Mechanisms underlying allosteric modulation of antiseizure medication binding to synaptic vesicle protein 2A (SV2A)

Anshumali Mittal; Matthew F Martin; Laurent Provins; Adrian Hall; Marie Ledecq; Christian Wolff; Michel Gillard; Peter S Horanyi; Jonathan A Coleman

doi:10.1073/pnas.2510239122

. 2025 Sep 2;122(36):e2510239122. doi: 10.1073/pnas.2510239122

Mechanisms underlying allosteric modulation of antiseizure medication binding to synaptic vesicle protein 2A (SV2A)

Anshumali Mittal ^a, Matthew F Martin ^a, Laurent Provins ^b, Adrian Hall ^b, Marie Ledecq ^b, Christian Wolff ^b, Michel Gillard ^b, Peter S Horanyi ^c,¹, Jonathan A Coleman ^a,¹

PMCID: PMC12435242 PMID: 40892927

Significance

This study reveals the structural basis for allosteric modulation of SV2A, a key synaptic vesicle protein and target for epilepsy. Structures of the SV2A-BRV-UCB1244283, -UCB-J, and -UCB7361 complexes show that simultaneous binding of certain primary site ligands such as LEV and BRV and an allosteric modulator, UCB1244283, induces a conformation that enhances binding of antiseizure medications (ASMs). UCB1244283 binds in a lumenal vestibule above the primary site, directly obstructing dissociation and altering transmembrane domain conformation. These findings offer a mechanistic explanation for increased ASM affinity and suggest broader relevance to other SV2 proteins and MFS transporters. Our work also provides a framework for designing ligands to enhance potency and selectivity, with implications for therapeutic development targeting SV2A and related transporters.

Keywords: pharmacology, structural biology, membrane protein, allostery

Abstract

Brivaracetam (BRV) and levetiracetam (LEV) are antiseizure medications (ASMs); UCB-J is a PET tracer targeting synaptic vesicle protein 2A (SV2A); UCB7361 is closely related to padsevonil, an experimental anticonvulsant; while UCB1244283 acts as an allosteric modulator for BRV and LEV binding but not for these other ligands. The SV2A-BRV-UCB1244283 structure reveals how UCB1244283 allosterically enhances BRV binding by occupying an allosteric site near the primary binding site, preventing BRV dissociation. This allosteric site, formed by hydrophobic and uncharged residues, is an uncharacterized small-molecule binding site in SV2A. Structural analysis and mutagenesis demonstrate that an allosteric network between the primary and allosteric sites governs high-affinity ASM binding. Our studies suggest that UCB1244283 selectively binds SV2A over SV2B and SV2C, with specific mutations disrupting binding. Structures of SV2A-UCB-J and SV2A-UCB7361 show that UCB1244283 binding is only possible when the primary site ligand does not overlap with the allosteric site, and that repositioning of Ser601, Thr605, and Leu655 is critical for allosteric ligand binding. Structural comparison of multiple SV2A complexes reveals that primary site occupancy shapes the conformation of the lumenal half of the transmembrane domain, influencing how UCB1244283 binds via a connected network that differentially stabilizes TM1 in either an open or closed conformation and repositions key allosteric and primary site residues. These insights provide a foundation for developing therapeutics targeting the allosteric site and modulating SV2A function.

Epilepsy is a common neurological disorder characterized by abnormal activity of neurons which leads to seizures. Approximately 1% of the population will develop a seizure disorder, which can cause exceptional challenges for affected individuals (1). Individuals with unresponsive seizures face a lifetime of physical, mental, and social challenges as well as mortality rates more than three times higher than the rest of the population (2). A large number of antiseizure medications (ASMs) are approved as therapeutics for different seizure types, however, 30% of patients cannot achieve seizure control with current available medications either due to poor efficacy or undesirable side-effects (3). Therefore, there is a pressing need to better understand the underlying etiologies of epilepsies and to develop more targeted treatments. Levetiracetam (LEV, Fig. 1A) is distinct from other ASMs due to its unique mechanism of action, involving specific binding to synaptic vesicle protein 2A (SV2A) (4). Since its FDA approval in 1999, LEV (tradename: Keppra) has emerged as a highly effective ASM for the treatment of myoclonic and generalized tonic-clonic seizures (5, 6). LEV and related compound brivaracetam (BRV, Fig. 1A), have been shown to reduce glutamatergic transmission, a mechanism that is critical for the generation of seizures (7, 8). SV2A belongs to the major facilitator superfamily (MFS) of neuronal transporters, which also includes SV2B and SV2C (9–13). LEV and BRV bind selectively to SV2A with at least 100-fold higher affinity compared to SV2B and SV2C (3, 5, 14, 15). Padsevonil (PSL, Fig. 1A) was rationally designed to bind all SV2 members (SV2s) with nanomolar affinity and was previously investigated in clinical trials (15, 16). The lumenal domain (LD) of SV2s serves as a receptor for mediating neuronal entry of exogenous toxins, such as botulinum toxin (BoNT) and tetanus toxin, through endocytosis (17–19). BoNT is used clinically to treat conditions like eyelid twitching and neck muscle spasm, and cosmetically for facial lines (20).

Recently, we determined single-particle cryoelectron microscopy (cryo-EM) structures of SV2A bound to UCB2500 (Fig. 1A), a PSL-related ligand, and SV2B bound to PSL. Our reconstructions revealed detailed insights into the architecture of the transmembrane domain (TMD), LD, and intracellular domain (ICD) along with density features at the primary binding site corresponding to these ligands (21). At the same time, four other groups have also determined structures of SV2A in complex with LEV and BRV and SV2B in an apo state (22–26). The structures of SV2A bound with LEV or BRV closely resemble our reconstructions, except for the conformation of the lumenal half of TM1 which adopts a more closed conformation in our structure. The SV2A and SV2B TMDs exhibit signature structural motifs of MFS transporters, including a canonical MFS fold composed of twelve TM segments organized into two six-TM bundles, referred to as N- (TM1-6) and C-terminal TMDs (TM7-12). Each of these domains is formed by two structurally inverted three-TM repeats. The first helices from each three-TM repeat, TMs 1, 4, and 7, 10, are positioned at the center of the TMD to form the primary binding site, harboring the density for ASMs. SV2A and SV2B adopt a conformation resembling a lumenal-facing occluded transporter when bound to UCB2500 and PSL, respectively (21). In this conformation, a dityrosine motif composed of Tyr461/404 and Tyr462/405 in TM7 of SV2A/SV2B serves as the “lower” lumenal gate, which is closed in our structures, preventing dissociation of the bound ASMs. However, the upper region of this cavity remains accessible from the lumenal side of SV2A and SV2B. Access to the primary ligand binding site from the intracellular side of the protein is blocked by the intracellular helix bundle and large residues in the primary binding site. In the primary binding site, two highly conserved tryptophan residues (27), Trp300, and Trp666 in SV2A are involved in recognition of the pyrrolidone of racetams and a protonated aspartate residue, Asp670, is involved in binding to the amide of LEV and BRV and thiadiazole ring of UCB2500 and PSL (21, 22, 26).

