Comparison of the patterns of altered cerebral glucose utilisation produced by competitive and non-competitive NMDA receptor antagonists

BRAIN RESEARCH Brain Research 735 (1996) 67-82

ELSEVIER

Research

report

Comparison of the patterns of altered cerebral glucose utilisation produced by competitive and non-competitive NMDA receptor antagonists J. S h a r k e y

a, *, I . M . R i t c h i e b, S . P . B u t c h e r

a, J . S . K e l l y c

a Fujisawa Institute of Neuroscience in Edinburgh, Department of Pharmacology, University of Edinburgh, 1, George Square, Edinburgh EH8 9JZ, UK b Department of Clinical Neurosciences, University of Edinburgh, Edinburgh, UK c Department of Pharmacology, University of Edinburgh, Edinburgh, UK

Accepted 7 May 1996

Abstract Recent studies indicate that competitive and non-competitive NMDA receptor antagonists can be readily distinguished by their effects on local cerebral glucose utilisation (1CGU). In the present study we compare the effects of the novel NMDA antagonist, (+)-l-methyllphenyl-l,2,3,4-tetrahydroisoquinoline (FR115427) on ICGU, comparing its metabolic profile with that of the non-competitive NMDA receptor antagonist, dizocilpine (MK801) and of the competitive NMDA receptor antagonist CGS19755, using the 2-deoxyglucose metabolic mapping approach. Local cerebral glucose utilisation was measured in 80 anatomically discrete regions of the conscious rat brain using [ 14C]2-deoxyglucose quantitative autoradiography. Studies were initiated 10 min after the administration of FR 115427 (0.1-3 m g / k g i.v.; n = 20), dizocilpine (0.03-0.3 mg/kg; n = 15), CGS19755 (1-30 mg/kg; n = 15) or saline (2 ml/kg; n = 5). Dizocilpine produced characteristic alterations in 1CGU with widespread increases in 1CGU in primary olfactory and limbic areas while reducing 1CGU in somatosensory and motor cortex. FR115427 produced a pattern of altered 1CGU which was broadly similar to that elicited by dizocilpine with increases in 1CGU in the pontine nuclei, presubiculum and hippocampus and reductions in somatosensory and motor cortex and within components of the auditory system. However, FR115427 was approximately 30-fold less potent than dizocilpine in this regard. In limbic structures, the effects of FRl15427 were less pronounced than those produced by dizocilpine. Increases in 1CGU of 62-98% were found in retrosplenial, piriform and entorhinal cortex of dizocilpine-treated rats whereas these areas appeared relatively unaffected following FR115427 administration. A comparison of the pattern of metabolic response produced by each of these agents was performed by constructing a hierarchy of regional responsiveness using the f statistic: while focal differences in the metabolic profiles of dizocilpine and FRl15427 were evident, a plot of the regional f values for dizocilpine and FRl15427 revealed a strong overall relationship between the metabolic responses with a Pearson's product moment correlation of 0.78. In contrast, the correlation between the patterns produced by CGS 19755 and that for dizocilpine or FR115427 was poor (r = 0.28 and 0.5 respectively). Keywords: Cerebral glucose metabolism; FR 115427; MK801; CGS 19755; NMDA receptor

1. I n t r o d u c t i o n Inhibition o f glutamatergic neurotransmission has been proposed for the treatment of a variety of clinical conditions including traumatic brain injury, subdural haemorrhage and stroke [1,19]. However, while numerous studies have shown the efficacy of agents which block N M D A receptor activation in animal models of brain injury, reports of neurotoxicity [26] and profound psychotomimetic effects with many of these agents make them unsuitable for human use [19]. Studies using electrophysiological, neurochemical and * Corresponding author. Fax: +44 (31) 667-9381.

more recently molecular biology techniques point to the existence o f a number o f modulatory sites within the N M D A receptor complex which are amenable to selective pharmacological manipulation [5]. N M D A receptor activation can be blocked in a competitive manner by ligands acting at the recognition site for the endogenous neurotransmitter. Agents acting at this site include AP-7, CPP and more recently, CGS19755 [12,16,27]. N M D A receptors can also be blocked in a non-competitive manner by agents such as ketamine, phencyclidine and ( + ) - 5 - m e t h y l 10,11-dihydro-5H-dibenzo[a,d]cyclo-hepten-5,10-imine maleate ( M K 8 0 1 / d i z o c i l p i n e ) which act at sites located within the N M D A receptor associated ligand channel [17,18,36].

0006-8993/96/$15.00 Copyright © 1996 Elsevier Science B.V. All rights reserved. PII S0006-8993(96)00574-4

68

J. Sharkey et al. / Brain Research 735 (1996) 67 82

Recently, we described the effects of ( + ) - l - m e t h y l 1phenyl- 1,2,3,4-tetrahydroisoquinoline (FR 115427) upon the NMDA receptor using radioligand binding and electrophysiological techniques [10,28]. These studies have demonstrated that FRl15427 acts as a non-competitive NMDA receptor antagonist which, although some 14-fold lower in affinity for the NMDA receptor, exhibits less marked use dependence and greater stereoselectivity than dizocilpine. Subsequent studies have demonstrated that, like dizocilpine, FRl15427 is a potent neuroprotectant in animals models of cerebral ischaemia [13,22]. However, while the profile of pharmacological specificity and neuroprotective efficacy is similar to that of dizocilpine, FR115427 appears to produce fewer of the unwanted effects such as weight loss, hypothermia or the increased animal mortality observed following the administration of dizocilpine at neuroprotective doses [13,22]. As yet, its effects upon brain function are unknown. 2-Deoxyglucose quantitative autoradiography has provided considerable insight into the effects of selective inhibition of the various sites within the NMDA receptor complex upon brain function. These studies have shown that non-competitive NMDA receptor antagonists produce a characteristic pattern of alterations in local cerebral glucose utilisation which is distinct from that produced by competitive NMDA receptor antagonists [15,20]. This technique is therefore of great potential value when evaluating the efficacy and potential side effects of novel NMDA receptor antagonists. It might reasonably be anticipated that the subtle differences observed between the pharmacology and toxicity of FRl15427 and dizocilpine be reflected in altered rates of local cerebral glucose utilisation. Therefore, in the present, we compared the metabolic profile of FR115427 to that of dizocilpine and of the competitive NMDA receptor antagonist CGS19755 using the 2-deoxyglucose fingerprinting approach [14].

2. Materials and methods

Fifty-five Sprague-Dawley rats (Harlan Olac, UK), weighing between 270 and 330 g at the time of study, were housed under standard laboratory (or natural day/night cycle) with food and water available ad libitum. All studies were performed in accordance with the UK Animals (Scientific Procedures) Act, 1986. As there are no alternative methods available for the in vitro study of integrated brain function, every effort was made to minimise animal suffering and to reduce the number of animals used. On the day of the study, the rats were anaesthetised with halothane (1.5% in a gaseous mixture of 30% oxygen, 70% nitrous oxide) to permit the insertion of polyethylene cannulae into both femoral arteries and veins. Upon securing the cannulae, the incision sites were infiltrated with a local anaesthetic gel (xylocaine, 2%) and the wound su-

tured closed. A loose-fitting Plaster of Paris cast was applied around the pelvis and lower abdomen and taped to a heavy block. A temperature probe was inserted into the rectum to monitor core temperature, and a pressure transducer attached to a chart recorder (Gould Stratham, Model 2202) was connected to the left arterial femoral catheter to monitor blood pressure. Anaesthesia was discontinued and the rat allowed to recover for at least 2 h prior to the initiation of any further experimental procedure. FR115427 (0.1, 0.3, 1.0, 3.0 mg/kg), dizocilpine (0.03, 0.1, 0.3 m g / k g ) , CGS19755 (1, 10 or 30 m g / k g ) or saline vehicle (2 m l / k g ) was administered by intravenous injection 10 min prior to the measurement of local cerebral glucose utilisation by the fully quantitative autoradiographic [ 14C]2-deoxyglucose technique [31]. 1CGU studies were initiated by an intravenous pulse of [14C]2-deoxyglucose (50 txCi in 0.7 ml of saline, infused over 30 s). Over the succeeding 45 min, 14 timed arterial blood samples (approximately 60 Ixl) were withdrawn for the determination of plasma [14C] and glucose timecourses. Plasma was prepared by the immediate centrifugation of the blood samples in a table top microcentrifuge (Beckman microfuge C). Aliquots of the plasma were taken from each sample for liquid scintillation analysis (20 txl) and for the determination of plasma glucose concentration (10 txl) (Beckman, Glucose Analyser II). At the end of the sampling period the rat was killed by rapid intravenous infusion of barbiturate (Euthetal, 1 ml), the animal decapitated, its brain excised and frozen in 2-methylbutane which had been precooled to - 4 2 ° C by a mixture of acetone and dry ice. The time taken from sacrifice to immersion in isopentane did not exceed 3 min. The frozen brain was transferred to a bed of dry ice where it was affixed to a swivel-headed microtome chuck with a plastic embedding matrix. The brain was sealed in a freezer bag and stored at - 7 0 ° C until sectioned (within 48 h). From each brain approximately 900 coronal brain sections (20 txm) were cut in a cryostat ( - 22°C). Of these, three sections in every ten were picked-up onto glass cover slips and rapidly dried on a hotplate (70°C). The remaining seven sections were discarded, or where necessary stained for histological identification of regions of interest. The coverslips were glued onto cards and placed, together with a series of [~4C]labelled plastic standards (40-1069 ~xCi/g tissue equivalents; Amersham) and autoradiograms obtained by exposing blue-sensitive X-ray film (Kodak SB-5) to the brain sections and standards in a light-tight cassette for five to seven days. Films were then processed according to the manufacturer's instructions.

