Neurobiology of Disease 62 (2014) 533–542

Contents lists available at ScienceDirect

Neurobiology of Disease

j ourna l homepage: www.e lsev ie r .com/ locate /ynbd i

Chronic blockade of extrasynaptic NMDA receptors ameliorates synapticdysfunction and pro-death signaling in Huntington diseasetransgenic mice

Alejandro Dau, Clare M. Gladding, Marja D. Sepers, Lynn A. Raymond ⁎Department of Psychiatry, Division of Neuroscience, Brain Research Centre, University of British Columbia, Vancouver, British Columbia V6T 1Z3, Canada

⁎ Corresponding author at: Department of Psychiatry,4N3-2255 Wesbrook Mall, Vancouver, BC V6T 1Z3, Canad

E-mail addresses: adau@alumni.ubc.ca (A. Dau), cgladmarjasepers@gmail.com (M.D. Sepers), lynn.raymond@ub

Available online on ScienceDirect (www.sciencedir

0969-9961/$ – see front matter © 2013 Elsevier Inc. All rihttp://dx.doi.org/10.1016/j.nbd.2013.11.013

a b s t r a c t

a r t i c l e i n f o

Article history:Received 5 August 2013Revised 17 October 2013Accepted 12 November 2013Available online 19 November 2013

Keywords:Huntington diseaseNMDA receptorGluN2BExtrasynapticp38pCREBCalpainCalcium signalingStriatumMemantine

In the YAC128 mouse model of Huntington disease (HD), elevated extrasynaptic NMDA receptor (Ex-NMDAR)expression contributes to the onset of striatal dysfunction and atrophy. A shift in the balance of synaptic–extrasynaptic NMDAR signaling and localization is paralleled by early stage dysregulation of intracellular calciumsignaling pathways, including calpain and p38 MAPK activation, that couple to pro-death cascades. However,whether aberrant calcium signaling is a consequence of elevated Ex-NMDAR expression in HD is unknown.Here, we aimed to identify calcium-dependent pathways downstream of Ex-NMDARs in HD. Chronic (2-month) treatment of YAC128 and WT mice with memantine (1 and 10 mg/kg/day), which at a low dose selec-tively blocks Ex-NMDARs, reduced striatal Ex-NMDAR expression and current in 4-month oldYAC128micewith-out altering synaptic NMDAR levels. In contrast, calpain activity was not affected by memantine treatment, andwas elevated in untreated YAC128 mice at 1.5 months but not 4 months of age. In YAC128 mice, memantineat 1 mg/kg/day rescued CREB shut-off, while both doses suppressed p38MAPK activation toWT levels. Taken to-gether, our results indicate that Ex-NMDAR activity perpetuates increased extrasynaptic NMDAR expression anddrives dysregulated p38 MAPK and CREB signaling in YAC128 mice. Elucidation of the pathways downstream ofEx-NMDARs in HD could help provide novel therapeutic targets for this disease.

© 2013 Elsevier Inc. All rights reserved.

Introduction

In Huntington disease (HD), progressive neurodegeneration is at-tributed to a polyglutamine (polyQ) expansion near the N-terminusof the protein huntingtin (mutant huntingtin, mtHtt) (Huntington'sDisease Collaborative Research Group, 1993). Whereas wildtypehuntingtin is vital for normal cellular function, mtHtt interfereswith essential intracellular processes, including gene expression,Ca2+ homeostasis and vesicular trafficking (Zuccato et al., 2010).The polyQ expansion primarily affects GABAergic medium-sizedspiny projection neurons (SPNs) of the striatum, which exhibit upto 95% neuronal loss at late stages of the disease (Vonsattel et al.,1985). However, the mechanisms underlying selective vulnerabilityof the striatum to mtHtt-induced death remain unclear.

Increased functional expression of N-methyl-D-aspartate glutamatereceptors (NMDARs) in HD could contribute to selective striatalexcitotoxicity (Cepeda et al., 2001; Chen et al., 1999; Fan et al.,2007; Li et al., 2003; Milnerwood et al., 2010; Starling et al., 2005;

University of British Columbia,a.ding@cdrd.ca (C.M. Gladding),c.ca (L.A. Raymond).ect.com).

ghts reserved.

Zeron et al., 2001, 2002; Zhang et al., 2008). Previously, we reportedelevated expression of GluN2B-containing extrasynaptic NMDARs(Ex-NMDARs) in presymptomatic YAC128 mice (Milnerwood et al.,2010), in which the yeast artificial chromosome is used to expressfull-length human huntingtin with 128 CAG repeats (Slow et al., 2003).Synaptic NMDARs activate pro-survival pathways, while Ex-NMDARstrigger cell death (Hardingham and Bading, 2010; Hardingham et al.,2002). A shift in the balance of synaptic to Ex-NMDAR signaling contrib-utes to HD pathology, as chronic Ex-NMDAR blockade attenuates mtHtt-induced striatal atrophy and motor learning deficits in YAC128 mice(Milnerwood et al., 2010; Okamoto et al., 2009).

Alongwith elevated Ex-NMDAR activity, intracellular Ca2+ signalingpathways that couple to survival or death are also dysregulated early inHD. Activity of the Ca2+-dependent protease calpain is elevated instriatal tissue of post-mortem HD human brains and presymptomatic1–2 month old YAC128 mice (Cowan et al., 2008; Gafni and Ellerby,2002; Gladding et al., 2012). Calpain potentiates HD-associated striataldegeneration by cleaving mtHtt into toxic fragments and triggeringpro-apoptotic cascades in parallel with caspases (Gafni et al., 2004;Kim et al., 2001). Calpain also contributes to Ex-NMDAR surfacemislocalization in 1–2 month old YAC128 mice by cleaving theGluN2B C-terminus and thus altering NMDAR surface stability(Gladding et al., 2012). Activity of the p38mitogen-activated protein ki-nase (MAPK), shown previously to be downstream of Ex-NMDARs (Xu

534 A. Dau et al. / Neurobiology of Disease 62 (2014) 533–542

et al., 2009), is also elevated in YAC128 mouse striatum at 1–2 monthsof age (Fan et al., 2012). The p38 MAPK mediates enhanced NMDA-induced toxicity in YAC128 cultured striatal neurons (Fan et al., 2012)and thus contributes to susceptibility of mtHtt-expressing striatal neu-rons to excitotoxic death. In addition, activity of the pro-survival tran-scription factor cAMP response element binding protein (CREB) isreduced in striatal tissue of 1 and 4 month-old YAC128 mice(Milnerwood et al., 2010). While synaptic NMDAR signaling pro-motes CREB activity, Ex-NMDARs trigger dephosphorylation andinactivation of CREB via dominant pathways (Hardingham et al.,2002). Additionally, CREB signaling is restored by chronic suppressionof Ex-NMDAR activity in YAC128 mice (Milnerwood et al., 2010), sug-gesting a link between Ex-NMDARs and CREB shut-off.

Together, elevated Ex-NMDAR activity and dysregulated intracellu-lar signaling could contribute to mtHtt-induced striatal degeneration.However, whether aberrant Ca2+ signaling in HD is a direct resultof enhanced Ex-NMDAR activity or a consequence of other effectsof mtHtt remains unclear. Here, we examined the role of Ex-NMDARs in HD-associated aberrant Ca2+ signaling.

Methods

Memantine treatment

WT and YAC128 (line 55)mice (Slow et al., 2003) bred on the FVB/Nbackground were maintained at the University of British Columbia(UBC) Faculty of Medicine Animal Resource Unit, according to guide-lines of the Canadian Council on Animal Care. Animals were housed inidentical conditions (2–4 mice/cage) in a 12-h light/dark cycle, withfull access to food and water. WT and YAC128 mice were treated withmemantine as described previously (Milnerwood et al., 2010;Okamoto et al., 2009). Briefly, memantine at 1 or 10 mg/kg/daywas provided to mice ad libitum in their drinking bottles, startingat 2 months of age (+10 days) for 2 months. Control mice receivedwater (vehicle). Memantine solution concentrations were adjustedon a semi-monthly basis according to mouse body weight anddaily solution intake to ensure consistent dosing throughout thetreatment period. Cohorts alternated between male and femalemice (a total of 4 female cohorts and 6 male cohorts). No differencesin daily solution intake were detected between treatment groupscompared to control groups for either genotype or sex (data notshown); thus, the mice had no apparent aversion to the taste ofmemantine.

