Uncoupling the NMDA receptor from death signals
12Channels2009; Vol. 3 Issue 1
[Channels 3:1, 12-15; January/February 2009]; ©2009 Landes Bioscience
NMDA receptors (NMDARs) mediate ischemic brain damage,
in part through interactions of the PDZ ligand of NR2 subunits
with the PDZ domain proteins PSD-95 and neuronal nitric oxide
synthase located within the NMDAR signaling complex. We have
recently shown that this PDZ ligand-dependent pathway promotes
neuronal death via p38 activation. A peptide mimetic of the NR2B
PDZ ligand (TAT-NR2B9c) reduces p38-mediated death in vitro
and p38-dependent ischemic damage in vivo. In the absence of
the PDZ ligand-p38 pathway, such as in TAT-NR2B9c-treated
neurons, or in NMDAR-expressing non-neuronal cells, NMDAR-
dependent excitotoxicity is mediated largely by JNK and requires
greater Ca2+ influx. A major reason for blocking pro-death signaling
events downstream of the NMDAR as an anti-excitotoxic strategy is
that it may spare physiological synaptic function and signaling. We
find that neuroprotective doses of TAT-NR2B9c do not alter the
frequency of spontaneous synaptic events within networks of cultured
cortical neurons nor is mini-EPSC frequency altered. Furthermore,
TAT-NR2B9c does not inhibit the capacity of synaptic NMDAR
activity to promote neuroprotective changes in gene expression,
including the upregulation of PACAP via CREB, and suppression
of the pro-oxidative FOXO target gene Txnip. Thus, while the NR2
PDZ ligand does not account for all the excitotoxic effects of excessive
NMDAR activity, these findings underline the value of the specific
targeting of death pathways downstream of the NMDAR.
It is long-established that high and prolonged levels of glutamate
kill neurons.1 During an ischemic episode, extracellular glutamate
builds up due to synaptic release and impaired/reversed uptake mech-
anisms.2 This glutamate induces excessive activation of the N-methyl
D-aspartate subclass of glutamate receptor (NMDAR) which results
in Ca2+-dependent cell death.3 The destructive effects of excessive
NMDAR activity are in no doubt, and nor is the protective effects of
NMDAR antagonists in blocking several animal models of neuronal
injury. However, NMDARs also have an important role in normal
physiology, so cannot be blocked with impunity. NMDARs are
heavily involved in synaptic transmission and synaptic plasticity, so
antagonists can have considerable CNS-adverse effects.4,5 Moreover,
modest levels of NMDAR activity can exert a neuroprotective
effect.6 In the adult CNS, NMDAR blockade exacerbates neuronal
loss when applied after traumatic brain injury and during ongoing
neurodegeneration,7 and prevents the survival of newborn neurons
in the adult dentate gyrus.8 Also, ischemic tolerance is thought to be
mediated, at least in part, by NMDAR activity.9
Neurons do not respond in a stereotypical way to Ca2+ influx: the
channel through which Ca2+ enters can also affect the response.3,10 In
the case of excitotoxicity, Ca2+ influx specifically through NMDARs
promotes cell death more efficiently than through voltage-gated Ca2+
channels.3,11 An explanation for this: the ‘source-specificity hypoth-
esis’ proposes that neuron-specific enzymes or substrates responsible
for Ca2+ dependent excitotoxicity are co-localized with NMDARs.
