Sensitivities of Benzodiazepine ReceptorBinding and Muscarinic AcetylcholineReceptor Binding for the Detection ofNeural Cell Death Caused by Sodium

Nitroprusside Microinjection in Rat BrainOSAMU INOUE,1* KAZUHIKO YANAMOTO,1 YOSHIKO FUJIWARA,1 RIE HOSOI,1

KAORU KOBAYASHI,1 AND HIDEO TSUKADA2

1Department of Medical Physics, School of Allied Health Sciences, Faculty of Medicine, Osaka University,Osaka 565-0871, Japan

2Central Research Laboratory, Hamamatsu Photonics K.K., Shizuoka 434-8601, Japan

KEY WORDS sodium nitroprusside (SNP); brain; neural cell death; benzodiazepinereceptor; muscarinic acetylcholine receptor

ABSTRACT Sodium nitroprusside (SNP) was microinjected into rat cerebral cortexand changes in muscarinic acetylcholine receptor (mAChR) binding and benzodiazepinereceptor (BZR) binding were followed for 24 h after the infusion using [3H]-N-methyl-4-piperidyl benzilate ([3H]-NMPB) and [3H]-flumazenil, respectively, as a radioligand.The microinjection of SNP dose-dependently caused significant neural cell death 3 hafter infusion, with the area of cell death becoming extensive 24 h after infusion. NeitherSIN-1 nor NOC-18, other types of NO donors, caused neural cell death. Together withthe result that deferoxamine, an iron-chelating agent, protected SNP-induced braininjury indicated important roles of iron-related radicals in SNP cytotoxicity in rat brain.In vitro [3H]-NMPB binding was significantly reduced in parallel with the time course ofneural cell death detected by TTC staining and Nissl staining. In contrast, [3H]-fluma-zenil binding was essentially unaltered during the 24-h period after the SNP infusion.Similar results were observed in in vivo binding experiments. In vivo [3H]-NMPBbinding was found to be much more sensitive at detecting cell death caused by SNP. Onthe other hand, [3H]-flumazenil binding in vivo was relatively insensitive to SNP-induced cell death. These results indicate that mAChR binding may be superior to BZRbinding for detecting cell death in brain tissue, in contrast to what was previouslythought. Synapse 49:134–141, 2003. © 2003 Wiley-Liss, Inc.

INTRODUCTION

Neuroreceptor imaging, using either positron emis-sion tomography (PET) or single photon emission com-puted tomography (SPECT), has been utilized to assessthe viability of neural cells in ischemic brain tissue(Baron, 2001; Heiss, 2001; Moriwaki et al., 1998; Na-kagawara et al., 1997). Measurements of [11C]-fluma-zenil binding or [123I]-iomazenil binding to benzodiaz-epine receptors (BZR) are known to reflect neural cellviability in ischemic regions of the brain. Other types ofreceptor imaging methods have also been reported,such as the detection of muscarinic acetylcholine recep-tors (mAChR) in neurodegenerative diseases (Araki etal., 1992; O’Neill et al., 1996; Ouchi et al., 1998). Themain purpose of this study was to compare the sensi-

tivity of BZR binding and mAChR binding for the de-tection of neural cell death in brain tissues.

Accumulating evidence indicates that radicals playimportant roles in the pathophysiology of neurologicaldiseases, such as Parkinson’s disease (Albers and Beal,2000), Alzheimer’s disease (Markesbery and Carney,1999), and brain trauma (Lewen et al., 2000). In brainischemia, neural cell death might also be related to

*Correspondence to: Osamu Inoue, Department of Medical Physics, School ofAllied Health Sciences, Faculty of Medicine, Osaka University, 1-7 Yamadaoka,Suita, Osaka 565-0871, Japan. E-mail: inoue@sahs.med.osaka-u.ac.jp

Received 14 January 2003; Accepted 16 March 2003

DOI 10.1002/syn.10217

SYNAPSE 49:134–141 (2003)