Striatal dissection, subcellular and nuclear fractionations

Striatal tissue from memantine-treated (1 and 10 mg/kg/day)and untreated WT and YAC128 mice was collected after the treat-ment period and paired on the day of dissection. Mice were decapi-tated following halothane vapour anesthesia. Brains were rapidlyremoved and striatal sections were dissected and homogenized in200 μL ice-cold sucrose buffer (0.32 M sucrose, 10 mM HEPES,pH 7.4). Striatal cytosolic, synaptosomal, and nuclear fractionswere ob-tained by subcellular and nuclear fractionation as described previously(Milnerwood et al., 2010). Synaptic (postsynaptic density, PSD) andextrasynaptic-enriched (non-PSD) synaptosomal fractions were isolat-ed based on the principle that the non-PSD is Triton-X-soluble, whereasthe PSD is Triton-X-insoluble. All buffers contained ‘complete’ proteaseand phosphatase inhibitor cocktails (Roche), as well as 15 μMcalpeptin(Calbiochem), 1 mM EDTA, 1 mM EGTA, 40 mM β-glycerophosphate,20 mM sodium pyrophosphate, 1 mM sodium orthovanadate, and30 mM sodium fluoride. The purity of the subcellular fractionationwas confirmed by enrichment of the PSD marker PSD-95 in the PSDfraction, and the presynaptic marker synaptophysin in the non-PSD(not shown). The purity of the nuclear fractionation was confirmed byenrichment of the nuclear marker histone deacetylase (HDAC) in the

nuclear matrix fraction (not shown). Fractions were stored at −80 °Cuntil use.

Western blotting

Protein concentration was assessed by a BCA protein assay (Pierce).Freshly-thawed samples were prepared for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) by heating (3–5 min,80–85 °C), in 3X protein sample buffer (PSB) (6% SDS, 0.4 mM Tris(pH 6.8), 30% glycerol, pyronin Y, 70 mg/mL DTT). Equal amountsof protein (5–15 μg for non-PSD, PSD, nuclear matrix fractions, or20–40 μg for cytosolic fractions) were separated in 10% (w/v) SDS-polyacrylamide gels, and transferred to polyvinylidene fluoride(PVDF) membranes by semi-dry electrophoresis (BioRad). Mem-branes were blocked in TBS with 0.5% Tween-20 (TBST) and 3% BSA(or 5% milk for spectrin blots) (1 h, RT), incubated in primary anti-bodies (overnight, 4 °C), then washed and incubated in horseradishperoxidase (HRP)-conjugated secondary antibodies (2 h, RT). Blotswere then washed and visualized using an enhanced chemilumines-cence substrate (ECL, Amersham) and developed by exposure to film(Amersham), except for pCREBSer13 and CREB blots which were devel-oped using an automated ChemiDoc XRS Molecular Imager (BioRad).Blots for which total p38 was probed were subsequently reprobed toquantify phosphorylated p38 (P-p38), using alkaline phosphatase(AP)-conjugated secondary antibodies and a Lumi-phosWB Chemilu-minescent Substrate detection system (Pierce).

The following primary antibodies were used: rabbit N-terminalanti-GluN2B (AGC-003; Alomone, 1:500), rabbit anti-spectrin (cleaved)(AB38, gift from Dr. David Lynch, University of Pennsylvania,Philadelphia, PA, 1:2000), rabbit anti- pCREBSer133 (06-519, Millipore,1:500), rabbit anti-CREB (9197, Cell Signaling, 1:500), rabbit anti-P-p38MAPK (4511S, Cell Signaling, 1:200), mouse anti-p38 MAPK (sc-7972,Santa Cruz, 1:200), mouse anti-PSD-95 (MA1-045, Pierce, 1:500), goatanti-β-actin (sc-1616, Santa Cruz, 1:1500), goat anti-α-tubulin (sc-9935, Santa Cruz, 1:1500), goat anti-HDAC (sc-6268, Santa Cruz, 1:500),and mouse anti-synaptophysin (S5768, Sigma, 1:1000). All primary anti-bodies were diluted in TBST with 3% BSA, except for anti-spectrin, whichwas diluted in TBST with 5% milk. The following secondary antibodieswere used: anti-mouse HRP-conjugated (NA931V, Amersham, 1:5000),anti-rabbit HRP-conjugated (NA934V, Amersham; 1:5000), anti-rabbitAP-conjugated (S372, Promega, 1:5000), and anti-goat HRP-conjugated(sc-2020; Santa Cruz, 1:5000). All secondary antibodies were diluted inTBST with 1% BSA.

Image analysis

For blots developed using film, the optical density of bands wasquantified using Image J software (NIH) after background subtraction.Band intensities of GluN2B in synaptosomal fractions were normalizedto β-actin (loading control), whereas bands for cytosolic spectrin werenormalized to α-tubulin, which gave a clearer signal in the cytosolicfraction. P-p38 bands were normalized to p38 bands probed on thesame membrane. For quantification of pCREBSer133/ CREB ratios, CREBand pCREBSer133 levels were probed on separate gels, and eachwas nor-malized to HDAC (loading control), as probing for CREB and pCREBSer133

on the same membrane did not provide clear results. pCREBSer133 andCREB blots, which were developed using the ChemiDoc XRS MolecularImager, were quantified using Image-Lab Analysis Software (4.1,BioRad).

Brain slice preparation

Ex-NMDAR currents were recorded from coronal brain slicesmade from memantine-treated (1 mg/kg/day) and untreated WT andYAC128 mice. Mice were halothane-anesthetized and decapitated.Brains were immersed in ice-cold oxygenated (95% O2, 5% CO2)

535A. Dau et al. / Neurobiology of Disease 62 (2014) 533–542

low-calcium artificial cerebrospinal fluid (aCSF), containing (in mM):125 NaCl, 2.5 KCl, 25 NaHCO3, 1.25 NaH2PO4, 2.5 MgCl2, 0.5 CaCl2, 25glucose, pH 7.3–7.4, 300–310 mosmol L−1. 300 μm thick coronal sliceswere cut on a vibratome (Leica VT1000), placed in a holding chamberwith continuously oxygenated standard aCSF (as above, but with1 mM MgCl2 and 2 mM CaCl2) at 37 °C for 45mins-1 h, then at RTuntil time of recording.

Slice electrophysiology

Slice electrophysiological recordings were conducted as describedpreviously (Milnerwood and Raymond, 2007; Milnerwood et al.,2010). Following incubation in a holding chamber, slices were per-fused continuously (1.0–1.5 ml/min, RT) with oxygenated standardaCSF containing 10 μMglycine, 2 μMstrychnine, and 100 μMpicrotoxin(Tocris), and allowed to equilibrate for 15–20 min prior to recording.EPSCs were evoked (eEPSCs) in the center of the striatum (150 μs,25–150 μA, every 20 s) using a glass micropipette (2–5MΩ) filledwith aCSF. NMDAR currents were recorded at a holding potential of+40mV from a randomly-selected SPN 150–250μm ventral to the siteof stimulation, with a micropipette filled with (in mM): 130 cesiummethanesulfonate, 5 CsCl, 4 NaCl, 1 MgCl2, 5 EGTA, 10 HEPES, 5QX-314, 0.5 GTP, 10 Na2-phosphocreatine and 5 Mg ATP, pH 7.3,280–290 mosmol/L. All recordings were made in 2,3-dihydroxy-6-nitro-7-sulfamoyl-benzo[f]quinoxaline-2,3-dione (NBQX, 10 μM, Tocris)to block AMPA receptors. After a stable baseline was established(N5 min), DL-threo-β-benzyloxyaspartic acid (DL-TBOA, 30 μM,Tocris) was bath-applied to block glutamate transporters and to in-duce extrasynaptic glutamate spillover. NMDAR-mediated eEPSCswere monitored in the presence of TBOA for 5 min, then TBOA waswashed off for 25 min. Signals were filtered at 1 kHz, digitized at10 kHz and analyzed in Clampfit 10 (Axon Instruments). The areaunder the curve for current in the first 3 s of each NMDAR-eEPSCwas normalized to peak current amplitude for that trace. Pipette resis-tance was 3–7 MΩ. Series resistance (Rs) was b25 MΩ and uncompen-sated, and monitored throughout the experiment; tolerance for ΔRswas b50% provided Rs b 30MΩ.

Statistical analysis

Statistical analyses were conducted using Prism 6 software(GraphPad). Data are presented as mean + SEM. All analyses wereperformed using a two-way ANOVA. For each treatment condition,significant differences between genotypes or age groups were testedby Bonferroni's multiple comparisons post-hoc test. For each genotype,significant differences between treatment dose (1 or 10 mg/kg/day)and the control (H2O) were examined by Dunnett's multiple compari-sons post-hoc test, unless indicated otherwise. Dunnett's post-hoc testwas selected for the latter analyses as it primarily compares differencesbetween a set of test groups and a control. Overall significant effects ofinteraction, treatment, genotype or age are indicated in the text.

