1
2
POST-ISCHEMIC TREATMENT WITH L-KYNURENINE SULFATE
3
EXACERBATES NEURONAL DAMAGE AFTER TRANSIENT MIDDLE
4
CEREBRAL ARTERY OCCLUSION
5 L.GELLE´RT,aL.KNAPP,aK.NE´METH,bJ.HERE´DI,a
6 D.VARGA,aG.OLA´H,aK.KOCSIS,aA´.MENYHA´RT,a
7 Z.KIS,aT.FARKAS,aL.VE´CSEIc,dANDJ.TOLDIa*
8 aDepartment of Physiology, Anatomy and Neuroscience, 9 University of Szeged, Ko¨ze´p fasor 52, H-6726 Szeged, Hungary 10 bDepartment of Cognitive Science, University of Technology and 11 Economics, Egry Jo´zsef utca 1, T building V. 506, H-1111 Budapest, 12 Hungary
13 cDepartment of Neurology, University of Szeged, Semmelweis u.
14 6, H-6725 Szeged, Hungary
15 dNeurology Research Group of the Hungarian Academy of 16 Science and University ofSzeged, Hungary
17 Abstract—Since brain ischemia is one of the leading causes of adult disability and death, neuroprotection of the ische- mic brain is of particular importance. Acute neuroprotective strategies usually have the aim of suppressing glutamate excitotoxicity and an excessive N-methyl-D-aspartate (NMDA)receptor function. Clinically tolerated antagonists should antagonize an excessive NMDA receptor function without compromising the normal synaptic function.
Kynurenic acid (KYNA) an endogenous metabolite of the tryptophan metabolism, may be an attractive neuroprotec- tant in this regard. The manipulation of brain KYNA levels was earlier found to effectively enhance thehistopatholo- gical outcome of experimental ischemic/hypoxic states.
The present investigation of the neuroprotective capacity of L-kynurenine sulfate (L-KYNs) administered systemically after reperfusion in a novel distal middle cerebral artery occlusion (dMCAO) model of focal ischemia/reperfusion revealed that in contrast with earlier results, treatment with L-KYNs worsened the histopathological outcome of dMCAO. This contradictory result indicates that post-ische- mic treatment with L-KYNs may be harmful.Ó2013 Published by Elsevier Ltd. on behalf of IBRO.
Key words: focalcerebral ischemia,neuroprotection,glycine co-agonist site, NMDAR, MCAO model,kynurenines.
18
INTRODUCTION 19
As a result of the high energy demands of the central 20
nervous system (CNS), a deprivation of oxygen and 21
glucose leads in a short time to abnormal glutamatergic 22
transmission. Malfunctioning of the ATP-dependent 23
transporters results in a disturbance of ionic 24
homeostasis, depolarization and the excessive release 25
of glutamate from neural and glial stores in the 26
extrasynaptic space. Acute or prolonged over-activation 27
ofN-methyl-D-aspartate receptors (NMDARs) allows the 28
excessive entry of Ca2+, initiating glutamate 29
excitotoxicity, the common core feature of many 30
neuropsychiatric disorders, including stroke, epilepsy, 31
Alzheimer’s disease and Huntington’s disease (Endres 32
and Dirnagl, 2002; Moskowitz et al., 2010). 33
Neuroprotective strategies usually have the aim of 34
suppressing an excessive NMDAR function. Indeed, a 35
number of NMDA antagonists have proven to be robust 36
neuroprotectants in animal models of an ischemic/ 37
hypoxic (I/H) state, but many failed in clinical trials in 38
consequence of their adverse side-effects (Ikonomidou 39
and Turski, 2002; Muir, 2006). 40
The destructive effect of NMDAR over-activity is in 41
contrast with the phenomenon that synaptic NMDAR 42
activity mediates the survival of several types of 43
neurons (Hetman and Kharebava, 2006; Hardingham, 44
2009). It has been reported that neurodegeneration in 45
the basal ganglia is exacerbated by NMDAR antagonists 46
(Ikonomidou et al., 2000), that an NMDAR antagonist 47
enhanced apoptotic cell loss in a head trauma model 48
(Pohl et al., 1999), and that synaptic NMDAR activity 49
boosts intrinsic antioxidant defenses (Papadia et al., 50
2008). Furthermore, the targeting of ischemic brain 51
areas by global NMDAR antagonism can confuse the 52
functioning of brain areas unaffected by ischemic 53
damage (Gunduz-Bruce, 2009). In this regard, a 54
clinically tolerated neuroprotectant should antagonize 55
the NMDAR function when it is excessive, but not later, 56
without compromising the normal synaptic function. 57
NMDAR activation requires the definite depolarization 58
of the cell and the presence of both glutamate and the full 59
co-agonists glycine or D-serine (Kussius and Popescu, 60
2009; Papouin et al., 2012). Furthermore, the glycine 61
co-agonist site is not saturated under physiological 62
conditions, but is in a hyperactive state (Li et al., 2009; 63
Fuchs et al., 2012). Glycine-site antagonists may be 64
attractive neuroprotectants in this respect. 65
0306-4522/13 $36.00Ó2013 Published by Elsevier Ltd. on behalf of IBRO.
http://dx.doi.org/10.1016/j.neuroscience.2013.04.063
*Corresponding author. Tel: +36-62-544153; fax: +36-63-544291.
