Glutamatergic Dysfunctioning in Alzheimer’s Disease and Related Therapeutic Targets

Uncorrected Author Proof

IOS Press

Review

1

Glutamatergic Dysfunctioning in

Alzheimer’s Disease and Related Therapeutic Targets

2

3

4

D´enes Z´adori^a, G´abor Veres^a, Levente Szal´ardy^a, P´eter Kliv´enyi^a, J´ozsef Toldi^b,cand L´aszl´o V´ecsei^a,c,∗

5

aDepartment of Neurology, Faculty of Medicine, Albert Szent-Gy¨orgyi Clinical Center, University of Szeged, Szeged, Hungary

6 7

bDepartment of Physiology, Anatomy and Neuroscience, Faculty of Science and Informatics, University of Szeged, Szeged, Hungary

8 9

cMTA-SZTE Neuroscience Research Group, Szeged, Hungary

10

Accepted 13 February 2014

Abstract. The impairment of glutamatergic neurotransmission plays an important role in the development of Alzheimer’s disease (AD). The pathological process, which involves the production of amyloid-␤peptides and hyperphosphorylated tau proteins, spreads over well-delineated neuroanatomical circuits. The gradual deterioration of proper synaptic functioning (via GluN2A-containing N-methyl-D-aspartate receptors, NMDARs) and the development of excitotoxicity (via GluN2B-containing NMDARs) in these structures both accompany the disease pathogenesis. Although one of the most important therapeutic targets would be glutamate excitotoxicity, the application of conventional anti-glutamatergic agents could result in further deterioration of synaptic transmission and intolerable side-effects. With regard to NMDAR antagonists with tolerable side-effects, ion channel blockers with low affinity, glycine site agents, and specific antagonists of polyamine site and GluN2B subunit may come into play.

However, in the mirror of experimental data, only the application of ion channel blockers with pronounced voltage dependency, low affinity, and rapid unblocking kinetics (e.g., memantine) and specific antagonists of the GluN2B subunit (e.g., ifenprodil and certain kynurenic acid amides) resulted in desirable symptom amelioration. Therefore we propose that these kinds of chemical agents may have therapeutic potential for present and future drug development.

11 12 13 14 15 16 17 18 19 20 21 22

Keywords: Alzheimer’s disease, glutamate excitotoxicity, kynurenic acid amides, memantine, neurodegeneration, neuroprotec- tion, therapy

23 24

INTRODUCTION

25

Alzheimer’s disease (AD) is a progressive neurode-

26

generative disorder, the main clinical feature of which

27

is dementia [1, 2]. Indeed, AD is the most common

28

type among dementia syndromes [3] and is responsible

29

for 60–80% of the cases [4], leading to a considerable

30

∗Correspondence to: L´aszl´o V´ecsei MD, PhD, DSc, Depart- ment of Neurology, Faculty of Medicine, Albert Szent-Gy¨orgyi Clinical Center, University of Szeged, Semmelweis u. 6, H- 6725 Szeged, Hungary. Tel.: +36(62)545351; Fax: +36(62)545597;

E-mail: vecsei.laszlo@med.u-szeged.hu.

socioeconomic burden. Although clinical diagnosis ³¹ can be determined during the disease course in most ³² cases, currently autopsy is necessary for a definite diag- ³³ nosis. The main pathological hallmark of AD is the ³⁴ presence of neurofibrillary tangles (NFTs) and senile ³⁵ plaques in specific brain areas [5]. With regard to ³⁶ the involvement of dysfunctional neurotransmission ³⁷ in disease pathogenesis, certain cholinergic and glu- ³⁸ tamatergic systems are the most affected [6, 7]. ³⁹ The aim of this short review is to highlight aspects of ⁴⁰ glutamatergic dysfunction in AD and to discuss some ⁴¹ possibilities of pharmaceutical interventions by target- ⁴² ing the glutamatergic system.

ISSN 1387-2877/14/$27.50 © 2014 – IOS Press and the authors. All rights reserved

Uncorrected Author Proof

ALTERATIONS IN GLUTAMATERGIC

43

SIGNALING IN ALZHEIMER’S DISEASE:

44

PATHOLOGICAL BASIS

45

With regard to the sensitivity and specificity for the

46

diagnosis of AD, the Braak staging system [5] gives

47

the best accuracy (79%) among the neuropatholog-

48

ical criteria systems [8]. This system classifies AD

49

into stages mainly by the temporal evolution of NFTs

50

(composed of intracellular aggregates of hyperphos-

51

phorylated tau protein), but it also takes into account

52

the loci of extracellular amyloid-␤(A␤) deposits in the

53

brain. The system distinguishes between the following

54

stages: transentorhinal/entorhinal (stage I, II), limbic

55

(stage III, IV), and neocortical (stage V, VI). This clas-

56

sification shows a good correlation with the severity of

57

dementia [9], though originally the pathological stages

58

were established by Braak irrespective of the clini-

59

cal stage of the dementia. Certain neuropathological

60

investigations have special significance in the assess-

61

ment of early stages of AD [10]. The most important

62

ones include the assessment of NFTs in the neurons of

63

the second layer of the entorhinal cortex in the slices

64

of the inferior temporal lobe. The entorhinal cortex

65

receives converging polysynaptic glutamatergic inputs

66

from the multimodal association cortices and limbic

67

areas including the hippocampal formation, while it

68

projects into the hippocampal formation and back to

69

the association cortices [11–13]. One of the main effer-

70

ent glutamatergic projections of the entorhinal cortex

71

is the perforant pathway, which predominantly orig-

72

inates from the second layer and serves as the main

73

excitatory input of the hippocampal formation. The

74

fourth layer of the entorhinal cortex in turn receives

75

excitatory input from the hippocampal formation. A

76

significant decrease was observed in the neuronal num-

77

ber of the fourth and especially the second layers

78

of the entorhinal cortex in clinically very mild AD

79

[14]. Another study likewise demonstrated a consid-

80

erable decrease in neuronal number and volume of

81

the entorhinal cortex (especially the second layer) and

82

those of the cornu ammonis (CA)1 region of the hip-

83

pocampus in preclinical AD cases [15]. It is important

84

to mention that the presence of NFTs can also be

85

observed in these early stages in the CA1-subiculum

86

part of the hippocampal formation and in the perirhinal

87

cortex, inferior temporal gyrus, amygdala, posterior

88

part of the parahippocampal gyrus, the cholinergic

89

basal forebrain and in the dorsal raphe nuclei, but in

90

a lesser extent compared to the second layer of the

91

entorhinal cortex [16]. In the next stages, almost all

92

the limbic structures, notably the hippocampal forma-

93

tion (consisting of the dentate gyrus, the hippocampus ⁹⁴ proper, and the subiculum) and the amygdala become ⁹⁵ considerably damaged [17] in addition to the more ⁹⁶ expressed involvement of the previously described ⁹⁷ brain structures. As partially mentioned above, the ⁹⁸ main glutamatergic input of the hippocampal forma- 99

tion comes from the second (toward the dentate gyrus) ¹⁰⁰ and the third (toward the subiculum and CA1 sector ¹⁰¹ of the hippocampus proper) layers of the entorhinal ¹⁰² cortex via the perforant and temporo-alvear pathways ¹⁰³ [18]. Scheff et al. [19] hypothesized that synaptic loss ¹⁰⁴ in the outer molecular layer (OML) of the dentate ¹⁰⁵ gyrus would be responsible for the transition from ¹⁰⁶ mild cognitive impairment to early AD. Total synaptic ¹⁰⁷ counts in the OML had a significant negative cor- ¹⁰⁸ relation with NFT density in the entorhinal cortex. ¹⁰⁹ Although there was a negative correlation between ¹¹⁰ the individual’s Braak score and total synaptic num- ¹¹¹ ber in the OML, this association was not significant ¹¹² and furthermore, this study did not find significant cor- ¹¹³ relation of Braak staging with the scores of any of 114

the applied psychometric tests. However, a high pos- ¹¹⁵ itive correlation of total synaptic number in the OML ¹¹⁶ with the values of tests of cognitive functions such ¹¹⁷ as the Mini-Mental State Examination and delayed ¹¹⁸ memory recall (one of the most sensitive measures of ¹¹⁹ hippocampal function) was demonstrated, which sug- ¹²⁰ gests that synaptic loss would be one of the strongest ¹²¹ predictive factors for cognitive decline. As a part of ¹²² the trisynaptic circuit, the information is transmitted ¹²³ further from the dentate gyrus via intrahippocampal ¹²⁴ association pathways (via mossy fibers toward the CA3 ¹²⁵ sector of the hippocampus, and then via Schaffer col- ¹²⁶ laterals toward the CA1 sector) [20]. The synaptic ¹²⁷ loss can also be observed in the CA1 sector of the ¹²⁸ hippocampus in mild AD cases [21]. The pyramidal 129

cells of the CA1 sector predominantly innervate the ¹³⁰ subiculum, which projects to the pre/parasubiculum ¹³¹ (parts of the subicular complex which also receives ¹³² neocortical inputs likewise the entorhinal cortex), the ¹³³ amygdala, the fourth layer of the entorhinal cortex, ¹³⁴ the anterior and midline thalamic and mammillary ¹³⁵ nuclei (via the fornix) [22]. Regarding further parts of ¹³⁶ the Papez circuit, the information processes from the ¹³⁷ mammillary nuclei to the anterior thalamic nuclei (via ¹³⁸ the mammilothalamic tract) and further to the cingu- ¹³⁹ lated gyrus (via the anterior thalamic radiation) and to ¹⁴⁰ the presubiculum (via the cingulum), which projects ¹⁴¹ to the fourth layer of the entorhinal cortex [23]. The ¹⁴² pre/parasubiculum also send minor projections to the ¹⁴³ dentate gyrus [24]. It is important to mention that parts 144

of the hippocampal formation in the two hemispheres ¹⁴⁵

Uncorrected Author Proof

are strongly interconnected via commissural fibers.

146

The amygdaloid complex, which consists of distinct

147

nuclei, receives inputs from multiple brain regions

148

via several kinds of transmitter systems, including

149

glutamatergic pathways [25]. The major sources of

150

sensory and polymodal information to the amygdala

151

are certain parts of the cerebral cortex, including the

152

association and prefrontal cortices [26]. The amyg-

153

dala also forms reciprocal and strong connections with

154

areas related to long-term declarative memory system,

155

including the perirhinal and entorhinal cortices and the

156

hippocampal formation [27]. Furthermore, the amyg-

157

daloid complex has widespread projections to certain

158

cortical, subcortical, and brainstem structures [25]. The

159

key feature of advanced stages of AD (stage V-VI)

160

is the occurrence of severe destruction of neocortical

161

association areas [28, 29]. Although NFT pathology

162

only becomes expressed in advanced stages of AD in

163

neocortical areas, the alteration in the level of some

164

molecular markers of synaptic dysfunctioning can be

165

observed even in early stages of AD. Accordingly,

166

vesicular glutamate transporter (VGLUT)1 expression

167

is found to be decreased in the prefrontal, parietal

168

and occipital and inferior temporal cortices, while it

169

was unaltered in the lateral temporal cortex [30–32].