© 2003 WILEY-LISS, INC.

radical formation (Love, 1999; Yamamoto et al., 1997),and several types of radical scavengers have been re-ported to protect the brain from damage in rodentmodels of brain stroke (Mizuno et al., 1998; Nakashimaet al., 1999). Other than stroke models, only a fewanimal models for investigating radical-induced celldeath in the central nervous system (CNS) have beenreported. Hironishi et al. (1999) showed that the mi-croinjection of FeCl2 into the right nigra of the ratbrain caused significant destruction of dopaminergicneurons in the right striatum, as demonstrated by bothbehavioral and pathological studies. However, no otherreports have been made on radical-induced neural celldeath in the cerebral cortex of intact animals. If suchan animal model could be successfully developed, itwould be of value for ascertaining the detailed rolesof radical reactions in the process of neural celldeath.

We chose sodium nitroprusside (SNP) as a cell death-inducing reagent to be used in the cerebral cortex ofawake rats. The cytotoxic effects of SNP on variouscells have been well documented (Boullerne et al.,1999; Chen et al., 1991; Yamada et al., 1996). However,most of these experiments were performed using tissueslices or cell cultures, and SNP cytotoxicity in intactbrain has not been well described. Therefore, we exam-ined whether the microinjection of SNP into the cere-bral cortex of awake rats caused cell death by using2,3,5-triphenyltetrazolium chloride (TTC) and Nisslstaining.

In these experiments, the in vitro binding of [3H]-flumazenil in brain slices was measured at variousintervals during the 24 h after SNP microinjection.Changes in mAChR binding in vitro, as demonstratedwith [3H]-N-methyl-4-piperidyl benzilate ([3H]-NMPB),were also followed, since the cerebral cortex is anmAChR-rich region (Hosoi et al., 1999). The in vivobinding of [3H]-flumazenil and [3H]-NMPB in rat brainwas also measured 3 h after the SNP infusion.

MATERIALS AND METHODSAnimals and chemicals

Male Sprague-Dawley rats (7–8 weeks old) were ob-tained from Clea Japan (Tokyo, Japan). The rats werehoused under a 12-h, light-dark cycle and allowed freeaccess to food and water. All experiments were per-formed with the permission of the Institutional AnimalCare and Use Committee, School of Allied Health Sci-ences, Osaka University.

Sodium nitroprusside (SNP) and deferoxamine me-sylate were purchased from Sigma Chemical Co. (St.Louis, MO, USA). 2,2-(hydroxynitrosohydrazino)bis-ethanamine (NOC-18), 3-morpholinosydnonimine chlo-ride (SIN-1) and 2,3,5-triphenyltetrazolium chloride

(TTC) were purchased from Dojin Laboratories (Kum-amoto, Japan), Tocris Cookson (Bristol, UK), and WakoPure Chemical Industries (Osaka, Japan), respectively.

[3H]-Flumazenil (specific activity, 2.9 TBq/mmol)and [3H]-N-methyl-4-piperidyl benzilate (NMPB, spe-cific activity, 3.1 TBq/mmol) were obtained from NENLife Sciences Products (Boston, MA). [3H]-Micro-scale(RPA 507) was obtained from Amersham Biotech(Buckinghamshire, UK).

Surgery on rat brains and infusionof SNP, SIN-1, or NOC-18

The rats were anesthetized with sodium pentobarbi-tal (50 mg/kg, i.p.) and placed in a stereotactic frame.Bilateral guide cannulas (26 gauge stainless steel)were implanted into the cerebral cortex (anterior, �0.2mm; lateral, �3.2 mm; ventral, –1.0 mm from thebregma) according to the atlas of Paxinos and Watson

Fig. 1. Time course of cell death in the rat brain, as detected byTTC staining following the microinjection of SNP (50 nmol/�l) into thecerebral cortex. No significant cell death was detected 1 h after theSNP infusion (A); however, a significant and time-dependent exten-sion of the area of cell death was apparent at 3 h (B) and 24 hours (C)after infusion. Three rats were used in each group.