E-mail address:toldi@bio.u-szeged.hu(J. Toldi).
Abbreviations: 3-HK, 3-hydroxykynurenine; CNS, central nervous system; dMCA, distal middle cerebral artery; dMCAO, distal middle cerebral artery occlusion; EEG, electroencephalography; I/H, ischemic/
hypoxic; KYNA, kynurenic acid; L-KYNs, L-kynurenine sulfate; NDS, normal donkey serum; NMDA, N-methyl-D-aspartate; NMDAR, N-methyl-D-aspartate receptors; PB, phosphate buffer; QUIN, quinolinic acid.
Q2 Q1
Neurosciencexxx (2013) xxx–xxx
1
66 Kynurenic acid (KYNA) is an endogenous metabolite
67 of the tryptophan metabolism. It is produced from its
68 precursor L-kynurenine (KYN) by the enzyme
69 kynurenine-aminotransferase II (KATII), and discharged
70 from the astrocytes in the CNS (Swartz et al., 1990).
71 KYNA is a competitive antagonist at theglycine/D-serine
72 co-agonist site of the NMDAR. Furthermore, it plays a
73 versatile role in pathological states, including
74 inflammatory (Moroni et al., 2012), vascular (Sas et al.,
75 2003) and antioxidant (Lugo-Huitron et al.) processes.
76 Acting on thea7 nicotinic acetylcholine receptor, KYNA
77 also influences the excitability of neurons (Banerjee et
78 al.). A huge body of evidence indicates that
79 manipulation of the brain KYNA levels can effectively
80 ameliorate the histopathological outcome of
81 experimental I/H state (Stone, 2000; Wu et al., 2000;
82 Schwarcz and Pellicciari, 2002; Stone and Addae, 2002;
83 Vamos et al., 2009; Zadori et al., 2009). The
84 neuromodulatory properties of KYNA are now well-
85 established (Vecsei et al., 2012).
86 In the present study, we investigated whether
87 L-kynurenine sulfate (L-KYNs) administered after
88 reperfusion (in a dose, formerly proved to be
89 neuroprotective) diminishes the neuronal damage
90 triggered by short-term occlusion of the distal middle
91 cerebral artery (dMCA) in the rat cerebral
92 somatosensory cortex. This novel dMCA occlusion
93 (dMCAO) model was recently developed and
94 characterized from histological and electrophysiological
95 aspects in our research group (L. Knapp, manuscript
96 under review).
97 EXPERIMENTAL PROCEDURES
98 Animals
99 Male Wistar rats (n= 23) weighing 200–250g were
100 used. The animals were kept under controlled laboratory
101 conditions with free access to food and water. The
102 experiments were carried out in accordance with the
103 protocol for animal care approved by both the
104 Hungarian Health Committee (1998) and the European
105 Communities Council Directive (86/609/EEC).
106 Surgical procedure
107 Experiments were carried out under Nembutal
108 anesthesia. The body temperature was maintained at
109 37 ± 0.5°C with a self-regulating heating pad and rectal
110 probe (Supertech TMP-5a). The animals were fixed in a
111 stereotaxic headholder (David Kopf Instr.) and the left
112 masticatory muscle was removed. The surface of the
113 temporal skull was cleaned and the brain was exposed
114 with a high-speed microdrill. The exposed cortical
115 surface involved the trunk and main branches of the
116 MCA. To induce ischemia, the MCA was carefully lifted
117 through 1200lm with a Fisher microsurgery hook with
118 the aid of a micromanipulator, and occluded for 30 min.
