate release as potential therapeutic target.
We would like to thank Drs Ilya Bezprozvanny, Lisa Monteggia and Michael Morris for advice, discussions and comments on the manuscript. This work was supported by grants from the National Institute of Mental Health to E.T.K. E.T.K. is an Established Investigator of the American Heart Association.