The cytoplasmic tail of NMDAR subunits are linked to a network
of neuronal proteins, the so-called NMDAR signaling complex
(NSC). A role for the NSC in mediating NMDAR-dependent death
was shown in the case of the PDZ proteins neuronal nitric oxide
synthase (nNOS) and PSD-95.12 PSD-95 is linked to the C-terminal
PDZ ligand of NR2, and also binds to nNOS. When the interaction
of NR2 and PSD-95 is disrupted (using TAT-NR2B9c, a cell-
permeable peptide mimetic of the NR2 PDZ ligand) the NMDAR
becomes uncoupled from nNOS activation, reducing (but not elimi-
nating) NMDAR-dependent excitotoxicity.12 The important role
of nNOS and PSD-95 above any other PDZ proteins in mediating
NMDAR-dependent excitotoxicity was recently demonstrated.13
Targeting interactions within the NSC in order to reduce excitotoxic
*Correspondence to: Giles E. Hardingham; Center for Integrative Physiology; Center
for Neuroscience Research; University of Edinburgh; Edinburgh, United Kingdom; Tel.:
+44.131.6507961; Fax: +44.131.6506527; Email: Giles.Hardingham@ed.ac.uk
Submitted: 01/02/09; Revised: 01/16/09; Accepted: 01/16/09
Previously published online as a Channels E-publication:
Addendum to: Soriano FX, Martel MA, Papadia S, Vaslin A, Baxter P, Rickman C,
Forder J, Tymianski M, Duncan R, Aarts M, Clarke P, Wyllie DJ, Hardingham GE.
Specific targeting of pro-death NMDA receptor signals with differing reliance on
the NR2B PDZ ligand. J Neurosci 2008; 28:10696–710; PMID: 18923045; DOI:
Inhibiting pro-death NMDA receptor signaling dependent on the NR2
PDZ ligand may not affect synaptic function or synaptic NMDA receptor
signaling to gene expression
Marc-Andre Martel,1,2,† Francesc X. Soriano,1,2,† Paul Baxter,1,2 Colin Rickman,1 Rory Duncan,1 David J.A. Wyllie1,2 and
Giles E. Hardingham1,2,*
1Center for Integrative Physiology; 2Center for Neuroscience Research; University of Edinburgh; Edinburgh, UK
†These authors contributed equally to this work.
Key words: NMDA receptor, PDZ domain, neuroprotection, excitotoxicity, CREB, cell death, FOXO
Uncoupling the NMDA receptor from death signals
signaling is therefore a potentially attractive proposition, since this
may be more selective than simple NMDAR blockade.
We recently showed that reducing the activation of the pro-death
p38 pathway is an important consequence of uncoupling NR2/
PSD-95 interactions with TAT-NR2B9c.14 TAT-NR2B9c is able to
achieve this without influencing NMDAR-dependent Ca2+ influx
or interfering with a model of NMDAR-dependent synaptic plas-
ticity,14 suggesting that targeting the NSC may spare some of the
normal functions of the NMDAR. Further investigations revealed
that TAT-NR2B9c does not affect either miniEPSC frequency, or
the frequency of spontaneous EPSCs (Fig. 1A and B). Thus, by these
metrics, synaptic properties are largely unaffected. Furthermore,
TAT-NR2B9c’s effect on the p38 pathway was specific to its activa-
tion by the NMDAR: activation of p38 by peroxide was unaffected
by TAT-NR2B9c (Fig. 2).
We also studied the effect of TAT-NR2B9c on the activation of
Akt by synaptic NMDAR activity, since this pathway is an impor-
tant mediator of activity-dependent neuroprotection.15 As with
previous studies, synaptic activity was enhanced by disinhibiting
the neuronal cultures by treatment with a GABAA receptor blocker,
bicuculline, which induces action potential bursting and concomi-
tant intracellular Ca2+ transients dependent on NMDAR activity
and augmented by release of Ca2+ from internal stores.15,16 We
found that Akt activation was not impeded by TAT-NR2B9c.14 One
pro-survival consequence of NMDAR-dependent Akt activation is
the nuclear export of FOXOs and the transcriptional suppression
of the pro-oxidant gene Txnip, a thioredoxin inhibitor.17,18 We
found that TAT-NR2B9c did not interfere with activity-dependent
export of FOXO1, nor suppression of Txnip expression, in contrast
to the NMDAR antagonist MK-801 (Fig. 3A and B). Activation
of CREB-mediated gene expression is another contributor to
activity-dependent neuroprotection.15,19 TAT-NR2B9c did not
interfere with activity-dependent induction of the CREB target gene
Adcyap1, which encodes the neuroprotective ligand, pituitary adeny-
late cyclase activating polypeptide (PACAP, Fig. 3C).
