Compartmental analysis of [11C]flumazenil kinetics for the estimation of ligand transport rate and receptor distribution using positron emission tomography.

The in vivo kinetic behavior of [11C]flumazenil ([11C]FMZ), a non-subtype-specific central benzodiazepine antagonist, is characterized using compartmental analysis with the aim of producing an optimized data acquisition protocol and tracer kinetic model configuration for the assessment of [11C]FMZ binding to benzodiazepine receptors (BZRs) in human brain. The approach presented is simple, requiring only a single radioligand injection. Dynamic positron emission tomography data were acquired on 18 normal volunteers using a 60- to 90-min sequence of scans and were analyzed with model configurations that included a three-compartment, four-parameter model, a three-compartment, three-parameter model, with a fixed value for free plus nonspecific binding; and a two-compartment, two-parameter model. Statistical analysis indicated that a four-parameter model did not yield significantly better fits than a three-parameter model. Goodness of fit was improved for three- versus two-parameter configurations in regions with low receptor density, but not in regions with moderate to high receptor density. Thus, a two-compartment, two-parameter configuration was found to adequately describe the kinetic behavior of [11C]FMZ in human brain, with stable estimates of the model parameters obtainable from as little as 20-30 min of data. Pixel-by-pixel analysis yields functional images of transport rate (K1) and ligand distribution volume (DV"), and thus provides independent estimates of ligand delivery and BZR binding.     

Autoradiography with saturation experiments of 11C-Ro 15-1788 binding to human brain sections

In vitro autoradiography of a 11C-labelled ligand, Ro 15-1788, was used in saturation experiments in order to quantify benzodiazepine receptor binding in whole human brain hemisphere cryo-sections. A special incubation chamber was developed with the aim of performing standardized quantitative studies with short-lived positron emitting isotopes. 11C-labelled ligand binding was studied in temporal cortex, cerebellum, white manner and pons incubated in vitro. White manner, lacking benzodiazepine receptor binding, was used as an estimated of nonspecific binding, for calculation of Saturability of binding was demonstrated in the neocortical and cerebellar regions. Radioactivity counting of incubated adjacent tissue sections was done as a control. Computerized densitometry of the autoradiograms gave similar results as the tissue counting. A comparison with in vivo saturation experiments using the same 11C-labelled ligand and positron emission tomography in healthy human subjects also gave binding characteristics of the same magnitude. The 11C-autoradiograms showed a good spatial resolution (about 180 microns). 11C-autoradiography with suitable radioligands should be a valuable technique of screening the distribution and characteristics of neuroreceptors in the human brain. This quantitative method will provide the PET investigator with preliminary binding data, and may thus supply valuable information concerning positioning in PET and concerning significant regions of interest.     

In vivo labeling of central benzodiazepine receptors with the partial inverse agonist [3H]Ro 15-4513

Ro 15-4513 is an imidazobenzodiazepine and a partial inverse agonist at the central benzodiazepine receptors (BZDr). It has been shown to antagonize behavioral and biochemical effects of ethanol. In vivo binding of [3H]Ro 15-4513 was evaluated in mouse brain. After intravenous injection [3H]Ro 15-4513 was readily taken up by the brain and distributed to brain areas enriched in benzodiazepine receptors. Binding was specific for central BZDr, saturable and reversible. A high degree of specific binding, relative to non-specific binding, was achieved. Analysis of dissociation kinetics revealed that [3H]Ro 15-4513 was retained significantly longer in hippocampus compared to other brain regions. In view of the known distribution of benzodiazepine receptor subtypes, this suggests that, in vivo, [3H]Ro 15-4513 has a higher affinity for benzodiazepine receptors type II and may explain quantitative differences in the regional distribution of this ligand compared to the antagonist [3H]Ro 15-1788. We conclude from these studies that Ro 15-4513 is a suitable ligand for in vivo studies of benzodiazepine receptors. Labeled with a positron-emitting isotope, it could be used with positron emission tomography to study BZDr in man under a variety of conditions.     