2.1. Quantitative densitometrical analysis Analysis of the resultant autoradiograms was performed using a computer-based image analysis system (Quantimet-970, Cambridge Instruments). The images were magnified and displayed on a video monitor. Optical density

J. Sharkey et al. / Brain Research 735 (1996) 67-82

measurements (OD) were obtained by placing a cursor (2.5 × 1 0 - 3 r n l T l 2 - 1800 mm 2) over the region of interest. Density values for each brain area were obtained by digitising the field of measurement into picture points (pixels 2.5 × 10 3 mm 2) with each pixel ascribed to one of 256 grey levels. Thereafter a optical density value was obtained by integrating the grey level value of each pixel on each of 4 scans of the structure. Bilateral measurements were made on at least six autoradiograms in which each brain area appeared. Identification of brain areas was made with reference to the stereotaxic atlas of Paxinos and Watson [28]. The [14C] concentration for each brain area was calculated from the mean OD value by reference to the calibration curve constructed from the OD values of the [14C] standards. Rates of glucose utilisation for each brain area were derived from: (a) the local tissue [14C] concentration, (b) the plasma histories of [14C] and glucose, and (c) the appropriate rate constants for 2-deoxyglucose in grey and white matter tissue, by means of the operational equation as derived by Sokoloff et al. [31]. 2.2. Statistical analysis For each of the 3 NMDA receptor antagonists, the local cerebral glucose use values for each of the treatment groups were compared with that of saline-treated control group using the Scheff6 extension of ANOVAR. Analysis of the metabolic patterns produced by FRl15427, dizocilpine and CGS19755 was performed using the deoxyglucose fingerprinting method of Kelly et al. [14]. This method ranks brain regions according to their responsiveness over a chosen concentration range. Hierarchies of regional responsiveness are derived from the function f which is defined as: J--1

: = E (Xc-Xj) 2 j=l

where X c - X j are the natural logarithms of the mean glucose use values obtained for the control group and the jth treatment group respectively. J-1 describes the number of drug concentrations examine for the drug under investigation (usually 3 or 4). For each of the 80 regions examined, f was calculated and used for subsequent inter-regional and inter-drug comparisons.

3. Results

3.1. Physiological and behavioural responses The effects of dizocilpine, FRl15427 and CGS19755 upon cardiovascular and respiratory parameters are described in Table 1. The administration of dizocilpine resulted in dose-related increases in mean arterial blood pressure. At the

69

lowest dose examined (0.03 m g / k g ) , moderate increases ( + 14%) in MABP occurred 6 - 8 min following the administration of dizocilpine and persisted approximately 20 min. At the higher doses of 0.1 and 0.3 m g / k g the increases in MABP occurred progressively (at 4 and 2 min post injection respectively), were greater in magnitude (with maximum increases of 24% and 27% above pretreatment levels) and persisted throughout the 55-min measurement period. These increases in MABP were accompanied both in timecourse and intensity by the characteristic head weaving stereotypies reported elsewhere [4,15,23]. The intravenous administration of FR115427 also produced dose-related alterations in cardiovascular and respiratory status (Table 1) and upon the overt behaviour of the animals tested. Significant increases in MABP were observed within 2 min following the injection of FR115427 at doses of 0.3, 1.0 and 3.0 m g / k g (12-20%, data not shown). The duration of the hypertensive response ranged from less than 1 min (0.3 m g / k g ) to approximately 20 min (3 m g / k g ) . At the lowest dose examined (0.1 m g / k g ) FRl15427 had little effect upon gross behaviour, body temperature or upon any of the cardiovascular or respiratory variables examined (Table 1). Following the administration of 0.3 m g / k g FRl15427, rats exhibited episodic head weaving over the first 2 - 3 min following injection. Thereafter, the animals remained quiescent for the rest of the experimental period. At higher doses the episodes of stereotyped head weaving became more protracted and intense, persisting for some 10-15 min at the 1 m g / k g dose and throughout the experimental period at the 3 m g / k g dose. CGS 19755 also produced a distinctive pattern of alterations in the cardiovascular and respiratory status and upon the overt behaviour of the animals tested. At 1 m g / k g CGS19755 produced no discernible change in the behaviour of the animals nor in any of the cardiovascular or respiratory parameters measured (Table 1). Rats receiving CGS19755 at 10 m g / k g exhibited episodic chewing behaviours which began 4 - 6 min post injection. In addition these animals became mildly hypotensive ( - 14%) approximately 20 min post injection. By contrast, rats receiving 30 m g / k g CGS19755 showed no evidence of a hypotensive response at any time during the experimental period. Indeed there was a transient hypertensive response associated with the injection of this dose of CGS19755. Rats receiving this dose of CGS19755 exhibited episodic chewing within 3 min of the injection. However, within 25 min the chewing behaviours had subsided with the animals remaining motionless for protracted periods. These behaviours were accompanied by a mild respiratory acidosis (Table 1). 3.2. Local cerebral glucose utilisation The effects of dizocilpine, FRl15427 and CGS19755 on the local rates of cerebral glucose utilisation in 81

36.9_+0.3 37.0_+0.3 37.0_+0.2 122_+3 142_+6 ** 130_+6 ** 1.34_+0.23 1.33_+0.20 2.18_+0.69 40.4_+2.1 39.2_+4.2 36.5+_2.9 92.8-+9.6 89.6_+7.6 90.9_+8.0 7.39_+0.05 7.38_+0.04 7.40_+0.02 23.8-+0.5 23.0-+2.9 22.6+ 1.2

37.2-+0.2 37.2_+0.2 37.2_+0.2 125-+3 133+_6 t21+3 1.23_+0.12 1.20_+0.12 1.38-+0.20 38.0_+4.3 37.5_+4.3 38.0_+2.8 88.0_+6.7 90.7_+5.5 86.7_+6.8 7.43_+0.02 7.44_+0.02 7.41-+0.03 25.0_+1.6 25.1 -+ 1.0 24.2_+ 1.0

26.5+2.4 26.1 - + 2 . 3 22.5-+2.0

7.45-+0.05 7.46_+0.06 7.38_+0.04

88.9_+10.0 91.0+_6.1 93.0_+4.9

40.7_+2.6 39.0-+1.4 37.9_+ 1.6

1.36_+0.18 1.42_+0.19 1.66_+0.21

119_+4 145_+2 ** 130+6"

37.2_+0.2 37.4_+0.2 37.2_+0.3

0.3

24.6-+0.6 25.1_+0.6 23.0-+0.9

7.40_+0.01 7.41_+0.01 7.40_+0.02

87.5+5.8 92.8-+2.2 92.3_+2.1

39.8_+1.2 39.9-+2.3 37.7_+3.2

1.28_+0.13 1.23_+0.07 1.44_+0.07

118_+2 121_+2 110_+6

37.3_+0.1 37.3_+0.1 37.2_+0.1

0.1

24.9-+0.3 25.4+1.3 23.5_+0.9

7.42_+0.02 7.42_+0.01 7.39_+0.02

85.5_+4.2 88.9-+4.8 89.0_+5.7

38.5+2.2 39.7-+1.9 38.8-+2.2

1.22-+0.12 1.21_+0.09 1.34_+0.11

124_+2 131-+6 119_+9

37.1 _+0.1 37.2_+0.1 37.0_+0.1

0.3

FR115427 (mg/kg)

Data are presented as mean glucose use _+one standard deviation for 5 animals per treatment group. * P < 0.05 when compared to saline vehicle. * P < 0.05 when compared to pretreatment values.

pre drug 37.2-+0.3 5 min post d r u g 37.4_+0.2 45 min post drug 37.1-+0.3 MABP (mmHg) pre drug 121_+8 5 min post drug 124_+6 45 min post drug 115_+10 Plasma glucose (mg/dl) pre drug 1.39_+0.24 5 min post d r u g 1.32_+0.21 45 min post d r u g 1.45-+0.32 pCO 2 (mmHg) pre drug 39.9_+3.0 5 min post drug 39.0_+3.2 45 min post drug 36.9_+3.0 pO 2 (mmHg) pre drug 83.1_+4.5 5 min post d r u g 92.9_+7.1 45 min post drug 97.2_+4.7 ptt (units) pre drug 7.42-+0.02 5 rain post d r u g 7.45_+0.05 45 min post d r u g 7.43_+0.03 HCO 3 (mM) pre dmg 25.9-+3.0 5 min post drug 26.8-+3.7 45 min post drug 24.3_+2.7