Whilst interfering with NMDAR-dependent p38 activation using
TAT-NR2B9c was neuroprotective, the effect was not complete:
death could still be achieved by using higher doses of NMDA.14
Other pro-death pathways remain unaffected by TAT-NR2B9c,
including one mediated by JNK.14 Targeting JNK with the cell-
permeable peptide inhibitor D-JNKI1,20 in addition to blocking
p38 signaling by TAT-NR2B9c was additively protective in in
vitro and in vivo models of excitotoxicity.14 These results indicate
that activation of JNK did not require signaling via the NR2 PDZ
ligand. Indeed, NMDAR-dependent JNK signaling appears not to
require the NSC at all. NMDAR-dependent neuronal death can be
recapitulated in non-neuronal cells simply by expression of func-
tional NMDARs.21,22 When we studied the pathway responsible for
this death in NMDAR-expressing AtT20 cells (NR-AtT20 cells),
we found the death to be JNK-dependent.14 Therefore, it appears
that the NMDAR can trigger pro-death pathways with differing
reliance on neuron-specific signaling molecules. The protective
effect of TAT-NR2B9c can be overcome in neurons by increasing
the NMDAR-dependent Ca2+ load, by exposing neurons to higher
concentrations of NMDA.14
Thus, while NMDAR-dependent cell death can be reconsti-
tuted in non-neuronal cells, the p38 route to death is absent. One
would predict therefore that the NMDAR would be less effective
at promoting death when placed outside its neuronal context, away
from neuron-specific signaling proteins. To obtain an indication
of this, we performed a series of measurements to calculate the
NMDAR-dependent Ca2+ load per unit cell volume required to kill
a neuron, compared to a NR-AtT20 cell. We first determined the45
Ca2+ load associated with a half-maximally toxic dose of glutamate
in neurons (Fig. 4A and B) and NR-AtT20 cells (see Fig. 1B and C
from ref. 14). Our comparison took into account AtT20 transfec-
tion efficiency (based on NR1 immunoreactivity). To correct for
Figure 1. TAT-NR2B9c does not alter basal miniEPSC frequency or basal
spontaneous EPSC frequency. Neurons were treated with TAT-NR2B9c
for at least 1 h and mEPSCs (A) and spontaneous EPSCs (B) recorded as
described,14 n = 5–8.
Figure 2. TAT-NR2B9c does not impair induction of p38 by peroxide.
Western analysis of p38 activation in cortical neurons stimulated with H2O2
(100 μM) for 5 min in the presence or absence of TAT-NR2B9c (1 h pretreat-
ment, n = 3). The example western shown below the graph ishows the level
of phosphorylated p38 (Pi-P38, upper) and total p38 (lower) detected under
each experimental condition.
Uncoupling the NMDA receptor from death signals
14Channels2009; Vol. 3 Issue 1
Figure 3. TAT-NR2B9c does not interfere with synaptic NMDAR-dependent pro-survival signaling to CREB or FOXO1. (A) TAT-NR2B9c does not impair
PI3K-dependent FOXO1 export. Neurons were transfected with vectors encoding eGFP and myc-tagged FOXO1. 24 h post-transfection the neurons were
pre-treated with TAT-NR2B9c (2 μM) or LY294002 (a selective inhibitor of PI3-kinase 50 μM) for 1 h prior to bicuculline stimulation (50 μM, 1 h) to stimulate
synaptic activity (reviewed in ref. 14). Cells were then fixed and processed for immunocytochemistry with an anti-myc antibody (9E10). FOXO1 subcellular
distribution was scored for around 150 cells per treatment across 3 independent experiments. Graph shows quantification of the data; example pictures
are also shown for each treatment. Immunofluorescent detection of proteins in neurons was performed as described previously.30 (B) TAT-NR2B9c does not
impair synaptic NMDAR-dependent suppression of Txnip transcription. Neurons were pre-treated with TAT-NR2B9c (2 μM) or MK-801 (10 μM) for 1 h prior
to bicuculline stimulation for 4 h. *p < 0.05 compared to unstimulated neurons (n = 5). For details of primers used see ref. 18 (C) TAT-NR2B9c does not
impair synaptic NMDAR-dependent activation of Adcyap1 transcription. Neurons were pre-treated with TAT-NR2B9c (2 μM) or MK-801 (10 μM) for 1 h
prior to bicuculline stimulation for 4 h. *p < 0.05 compared to unstimulated neurons (n = 5).