Results

Memantine decreases Ex-NMDAR expression in YAC128 mice

Elevated expression of GluN2B-containing Ex-NMDARs contributesto enhanced susceptibility of striatal neurons to excitotoxic deathin YAC128mice (Milnerwood et al., 2010, 2012). However, themech-anisms underlying NMDARmislocalization in HD remain unclear. Previ-ous studies indicate that Ex-NMDAR currents activate signalingpathways, including calpains and protein phosphatases/kinases, thatcan in turn modulate synaptic vs. extrasynaptic targeting of GluN2B-containing receptors (Gladding and Raymond, 2011; Gladding et al.,2012; Xu et al., 2009). To determine whether Ex-NMDAR activity regu-lates GluN2B-NMDAR localization, 2 month-old WT and YAC128 mice

were treated with memantine (1 and 10 mg/kg/day) for two months,as described previously (Milnerwood et al., 2010; Okamoto et al.,2009). The 1mg/kg/day dose is estimated to result in an effective con-centration at the NMDAR channel mouth of ~5–10 μM,which selective-ly blocks Ex-NMDARs, relatively sparing synaptic receptors in culturedneurons, and closely mimics the therapeutic dose of 10–20 mg/day inhumans (Okamoto et al., 2009; Raina et al., 2008; Xia et al., 2010). Weadditionally selected the higher 10 mg/kg/day dose as it has beenshown to be neuroprotective in several neuroinflammatory disorders(Rammes et al., 2008; Rosi et al., 2006).

Aftermaintainingmicewith eithermemantine in the drinkingwaterorwater alone (untreated control), we isolated striatal tissue and exam-ined extrasynaptic GluN2B subunit levels by subcellular fractionationand western blotting. The subcellular fractionation separates synaptic(postsynaptic density, PSD) from extrasynaptic-enriched (non-PSD)membranes, and thus allowed us to directly examine protein expressionin each synaptosomal compartment (Gladding et al., 2012; Goebel-Goody et al., 2009; Pacchioni et al., 2009). We used an N-terminalGluN2B antibody that detects both calpain-cleaved (115kDa) andfull-length (180kDa) subunits. There was an enrichment of calpain-cleaved relative to full-length GluN2B in the non-PSD fraction in allgroups examined, as reported previously (Gladding et al., 2012)(Fig. 1A). Total (full-length plus cleaved) GluN2B levels were signif-icantly increased in the non-PSD fraction of untreated YAC128 com-pared to WT mice (p b 0.01) (GluN2B/β-actin ratios for untreatedmice: WT, 1.51 + 0.12; YAC128, 2.04 + .11) (Fig. 1B). Interestingly,YAC128 extrasynaptic GluN2B expression was significantly reduced toWT levels bymemantine at both doses (p b 0.01) (GluN2B/β-actin ratiosfor treated YAC128 mice: 1 mg/kg/day, 1.61 + 0.13; 10 mg/kg/day,1.56 + 0.l7). This trend was also observed when full-length andcalpain-cleaved GluN2B bands were quantified separately (Fig. 1C,D).However, significant differences were only detected in full-lengthGluN2B levels between memantine-treated (1 mg/kg/day) and untreat-ed YAC128 mice (p b 0.05), as separate quantification of full-length andcalpain-cleaved GluN2B bands yielded higher variability. Moreover,when normalized to total receptor levels, calpain-cleaved GluNB wasnot different between genotypes or treatments, suggesting that themtHtt- and memantine-induced changes in Ex-NMDAR localizationwere not associated with changes in the extent of C-terminal GluN2Bcleavage (calpain-cleaved/total GluN2B levels: WTH2O, 0.55 + .04;YAC128H2O, 0.53 + .05; WT1 mg/kg/day, 0.53 + .05; YAC1281 mg/kg/day,0.56 + .04;WT10 mg/kg/day, 0.56 + .05; YAC12810 mg/kg/day, 0.60 + 0.04).

To more definitively determine the effect of the 2-monthmemantine treatment on functional Ex-NMDAR expression, we exam-ined striatal spiny projection neurons (SPNs) from untreated andmemantine-treated (1 mg/kg/day) WT and YAC128 mice by whole-cell electrophysiology in acute cortico-striatal slices. As brain sliceswere incubated in recording solution lacking memantine for at leastan hour prior to recording, any residual memantine should have beenremoved from the tissue. We evoked NMDAR-mediated EPSCs beforeand during application of the glutamate transporter inhibitor DL-TBOA(30 μM), to induce glutamate spillover and activate extrasynaptic re-ceptors (Tzingounis and Wadiche, 2007) as described previously(Milnerwood et al., 2010). The effect of TBOA on NMDAR charge-transfer reflects activation of Ex-NMDARs (Milnerwood et al., 2010).In TBOA, NMDAR peak-normalized charge increased (Fig. 1E), as previ-ously observed (Milnerwood et al., 2010), and this effect wasmore pro-nounced for untreated YAC128 mice compared to other groups. In fact,maximal TBOA effects were significantly greater in untreated YAC128mice compared to WT (p b 0.05) (Fig. 1F), consistent with increasedEx-NMDAR levels (maximal TBOA increase over baseline for untreatedmice: WT, 23.8 + 6.0%; YAC128, 46.1 + 5.7%). Moreover, YAC128maximal TBOA effects were significantly reduced by 1mg/kg/daymemantine treatment (p b 0.05), to levels similar to WT, in agreementwith our biochemical analysis of non-PSDGluN2B expression (maximalTBOA increase for memantine-treated mice: WT, 29.29 + 9.2%;

Fig. 1. Memantine decreases functional Ex-NMDAR expression in YAC128 mice. WT and YAC128 mice were treated with water (H2O) only or memantine (1 and 10 mg/kg/day) in thedrinking water for 2 months, beginning at 2 months of age. A) Representative blots of non-PSD fractions probed for full-length (180kDa) and calpain-cleaved (115 kDa) GluN2B and β-actin (loading control), in memantine-treated (1 and 10 mg/kg/day) and untreated WT and YAC128 mice. Calpain-cleaved GluN2B bands were highly enriched relative to full-lengthGluN2B bands in the non-PSD; thus higher (top panel) and lower (bottom panel) exposures were used to quantify full-length and cleaved GluN2B levels, respectively. B–D) Quantificationof total (cleaved plus full-length) (B), full-length (C), and calpain-cleaved (D) GluN2B subunit levels normalized to β-actin. Analyzed by a two-way ANOVA (**p b 0.01,Bonferroni's post-hoc test; #p b 0.05, ## p b 0.01, Dunnett's post-hoc test). The number of independent experiments is indicated inside each bar. B) The interaction betweengroups was significant (F(2,55) = 5.462, p b 0.01). E) Representative peak-normalized NMDAR-mediated EPSCs in SPNs of acute coronal brain slices before (gray) and during (black)bath application ofDL-TBOA (30 μM), for untreatedWT (i) and YAC128 (ii) and 1 mg/kg/daymemantine-treatedWT (iii) and YAC128 (iv)mice. F) Quantification ofmaximal TBOA effectson peak-normalized NMDAR charge as a percentage of baseline (*p b 0.05; two-way ANOVA, Bonferroni's post-hoc test). WTH20, n = 14; YAC128H2O, n = 17; WT1 mg/kg/day, n = 9;YAC1281 mg/kg/day, n = 12; 5–7 mice per group. The interaction between groups was significant (F(1,48) = 4.984, p b 0.05).

536 A. Dau et al. / Neurobiology of Disease 62 (2014) 533–542

YAC128, 22.9 + 4.6%). Together, these results suggest that chronicblockade of Ex-NMDARs with memantine corrects the elevated func-tional expression of Ex-NMDARs in striatal neurons from YAC128mice to WT levels.

Synaptic NMDAR levels are unaffected by mtHtt or memantine

The mtHtt- and memantine-induced changes in Ex-NMDAR expres-sion could be mediated by shifts in synaptic-extrasynaptic localization

of surface receptors. To observe whether synaptic NMDAR levelsare inversely correlated with Ex-NMDAR expression, we examinedsynaptic GluN2B expression by western blotting of the striatal PSDfraction from memantine-treated (1 and 10 mg/kg/day) and un-treated WT and YAC128 mice. In contrast to the non-PSD, full-length GluN2B levels were enriched compared to calpain-cleavedsubunits in the PSD of all groups tested, as reported previously(Gladding et al., 2012) (Fig. 2A). However, no average changeswere detected in total (full-length plus cleaved) synaptic GluN2B

Fig. 2. Synaptic NMDAR expression is not altered between genotypes or treatments. PSDfractions of memantine-treated (1 and 10 mg/kg/day) and untreated WT and YAC128mice were isolated from striatal tissue and probed for synaptic GluN2B subunit levels.A) Representative blots of PSD fractions probed for full-length and calpain-cleavedGluN2B and β-actin (loading control). An enrichment of full-length GluN2B was detectedin the PSD subcellular fraction. B) Quantification of total (cleaved plus full-length)GluN2B/β-actin ratios. Analyzed by a two-way ANOVA. The number of independent ex-periments is indicated inside each bar.