0. l

0.03

vehicle

Body temperature (°C)

Dizocilpine (mg/kg)

Saline

Table 1 Physiological variables

24.3_+1.8 25.7_+ 1.1 22.2_+ 1.9

7.41_+0.02 7.42_+0.01 7.37_+0.01 *

87.1_+2.5 90.5+4.9 86.1 - + 6 . 2

38.7_+2.6 40.3_+1.9 38.7-+3.5

1.3//+0.(/6 1.28_+0.08 1.47_+0.08

123_+4 145-+6 ** 119_+8

37.5_+0.2 37.4_+0.2 37.4_+0.1

1.0

23.1_+1.2 24.3_+0.7 20.9+ 1.9

7.39_+0.03 7.36_+0.06 7.36_+0.03 *

90.1_+4.8 88.8+4.0 90.5_+6.6

38.3_+2.1 39.9_+1.9 37.0_+2.8

1.22+0.09 1.15_+0.08 1.73_+0.13

122_+2 146-+4 ** 126+_6

37.1 _+0.1 37.1 _+0.1 37.2_+0.2

3.0

24.4-+ 1.3 24.5_+2.0 23.2-+ 1.9

7.40_+0.02 7.40_+0.02 7.38_+0.03

89.1_+5.5 86.5+_3.3 87.4+4.8

38.4_+1.2 39.6_+2.9 39.0-+1.9

1.28_+0.18 1.30+0.18 1.43+_0.20

124_+3 125_+3 117_+7

37.1 _+0.3 37.1 _+0.3 37.0_+0.5

1

24.1_+1.7 24.9_+1.5 23.9-+ 1.2

7.42_+0.02 7.41_+0.01 7.38-+0.03

87.2_+7.8 88.9-+3.5 90.2-+7.5

37.5_+2.1 39.4_+2.3 40.4+2.0

1.35+0.17 1.34-+(/.16 1.30_+0.16

120+_4 121_+4 106_+8"

37.0+0.3 37.2_+0.1 37.0_+0.2

l0

CGS 19755 (mg/kg)

25.1-+1.5 25.6-+0.8 23.9-+ 1.3

7.39_+0.02 7.38_+0.01 7.34-+0.02 *

86.2_+7.6 87.9-+8.7 89.1 -+3.8

41.7_+1.6 43.6_+1.4 44.9_+3.5*

1.07_+0.14 1.08-+0.11 1.28+0.28

122_+2 129_+7 129_+7

37.1 _+0.1 37.0_+0.3 37.0_+0.2

30

"-4 i

t.n

2'

95±9 105_+6 111_+10 74-+8 74_+11

129_+10 111_+14 151-+7 109-t-9 145_+11 139_+12

114_+11 107_+6 172_+16 112-+8 154-+19 161_+13

87_+15 108-t-10 112+7 56±9 84_+9

172_+19

82 ± 7 76-+13 100+12

114_+10 119-+7 87+11

0.1

88+13 113-+16 113_+7 65-+7 72_+4

116-+13 95_+12 160_+11 107_+10 158_+10 150+15

157_+8

66 -+ 6 73+7 98_+11

122_+12 120+19 76_+9

0.3

FR115427(mg/kg)

183±20

71 ± 5 74-+5 96_+8

110_+8 118+6 81_+3

2

(ml/kg)

92_+7 106_+14 116_+16 63±2 75_+6

95+11 87_+6 156_+21 122-+11 177-t-13 139_+11

123_+13 *

65 ± 6 75-+8 93-+10

109_+11 126-+14 80_+4

1

99_+7 106-+8 116_+10 54_+8 92_+5

72_+5 * 73_+6 * 104_+5 * 92±9 129_+7 130-+14 *

95_+9 *

62 -+ 4 78_+3 111_+17

141_+21 107+13 72+4

3

97_+8 114_+10 112_+4 70_+5 85+9

88_+10 * 91_+10 136_+20 * 107±15 145_+8 140_+20

118_+14 *

68 -+ 4 76_+5 90-+9

111+9 121+9 79_+4

0.03

101+12 105-+7 141+15 * 67_+8 79_+10

72_+6 * 76_+8 * 99_+17 * 84_+11 * 131_+4 134-+17

93-+12 *

70 -+ 5 79+8 106-+16

104_+8 116-+8 81_+6

0.1

Dizocilpine (mg/kg)

160+16 152±12 179_+16 111_+17 148_+29

83_+16 * 105_+11 108_+5 * 98_+10 138_+11 129-+9 *

94_+12 *

* * * * *

66 _+ 7 84_+7 132_+10 *

113_+13 155_+9 * 87+7

0.3

82+5 101-+7 106_+18 57_+4 63-+9 *

112_+3 101-+6 150+14 92_+10 130_+17 151±15

132-+10 *

66 ± 5 71+2 89_+7

97+6 101_+6 77-+4

1

93-+9 90-+9 * 138_+7 * 64_+3 79-+13

87_+9 * 82-+8 * 135-]-15 * 102_+13 151_+15 126_+11

103±16 *

59 -+ 4 * 59_+4 * 73_+11 *

92_+9 * 95+13 * 65-+8 *

10

CGS19755 (mg/kg)

Data are presented as mean glucose use _ one standard deviation for 5 animals per treatment group. * P < 0.05 as revealed by A N O V A with post-hoc Scheff6 test.

Cortical amygdaloid nucleus Mediodorsal thalamic n. Olfactory tubercle Entorhinal cortex Prepyriform cortex

Primary olfactory

Primary auditory cortex Medial geniculate (lateral) (medial) Inferior colliculus n. lateral lemniscus Superior olivary body Cochlear n.

Primary auditory

(superficial layers) (profundum) Anterior pretectal area

Superior colliculus

Visual cortex: (Areal) (Area 2) Lateral geniculate body

Primary visual

Saline

Table 2 Local glucose ufilisation in sensory structures

84_+4 83_+6 * 146_+ 10 58 _+ 10 78+6

*

69+6 * 74_+ 11 * 93_+9 * 78 _+ 10 * 120_+25 116_+24 *

82_+8 *

50_+6 * 54_+7 * 70+4 *

83+3 * 82+_11 * 67_+6 *

30

72

J. Sharkey et al. / Brain Research 735 (1996) 67-82

regions of the conscious rat brain are described in Tables 2-5. 3.2.1. P r i m a r y s e n s o r y s t r u c t u r e s

The effects of dizocilpine, F R l 1 5 4 2 7 and CGS19755 upon glucose utilisation in primary auditory, visual and olfactory areas are described in Table 2. Dizocilpine, F R l 1 5 4 2 7 and CGS19755 produced widespread reductions in 1CGU within components of the auditory system (Table 2). Following the administration of 0.03 m g / k g dizocilpine, significant reductions in 1CGU were evident in primary auditory cortex, medial geniculate body and inferior colliculus. Further and more widespread reductions in 1CGU were evident at the higher doses of 0.1 and 0.3 m g / k g (Figs. 1 and 2). FR115427 and CGS19755 produced a similar pattern of altered 1CGU in auditory structures to that of dizocilpine, although both agents required a 30 and 100 fold higher doses than dizocilpine, respectively (Figs. 1 and 2). Dizocilpine had little effects upon the various anatomical components of the visual system with the exception of Area 2 of visual cortex and anterior pretectal area which exhibited significantly increased 1CGU (of 31% and 37% respectively) at the 0,3 m g / k g dose. No significant alterations in 1CGU were observed within the visual components of animals treated with F R l 1 5 4 2 7 . In contrast, CGS19755 produced dose-related reductions in 1CGU throughout the visual areas examined. Dizocilpine produced marked increases in 1CGU in each of the olfactory areas examined, though only at the highest dose examined (0.3 m g / k g ) . The most marked elevations in 1CGU were found in entorhinal cortex ( + 98%), cortical amygdaloid nucleus ( + 84%) and primary olfactory cortex ( + 7 6 % ) . CGS19755 produced a different pattern of altered glucose use in primary olfactory areas with significant reductions ( - 2 5 % ) in primary olfactory cortex of animals treated with l m g / k g CGS19755. Dose-related reductions in 1CGU were also found within the mediodorsal thalamic nucleus. In contrast, FR115427 failed to significantly alter 1CGU in olfactory areas. 3.2.2. C e r e b r a l c o r t e x