Figure 4. Increased NMDAR-mediated Ca2+ influx is required to kill NMDAR-expressing AtT20 cells than is required to kill neurons. (A) Dose-dependent
NMDAR-dependent cell death in cortical neurons. Cortical neurons were treated with glutamate for 1 h and then returned to glutamate-free medium.
Cells were fixed at 24 h, DAPI stained, and cell death counted (n = 3). Line indicates the concentration of glutamate required to achieve 50% death. (B)
Dose-dependent NMDAR-dependent Ca2+ influx. Cortical neurons were incubated in 45Ca2+-containing medium and treated with glutamate (10 min). Line
indicates the Ca2+ load associated with 50% death. (C) Example 3-D representations of eGFP-expressing AtT20 cells and neuron (cell body), obtained by
confocal microscopy and image deconvolution (Scale bar 5 μm). Also shown is a maximum intensity projection (MIP) of the confocal stack of a region of
neuronal processes (90 μm x 90 μm). D) Ca2+ entry through NMDARs is more efficient at killing cortical neurons than NR-AtT20 cells. Relative Ca2+ load
required to half-maximally kill NR-AtT20 cells vs. cortical neurons, taking into account relative cell volume. *p < 0.05, 2-tailed Student’s t-test.
Uncoupling the NMDA receptor from death signals
the differences in cell volume, which was measured by acquiring
3-D deconvolved models of eGFP-expressing cells (examples in
Fig. 4C, see Methods). AtT20 cell volume was found to be 1.84
± 0.16 pl (n = 6) and neuronal volume found to be 10.4 ± 0.9 pl
(neuron, n = 6). Based on this, NR-AtT20 cells need significantly
more NMDAR-dependent Ca2+ load to kill them than neurons
(Fig. 4D). This is consistent with the absence of NMDAR induction
of p38-dependent pro-death signaling in NR-AtT20 cells.14
To conclude, pro-death signaling by excessive NMDAR activity is
mediated by multiple pathways, including p38 and JNK, which have
differing requirements for the NSC. Using combined approaches to
block specifically individual pathways may provide optimal protec-
tion against excitotoxic insults, while sparing physiological NMDAR
signaling. At present it is not clear whether specific interactions
involving NMDAR-associated proteins are important for activation
of other mediators of excitotoxicity, such as calpains23 and TRPM7
activation.24 Moreover, it remains to be seen whether differential
coupling of the NMDAR to Ca2+ effectors of cell death underlies
the reported differences in pro-death signaling by synaptic versus
extrasynaptic NMDARs18,25-27 or NR2A- vs. NR2B-containing
Materials and Methods
All methods except those listed below were described previously.14
Cell volume measurements. Cell volume measurements were
performed using cortical neurons or AtT20 cells transfected with
peGFP-N1 and were imaged using a Zeiss LSM 510 Axiovert
confocal laser-scanning microscope. Data were acquired at Nyquist
sampling rates with voxel dimensions of 59.509 nm (X) x 59.509
nm (Y) x 170 nm (Z), using a Zeiss C-Apochromat 1.2 NA 63x
water corrected immersion objective lens. Z-stacks were acquired
optimally with intensity values falling within the dynamic range of
the detectors and extending above and below the cell to obtain all
available fluorescence. Imaging was performed on living cells, main-
tained at 37°C in 5% (v/v) CO2, 95 % (v/v) air. Image data were
deconvolved using Huygens software (Scientific Volume Imaging)
before volume calculation using Volocity (Improvision). In the
case of neurons, this technique was used to calculate the volume
of the cell body. To estimate the mean diameter of a process, the
total area contained within a network of processes, from a most
intense pixel projection of a z-stack, was divided by the total process
length calculated using ImageJ (http://rsb.info.nih.gov/ij/) and the
NeuronJ plugin.29 Using this technique, the mean process diameter
was found to be 0.74 μm, comfortably above the resolution of the
technique which is 223 nm (0.61*wavelength (512 nm)/NA(1.4)).