537A. Dau et al. / Neurobiology of Disease 62 (2014) 533–542

expression between YAC128 and WT mice with or without memantinetreatment (Fig. 2B) (total GluN2B/β-actin ratios: WTH2O, 1.64 + .11;YAC128H2O, 1.82 + .09; WT1 mg/kg/day, 1.69 + 0.10; YAC1281 mg/kg/day,1.61 + 0.21; WT10 mg/kg/day, 1.84 + 0.21; YAC12810 mg/kg/day, 1.79 +0.15). Moreover, when quantified separately, full-length and cleavedGluN2B bands were not different between genotypes or treatments (notshown).

Striatal calpain activity is not different between 4 month-old YAC128 andWT mice, and is unaffected by memantine

The modulation of YAC128 Ex-NMDAR expression by memantinesuggests that receptor mislocalization in HD is Ex-NMDAR-activitydependent. The protease calpain is activated by Ca2+ influx throughEx-NMDARs (Xu et al., 2009), and elevated calpain activity contributesto Ex-NMDAR mislocalization in presymptomatic (1–2 month-old)YAC128mice (Gladding et al., 2012). Hence, calpain signaling is a puta-tive Ex-NMDAR activity-dependent pathway driving HD-associatedreceptor mislocalization. To determine whether calpain contributesto elevated Ex-NMDAR expression in 4 month-old YAC128 mice,we examined calpain activity in the cytosolic fraction of striatal tis-sue from memantine-treated and untreated WT and YAC128 mice.We quantified cleavage levels of the calpain substrate spectrin witha neo-epitope antibody specific for its calpain cleavage product(150 kDa) (a gift from Dr. David Lynch, University of Pennsylvania,Philadelphia, PA), as used previously (Cowan et al., 2008). Interestingly,the level of calpain-cleaved spectrin was not different between 4month-old untreated WT and YAC128 mice, and was not significantlyaltered by memantine in either genotype (cleaved spectrin/α-tubulin

ratios: WTH2O, 0.82 + 0.06; YAC128H2O, 0.88 + 0.08; WT1 mg/kg/day,0.99 + 0.07; YAC1281 mg/kg/day, 0.89 + 0.09; WT10 mg/kg/day, 0.74 +0.07; YAC12810 mg/kg/day, 0.83 + 0.09; p N 0.05) (Fig. 3A,B).

We previously observed elevated calpain activity in untreated1–2 month-old YAC128 mice (Cowan et al., 2008; Gladding et al.,2012); this apparent discrepancy with the lack of difference between4 month-old YAC128 and WT mice could be due to differences in theages of mice examined in each study. To confirm that calpain activityis elevated in 1–2 month-old but not 4 month-old YAC128 mice, wecompared spectrin cleavage in untreated animals at both ages.Calpain-cleaved spectrin levels were significantly elevated in YAC128compared to WT mice at 1.5 months (p b 0.05) but not 4 months ofage (cleaved spectrin/α-tubulin ratios at 1.5 months: WT, 0.98 +0.05; YAC128, 1.31 + 0.10; at 4 months: WT, 0.87 + 0.13; YAC128,0.82 + 0.11) (Fig. 3C,D). In fact, YAC128 calpain-cleaved spectrin levelswere significantly decreased at 4 months compared to 1.5 months(p b 0.01).

Low-dose memantine rescues mtHtt-induced CREB shut-off

While synaptic NMDARs promote CREB activation by stimulatingpathways that result in phosphorylation of the active site Ser133 residue(pCREBSer133), Ex-NMDARs trigger dominant CREB shut-off pathways(Hardingham and Bading, 2010; Hardingham et al., 2002; Papadia andHardingham, 2007). In a previous study, suppression of Ex-NMDAR activ-ity with low-dose memantine (1 mg/kg/day, 2 months) fully rescuedmtHtt-associatedCREB shut-off in 4month-old YAC128mice toWT levels(Milnerwood et al., 2010). However, at higher doses memantine alsoblocks synaptic receptors (Okamoto et al., 2009) and could thereby sup-press synaptic NMDAR-mediated CREB phosphorylation. We thus aimedto confirm that low-dose (1 mg/kg/day)memantine fully restores striatalCREB activity as previously reported, and to investigate whether thehigher dose (10 mg/kg/day) rescues or further suppresses CREBSer133

phosphorylation. We compared nuclear pCREBSer133 relative tototal CREB levels between untreated and memantine-treated WTand YAC128 mice. In agreement with previous data (Milnerwoodet al., 2010), untreated YAC128 pCREBSer133/CREB ratios were signifi-cantly decreased compared to untreated WT mice at 4 months of age(p b 0.05) and were rescued to WT levels by memantine treatment at1 mg/kg/day (p b 0.05) (Fig. 4A,B). At the 10 mg/kg/day dose, however,YAC128 pCREBSer133/CREB levels remained reduced. Additionally, therewas a trend towards a decrease in WT pCREBSer133/CREB levels ina dose-dependent manner with memantine treatment, althoughthis did not reach significance (pCREBSer133/CREB ratios: WTH2O,1.00 + 0.05; YAC128H2O, 0.70 + 0.05; WT1mg/kg/day, 0.85 + 0.09;YAC1281 mg/kg/day, 0.99 + 0.12; WT10 mg/kg/day, 0.69 + 0.15;YAC1281 mg/kg/day, 0.69 +0.08). Expression levels of HDAC also ap-peared to vary between conditions in the blot shown in Fig. 4A;however, this variability was not consistent between experiments.Moreover, average HDAC levels were not significantly different betweengenotypes or treatments (average HDAC band intensity (×106): WTH2O,1.83 + 0.21; YAC128H2O, 1.70 + 0.28; WT1mg/kg/day, 2.06 + 0.34;YAC1281 mg/kg/day, 1.71 + 0.15; WT10 mg/kg/day, 2.15 + 0.34;YAC1281 mg/kg/day, 1.92 +0.25).

Memantine reduces elevated p38 MAPK activity in YAC128 mice

Ex-NMDARs activate p38MAPK (Xu et al., 2009),which is associatedwith neuronal death (Soriano et al., 2008). Additionally, p38 MAPKactivity is elevated in 1–2 month-old YAC128 mice, and contributesto NMDA-induced toxicity in mtHtt-expressing SPNs (Fan et al.,2012). We thus examined whether YAC128 p38 MAPK activity isalso elevated at 4 months of age, and whether Ex-NMDAR activitydrives mtHtt-induced p38 MAPK activation in vivo. As a measure ofp38 MAPK activity, phosphorylation levels of p38 MAPK at Thr180and Tyr182 (herein P-p38) were quantified in cytosolic fractions

Fig. 3. Calpain activity is unaffected bymemantine treatment, and is elevated in 1.5 month- but not 4month-old YAC128mice. Striatal cytosolic fractionswere probed for calpain-cleavedspectrin (150kDa). A) Representative blots of calpain-cleaved spectrin and α-tubulin (loading control) in memantine-treated (1 and 10 mg/kg/day) and untreated 4 month-oldWT andYAC128 mice. B) Quantification of calpain-cleaved spectrin/α-tubulin ratios. C) Representative blots for calpain-cleaved spectrin and α-tubulin in untreated 1.5 and 4 month-old mice.D)Quantification of calpain-cleaved spectrin/α-tubulin ratios between 1.5 and 4months. Analyzedby a two-wayANOVA (*p b 0.05, **p b 0.01, Bonferroni'spost-hoc test). Themain effectof age was significant (F(1,21) = 8.664, p b 0.01). B, D) The number of independent experiments is indicated inside each bar.

538 A. Dau et al. / Neurobiology of Disease 62 (2014) 533–542

frommemantine-treated and untreatedWTand YAC128mice. The ratioof P-p38/p38 was significantly elevated in untreated 4 month-oldYAC128 compared to WT mice (p b 0.05) (Fig. 5A,B), as previouslyreported in 1–2 month-old mice (Fan et al., 2012). Notably,YAC128 p38 MAPK activity was significantly attenuated to WT levelsby memantine treatment at both doses (p b 0.01). (P-p38/p38 ra-tios: WTH2O, 1.03 + 0.04; YAC128H2O, 1.39 + 0.12; WT1 mg/kg/day,0.94 + 0.10; YAC1281 mg/kg/day, 0.88 + 0.10; WT10 mg/kg/day,1.04 + 0.15; YAC12810 mg/kg/day, 0.92 + 0.10).