119 To terminate the occlusion, the hook was carefully
120 removed, and restoration of the blood flow was
121 confirmed under an operating microscope. Finally, the
122 dura and the temporal muscle were replaced, the skin
was closed with a silk suture and the wound was 123
cleaned with iodine solution. All interventions were 124
strictly synchronized in time, to make the effect of 125
Nembutal on the experiment uniform. 126
Electrophysiology 127
60sofelectroencephalography (EEG) was recorded on 128
the surface of the skull with a silver electrode (2 mm 129
lateral to the sutura sagittalis and 3 mm behind the 130
bregma), promptly before and in the29th–30thmin after 131
dMCAO (sampling rate: 1024 Hz; gain: 1000) with 132
Experimetria NeuroSys software (Experimetria Ltd., 133
Hungary). 134
EEG power analysis was performed with the EEGLab 135
toolbox (Delorme and Makeig, 2004) and custom-written 136
MATLAB 7.1 (Mathworks, Natick, Massachusetts, USA) 137
software. 138
The range of frequency of interest was assigned to 139
2–20Hz and further analysis was performed within this 140
range. 141
Histology 142
Tissue processing. For the histological study, 5 days 143
after dMCAO, animals were anesthetized with an 144
overdose of urethane and perfused transcardially with 145
ice-cold phosphate buffer (PB, 0.1 M, pH 7.4) and 4% 146
paraformaldehyde (dissolved in 0.1 M PB, pH 7.4). The 147
brains were removed and postfixed overnight in 148
paraformaldehyde. On the next day, 20-lm coronal 149
sections were obtained with a vibratome (Leica VT1000 150
S) between0.5 and 4 mm behind the bregma (Paxinos 151
et al., 1980). Two adjacent slices were collected in 152
500-lm steps, one for double immunostaining and the 153
other for Fluoro Jade-C staining. Fluorescent 154
photomicrographs were obtained with an Olympus BX51 155
microscope fitted with a DP70 digital imaging system. 156
Fluoro Jade-C staining. Fluoro Jade-C (FJ-C) staining 157
was performed with the literature protocol (Schmued 158
et al., 2005) with some modification. The slices were 159
mounted on gelatine-coated slides, then coverslipped 160
with Fluoromount. FJ-C-positive (FJ-C+) cells were 161
counted in the ispilateral cortex at 40 magnification. 162
Automated counting of FJ-C+ cells was performed with 163
custom-written software in MATLAB 7.1 (Mathworks, 164
Natick, Massachusetts, USA). After automated threshold 165
adjustment and noise reduction, 25–400-lm2 166
fluorescent objects wereacceptedas cells and counted 167
in binary images. 168
Immunohistochemistry. Glial reaction was detected 169
with an indirect immunohistochemical method. 20-lm- 170
thick free-floating sections were washed in PB, and then 171
incubated in 10% normal donkey serum (NDS). For the 172
detection of activated microglia (mouse anti-CD11b, 173
clone OX42, 1:1000, Millipore) and reactive astrocytes 174
(rabbit anti-S100, 1:2000, DAKO), sections were 175
exposed to the primary antibodies overnight at 4°C, and 176
to the appropriate secondary antibodies for 2 h at room 177
178 temperature. Primary and secondary antibodies were
179 diluted in 0.1 M PB containing 0.4% Triton-X100, 2%
180 NDS and 0.01 % sodium azide. The sections were
181 coverslipped with an aqueous mounting medium.
182 Drug administration
183 The rats were divided into two groups: L-KYNs-treated
184 animals(n= 11)received 300 mg/kg L-KYNs (dissolved
185 in 5% NaOH, pH 7.4) intraperitoneally, immediately after
186 reperfusion, while the control animals (n= 12) were
187 treated with the vehicle.
188 All chemicals were purchased from VWR Ltd.,
189 Hungary, and Sigma, St. Louis,MO, USA.
190 Statistical analysis
191 Electrophysiology. EEG power spectra filtered at
192 2–20Hz were decomposed at 1-Hz intervals. The EEG
193 power of a given frequency was considered as an
194 individual case. Analysis was performed with General
195 LinearModel/Repeated measures (IBM SPSS Statistics
196 version 20).
197 Histology. Numbers of FJ-C+ cells were compared
198 with the General Linear Model. The effects of the
199 different rats were used as random effects and the
200 different treatments were used as fixed effects in the
201 mixed effect linear model (IBM SPSS Statistics version
202 20).
203 RESULTS
204 Electrophysiology
205 The EEG registered for 60 s filtered for2–20Hz revealed
206 a marked and characteristic change in EEG during
207 dMCAO (Fig. 1). The power values in each frequency
208 bin were submitted to separate repeated-measures
209 analysis of variance, with period and frequency as
210 within-subject factors. All effects with two or more
211 degrees of freedom were adjusted for violations of
212 sphericity according to the Greenhouse–Geisser
213 correction.