The effects of dizocilpine, F R l 1 5 4 2 7 and CGS19755 upon glucose utilisation in cerebral cortex and myelinated fibre tracts are described in Table 2. Dizocilpine administration resulted in a heterogeneous pattern of altered glucose use within cerebral cortex with

marked dose-related increases in glucose utilisation measured in insular, piriform and retrosplenial cortices, reductions in frontoparietal, somatosensory and forelimb area of motor cortex and little change in cingulate, frontal cortex and in the hindlimb area of motor cortex. FR115427 produced dose-related reductions in 1CGU in frontoparietal and forelimb area of motor cortex. Within somatosensory cortex, there was evidence of biphasic changes with significant increases in 1CGU (25%) evident at 0.3 m g / k g and reductions of 40% at 3 m g / k g . No significant alterations in lCGU were observed in retrosplenial, piriform and insular cortices. CGS19755 significantly altered 1CGU in 9 of the 11 cortical areas examined. Significant increases in 1CGU were observed in the piriform cortex of rats which had received CGS19755 at 10 and 30 m g / k g while concomitant reductions were noted in both the hindlimb and forelimb areas of motor cortex. Five areas of cortex exhibited significant alterations in ICGU at only one of the three doses examined. In retrosplenial cortex significant reductions in 1CGU were observed only at the 10 m g / k g dose, while in anterior cingulate, frontal and parietal cortices CGS19755 significantly reduced 1CGU, but only at the highest dose examined. Dizocilpine, F R l 1 5 4 2 7 and CGS19755 produced distinct alterations in 1CGU within myelinated fibre tracts (Table 3). Dizocilpine and FR115427 produced dose-related increases in 1CGU within the fornix (Table 3). Increases in 1CGU of 26% and 93% were observed in the fornix of animals which had received 0.03 and 0.1 m g / k g dizocilpine whereas significant increases in fornix glucose use were observed only at the 30-fold higher doses of 1 m g / k g ( + 2 4 % ) and 3 m g / k g F R l 1 5 4 2 7 ( + 5 5 % ) . In contrast, CGS19755 failed to significantly alter glucose use in the fornix. Marked increases ( + 107%) in glucose use were evident in the anterior commissure of rats which had received 0.3 m g / k g dizocilpine whereas neither F R l 1 5 4 2 7 nor CGS19755 effecteJ significant alterations in commissural glucose use at any of the doses examined. Significant reductions in glucose use were found in the internal capsule of animals which had received CGS19755 at every dose tested. 3.2.3. H i p p o c a m p u s a n d s u b c o r t i c a l limbic a r e a s

The effects of dizocilpine, F R l 1 5 4 2 7 and CGS19755 upon local glucose use within hippocampus and subcortical limbic areas are presented in Table 4.

Fig. 1. Effects of dizocilpine (0.03-0.3 mg/kg), FR115427 (0.3-3 mg/kg) and CGS19755 (1-30 mg/kg) upon local cerebral glucose use in auditory cortex, inferior colliculus, fornix, presubiculum, pontine nuclei and entorhinal cortex. Data are presented as mean ± one standard deviation for 5 rats per dose. All 3 NMDA receptor antagonists produced parallel reductions in local glucose use within the auditory structures (auditory cortex and inferior colliculus). Within components of the limbic (presubiculum), sensorirnotor (pontine nuclei) systems and in the white matter tract of the fornix, both dizocilpine and FR115427 produced marked dose-related increases in ICGU, although FR115427 appeared some 30-fold less potent. In contrast, CGS19755 had little effect at any of the doses examined. Dizocilpine produced marked increases in 1CGU (+98%) within entorhinal cortex at the 3 mg/kg dose. However, neither FR115427 or CGS19755 significantly effected ICGU within this allocortical region at any dose examined.

73

J. Sharkey et al. / Brain Research 735 (1996) 67-82

Dizocilpine produced significant alterations in 4 of the 7 areas of the hippocampal complex examined; FRl15427 in 3 of the areas and CGS19755 in 2 areas. In dizocilpine-

Auditory Cortex

200

_=

treated rats the most marked alterations were found in the presubiculum where increases in ICGU of 33% were observed at 0.03 m g / k g rising to 110% at 0.3 m g / k g (Figs.

Inferior Colliculus

200

180

180

160

160

140

140

120

120

100

100

80

80

60

60

Presubiculum

200

Pontine Nuclei

140

~g 180

~g

120 160 100

140 120

80

100 60 80 60

F

40

Fornix

140

Entorhinal Cortex

140

120

120

100

100

80

80

60

60

o

_=

40

r--~l

, 0.01

, 0.10

, 1.00

Dose (mg/kg)

, 10.00

i

100.00

40

~ ~

1

r--~l

, 0.01

9

I 7

, 0.I0

.

5

, 1.00 Dose (mg/kg)

5

, 10.00

, 100.00

.

•

,,%

=

=R115427

"':'-

Vehicle

S

#.

. .~-S

•

CGS19755

3izocilpine

I

"4

L.~

P~

#

J. Sharkey et al. / Brain Research 735 (1996) 67-82

75

1 and 2). Increases in presubicular glucose use observed in rats treated with FR115427 paralleled those produced by dizocilpine; however, the doses of FR115427 required to produce similar increases in 1CGU were some 30-fold higher than for dizocilpine. In contrast, CGS19755 failed to increase glucose use in the presubiculum; indeed reductions of - 16% were observed in rats receiving 10 m g / k g of the competitive NMDA receptor antagonist. A similar pattern of parallel increases in ICGU in dizocilpine and FRl15427, but not CGS19755-treated rats was observed within the molecular layer of the dorsal hippocampus and within the dentate gyrus (Fig. 2). Conversely, dose-related reductions in glucose use within the subiculum were observed in CGS19755-treated rats whereas in FR115427and dizocilpine-treated animals glucose use remained relatively unaffected. Dizocilpine significantly altered local glucose use in 9 of the 18 subcortical limbic areas examined. Prominent increases in 1CGU were found within the mamillary body, basolateral amygdaloid nucleus and in a number of limbic areas of the thalamus (lateroposterior n., laterodorsal n., posterior n., anteroventral n. and nucleus gelatinosus). Concomitant reductions in ICGU were observed within the dorsal tegmental nucleus and mediolateral nucleus of the habenula. A similar pattern of altered glucose use was tbund in FR115427-treated rats, although again, a 30-fold higher dose was required. In contrast, CGS 19755 produced widespread reductions in limbic glucose use. Significant reductions in 1CGU were evident within the dorsal raphe nucleus of rats which had received the lowest dose of CGS19755 (1 m g / k g ) . At the 10 m g / k g dose, significant reductions in 1CGU were noted in dorsal tegmental nucleus ( - 21%), median raphe ( - 20%), mamillary body ( - 27%), mediolateral habenular nucleus ( - 2 6 % ) , lateroposterior ( - 16%), anteromedial ( - 24%), anteroventral ( - 35%) thalamic nuclei. At the highest dose (30 m g / k g ) , CGS19755 tended to produce further reductions in limbic glucose use within these limbic areas with significant reductions also found in the pontine reticular formation, basolateral amygdala, lateral hypothalamus and lateral septal nucleus. CGS 19755 significantly increase local glucose use in only one limbic area, the posterior thalamic nucleus.

Dizocilpine effected dose-related alterations in 11 of the 23 areas examined. Significant increases in 1CGU were observed within the cerebellum ( + 36%), globus pallidus ( + 4 7 % ) and pontine nuclei ( + 32%; Fig. 1) of rats receiving the lowest dose of dizocilpine (0.03 m g / k g ) . Further increases were noted in these areas at the higher dose of 0.1 m g / k g and these were accompanied by significant elevations in ICGU within the pars reticulata of the substantia nigra ( + 54%) and entopeduncular nucleus ( + 43%). Significant alterations in ICGU were also observed within the red nucleus ( + 20%) and posterioventrolateral thalamic nucleus ( - 2 5 % ) , but only at the 0.1 m g / k g dose. At the highest dose tested, additional structures exhibiting elevated 1CGU included the rostral (but not caudal) aspects of the nucleus accumbens ( + 4 0 % ) , dorsolateral caudateputamen ( + 26%), ventral pallidum ( + 34%) and subthalamic nucleus ( + 23%). FR115427 failed to alter 1CGU in any of the extrapyramidal and somatosensory structures examined at the 0.1 and 0.3 m g / k g doses. At 1.0 m g / k g FRl15427, significant alterations were measured in one area, the pontine nuclei ( + 4 2 % ) while at 3.0 m g / k g , increases in 1CGU were also observed in the cerebellum ( + 47%), pars reticulata of the substantia nigra ( + 56%) and both the internal and external segment of the pallidus (34% and 47% respectively). Significant reductions within the pars compacta of the substantia nigra ( - 2 2 % ) and vpl thalamic nucleus ( - 21%) were also noted at this dose of FR 115427. CGS19755 significantly altered 1CGU in 9 of the 23 extrapyramidal areas depicted in Table 4. At the lowest dose of 1 m g / k g CGS19755 reduced 1CGU within the pvl thalamic nucleus ( - 19%), with further reductions in 1CGU observed with 10 and 30 m g / k g . At the 10 m g / k g dose, significant reductions in 1CGU were also noted within the cerebellum (cortex and flocculus), vestibular nucleus, nucleus accumbens and ventromedial thalamic nucleus At the highest dose of CGS19755 examined (30 m g / k g ) , additional structures exhibiting suppressed glucose use include the central nucleus of the inferior olivary body, nucleus accumbens and pars compacta of the substanta nigra. CGS 19755 administration result in an increase in 1CGU in only one area, the pars reticulata of the substantia nigra.