Benzodiazepine receptor binding of nonbenzodiazepines in vivo: Alpidem, zolpidem and zopiclone

Several classes of nonbenzodiazepine compounds, including imidazopyridines such as alpidem and zolpidem and cyclopyrrolones, e.g., zopiclone, have effects similar to benzodiazepines and may act at the benzodiazepine receptor in brain. We characterized the binding of these compounds to the benzodiazepine site in three brain regions using specific uptake of the high-affinity ligand [3H]Ro15-1788 (flumazenil). For alpidem, benzodiazepine binding was decreased in cortex and hippocampus with increasing drug dose. For zolpidem, receptor binding was reduced in cortex without a dose-response effect and no effect was observed on cerebellar binding. Zopiclone did not alter binding except for a decrease in binding at the lowest dose evaluated and an increase in binding above control at the highest dose. These data corroborate prior studies indicating that the imidazopyridines appear to act at the benzodiazepine receptor, but do not support receptor subtype selectivity of zolpidem. The limited effect of zopiclone except for increased binding at high doses is also consistent with prior studies suggesting that zopiclone acts at a site distinct from the benzodiazepine receptor.     

Imaging of [C]-labelled RO 15-1788 binding to benzodiazepine receptors in the human brain by positron emission tomography

The benzodiazepine antagonist Ro 15-1788 was labelled with [¹¹C] and examined for possible use as ligand for PET scan studies on benzodiazepine receptors in the brain of cynomolgus monkeys and human subjects. [¹¹C] Ro 15-1788 allowed the in vivo visualization of benzodiazepine receptor binding in cerebral and cerebellar cortical areas as well as in basal brain nuclei in PET scan images. [¹¹C] Ro 15-1788 exhibited a high ratio of specific benzodiazepine receptor binding (cerebral cortex) to non-specific binding (pons) and the kinetics of binding should be satisfactory for quantitative clinical PET scan studies using [¹¹C]. The in vivo binding of [¹¹C] Ro 15-1788 in the cerebral cortex of cynomolgus monkeys and healthy human subjects was reduced by approximately 90% within 10 min after the intravenous injection of a high dose of unlabelled Ro 15-1788 (0.5 mg/kg i.v.). Different areas of the healthy human brain showed an approximately 10-fold variation in maximal [¹¹C] Ro 15-1788 binding that corresponded to the previously known distribution of benzodiazepine receptors in these regions. The highest degree of binding was obtained in the medial occipital cerebral cortex followed by frontal cortex, cerebellum, thalamus, striatum and pons. Two psychiatric patients with anxiety syndromes who had been treated for a long time with high doses of benzodiazepines had roughly the same degree of maximal [¹¹C] Ro 15-1788 binding in brain regions as the healthy subjects but the rate of decline of [¹¹C] Ro 15-1788 in the brain was higher. This indicates that there is measurable competition between [¹¹C] Ro 15-1788 binding and clinical benzodiazepine concentrations in the body fluids of psychiatric patients. The results demonstrate that [¹¹C] Ro 15-1788 should be a valuable tool for quantitative analyses of benzodiazepine receptor characteristics and receptor occupancy in the brain of patients with neuropsychiatric disorders.     

Adrenalectomy prevents the stress-induced decrease in in vitro [H]Ro15-1788 binding to GABAA benzodiazepine receptors in the mouse

The effect of a single or repeated swim stress on in vivo benzodiazepine receptor binding to various brain regions in adrenalectomized and sham-operated (control) mice was assessed using the benzodiazepine receptor antagonist, [³H]Ro15-1788. In sham-operated mice the binding of [³H]Ro15-1788 to benzodiazepine receptors was reduced in the hippocampus and hypothalamus (single or repeated stress) and cerebral cortex (repeated swim stress) compared to non-stressed mice. In contrast, no alterations in [³H]Ro15-1788 binding were observed in any brain region in adrenalectomized mice after either single or repeated swim stress. These data suggest that an intact hypothalamic-pituitary-adrenal axis is required for the stress-induced decrease in benzodiazepine receptor occupancy measured using the in vivo binding method.     

Biological substrates of anxiety: benzodiazepine receptors and endogenous ligands

Benzodiazepines have been shown to produce most, if not all, of their pharmacological effects by directly interacting with specific recognition or receptor sites within the CNS. The presence of benzodiazepine receptors has prompted many studies as to their possible physiological significance, including attempts at isolating an endogenous ligand. To data a number of substances including the purines inosine and hypoxanthine, nicotinamide, beta-carbolines, and an unidentified peptide factor have been isolated and postulated as being endogenous ligands. A number of these compounds have also been shown to either mimic or antagonize the behavioral effects of benzodiazepines, although it is still unclear whether these occur under physiological conditions. More recent biochemical studies have established a functional (and perhaps structural) relationship between te benzodiazepine receptor and the receptor for gamma amino butyric acid (GABA), the major inhibitory neurotransmitter in brain. It now appears that the benzodiazepine receptor actually exists as a "supramolecular complex" consisting of a GABA receptor and an associated chloride channel. A number of anxiolytic drugs including the barbiturates and pyrazolopyridines appear to act through these associated "regulatory" sites rather than directly on the benzodiazepine recognition site.     