Moreover, the average diameter of the top 3% of widest processes was
only 0.81 μm, demonstrating that the processes had a fairly narrow
distribution of diameters, and that the average was not skewed by
outliers. The average process volume per neuron was calculated by
determining the average length of process per neuron. To calculate
this, a number of images were taken and the total process length
traced (as above). This length was then divided by the number of
cell bodies in the fields (a total of 25, within six fields) to give the
average process length per neuron. For display, representative images
are shown surface rendered using Imaris (Bitplane).
1. Olney JW. Brain lesions, obesity and other disturbances in mice treated with monosodium
glutamate. Science 1969; 164:719-21.
2. Rossi DJ, Oshima T, Attwell D. Glutamate release in severe brain ischaemia is mainly by
reversed uptake. Nature 2000; 403:316-21.
3. Arundine M, Tymianski M. Molecular mechanisms of glutamate-dependent neurodegen-
eration in ischemia and traumatic brain injury. Cell Mol Life Sci 2004; 61:657-68.
4. Ikonomidou C, Turski L. Why did NMDA receptor antagonists fail clinical trials for stroke
and traumatic brain injury? The Lancet Neurology 2002; 1:383-6.
5. Muir KW. Glutamate-based therapeutic approaches: clinical trials with NMDA antagonists.
Curr Opin Pharmacol 2006; 6:53-60.
6. Hardingham GE. Pro-survival signaling from the NMDA receptor. Biochem Soc Trans
7. Ikonomidou C, Stefovska V, Turski L. Neuronal death enhanced by N-methyl-D-aspartate
antagonists. Proc Natl Acad Sci USA 2000; 97:12885-90.
8. Tashiro A, Sandler VM, Toni N, Zhao C, Gage FH. NMDA-receptor-mediated, cell-
specific integration of new neurons in adult dentate gyrus. Nature 2006; 442:929-33.
9. Shpargel KB, Jalabi W, Jin Y, Dadabayev A, Penn MS, Trapp BD. Preconditioning para-
digms and pathways in the brain. Cleveland Clinic journal of medicine 2008; 75:77-82.
10. Soriano FX, Hardingham GE. Compartmentalized NMDA receptor signaling to survival
and death. J Physiol 2007; 584:381-7.
11. Tymianski M, Charlton MP, Carlen PL, Tator CH. Source specificity of earlly calcium neuro-
toxicity in cultures embryonic spinal neurons. Journal of Neuroscience 1993; 13:2085-104.
12. Aarts M, Liu Y, Liu L, Besshoh S, Arundine M, Gurd JW, et al. Treatment of ischemic
brain damage by perturbing NMDA receptor-PSD-95 protein interactions. Science 2002;
13. Cui H, Hayashi A, Sun HS, Belmares MP, Cobey C, Phan T, et al. PDZ protein interac-
tions underlying NMDA receptor-mediated excitotoxicity and neuroprotection by PSD-95
inhibitors. J Neurosci 2007; 27:9901-15.