Discussion

Identification of the pathways downstream of Ex-NMDARs mayhelp to develop new targets for therapy in HD. Elevated Ex-NMDAR activity occurs early in the YAC128 mouse model of HDand contributes to mtHtt-induced striatal pathology (Milnerwoodet al., 2010; Okamoto et al., 2009). Ca2+ dependent pathways close-ly linked to cell death are also dysregulated prior to phenotype onsetand contribute to neuronal toxicity in HD (Cowan et al., 2008; Fanet al., 2012; Gafni and Ellerby, 2002; Gladding et al., 2012;Milnerwood et al., 2010; Wu et al., 2011; Zhang et al., 2008). Here,we aimed to demonstrate a causal link between Ex-NMDAR activityand aberrant intracellular Ca2+ signaling in HD. Somewhat surpris-ingly, we found that chronic suppression of Ex-NMDAR activity withmemantine reduced striatal extrasynaptic GluN2B subunit expres-sion as well as Ex-NMDAR current in YAC128 mice to WT levelswithout altering synaptic GluN2B-NMDAR expression. These chang-es were not associated with calpain signaling, which was unaffectedby memantine treatment, and was only elevated in YAC128 mice at1.5 months but not 4 months of age. Finally, memantine treatmentrescued YAC128 CREB shut-off at the lower but not higher dose,while both doses decreased mtHtt-induced p38 MAPK activation toWT levels.

Considerations for in vivo memantine treatment

Ex-NMDAR blockade with low-dosememantine is a valuable tool todiscriminate pathways mediated by Ex-NMDAR activity from thosewhich are a direct result of mtHtt-induced Ca2+ dyshomeostasis.Prevous dose-finding pilot studies have determined that memantineat 1 mg/kg/day in mice results in an approximate concentration of5–10 μM at the NMDAR channel mouth (Okamoto et al., 2009; Xiaet al., 2010). To confirm the selectivity of this concentration for Ex-NMDAR blockade, Okamoto et al. (2009) compared the effect of low-dose memantine on extrasynpatic vs. synaptic NMDAR-mediatedEPSCs in cultured cortical neurons. They isolated Ex-NMDAR currentsusing MK-801 to block synaptic NMDAR activity, followed by bath ap-plication of NMDA to activate extrasynaptic receptors, as done previ-ously (Hardingham et al., 2002). Bath application of memantine at 1–10 μM blocked N85% of isolated Ex-NMDAR-mediated evoked EPSCs.Moreover, memantine at the same concentration had no effect onNMDAR-mediated sEPSCs in a separate experiment. Together, theseexperiments suggest that memantine preferentially blocks extrasynapticover synaptic NMDARs. Using a similar experimental approach, Xia et al.(2010) compared the effect of memantine on isolated extrasynaptic andsynaptic NMDAR-mediated EPSCs. At physiologicalMg2+ concentrations,bath application of low-dose memantine (10μM) blocked significantlymore extrasynaptic than synaptic NMDAR-mediated currents in thesame neuron (1.5-fold higher blockade of extrasynaptic over synapticNMDARs).

Notably, the selectivity of memantine for Ex-NMDAR blockade ishighly dose-dependent: at higher concentrations (30 μM) memantineblocks similar proportions of synaptic and extrasynpatic NMDARcurrents in culture (Okamoto et al., 2009). Moreover, treatmentof YAC128 mice with high-dose memantine (30 mg/kg/day) for10 months exacerbates mtHtt-associated neuropathology and behav-ioral deficits at 12 months of age (Okamoto et al., 2009). Althoughless is known about the 10 mg/kg/day dose used in our study, our

Fig. 4. Low-dosememantine rescues mtHtt-induced CREB shut-off. Nuclear fractions fromstriatal tissuewereprobed for pCREBSer133 and total CREBprotein levels. A)Representativeblots for pCREB Ser133 and total CREB in memantine-treated (1 and 10 mg/kg/d) and un-treatedWT andYAC128mice. pCREBSer133 and total CREB proteinwere probed in separategels and normalized to HDAC (loading control). The top band was analyzed to quantifypCREB Ser133 levels. B) Quantification of pCREBSer133/CREB levels. Analyzed by a two-wayANOVA (*p b 0.05, Bonferroni's post-hoc test; #p b 0.05, Dunnett's post-hoc test). Thenumber of independent experiments is indicated inside each bar.

Fig. 5.Memantine reduces elevated p38MAPK activity in YAC128mice. Cytosolic fractionsisolated from striatal tissuewere probed for P-p38 and total p38 levels, as ameasure of rel-ative p38 MAPK activity. A) Representative blots for P-p38 and p38 levels in memantine-treated (1 and 10 mg/kg/day) and untreatedWT and YAC128mice. After probing for totalp38 protein, blots were reprobed for P-p38 using an antibody directed against p38 phos-phorylated at Thr180 and Tyr182. P-p38 levelswere normalized to total p38 expression ineach lane. B) Quantification of P-p38/p38 ratios. Analyzed by a two-way ANOVA(*p b 0.05, Bonferroni's post-hoc test; ##p b 0.01, Dunnett's post-hoc test). Themain effectof treatment was significant (F(2,46) = 4.085, p b 0.05). The number of independent ex-periments is indicated inside each bar.

539A. Dau et al. / Neurobiology of Disease 62 (2014) 533–542

observation that memantine at this higher dose did not rescueYAC128 CREB activity could indicate that this dose also blocks asubstantial proportion of synaptic NMDARs. Future studies will berequired to determine the selectivity of memantine at both dosesfor Ex-NMDARs in vivo, and to examine the efficacy of both dosesin ameliorating striatal excitotoxicity and cognitive dysfunction inHD.

Certain considerations should be taken into account when usingmemantine treatment in vivo. First, several studies examining the af-finity of memantine for the NMDAR have been largely performedin vitro, in the absence of Mg2+ (Dravid et al., 2007; Johnson andKotermanski, 2006). However, in vivo at physiological Mg2+ concen-trations (~1 mM), memantine likely exhibits a higher IC50 than thatreported in vitro (Kotermanski and Johnson, 2009; Otton et al.,2011), as memantine and Mg2+ compete for NMDAR channel bind-ing (Sobolevsky et al., 1998). As indicated above, more recent studieshave confirmed that low-dosememantine (10 μM) effectively blocksthe majority of Ex-NDMAR currents in culture and remains selectivefor these receptors at physiological Mg2+ concentrations (1mM)(Xia et al., 2010). Second, memantine blocks 5-HT3 and acetylcholinereceptors (Aracava et al., 2005; Rammes et al., 2001) and activatesdopamine D2 receptors (D2Rs) by binding them with similar affinityas Ex-NMDARs (Seeman et al., 2008). Agonism of modulatory dopa-minergic signaling in D2R-expressing indirect pathway SPNs could

suppress excitatory corticostriatal neurotransmission in this pathway(Gerfen and Surmeier, 2011). However, whether this would resultin a net neuroprotective or neurotoxic effect in HD remains to be deter-mined. Moreover, the potential off-target effects of memantine are like-ly indirect and may not affect Ex-NMDAR-mediated Ca2+ signaling.

Ex-NMDAR surface expression is Ex-NMDAR activity-dependent

Previously, we reported elevated striatal expression of GluN2B-containing Ex-NMDARs in YAC128 mice both at early presymptomatic(1–2 months of age) and late stages of HD (1 year of age) (Gladdinget al., 2012;Milnerwood et al., 2010). Supporting these findings, we ob-served a significant increase in total (calpain-cleaved and full-length)GluN2B expression in the non-PSD of 4 month-old untreated YAC128compared to WT mice. Hence, increased Ex-NMDAR expression in HDis present at early, mid, and late stages of the disease, and could contrib-ute to striatal excitotoxicity at both early andmore advanced periods ofdisease progression. Although the non-PSD fraction contains presynap-tic, endosomal, and extrasynaptic membranes, our electrophysiologicaldata showing elevated Ex-NMDAR currents in untreated YAC128 micesuggest that the increase in non-PSD GluN2B subunits likely involvesfunctional receptors.

Interestingly, memantine (1 and 10 mg/kg/day) reduced functionalEx-NMDAR expression in YAC128 mice to WT levels. This suggests

540 A. Dau et al. / Neurobiology of Disease 62 (2014) 533–542

that the mtHtt-induced increase in extrasynaptic receptor expression isEx-NMDAR activity-dependent. Hence, Ex-NMDAR mislocalizationcould be driven by a feed-forward loop, whereby Ex-NMDAR activitytriggers pathways that in turnpotentiate expression of extrasynaptic re-ceptors. The protease calpain,which is activated early inHD, contributesto Ex-NMDAR surface mislocalization in 1–2 month old YAC128 mice(Gladding et al., 2012). Notably, calpain activity is downstream of Ex-NMDARs (Xu et al., 2009), and could thus drive Ex-NMDAR-mediatedreceptor mislocalization. Although we detected an increase in calpain-cleaved spectrin levels in 1.5 month-old YAC128 mice compared toWT littermates, calpain cleavage of spectrin was similar between4 month old YAC128 and WT mice, and was unaffected by memantinetreatment. Hence, in contrast to presymptomatic stages of HD, calpainactivity is attenuated with HD progression and may not be associatedwith receptor mislocalization or striatal excitotoxicity at later stages ofthe disease.