Despite these advances, structural and biochemical understanding of the SV2 ligand binding sites, including the mechanisms that regulate binding, is still limited. Previous studies identified an allosteric modulator, UCB1244283, (Fig. 1A) which binds to a different site and modulates the binding of LEV and BRV (28–30). The administration of UCB1244283 in sound-sensitive mice has been shown to provide protection against both tonic and clonic seizures (28). UCB1244283 does not bind to the PSL-bound SV2A complex (15), however, incubation of UCB1244283 with SV2A results in a 2.5 or 10-fold increase in binding affinity for LEV or BRV, respectively, by slowing the rate of dissociation. Additionally, UCB1244283 increases the maximum binding capacity (B_max) by twofold for LEV and 1.4-fold for BRV (29). Interestingly, experiments using human brain membranes have shown that UCB1244283 only increases the B_max for LEV, while it increases both the B_max and binding affinity for BRV (29). These differential effects of modulator led to the hypothesis that UCB1244283 acts differently on the BRV- and LEV-bound SV2A complexes and may expose a second LEV binding site in the SV2A-LEV complex (28). Altogether, these studies demonstrate that there is an allosteric binding site in SV2s that modulates ligand binding to the primary site and displays distinct chemical specificity. However, the location of the allosteric site, the selectivity of UCB1244283 toward other SV2s, the possibility of a second LEV binding site, and the structural basis for why UCB1244283 does not bind to the larger ligand complexes like PSL are not known.

Here, we report structural and functional analysis of UCB1244283 binding to SV2A using cryo-EM and binding experiments. Our studies provide an explanation for how allosteric ligand binding slows dissociation of primary site ligands and increases binding affinity. UCB1244283 sits directly in the lumenal pathway that leads to the primary binding site and induces a conformation of TM1 that further blocks both binding sites. Our studies also uncover the underlying basis of selectivity of UCB1244283 for SV2s and other ligands like UCB-J (31) and PSL. Additionally, they suggest that the distinct effects of UCB1244283 on the BRV- and LEV-bound SV2A complexes may not involve the presence of a second LEV binding site.

Results

UCB1244283 Modulates the SV2A-LEV Complex Through an Allosteric Mechanism.

We focused on functional characterization of SV2A, using LEV and BRV, which bind to the primary site of SV2A, and UCB1244283 that binds to an unidentified allosteric site (Fig. 1A). The SV2A allosteric modulator UCB1244283 is known to bind to SV2A in the presence of smaller racetam ligands, such as LEV and BRV, but not to PSL (15). These interactions appear to induce conformational changes in SV2A (15, 29). Therefore, we started by investigating the thermostability of SV2A-UCB1244283 complexes in the presence of LEV or BRV to identify complexes which are suitable for subsequent structural and functional studies (Fig. 1B). The BRV-UCB1244283 bound SV2A complex exhibits enhanced stabilization compared to SV2A in its apo state, as well as to SV2A complexes bound with LEV-UCB1244283 or UCB1244283. Since labeled LEV is commercially available but BRV is not, we performed our binding experiments with ³H-LEV in the presence of UCB1244283. We assessed LEV binding using saturating concentrations of ³H-LEV, both with and without UCB1244283. We found that UCB1244283 induces a ~2-fold increase in specific counts, consistent with previous functional experiments (Fig. 1C) (29). Increased LEV binding upon addition of UCB1244283 enabled measurement of the half maximal effective concentration (EC₅₀) of UCB1244283 binding, which is 0.7 ± 0.3 µM for wild-type SV2A (Fig. 1D). We also measured the binding affinity of LEV in the presence of saturating concentrations of UCB1244283, finding that LEV binds with a K_d of 3.8 ± 0.8 µM for wild-type SV2A in agreement with earlier studies (29) (Fig. 1E).

Cryo-EM of the SV2A-BRV-UCB1244283 and SV2A-LEV-UCB1244283 Complexes.

Our thermostability and binding data indicated that the UCB1244283 bound SV2A-BRV and SV2A-LEV complexes exhibit higher thermostability and binding affinity (30), consequently we selected these complexes for our structural studies. We purified SV2A together with LEV or BRV and UCB1244283 by previously established methods along with 8783-Nb to stabilize the SV2A lumenal domain (LD) (21). Guided by our functional analysis, we used saturating concentrations of LEV or BRV and UCB1244283 throughout the SV2A purification. The purified SV2A complexes were subsequently analyzed by single particle cryo-EM, yielding a low-resolution map of the SV2A-LEV-UCB1244283 complex, and a 3.05 Å map of the SV2A-BRV-UCB1244283 complex after local refinement focusing on the TMD (Fig. 2A and SI Appendix, Figs. S1 and S2 Table S1). The refinement of the SV2A-BRV-UCB1244283 complex enabled the accurate modeling of all the residues in the TMD and intracellular domain (ICD) (SI Appendix, Fig. S3). Notably, we observed two distinct density features in our map, one for BRV at the primary site and another for UCB1244283 at a site located “above” the primary binding site toward the lumenal vestibule (Fig. 2A). When viewed from the lumenal side of SV2A, the binding sites of BRV and UCB1244283 appear to be positioned approximately in the middle of N-terminal and C-terminal TMDs, with a slight preference toward the C-terminal TMD, indicating that in the lumenal facing conformation these ligands bind SV2A predominantly with residues located in the C-terminal TMD. Furthermore, UCB1244283 binding to the allosteric site appears to directly block the dissociation of BRV from the primary site (Fig. 2B). Fitting of the SV2A-BRV-UCB1244283 model into the SV2A-LEV-UCB1244283 map shows that the TMD adopts a conformation that resembles the BRV-UCB1244283 state but the ICD appears disordered in the SV2A-LEV-UCB1244283 map (SI Appendix, Fig. S2), either due to its lower intrinsic thermostability (Fig. 1B) or because the disordered ICD limits particle alignment to higher resolution.