The ischemic period significantly reduced the power of 214
the signal as compared with the power of the EEG 215
registered before ischemia (main effect of period: 216
F(1,21) = 32.989, p< 0.0001, g2= 0.61; Fig. 2, panel 217
A, B). It was earlier observed that somatosensory- 218
evoked responses disappearcompletely during dMCAO 219
(L. Knapp, manuscript under review). Together, these 220
data indicate, that the dMCAO in our model resulted in 221
a clean-cut decay of activity in the somatosensory 222
cortices, i.e. the animals underwent a 30-min I/H period. 223
Histology 224
After a 5-day survival period, definite FJ-C staining and 225
astrocyte/microglial activation throughout the 226
somatosensory cortices emerged in approximately half 227
of the animals, ipsilateral to the dMCAO (6/12 of the 228
saline-treated animals; 5/11 of the L-KYNs-treated 229
animals). In the remaining animals, no FJ-C staining 230
and no glial reaction were observed, i.e. complete 231
staining negativity. Ipsilateral to the dMCAO, astrocyte 232
activation was characterized by hypertrophic astrocytes 233
with prominent, thick processes and small vacuoles in 234
the cell bodies as compared with the contralateral cortex 235
(Fig. 3, panel A and insert). The microglia also revealed 236
the activated phenotype ipsilateral to the dMCAO. 237
Enlarged somata and the loss of secondary and tertiary 238
branching were characteristic (Fig 3, panel B and 239
insert). The glial reaction was more prominent in the 240
L-KYNs-treated group (visual observation). The FJ-C 241
staining distribution was similar to that in the activated 242
microglia (compare Fig. 3, panels B, C). The groups 243
were compared quantitatively for FJ-C staining. The 244
number of FJ-C+ neurons was significantly higher in 245
the L-KYNs-treated group (Fig. 4, General Linear Model; 246
p= 0.023). 247
DISCUSSION 248
Physiological glutamatergic transmission through 249
NMDARs is essential in the brain, playing a key role in 250
development and synaptic plasticity. Due to its high 251
permeability for Ca2+, the NMDAR is linked to several 252
cell-signaling pathways, and to learning and memory 253
(Nakazawa et al., 2004; Zhang et al., 2007). In certain 254
acute and chronic neuropsychiatric disorders, however, 255
Ca2+ entry is the key mediator of glutamate 256
excitotoxicity and the NMDAR is the primary source of a 257
toxic Ca2+ influx (Stanika et al., 2012). NMDAR 258
antagonism is therefore an obvious neuroprotective 259
approach. 260
The failure of numerous antagonists in clinical trials is 261
due in part to the different roles of synaptic and 262
extrasynaptic NMDARs during excitotoxic processes. 263
The hypothesis that extrasynaptic NMDARs mediate cell 264
death, while synaptic NMDARs may promote survival 265
was recently discussed (Hardingham and Bading, 2010; 266
Li and Ju, 2012). From this respect, the selective 267
targeting of extrasynaptic receptors without interfering 268
with the normal synaptic function will involve a great 269
advance (Chen and Lipton, 2006). 270 Fig. 1.EEG recordings from a rat somatosensory cortex ipsilateral to
the dMCAO. 60-s EEG recordings during control (black) and ischemic (red) periods are superimposed. The EEG filtered for 2–20 Hz revealed a marked change during dMCAO. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
271 It has been argued that systemically administered
272 L-KYNs is neuroprotective in different I/H states (Gigler
273 et al., 2007; Sas et al., 2008). In such experiments, the
I/H model triggered massive excitotoxicity and a high- 274
level, long-lasting glutamate spillover. On the other 275
hand, pre-ischemic treatment was effective, since the 276 Fig. 2.(Panel A) EEG power decomposed at 1-Hz intervals. Lines demonstrate the EEG power of given frequencies during the control (line) and ischemic (dashed line) period (mean ± S.E.M.). (Panel B) The EEG power decreased significantly during the ischemic period (Repeated measures:
F(1,21) = 32.989,p< 0.0001,g2= 0.61; mean ± S.E.M.).
Fig. 3.Representative photomicrograph of the rat somatosensory cortex ipsilateral to the dMCAO after L-KYNs treatment. Double immunostaining of reactive astrocytes and microglia from the same slice; FJ-C staining from the adjacent slice (100magnification, scale bars = 500lm).
Astrocyte activation was characterized by hypertrophic astrocytes with prominent, thick processes and small vacuoles in the cell bodies (panel A, and insert). The microglia also revealed an activated phenotype, enlarged somata and the loss of secondary and tertiary branching (panel B, and insert). A high number of FJ-C+ neurons were seen throughout the cortex (panel C). The FJ-C staining pattern closely followed the microglia distribution (compare panels B and C).
Fig. 4.FJ-C+ cells counted in the rat somatosensory cortex ipsilateral to the dMCAO. The numbers of FJ-C+ cells were compared with the General Linear Model, and plotted in a bar chart. The cell number was significantly higher in the L-KYNs-treated group (General Linear Model;
p= 0.023; mean ± S.E.M.).
277 KYN/KYNA transition in the astrocytes is time-consuming
278 (Swartz et al., 1990).
