SV2 proteins are integral transmembrane proteins expressed on synaptic dense core vesicles [1, 2] but are also found on small clear vesicles containing neurotransmitters [3]. Cloning of the individual family members resulted in the identification of three different isoforms, SV2A [1], SV2B [4] and SV2C [5]. The overall homology between the three rat isoforms is approximately 60%, with SV2A and SV2C being more similar to each other than SV2B [5]. SV2A is ubiquitously expressed in the central nervous system (CNS), neuroendocrine cells [6, 7] and at the neuromuscular junction [8]. SV2B is widely expressed in the brain [7–12] while SV2C has a more restricted distribution in the CNS [5, 8, 11, 13, 14].

All SV2 isoforms are characterized by twelve transmembrane domains and three N-glycosilation sites in the intravesicular loop [5, 7, 9]. SV2 proteins show similarities to members of the membrane transporter family and belong to the major facilitator superfamily (MFS), although the transported substrate has not been identified yet, and the molecular function of SV2 proteins remains elusive.

However, interactions with several synaptic proteins and hypothesis about their possible function have been reported. Botulinum Neurotoxin E was shown to interact with glycosilated forms of SV2A and SV2B [8, 15]. Moreover, Mahrhold et al. reported that the carboxy-terminal region of SV2C intravesicular domain mediates the Botulinum Neurotoxin A entry leading to its toxic effects [16].

Several groups suggested a role of SV2 proteins in the regulation of presynaptic calcium concentration [17–19]. In contrast, other authors have reported that SV2 protein activities are not related to a change in presynaptic calcium [20, 21]. However, there is evidence linking SV2 and calcium given the demonstrated interaction of SV2 with synaptotagmin considered to be the primary calcium sensor triggering calcium-dependent exocytosis [22–24]. More recently, Wan et al. used mouse rod bipolar cell preparations for direct biophysical studies measuring exocytosis and presynaptic calcium concentration in the mammalian central nervous system. In these neurons, SV2B is the main SV2 isoform expressed [25, 26]. As SV2B knockout (KO) mice are viable [17], Wan et al. were able to show in rod cells from SV2B KO animals, an elevation of Ca²⁺ in nerve terminal both in resting and evoked presynaptic signals. This increase of Ca²⁺ concentration results in changes in synaptic vesicle dynamics, synaptic plasticity, and synaptic strength.

SV2A, the major isoform in the CNS has been linked with seizures and an epileptic phenotype. SV2A KO animals are characterized by the onset of epileptic seizures in early postnatal stage (postnatal day 7 or P7) leading to death around P15 [17, 27]. Moreover, it appears that SV2A expression decreases during epileptogenesis and chronically epileptic animals [28–30] as well as in patients with temporal lobe epilepsy [31, 32]. The therapeutic interest of SV2 proteins has been demonstrated with the antiepileptic drug levetiracetam. Consecutively SV2A was identified as the molecular target of the previously described brain-binding site of levetiracetam [27, 29]. Levetiracetam has a unique activity profile in animal models of seizure and epilepsy, favourable side-effect profile and straightforward pharmacokinetics [33]. Furthermore, brivaracetam, a novel antiepileptic drug (AED) with higher SV2A affinity than levetiracetam [34], shows higher potency in several preclinical models of epilepsy [35] and is currently in clinical development [36].

In this study, we therefore systematically analyzed the expression of all SV2 isoforms during mouse brain development, to contribute to the understanding of their role during development and the onset of epileptic seizures.

