- Top of page
- Materials and methods
By using a combination of an original β+-sensitive␣intracerebral probe and microdialysis, the effect of increased endogenous serotonin on specific binding of 18F-MPPF [4-(2′-methoxyphenyl)-1-[2′-[N-(2′′-pyridinyl)-p-fluorobenzamido]ethyl]piperazine] to the serotonin-1A (5-HT1A) receptors was investigated in the hippocampus of the anaesthetized rat. Our β-sensitive probe prototype was sensitive enough to obtain specific 18F-MPPF time–activity curves in the rodent (hippocampus/cerebellum ratio ≈ 2). The serotonin neuronal release was pharmacologically enhanced using fenfluramine at three different doses (1, 2 and 10 mg/kg intravenous) multiplying by 2–15 the extracellular serotonin in the hippocampus. These extracellular variations of extracellular serotonin resulted in dose-ranging decreases in 18F-MPPF-specific binding in the same rat. Our results showed for the first time that 18F-MPPF binding could be modulated by modifications of extracellular serotonin in the rat hippocampus. These results were confirmed by the enhancement of extracellular radioactivity collected in dialysates after the displacement of 18F-MPPF by fenfluramine. After modelization, 18F-MPPF binding could constitute an interesting radiotracer for positron emission tomography in evaluating the serotonin endogenous levels in limbic areas of the human brain.
Positron emission tomography (PET) has the unique ability to quantitatively monitor physiological changes in the living brain. Several radioligands have been developed for the imaging and quantification of 5-HT1A receptors with PET, and several of these have been tested in humans. In particular, [carbonyl-11C] (N-2-(4-(2-methoxyphenyl-1-piperazinyl)ethyl)-N-(2-pyridil)-cyclohexanecarboxamide␣([carbony-11C]WAY 100635) has been reported to bind to the 5-HT1A receptors [inhibition constant (Ki) = 0.8 nmol/L, Zhuang et al. 1994].
Recently, the selective 5-HT1A antagonist 4-2′-(methoxyphenyl)-1-[2′-(N-2′′-pyridinyl)-p-fluorobenzamido]ethylpiperazine (MPPF) has successfully been labelled with 18F-fluorine, resulting in the 18F-fluoro analogue 18F-MPPF (Shiue et al. 1997). Animal experiments have shown a regional distribution of this radioligand that concurs well with known 5-HT1A receptor densities (Hamon et al. 1990), with a high uptake in the hippocampus and a low uptake in the receptor-poor cerebellum (Shiue et al. 1997; Le Bars et al. 1998; Ginovart et al. 2000; Plenevaux et al. 2000b).
It is known that 18F-MPPF has a relatively low affinity for the 5-HT1A receptor (Ki = 3.3 nmol/L) (Zhuang et al. 1994). Moreover, the relative affinity of 5-HT for the 5-HT1A receptors is Ki= 4.17 nm (Van Wijngaarden et al. 1990), and 18F-MPPF may therefore be more suitable for the detection of changes in endogenous 5-HT. While the affinity characteristics of 18F-MPPF are well-known, to our knowledge no data are available in the literature documenting the displacement of this radioligand by endogenous 5-HT. Therefore, the aim of our study was to demonstrate that 18F-MPPF could be displaced by endogenous 5-HT in the rat hippocampus.
We adopted an original approach using the prototype of a new detector dedicated to the measurement of the kinetics of PET radioligands in the rat brain (Pain et al. 2000; Zimmer et al. 2002). This consists of a β-sensitive intracerebral probe (SIC, French acronym for: ‘sonde intracérébrale’) stereotaxically implanted in the hippocampus and allowing local counting of radioactivity. This approach was coupled with microdialysis which allows the measurement of the extracellular 5-HT. To assess the validity of the displacement of 18F-MPPF by endogenous 5-HT, the following investigations were conducted: (i) we verified the ability of SIC to measure the 18F-MPPF specific binding; (ii) we studied the relationship between changes in 5-HT concentration (measured by microdialysis) and changes in binding parameters of 18F-MPPF (measured with SIC); (iii) we validated our results by the measurement of the extracellular radioactivity release (measured by microdialysis). The main prospect of this study is the measurement of changes in 5-HT levels in the human brain using PET.
- Top of page
- Materials and methods
Recently, several radiotracers have been developed for the imaging and quantification of 5-HT1A receptors with PET (Passchier and van Waarde 2001). The selective 5-HT1A antagonist MPPF has successfully been labelled with fluorine-18, resulting in fluoro-MPPF (Shiue et al. 1997; Le Bars et al. 1998). 18F-MPPF is a well-known 5-HT1A radioligand for 5-HT1A imaging. Pharmacological experiments have shown that MPPF is an antagonist of the pre and postsynaptic 5-HT1A receptors (Zhuang et al. 1994; Thielen et al. 1996) and that MPPF presents a high selectivity for this receptor (Kung et al. 1996). Radiopharmacological experiments have demonstrated that the fluorine-labelled MPPF presents an important initial uptake in the brain and that there are few radioactive metabolites in the brain (Plenevaux et al. 2000b). 5-HT1A receptors are concentrated mainly in limbic areas, notably in the hippocampus (Kia et al. 1996). The autoradiographies of 18F-MPPF are consistent with the known distribution of 5-HT1A receptors (Palacios et al. 1990) with a high labelling in the hippocampus and low labelling in the cerebellum, which is poor in 5-HT1A receptors (Plenevaux et al. 2000a,b). Moreover, it is known that 18F-MPPF has a relatively low affinity for the 5-HT1A receptor (Ki = 3.3 nm, Zhuang et al. 1994) and that the relative affinity of 5-HT for the 5-HT1A receptors is comparable (Ki = 4.17 nm, Van Wijngaarden et al. 1990). Our hypothesis was therefore that 18F-MPPF-specific binding may be influenced by the changes in cerebral 5-HT concentrations.