279 However, in a recent study we showed that a KYNA
280 derivative significantly diminished hippocampal
281 neurodegeneration, even if administered at the time of
282 reperfusion (Gellert et al., 2011).
283 A relatively brief MCAO evokes clean-cut
284 neurodegeneration in only a fraction of the animals
285 (Memezawa et al., 1992; Aspey et al., 2000; Popp
286 et al., 2009). Similarly, in our experiment only half of the
287 animals exhibited neurodegeneration, irrespectively of
288 whether they received L-KYNs or saline treatment.
289 However, the amplitude of the evoked responses (L.
290 Knapp, manuscript under review) and the EEG power
291 decreased markedly during dMCAO, and it may be
292 therefore postulated, that the somatosensory cortices
293 were subjected to an I/H state. This indicates that
294 endogenous protective processes are able to withstand
295 a short I/H state in this cortical area.
296 Systemic treatment with L-KYNs in our experiment did
297 not alter the probability of occurrence of
298 neurodegeneration, but extended the damaged area,
299 the glial activation and the number of FJ-C+ cells in the
300 animals, which ignited cell-death pathways.
301 Around one-quarter of the extrasynaptic NMDARs in
302 adult hippocampal slices are perisynaptic (within 100 nm
303 of the postsynaptic density). Of the dendritically
304 localized extrasynaptic NMDARs, around one-third is
305 adjacent to glia-like processes (Petralia et al., 2010).
306 KYNA produced in the glia may therefore, antagonize
307 both synaptic and extrasynaptic NMDARs, influencing
308 pro-death or survival mechanisms, respectively.
309 The emergence of KYNA produced de novo from
310 systemically administered L-KYNs takes time that is
311 considerable from the aspect of an excitotoxic process
312 (Swartz et al., 1990). However, KYNA or KYNA analogs
313 can act quickly after administration. Furthermore, during
314 a brief I/H state the presence of excessive glutamate
315 and concomitant extrasynaptic NMDAR activation can
316 last for minutes (Benveniste et al., 1984; Ikonomidou
317 and Turski, 2002). The phenomenon that the KYNA
318 analog, but not L-KYNs, is neuroprotective when
319 administered after reperfusion may depend on the
320 intensity and duration of the I/H state, the concomitant
321 glutamate spillover, and the duration of the KYN-KYNA
322 turnover.
323 Another possible explanation would be that L-KYNs
324 administration led to the increased concentrations of
325 quinolinic acid (QUIN) and 3-hydroxykynurenine (3-HK),
326 neurotoxic components of the kynureninepathway.
327 Several studies observed that increased brain KYNA
328 levels follow systemic administration of L-KYNs. Swartz
329 et al. found that striatal KYNA level increased gradually
330 as a result of L-KYNs administered systemically in
331 gradually increased doses. The main conclusion of this
332 study was that extracellular levels of KYNA can be
333 dramatically increased by pharmacologic manipulation
334 of precursor levels (Swartz et al., 1990).
335 In another study concerning the effect of systemically
336 administered L-KYNs on cortical spreading depression,
337 intraperitoneal injections of L-KYNs were found to
increase cortical KYNA level about 40-fold in rats 338
(Chauvel et al., 2012). 339
Investigating the effect of systemically administered L- 340
KYNs on sensory gating, Shepard and associates found 341
that systemic administration of L-KYNs was not followed 342
by an increase of the harmful L-KYN metabolite, QUIN 343
(Shepard et al., 2003). 344
Astrocytes do not contain kynurenine 3-hydroxylase 345
and therefore cannot produce 3-HK, but are able to 346
produce large amounts of KYN and KYNA, whereas 347
microglial cells preferentially produce intermediates of 348
the quinolinic branch of the KYN pathway. It has also 349
been demonstrated that the other main source of QUIN 350
is the macrophage, infiltrated during inflammatory 351
processes (Guillemin et al., 2001; Wonodi and 352
Schwarcz, 2010). 353
In the main, the activation of the microglia increases 354
extracellular levels of QUIN or other kynurenines that 355
exacerbate neuronal damage (Schwarcz and Pellicciari, 356
2002). 357
In gerbils subjected to a period of cerebral ischemia, 358
50-fold QUIN level increases were observed 7days 359
after the onset of ischemia (Heyes and Nowak, 1990). 360
Finally, increased L-kynurenine influx from the blood 361
exceeds the catabolic capacity of kynurenine 3- 362
hydroxylase in microglia, promoting KYNA production in 363
the astrocytes (Wonodi and Schwarcz, 2010). 364