3.2.4. Extrapyramidal areas The effects of dizocilpine, FRl15427 and CGS19755 upon local glucose use within extrapyramidal and somatosensory areas are presented in Table 5.

3.3. Metabolic fingerprinting Plots of the regional f values derived from the dose-response curves for each of the 80 brain regions examined

Fig. 2. Pseudocolour reconstructions of [14C]2-deoxyglucose autoradiographs prepared form coronal sections of rat brain taken at the level of the medial geniculate bodies. Top panel: saline-treated control animal (left) and dizocilpine- (0.3 m g / k g ) treated rat (right). Bottom panel: FR115427- (3 m g / k g : left) and CGS19755- (30 m g / k g : right) treated animals. The location of retrosplenial cortex (RSG), auditory cortex (AC), medial genicnlate bodies (Mg) hippocampus molecular layer (Mol) and entorhinal cortex tEnt) are indicated by arrows. Note the relative increase in glucose use within the retrosplenial cortex and entorhinal cortex of the dizocilpine-treated rat compared with the other three treatments. Increases in glucose utilisation within the molecular layer of the bippocampus are increased by dizocilpine and FRl15427, but not by CGS19755. Glucose use within the auditory cortex and medial geniculates are reduced by all three NMDA receptor antagonists.

106-+8 121_+10 116_+12 103_+15 68-+7 108_+10 108_+8 126+12 99_+ 12 99_+11 33 -+ 2 32 -+ 5 23-+2 59_+4 36_+4

107_+5 112+7 107_+ 12

93_+6 59_+5

107+8 107_+7 120+5

101 _+8 102_+7

35 _+4 30 _+2 23_+3 58+3 41 + 2

35 -+ 5 32 + 2 21 -+2 61_+9 41 _+6

103_+ 16 101_+9

105+9 101_+8 150+17 *

1(/3-+4 63-+2

103-+ 10 114-+11 119+8

32 + 2 34 + 7 22-+3 72_+7 * 32_+5

105_+5 105+14

99+13 86_+7 125+12

98+13 65_+5

103_+7 108-+13 lll-+ll

1.0

30 -+ 1 34 _+4 21 + I 90_+5 * 43-+3

104_+6 116_+9

91-+7 75+7 * 71_+4 *

80_+8 64-+5

101 + 6 65_+5 * 104_+7

3.0

34 _+3 29 _+3 21 _+3 73+9 37-+3

105_+5 103+7

102_+4 88_+8 * 111_+10

89_+8 61_+6

99_+8 99_+6 110+8

0.03

32 + 5 31 + 4 22_+3 112+_11 * 40_+8

121 _+ 10 * 115+11

95_+12 65_+7 * 85_+16 *

91+14 67_+12

96_+9 70-+13 * 110_+6

O. 1

Dizocilpine ( m g / k g )

37 + 4 33 _+4 28_+2 104_+12 * 85-+ 11 ~

130_+6 * 165_+9 *

105_+10 74_+8 * 77_+6 *

117_+13 * 101_+10 *

96_+8 69-+6 * 130_+22

0.3

29 _+2 * 32 _+ 2 20+ 1 57_+6 38_+2

94_+6 89-+6

93_+12 92+15 108_+15

96+9 59_+8

97+6 98_+11 104_+8

1

27+ 1 * 25_+3 23_+2 52+3 42_+2

78-+11 ~ 81_+9

86-+8 * 86-+8 * 111_+14

98_+ 15 60_+4

97 + 13 104_+ 16 89 + 16

10

CGS19755 ( m g / k g )

Data are presented as mean glucose use_+ one standard deviation for 5 animals per treatment group. * P < 0.05 as revealed by A N O V A with post-hoc Scheff~ test.

Neocortex Frontal cortex: area 1 Frontoparietal cortex Anterior cingulate cortex Insular cortex granular agranular Motor cortex Hindlimb area Forelimb area Parietal:areal Retrosplenial cortex granular layers agranular layers Subeortical white matter Internal capsule Cerebellar white matter Genu of corpus callosum Fornix Anterior commissure

0.1

2

0.3

FR115427 ( m g / k g )

Saline ( m l / k g )

Table 3 Local glucose utilisation in neocortex and subcortical white matter

27_+2 * 28_+2 22+ 1 58_+11 45_+2

86_+9 93+10

60_+8 71_+4 * 81+8 *

85_+10 61 + 10

8 6 + 10 * 72_+ 12 * 77±5 *

30

Ix)

45±4 61_+2 65_+5 50_+2 84_+7 69 _+8 86_+8 92+7 75_+7 98_+5 56_+2 100_+18 96±11 40+4 73_+6 38+_2 42_+7 113_+13 89_+8 106+13 84_+10 112_+7 114_+11 116_+13 93_+14

46_+2 67-+5 65_+1 47_+2 79___4 72 _+4 83+_6 102_+6 81+3 91_+6 57_+5 99_+7 100_+8

39_+4 76_+5

38_+2 43_+6

105_+9 102_+8

93_+5 80_+2 118±9 126_+5 116_+10 88_+7

99_+12 96_+9

38±5 42_+3

38_+5 80_+12

45-+5 67-+5 65_+8 54-+5 92-+1 76 _+6 111_+13* 93±4 87_+11 92+14 64-+5 94_+10 108+7

99_+9 106_+5 78+_8 80_+6 1 2 2 _ + 1 7 114_+11 114_+5 138+13 131_+9 116_+11 100±9 102_+11

104_+6 90+9

36_+3 38_+4

39_+3 74_+6

42_+5 63-+4 65_+4 47_+5 84-+7 69 _+6 91_+6 95_+6 76_+4 84-+3 55+4 96_+11 103_+9

113+_11 * 82_+5 113_+5 150_+9 105+9 137-+12 *

103_+8 80_+3 *

37_+2 36_+4

37+_1 98_+5 *

40-+2 70_+4 64_+5 62+5 * 94_+7 * 70 _+8 135_+11 * 87_+6 * 79_+2 83+4 69-+3 96_+7 133+_9 *

108_+7 74_+4 111_+6 124_+10 123+_6 114±12 *

100_+8 83_+3

42_+3 44+1

42_+4 86_+8

50-+1 72_+4 68_+8 57_+4 89_+4 74 ___8 110_+8 * 98_+9 79_+4 90+-7 62+-2 97_+9 108_+8

126_+20 * 79_+7 102+_11 130_+10 114_+10 126-+19 *

107_+15 88+_11

41 _+5 43+7

39_+4 95_+11 *

49±7 73_+9 69_+8 66_+9 * 103_+14 * 82 + 7 135±11 * 89_+6 77_+7 82_+6 63_+4 103_+7 131_+14 *

0.1

141_+8 * 106_+12 * 124_+9 180+_20 * 141-+13 * 162_+11 *

102_+14 96_+11

42_+2 45-+5

42_+4 108_+9 *

46+6 85±10 * 71+6 77_+12 * 137_+16 * 79 _+6 174_+6 * 91_+7 77_+7 85_+7 67_+7 101+__11 139+_16 *

0.3

94_+6 75_+2 107_+9 102+_8 104_+2 96+9

101-+12 88_+10

38_+3 39-+2

38_+1 72_+3

47-+3 66_+3 64_+3 51_+3 80_+8 69 _+2 77_+4 91+11 71+4 * 83_+6 52_+5 95_+8 95_+8

1

34_+4 64_+6 * 32+3 * 32+3 * 121 -+ 13 61 -+5 * 88_+3 112±11 * 87_+6 * 9 6 + 18 * 105 -+_13 91_+5

36_+ 1 36_+3 99_+8 75_+9 * 78+9 * 9O_+6 92+9 * 82_+ 10 * 99_+ 12 80_+9

*

* * * *

38+3 71_+6

*

44_+4 65_+5 62_+4 45_+4 75±5 59_+6 79_+9 71_+4 54_+3 63_+3 45+_3 85-+4 75_+9

30

41-+4 60-+7 60_+3 47_+2 73_+7 61_+6 * 70_+5 * 80_+ 10 * 66_+6 * 73+7 * 51_+2 85_+ 13 73_+8 *

10

CGS 19755 (mg/kg)

Data are presented as mean glucose use±one standard deviation for 5 animals per treatment group. * P < 0.05 as revealed by ANOVA with post-hoc Scheff4 test.