Molecular Architecture of the UCB1244283 and BRV Binding Sites.

BRV and UCB1244283 were modeled into their corresponding density features in our cryo-EM map. (Fig. 2 C and D). Consistent with recent findings, the pyrrolidone group of BRV interacts with the primary binding site through hydrophobic interactions between Trp300 and Trp666, while the carbonyl of the pyrrolidone hydrogen bonds with Tyr462. The aromatic residues Trp300, Trp666, and Tyr462 have been identified in several studies as key residues for binding both LEV and BRV (21, 22, 32, 33). The propyl group attached to the pyrrolidone extends into a subpocket near Tyr461 and TM10. Additionally, the amide group of BRV interacts with Asp670, playing a crucial role in the binding of both LEV and BRV, while the attached ethyl group forms van der Waals interactions with Leu176, Ile273, Phe277, and Cys297, which are important residues for racetam binding (21, 22). A group of conserved charged residues that are present in SV2 family members are found in the N-terminal TMD near the primary binding site: Arg262 on the lumenal end of TM4, Asp179, and Glu182 in TM1, and we observe that these residues coordinate density features for putative water molecules in our structure (SI Appendix, Fig. S4).

The binding site of UCB1244283 is located in a lumenal vestibule between the two halves of the TMD (Fig. 2 C and E), primarily constituted by residues from TM1, 5, 7, and 10. Since UCB1244283 is a racemic mixture of both R- and S-enantiomers, we modeled both isomers into the ligand density. The R- and S-forms fit equally well to the density (correlation coefficient: 0.79 vs. 0.77; SI Appendix, Fig. S5), so it is possible that both may bind and therefore, we have included both fits in our PDB model but have primarily focused our description and analysis on (R)-UCB1244283 since the interactions of each isomer are similar. Further investigations will be required to clarify the enantiomeric specificity of the allosteric site in SV2A. EC₅₀ measurements with (R)- or (S)-UCB1244283 would provide direct insight into whether one enantiomer binds with higher affinity or if there is no preference. In TM1, Phe184, Phe188, and Val597 in TM8 cap the upper part of the binding site on the lumenal side, positioned “above” the 2-methoxyphenyl group of UCB1244283. Trp300 and Met301 in TM5 block access from the “lower” part of the allosteric side to the primary binding site, while Ala308 is in the lumenal half of TM5, near Phe184 and Val597. Ser601 in TM8 is positioned near the oxygen of the methoxy group and may also form a hydrogen bond with the amide (NH) moiety of UCB1244283 while Leu655 in TM10 is located adjacent to the dimethylphenyl group. TM10 contributes several key residues for UCB1244283 binding, with Thr465 and Tyr461 forming a polar network that interacts with the oxygen of the amide group in UCB1244283 (Fig. 2E).

Analysis of the Allosteric Site Among SV2 Family Members.

To understand whether UCB1244283 can bind other members of the SV2 family, we compared the sequence of the UCB1244283 binding site in SV2A with those of SV2B and SV2C (Fig. 3A). We also compared our structure with the SV2B-PSL complex (21) and the AlphaFold-predicted (34) structure of SV2C (Fig. 3 B and C). Since our experimentally determined structures of SV2A (in complex with UBC2500) and SV2B (bound with PSL) were compared with the AlphaFold prediction of SV2C (apo), we note that the conformation of Phe170 and 174 in SV2C is different. Phe184 and Phe188 in SV2A are absolutely conserved across the SV2 family members, while several other key residues involved in UCB1244283 binding are divergent. In brief, the residue equivalent to Ala308 in SV2A is a serine in both SV2B and SV2C, and Thr465 is conserved as a threonine in SV2B and replaced by a serine in SV2C. We observed more substantial differences near the dimethylphenyl group, where the equivalent residue to Leu655 in SV2A is a glutamine in SV2B and a leucine in SV2C. Finally, Gly659 which straddles the primary and allosteric sites and contributes to a subpocket in the primary site for larger ligands, such as UCB2500, is a cysteine in SV2B and an asparagine in SV2C. Superposition of structures shows that the larger side chains of the asparagine and cysteine in SV2C and SV2B, respectively, would likely occupy the allosteric site near the dimethylphenyl group of UCB1244283.

Fig. 3. — Molecular determinants of UCB1244283 binding and selectivity. (A) Alignment of residues around the allosteric binding sites of SV2A, SV2B, and SV2C. Conserved residues are highlighted in green, residues directly involved in binding are shown in purple, and pink indicate residues directly involved in binding that are not absolutely conserved. (B) Superposition of SV2A (green) and SV2B (pink) residues in the UCB1244283 binding site. (C) Superposition of SV2A (green) and SV2C (pink) residues in the UCB1244283 binding site. (D) Concentration–response curves illustrating the effects of UCB1244283 on the binding of ³H-LEV to SV2A and allosteric-binding site mutants. UCB1244283 was used at concentrations of 30 µM, 1 µM, 100 nM, and 1 nM, and incubated with ³H-LEV (3.3 µM) in the presence of SV2A and each mutant to determine the EC₅₀ values in triplicate. The measured EC₅₀ values were 0.7 ± 0.3 µM for wild-type SV2A (black circles) and 0.20 ± 0.04 µM, 0.07 ± 0.01 µM, and 1.0 ± 0.8 µM for SV2A-F184A (blue inverted triangles), SV2A-F188A (lime stars), and SV2A-G659C (brown hexagons) mutants, respectively. EC₅₀ values for A308S (green open circles), T465A (cyan diamonds), L655Q (purple triangles), and G659N (red squares) could not be confidently determined by fitting due to their low activation binding counts. Data are shown as mean ± SEM (n = 7). (E) UCB1244283 (30 µM) binding does not enhance the maximum specific binding (B_max) of ³H-LEV (3.3 µM) to SV2A-D670A and SV2A-W300A mutants.