BENZODIAZEPINE AND MUSCARINIC BINDING ACETYLCHOLINE 135

(1998) and affixed to the skull with acrylic cement. Theguide cannulas were occluded with a dummy cannulaof the same length. The rats were then allowed torecover for several days.

SNP (10, 25 and 50 nmol/�l; dissolved in saline),SIN-1 (50 nmol/�l), or NOC-18 (50 nmol/�l) was in-fused through the infusion cannulas (33 gauge, 1.5 mmlonger than the guide cannulas) and into the rightcerebral cortex of each rat while the rat was awake.The infusion was performed for 4 min at a flow rate of0.25 �l/min and the infusion cannulas were left in placefor an additional 3 min to allow the chemicals to com-pletely diffuse. Saline solution (1 �l) was infused intothe left cerebral cortex using the same protocol as de-scribed above. In the experiment on the effect of aniron-chelating agent to SNP cytotoxicity, deferoxamine(10 nmol/�l) was preinfused into the right cerebralcortex 10 min before the SNP (50 nmol/�l) infusion,and SNP (50 nmol/�l) alone was infused into the leftcerebral cortex as a control side.

TTC and Nissl staining of rat brain

The rats were killed by decapitation at 1, 3, and 24 hafter the infusion of SNP. The brains were quicklyremoved and coronal slices (approximately 3 mm inthickness) were prepared. The slices were subse-quently stained with TTC (2% solution in PBS) at 37°Cfor 30 min.

Slices (15 �m in thickness) of rat brain infused withSNP were also stained with cresyl violet and the num-

ber of living neural cells or glia cells in the right cere-bral cortex were counted and compared with the num-ber in the left cerebral cortex. In experiments on theeffects of the SIN-1, NOC-18, or deferoxamine infusion,and on the dose-dependency of SNP, rats were killed bydecapitation at 3 h after the infusion.

In vitro autoradiography

At 1, 3, and 24 h after the infusion of SNP, the ratswere killed by decapitation under a light anesthesiawith diethyl ether. The brains were quickly removed,frozen in chilled hexane, and stored at –80°C until use.Coronal sections (30 �m in thickness) were preparedusing a cryostat at –20°C and mounted on poly-L-ly-sine-coated slides. The sections were preincubated at25°C for 10 min in 50 mM Na-K phosphate buffer (pH7.4), and incubated at 25°C for 60 min in the same

Fig. 2. Number of living neural cells (filled circles) or glia cells(crosses) in the right cerebral cortex of the rat brain, as measuredusing Nissl staining. The values are expressed as percent (number ofliving cells/mm2, average � 1 SD of three rats) on the SNP-infusedside relative to that on the control side of the cerebral cortex. *P �0.05, ***P � 0.001 vs. control side of the cerebral cortex (Student’spaired t-test).

Fig. 3. Dose-dependency of SNP-induced cell death detected byTTC staining. Rats were infused with different amounts of SNP (10,25, and 50 nmol/�l) and killed by decapitation at 3 h after the SNPinfusion. The area of cell death was increased with increasing SNPdose. Three to five rats were used in each group.

136 O. INOUE ET AL.

buffer with 1 nM of [3H]-flumazenil or [3H]-NMPB.After incubation, the sections were washed three timesfor 5 min with a 50 mM Na-K phosphate buffer at 4°Cand rinsed for a few seconds in cold water. After drying,the sections were exposed to an imaging plate (BAS-TR, Fuji Photo Film) together with a [3H]-Micro-scale.The autoradiograms were quantified using the photo-stimulated luminescence (PSL) values with a Bio-Im-aging Analyzer System (BAS1500, Fuji Photo Film).The radioactivity concentrations in the regions of in-terest (ROIs) were determined as PSL/area (mm2), andall values were expressed as the relative ratio of theradioactivity concentration in the right cerebral cortex(SNP-infused side) to that in the left cerebral cortex(saline-infused side).