170

With regard to the murine models of AD, a signif-

171

icant reduction of VGLUT1 was observed in both

172

the frontal cortex and the hippocampus [33, 34]. The

173

expression of VGLUT2 and synaptophysin was altered

174

only in the prefrontal cortex in human AD cases [30].

175

Loss of VGLUT1 and VGLUT2 in the prefrontal

176

cortex correlated with cognitive status even at early

177

phases of cognitive decline [30]. Although the typi-

178

cal spreading of neuropathological alterations over the

179

above-mentioned glutamatergic structures with strong

180

connections (Fig. 1) can be well observed in most

181

cases, limbic-predominant and hippocampal-sparing

182

subtypes of AD cases were also reported [35].

183

ALTERATIONS IN GLUTAMATERGIC

184

SIGNALING IN ALZHEIMER’S DISEASE:

185

MOLECULAR BASIS

186

The main culprits responsible for the discon-

187

nection of the previously delineated glutamatergic

188

networks would be the A␤ peptide and the tau pro-

189

tein [36]. A␤1-42 aggregates are capable of inducing

190

tau hyperphosphorylation [36] and promote in vitro

191

tau aggregation in a dose-dependent manner [37].

192

In addition to NFTs, soluble tau also would have

193

neurotoxic properties [38]. A␤ can influence gluta-

194

matergic neurotransmission in several ways. Although ¹⁹⁵ under physiological concentrations, endogenous A␤ ¹⁹⁶ is necessary for proper neurotransmitter release [39], ¹⁹⁷ in excess it weakens synaptic transmission affecting ¹⁹⁸ the synaptic vesicle pools [40]. Accordingly, A␤ is ¹⁹⁹ co-localized in glutamatergic boutons immunoreac- 200

tive for VGLUT1 and VGLUT2 in postmortem AD ²⁰¹ brains [41]. Furthermore, soluble A␤oligomers induce ²⁰² the disruption of dendritic spines, resulting in severe ²⁰³ neuropil damage [42]. The degeneration of synapses ²⁰⁴ and dendritic spines is one of the earliest feature of ²⁰⁵ AD [43]. Glutamatergic synapses contain ␣-amino- ²⁰⁶ 3-hydroxy-5-methyl-4-isoxazolepropionic acid recep- ²⁰⁷ tors (AMPARs) and N-methyl-D-aspartate receptors ²⁰⁸ (NMDARs) localized on dendritic spines. The basal ²⁰⁹ synaptic transmission is mainly mediated by AMPARs. ²¹⁰ However, in view of receptor dysfunction in AD, ²¹¹ the NMDAR would be the major site of A␤ action, ²¹² and in turn, NMDAR activation enhances A␤ pro- ²¹³ duction [44]. A conventional NMDAR is composed ²¹⁴ of two glycine or D-serine-binding GluN1 and 2 215

glutamate-binding GluN2 (A-D) subunits, forming ²¹⁶ a heterotetramer. The GluN1 subunits form the ion ²¹⁷ channel, while the GluN2 subunits have more of a ²¹⁸ regulatory and refining role. It has been shown that ²¹⁹ the GluN2B subunit-containing NMDARs predomi- ²²⁰ nate at the extrasynaptic site [45], which preferential ²²¹ localization becomes more predominant by the phos- ²²² phorylation at Tyr1336 [46]. Oligomeric A␤promotes ²²³ Fyn kinase activation via binding to the post-synaptic ²²⁴ prion protein (PrP^C), resulting in the increased phos- ²²⁵ phorylation of the GluN2B subunits at Tyr1472 [47]. ²²⁶ This activation induces altered NMDAR localiza- ²²⁷ tion with destabilization of dendritic spines and the ²²⁸ loss of surface NMDARs. It is important to mention ²²⁹ that several other receptors are regulated by PrP^C, 230

including metabotropic glutamate receptor (mGluR) ²³¹ 1 and 5 [48]. The available data suggest that the ²³² activation of NMDARs at the synaptic site promotes ²³³ neuronal survival, while activation at the extrasynap- ²³⁴ tic site mediates neurotoxic effects [49]. However, ²³⁵ some recent findings suggest that the simultaneous ²³⁶ activation of synaptic NMDARs are also necessary ²³⁷ for the initiation of cell death program [50]. So ²³⁸ in brief, the inactivation of glutamatergic synap- ²³⁹ tic transmission and the activation of that at the ²⁴⁰ extrasynaptic sites would both accompany the path- ²⁴¹ omechanism of AD. Oligomeric A␤impairs long-term ²⁴² potentiation (LTP; a form of synaptic strengthening ²⁴³ following brief, high frequency stimulation [51]) and ²⁴⁴ enhances long-term depression (LTD; a form of synap- 245

tic weakening following low frequency stimulation ²⁴⁶

Uncorrected Author Proof

Fig. 1. The schematic depiction of the predominant connections between the affected glutamatergic brain areas in Alzheimer’s disease. (CA, cornu ammonis).

