Monoaminergic neuronal integrity measured in vivo by ligands of the vesicular monoamine transporter

Reserpine requires intact vesicular energy charge and pH gradient for avid and essentially irreversible VMAT2 binding. The selection of benzisoquinoline derivates for in vivo imaging of VMAT2 was based n the uncomplicated, raid, high-affinity, reversible binding properties of these ligands with the VMAT2 bindings sites (Scherman and Henry, 1984). Based in the results obtained with [11C]tetrabenazine and its known metabolism, we investigated methoxytetrabenzine (MTBZ) as a potential tracer of VMAT2. the binding properties, distribution and pharmacological profile of high specific activity tritium-labeled ligand ([3H]MTBZ) were studied on rat brain sections. Saturation analysis revealed interaction with a homogeneous population of stritial sites (Hill coefficient = 1.00 ± 0.05), with an apparent dissociation constant (Kd) of 3.9 ± 0.4 nM and a maximal binding capacity (Bmax) of 1.2 ± 0.1 fmol/μg protein determined at equilibrium. Distribution of [3H]MTBZ binding sites was observed in regions richly innervated by the monoamine systems. In presence of 1μM of various neruoactive drugs, only reserpine inhibited significantly [3H]MTBZ binding. Chapter II Is devoted to the validation of the VMAT2 as an index of dopaminergic neuronal density (Vander Borght et al., 1995c). Presynaptic location of MTBZ bindings sites demonstrated following unilateral 6-OHDA lesion of the median forebrain bundle. The resulting decrease of [3H]MTBZ stratial binding assessed by quantitative autoradiography (average 60%, range 14 to 93%; n=8). Integrity of the postsynaptic site was demonstrated by the absence of significant modification of striatal [3H]raclopride binding compared to the contralateral side.
The chapter III investigates the possible regulation of brain VMAT2 expression using both in vitro and in vivo experimental protocols. Based on the fact that the number of vesicles involved in synaptic transmission can be regulated by mobilisation from a voluminous pool of reserve, rather than by an increase of the number of transporters per vesicle (Kelly, 1993), we hypothesized that VMAT2 binding density may not be influences by disease progression or treatment. For in vitro experiments, rats were treated for 2 weeks with drugs known to influence dopaminergic neurotransmission, including those commonly used in the treatment of Parkinson’s disease (Vander Broght et al., 1995a). Autoradiographic assays were performed using [3H]MTBZ, [3H]raclopride, and [3H]WIM 35, 428 to measure VMAT2, dopamine D2 receptor and DAT bindings, respectively. None of the drug treatments significantly modified levels of VMAT2 binding. In contrast, both D2 receptors and DAT sites were altered by some of the treatment regimens. This data extends preliminary results that suggest that VMAT2 is not easily regulated and confirm the plasticity of D2 receptors and the DAT site. A second series of experiments realized in vivo in rodents (Kilbourn et al., 1996) demonstrate that dopaminergic treatments used in Parkinson’s disease do not influence VMAT2 measurements in the striatum. Striatal VMAT2 density may, thus, provide an objective estimate of monoaminergic innervation in neurodegenerative diseases, unaffected by the use of symptomatic therapies.
In chapter IV, the development of specific in vivo VMAT2 radiotracers is discussed. The structure-pharmacological activity relationship of benzisoquinoles (Lee et al., 1996). Binding of α-TBZOH to VMAT2 was demonstrated to be stereospecific : the (+)-isomer showed high affinity in vitro (Ki = 0.97 ± 0.48 nM) for rat stritial VMAT2, whereas the (-)-isomer was inactive (Ki = 2.2 ± 0.3 nM). Each isomer was then synthesized in carbon-11 labeled form, and regional brain biodistriductions in mice determined after intravenous injection. Only (+)-α-TBZOH showed selective and specific accumulations in regions of dense monoaminergic innervation (e.g. striation, hypothalamus), which could be blocked by co-injection of unlabeled tetrabenazine. The prerequisites for successful in vivo applications of an in vitro radioligand were illustrated through the development of [11C]MTBZ. Its biodistribution, metabolism and in vivo specificity were first evaluated in rodents, and human dosimetry estimated. Regional rat brain localization of [3H]MTBZ 15 min post-injection was consistent with the known monoamine nerve terminal density, demonstrating highest activity in the striatum, lateral setum, substantia nigra pars compacta, the raphe nuclei, and the locus coerulus. In vivo [11C]MTBZ binding in the mouse brain was inhibited by c)injection of excess unlabeled TBZOH. Chromatography revealed over 82% of brain activity, but less than 47% of plasma activity, corresponded to authentic MTBZ. In contrast to rodents, MTBZ appears to be metabolized only to polar derivatives in primates. Subsequently, the human brain distribution of VMAT2 binding was determined in normal volunteers following administration of [11C]MTBZ. [11C]MTBZ had high initial brain uptake and rapid clearance from all regions, with longest retention in areas of high VMAT2 concentration. After correction for labelled blood metabolites, parametric transport and binding images were calculated using arterial blood sampling and a two-compartment tracer kinetic model. Parametric quantification of VMAT2 density revealed highest distribution volumes in the putamen and caudate nucleus with lower values in cerebral cortex and cerebellum. For in vivo applications, [11C]TBZOH rather than the [11C]MTBZ was selected due to its more reliable synthesis, lesser lipophilicity reducing the plasma protein binding and nonspecific activity, and easier metabolism that exclude all labeled degradation products from the brain. This choice is again strengthned by the discovery of a stereospecificity of [11C]TBZOH increasing significantly the signal-to-noise ratio (Kilbourn et al., 1995a). With [11C]TBZOH, a tree-compartment model appears to describe better the experimental data.