Molecular Basis of Binding and Selectivity of UCB1244283.

To investigate the molecular determinants of UCB1244283 binding and selectivity, we studied the effects of mutants in the UCB1244283 binding site. We used two different strategies for mutating residues: first, residues which are predicted to be important for binding and are invariant or highly conserved among all SV2s were mutated to alanine, this includes F184A, F188A, and T465A; and second, residues which differ among SV2s were mutated to the corresponding residue in either SV2B or SV2C, which includes A308S, L655Q, G659C, and G659N. We then determined the binding affinity by measuring the EC₅₀ for UCB124483 binding for each mutant (Fig. 3D).

For many of the mutants that we analyzed, a substantial reduction in the B_max was observed which precluded accurate EC₅₀ fitting. The binding data demonstrate that A308S, L655Q, T465A, and G659N mutants do not show a significant change in LEV binding. The G659C mutant by contrast displayed a ~4-fold reduction in B_max but exhibited a similar EC₅₀ for UCB1244283 compared to wild-type SV2A. Finally, the F184A and F188A mutants showed minor reductions in B_max and exhibited enhanced EC₅₀ values for UCB1244283 by ~3- and 9-fold, respectively.

Previous studies suggest that UCB1244283 acts differently on the BRV and LEV complexes. UCB1244283 slows association and dissociation rates of BRV and increases its K_d, while LEV binding increases the B_max by twofold without significantly affecting LEV binding kinetics (28–30). These observations led to the hypothesis that UCB1244283 binding may expose a second LEV binding site. To address this hypothesis, we measured LEV binding in the presence of UCB1244283 using the W300A and D670A mutants, which are known to disrupt LEV binding at the primary binding site (21, 22). In these experiments, we were unable to measure any significant LEV binding to either mutant with or without UCB1244283 (Fig. 3E). We also measured the thermostability of the D670A and W300A mutants (SI Appendix, Fig. S6 A–C), finding that both mutants were less stable than the wild-type SV2A, and that only the D670A mutant exhibited a modest increase in thermostability upon addition of LEV and UCB1244283. The increase in stability for D670A with LEV and UCB1244283 may result from a partial retention of binding affinity for LEV (30).

SV2A-UCB-J and SV2A-UCB7631 Complexes and Analysis of the Molecular Architecture of Binding Sites.

Building on the finding that the SV2A-BRV and -LEV complexes allow for binding of UCB1244283 while the SV2A-PSL (15, 29, 30) and -UCB2500 complexes do not, we explored interactions of UCB-J and UCB7361 (Fig. 4A) with UCB1244283, to understand the structural determinants of UCB1244283 binding. UCB7361 is similar to PSL and UCB2500, containing one additional carbon compared with PSL and a fluorine instead of a chlorine (trifluorinated propyl group). We began by investigating the binding affinity between SV2A and ³H-UCB-J in the presence of saturating concentration of allosteric modulator UCB1244283 to determine whether it also enhances UCB-J binding. SV2A displayed a similar binding affinity for UCB-J under both conditions, with a K_d of 6 ± 1 nM in the presence and 5.3 ± 0.8 nM in the absence of UCB1244283 (Fig. 4B), indicating that UCB1244283 does not bind to the SV2A-UCB-J complex. Thus, we purified SV2A bound with UCB-J alone or UCB7361 together with UCB1244283 and analyzed the resulting complexes using single particle cryo-EM, determining a 3.42 Å and 2.91 Å map of SV2A-UCB-J and SV2A-UCB7361, respectively (Fig. 4 C and D and SI Appendix, Fig. S7 A–H and Table S1). The maps of both complexes enabled the accurate modeling of residues in the TMD and ICD. Notably, we observed density features of UCB-J or UCB7361 at the primary binding site in both complexes. However, no density was observed at the allosteric site for UCB1244283 in the SV2A-UCB7361 complex. The pyrrolidone group of UCB-J and UCB7361 engages the primary binding site through hydrophobic interactions with Trp300 and Trp666. Additionally, the hydroxyl side chain of Tyr462 forms a hydrogen bond with the pyrrolidone carbonyl group, consistent with interactions observed in other SV2A complexes. In the SV2A-UCB7361 and -UCB-J complexes, the trifluorinated propyl group of UCB7361 occupies a hydrophobic pocket formed by Tyr461, Ile663, and Gly659, while the trifluorophenyl group of UCB-J fits into the same pocket created by Tyr461, Ile 663, Gly659, and Val608. The nitrogen atom of the thiadiazole ring in UCB7361 and the pyridine ring in UCB-J form a hydrogen bond with the protonated side chain of Asp670. Similar to the SV2A-UCB2500 complex, the trifluorinated methyl group of UCB7361 also interacts with Cys297.

Fig. 4. — Structures of SV2A with UCB-J and UCB7361, and the capability of UCB1244283 for binding these complexes. (A) Chemical structures of UCB-J [(4R)-1-((3-Methyl-4-pyridinyl)methyl)-4-(3,4,5-trifluorophenyl)-2-pyrrolidinone] and UCB7361 [(4S)-1-((2-(methoxymethyl)-6-(trifluoromethyl)imidazo(2,1-b)(1,3,4)thiadiazol-5-yl)methyl)-4-(3,3,3-trifluoropropyl)pyrrolidin-2-one]. (B) Binding of ³H-UCB-J in the presence (*K_d* = 6 ± 1 nM) or absence of UCB1244283 (*K_d* = 5.3 ± 0.8 nM). (C) Overview of the SV2A-UCB-J structure showing transmembrane helices TM1-6 (green) and TM7-12 (lilac). UCB-J (brown) binds at the primary site. Inset shows residues involved in UCB-J binding. (D) Overview of the SV2A-UCB7361 structure showing transmembrane helices TM1-6 (green) and TM7-12 (lilac). UCB7361 (purple) binds at the primary site. Inset shows residues involved in UCB7361 binding. (E) Sterics of the UCB1244283 (carbons, blue; hydrogens, white) and UCB2500 (pink) ligand binding sites, compared using the SV2A-BRV-UCB1244283 and SV2A-UCB2500 (PDB: 8UO9) complexes. Dotted lines indicate atoms that are predicted to be sterically incompatible. (F) Sterics of the UCB1244283 and UCB7361 (purple) binding sites. (G) Sterics of the UCB1244283 and UCB-J (brown) binding sites. These models demonstrate that UCB2500, UCB7361, and UCB-J at the primary binding site are in close proximity to UCB1244283 binding at the allosteric site. (H) Allosteric site residues of the UCB7361 complex (gray) that exhibit differences in positioning relative to the BRV-UCB1244283 complex (green) are depicted as sticks.