or synaptic inactivity [52]) and the depotentiation

247

of LTP, thereby causing synaptic dysfunctioning

248

[53, 54]. Oligomeric A␤-induced internalization of

249

synaptic AMPARs and NMDARs [55, 56] and non-

250

apoptotic caspase activation [57] both accompany LTD

251

enhancement. Although several forms of synaptic plas-

252

ticity depend on NMDAR-driven calcium flux [58],

253

some recent data indicate that A␤-mediated synap-

254

tic AMPAR depression requires NMDAR activation

255

in a metabotropic manner, i.e., without ion flow via

256

the NMDAR [59]. NMDARs also have an important

257

role in spontaneous glutamate release-induced depres-

258

sion of evoked neurotransmission, thereby influencing

259

synaptic efficacy as well [60]. In addition to the demon-

260

strated alteration of glutamatergic neurotransmission

261

via postsynaptic and extrasynaptic NMDARs in AD,

262

recent experimental data provide increasing evidence

263

of the involvement of presynaptic NMDARs in the

264

enhancement of timing-dependent LTD, resulting in

265

impaired memory functions, which phenomenon may

266

have implications in the development of cognitive

267

decrement in AD [61–63]. With regard to caspase-

268

3 activation, the increased activity of the pyramidal

269

neurons of the entorhinal cortex, the subiculum, and

270

the CA1-3 sector of the hippocampus was found in

271

early stages of AD [64]. The second layer of the

272

entorhinal cortex showed the highest activity. A␤accu-

273

mulation activates NMDARs at early stages of AD

274

[65], andin vitrostudies suggest that this activation

275

might be mediated by GluN2B-containing NMDARs

276

[66]. It has been also demonstrated that NMDARs 277

are connected to neuronal nitric oxide synthase by a ²⁷⁸ scaffolding protein PSD-95 (postsynaptic density pro- ²⁷⁹ tein of molecular weight 95 kDa), which binds to the ²⁸⁰ GluN2B subunit of the NMDAR [67]. Thus, PSD- ²⁸¹ 95 would have an important role in the evocation of ²⁸² downstream excitotoxic events mediated by GluN2B ²⁸³ subunit-containing NMDARs via the production of ²⁸⁴ nitric oxide in an excessive amount [68]. Recent data ²⁸⁵ indicate that the activation of NMDARs by A␤1-42may ²⁸⁶ be secondary to its binding to postsynaptic anchoring ²⁸⁷ proteins such as PSD-95 [42]. Extrasynatptic NMDAR ²⁸⁸ activation triggers the increased production of A␤due ²⁸⁹ to the shift of amyloid␤-protein precursor (A␤PP) pro- ²⁹⁰ duction from A␤PP695 to Kunitz protease inhibitory 291

domain-containing isoforms with higher amyloido- 292

genic potential [69]. This kind of positive feedback ²⁹³ leads to the formation of a vicious circle [70]. GluN2B- ²⁹⁴ mediated neurotransmission also seems to be involved ²⁹⁵ in tau-induced neurotoxicity [71]. Tau phosphorylation ²⁹⁶ causes tau mislocalization and subsequent synaptic ²⁹⁷ impairment as phosphorylated tau can accumulate in ²⁹⁸ dendritic spines, where it may affect the synaptic traf- ²⁹⁹ ficking and/or anchoring of glutamate receptors [72]. ³⁰⁰ The interaction of tau with fyn targets fyn to dendritic ³⁰¹ spines, where it can exert the above-mentioned phos- ³⁰² phorylation of GluN2B subunit of NMDAR, thereby ³⁰³ enhancing the excitotoxic process [73]. In addition to ³⁰⁴ its neuronal effects, A␤also downregulates glutamate ³⁰⁵ uptake capacity of astrocytes and thereby induces a 306

Uncorrected Author Proof

dysfunctional extracellular glutamate clearance [74].

307

Besides the elevated levels of glutamate in the extra-

308

cellular space, the presence of an energy impairment,

309

as a consequence of mitochondrial dysfunction and

310

oxidative stress, would be another causative factor in

311

glutamate excitotoxicity, which leads to a partial mem-

312

brane depolarization resulting in relief of the Mg²⁺

313

blockade of the NMDAR channel and calcium over-

314

load [75].

315

THERAPEUTIC APPROACHES

316

TARGETING THE GLUTAMATERGIC

317

NEUROTRANSMISSION SYSTEM WITH A

318

SPECIAL VIEW OF NMDA RECEPTORS IN

319

ALZHEIMER’S DISEASE: PITFALLS AND

320

POSSIBILITIES

321

The application of agents that completely block

322

NMDAR activity has limited usefulness due to severe

323

clinical side-effects such as hallucinations, agitation,

324

memory impairment, catatonia, nausea, vomiting, a

325

peripheral sensory disturbance, and sympathomimetic

326

effects such as increased blood pressure [76, 77].

327

In order to achieve neuroprotection by targeting the

328

NMDARs in AD, the best therapeutic strategy could

329

be the normalization of synaptic GluN1/GluN2A activ-

330

ity and the abolishment of excitotoxicity mediated

331

by extrasynaptic GluN1/GluN2B subunits. In view of

332

NMDAR antagonists with tolerable side-effects, ion

333

channel blockers with lower affinity, glycine site agents

334

as well as specific antagonists of the polyamine site

335

or the GluN2B subunit may come into play (Fig. 2)

336

[78]. Memantine (3,5-dimethyladamantan-1-amine) is

337

a low affinity open channel blocker, which prefer-

338

entially antagonizes excessively activated NMDARs

339

without affecting physiological NMDAR activity [79].

340

Accordingly, this substance has recently been demon-

341

strated to selectively target mainly GluN2B-containing

342

extrasynaptic NMDARs [80], i.e., it is three times

343

more potent in the inhibition of calcium influx via

344

GluN1/GluN2B than via GluN1/GluN2A subunit-

345

containing NMDARs [81]. Furthermore, memantine

346

concentration-dependently inhibited the expression of

347

Kunitz protease inhibitory domain-containing A␤PP

348

isoforms as well as neuronal production and release

349

of A␤[69, 82]. Accordingly, memantine is a widely

350

applied medicament in the treatment of moderate-

351

advanced stages of AD with beneficial effects as

352

regards language, memory, praxis, and communication

353

dysfunction as well as the activities of daily living [83].