In vivo autoradiography

The rats were injected i.v. with either 1.85 MBq of[3H]-flumazenil or [3H]-NMPB 3 h after the infusion ofSNP. Rats were killed by decapitation 10 min aftertracer injection for [3H]-flumazenil binding or 60 minafter tracer injection for [3H]-NMPB binding. Brainslices (30 �m in thickness) were prepared and con-tacted with imaging plates. The radioactivity concen-trations in the ROIs were determined using the samemethod described above.

RESULTS

Figure 1 shows the results of TTC staining in ratbrain slices prepared 1, 3, and 24 h after SNP micro-injection into the cerebral cortex. No significant celldeath was seen 1 h after the SNP infusion, whereas celldeath was significant 3 h after the infusion. The areasof cell death showed a time-dependent distribution,and part of the striatum also revealed degeneration24 h after the infusion. The number of living neuralcells in the right cerebral cortex decreased with timeafter the SNP infusion, as shown in Figure 2. SNPcaused cell death in a dose-dependent manner asshown in Figure 3.

Figure 4 shows the effects of infusion of other types ofNO donors and the effect of an iron-chelating agent onSNP-induced brain injury. Neither SIN-1 nor NOC-18caused cell death. A pronounced protective effect ofdeferoxamine to SNP cytotoxicity was observed. Theautoradiographic results for [3H]-NMPB binding and[3H]-flumazenil binding in vitro are summarized inFigure 5. A significant reduction in [3H]-NMPB bindingwas essentially consistent with the area of cell deathdetected by TTC staining and the degree of changes inbinding was time-dependent. In contrast, [3H]-fluma-zenil binding was found to be resistant to cell deathcaused by the SNP infusion, and more than 80% of BZRbinding remained even 24 h after the SNP infusion.

Fig. 4. Photographs of coronal sectionsat the level of the striatum in rats infusedwith SNP (A), SIN-1 (B), NOC-18 (C), orSNP plus deferoxamine (D). SNP, SIN-1,or NOC-18 (50 nmol/�l) was microinjectedinto the right cerebral cortex. In the exper-iment on the effect of an iron-chelatingagent to SNP cytotoxicity, deferoxamine(10 nmol/�l) was preinfused into the rightcerebral cortex 10 min before the SNP (50nmol/�l) infusion and SNP (50 nmol/�l)alone was infused into the left cerebral cor-tex as a control side. Rat brains were dis-sected at 3 h after the infusion, then eachsection was stained with TTC. Three ratswere used in each group.

BENZODIAZEPINE AND MUSCARINIC BINDING ACETYLCHOLINE 137

The time courses for the changes in [3H]-NMPB and[3H]-flumazenil binding in rat cerebral cortex infusedwith SNP are shown in Figure 6.

[3H]-NMPB binding decreased significantly withtime after the infusion of SNP, and less than 20% of[3H]-NMPB binding remained 24 h after the infusion.In contrast, almost no change in [3H]-flumazenil bind-ing was observed in the rat cerebral cortex during the24 h after the SNP infusion, despite the significant andwidespread cell death that was apparent at this time.

Figure 7 shows the in vivo binding of [3H]-flumazeniland [3H]-NMPB in rat brain tissue obtained 3 h afterthe SNP infusion. A significant decrease in [3H]-NMPBbinding in the right cerebral cortex was detected; how-ever, a significant increase in binding was observed inthe rim of the cell death area. In contrast, the in vivobinding of [3H]-flumazenil in the area of cell death wasslightly increased 3 h after the SNP infusion.