PET imaging of human gliomas with ligands for the peripheral benzodiazepine binding site

Human gliomas were imaged in vivo using ligands for the peripheral-type benzodiazepine binding site (or omega 3 binding site) and positron emission tomography (PET). Although gliomas have a high density of the peripheral-type benzodiazepine binding site, PET scans with a selective ligand for this site, [11C] Ro5-4864, failed to demonstrate higher radioactivity levels in human gliomas than in brain. In vitro studies of surgically removed specimens of human glioma demonstrated little binding of Ro5-4864 but high levels of binding of another selective ligand, PK 11195. Scans with [11C]PK 11195 demonstrated increased radioactivity in glioma compared to brain in 8 of 10 patients. Radioactivity in tumor and the ratios of radioactivity in tumor to that in remote gray and in white matter correlated significantly with the specific activity of [11C]PK 11195, suggesting that accumulation represents saturable high-affinity binding. We conclude that the PK 11195 manifests greater binding than Ro5-4864 to the peripheral-type benzodiazepine binding site on human gliomas and that human gliomas can be successfully imaged using [11C]PK 11195 and PET.     

Relationships between benzodiazepine receptors, impairment of GABAergic transmission and convulsant activity of beta-CCM: a PET study in the baboon Papio papio.

Central type benzodiazepine receptors were studied in vivo by positron emission tomography in brain areas of 2 different groups of the baboon Papio papio: non-photosensitive (group 1) and those with an allylglycine-induced decrease in GABA-mediated inhibition (group 2). Further, a naturally photosensitive Papio papio (+3 level of photosensitive response) was compared to both groups. Regional brain binding of the specific benzodiazepine receptor ligand, [11C]Ro 15-1788, was not significantly different between groups 1 and 2. In addition, the data from the naturally photosensitive Papio papio did not seem to differ markedly from groups 1 and 2 either. Pharmacological effects of increasing doses of beta-CCM (0.05-3 mg/kg i.v.) and regional benzodiazepine receptor occupancy by the drug were simultaneously studied using electroencephalographic activity recording and positron emission tomography. A positive correlation was observed between the degree of photosensitivity of the baboon and sensitivity to the action of beta-CCM, with increasing convulsant efficacy of beta-CCM in going from group 1 to the naturally photosensitive baboon, then to group 2. Dose-related displacement curves of [11C]Ro 15-1788 binding by beta-CCM revealed that reduction in brain GABA concentration did not modify the inhibitory potency of beta-CCM on [11C]Ro 15-1788 binding in cerebral cortex. This suggests a lack of detectable in vivo allosteric effects of GABA on beta-CCM binding during beta-CCM-induced seizures. Thus, a given dose of beta-CCM displayed increasing pharmacological potency in going from baboons with the lowest photosensitivity to those with the highest, whereas benzodiazepine receptor occupancy by beta-CCM was similar in the cerebral cortex of the different baboons. Conversely, a given level of convulsant activity of beta-CCM was related to a different benzodiazepine receptor occupancy by the drug, depending on the photosensitivity of Papio papio. A given dose of a drug may, thus, have a different pharmacological potency when occupying the same number of receptors, depending on the physiopathological state of the subject.     

The positron emission tomography ligand DAA1106 binds with high affinity to activated microglia in human neurological disorders