354

Although memantine has some potential side-effects

355

such as somnolence, weight gain, confusion, hyper- ³⁵⁶ tension, nervous system disorders, and falling [84], to ³⁵⁷ date this is the only commercially available NMDAR ³⁵⁸ antagonist in the treatment of AD. In summary, the ³⁵⁹ good effect/side-effect profile would be explained by ³⁶⁰ its pronounced voltage dependency, low affinity, and 361

rapid unblocking kinetics, properties which make the ³⁶² restoration of the desired signal-to-noise ratio in glu- ³⁶³ tamatergic neurotransmission available [85]. ³⁶⁴ Kynurenic acid (KYNA; produced by kynurenine ³⁶⁵ aminotransferases, KATs), a side-product of the main ³⁶⁶ pathway of the tryptophan metabolism, can influ- ³⁶⁷ ence glutamatergic neurotransmission at several levels ³⁶⁸ [86], and exerted neuroprotective effects in several ³⁶⁹ paradigms [86–90]. On the one hand, KYNA can exert ³⁷⁰ wide-spectrum endogenous antagonism of ionotropic ³⁷¹ excitatory amino acid receptors [91], mainly target- ³⁷² ing the strychnine-insensitive glycine-binding site on ³⁷³ the GluN1 subunit of the NMDA receptor [92]. This ³⁷⁴ action requires relatively high (∼10–20␮M) concen- ³⁷⁵ trations of KYNA under physiological conditions [93]; 376

the basal extracellular concentration of KYNA in rats ³⁷⁷ (15–23 nM) [94, 95] is far below the required level ³⁷⁸ to directly interfere with glutamate receptor functions. ³⁷⁹ Accordingly, only excessive elevation of the KYNA ³⁸⁰ level could be accompanied by adverse effects in rats, ³⁸¹ such as reduced exploratory activity, ataxia, stereotypy, ³⁸² sleeping, and respiratory depression, while there was ³⁸³ only a slight effect on the learning ability [96]. How- ³⁸⁴ ever, human postmortem analyses revealed elevated ³⁸⁵ levels of KYNA in the striatum and hippocampus of ³⁸⁶ AD patients [97], alteration of which is suggested to ³⁸⁷ accompany to the cognitive dysfunction in AD rather ³⁸⁸ than to exert a compensatory protective role. Accord- ³⁸⁹ ingly, the achievement of lowering brain KYNA levels ³⁹⁰ by knocking out one of its producing enzyme (KAT 391

II) resulted in the improvement of cognitive functions ³⁹² in mice [98]. With regard to the mechanisms of influ- ³⁹³ encing glutamatergic transmission, on the other hand, ³⁹⁴ KYNA non-competitively blocks the alpha7-nicotinic ³⁹⁵ acetylcholine receptors [99], thereby inhibiting glu- ³⁹⁶ tamate release at the presynatptic site [100]. This ³⁹⁷ blockade can be effective at high nanomolar con- ³⁹⁸ centrations (IC50=∼7␮M), and can also influence ³⁹⁹ hippocampus-dependent cognitive functions [101]. In ⁴⁰⁰ addition to the multiplex receptor antagonism, recent ⁴⁰¹ studies showed that KYNA is capable of facilitating ⁴⁰² AMPA receptor responses in nanomolar concentra- ⁴⁰³ tions [102, 103]. The significance of this phenomenon ⁴⁰⁴

is not really known yet. ⁴⁰⁵

The selective inhibition of GluN2B subunit- 406

containing NMDARs could be another successful ⁴⁰⁷

Uncorrected Author Proof

Fig. 2. Some possibilities of influencing glutamatergic dysfunctioning in Alzheimer’s disease. (␣7-nAChR, alpha7-nicotinic acetylcholine receptors; KYNA, kynurenic acid; NMDAR, N-methyl-D-aspartate receptor;, glutamate, the thickness of the lines represents the extent of inhibition, while dashed lines refers to possible mechanism of action).

strategy in the amelioration of neurodegenerative

408

processes [104]. Ifenprodil (␣-(4-hydroxyphenyl)-␤-

409

methyl-4-benzyl-1-piperidineethanol) is a synthetic

410

negative allosteric modulator of such of receptors,

411

with relatively high affinity (IC50=∼150 nM) [105].

412

Ifenprodil binding seems to interact with polyamine

413

binding in a negative allosteric manner, i.e., it can

414

inhibit the potentiation of NMDAR currents evoked

415

by certain polyamines [106, 107]. It has a consid-

416

erably good side-effect profile: only mouth dryness,

417

nausea, headache, and palpitations were observed.

418

Accordingly, several derivatives, including Ro 25-

419

6981 ([R-(R^∗,S^∗)]-␣-(4-hydroxyphenyl)-␤-methyl-4-

420

benzyl-1-piperidinepropanol), have been synthesized

421

with the aim of presenting lead compounds in pharma-

422

ceutical development in the field of neurodegenerative

423

disorders [104]. With regard to AD, A␤-induced endo-

424

plasmic reticulum and oxidative stress was prevented

425

by ifenprodil [108]. Furthermore, this substance and

426

Ro 25-6981 also prevented the A␤-mediated inhibi-

427

tion of LTP in rodent hippocampal slices [109–112].