DISCUSSION

The microinjection of SNP into the cerebral cortex ofan awake rat can cause cell death. This process israpid and can be detected as early as 3 h after theinfusion. In this study, the cell death process was con-tinuous, progressive, and widespread, with part of thestriatum being altered 24 h after the infusion. In ad-dition, dose-dependency of SNP cytotoxicity in rat ce-rebral cortex was observed. SNP is a traditional NOdonor compound and has been reported to decomposeinto various compounds, including cyanide anion and apotentially cytotoxic pentacyanoferrate complex (Bateset al., 1991). Cyanide anion does not appear to be a majorcause of cell death, because cyanide anions do not exhibitcytotoxicity in cell cultures (Boullerne et al., 1999). SNPmetabolites, [Fe(CN)5NO]3- and [Fe(CN)4NO]2-, undergoredox cycling with oxygen (Bates et al., 1991). During this

Fig. 5. In vitro autoradiograms of [3H]-NMPB binding (left side) and [3H]-flumazenilbinding (right side) obtained using rat brainslices prepared at 1, 3, and 24 h after the SNPinfusion. SNP (50 nmol/�l) was microinjectedinto the rat cerebral cortex. The rats were thenkilled by decapitation 1, 3, or 24 h after theinfusion. The brains were quickly removedand slices were prepared and incubated with[3H]-NMPB or [3H]-flumazenil, as described inthe text. Four or six rats were used in eachgroup.

138 O. INOUE ET AL.

cycle, DNA damage in cells could be triggered by theformation of reactive oxygen species, such as superoxide,hydrogen peroxide, and hydroxyl radical (Yamada et al.,1996). Iron and related radicals might have importantroles in the cell death that is induced by SNP microinjec-tion into the rat cerebral cortex. In fact, a pronouncedprotective effect of deferoxamine to SNP-induced celldeath was observed, which supported the above hypoth-esis that iron-related radicals had important roles in SNPcytotoxicity. Several lines of evidence indicate that syn-ergistic cytotoxicity between NO and iron or iron-relatedradicals in SNP induces apoptotic cell death (Boullerne etal., 1999; Yamada et al., 1996). Such synergism appearsto be an important factor in the current observation of cell

death in intact rat brain, since the microinjection of eitherSIN-1 or NOC-18, other types of NO donors, did not causecell death.

The most important and unexpected finding in thecurrent study was that [3H]-flumazenil binding was notaltered in the area of cell death in the cerebral cortex.Central-type BZR is part of the postsynaptic GABAA

receptor complex (Schwartz, 1988), and its binding isapparently a marker of the number of neural cells. InPET or SPECT imaging, BZR binding has been used asa marker of cell viability in patients with brain isch-emia (Dong et al., 1997; Heiss et al., 1998, 2000). Heisset al. (2001b) reported the penumbral probabilitythresholds for cortical [11C]-flumazenil binding in pa-tients with cerebral ischemia. The quantification ofincomplete brain infarction of reperfused cortex using[123I]-iomazenil binding has also been proposed (Naka-gawara et al., 1997). In animal experiments on brainischemia, significant decreases in BZR binding havebeen reported (Heiss et al., 1997, 2001a; Matsuda et al.,1995; Onodera et al., 1987). Sette et al. (1993) showedthat in vivo BZR binding decreased in infarct areas inthe baboon brain. The loss of central-type BZR bindingwas also observed in a rat model for lateral fluid-percussion brain injury (Toulmond et al., 1993). In thecurrent study, however, [3H]-flumazenil binding wasnot sensitive to cell death induced by SNP microinjec-tion into the rat cerebral cortex. Almost no reduction inin vitro [3H]-flumazenil binding was seen at 3 h afterthe SNP infusion, and only about a 20% decrease inbinding was detected 24 h after infusion. In addition, aslight increase in in vivo [3H]-flumazenil binding wasnoticed. The current observations indicate that BZRbinding sites seem to be considerably stable to radical-induced neural cell death.

