A specific binding site for the antiepileptic drug levetiracetam (2S-(oxo-1-pyrrolidinyl)butanamide, Keppra) in rat brain, referred to as the levetiracetam binding site, was discovered several years ago. More recently, this binding site has been identified as the synaptic vesicle protein 2A (SV2A), a protein present in synaptic vesicles [Lynch, B., Lambeng, N., Nocka, K., Kensel-Hammes, P., Bajjalieh, S.M., Matagne, A., Fuks, B., 2004. The synaptic vesicle protein SV2A is the binding site for the antiepileptic drug levetiracetam. Proc. Natl. Acad. Sci. USA, 101, 9861-9866.]. In this study, we characterized the binding properties of levetiracetam in post-mortem human brain and compared them to human SV2A expressed in Chinese hamster ovary (CHO) cells. The results showed that the binding properties of levetiracetam and [3H]ucb 30889, an analogue that was previously characterized as a suitable ligand for levetiracetam binding site/SV2A in rat brain [Gillard, M., Fuks, B., Michel, P., Vertongen, P., Massingham, R. Chatelain, P., 2003. Binding characteristics of [3H]ucb 30889 to levetiracetam binding sites in rat brain. Eur. J. Pharmacol. 478, 1-9.], are almost identical in human brain samples (cerebral cortex, hippocampus and cerebellum) and in CHO cell membranes expressing the human SV2A protein. Moreover, the results are also similar to those previously obtained in rat brain. [3H]ucb 30889 binding in human brain and to SV2A was saturable and reversible. At 4 degrees C, its binding kinetics were best fitted assuming a two-phase model in all tissues. The half-times of association for the fast component ranged between 1 to 2 min and represent 30% to 36% of the sites whereas the half-times for the slow component ranged from 20 to 29 min. In dissociation experiments, the half-times were from 2 to 4 min for the fast component (33% to 49% of the sites) and 20 to 41 min for the slow component. Saturation binding curves led to Kd values for [3H]ucb 30889 of 53+/-7, 55+/-9, 70+/-11 and 75+/-33 nM in human cerebral cortex, hippocampus, cerebellum and CHO cells expressing SV2A respectively. Bmax values around 3-4 pmol/mg protein were calculated in all brain regions. Some of the saturation curves displayed curvilinear Scatchard plots indicating the presence of high and low affinity binding sites. When this was the case, Kd values from 25 to 30 nM for the high affinity sites (24% to 34% of total sites) and from 200 to 275 nM for the low affinity sites were calculated. This was observed in all brain regions and in CHO cell membranes expressing the SV2A protein. It cannot be explained by putative binding of [3H]ucb 30889 to SV2B or C isoforms but may reflect different patterns of SV2A glycosylation or the formation of SV2A oligomers. Competition experiments were performed to determine the affinities for SV2A of a variety of compounds including levetiracetam, some of its analogues and other molecules known to interact with levetiracetam binding sites in rat brain such as bemegride, pentylenetetrazol and chlordiazepoxide. We found an excellent correlation between the affinities of these compounds measured in human brain, rat brain and CHO cells expressing human SV2A. In conclusion, we report for the first time that the binding characteristics of native levetiracetam binding sites/SV2A in human brain and rat brain share very similar properties with human recombinant SV2A expressed in CHO cells.