Next, we conducted detailed comparisons of the binding sites of each complex to elucidate why UCB1244283 binds to some primary site ligand complexes but not others. When we compared our SV2A-BRV-UCB1244283 structure to SV2A bound to UCB2500 and SV2B bound to PSL, our superpositions revealed that the fluorines of the trifluorobutyl and chloro-difluoroethyl groups in UCB2500 or PSL are located within ~4 Å of the dimethylphenyl group in UCB1244283, whereas the distance of the fluorine of UCB7361 to the dimethylphenyl group is 3.7 Å. (Fig. 4 E and F). We also examined whether the binding sites of UCB-J and UCB1244283 overlap, finding that the fluorine atoms of the trifluorinated phenyl group of UCB-J directly overlaps with the dimethylphenyl group in UCB1244283 (Fig. 4G), indicating a shared binding region. We also compared allosteric site residues between the SV2A-BRV-UCB1244283 and SV2A-UCB7361 complexes to understand whether there are other differences in conformation which prevent UCB1244283 from binding. Notably, binding of UCB1244283 and BRV to SV2A leads to repositioning of Ser601 and Thr605 in TM8 and Leu655 in TM10, compared with other SV2A structures, including SV2A-UCB2500, SV2A-UCB-J, and SV2A-UCB7361 (Fig. 4H). These alterations facilitate the remodeling of the allosteric site that allows UCB1244283 binding. Interestingly, Thr465, Pro469, Val608, and Cys656, which are in proximity to the repositioned residues, show no significant changes. We find that changes in the sidechain positions of Val600, Ser601, Thr605, and Leu655 would render the allosteric site incompatible with binding of UCB1244283 in the SV2A-UCB7361 and -UCB-J complexes since this would result in significant clashes with UCB1244283 (Fig. 4H).

Ligand-Induced Remodeling of the SV2A TMD Conformation.

Next, to understand the basis of ligand-induced conformational changes, we compared the TMDs and ICDs between the SV2A-BRV-UCB1244283 complex and other SV2A structures (21, 22), bound with UCB2500, BRV, LEV, UCB-J, and UCB7361 and the SV2B-PSL complex (Table 1), finding that the conformation of the SV2A-BRV-UCB1244283 complex most closely resembles the SV2A-UCB2500, UCB-J, and UCB7361 complexes. Comparison of the SV2A-BRV-UCB1244283 structure with the LEV-bound (PDB: 8JS8) or BRV-bound (PDB: 8K77) SV2A structures revealed substantial differences in the conformation of the lumenal half of TM1 (Fig. 5A). In the BRV- and LEV-bound structures of SV2A, the TM1 helix is substantially longer with a kink, featuring a 3₁₀ helix spanning residues 181 to 184 (Fig. 5B). In contrast, this region is a continuous α-helix and is shorter in the SV2A-UCB2500, UCB-J, UCB7361, SV2B-PSL, or SV2A-BRV-UCB1244283 structures (Fig. 5B and SI Appendix, Fig. S8 A–E). In the SV2A-BRV-UCB1244283 structure, this α-helical region undergoes a helical displacement of 4.7 Å (Cα to Cα distance for Gly187) compared to the SV2A-BRV structure. This conformational change results in blocking of both the allosteric and primary binding sites by Phe188. In the SV2A-BRV structure, Pro191 forms a hinge point after the 3₁₀ helix and residues 191 to 197 are α-helical, whereas in our SV2A-BRV-UCB1244283, SV2A-UCB2500, SV2A-UCB-J, SV2A-UCB7361, and SV2B-PSL structures, this region is modeled as a structured loop. We also observe shifts in the lumenal halves of TM2, 5, 7, 8, 10, and 11 in our SV2A-BRV-UCB1244283 and SV2A-UCB2500, UCB-J, and UCB7361 structures relative to the BRV-bound SV2A structure (Fig. 5 C and D and SI Appendix, Fig. S8 F–I). These TMs contribute the majority of the residues responsible for binding both ASMs and UCB1244283, and these rearrangements modify the helical packing between TMs. They also alter the hydrogen bonding network in TM8, which is more extensive in the SV2A-BRV-UCB1244283 complex, involving Lys621, Asn667, Ser294, Asn612, Thr605, and Met301 (SI Appendix, Fig. S8 J–M). In TM8, residues 603 to 609 which line both primary and allosteric binding sites, we observe a distinct shift in the lumenal half of TM8 in the SV2A-BRV-UCB1244283 complex compared to other SV2A complexes. This conformational adjustment appears to facilitate binding of UCB1244283, preventing steric clashes at the allosteric site and likely enhances binding of UCB1244283. Additionally, several significant changes in the position of primary site residues between various complexes were also found, primarily in the positioning of Trp300 and Met301 which are shifted toward the primary binding site in the SV2A-BRV-UCB1244283 and SV2A-UCB2500 complexes vs. the SV2A-BRV and SV2A-LEV complexes (Fig. 5E) which may also contribute toward inducing these conformational changes.

Table 1.

Comparison of RMSD values between SV2A complexes

Complex name	RMSD (Å)
SV2A-UCB2500 (PDB: 8UO9)	0.82
SV2A-LEV (PDB: 8JS8)	0.95
SV2A-BRV (PDB: 8K77)	0.98
SV2B-PSL (PDB: 8UO8)	0.96
SV2A-UCB-J	0.78
SV2A-UCB7361	0.79

Open in a new tab

The TMD and ICD of each complex were compared to the SV2A-BRV-UCB1244283 complex.