Dorsal hippocampus CAI CA2 CA3 Dentate gyrus CAI: molecular layer Subiculum Presubiculum Dorsal tegmental nucleus Dorsal raphe nucleus Median raphe nucleus Pontine reticular formation Interpeduncular nucleus Mamillary body Amygdala lateral nucleus basolateral nucleus Hypothalamus lateral nucleus Lateral septal nucleus Habenula lateral nucleus mediolateral nucleus Thalamus lateroposterior n. posterior n. anteromedial n. anteroventral n. gelatinosus laterodorsal n.

3

0.03

1

0.1

2

0.3

Dizocilpine (mg/kg)

FR 115427 (mg/kg)

Saline (ml/kg)

Table 4 Local glucose utilisation in limbic structures

I

t.n

59_+8 82_+8

104-+3 90 _+7

75_+2 110_+12 85_+7 44 _+5 49 _+5 59 _+5

73 _+5 48 _+4 79 _+5

49 -+ 3 74_+2 109_+13 88 + 7 106_+2 59-+6 69 _+4

76 _+6 70_+8

51 -+6 86_+9

104_+7 87 _+4

79_+6 109_+7 80_+6 49 _+3 50 _+ 2 60 _+6

72_+7 49 _+4 84 -+ 9

47 + 6 76_+11 93_+8 96 _+9 110_+8 74-+10 74 + 6

78 + 9 74_+13

52+8 80_+4

110_+6 88 _+4

75_+4 110_+ 11 78_+ 11 52 + 4 56 _+3 57 -+ 4

71 _+9 50 -+ 7 85 _+6

46 _+3 80-+6 108_+8 96 _+6 107+7 70_+4 69 + 4

80 + 5 72_+7

82_+11

0.3

0.1 65_+4

2

65_+9

1

52_+6 83+_14

112+8 91 _+ 11

73_+8 110±10 81 _+5 50 _+7 56 _+5 58 + 9

66_+4 54 + 4 90 _+9

47 -+ 6 87+13 107_+11 96 _+ 10 104_+10 84_+8 * 79 -+ 8

87 -+ 11 80_+7

77_+15

50_+5 94_+9

103+4 86 _+ 8

59_+4 * 105_++6 71 _+8 59 _+3 * 72 _+7 * 66 _+6

57_+4 * 75 -+ 6 * 95 + 6

44 -+ 10 109-+17 * 124_+4 79 + 8 113_+7 107-+10 * 77 _+3

79 -+ 5 72_+4

81_+7

3

61 _+ 11 90_+7

107-+9 85 _+4

72_+4 115_+5 78_+6 59 _+7 72 _+3 * 64 -+ 5

70_+8 62 _+4 84 _+4

44 -+ 3 101-+9 108+10 91 + 7 102_+5 78_+10 * 76 + 6

75 + 3 73_+5

74_+6

0.03

53 _+5 99+12

117_+12 92 _+ 8

66_+7 111_+16 64-+9 * 63 _+7 * 72 + 2 * 69 -+ 5

71 + 5 74 _+4 * 89 _+7

47 + 7 121_+22 * 118_+17 84 _+ 10 105_+7 104_+7 * 83 _+7 *

83 -+ 11 80_+10

69+_8

0.1

Dizocilpine ( m g / k g )

64_+ 10 115+11 *

131_+ II ' 103 _+ 11

68_+8 133_+9 72_+4 72 _+ 8 * 91 _+ 13 * 79 _+7 *

61 _+5 95 _+ 16 * 97 _+ 11 *

48 -+ 3 113_+13 * 115_+14 88 + 8 112_+7 119_+11 * 76 _+ 8

79 _+9 75_+10

70+5

0.3

53_+5 76_+10

101 -+7 81 _+9

72_+3 94_+3 69+3 * 44 _+2 55 _+2 54 _+3

70_+6 52 _+ 2 76 + 6

47 _+2 73_+2 96_+10 88 _+4 106_+10 60_+5 70 + 8

72 + 6 66-+4

66+9

1

*

* * *

57 _+5 72_+10

90_+11 86 _+9

73_+7 87+11 * 68_+9 * 47 _+6 53 + 3 54 _+4

68_+8 48 + 8 81 _+ 11

41 + 3 57_+4 77_+8 88 _+7 90_+8 50_+4 63 _+3

76 _+6 63_+4

71_+4

10

CGS19755 ( m g / k g )

Data are presented as mean glucose use ± one standard deviation for 5 animals per treatment group. * P < 0.05 as revealed by A N O V A with post-hoc Scheffd test.

(caudal) (mstral)

Nucleus accumbens

(dorsolateral) (ventrolateral)

Caudate-putamen

ventroposteriomedial n. ventromedial n. posterioventrolateral Entopeduncular nucleus Globus pallidus Ventral pallidum

Thalamus

(pars compacta) (pars reticulata) Subthalamic nucleus

Substantia nigra

(minimum) (maximum) Flocculus Cerebellar nuclei Vestibular nucleus Pontine nuclei Red nucleus

Cerebellar hemisphere

(lateral nuclei) (central nuclei)

Inferior olivary body

Spinal trigeminal nucleus

FRt 15427 ( m g / k g )

Saline ( m l / k g )

Table 5 Local glucose utilisation in extrapyramidal and somatosensory structures

42_+3 * 74_+ 10

100+8 90+_3

73+12 94_+8 56_+4 * 45_+4 55_+4 71_+8

60_+5 * 62_+5 * 83_+9

38_+2 * 79_+7 78_+9 * 82_+5 88+2 ' 59_+3 65_+2

72_+4 58_+3 *

67+3

30

I

"-1

van

Q

79

J. Sharkey et a l . / Brain Research 735 (1996) 67-82

metabolic fingerprinting approach can be used to identify drugs acting at a specific receptor population (e.g. the NMDA receptor) by means of their metabolic profile, and to differentiate between agents acting at different sites within the receptor.

for dizocilpine, FR 115427 and CGS 19755 are presented in Fig. 3. A rank-order of the f values provides a hierarchy of regional responsiveness to each of the three drugs. These plots reveal that in each case, the majority of brain regions are relatively unaffected by drug treatment, with 54 of the 80 areas exhibiting an f value < 0.15 for dizocilpine; 67 areas with FRl15427 and 60 areas with CGS19755. The overall profiles of metabolic responsiveness produced by dizocilpine and FR115427 are similar: in each case auditory cortex is most sensitive followed by the pontine nuclei, presubiculum and fornix. However, some regions (e.g. entorhinal cortex) exhibited marked alterations in 1CGU in dizocilpine-treated rats (Fig. 1) but exhibited little change in FR115427-treated rats. Nevertheless, regression analysis confirmed the strong overall relationship between the metabolic responses produced by dizocilpine and FRl15427 with a Pearson's product moment correlation of 0.78 ( P < 0.0001). The highest f values in CGS19755-treated rats were also found in auditory structures (auditory cortex and inferior colliculus). However, unlike dizocilpine and FR115427, CGS19755 had little effect upon 1CGU within pontine nuclei, presubiculum or the white matter tracts of the fornix and overall there were poor correlations between the pattern of regional responsiveness produced by CGS 19755 and either dizocilpine ( r = 0.28) or FR115427 ( r = 0.52). Thus, the present data support the view that the

4. D i s c u s s i o n

The present data highlight the power of the metabolic fingerprinting approach [14] in characterising the effects of novel pharmacological agents upon brain function. This method has previously been used to examine the differences in the functional profiles between benzodiazepines and agents acting directly via the GABA receptor [7,14] and those produced by dopamine D 1 and D 2 receptor agonists (Sharkey and McCulloch, unpublished). In the present study, we have shown that this method can be successfully used to investigate the metabolic profile of the novel NMDA receptor antagonist, FRl15427. We observed that the pattern of regional f values observed in rats treated with FR115427 is similar to that produced by dizocilpine, but different from that produced by the competitive NMDA receptor antagonist, CGS 19755. Moreover, in general, FRl15427 appeared some 30-fold less potent than dizocilpine with regard to function-related alterations in glucose utilisation. These data closely parallel radioligand binding and electrophysiological data which show

1.4

1.4

m

~.2

m

--

1.2 1 1

1.0

--

2 1°0

--

°4

e~

-,e ~ °1,i

4

•

3

•

~

0.8-

~

0.6

43

0.8•

•

6

0.6-

S

~.

•

•

O

•

•

•

•

0.2

--go • 0.2

- -

WOQ • •

"

•

0.0

05

•

0.4

•

0.4-

6

|0

0~•• • • fl.~p ° 0

•

0.0

I

I

I

~

I

I

I

I

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

"f" FRl15427

I' 0.0

I

I

I

I

I

I

I

0,2

0.4

0.6

0.8

1.0

1.2

1.4

"f" CGS19755

Fig. 3. Comparison of the profile of regional responsiveness produced by dizocilpine with that of FR11527 and of CGS19755. Each data point represents the f value generated from the dose-glucose use curves for a single brain structure in response to (left) the non-competitive NMDA receptor antagonist, dizocilpine and FR115427, and (right) dizocilpine and the competitive NMDA receptor antagonist CGS19755. The individual f values for the six brain regions depicted in Fig. 1 are numbered as follows: (1) auditory cortex, (2) pontine nuclei, (3) presubiculum, (4) fornix, (5) inferior colliculus, (6) entorhinal cortex.