Microglia activation and the infiltration of the 365
macrophages follow the ischemic insult with a certain 366
delay. So we might reasonably conclude that the 367
extension of the damaged area in our experiments is 368
not the result of high 3-HK or QUIN levels originated 369
from L-KYNs administered promptly after reperfusion. 370
Extension of the neural damage is attributable to the 371
disturbedNMDAR-mediatedsurvival mechanisms. 372
These data indicate that kynurenergic manipulation 373
remains a potent strategy against excitotoxic cell death, 374
but the excitotoxic state and treatment pattern should be 375
well-tuned. 376
CONCLUSION 377
Suppression of excessive NMDA function has long been 378
the focus of research aimed at neuroprotection after 379
brain ischemia. However, robust NMDA antagonism is 380
not acceptable from the clinical point of view, since 381
normal synaptic NMDA function should not be inhibited, 382
even in the ischemic brain. The endogenous KYNA 383
acting at the glycine/D-serine co-agonist site of the 384
NMDA receptors is a pharmacon that might potentially 385
absolve this contradiction. Indeed, a huge body of 386
evidence indicates that manipulation of the brain KYNA 387
levels can effectively enhance the histopathological 388
outcome of the experimental I/H state. However, the 389
neuroprotective potential of L-KYNs administered after 390
brief focal ischemia has not yet been tested; 391
surprisingly, treatment with L-KYNs worsened the 392
histopathological outcome in our experiments. This 393
contradictory result indicates that post-ischemic 394
treatment with L-KYNs may be harmful. 395
396 Acknowledgement—This work was supported by OTKA grant 397 K105077.The publication is supported by the European Union 398 and co-funded by the European SocialFund.Project title:
399 ‘‘Broadening the knowledge base and supporting the long term 400 professional sustainability of the Research University Centre of 401 Excellence at the University of Szeged by ensuring the rising 402 generation of excellentscientists.’’
403 Project number: TA´MOP4.2.2-A-11/KONV-2012-0052.
404 Research project TA´MOP 4.2.4.A/2-11-1-2012-0001 is assisted 405 by the National Excellence Program–a convergence program 406 to support domestic students and researchers, funded by the 407 EU and co-financed by the European SocialFund.
408 T.F. was a Bolyai Fellow of the Hungarian Academy ofSciences.
409 This research was realized in the frames of TA´MOP 4.2.4. A/1- 410 11-1-2012-0001 ‘‘National Excellence Program – Elaborating 411 and operating an inland student and researcher personal support 412 system’’.The project was subsidized by the European Union and 413 co-financed by the European Social Fund.’’
414 REFERENCES
415 Aspey BS, Taylor FL, Terruli M, Harrison MJ (2000) Temporary 416 middle cerebral artery occlusion in the rat: consistent protocol for 417 a model of stroke and reperfusion. Neuropathol Appl Neurobiol 418 26:232–242.
419 Banerjee J, Alkondon M, Pereira EF, Albuquerque EX Regulation of 420 GABAergic inputs to CA1 pyramidal neurons by nicotinic 421 receptors and kynurenic acid. J Pharmacol Exp Ther 341:500–
422 509.
423 Benveniste H, Drejer J, Schousboe A, Diemer NH (1984) Elevation of 424 the extracellular concentrations of glutamate and aspartate in rat 425 hippocampus during transient cerebral ischemia monitored by 426 intracerebral microdialysis. J Neurochem 43:1369–1374.
427 Chauvel V, Vamos E, Pardutz A, Vecsei L, Schoenen J, Multon S 428 (2012) Effect of systemic kynurenine on cortical spreading 429 depression and its modulation by sex hormones in rat. Exp 430 Neurol 236:207–214.
431 Chen HS, Lipton SA (2006) The chemical biology of clinically 432 tolerated NMDA receptor antagonists. J Neurochem 433 97:1611–1626.
434 Delorme A, Makeig S (2004) EEGLAB: an open source toolbox for 435 analysis of single-trial EEG dynamics including independent 436 component analysis. J Neurosci Methods 134:9–21.
437 Endres M, Dirnagl U (2002) Ischemia and stroke. Adv Exp Med Biol 438 513:455–473.
439 Fuchs SA, Peeters-Scholte CM, de Barse MM, Roeleveld MW, Klomp 440 LW, Berger R, de Koning TJ (2012) Increased concentrations of 441 both NMDA receptor co-agonists D-serine and glycine in global 442 ischemia: a potential novel treatment target for perinatal asphyxia.
443 Amino Acids 43:355–363.
444 Gellert L, Fuzik J, Goblos A, Sarkozi K, Marosi M, Kis Z, Farkas T, 445 Szatmari I, Fulop F, Vecsei L, Toldi J (2011) Neuroprotection with 446 a new kynurenic acid analog in the four-vessel occlusion model of 447 ischemia. Eur J Pharmacol 667:182–187.