Chronic microglial activation is an important component of many neurological disorders, and imaging activated microglia in vivo will enable the detection and improved treatment of neuroinflammation. 1-(2-chlorphenyl)-N-methyl-N-(1-methylpropyl)-3-isoquinoline-carbox-amide (PK11195), a peripheral benzodiazepine receptor ligand, has been used to image neuroinflammation, but the extent to which PK11195 binding distinguishes activated microglia and reactive astrocytes is unclear. Moreover, PK11195 may lack sufficient sensitivity for detecting mild neuroinflammation. We hypothesized that N-(2,5-dimethoxybenzyl)-N-(4-fluoro-2-phenoxyphenyl) acetamide (DAA1106), a new ligand that binds specifically to peripheral benzodiazepine receptor, binds to activated microglia in human neurological diseases with higher affinity than does PK11195. We therefore compared the pharmacological binding properties of [3H](R)-PK11195 and [3H]DAA1106 in postmortem tissues from patients with cerebral infarcts, amyotrophic lateral sclerosis, Alzheimer disease, frontotemporal dementia, and multiple sclerosis (n=10 each). In all diseases, [3H]DAA1106 showed a higher binding affinity as reflected by lower dissociation constant (KD) values than that of [3H](R)-PK11195. Moreover, specific binding of both ligands correlated with the presence of activated microglia identified by immunohistochemistry in situ. We conclude that 1) ligands that bind peripheral benzodiazepine receptor mainly label activated microglia in human neurological disorders and that 2) DAA1106 may possess binding characteristics superior to those of PK11195, which may be beneficial for in vivo positron emission tomography imaging.     

The high affinity peripheral benzodiazepine receptor ligand DAA1106 binds to activated and infected brain macrophages in areas of synaptic degeneration: implications for PET imaging of neuroinflammation in lentiviral encephalitis

HIV encephalitis (HIVE) is characterized by neurodegeneration mediated by toxins derived from infected and activated brain macrophages. Since the peripheral benzodiazepine receptor (PBR) is abundant on brain macrophages, we hypothesized that [(3)H]DAA1106, a new PBR ligand, can label infected and activated brain macrophages in HIVE. Using cell culture and postmortem brain tissues from HIVE and a macaque model of HIVE, we show that [(3)H]DAA1106 binds with high affinity to activated and infected macrophages in regions of synaptic damage. Further, binding affinity reflected by lower K(D) (dissociation constant) values and the B(max) (total number of binding sites) to K(D) ratios reflective of ligand-binding potential was significantly higher with [(3)H]DAA1106 compared to the extensively characterized PBR ligand [(3)H](R)-PK11195. These data suggest that DAA1106 binds with high affinity to activated and infected brain macrophages and possesses binding characteristics beneficial for in vivo use in the detection and clinical monitoring of HIVE using positron emission tomography.     

Reversible dysfunction of receptors in traumatic brain injury?

In many brain disorders reduced binding of central benzodiazepine receptor ligands indicates irreversible neuronal damage. The data presented by Koizumi et al (2010) demonstrate that this is not the case in traumatic brain injury suggesting different pathogenetic mechanisms leading to tissue damage. The proof for this hypothesis requires further studies that should also consider thresholds of ligand binding as indicators of irreversible damage.     

Differentiation of radioligand delivery and binding in the brain: validation of a two-compartment model for [11C]flumazenil

We recently developed a two-compartment, two-parameter tracer kinetic model to estimate the in vivo ligand transport rate (K1) and distribution volume (DV) for the benzodiazepine antagonist [11C]flumazenil (FMZ) as measured by positron emission tomography (PET). The aim of the present study was to validate that this simplified model provides a stable measure of regional benzodiazepine receptor availability even when ligand delivery is altered. Six young normal volunteers underwent two PET studies subsequent to intravenous injections of [11C]FMZ. Each FMZ study was immediately preceded by measurements of CBF following injection of [15O]water. One set of scans (water/FMZ) was acquired under resting conditions and the other set during audiovisual stimulation. Six additional volunteers underwent two FMZ studies under identical resting conditions. Parametric images were analyzed and a comparison of test-retest studies in the stimulation group revealed a significant increase of CBF and K1 of FMZ in the occipital cortex evoked by visual activation, whereas no regional changes were noted for the DV of FMZ. No significant changes were noted for either K1 or DV of FMZ when comparing studies in the rest-rest setting. The results indicate that the use of a simple two-compartment model for the tracer kinetic analysis of [11C]FMZ makes it possible to separate high-affinity binding from altered radio-ligand delivery to the human brain.     