Fig. 5. — Conformation differences between the SV2A BRV-UCB1244283 and other ligand complexes. (A) Comparison of the SV2A-BRV (PDB: 8K77) and SV2A-BRV-UCB1244283 complexes highlight conformational rearrangements in the transmembrane helices surrounding the primary and allosteric binding site observed from the lumenal side. (B) Structural alignments of the SV2A-BRV (gray), UCB2500 (green), and SV2A-BRV-UCB1244283 (pink) complexes reveal a conformational transition, from a 3₁₀-helix to α-helix, in the TM1 helical region near the UCB1244283-binding site. (C) Comparison of the allosteric site between SV2A-BRV-UCB1244283 (green) and the SV2A-BRV structures (gray). The positions of BRV and UCB1244283 are shown in tan and blue sticks respectively. (D) Comparison of the allosteric site between SV2A-BRV-UCB1244283 (green) and the SV2A-UCB2500 structures (pink). (E) The structural superposition between the SV2A-BRV-UCB1244283 complex (green sticks, BRV shown in tan) and the SV2A-BRV (gray sticks, BRV shown in light brown) or SV2A-LEV (purple sticks, LEV shown in dark brown) structures revealed alterations in primary site residue positioning.

Discussion

The SV2A-BRV-UCB1244283 complex traps SV2A in a conformation where both the primary and allosteric binding sites are occupied by ligands, allowing for detailed mechanistic insights into allosteric modulation of ASM binding. UCB1244283 binds to a lumenal site in a vestibule located in the pathway which leads to the primary binding site, where ASMs like LEV and BRV bind (Fig. 6A). The allosteric site is positioned “above” the primary binding site, directly obstructing the dissociation of ASMs from the primary binding site (Fig. 6B). Therefore, our structure provides a direct explanation of the mechanism by which UCB1244283 slows the dissociation rate of ASMs and increases their binding affinities. The molecular basis for how MFS transporters are allosterically controlled by small-molecule binding is also not well understood. While allosteric sites have been identified in other MFS transporters, they typically bind either lipid, detergent, or ions (35). The allosteric site in SV2A is a novel binding site for small-molecule ligands, and our results suggest that similar mechanisms and binding sites may be present in other SV2 family members and related MFS transporters. This allosteric mechanism closely parallels that of the unrelated serotonin transporter (SERT), where allosteric ligands similarly influence the binding of inhibitors to the primary site (36).

Fig. 6. — Model illustrating the effects of the SV2A allosteric modulator UCB1244283 on the enhanced binding kinetics of antiseizure medications (ASMs). (A) SV2A is composed of two six-transmembrane (TM) bundles, N-terminal (TM1–6, green) and C-terminal domains (TM7–12, purple). Structural analyses of SV2A-BRV (PDB: 8K77) and SV2A-BRV-UCB1244283 reveal a single binding site for BRV, located between the two halves of the transporter. The TM1 helix, toward the lumenal side, is oriented away from the C-terminal domain due to a kink created by the 3₁₀-helix, which results in an open upper lumenal gate. (B) UCB1244283 binding to the allosteric site stabilizes BRV at the primary binding site by sterically obstructing its unbinding, while inducing conformational changes in the surrounding TMs, notably, inducing a conformation change in the lumenal half of TM1 from a 3₁₀-helix to an α-helix, resulting in reorientation of TM1 toward the C-terminal half of the TMD. Key phenylalanine residues lining the allosteric site are shown in green. This conformation results in closure of the upper lumenal gate in the SV2A-BRV-UCB1244283 structure, thereby enhancing the binding affinity of BRV at the primary binding site.

UCB1244283 may serve as a useful tool for trapping endogenous molecules at the primary site, aiding the investigation of SV2A function (37). SV2s have been thought to be transport proteins, but the endogenous substrate has not yet been identified (9, 10, 14, 38–40). For most transporters, the substrate affinity is in the micromolar range, which makes purifying substrate-bound transporters, especially from native sources, challenging (35). Transported substrates are predicted to bind to the primary site (35), and since UCB1244283 blocks unbinding from this site, it could be used to facilitate the isolation of SV2A bound to its natural substrate (21, 22, 25, 37, 38).

Several strategies have been employed in other transporters, GPCRs, and other proteins to effectively target neighboring binding sites, which may be applicable to SV2s given the close proximity of the primary and allosteric sites (41–45). For example, in SERT, bivalent serotonin molecules have been developed to bridge both sites (42–44). Bitopic ligands have been designed in GPCRs to target both allosteric and orthosteric sites; these ligands are functionally selective which may allow for improved efficacy (43–45). Another approach would be to develop high-affinity versions of UCB1244283 that target the allosteric site of SV2A with optimized pharmacological properties that could be used in conjunction with existing therapies to improve potency and specificity. It is conceivable that such strategies may also prove useful in further development of novel molecules against SV2s and other related SLC transporters.

Previous saturation binding studies have indicated that UCB1244283 differentially affects BRV binding to SV2A compared to LEV (28–30). UCB1244283 increases BRV binding primarily by increasing its affinity and B_max, while it increases LEV binding mainly through an increase in B_max. This observation has led to the hypothesis that the number of LEV binding sites increases from one to two, suggesting that LEV and BRV either bind to different sites or modulator induces significant conformational changes in SV2A, altering its interactions with BRV or LEV. Our binding experiments with W300A and D670A mutants, together with UCB1244283, suggest that there is not a second LEV binding site (Fig. 3E). However, kinetic association experiments performed by adding UCB1244283 prior to the addition of labeled primary site ligands would suggest that binding of a primary site ligand is required for UCB1244283 binding (28). Therefore, our results do not definitively rule out the presence of a second LEV binding site, particularly given that the resolution of our SV2A-LEV-UCB1244283 map is insufficient to clearly visualize bound ligands. Our studies do offer an alternative explanation for the observed relative increase in B_max for the SV2A-LEV-UCB1244283 and BRV-UCB1244283 complexes: rather than indicating a second LEV binding site, these differences may reflect variations in the relative populations of open and closed TMD conformations of the tripartite complexes compared with the SV2A-LEV and SV2A-BRV complexes.