With a view to demonstrating our hypothesis, we used a prototype of a new detector dedicated to the measurement of kinetic radiotracer in a small animal brain area (‘SIC’). This was done on the basis of a small β-range-sensitive scintillating probe coupled to an ultra-compact and low-noise photomultiplier tube (Pain et al. 2000). The detector takes advantage of the limited range of β particles within biological tissues to determine a limited ‘detectable thickness’ surrounding the probe. In vitro calibrations showed that for fluorine-18-labelled radiotracers, 90% of the count rate measured by the probe corresponds to β+ emitted at a distance of 1.1 mm from the tip of the probe (Pain et al. 2000). Most particles emitted from a distance greater than 1.1 mm from the probe will be stopped in tissue and will not be detected. The probe sensitivity was experimentally determined to be 0.55 cp/KBq/mL for fluorine-18 (Pain et al. 2000). In a previous study, we demonstrated that the sensitivity and selectivity of the device make it suitable for use in the study of the kinetic properties of established or potential PET ligands in small animals (Zimmer et al. 2002).
In this study, the SIC β-microprobe, stereotaxically implanted in the hippocampus and the cerebellum, makes it possible to count the MPPF radioactivity locally. In our study, it was assumed that there were no regional differences in radioligand delivery and non-specific binding (Ichise et al. 2001). According to previous 18F-MPPF studies (Passchier et al. 2000a,b), the cerebellum was used as reference tissue because this region is practically devoid of 5-HT1A receptors (Pazos et al. 1987; Burnet et al. 1995). Thus the specific binding was estimated as the difference between the concentration of radioligand in the reference region (cerebellum, non-specific) and in the region of interest (hippocampus, specific + non-specific). This approach was coupled with microdialysis which allows the measurement of the extracellular 5-HT after electrochemical HPLC and also the measurement of the extracellular 18F-MPPF radioactivity by counting the dialysis samples using a gamma counter in the same animal.
The 18F-MPPF radioactivity curves are well documented in PET scans in humans (Passchier et al. 2000a). After i.v. administration, 18F-MPPF accumulates preferentially in the medial temporal cortex, especially in the hippocampus area, with a ratio of approximately 3 with the cerebellum (Passchier et al. 2000b). Tomography of the rat brain with microPET revealed an average hippocampus/cerebellum ratio of 2, 30 min after injection (Plenevaux et al. 2000a). Although SIC does not deliver images but defines only a detection volume drawn around the probe, our results demonstrate the validity and the feasibility of SIC for 18F-MPPF studies in rats. In our experiments, the hippocampus to cerebellum ratio increased linearly with time, reaching a value of 2 after 20 min and being reproducible between rats. No initial peak of radioactivity in the hippocampus or in the cerebellum was detected with the SIC probe, in contrast with that observed in PET scans. The␣initial peak of radioactivity is generally considered as the␣vascular bolus. Its absence in the SIC data suggest that SIC is less sensitive to the circulating radioactivity than a PET camera (Zimmer et al. 2002). The specificity of the 18F-MPPF binding on the 5-HT1A receptors was confirmed by the total displacement of 18F-MPPF after injection of unlabelled WAY 100635, a specific 5-HT1A antagonist. Simultaneously, after a 18F-MPPF injection, the measurement of the extracellular radioactivity after microdialysis collection revealed a higher radioactivity in the cerebellum than in the hippocampus. This radioactivity can be attributed to the 18F-MPPF itself since more than 90% of the radioactivity in the hippocampus and the cerebellum is due to the parent compound (Plenevaux et al. 2000b). This result could be interpreted as a lower free 18F-MPPF in extracellular space of the hippocampus as a fraction of 18F-MPPF is specifically bound on 5-HT1A receptors.
In order to observe a 18F-MPPF displacement, we increased synaptic 5-HT levels by using fenfluramine. Microdialysis studies in rats have shown that fenfluramine increases hippocampal 5-HT concentrations to the micromolar range (Thomas et al. 2000). Mechanistically, fenfluramine is known to cause a release in cytosolic 5-HT stores and inhibit its re-uptake (Rowland and Carlton 1986; Bonanno et al. 1994). In our studies, the microdialysis data and the β-kinetic measurement of 18F-MPPF were obtained simultaneously, thus providing the opportunity to examine the relationship between microdialysis and SIC measurements within each experiment. Our microdialysis results show the existence of a dose–effect relationship between challenge drug and extracellular 5-HT concentrations. The SIC measurements revealed that the magnitude of the 18F-MPPF-specific binding decreased gradually and proportionally with the magnitude of 5-HT concentration increase. A large increase in extracellular 5-HT release was therefore associated with a total displacement of the specific binding of 18F-MPPF, whereas the non-specific binding of 18F-MPPF in the cerebellum was unchanged.