448 Gigler G, Szenasi G, Simo A, Levay G, Harsing Jr LG, Sas K, Vecsei 449 L, Toldi J (2007) Neuroprotective effect of L-kynurenine sulfate 450 administered before focal cerebral ischemia in mice and global 451 cerebral ischemia in gerbils. Eur J Pharmacol 564:116–122.
452 Guillemin GJ, Kerr SJ, Smythe GA, Smith DG, Kapoor V, Armati PJ, 453 Croitoru J, Brew BJ (2001) Kynurenine pathway metabolism in 454 human astrocytes: a paradox for neuronal protection. J 455 Neurochem 78:842–853.
456 Gunduz-Bruce H (2009) The acute effects of NMDA antagonism:
457 from the rodent to the human brain. Brain Res Rev 60:279–286.
458 Hardingham GE (2009) Coupling of the NMDA receptor to 459 neuroprotective and neurodestructive events. Biochem Soc 460 Trans 37:1147–1160.
Hardingham GE, Bading H (2010) Synaptic versus extrasynaptic 461 NMDA receptor signalling: implications for neurodegenerative 462 disorders. Nat Rev Neurosci 11:682–696. 463
Hetman M, Kharebava G (2006) Survival signaling pathways 464 activated by NMDA receptors. Curr Top Med Chem 6:787–799. 465 Heyes MP, Nowak Jr TS (1990) Delayed increases in regional brain 466
quinolinic acid follow transient ischemia in the gerbil. J. Cerebral 467 Blood Flow Metab. 10:660–667. 468
Ikonomidou C, Stefovska V, Turski L (2000) Neuronal death 469 enhanced by N-methyl-D-aspartate antagonists. Proc Natl Acad 470 Sci U S A 97:12885–12890. 471
Ikonomidou C, Turski L (2002) Why did NMDA receptor antagonists 472 fail clinical trials for stroke and traumatic brain injury? Lancet 473 Neurol 1:383–386. 474
Kussius CL, Popescu GK (2009) Kinetic basis of partial agonism at 475 NMDA receptors. Nat Neurosci 12:1114–1120. 476
Li ST, Ju JG (2012) Functional roles of synaptic and extrasynaptic 477 NMDA receptors in physiological and pathological neuronal 478 activities. Curr Drug Targets 13:207–221. 479
Li Y, Krupa B, Kang JS, Bolshakov VY, Liu G (2009) Glycine site of 480 NMDA receptor serves as a spatiotemporal detector of synaptic 481 activity patterns. J Neurophysiol 102:578–589. 482
Lugo-Huitron R, Blanco-Ayala T, Ugalde-Muniz P, Carrillo-Mora P, 483 Pedraza-Chaverri J, Silva-Adaya D, Maldonado PD, Torres I, 484 Pinzon E, Ortiz-Islas E, Lopez T, Garcia E, Pineda B, Torres- 485 Ramos M, Santamaria A, La Cruz VP On the antioxidant 486 properties of kynurenic acid: free radical scavenging activity and 487 inhibition of oxidative stress. Neurotoxicol Teratol 33:538–547. 488 Memezawa H, Smith ML, Siesjo BK (1992) Penumbral tissues 489
salvaged by reperfusion following middle cerebral artery occlusion 490 in rats. Stroke 23:552–559. 491
Moroni F, Cozzi A, Sili M, Mannaioni G (2012) Kynurenic acid: a 492 metabolite with multiple actions and multiple targets in brain and 493 periphery. J Neural Transm 119:133–139. 494
Moskowitz MA, Lo EH, Iadecola C (2010) The science of stroke: 495 mechanisms in search of treatments. Neuron 67:181–198. 496 Muir KW (2006) Glutamate-based therapeutic approaches: clinical 497
trials with NMDA antagonists. Curr Opin Pharmacol 6:53–60. 498 Nakazawa K, McHugh TJ, Wilson MA, Tonegawa S (2004) NMDA 499
receptors, place cells and hippocampal spatial memory. Nat Rev 500 Neurosci 5:361–372. 501
Papadia S, Soriano FX, Leveille F, Martel MA, Dakin KA, Hansen HH, 502 Kaindl A, Sifringer M, Fowler J, Stefovska V, McKenzie G, 503 Craigon M, Corriveau R, Ghazal P, Horsburgh K, Yankner BA, 504 Wyllie DJ, Ikonomidou C, Hardingham GE (2008) Synaptic NMDA 505 receptor activity boosts intrinsic antioxidant defenses. Nat 506 Neurosci 11:476–487. 507
Papouin T, Ladepeche L, Ruel J, Sacchi S, Labasque M, Hanini M, 508 Groc L, Pollegioni L, Mothet JP, Oliet SH (2012) Synaptic and 509 extrasynaptic NMDA receptors are gated by different endogenous 510 coagonists. Cell 150:633–646. 511