80

J. Sharkey et al. / Brain Research 735 (1996) 67-82

FR115427 to be a use-dependent, non-competitive NMDA receptor antagonist with approximately 10-30 fold lower affinity for the NMDA ion channel than dizocilpine [10,29]. The metabolic fingerprinting approach provides a technique for evaluating novel pharmacological compounds that is only as good as the reference compounds used to define the receptor of interest. Dizocilpine is the prototypic non-competitive NMDA receptor antagonist whose effects upon the NMDA-associated ion channel have been extensively characterised [3,9,34-36]. Similarly, CGS19755 is a highly selective competitive NMDA receptor antagonist which has been shown to exhibit some three-fold higher affinity than CPP for the NMDA receptor [16,27]. Thus, the in vitro potency and selectivity of these compounds make these useful reference compounds. However, in vivo pharmacokinetic factors such as blood-brain barrier penetration and biological half-live must also be taken into account. While hydrophobic compounds such as dizocilpine rapidly enter and egress from the CNS [32], competitive NMDA receptor antagonists such as CGS19755, which are hydrophilic, exhibit low brain uptake indices [11]. An inability to cross the blood-brain barrier would obviously negate the use of CGS 19755 as a reference compound for metabolic fingerprinting studies. However, despite its low brain uptake index, there is evidence that CGS 19755 does penetrate sufficiently rapidly and in sufficient concentrations to produce biochemical and neuroprotective effects within the 10-55 rain post injection measurement period used in the present study. Significant reductions in cerebellar cyclic GMP levels have been reported by Lehman et al., 20 min following intraperitoneal administration of CGS19755 (4 mg/kg) [16]. These authors have also reported significant anticonvulsant activity within 30 min of a subcutaneous injection of CGS19755. Furthermore, previous studies with CPP, a structural analogue of CGS19755 [27], have shown dosedependent alterations in 1CGU when using the same measurement period and route of administration as in the present study [16]. Similarly, the choice of dose range could have a significant impact upon the usefulness of any drug comparison. Since drugs with similar mechanisms of action may, as in the present study, produce differing behavioural effects, some other (physiological, biochemical or neuropathological) common reference point should be chosen. In the present studies we chose to examine doses of the two reference compounds which were up to that reported to reduce ischaemic damage by 40-50% [1,8,19,30]. The choice of dose range for FR115427 reflected the approximately 10 fold higher dose required for a comparable degree of neuroprotection to that produced by dizocilpine in the rat MCA occlusion model [13]. Metabolic mapping studies, using 2-deoxyglucose autoradiography, have highlighted the broad similarities in the functional consequences of the NMDA receptor blockade by non-competitive antagonists [15,19,20]. Patterns of al-

tered glucose use, remarkably similar to that produced by dizocilpine and FR115427 (in the present study) have also been reported following sub-anaesthetic doses of phencyclidine (PCP) and ketamine; characterised by profound elevations in 1CGU within anatomical components of the limbic system (mammillary bodies, fornix, anterior thalamus, cingulum, entorhinal and subicular complex), olfactory system and parts of the basal ganglia (globus pallidus, entopeduncular nucleus and pars reticulata of the substantia nigra); with concomitant reductions in 1CGU within somatosensory cortex, habenula and inferior colliculus [6,21,24,25,33]. The magnitude of these changes, particularly within allocortex (cingulum, entorhinal cortex and subicular complex) are amongst the largest ever observed following drug administration and have been attributed to an increase in neuronal activity in a futile attempt to overcome the use-dependent receptor blockade produced by these agents [15,19]. Moreover, the magnitude of the metabolic alterations produced by non-competitive NMDA receptor antagonists like dizocilpine and FR115427 contrasts markedly with the relatively modest alterations in 1CGU elicited by competitive NMDA receptor antagonists, AP7 [4], CGP39551 [2], CPP [15] and CGS19755 (present study) whose blockade of the NMDA receptor is not use-dependent. While the patterns of altered 1CGU produced by noncompetitive NMDA receptor antagonists are broadly similar, those produced by competitive NMDA receptor antagonists differ, not only from those of the non-competitive antagonists, but they also appear to differ substantially from each other. For example, high doses of the prototypic competitive NMDA receptor antagonist, AP7 has been shown to produce widespread reductions in 1CGU throughout allo- and neo-cortex [4]. CGP39551, a more potent analogue of AP7 resulted in focal increases in ICGU within piriform and entorhinal cortex, reduced glucose use in the subiculum and mamillary bodies, but failed to significantly alter 1CGU within retrosplenial cortex or hippocampus [2]. Although in the present study CGS19755 produced widespread reductions in neocortical glucose use, with the notable exception of piriform cortex, allocortical areas appeared relatively unaffected. A more widespread pattern of increased glucose use in allocortical areas (piriform cortex, entorhinal cortex, amygdala and molecular layer of the hippocampus) has been reported following administration of CPP while neocortical areas appeared relatively unaffected [ 15]. Differences in the magnitude of the metabolic responses elicited by competitive and non-competitive NMDA receptor antagonists have been attributed to the use-dependent nature of the glutamate receptor blockade produced by agents such as dizocilpine which is intensified by increasing glutamatergic transmission [19]. However, the mechanisms underlying the apparent qualitative differences in the metabolic patterns produced by these agents are less clear. That competitive and non-competitive NMDA receptor

J. Sharkey et a l . / Brain Research 735 (1996) 67-82

antagonists should produce such marked differences in functional responses is surprising given the high degree of register in the anatomical distribution of their respective binding sites [15]. Moreover, blockade of either of these sites ultimately leads to a reduction in glutamatergic transmission. Differences in the metabolic responses elicited by competitive and non-competitive NMDA receptor antagonists have been attributed to their relative affinities for other receptor systems (dopaminergic or (r-receptors) [15,19]. Differences in receptor specificity could also account for the observed differences between competitive NMDA receptor antagonists, particularly at the high doses which are required in vivo. Both FR115427 and dizocilpine exhibit similar binding characteristics, with unusually slow association and dissociation kinetics, and potentiation by the presence of glutamate and glycine in the incubation medium [29]. Both inhibit NMDA-induced depolarisation and spontaneous epileptiform activity in rat cortical slice preparations in a non-competitive manner [10]. Furthermore, evidence of a high correlation between the affinities ( K i values) exhibited by the stereoisomers of FR115427 and dizocilpine, and by the non-competitive NMDA receptor antagonists, phencyclidine, ketamine and N-allylnormetazocine for the sites labelled by [3H]dizocilpine and [3H]FRl15427 provide further support for the view that both FRl15427 and dizocilpine bind to a similar, if not identical, site within the NMDA receptor ion channel [29]. The present data, showing a highly significant correlation between the patterns of altered 1CGU by FR115427 and dizocilpine, is therefore consistent with the in vitro electrophysiological and radioligand binding data. The requirement of 10-30 fold higher doses of FRl15427 to produce alterations in 1CGU of a similar magnitude to that seen in dizocilpine-treated rats, is also consistent with the 10-20 fold lower affinity exhibited by FR115427 in binding studies [10,29]. In the present study, these differences in potency can readily be seen in areas such as the presubiculum, dentate gyms, mammillary bodies and auditory cortex. However, in a few brain areas (most notably the entorhinal and prepyriform cortex) where dizocilpine administration resulted in marked, dose-related increases in 1CGU, FR115427 failed to significantly alter glucose use at any of the doses examined. Since these structures have been implicated in the psychotomimetic actions of NMDA receptor antagonists [26] such findings may have bearing on the usefullness of FR115427 for the treatment of CNS injury. Dizocilpine and phencyclidine have been reported to cause morphological changes in neurones within the retrosplenial cortex at doses which are neuroprotective in animal models of focal cerebral ischaemia [26]. The neuronal swelling and vacuolisation observed in retrosplenial cortex has been attributed to local metabolic hyperactivity resulting from increased neuronal activity in a futile attempt to overcome the use-dependent blockade of glutamatergic

81

transmission by these agents [19,26]. In this context, it is interesting to note that FRl15427 exhibits significantly less use-dependence than dizocilpine [10] and that while approximately 10-fold higher doses of FRl154427 are required to produce a similar degree of neuroprotection as dizocilpine in rat models of focal cerebral ischaemia the dose required to produce vacuolisation is at least 30-fold greater [13]. In conclusion, the results of the present study are consistent with the view that FR115427 is a non-competitive NMDA receptor antagonist which effects brain function in a manner which is broadly similar to that exhibited by the class standard, dizocilpine. However, FR115427 did not produce the marked increases in glucose use observed in retrosplenial, piriform and entorhinal cortices which have been reported following administration of dizocilpine. The significance of these differences in the pattern of glucose use within these allocortical structures remains to be determined.