428

Indeed, Ro 25-6981 abolished LTD enhancement and

429

learning impairment in rats as well [113]. Evotect’s ⁴³⁰ EVT 101, another GluN2B antagonist which has been ⁴³¹ shown to penetrate into the human brain, was well ⁴³² tolerated in a double-blind, 4-week phase Ib study ⁴³³

(http://www.evotec.com). ⁴³⁴

A possible pharmaceutical modification of KYNA ⁴³⁵ is amidation at the carboxyl moiety [114, 115]. ⁴³⁶ The resulting KYNA amides may be of special ⁴³⁷ interest since they have been shown to preferen- 438

tially act on GluN2B subunit-containing extrasynaptic 439

NMDARs [116]. This feature may also offer the ⁴⁴⁰ opportunity to establish an extracellular concen- ⁴⁴¹ tration that is capable of inhibiting the tonic ⁴⁴² extrasynaptic NMDAR currents without impairing ⁴⁴³ synaptic glutamatergic neurotransmission. Accord- ⁴⁴⁴ ingly, one of the KYNA amide compounds synthesized ⁴⁴⁵ by our group, N-(2-N,N-dimethylaminoethyl)-4-oxo- ⁴⁴⁶ 1H-quinoline-2-carboxamide hydrochloride exerted ⁴⁴⁷ protective effects both in the four-vessel occlu- ⁴⁴⁸ sion model of cerebral ischemia (rats; [117]) and ⁴⁴⁹ in the N171-82Q transgenic mouse model of HD ⁴⁵⁰

[118]. ⁴⁵¹

Uncorrected Author Proof

Finally, in addition to directly influencing

452

NMDARs, it is important to mention that there

453

are some indirect regulators of NMDAR function-

454

ing, targeting of which can be used as alternative

455

therapeutic approaches in the amelioration of gluta-

456

matergic dysfunction in AD. These targets include

457

some metabotropic glutamatergic receptors [119] and

458

certain adenosine receptors [120, 121].

459

CONCLUSION

460

Although more and more details are being revealed

461

regarding the pathomechanism of AD, the recent

462

therapeutic strategies are restricted only to few

463

pharmaceutical agents. The glutamatergic system is

464

presumed to be the major altered neurotransmitter

465

system in AD; therefore, there is a great need for

466

the development of pharmakons targeting this system

467

with acceptable side-effect profile. From this respect,

468

ion channel blockers with lower affinity as well as

469

GluN2B subunit specific antagonists might be the

470

most promising candidates for future AD therapy.

471

Although the present short review focused on the pos-

472

sibilities of therapeutic amelioration via targeting the

473

glutamatergic neurotransmission system with special

474

attention to NMDARs, it should be noted that achiev-

475

ing neuroprotection in AD—especially in terms of

476

‘synaptoprotection’—is a complex issue, with phar-

477

macological targets and approaches we could not detail

478

here, but have already been comprehensively discussed

479

by others [122, 123].

480

ACKNOWLEDGMENTS

481

This work was supported by the projects OTKA (K

482

75628), KTIA NAP 13 – Hungarian National Brain

483

Research Program and T ´AMOP-4.2.2.A-11/1/KONV-

484

2012-0052. Furthermore, this research was realized

485

in the frames of T ´AMOP 4.2.4. A/1-11-1-2012-0001

486

“National Excellence Program – Elaborating and oper-

487

ating an inland student and researcher personal support

488

system”. The project was subsidized by the European

489

Union and co-financed by the European Social Fund.

490

Authors’ disclosures available online (http://www.

491

j-alz.com/disclosures/view.php?id=2158).

492

REFERENCES

493

[1] Blennow K, de Leon MJ, Zetterberg H (2006) Alzheimer’s

494

disease.Lancet368, 387-403.