On the other hand, changes in [3H]-NMPB bindingwere much more sensitive to cell death induced bySNP, as shown in Figures 5 and 6. The time course forthe degree of [3H]-NMPB reduction as well as the sizeof the affected area in the rat brain were proportionalto the cell death findings detected by TTC staining.These results strongly suggest that mAChR binding is

Fig. 7. In vivo binding of [3H]-NMPB (A) and[3H]-flumazenil (B) binding in rat brain infusedwith 50 nmol/�l of SNP. Rats were injected i.v.with [3H]-NMPB or [3H]-flumazenil 3 h after theSNP infusion and decapitated 10 min aftertracer injection for [3H]-flumazenil binding or 60min after tracer injection for [3H]-NMPB bind-ing. A significant reduction in [3H]-NMPB bind-ing in the core region of cell death and an in-crease in binding in the rim of the cell deathregion were observed. In contrast, a slight in-crease in in vivo [3H]-flumazenil binding wasseen. Four or five rats were used in each group.

Fig. 6. Results of the quantitative analysis of in vitro autoradio-grams. The regions of interest (ROIs) were identified on autoradio-grams and the radioactivity concentrations were determined by esti-mating the photostimulated luminescence (PSL) values per squaremm (PSL/mm2). The values are expressed as percent (PSL/mm2,average � 1 SD of four or six rats) on the SNP-infused side relative tothat on the control side of the cerebral cortex. ***P � 0.001 vs. controlside of the cerebral cortex (Student’s paired t-test).

BENZODIAZEPINE AND MUSCARINIC BINDING ACETYLCHOLINE 139

much more sensitive at detecting cell death in thebrain than BZR binding, although the reason for thisdifference is unclear. Several reports have described asignificant reduction in mAChR binding in ischemicbrain (O’Neill et al., 1996; Ouchi et al., 1998). Araki etal. (1992) reported that [3H]-QNB binding in the stri-atum and hippocampus decreased significantly 7 daysafter ischemia in the gerbil brain. To our knowledge,however, this report describes the first observation of adifference in sensitivity to neural loss between mAChRbinding and BZR binding.

The increase in in vivo [3H]-NMPB binding that wasobserved in the rim of the region of cell death is anotherinteresting finding (Fig. 7). The regional cerebral bloodflow, as measured with [14C]-iodoantipyrine, was notsignificantly altered by the SNP infusion (data notshown). Therefore, other factors, such as an alterationin the permeability of [3H]-NMPB from the plasma tothe brain or nonspecific binding, should be consideredas subjects of future study.

[11C]-NMPB is a nonsubtype-selective ligand (Mul-holland et al., 1995) and has been used for the mea-surement of mAChR binding in living monkey (Buck etal., 1996; Nishiyama et al., 2000) and human brains(Asahina et al., 1998; Zubieta et al., 2001). PET studieswith [11C]-NMPB would be of value for demonstratingcell death induced by radical-related reactions in thebrain. The efficacy of neuroprotective drugs could befollowed using this SNP-induced cell death model inboth rodent and primate brain.

REFERENCES

Albers DS, Beal MF. 2000. Mitochondrial dysfunction and oxidativestress in aging and neurodegenerative disease. J Neural TransmSuppl 59:133–154.

Araki T, Kato H, Kogure K. 1992. Postischemic alteration of musca-rinic acetylcholine and adenosine A1 binding sites in gerbil brain.Protective effects of a novel vinca alkaloid derivative, vinconate,and pentobarbital using an autoradiographic study. Res Exp Med(Berl) 192:79–88.

Asahina M, Suhara T, Shinotoh H, Inoue O, Suzuki K, Hattori T.1998. Brain muscarinic receptors in progressive supranuclear palsyand Parkinson’s disease: a positron emission tomographic study.J Neurol Neurosurg Psychiatry 65:155–163.

Baron JC. 2001. Mapping the ischemic penumbra with PET: a newapproach. Brain 124:2–4.

Bates JN, Baker MT, Guerra Jr R, Harrison DG. 1991. Nitric oxidegeneration from nitroprusside by vascular tissue. Evidence thatreduction of the nitroprusside anion and cyanide loss are required.Biochem Pharmacol 42(Suppl):S157–S165.