References [1] Bullock, R. and Fujisawa, H., The role of glutamate antagonists for the treatment of CNS injury, J. Neurotrauma, 9 (1992) s443-s462. [2] Chapman, A.G., Graham, J.L., Swan, J.H. and Meldrum, B.S., Regional glucose utilisation in rat brain following the administration of CGP39551, a competitive NMDA antagonist, J. Cereb. Blood Flow Metab., 11 (Suppl. 2) (1991) s226. [3] Clineschmidt, B.V., Martin, G.E. and Bunting, P.R., Anticonvulsant activity of ( + )-5-methyl- 10,11-dihydro-54H-dibenzo [a,d]cyclohepten-5,10-imine (MK801), a substance with potent anticonvulsant, central sympathomimetic and apparent anxiolytic properties, Drug Dev. Res., 2 (1982) 123-134. [4] Clow, D.W., Lee, S.J. and Hammer, R.P., Competitive (AP-7) and non-competitive (MK-801) NMDA receptor antagonists differentially alter glucose utilization in rat cortex, Synapse, 7 (1991) 260-268. [5] Collingridge, G.L. and Watkins, J.C. In G.L. Collingridge and J.C. Watkins (Eds.), The NMDA Receptor, IRL Press, Oxford, 1994. [6] Crosby, G., Crane, A.M. and Sokoloff, L., Local changes in cerebral glucose utilization during ketamine anaesthesia, Anaesthesiology, 56 (1982) 437-443. [7] Cudennec, A., Duverger, D., Lloyd, K.G., MacKenzie, E.T., McCulloch, J., Motohashi, N., Nishikawa, T. and Scatton, B., Effects of the GABA receptor agonist, progabide, upon local cerebral glucose utilization, Brain Res., 423 (1987) 162-172. [8] Grotta, J.C., Picone, C.M., Ostrow, P.T., Strong, R.A., Earls, R.M., Yao, L.P., Rhoades, H.M. and Dedman, J.R., CGS-19755, a competitive NMDA receptor antagonist, reduces calcium-calmodulin binding and improves outcome after global cerebral ischemia, Ann. Neurol., 27 (1990), 612-619. [9] Halliwell, R.F., Peters, J.A. and Lambert, J.J., The mechanism of action and pharmacological specificity of the anticonvulsant NMDA antagonist, MK801: a voltage clamp study on neuronal cells in culture, Br. J. Pharmacol., 96 (1989) 480-494. [10] Hodgkiss, P., Sherriffs, H.J., Cottrell, D.A., Shirakawa, K., Kelly, J.S., Kuno, A., Ohkubo, M., Butcher, S.P. and Olverman, H.J., Neurochemical and electrophysiological studies on FRl15427, a novel non-competitive NMDA receptor antagonist, Eur. J. Pharmacol., 240 (1993) 219-227. [11] Hogan, M.J., Gjedde, A. and Hakim, A.M., In vivo distribution of

82

[12]

[13]

[14]

[15]

[16]

[17] [18]

[19]

[20]

[21]

[22]

[23]

[24]

J. Sharkey et al. / Brain Research 735 (19961 67-82

CGS19755 within brain in a model of focal cerebral ischemia, J. Neurochem., 58 (1992) 186-191. Jane, D.E., Olverman, H.J. and Watkins, J.C., Agonists and competitive: structure-activity and molecular modelling studies. In G.L. Collingridge and J.C. Watkins (Eds.), The NMDA Receptor, IRL Press, Oxford, 1984. Katsuta, K., Nakanishi. H., Shirakawa, K., Yoshida, K., Takagi, K., Tamura, A., The neuroprotective effect of the novel noncompetitive NMDA antagonist, FR115427 in focal cerebral ischaemia in rats, J. Cereb, Blood Flow Memb., 15 (1995) 345-348. Kelly, P.A.T., Ford, I. and McCulloch, J., The effect of diazepam upon local cerebral glucose use in the conscious rat, Neuroscience, 19 (1986) 257-265. Kurumaji, A., Nehls, D.G., Park, C.K. and McCulloch, J., Effects of the NMDA antagonists, MK801 and CPP upon local cerebral glucose use, Brain Res., 496 (1989) 268-284. Lehmann, J., Hutchison, A.J., McPherson, S.E., Mondadori, C., Schmutz, M., Sinton, C.M., Tsai, C., Murphy, D.E., Steel, D.J., Williams, M., Cheney, D.L. and Wood, P.L.. CGS19755, a selective and competitive N-metbyl-D-aspartate-type excitatory amino acid receptor antagonist, J. Pharmacol. Exp. Ther., 246 (1988) 65-75. Lodge, D. and Johnson, K.M., Non-competitive excitatory amino acid antagonists, Trends Pharmacol. Sci., 11 (1990) 81-86. MacDonald, J.F. and Nowak, L.M., Mechanism of blockade of excitatory amino acid receptor channels, Trends Pharmacol. Sei., I 1 (1990) 167 172. McCulloch. J., Bullock, R. and Teasdale. G.M., Excitatory amino acid antagonists: opportunities for the treatment of ischaemic brain damage in man. In B.S. Meldrum (Ed.), Excitatory Amino Acid Antagonists, Blackwell, 1991. McCulloch. J. and Iversen, L.L., Autoradiographic assessment of the effects of N-methyl-D-asparate receptor antagonists in vivo. Neuiv~hem. Res., 16 (1991 ) 951-963. Meibach, R.C., Glick, S.D.. Cox. R. and Maayani. S.. Localisation of phencyclidine-induced changes in brain energy metabolism, Nature, 282 (1979) 625-626. Nakanishi, H., Katsuta, K., Koide, T., Shirakawa, K. and Yoshida, K., Protective effect of FR115427 against ischemic hippocampal damage in gerbils, Jpn. J. Pharmacol., 64 (1994) 189-193. Nehls, D.G., Kurumaji, A., Park, C.K., MacCormack, A.G. and McCulloch, J., The effects of N-methyl-D-aspartate receptor blockade upon the relationship between local cerebral blood flow and glucose utilisation, Brain Res., 511 (1990) 271-279. Nelson, S.R., Howard, R.B., Cross, R.S. and Samson, R.,

[25]

[26]

[27]

[28] [29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

Ketamine-induced changes in regional glucose utilisation in the rat brain, Anaesthesiology, 52 (1980) 330 334. Oguchi, K,, Arakawa. K., Nelson, S.R. and Samson, F., The influence of droperidol, diazepam and physostigmine on ketamine-induced behavior and regional glucose utilisation in the rat, Anaesthesiology, 57 (1982) 353-358. Olney, J.W., Labruyere, J., Wang, G., Wozniak, D.F., Price, M. and Sesma, M.A., NMDA antagonist neurotoxicity: mechanism and prevention, Science, 254 ( 1991 ) 1515-1518. Olverman, H.J. and Watkins, J.C., NMDA agonists and competitive. In J.C. Watkins and G.L. Collingridge (Eds.), The NMDA Receptor, IRL Press, Oxford, 1989. Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, Academic Press, London, 1986. Sherriffs, H.J., Shirakawa, K., Kelly, J.S., Olverman, H.J., Kuno, A., Okubo, M. and Butcher, S.P., Characterisation of the binding of [3H]_FR115427, a novel, non-competitive NMDA receptor antagonist, to rat brain membranes, Eur. J. Pharmacol., 247 (1993) 319-324. Simon, R. and Shiraishi, K., N-Methyl-D-aspartate antagonist reduces stroke size and regional glucose metabolism, Ann. Neurol., 27 (1990) 606 611. Soknloff, L., Reivich, M., Kennedy, C., Des Rosier, M.H., Patlak, C.S., Pettigrew, K.D., Sakurada, O. and Shinohara. M., The [14C]deoxyglucose method for the measurement of local cerebral glucose utilisation: theory, procedure, and normal values in the conscious and anaesthetized albino rat, J. Neuroehem., 28 (1977) 897-916. Swan, J.H. and Meldrum. B.S., Protection by NMDA antagonist against cell loss following transient ischaemia, J. Cereb. Blood Flow Metab., 10 (19901 343-351. Weissman, A.D.. Dam, M. and London. E.D., Alterations in local cerebral glucose utilisation induced by phencyclidine, Bruin Res., 435 (1987) 29-4(I. Wong, E.H.F. and Kemp, J.A., Sites for antagonism on the Nmethyl-~aspartate receptor channel complex, Annu. Ret:. Pharmacol. Toxicol., 31 (1991) 401 425. Wong, E.H.F., Kemp, J.A., Priestly. T., Knight, A.R.. Woodruff, G.N. and lversen, L.L., The anticonvulsant MKS(/I is a potent N-methyl-D-aspartate antagonist, Proc. Natl. Acad. Sci. USA, 83 (19861 7104-7108. Wong, E.H.F., Knight, A.R. and Woodruff, G.N., [3H]MK801 labels a site on the N-methyl-D-aspartate receptor channel complex in rat brain membranes, .I. Neurochem., 50 (1988) 274-281.