Mutagenesis of residues around the primary binding site reduced the binding of both BRV and LEV (22, 30, 32, 33). However, for other mutants, the modulator’s effect on increasing the B_max for LEV was abolished, but not for BRV, further suggesting that SV2A ligands interact distinctly with the primary or other binding sites (28–30). Alignment of the primary binding site in the SV2A-BRV- UCB1244283 complex with the structures of SV2A-BRV (PDB 8K77), and SV2A-LEV (PDB 8JS8), shows what appear at first glance to be only minor changes, primarily around Trp300, Asn667, and Met301 (Fig. 5E). Subtle differences in the conformation of primary binding site residues and/or the surrounding TMs may also be responsible for the previously observed binding differences between the UCB1244283-BRV and -LEV complexes (Figs. 4G and 5 C–E). Trp300 and Met301 reside in the primary binding site sitting “beneath” the allosteric site. The methoxyphenyl ring of UCB1244283 pi-stacks with the sidechain of Trp300, and this interaction may be partially responsible for repositioning of this residue in comparison to other SV2A-ASM complexes. These residues may be part of an allosteric network that connects the primary site to the allosteric site and is responsible for transmitting the binding differences that are observed between the UCB1244283-LEV and BRV complexes (28–30) as well as the SV2A-LEV and SV2A-BRV complexes without UCB1244283 (22).

Superposition of the SV2A-BRV-UCB1244283 structure with other SV2 structures, such as the SV2A-BRV, -LEV, -UCB2500, UCB7361, UCB-J, or SV2B-PSL complexes, demonstrates no large-scale conformational changes upon occupancy of the allosteric site by UCB1244283. However, our structural analysis highlights several critical conformational changes in the primary binding site, allosteric site, and the lumenal half of the TMD, that distinguish the SV2A-BRV-UCB1244283 complex from other previously characterized complexes. Comparison of the SV2A-BRV-UCB1244283 and SV2A-UCB2500, -UCB-J, and -UCB7361 structures with SV2A-BRV and SV2A-LEV revealed significant conformational differences in the upper lumenal part of TMDs, particularly in TM1 and TM8, which result in a closed TMD conformation. In addition to the blockade of the primary site by UCB1244283 binding, the conformational changes in TM1 involving Phe184 and Phe188 further obstruct the lumenal side of the allosteric site in both the SV2A-BRV-UCB1244283 and SV2A-UCB2500, -UCB-J, and -UCB7361 complexes, which likely also prevents primary ligand dissociation. Changes in the position of primary site residues in the SV2A-BRV-UCB1244283, -UCB2500, -UCB-J, and -UCB7361 complexes also alter the hydrogen bonding network in TM8 relative to the SV2A-BRV and SV2A-LEV complexes, and these changes appear to be transmitted from the primary site to the lumenal half of the TMD, thereby repositioning key residues in the allosteric site. Therefore, these networks may control the conformation of the TMD, and conformational change can be regulated both by primary or allosteric ligand binding, thus explaining why the SV2A-BRV-UCB1244283 and SV2A-UCB2500, -UCB-J, and -UCB7361 complexes are more closely related to one another vs. SV2A-BRV and SV2A-LEV complexes.

We predict that nonconserved substitutions of residues in SV2B would induce steric clashes with the dimethylphenyl group of UCB1244283 in the allosteric site and based on our binding data for these mutants in SV2A, we conclude that UCB1244283 likely does not bind SV2B. For SV2C, the most notable difference in the allosteric site is the replacement of the residue equivalent to Gly659 with asparagine. The longer side chain of the asparagine provides an explanation for why the G659N mutant also disrupts UCB1244283 binding. Future studies will be required with wild-type SV2B and SV2C to understand whether UCB1244283 is an SV2A-specific allosteric modulator.

UCB1244283 is known to selectively bind to the allosteric site in the presence of certain primary site ligand complexes, while binding to others, such as, PSL, UCB-J, and UCB2500, are not compatible (15). Our structural analysis provides two explanations for this selective binding behavior. First, we find that the chloro-difluoro, trifluoro, and trifluorophenyl groups of PSL, UCB2500, UCB7361, and UCB-J would likely create steric hindrance by interfering with the dimethylphenyl group of UCB1244283, disrupting binding (Fig. 4 E–G). The SV2A-UCB7361 complex stabilizes the same closed conformation of TM1 observed in other complexes. Interestingly, the SV2A-UCB7361 complex does not contain density for UCB1244283 despite it being present during purification. Further analysis of the SV2A-UCB7361 complex reveals a second reason for lack of UCB1244283 binding to these complexes: occlusion of allosteric site by sidechains of Ser601, Thr605, and Leu655. Closer inspection of the allosteric site shows that these residues adopt positions similar to those observed in the SV2A-UCB-J or SV2A-UCB2500 structures, rather than conformation seen in the SV2A-BRV-UCB1244283 structure. These findings suggest that these residues are critical for binding UCB1244283. Based on these observations, ligands like BRV and LEV stabilize SV2A in an open conformation that permits access to the allosteric site, where subsequent UCB1244283 binding induces conformational rearrangements in the TMD and repositions key residues to optimize interactions. In contrast, ligands such as UCB-J, UCB2500, and UCB7361 favor the closed conformation of SV2A, which not only restricts access to the allosteric site but also sterically hinders UCB1244283 binding through altered positioning of binding residues and, in some cases, the overlap of binding sites.