Boullerne AI, Nedelkoska L, Benjamins JA. 1999. Synergism of nitricoxide and iron in killing the transformed murine oligodendrocytecell line N20.1. J Neurochem 72:1050–1060.

Buck A, Mulholland GK, Papadopoulos SM, Koeppe RA, Frey KA.1996. Kinetic evaluation of positron-emitting muscarinic receptorligands employing direct intracarotid injection. J Cereb Blood FlowMetab 16:1280–1287.

Chen J, Chang B, Williams M, Murad F. 1991. Sodium nitroprus-side degenerates cultured rat striatal neurons. Neuroreport2:121–123.

Dong Y, Fukuyama H, Nabatame H, Yamauchi H, Shibasaki H,Yonekura Y. 1997. Assessment of benzodiazepine receptors usingiodine-123-labeled iomazenil single-photon emission computed to-mography in patients with ischemic cerebrovascular disease. Acomparison with PET study. Stroke 28:1776–1782.

Heiss WD. 2001. Imaging the ischemic penumbra and treatmenteffects by PET. Keio J Med 50:249–256.

Heiss WD, Graf R, Fujita T, Ohta K, Bauer B, Lottgen J, Wienhard K.1997. Early detection of irreversibly damaged ischemic tissue byflumazenil positron emission tomography in cats. Stroke 28:2045–2052.

Heiss WD, Grond M, Thiel A, Ghaemi M, Sobesky J, Rudolf J, BauerB, Wienfard K. 1998. Permanent cortical damage detected byflumazenil positron emission tomography in acute stroke. Stroke29:454–461.

Heiss WD, Kracht LW, Grond M, Rudolf J, Bauer B, Wienhard K,Pawlik G. 2000. Early [11C]Flumazenil/H2O positron emissiontomography predicts irreversible ischemic cortical damage instroke patients receiving acute thrombolytic therapy. Stroke 31:366 –369.

Heiss WD, Graf R, Wienhard K. 2001a. Relevance of experimentalischemia in cats for stroke management: a comparative reevalua-tion. Cerebrovasc Dis 11:73–81.

Heiss WD, Kracht LW, Thiel A, Grond M, Pawlik G. 2001b. Penum-bral probability thresholds of cortical flumazenil binding and bloodflow predicting tissue outcome in patients with cerebral ischemia.Brain 124:20–29.

Hironishi M, Ueyama E, Senba E. 1999. Systemic expression of im-mediate early genes and intensive astrocyte activation induced byintrastriatal ferrous iron injection. Brain Res 828:145–153.

Hosoi R, Kobayashi K, Watanabe Y, Inoue O. 1999. Evaluation of invivo binding properties of 3H-NMPB and 3H-QNB in mouse brain.J Neural Transm 106:583–592.

Lewen A, Matz P, Chan PH. 2000. Free radical pathway in CNSinjury. J Neurotrauma 17:871–890.

Love S. 1999. Oxidative stress in brain ischemia. Brain Pathol 9:119–131.

Markesbery WR, Carney JM. 1999. Oxidative alterations in Alzhei-mer’s disease. Brain Pathol 9:133–146.

Matsuda H, Tsuji S, Kuji I, Shiba K, Hisada K, Mori H. 1995. Dual-tracer autoradiography using 125I-iomazenil and 99mTc-HMPAO inexperimental brain ischemia. Nucl Med Commun 16:581–590.

Mizuno A, Umemura K, Nakashima M. 1998. Inhibitory effect ofMCI-186, a free radical scavenger, on cerebral ischemia followingrat middle cerebral artery occlusion. Gen Pharmacol 30:575–578.

Moriwaki H, Matsumoto M, Hashikawa K, Oku N, Ishida M, Seike Y,Fukuchi K, Hori M, Nishimura T. 1998. Iodine-123-iomazenil andiodine-123-iodoamphetamine SPECT in major cerebral artery oc-clusive disease. J Nucl Med 39:1348–1353.