The peroxisome proliferator-activated receptor γ coactivator-1α(PGC-1α), a transcriptional coactivator that regulates mitochondrialfunction (Finck and Kelly, 2006), has recently been reported to modu-late Ex-NMDAR surface expression: reduced levels of PGC-1α are asso-ciated with a shift in NMDAR distribution to extrasynaptic sites,whereas increased levels reduce Ex-NMDARs (Puddifoot et al., 2012).CREB-mediated PGC-1α expression is reduced in HD, which may inpart result from Ex-NMDAR-mediated CREB shut-off (Cui et al., 2006;Okamoto et al., 2009; Puddifoot et al., 2012). Thus, suppressed PGC-1α signaling could be a putative mechanism linking Ex-NMDAR activityto receptor mislocalization.

Synaptic (PSD) GluN2B levels were unchanged between untreatedYAC128 and WT mice and were unaffected by memantine treatment.Hence, the effect of mtHtt and memantine on NMDAR distribution isspecific to an increase at extrasynaptic sites of neurons and maynot involve a shift in synaptic–extrasynaptic NMDAR localization.This is consistent with reports that synaptic NMDARs are less mobilethan Ex-NMDARs due to tight anchoring in the PSD (Gladding andRaymond, 2011). Moreover, in agreement with these data, we havepreviously detected similar levels of synaptic NMDAR activity andGluN2B expression between WT and YAC128 SPNs (Milnerwoodet al., 2010, 2012).

Elevated Ex-NMDAR expression in HD could be attributed to anaccelerated rate of NMDAR forward trafficking and translocation ofsynaptic receptors to extrasynaptic sites, or increased surface reten-tion of Ex-NMDARs (Fan et al., 2007; Gladding and Raymond, 2011;Gladding et al., 2012; Tovar and Westbrook, 2002). As we did notdetect changes in synaptic NMDAR levels between YAC128 micecompared to WT, a shift in NMDAR localization from synaptic toextrasynaptic sites would have to occur simultaneously with increasedsurface delivery at the PSD, such that net synaptic NMDAR levels remainunchanged. Further studies will be required to elucidate the pathwaysdriving activity-dependent Ex-NMDAR mislocalization in HD.

Calcium signaling is age-dependent in HD

The observation that elevated calpain activity in YAC128 mice isattenuated with age suggests that Ca2+-dependent signaling mech-anisms in HD may evolve with disease progression. Previous workhas also detected age-dependent alterations in Ca2+ signaling path-ways in HD. For instance, activity and expression of the Ca2+-dependent STriatal-Enriched tyrosine Phosphatase (STEP) is de-creased with age in a late-stage HD mouse model (Saavedra et al.,2011). Additionally, striatal neurons in an excitotoxin-resistantmouse model develop enhanced cytosolic Ca2+ buffering with age,which could account for their resistant phenotype (Hansson et al.,2001). Notably, these age-dependent alterations in Ca2+ signalingappear to be specific to particular pathways, as Ex-NMDAR expres-sion, as well as p38 MAPK and CREB activity, remain dysregulatedat both early (1–2 months) and later (4 months) stages of HD, as

shown here and elsewhere (Fan et al., 2012; Milnerwood et al.,2010). Overall, more work is required to further examine the time-dependent alterations in aberrant Ca2+ signaling in HD.

Ex-NMDARs couple to pro-death signaling in HD

In culture, activation of Ex-NMDARs induces CREB shut-off(Hardingham et al., 2002; Kaufman et al., 2012; Papadia andHardingham, 2007). Here, we observed a significant decrease in nuclearCREB activity in 4 month-old YAC128 mice compared to WT litter-mates, which was rescued by memantine treatment at the lower dose(1 mg/kg/day), as reported previously (Milnerwood et al., 2010).Thus, our data confirms that Ex-NMDAR activity drives suppression ofneuroprotective CREB activity, which could contribute to striatal neuro-degeneration in HD. Strikingly, memantine at the higher dose did notrescue YAC128 CREB activity and trended to decrease pCREBSer133/CREB levels in WT mice. Synaptic NMDARs, which activate CREB viaRas-ERK or CaMKIV pathways (Impey et al., 2002; Wu et al., 2001),are blocked by memantine at higher doses (Okamoto et al., 2009).Thus, partial inhibition of synaptic NMDARs by memantine at the10 mg/kg/day dose could explain its inability to rescue CREB signalingin YAC128 mice.

P-p38/p38 levels were significantly increased in untreated4 month-old YAC128 mice compared to WT. This is consistent withreports of elevated p38 MAPK activity in 1–2 month old YAC128mice (Fan et al., 2012), and suggests that unlike calpain, aberrantp38 signaling remains elevated as HD progresses. Moreover, p38phosphorylation in YAC128 mice was suppressed to WT levels bymemantine, indicating that p38 signaling is downstream of Ex-NMDAR activity in HD. In fact, stimulation of Ex-NMDARs in cultureinduces a rapid and prolonged increase in p38 activity (Xu et al.,2009). During periods of oxidative stress, p38 activation stimulatesmitochondrial translocation of Bax, cytochrome-c release and cas-pase activation thus potentiating pro-apoptotic signaling (Ghatanet al., 2000; Gomez-Lazaro et al., 2007). Hence, p38 MAPK signalingcould be a pathway bywhich Ex-NMDARs facilitate later striatal neuro-degeneration. In neurons, Ex-NMDARs could couple to p38MAPK activ-ity via distinct pathways. In cultured cortical neurons, Ex-NMDARspotentiate p38 phosphorylation by inducing calpain-mediated cleavageand inactivation of the striatal enriched tyrosine phosphatase (STEP),which dephosphorylates p38 (Xu et al., 2009).Wedid not detect chang-es in calpain activity in 4 month-old YAC128 mice; thus, other mecha-nisms must mediate Ex-NMDAR-dependent enhanced p38 activationat this age. Alternatively, GluN2B-PSD-95 interactions couple Ca2+ in-flux through the receptor to SynGAP and nNOS, both of which triggerp38 signaling (Aarts et al., 2002; Cao et al., 2005; Rumbaugh et al.,2006; Soriano et al., 2008). Indeed, associations between Ex-NMDARsand PSD-95 are increased in YAC128mice, and these interactionsmedi-ate p38 activity and acute excitotoxic death in mtHtt-expressing SPNs(Fan et al., 2009, 2012). Hence, Ex-NMDAR-induced p38 MAPK acti-vation in HD could be associated with aberrant GluN2B-PSD-95-mediated signaling.

Conclusions

By chronically blocking Ex-NMDARs in vivo, we show that Ex-NMDAR activity contributes to receptor mislocalization as well asaberrant p38 and CREB signaling pathways in the YAC128 mousemodel of HD. As these mechanisms promote mtHtt-induced neuro-nal dysfunction and death (Fan et al., 2012; Milnerwood et al.,2010; Okamoto et al., 2009; Zeron et al., 2004), our data shedlight on the signaling pathways linking Ex-NMDARs to cognitivedecline and striatal atrophy in HD, and help clarify why low-dosememantine treatment is neuroprotective in YAC128 HD pathology.

541A. Dau et al. / Neurobiology of Disease 62 (2014) 533–542

Acknowledgments

Wewould like to thankD. Lynch for the anti-spectrin AB28 antibody.Wewould also like to thank L.Wang and L. Zhang for technical support,as well as M. Parsons and K. Kolodziejczyk for experimental advice andinput. Funding for LAR was provided by the Cure Huntington's DiseaseInitiative (CHDI) and the Canadian Institutes of Health Research(CIHR) (MOP-12699). Funding for CMG was provided by a CIHR–Hun-tington Society of Canada Fellowship, and CMG also held a HereditaryDisease Foundation Fellowship.

References

Aarts, M., Liu, Y., Liu, L., Besshoh, S., Arundine, M., Gurd, J.W., Wang, Y.-T., Salter, M.W.,Tymianski, M., 2002. Treatment of ischemic brain damage by perturbing NMDAreceptor- PSD-95 protein interactions. Science 298, 846–850.

Aracava, Y., Pereira, E.F.R., Maelicke, A., Albuquerque, E.X., 2005. Memantine blocks α7*nicotinic acetylcholine receptors more potently than N-methyl-D-aspartate receptorsin rat hippocampal neurons. J. Pharmacol. Exp. Ther. 312, 1195–1205.