495

[2] Zadori D, Datki Z, Penke B (2007) [The role of chronic

496

brain hypoperfusion in the pathogenesis of Alzheimer’s

497

disease–facts and hypotheses]. Ideggyogy Sz 60, 498

428-437. 499

[3] Scott KR, Barrett AM (2007) Dementia syndromes: Evalu- 500

ation and treatment.Expert Rev Neurother7, 407-422. 501

[4] Hebert LE, Scherr PA, Bienias JL, Bennett DA, Evans DA 502

(2003) Alzheimer disease in the US population: Prevalence 503

estimates using the 2000 census.Arch Neurol60, 1119- 504

1122. 505

[5] Braak H, Braak E (1991) Neuropathological stageing of 506

Alzheimer-related changes.Acta Neuropathol82, 239-259. 507

[6] Vogels OJ, Broere CA, ter Laak HJ, ten Donkelaar HJ, 508

Nieuwenhuys R, Schulte BP (1990) Cell loss and shrink- 509

age in the nucleus basalis Meynert complex in Alzheimer’s 510

disease.Neurobiol Aging11, 3-13. 511

[7] Neill D (1995) Alzheimer’s disease: Maladaptive synapto- 512

plasticity hypothesis.Neurodegeneration4, 217-232. 513

[8] Geddes JW, Tekirian TL, Soultanian NS, Ashford JW, Davis 514

DG, Markesbery WR (1997) Comparison of neuropatho- 515

logic criteria for the diagnosis of Alzheimer’s disease. 516

Neurobiol Aging18, S99-105. 517

[9] Mortimer JA, Borenstein AR, Gosche KM, Snowdon DA 518

(2005) Very early detection of Alzheimer neuropathology 519

and the role of brain reserve in modifying its clinical expres- 520

sion.J Geriatr Psychiatry Neurol18, 218-223. 521

[10] Perl DP, Purohit DP (1997) Proposal to revise the mor- 522

phologic criteria for the diagnosis of Alzheimer’s disease. 523

Neurobiol Aging18, S81-S84. 524

[11] Van Hoesen G, Pandya DN (1975) Some connections of the 525

entorhinal (area 28) and perirhinal (area 35) cortices of the 526

rhesus monkey. I. Temporal lobe afferents.Brain Res95, 527

1-24. 528

[12] Van Hoesen G, Pandya DN, Butters N (1975) Some con- 529

nections of the entorhinal (area 28) and perirhinal (area 35) 530

cortices of the rhesus monkey. II. Frontal lobe afferents. 531

Brain Res95, 25-38. 532

[13] Van Hoesen GW, Pandya DN (1975) Some connections of 533

the entorhinal (area 28) and perirhinal (area 35) cortices of 534

the rhesus monkey. III. Efferent connections.Brain Res95, 535

39-59. 536

[14] Gomez-Isla T, Price JL, McKeel DW Jr, Morris JC, Growdon 537

JH, Hyman BT (1996) Profound loss of layer II entorhinal 538

cortex neurons occurs in very mild Alzheimer’s disease.J 539

Neurosci16, 4491-4500. 540

[15] Price JL, Ko AI, Wade MJ, Tsou SK, McKeel DW, Morris JC 541

(2001) Neuron number in the entorhinal cortex and CA1 in 542

preclinical Alzheimer disease.Arch Neurol58, 1395-1402. 543

[16] Arriagada PV, Growdon JH, Hedley-Whyte ET, Hyman BT 544

(1992) Neurofibrillary tangles but not senile plaques parallel 545

duration and severity of Alzheimer’s disease.Neurology42, 546

631-639. 547

[17] Hooper WM, Vogel FS (1976) The limbic system in 548

Alzheimer’s disease.Am J Pathol85, 1-19. 549

[18] Augustinack JC, Helmer K, Huber KE, Kakunoori S, Zollei 550

L, Fischl B (2010) Direct visualization of the perforant 551

pathway in the human brain withex vivodiffusion tensor 552

imaging.Front Hum Neurosci4, 42. 553

[19] Scheff SW, Price DA, Schmitt FA, Mufson EJ (2006) 554

Hippocampal synaptic loss in early Alzheimer’s disease 555

and mild cognitive impairment. Neurobiol Aging 27, 556

1372-1384. 557

[20] Witter MP, Wouterlood FG, Naber PA, Van Haeften T 558

(2000) Anatomical organization of the parahippocampal- 559

hippocampal network.Ann N Y Acad Sci911, 1-24. 560

[21] Scheff SW, Price DA, Schmitt FA, DeKosky ST, Mufson 561

EJ (2007) Synaptic alterations in CA1 in mild Alzheimer 562

Uncorrected Author Proof

disease and mild cognitive impairment.Neurology68, 1501-

563

1508.

564

[22] Braak H, Braak E, Yilmazer D, Bohl J (1996) Functional

565

anatomy of human hippocampal formation and related struc-

566

tures.J Child Neurol11, 265-275.

567

[23] Shah A, Jhawar SS, Goel A (2012) Analysis of the anatomy

568

of the Papez circuit and adjoining limbic system by fiber

569

dissection techniques.J Clin Neurosci19, 289-298.

570

[24] Kohler C (1985) Intrinsic projections of the retrohippocam-

571

pal region in the rat brain. I. The subicular complex.J Comp

572

Neurol236, 504-522.

573

[25] Sah P, Faber ES, Lopez De Armentia M, Power J (2003) The

574

amygdaloid complex: Anatomy and physiology.Physiol Rev

575

83, 803-834.

576

[26] McDonald AJ (1998) Cortical pathways to the mammalian

577

amygdala.Prog Neurobiol55, 257-332.

578

[27] Milner B, Squire LR, Kandel ER (1998) Cognitive neuro-

579

science and the study of memory.Neuron20, 445-468.

580

[28] Khachaturian ZS (1985) Diagnosis of Alzheimer’s disease.

581

Arch Neurol42, 1097-1105.

582

[29] Mirra SS, Heyman A, McKeel D, Sumi SM, Crain BJ,

583

Brownlee LM, Vogel FS, Hughes JP, van Belle G, Berg

584

L (1991) The Consortium to Establish a Registry for

585

Alzheimer’s Disease (CERAD). Part II. Standardization

586

of the neuropathologic assessment of Alzheimer’s disease.

587

Neurology41, 479-486.

588

[30] Kirvell SL, Esiri M, Francis PT (2006) Down-regulation

589

of vesicular glutamate transporters precedes cell loss and

590

pathology in Alzheimer’s disease.J Neurochem98, 939-

591

950.

592

[31] Kashani A, Lepicard E, Poirel O, Videau C, David JP, Fallet-

593

Bianco C, Simon A, Delacourte A, Giros B, Epelbaum J,

594

Betancur C, El Mestikawy S (2008) Loss of VGLUT1 and

595

VGLUT2 in the prefrontal cortex is correlated with cognitive

596

decline in Alzheimer disease.Neurobiol Aging29, 1619-

597

1630.

598

[32] Mitew S, Kirkcaldie MT, Dickson TC, Vickers JC (2013)

599

Altered synapses and gliotransmission in Alzheimer’s dis-

600

ease and AD model mice.Neurobiol Aging34, 2341-2351.

601

[33] Cassano T, Serviddio G, Gaetani S, Romano A, Dipasquale

602

P, Cianci S, Bellanti F, Laconca L, Romano AD, Padalino

603

I, LaFerla FM, Nicoletti F, Cuomo V, Vendemiale G (2012)

604

Glutamatergic alterations and mitochondrial impairment in

605

a murine model of Alzheimer disease.Neurobiol Aging33,

606

1121 e1121-1112.

607

[34] Canas PM, Simoes AP, Rodrigues RJ, Cunha RA (2014)

608

Predominant loss of glutamatergic terminal markers in a

609

beta-amyloid peptide model of Alzheimer’s disease.Neu-

610

ropharmacology76(Pt A), 51-56.