Paxinos G, Watson CR, Emson PC (1980) AChE-stained horizontal 512 sections of the rat brain in stereotaxic coordinates. J Neurosci 513 Methods 3:129–149. 514
Petralia RS, Wang YX, Hua F, Yi Z, Zhou A, Ge L, Stephenson FA, 515 Wenthold RJ (2010) Organization of NMDA receptors at 516 extrasynaptic locations. Neuroscience 167:68–87. 517
Pohl D, Bittigau P, Ishimaru MJ, Stadthaus D, Hubner C, Olney JW, 518 Turski L, Ikonomidou C (1999) N-Methyl-D-aspartate antagonists 519 and apoptotic cell death triggered by head trauma in developing 520 rat brain. Proc Natl Acad Sci U S A 96:2508–2513. 521
Popp A, Jaenisch N, Witte OW, Frahm C (2009) Identification of 522 ischemic regions in a rat model of stroke. PLoS ONE 4:e4764. 523 Sas K, Csete K, Vecsei L, Papp JG (2003) Effect of systemic 524
administration of L-kynurenine on corticocerebral blood flow 525 under normal and ischemic conditions of the brain in conscious 526 rabbits. J Cardiovasc Pharmacol 42:403–409. 527
Sas K, Robotka H, Rozsa E, Agoston M, Szenasi G, Gigler G, Marosi 528 M, Kis Z, Farkas T, Vecsei L, Toldi J (2008) Kynurenine diminishes 529 the ischemia-induced histological and electrophysiological deficits 530 in the rat hippocampus. Neurobiol Dis 32:302–308. 531
Q3
532 Schmued LC, Stowers CC, Scallet AC, Xu L (2005) Fluoro-Jade C 533 results in ultra high resolution and contrast labeling of 534 degenerating neurons. Brain Res 1035:24–31.
535 Schwarcz R, Pellicciari R (2002) Manipulation of brain kynurenines:
536 glial targets, neuronal effects, and clinical opportunities. J 537 Pharmacol Exp Ther 303:1–10.
538 Shepard PD, Joy B, Clerkin L, Schwarcz R (2003) Micromolar brain 539 levels of kynurenic acid are associated with a disruption of 540 auditory sensory gating in the rat. Neuropsychopharmacology 541 28:1454–1462.
542 Stanika RI, Villanueva I, Kazanina G, Andrews SB, Pivovarova NB 543 (2012) Comparative impact of voltage-gated calcium channels 544 and NMDA receptors on mitochondria-mediated neuronal injury. J 545 Neurosci 32:6642–6650.
546 Stone TW (2000) Development and therapeutic potential of kynurenic 547 acid and kynurenine derivatives for neuroprotection. Trends 548 Pharmacol Sci 21:149–154.
549 Stone TW, Addae JI (2002) The pharmacological manipulation of 550 glutamate receptors and neuroprotection. Eur J Pharmacol 551 447:285–296.
552 Swartz KJ, During MJ, Freese A, Beal MF (1990) Cerebral synthesis 553 and release of kynurenic acid: an endogenous antagonist of 554 excitatory amino acid receptors. J Neurosci 10:2965–2973.
Vamos E, Pardutz A, Klivenyi P, Toldi J, Vecsei L (2009) The role of 555 kynurenines in disorders of the central nervous system: 556 possibilities for neuroprotection. J Neurol Sci 283:21–27. 557 Vecsei L, Szalardy L, Fulop F, Toldi J (2012) Kynurenines in the CNS: 558
recent advances and new questions. Nat Rev Drug Discov 559 12:64–82. 560
Wonodi I, Schwarcz R (2010) Cortical kynurenine pathway 561 metabolism: a novel target for cognitive enhancement in 562 Schizophrenia. Schizophr Bull 36:211–218. 563
Wu HQ, Guidetti P, Goodman JH, Varasi M, Ceresoli-Borroni G, 564 Speciale C, Scharfman HE, Schwarcz R (2000) Kynurenergic 565 manipulations influence excitatory synaptic function and 566 excitotoxic vulnerability in the rat hippocampus in vivo. 567 Neuroscience 97:243–251. 568
Zadori D, Klivenyi P, Vamos E, Fulop F, Toldi J, Vecsei L (2009) 569 Kynurenines in chronic neurodegenerative disorders: future 570 therapeutic strategies. J Neural Transm 116:1403–1409. 571 Zhang SJ, Steijaert MN, Lau D, Schutz G, Delucinge-Vivier C, 572
Descombes P, Bading H (2007) Decoding NMDA receptor 573 signaling: identification of genomic programs specifying 574 neuronal survival and death. Neuron 53:549–562. 575
576
577 (Accepted 30 April 2013)
578 (Available online xxxx)