Cao, J., Viholainen, J.I., Dart, C., Warwick, H.K., Leyland, M.L., Courtney, M.J., 2005. ThePSD95-nNOS interface: a target for inhibition of excitotoxic p38 stress-activated pro-tein kinase activation and cell death. J. Cell Biol. 168, 117–126.

Cepeda, C., Ariano, M.A., Calvert, C.R., Flores-Hernández, J., Chandler, S.H., Leavitt, B.R.,Hayden, M.R., Levine, M.S., 2001. NMDA receptor function in mouse models of Hun-tington disease. J. Neurosci. Res. 66, 525–539.

Chen, N., Luo, T., Wellington, C., Metzler, M., McCutcheon, K., Hayden,M.R., Raymond, L.A.,1999. Subtype-specific enhancement of NMDA receptor currents by mutanthuntingtin. J. Neurochem. 72, 1890–1898.

Cowan, C.M., Fan, M.M.Y., Fan, J., Shehadeh, J., Zhang, L.Y.J., Graham, R.K., Hayden, M.R.,Raymond, L.A., 2008. Polyglutamine-modulated striatal calpain activity in YAC trans-genic Huntington disease mouse model: impact on NMDA receptor function and tox-icity. J. Neurosci. 28, 12725–12735.

Cui, L., Jeong, H., Borovecki, F., Parkhurst, C.N., Tanese, N., Krainc, D., 2006. Transcriptionalrepression of PGC-1alpha by mutant huntingtin leads to mitochondrial dysfunctionand neurodegeneration. Cell 127, 59–69.

Dravid, S.M., Erreger, K., Yuan, H., Nicholson, K., Le, P., Lyuboslavsky, P., Almonte, A.,Murray, E., Mosely, C., Barber, J., et al., 2007. Subunit-specific mechanisms and protonsensitivity of NMDA receptor channel block. J. Physiol. 581, 107–128.

Fan, M.M.Y., Fernandes, H.B., Zhang, L.Y.J., Hayden, M.R., Raymond, L.A., 2007. AlteredNMDA receptor trafficking in a yeast artificial chromosome transgenic mousemodel of Huntington's disease. J. Neurosci. Off. J. Soc. Neurosci. 27, 3768–3779.

Fan, J., Cowan, C.M., Zhang, L.Y.J., Hayden, M.R., Raymond, L.A., 2009. Interaction of post-synaptic density protein-95 with NMDA receptors influences excitotoxicity in theyeast artificial chromosome mouse model of Huntington's disease. J. Neurosci. Off.J. Soc. Neurosci. 29, 10928–10938.

Fan, J., Gladding, C.M., Wang, L., Zhang, L.Y.J., Kaufman, A.M., Milnerwood, A.J., Raymond,L.A., 2012. P38 MAPK is involved in enhanced NMDA receptor-dependentexcitotoxicity in YAC transgenic mouse model of Huntington disease. Neurobiol.Dis. 45, 999–1009.

Finck, B.N., Kelly, D.P., 2006. PGC-1 coactivators: inducible regulators of energy metabo-lism in health and disease. J. Clin. Invest. 116, 615–622.

Gafni, J., Ellerby, L.M., 2002. Calpain activation in Huntington's disease. J. Neurosci. Off.J. Soc. Neurosci. 22, 4842–4849.

Gafni, J., Hermel, E., Young, J.E., Wellington, C.L., Hayden, M.R., Ellerby, L.M., 2004. Inhibi-tion of calpain cleavage of huntingtin reduces toxicity: accumulation of calpain/cas-pase fragments in the nucleus. J. Biol. Chem. 279, 20211–20220.

Gerfen, C.R., Surmeier, D.J., 2011. Modulation of striatal projection systems by dopamine.Annu. Rev. Neurosci. 34, 441–466.

Ghatan, S., Larner, S., Kinoshita, Y., Hetman, M., Patel, L., Xia, Z., Youle, R.J., Morrison, R.S.,2000. p38 MAP kinase mediates bax translocation in nitric oxide-induced apoptosisin neurons. J. Cell Biol. 150, 335–347.

Gladding, C.M., Raymond, L.A., 2011. Mechanisms underlying NMDA receptor synaptic/extrasynaptic distribution and function. Mol. Cell. Neurosci. 48, 308–320.

Gladding, C.M., Sepers, M.D., Xu, J., Zhang, L.Y.J., Milnerwood, A.J., Lombroso, P.J.,Raymond, L.A., 2012. Calpain and STriatal-Enriched protein tyrosine phosphatase(STEP) activation contribute to extrasynaptic NMDA receptor localization in aHuntington's disease mouse model. Hum. Mol. Genet. 21, 3739–3752.

Goebel-Goody, S.M., Davies, K.D., Alvestad Linger, R.M., Freund, R.K., Browning, M.D.,2009. Phospho-regulation of synaptic and extrasynaptic N-methyl-D-aspartate recep-tors in adult hippocampal slices. Neuroscience 158, 1446–1459.

Gomez-Lazaro, M., Galindo, M.F., Melero-Fernandez deMera, R.M., Fernandez-Gómez, F.J.,Concannon, C.G., Segura, M.F., Comella, J.X., Prehn, J.H.M., Jordan, J., 2007. Reactive ox-ygen species and p38 mitogen-activated protein kinase activate Bax to induce mito-chondrial cytochrome c release and apoptosis in response to malonate. Mol.Pharmacol. 71, 736–743.

Hansson, O., Guatteo, E., Mercuri, N.B., Bernardi, G., Li, X.J., Castilho, R.F., Brundin, P., 2001.Resistance to NMDA toxicity correlates with appearance of nuclear inclusions, behav-ioural deficits and changes in calcium homeostasis in mice transgenic for exon 1 ofthe huntington gene. Eur. J. Neurosci. 14, 1492–1504.

Hardingham, G.E., Bading, H., 2010. Synaptic versus extrasynaptic NMDA receptor signal-ling: implications for neurodegenerative disorders. Nat. Rev. Neurosci. 11, 682–696.

Hardingham, G.E., Fukunaga, Y., Bading, H., 2002. Extrasynaptic NMDARs oppose synapticNMDARs by triggering CREB shut-off and cell death pathways. Nat. Neurosci. 5,405–414.

Huntington's Disease Collaborative Research Group, 1993. A novel gene containing a tri-nucleotide repeat that is expanded and unstable on Huntington's disease chromo-somes. Cell 72, 971–983.

Impey, S., Fong, A.L., Wang, Y., Cardinaux, J.R., Fass, D.M., Obrietan, K., Wayman, G.A.,Storm, D.R., Soderling, T.R., Goodman, R.H., 2002. Phosphorylation of CBP mediatestranscriptional activation by neural activity and CaM kinase IV. Neuron 34, 235–244.

Johnson, J.W., Kotermanski, S.E., 2006. Mechanism of action of memantine. Curr. Opin.Pharmacol. 6, 61–67.

Kaufman, A.M., Milnerwood, A.J., Sepers, M.D., Coquinco, A., She, K., Wang, L., Lee, H.,Craig, A.M., Cynader, M., Raymond, L.A., 2012. Opposing roles of synaptic andextrasynaptic NMDA receptor signaling in cocultured striatal and cortical neurons.J. Neurosci. 32, 3992–4003.

Kim, Y.J., Yi, Y., Sapp, E., Wang, Y., Cuiffo, B., Kegel, K.B., Qin, Z.-H., Aronin, N., DiFiglia, M.,2001. Caspase 3-cleaved N-terminal fragments of wild-type and mutant huntingtinare present in normal and Huntington's disease brains, associate with membranes,and undergo calpain-dependent proteolysis. Proc. Natl. Acad. Sci. 98, 12784–12789.

Kotermanski, S.E., Johnson, J.W., 2009. Mg2+ imparts NMDA receptor subtype selectivityto the Alzheimer's drug memantine. J. Neurosci. Off. J. Soc. Neurosci. 29, 2774–2779.

Li, L., Fan, M., Icton, C.D., Chen, N., Leavitt, B.R., Hayden, M.R., Murphy, T.H., Raymond, L.A.,2003. Role of NR2B-type NMDA receptors in selective neurodegeneration in Hunting-ton disease. Neurobiol. Aging 24, 1113–1121.

Milnerwood, A.J., Raymond, L.A., 2007. Corticostriatal synaptic function in mouse modelsof Huntington's disease: early effects of huntingtin repeat length and protein load.J. Physiol. 585, 817–831.

Milnerwood, A.J., Gladding, C.M., Pouladi, M.A., Kaufman, A.M., Hines, R.M., Boyd, J.D., Ko,R.W.Y., Vasuta, O.C., Graham, R.K., Hayden, M.R., et al., 2010. Early increase inextrasynaptic NMDA receptor signaling and expression contributes to phenotypeonset in Huntington's disease mice. Neuron 65, 178–190.