Chapter V illustrates the first clinical applications of the [11C]TBZOH PET method. An aging effect on the striatal VMAT2 density has been observed in our population sample, with a 4% loss per decade. Two pathologies characterized by a decrease of dopaminergic neurons have been then investigated : Parkinson’s disease (PD) patients to correlate the striatal VMAT2 density to the clinical status (Frey et al., 1995, 1996), and patients with olivopontocerebellar atrophy (OPCA) to anticipate those that might evolve to a multisystemic degeneration (MSA; [Gilman et l., 1995, 1996]). PD patients demonstrate a decreased striatal VMAT2 density more pronounced in the posterior aspect of the putamen, richer in dopaminergic projections from the sustantia nigra pars compacta. While all MSA patients have a decrease of VMAT2 density in the striatum, only few aptients with OPCA reveal a similar diminution that renders them susceptible to develop symptomatic extrapyramidal disease. These examples illustrate clearly the interest of a non-invasive measurement of the monoaminergic neuronal density for diagnosis, prognosis and evaluation of therapeutic response. This last element becomes increasingly important to the emerging view that some treatments may not be solely symptomatic, but equally neuroprotective

In vivo measurement of dopaminergic neuronal density constitutes an objective of many neurologists interested in the non-invasive potential of nuclear medicine. The development of such a measurement would increase our understanding of neuropathophysiology and offer a unique tool to evaluate disease progression and possible modifications by therapy. After a short introduction reviewing the interest of the in vivo measurement of monoamine neuronal density, chapter I summarizes the different PET and SPET approaches developed for this purpose.
6[18F]Fluoro-L-dopa (FDOPA), a dopamine precursor analogue, was first used to image successively with Et presynaptic monoaminergic nerve terminals in human striatum (Garnett et al., 1983b). The tracer is converted to 6-[18F]fluorodopamine by dopa decarboxylase (DDC) and retained in the striatum. Its in vivo application is, however, complicated by the presence of labelled metabolites that contribute to the radioactivity measured by PET (Luxen et al., 1992). Despite the development of mathematical methods to correct for labelled degradative products or the synthesis of alternative dopamine precursors less prone to give rise to brain labelled metabolites, the uptake of dopamine analogues depends on the activity of DDC, an enzyme regulated via D2-dopamine receptors (Zhu et al, 1992; Hadjiconstantinou et al., 1993). Therefore, its uptake may reflect compensatory regulation mechanisms as well as real dopaminergic neuronal loss. The plasma membrane dopamine transporter (DAT) donctitutes an interesting alternative to measure the neuronal integrity since a diminution of their numbers has been demonstrated in vitro (Pimoule et al., 1983; Janowsky et al., 1987; Kaufman and Madras, 1991; Chinaglia et al., 1992) as well as in vivo (Leenders et al., 1990; Frost et al., 1993) following lesions of the nigrostriatal system. Nevertheless, as for the dopamine metabolite analoguesd, animal studies demonstrate regulation of the DAT site after pharmacological treatments (Scheffel et al., 1991; Kilbourn et al, 1992) or manipulation of endogenous dopamine concentration (Ikegami and Prasad, 1990; Sharpe et al., 1991; Wilson et al., 1994).
The number of neuronal synaptic vesicular monoamine transporters (vesicular monoamine transporter type 2; VMAT2) has been proposed as an index of monoaminergic neuronal innervation. WMAT2 is responsible for movement of cytoplasmic monoamines into synaptic vesicles in exchange for two intravesicular protons (for reviews, see [Schuldiner, 1994]), and is the only isoform present in the brain. Reductions of VMAT2 binding density have been demonstrated in vitro after 6-OHDA lesions of the rat striatum (Darchen et al., 1989b) and in post-mortem tissues from Parkinson’s patients (Scherman et al., 1989; Lehericy et al., 1994; Ruberg et al., 1995). Despite a lack of transport selectivity, VMAT2 binding may be considered specific to dopaminergic neurons ion the striatum, which is relatively devoid of noradrenergic or serotonergic contents (Vander Borght et al., 1995c). Three major families of VMAT inhibitors have been identified: benzisoquinolines, reserpine and ketanserin. Among these the benzisoquinolines, particularly dihydrotetrabenazine (TBZOH), have been widely used for in vitro binding studies (Scherman and Henry, 1984; Scherman, 1986; Scherman et al., 1986 and 1988b; Darchen et al., 1989b) Despite a D2 dopamine receptor antagonist action at micromolar concentrations (Reches et al., 1983), benzisoquinolines binding has been shown to be a specific marker of monoamine synaptic vesicles. In contrast, ketanserin recognizes a population of serotonergic receptors (5-HT2; [Carchen et al., 1989b]).