611

[35] Jellinger KA (2012) Neuropathological subtypes of

612

Alzheimer’s disease.Acta Neuropathol123, 153-154.

613

[36] Mota SI, Ferreira IL, Rego AC (2014) Dysfunctional

614

synapse in Alzheimer’s disease – A focus on NMDA recep-

615

tors.Neuropharmacology76(Pt A), 16-26.

616

[37] Rank KB, Pauley AM, Bhattacharya K, Wang Z, Evans DB,

617

Fleck TJ, Johnston JA, Sharma SK (2002) Direct interaction

618

of soluble human recombinant tau protein with Abeta 1-42

619

results in tau aggregation and hyperphosphorylation by tau

620

protein kinase II.FEBS Lett514, 263-268.

621

[38] Crimins JL, Pooler A, Polydoro M, Luebke JI, Spires-Jones

622

TL (2013) The intersection of amyloid beta and tau in gluta-

623

matergic synaptic dysfunction and collapse in Alzheimer’s

624

disease.Ageing Res Rev12, 757-763.

625

[39] Puzzo D, Privitera L, Fa M, Staniszewski A, Hashimoto G,

626

Aziz F, Sakurai M, Ribe EM, Troy CM, Mercken M, Jung

627

SS, Palmeri A, Arancio O (2011) Endogenous amyloid- 628

beta is necessary for hippocampal synaptic plasticity and 629

memory.Ann Neurol69, 819-830. 630

[40] Abramov E, Dolev I, Fogel H, Ciccotosto GD, Ruff E, 631

Slutsky I (2009) Amyloid-beta as a positive endogenous reg- 632

ulator of release probability at hippocampal synapses.Nat 633

Neurosci12, 1567-1576. 634

[41] Sokolow S, Luu SH, Nandy K, Miller CA, Vinters HV, 635

Poon WW, Gylys KH (2012) Preferential accumulation 636

of amyloid-beta in presynaptic glutamatergic terminals 637

(VGluT1 and VGluT2) in Alzheimer’s disease cortex.Neu- 638

robiol Dis45, 381-387. 639

[42] Lacor PN, Buniel MC, Furlow PW, Clemente AS, Velasco 640

PT, Wood M, Viola KL, Klein WL (2007) Abeta oligomer- 641

induced aberrations in synapse composition, shape, and 642

density provide a molecular basis for loss of connectivity 643

in Alzheimer’s disease.J Neurosci27, 796-807. 644

[43] Yu W, Lu B (2012) Synapses and dendritic spines as 645

pathogenic targets in Alzheimer’s disease. Neural Plast 646

2012, 247150. 647

[44] Lesne S, Ali C, Gabriel C, Croci N, MacKenzie ET, Glabe 648

CG, Plotkine M, Marchand-Verrecchia C, Vivien D, Buisson 649

A (2005) NMDA receptor activation inhibits alpha-secretase 650

and promotes neuronal amyloid-beta production.J Neurosci 651

25, 9367-9377. 652

[45] Tovar KR, Westbrook GL (1999) The incorporation of 653

NMDA receptors with a distinct subunit composition at 654

nascent hippocampal synapsesin vitro.J Neurosci19, 4180- 655

4188. 656

[46] Goebel-Goody SM, Davies KD, Alvestad Linger RM, 657

Freund RK, Browning MD (2009) Phospho-regulation of 658

synaptic and extrasynaptic N-methyl-d-aspartate receptors 659

in adult hippocampal slices.Neuroscience158, 1446-1459. 660

[47] Um JW, Nygaard HB, Heiss JK, Kostylev MA, Stagi M, 661

Vortmeyer A, Wisniewski T, Gunther EC, Strittmatter SM 662

(2012) Alzheimer amyloid-beta oligomer bound to postsy- 663

naptic prion protein activates Fyn to impair neurons.Nat 664

Neurosci15, 1227-1235. 665

[48] Beraldo FH, Arantes CP, Santos TG, Machado CF, Roffe 666

M, Hajj GN, Lee KS, Magalhaes AC, Caetano FA, Mancini 667

GL, Lopes MH, Americo TA, Magdesian MH, Ferguson 668

SS, Linden R, Prado MA, Martins VR (2011) Metabotropic 669

glutamate receptors transduce signals for neurite outgrowth 670

after binding of the prion protein to laminin gamma1 chain. 671

FASEB J25, 265-279. 672

[49] Hardingham GE, Fukunaga Y, Bading H (2002) Extrasy- 673

naptic NMDARs oppose synaptic NMDARs by triggering 674

CREB shut-off and cell death pathways.Nat Neurosci5, 675

405-414. 676

[50] Zhou Q, Sheng M (2013) NMDA receptors in nervous sys- 677

tem diseases.Neuropharmacology74, 69-75. 678

[51] Baudry M, Lynch G (2001) Remembrance of arguments 679

past: How well is the glutamate receptor hypothesis of LTP 680

holding up after 20 years?Neurobiol Learn Mem76, 284- 681

297. 682

[52] Escobar ML, Derrick B (2007) Long-term potentiation and 683

depression as putative mechanisms for memory formation. 684

In Neural Plasticity and Memory: From Genes to Brain 685

Imaging, Berm´udez-Rattoni F, ed. CRC Press, Boca Raton 686

(FL), pp. 15-45. 687

[53] Wang HW, Pasternak JF, Kuo H, Ristic H, Lambert MP, 688

Chromy B, Viola KL, Klein WL, Stine WB, Krafft GA, 689

Trommer BL (2002) Soluble oligomers of beta amyloid 690

(1-42) inhibit long-term potentiation but not long-term 691

depression in rat dentate gyrus.Brain Res924, 133-140. 692