Opiate antagonist receptor binding in vivo: evidence for a new receptor binding model

The in vivo accumulation and retention of the opiate antagonist tracers [3H]diprenorphine and [3H]naloxone at cerebral opiate receptor sites in rats exceed that expected from their known in vitro receptor affinities. The [3H]diprenorphine serum and brain levels can be stimulated with a pharmacokinetic model that contains the receptors in a micro-compartment. The receptor micro-compartment consists of a population of binding sites next to a diffusion boundary which restricts ligand diffusion away from the receptor. Such an arrangement introduces a delay in the binding equilibrium of potent antagonists with the receptor sites and an increase in the apparent in vivo receptor affinity at subsaturating doses of the ligand; at saturating ligand concentrations these functions of the receptor micro-compartment are abolished. A physiological interpretation of the receptor micro-compartment could be the location of clustered opiate receptor sites on the exterior cell surface next to the synaptic cleft as the diffusion boundary. This kinetic approach involving a combination of pharmacokinetics and drug-receptor interactions permits the quantitative analysis of receptor site availability in the intact animal. Our results support the hypothesis that only one receptor population affects the in vivo disposition of the antagonist tracers, while they do not exclude the presence of low affinity binding sites that have been observed with the use of [3H]naloxone in vitro. Moreover, the binding site population observed in vivo may be responsible for mediating opiate agonist analgesia.     

Novel peripheral benzodiazepine receptor ligand [11C]DAA1106 for PET: an imaging tool for glial cells in the brain

Peripheral benzodiazepine receptor (PBR) is expressed in most organs and its expression is reported to be increased in activated microglia in the brain. [(11)C]PK11195 has been widely used for the in vivo imaging of PBRs, but its signal in the brain was not high enough for stable quantitative analysis. We synthesized a novel positron emission tomography (PET) ligand, [(11)C]DAA1106, for PBR and investigated its in vivo properties in rat and monkey brain. High uptake of [(11)C]DAA1106 was observed in the olfactory bulb and choroid plexus area, followed by the pons/medulla and cerebellum by in vivo autoradiography of rat brain, correlating with the binding in vitro. [(11)C]DAA1106 binding was increased in the dorsal hippocampus with neural destruction, suggesting glial reaction. [(11)C]DAA1106 binding was both inhibited and displaced by 1.0 mg/kg of DAA1106 and 5 mg/kg of PK11195 by 80% and 70%, respectively. Specific binding was estimated as 80% of total binding. [(11)C]DAA1106 binding was four times higher compared to the binding of [(11)C]PK11195 in the monkey occipital cortex. These results indicated that [(11)C]DAA1106 might be a good ligand for in vivo imaging of PBR.     

Chronic brief restraint decreases in vivo binding of benzodiazepine receptor ligand to mouse brain

This study examines the effects of chronic brief restraint on in vivo benzodiazepine (BZD) receptor binding in mouse brain. Three groups of mice were used. Mice in group 1 were neither restrained nor injected (ACUTE control). Mice in group 2 were restrained for 5–6 s by grabbing the back skin and holding the subject upside-down at a 45° angle as if to be injected (CHRONIC SHAM control) for 7 d. Mice in group 3 (CHRONIC SALINE) received daily single intraperitoneal (ip) injections of saline (5 mL/kg) for 7 d. On d 8 BZD receptors were labeled in vivo by administration of 3 μCi [³H]flumazenil (ip). The levels of ligand bound in vivo to cerebral cortex (CX), cerebellum (CB), brain stem (BS), striatum (ST), hippocampus (HP), and hypothalamus (HY) were determined. Results indicated that the level of binding was significantly (p<0.01) lower by 30–50% (depending on the brain region) in saline-injected or sham control groups compared to acute control animals. Furthermore, the values for sham control were similar to the saline-treated group. Our data suggest that exposure to chronic mild restraint produces a decrease in in vivo binding of [³H]flumazenil in mouse brain and supports the hypothesis that chronic mild stress produces a decrease in BZD receptor binding sites.     

Profound stress-induced alterations in flurazepam's antiseizure efficacy can be attenuated

A new procedure is described for assessing the sensitivity of the benzodiazepine receptor in intact animals. This procedure measures the ability of benzodiazepines to antagonize electrically-induced seizures precipitated by incremental increases in voltage. This functional measure detected stress-induced alterations in benzodiazepine receptor sensitivity, an alteration shown previously with in vitro and in vivo receptor binding techniques. In contrast, mouse rotorod performance, which has been proposed as a behavioral measure of benzodiazepine receptor sensitivity did not show stress-induced alterations. The mechanism of these stress-induced alterations was explored using this incremental electroconvulsive shock procedure (IECS). A ‘GABA-negative’ benzodiazepine mimicked the effects of swim stress, whereas treatment of animals prior to swim stress with Ro 15–1788, a ‘GABA-neutral’ benzodiazepine, attenuated these stress-induced alterations. These data suggest both that these stress effects may be mediated by an endogenous ligand, and that Ro 15–1788 may have a novel indication as a prophylactic intervention for individuals at risk for exposure to severe and unusual stress.     