Our study focuses on the expression of the three SV2 protein isoforms in the mouse developing brain from embryonic day 12 (first differentiated neurons) to post-natal day 30 (P30). We analyzed several regions in the central nervous system, focusing the analysis of the SV2 protein expression in telencephalic regions given the relationships of SV2A with epilepsy. Interestingly, non quantitative immunofluorescence suggests a decrease of the SV2A signal in the hippocampus around P7 which is precisely the age at which epileptic seizures are observed in SV2A KO mice. A decrease of the SV2A signal was also observed using this semi-quantitative approach at P7 in various telencephalic regions (sub-cortical nuclei and pallidum), including the thalamus, as well the hypothalamus, the mesencephalon in the pontic and the bulbar regions suggesting potentially a technical reason for this reduction. However, this technical problem is unlikely as the decreased SV2A signal at P7 was not observed in other brain regions such as olfactory bulbs or in the cortical grey matter. Interestingly, the growth of all these telencephalic structures are characterized by a rapid expansion at P7 given the myelinogenesis which takes place at that age [37]. For this reason, we performed a precise quantification of the SV2A protein and mRNA expression levels in various hippocampal regions around P7. This precise quantification did not confirm the apparent reduction of SV2A at P7 observed by immunohistology. Similarly, a systematic analysis of SV2B expression in various brain regions by the same methodological approach gave similar results: an apparent decreased expression of SV2B in hippocampus around P7 (data not shown) which was not confirmed by the quantitative SV2B expression measurements. Thus, we believe that this decrease in SV2A and in SV2B expression which is apparent for hippocampus, is a consequence of a phenomenon which is not related to the SV2A or SV2B expression per se but rather to the enlargement and growth of tissue as myelin appears at that age in these various mouse brain regions [37], potentially affecting the semi-quantitative immunohistological read out suggesting a decrease in SV2A or SV2B expression.

A more precise quantification of SV2A, SV2B and SV2C in whole hippocampus showed an increase in the expression of SV2A and SV2B between P5 and P7, and a stabilization between P7 and P10, while SV2C showed low expression at P5, P7 and P10. Majority of the protein quantification data were confirmed by mRNA measurements using qPCR indicating elevated levels for SV2A and SV2B at later stages except for SV2B mRNA which was increased between P7 and P10. The increase of SV2A expression in the hippocampus observed by confocal microscopy measurements, between P5 and P7 seems to be restricted exclusively to the CA1 region. Indeed, we confirmed a higher SV2A expression by western blot in the CA1 after laser micro-dissection. The SV2A expression in hippocampus is thus different than the SV2B expression which progressively increases in CA1. An observation that could possibly explain that the absence of SV2A in KO mice at that age is followed by onset of epileptic seizures onset, while the absence of SV2B has no effect on seizure susceptibility [38, 39]. In this context, as presynaptic calcium regulation by SV2 proteins have been hypothesized by studying SV2B KO mice, the question of the exact role of SV2A, the target of the anti epileptic drug levetiracetam, however remains despite intense research elusive.

We thus observed a sharp increase of SV2A expression in CA1 region at P7, the age of onset of epileptic seizures in SV2A KO animals [17]. SV2A is expressed as early as neurons differentiated, the question of the appearance of seizures only at a late developmental stage has to be addressed. On the other hand, the role of SV2A in seizure susceptibility could be linked to its presence in inhibitory neurons. When the seizures appear, GABA is changing from being an excitatory to an inhibitory transmitter [40]. It was previously described, that P7 is the earliest developmental stage where a long-term potentiation in CA1 region in rats can be demonstrated suggesting an increase of synaptic plasticity at that age [41]. More recently, it has been demonstrated that levetiracetam controls more efficiently epileptic seizures when SV2A is highly expressed in tissue surrounding resected glioma [42]. This observation together with our study suggests that a high level of SV2A expression is required to protect animals from epileptic seizures and for the efficacy of levetiracetam. This conclusion is consistent with the observation that SV2A (+/−) heterozygous mice are more prone to seizures in a range of various acute models as well as accelerated epileptogenesis [43, 44]. Starting at P7, an age characterized by myelinogenesis in several telencephalic regions and by synaptic plasticity in CA1 region, a minimum SV2A level is required to protect animals from seizures. This potential “protective” effect of SV2A increased expression in CA1 at P7 is not compensate by SV2B either because SV2A exhibit specific function(s) different to SV2B or because SV2B expression does not increase in hippocampus at the same age. In this context, it can be speculated that levetiracetam could potentially increase the SV2A activity, whatever this activity is and consequently, thereby decreasing epileptic seizures. This putative role of levetiracetam is also sustained by the observation that brivaracetam, a compound which binds SV2A with a higher affinity than levetiracetam, exhibits more potent anticonvulsant properties in various acute epilepsy models [45].

In this work, we study the pattern of SV2 isoforms expression during mouse brain development. By non-quantitative and quantitative approaches, we show an increased expression of SV2A restricted to the CA1 region at P7. These findings provide new elements to better understand the role of SV2A in the epileptic seizures.