Between 0 and 2 mg/kg fenfluramine injected, each per cent decrement in 18F-MPPF specific binding corresponded to a 1% increase of 5-HT concentration following fenfluramine injection. But this quantification must be interpreted with care as there is a 5-HT concentration gradient between the neuronal level and the extracellular fluid collected by microdialysis (Fisher et al. 1995; Di Chiara et al. 1996). Moreover, the extrasynaptic location of the 5-HT transporters (Zhou et al. 1998) allows the neurotransmittter to diffuse into the brain extracellular space, characterizing a paracrine neurotransmission (Bunin and Wightman 1999).
According to our results, our methodological approach allows the measurement of 18F-MPPF displacement at different levels in the rat hippocampus. Our interpretation is that 18F-MPPF is displaced from the 5-HT1A receptors after a fenfluramine-increase of 5-HT release. The consequence is that 18F-MPPF is extracted from the neuronal level and is no longer detected by SIC. At this time, the displaced 18F-MPPF appears in the extracellular level, is collected by microdialysis and detected by gamma counting.
To our knowledge few data are available in the literature documenting the displacement of a PET radioligand by cerebral 5-HT. The available works concern mainly the widely used 11C-WAY 1000635 and no data with 18F-MPPF and endogenous 5-HT are published. Moreover, it must be noticed that a direct comparison with these works has to be done cautiously since different animal models or techniques were used. A first study had shown that pre-treatment by fenfluramine decreased the specific binding of the 11C-WAY 100635 in rats and rhesus monkeys (Mathis et al. 1995). However, these results were limited as the dose of fenfluramine was high (8 mg/kg, i.v.) and the resulting tissular displacement low (< 20%). In another study in baboons, administration of the 5-HT releaser fenfluramine failed to decrease 11C-WAY 100635-specific binding (Parsey et al. 1999). In a recent paper using a combination of microPET and microdialysis (Hume et al. 2001), it has been shown that the massive release of 5-HT after 10 mg/kg fenfluramine injection (15-fold increase, according to our results) induced a reduction of the 11C-WAY 100635 of only 10–20% in the rat hippocampus. The author claimed that this minimal effect was explained by a low baseline occupancy of the 5-HT1A receptors by 5-HT, so that only a large change in endogenous agonist concentration would affect radioligand binding. Although our radioactive measurement technique is different (SIC vs. microTEP), we cannot agree with this affirmation as our results showed that a small modification of the 5-HT extracellular concentration induced a significant displacement of 18F-MPPF, in the same animal model and with the same pharmacological paradigm. The initial proposition of Seeman et al. (1989), to the effect that low affinity radiotracers are more vulnerable to endogenous neurotransmitter competition, has gained wide acceptance. The dissociation of 18F-MPPF from the 5-HT1A receptors is therefore clearly much more rapid than the dissociation of 11C-WAY 100635, possibly explaining the lower affinity of MPPF (Zhuang et al. 1994). In this view, the high affinity of 11C-WAY 100635 may restrict its use for measuring changes in endogenous 5-HT levels. However, according to Laruelle (2000), the simple binding competition might not be the only phenomenon regulating transmitter–radioligand interactions in vivo. The concept of receptor trafficking (internalization) might also be involved (Laruelle 2000; Laruelle and Huang 2001).
Our study constitutes the first encouraging observation for the serotonergic system, allowing a better understanding of the radiotracer binding vulnerability to changes in endogenous transmitter levels. A complementary study would be to investigate the effect of phasic or tonic effect of diminution of 5-HT content on the 18F-MPPF binding. This work is now in progress in our laboratory. Since it is known that some spontaneous behaviours (sleep, arousal, stress response, etc.) induce a 30–100% increase in 5-HT release in the rat hippocampus (for a review see Rueter et al. 1997), it would be of great interest to evaluate these paradigms during 18F-MPPF injection. Moreover, the possibility of the behavioural state of a patient during a PET scan having an indirect effect of the 18F-MPPF binding cannot be excluded.
In conclusion, we have established for the first time that 18F-MPPF-specific binding could reflect changes in cerebral 5-HT in the rat hippocampus. Our experiments presented showed that this displacement is well correlated with the enhanced extracellular 5-HT concentrations. This approach, combined with an appropriate pharmacological challenge paradigm, can be used to address a wide range of issues relevant to the regulation of neurotransmitter activity in vivo, mechanisms of neuropsychiatric disease or drug development in animal models. After modelization 18F-MPPF could constitute an original tool to quantify in vivo the cerebral 5-HT in limbic areas. Consequently, 18F-MPPF would be essential to realising the unique potential of the challenge techniques for measurement of 5-HT synaptic transmission in the living human brain.