Milnerwood, A.J., Kaufman, A.M., Sepers, M.D., Gladding, C.M., Zhang, L., Wang, L., Fan, J.,Coquinco, A., Qiao, J.Y., Lee, H., et al., 2012. Mitigation of augmented extrasynapticNMDAR signaling and apoptosis in cortico-striatal co-cultures from Huntington's dis-ease mice. Neurobiol. Dis. 48, 40–51.

Okamoto, S., Pouladi, M.A., Talantova, M., Yao, D., Xia, P., Ehrnhoefer, D.E., Zaidi, R.,Clemente, A., Kaul, M., Graham, R.K., et al., 2009. Balance between synaptic versusextrasynaptic NMDA receptor activity influences inclusions and neurotoxicity of mu-tant huntingtin. Nat. Med. 15, 1407–1413.

Otton, H.J., Lawson McLean, A., Pannozzo, M.A., Davies, C.H., Wyllie, D.J.A., 2011. Quantifi-cation of the Mg2+-induced potency shift of amantadine and memantine voltage-dependent block in human recombinant GluN1/GluN2A NMDARs. Neuropharmacol-ogy 60, 388–396.

Pacchioni, A.M., Vallone, J., Worley, P.F., Kalivas, P.W., 2009. Neuronal pentraxins modu-late cocaine-induced neuroadaptations. J. Pharmacol. Exp. Ther. 328, 183–192.

Papadia, S., Hardingham, G.E., 2007. The dichotomy of NMDA receptor signaling. Neuro-scientist 13, 572–579.

Puddifoot, C., Martel, M.-A., Soriano, F.X., Camacho, A., Vidal-Puig, A., Wyllie, D.J.A.,Hardingham, G.E., 2012. PGC-1a negatively regulates extrasynaptic NMDAR activityand excitotoxicity. J. Neurosci. Off. J. Soc. Neurosci. 32, 6995–7000.

Raina, P., Santaguida, P., Ismaila, A., Patterson, C., Cowan, D., Levine, M., Booker, L.,Oremus, M., 2008. Effectiveness of cholinesterase inhibitors and memantine fortreating dementia: evidence review for a clinical practice guideline. Ann. Intern.Med. 148, 379–397.

Rammes, G., Rupprecht, R., Ferrari, U., Zieglgänsberger, W., Parsons, C., 2001. The N-methyl-D-aspartate receptor channel blockers memantine, MRZ 2/579 and otheramino-alkyl-cyclohexanes antagonise 5-HT3 receptor currents in cultured HEK-293and N1E-115 cell systems in a non-competitive manner. Neurosci. Lett. 306, 81–84.

Rammes, G., Danysz, W., Parsons, C.G., 2008. Pharmacodynamics of memantine: an up-date. Curr. Neuropharmacol. 6, 55–78.

Rosi, S., Vazdarjanova, A., Ramirez-Amaya, V., Worley, P.F., Barnes, C.A., Wenk, G.L., 2006.Memantine protects against LPS-induced neuroinflammation, restores behaviorally-induced gene expression and spatial learning in the rat. Neuroscience 142,1303–1315.

Rumbaugh, G., Adams, J.P., Kim, J.H., Huganir, R.L., 2006. SynGAP regulates synapticstrength and mitogen-activated protein kinases in cultured neurons. Proc. Natl.Acad. Sci. U. S. A. 103, 4344–4351.

Saavedra, A., Giralt, A., Rue, L., Xifro, X., Xu, J., Ortega, Z., Lucas, J.J., Lombroso, P.J., Alberch,J., Perez-Navarro, E., 2011. STriatal-Enriched protein tyrosine Phosphatase expressionand activity in Huntington's disease — a STEP in the resistance to excitotoxicity.J. Neurosci. Off. J. Soc. Neurosci. 31, 8150–8162.

Seeman, P., Caruso, C., Lasaga, M., 2008. Memantine agonist action at dopamine D2Highreceptors. Synapse 62, 149–153.

Slow, E.J., van Raamsdonk, J., Rogers, D., Coleman, S.H., Graham, R.K., Deng, Y., Oh,R., Bissada, N., Hossain, S.M., Yang, Y.-Z., et al., 2003. Selective striatal neuronalloss in a YAC128 mouse model of Huntington disease. Hum. Mol. Genet. 12,1555–1567.

Sobolevsky, A.I., Koshelev, S.G., Khodorov, B.I., 1998. Interaction of memantine and aman-tadine with agonist-unbound NMDA-receptor channels in acutely isolated rat hippo-campal neurons. J. Physiol. 512 (Pt 1), 47–60.

Soriano, F.X., Martel, M.-A., Papadia, S., Vaslin, A., Baxter, P., Rickman, C., Forder, J.,Tymianski, M., Duncan, R., Aarts, M., et al., 2008. Specific targeting of pro-deathNMDA receptor signals with differing reliance on the NR2B PDZ ligand. J. Neurosci.Off. J. Soc. Neurosci. 28, 10696–10710.

Starling, A.J., André, V.M., Cepeda, C., de Lima, M., Chandler, S.H., Levine, M.S., 2005. Alter-ations in N-methyl-D-aspartate receptor sensitivity and magnesium blockade occur

542 A. Dau et al. / Neurobiology of Disease 62 (2014) 533–542

early in development in the R6/2 mouse model of Huntington's disease. J. Neurosci.Res. 82, 377–386.

Tovar, K.R., Westbrook, G.L., 2002. Mobile NMDA receptors at hippocampal synapses.Neuron 34, 255–264.

Tzingounis, A.V., Wadiche, J.I., 2007. Glutamate transporters: confining runaway excita-tion by shaping synaptic transmission. Nat. Rev. Neurosci. 8, 935–947.

Vonsattel, J.P., Myers, R.H., Stevens, T.J., Ferrante, R.J., Bird, E.D., Richardson Jr., E.P., 1985.Neuropathological classification of Huntington's disease. J. Neuropathol. Exp. Neurol.44, 559–577.

Wu, G.-Y., Deisseroth, K., Tsien, R.W., 2001. Activity-dependent CREB phosphorylation:convergence of a fast, sensitive calmodulin kinase pathway and a slow, less sensitivemitogen-activated protein kinase pathway. Proc. Natl. Acad. Sci. U. S. A. 98,2808–2813.

Wu, J., Shih, H.-P., Vigont, V., Hrdlicka, L., Diggins, L., Singh, C., Mahoney, M., Chesworth, R.,Shapiro, G., Zimina, O., et al., 2011. Neuronal store-operated calcium entry pathway asa novel therapeutic target for Huntington's disease treatment. Chem. Biol. 18, 777–793.

Xia, P., Chen, H.V., Zhang, D., Lipton, S.A., 2010. Memantine preferentially blocksextrasynaptic over synaptic NMDA receptor currents in hippocampal autapses.J. Neurosci. Off. J. Soc. Neurosci. 30, 11246–11250.

Xu, J., Kurup, P., Zhang, Y., Goebel-Goody, S.M., Wu, P.H., Hawasli, A.H., Baum, M.L., Bibb,J.A., Lombroso, P.J., 2009. Extrasynaptic NMDA receptors couple preferentially toexcitotoxicity via calpain-mediated cleavage of STEP. J. Neurosci. Off. J. Soc. Neurosci.29, 9330–9343.

Zeron, M.M., Chen, N., Moshaver, A., Lee, A.T., Wellington, C.L., Hayden, M.R., Raymond, L.A.,2001. Mutant huntingtin enhances excitotoxic cell death. Mol. Cell. Neurosci. 17, 41–53.

Zeron, M.M., Hansson, O., Chen, N.,Wellington, C.L., Leavitt, B.R., Brundin, P., Hayden, M.R.,Raymond, L.A., 2002. Increased sensitivity to N-methyl-D-aspartate receptor-mediated excitotoxicity in a mouse model of Huntington's disease. Neuron 33,849–860.

Zeron, M.M., Fernandes, H.B., Krebs, C., Shehadeh, J., Wellington, C.L., Leavitt, B.R.,Baimbridge, K.G., Hayden, M.R., Raymond, L.A., 2004. Potentiation of NMDAreceptor-mediated excitotoxicity linked with intrinsic apoptotic pathway in YACtransgenic mouse model of Huntington's disease. Mol. Cell. Neurosci. 25, 469–479.

Zhang, H., Li, Q., Graham, R.K., Slow, E., Hayden, M.R., Bezprozvanny, I., 2008. Full lengthmutant huntingtin is required for altered Ca2+ signaling and apoptosis of striatalneurons in the YAC mouse model of Huntington's disease. Neurobiol. Dis. 31, 80–88.

Zuccato, C., Valenza, M., Cattaneo, E., 2010. Molecular mechanisms and potentialtherapeutical targets in Huntington's disease. Physiol. Rev. 90, 905–981.