Mulholland GK, Kilbourn MR, Sherman P, Carey JE, Frey KA, Ko-eppe RA, Kuhl DE. 1995. Synthesis, in vivo biodistribution anddosimetry of [11C]N-methylpiperidyl benzilate ([11C]NMPB), a mus-carinic acetylcholine receptor antagonist. Nucl Med Biol 22:13–17.

Nakagawara J, Sperling B, Lassen NA. 1997. Incomplete brain in-farction of reperfused cortex may be quantitated with iomazenil.Stroke 28:124–132.

Nakashima M, Niwa M, Iwai T, Umematsu T. 1999. Involvement offree radicals in cerebral vascular reperfusion injury evaluated in atransient focal cerebral ischemia model of rat. Free Radic Biol Med26:722–729.

Nishiyama S, Sato K, Harada N, Kakiuchi T, Tsukada H. 2000.Development and evaluation of muscarinic cholinergic receptor li-gands N-[11C]ethyl-4-piperidyl benzilate and N-[11C]propyl-4-pip-eridyl benzilate: a PET study in comparison with N-[11C]methyl-4-piperidyl benzilate in the conscious monkey brain. Nucl Med Biol27:733–740.

O’Neill M, Canney M, Earley B, Junien JL, Leonard BE. 1996. Thenovel sigma ligand JO 1994 protects against ischemia-induced be-havioural changes, cell death and receptor dysfunction in the gerbil.Neurochem Int 28:193–207.

Onodera H, Sato G, Kogure K. 1987. GABA and benzodiazepine re-ceptors in the gerbil brain after transient ischemia: demonstrationby quantitative receptor autoradiography. J Cereb Blood FlowMetab 7:82–88.

Ouchi Y, Tsukada H, Kakiuchi T, Nishiyama S, Futatsubashi M.1998. Changes in cerebral blood flow and postsynaptic muscariniccholinergic activity in rats with bilateral carotid artery ligation.J Nucl Med 39:198–202.

Paxinos G, Watson C. 1998. The rat brain in stereotaxic coordinates,4th ed. San Diego: Academic Press.

Schwartz RD. 1988. The GABAA receptor-gated ion channel: biochem-ical and pharmacological studies of structure and function. BiochemPharmacol 37:3369–3375.

140 O. INOUE ET AL.

Sette G, Baron JC, Young AR, Miyazawa H, Tillet I, Barre L, TravereJM, Derlon JM, MacKenzie ET. 1993. In vivo mapping of brainbenzodiazepine receptor changes by positron emission tomographyafter focal ischemia in the anesthetized baboon. Stroke 24:2046–2058.

Toulmond S, Serrano A, Benavides J, Scatton B. 1993. Prevention byeliprodil (SL 82.0715) of traumatic brain damage in the rat. Exis-tence of a large (18 h) therapeutic window. Brain Res 620:32–41.

Yamada M, Momose K, Richelson E, Yamada M. 1996. Sodium nitro-prusside-induced apoptotic cellular death via production of hydro-

gen peroxide in murine neuroblastoma N1E-115 cells. J PharmacolToxicol Methods 35:11–17.

Yamamoto T, Yuki S, Watanabe T, Mitsuka M, Saito KI, Kogure K.1997. Delayed neuronal death prevented by inhibition of increasedhydroxyl radical formation in a transient cerebral ischemia. BrainRes 762:240–242.

Zubieta JK, Koeppe RA, Frey KA, Kilbourn MR, Mangner TJ, FosterNL, Kuhl DE. 2001. Assessment of muscarinic receptor concentra-tions in aging and Alzheimer disease with [11C]NMPB and PET.Synapse 39:275–287.

BENZODIAZEPINE AND MUSCARINIC BINDING ACETYLCHOLINE 141