14. Soriano FX, Martel MA, Papadia S, Vaslin A, Baxter P, Rickman C, et al. Specific targeting
of pro-death NMDA receptor signals with differing reliance on the NR2B PDZ ligand. J
Neurosci 2008; 28:10696-710.
15. 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; 25:4279-87.
16. Hardingham GE, Arnold FJ, Bading H. Nuclear calcium signaling controls CREB-
mediated gene expression triggered by synaptic activity. Nat Neurosci 2001; 4:261-7.
17. Soriano FX, Papadia S, Hofmann F, Hardingham NR, Bading H, Hardingham GE.
Preconditioning doses of NMDA promote neuroprotection by enhancing neuronal excit-
ability. J Neurosci 2006; 26:4509-18.
18. Papadia S, Soriano FX, Leveille F, Martel MA, Dakin KA, Hansen HH, et al. Synaptic NMDA
receptor activity boosts intrinsic antioxidant defenses. Nat Neurosci 2008; 11:476-87.
19. Lee B, Butcher GQ, Hoyt KR, Impey S, Obrietan K. Activity-Dependent Neuroprotection
and cAMP Response Element-Binding Protein (CREB): Kinase Coupling, Stimulus
Intensity and Temporal Regulation of CREB Phosphorylation at Serine 133. J Neurosci
20. Borsello T, Clarke PG, Hirt L, Vercelli A, Repici M, Schorderet DF, et al. A peptide inhibi-
tor of c-Jun N-terminal kinase protects against excitotoxicity and cerebral ischemia. Nat
Med 2003; 9:1180-6.
21. Cik M, Chazot PL, Stephenson FA. Optimal expression of cloned NMDAR1/NMDAR2A
heteromeric glutamate receptors: a biochemical characterization. Biochem J 1993;
22. Anegawa NJ, Guttmann RP, Grant ER, Anand R, Lindstrom J, Lynch DR. N-Methyl-
D-aspartate receptor mediated toxicity in nonneuronal cell lines: characterization using
fluorescent measures of cell viability and reactive oxygen species production. Brain Res Mol
Brain Res 2000; 77:163-75.
23. Bano D, Young KW, Guerin CJ, Lefeuvre R, Rothwell NJ, Naldini L, et al. Cleavage of the
plasma membrane Na+/Ca2+ exchanger in excitotoxicity. Cell 2005; 120:275-85.
24. Aarts M, Iihara K, Wei WL, Xiong ZG, Arundine M, Cerwinski W, et al. A key role for
TRPM7 channels in anoxic neuronal death. Cell 2003; 115:863-77.
25. Hardingham GE, Fukunaga Y, Bading H. Extrasynaptic NMDARs oppose synaptic
NMDARs by triggering CREB shut-off and cell death pathways. Nat Neurosci 2002;
26. Zhang SJ, Steijaert MN, Lau D, Schutz G, Delucinge-Vivier C, Descombes P, Bading H.
Decoding NMDA Receptor Signaling: Identification of Genomic Programs Specifying
Neuronal Survival and Death. Neuron 2007; 53:549-62.
27. Papadia S, Hardingham GE. The Dichotomy of NMDA Receptor Signaling. Neuroscientist
28. Liu Y, Wong TP, Aarts M, Rooyakkers A, Liu L, Lai TW, et al. NMDA receptor subunits
have differential roles in mediating excitotoxic neuronal death both in vitro and in vivo. J
Neurosci 2007; 27:2846-57.
29. Meijering E, Jacob M, Sarria JC, Steiner P, Hirling H, Unser M. Design and validation of a
tool for neurite tracing and analysis in fluorescence microscopy images. Cytometry A 2004;
30. Mckenzie GJ, Stephenson P, Ward G, Papadia S, Bading H, Chawla S, et al. Nuclear Ca2+
and CaM kinase IV specify hormonal- and Notch-responsiveness. J Neurochem 2005;