Involvement of benzodiazepine receptors in neuroinflammatory and neurodegenerative diseases: evidence from activated microglial cells in vitro

Increased binding of a ligand for the peripheral benzodiazepine binding receptor is currently used in PET studies as an in vivo measurement of inflammation in diseases like multiple sclerosis and Alzheimer's disease. Although peripheral-type benzodiazepin receptors (PBRs) are abundant in many cell types and expressed in the CNS physiologically only at low levels, previous reports suggest that after experimental lesions in animal models and in human neurodegenerative/-inflammatory diseases upregulated PBR expression with increased binding of its ligand PK11195 is confined mainly to activated microglia in vivo/in situ. Because the functional role of the PBR is unknown, we confirm by immunohistochemistry and PCR (I) that this receptor is expressed on microglia in vitro and (II) that benzodiazepines modulate proliferation of microglial cells and the release of the inflammatory molecules nitric oxide (NO) and tumor necrosis factor-alpha (TNF-alpha) in cell culture supernatants of primary rat microglia. Compared to lipopolysaccharide-activated controls the release of NO was markedly decreased in cultures treated with benzodiazepines (clonazepam, midazolam, diazepam) and the PBR ligand PK11195. Moreover, release of TNF-alpha and proliferation was significantly inhibited in the benzodiazepine-treated groups. These findings link the in vivo data of elevated PBR levels in neurodegenerative/-inflammatory diseases to a functional role and opens up possible therapeutic intervention targeting the PBR in microglia.     

35S]-t-butylbicyclophosphorothionate binding sites are constituents of the gamma-aminobutyric acid benzodiazepine receptor complex

t- Butylbicyclophosphorothionate ( TBPS ), a derivative of potent GABA antagonistic cage convulsants, has recently been introduced ( Squires , R. F., J.E. Casida , M. Richardson, and E. Saederup (1983) Mol. Pharmacol. 13:326-336) as ligand for a GABA-A receptor-linked drug receptor. Using conventionally prepared washed membrane fractions from rat cerebral cortex, we have confirmed that in the presence of 200 mM NaBr [35S] TBPS binds to a high affinity population of binding sites (Kd 26 +/- 5 nM) and that muscimol inhibits [35S] TBPS binding (IC50 0.32 microM) allosterically. In 200 mM NaCl the apparent affinity of [35S] TBPS binding sites is lower (Kd 60 +/- 5 nM), and muscimol has biphasic effects with stimulation at low concentrations of muscimol (EC50 0.023 microM) followed by inhibition at high concentrations (IC50 0.72 microM). Both base line [35S] TBPS binding (in 200 mM NaCl) and muscimol inhibition of [35S] TBPS binding (in 200 mM NaBr) are bidirectionally modulated by the occupancy of benzodiazepine receptors with its ligands. Benzodiazepine receptor agonists, regardless of their structure, enhance and inverse benzodiazepine receptor agonists inhibit base line [35S] TBPS binding and muscimol inhibition of [35S] TBPS binding. Fourteen ligands for benzodiazepine receptors display a similar in vitro profile as benzodiazepine receptor agonists or inverse benzodiazepine receptor agonists on [35S] TBPS binding as their anti- or proconvulsive effects in vivo suggest (Jensen, L. H., E. N. Petersen, and C. Braestrup (1983) Life Sci. 33: 393-399). That [35S] TBPS binding sites are constituents of a GABA benzodiazepine receptor complex is also suggested by a number of membrane pretreatments.(ABSTRACT TRUNCATED AT 250 WORDS)     

Age-related differences of cholecystokinin receptor binding in the rat brain.

1. Cholecystokinin and benzodiazepine receptor binding was evaluated in 2-, 9- and 18-month old rats in the brain regions where cholecystokinin octapeptide and gamma-aminobutyric acid are known to coexist in common nerve cells (frontal cortex, hippocampus). 2. There was a regionally selective alteration of hippocampal 3H-CCK-8 binding in the oldest age group, if compared to both young and adult animals. Non-linear regression analysis of binding data revealed significantly lower apparent number of binding sites (Bmax), and twofold (but not statistically significantly) higher binding affinity for the radiolabelled ligand. No differences between any age groups in 3H-flunitrazepam binding to benzodiazepine receptors were found. 3. The results suggest that changes in cholecystokinin receptor characteristics might contribute to the behavioural impairments in aged rats.