In conclusion, this study provides a detailed structural analysis of the essential synaptic vesicle protein, SV2A, bound to the ASM, BRV, and an allosteric modulator, UCB1244283. We demonstrate how UCB1244283 binds in a hydrophobic pocket, only in combination with LEV and BRV and not with UCB-J or UCB7361 and selectively binds to SV2A within the SV2 family. BRV and LEV bind to a shared primary site but interact distinctly with SV2A. Binding experiments with mutants at the primary binding site in the presence of UCB1244283 suggest that SV2A may not contain a second cryptic LEV binding site (28–30). We further highlight the role of conserved residues in the human SV2 family, which are not directly part of the primary or allosteric sites but contribute to a network involved in mediating crosstalk between these two functional sites. These differences were primarily localized to the lumenal regions of the TMD, particularly in TM1 and TM8, which resulted in a closed TMD conformation in our structures. The binding of UCB1244283 to the allosteric site induces changes in primary site residues which alters the local conformation of the TMD. These structural changes are transmitted between both sites, affecting the conformation of key residues and facilitating ligand interactions. Notably, the enhanced binding affinity of BRV in the SV2A-BRV-UCB1244283 complex is likely driven by these changes and by obstruction of the primary site by UCB1244283. Collectively these interactions are distinct from those observed in the SV2A-LEV or SV2A-BRV complexes (22) without UCB1244283 and explain the mechanism of allosteric modulation of ASM binding.

Materials and Methods

Small-Molecule Ligands.

PSL, ³H-LEV, ³H-UCB-J, UCB-J, UCB7361, and UCB1244283 were obtained from UCB Pharma through a Material Transfer Agreement. ³H-LEV was purchased through this agreement from American Radiolabeled Chemicals (ART 1832). BRV (Cat No. R020192) and LEV (Cat No. L8668) were purchased from MuseChem and Sigma, respectively.

Construct Design and Cloning.

The SV2A construct used in this study has been described previously and behaves similar to full-length SV2A in ligand binding experiments (21). In brief, the first 64 amino acids were deleted from the 5’ end of human SV2A, and the 3’ end was extended with sequences coding a 3C protease site (L-E-V-L-F-Q-G-P), mVenus (21), TwinStrep, and 10-His-tags. This sequence was cloned into pEG-BacMam (46), resulting in SV2AΔ64-mVenus construct. The SV2AΔ64-mVenus construct was subjected to site-directed mutagenesis to create F184A, F188A, A308S, T465A, L655Q, G659C, and G659N mutants. The mutants were verified by sequencing.

Fluorescence-Detection Size-Exclusion Chromatography.

Fluorescence-detection size-exclusion chromatography (FSEC) analysis (47, 48) was conducted using a Shimadzu HPLC instrument equipped with a refrigerated autosampler, multiwavelength detector, and a Superose 6 Increase 5 × 150 column. The fluorescence of mVenus (excitation at 515 nm and emission at 528 nm) was monitored in whole cell lysates solubilized using 10 mM lauryl maltose neopentyl glycol (LMNG) detergent and 1 mM cholesteryl hemisuccinate (CHS) for mVenus-tagged SV2A. For purified proteins, tryptophan fluorescence (excitation at 280 nm and emission at 325 nm) was monitored.

Expression and Purification of 8783 and 8667-Nbs.

The expression and purification of 8783-Nb and 8667-Nb was performed following our previously described protocol (21). In brief, 8783-Nb or 8667-Nb expression was carried out in the periplasm of BL21-DE3 E.coli cells from pET26-8783-Nb or pET-8667-Nb plasmid harboring a PelB sequence. The periplasmic fraction of the cells was prepared by incubating the harvested cells with ice-cold hypertonic solution (20mM Tris-HCl, pH 8.0, 20% sucrose, 1 mM EDTA) (49) for 30 min, followed by centrifugation at 7600×g. The 8783-Nb or 8667-Nb was purified from the diluted supernatant by binding to a Ni²⁺-NTA gravity flow column, followed by size-exclusion chromatography (SEC) of eluted fractions using a Superdex 75 Increase 10/300 column (Cytiva).

Expression and Purification of SV2A-LEV-UCB1244283-8783-Nb, -BRV-UCB1244283-8783-Nb, -UCB-J-8667-Nb, -UCB7361-8783-Nb Complexes.

The recombinant expression of the SV2AΔ64-mVenus protein was carried out in mammalian tsA201 cells using the baculovirus-mammalian (BacMam) cell expression system, using Sf9 cells to produce baculovirus (46). Harvested cell pellets were resuspended in TBS₁₅₀ (20 mM Tris-HCl pH 8.0, 150 mM NaCl), protease inhibitor cocktail (200 nM aprotinin, 2 µM leupeptin, 2 µg/mL pepstatin) and were lysed using high-frequency ultrasonic waves (Misonix Sonicator 3000). Unbroken cells and cell debris were removed by centrifugation at 10,000 rpm [SS(21) rotor, Sorvall RC6] for 10 min, followed by ultracentrifugation of supernatant fraction at 40,000 rpm (Ti45 rotor, Beckman Coulter) for 90 min at 4 °C to isolate membranes. Cell membranes were resuspended in TBS₁₅₀ and incubated with 30 µM UCB1244283 and 100 µM of levetiracetam (LEV) (Sigma, Cat No. L8668) or BRV (MuseChem, Cat. Number R020192) or 1 µM of UCB7361 for 30 min on ice, followed by solubilization using 10 mM lauryl maltose neopentyl glycol (LMNG) and 1 mM cholesteryl hemisuccinate (CHS) at 4 °C for 90 min. For UCB-J the same procedure was followed using 10 µM UCB-J but without UCB1244283. The insolubilized fraction was removed by ultracentrifugation at 40,000 rpm for 60 min at 4 °C, and the cleared supernatant was loaded onto a gravity-flow column containing 4 ml GFP-nanobody conjugated sepharose resin. The resin was washed with 20 column volumes of TBS₁₅₀ containing 0.020 mM CHS, 0.2 mM glyco-diosgenin (GDN), 30 µM UCB1244283, and 10 µM of LEV or BRV or 1 µM of UCB7361. For UCB-J, 1 µM UCB-J was added without UCB1244283. The resin bound with SV2AΔ64 was incubated overnight with 3C protease to remove mVenus, TwinStrep, and 10-His-tags. The next day, the flow-through containing SV2AΔ64 was mixed with 4 to 5 fold molar excess of 8783-Nb or 8667-Nb, followed by concentration to 500 µl using a 100 kDa concentrator (Amicon). The SV2AΔ64-8783-Nb or SV2A-8667-Nb complex was injected onto a Superose 6 10/300 column (Cytiva) in TBS containing 0.020 mM CHS, 0.2 mM GDN, and 30 µM UCB1244283 and 25 µM LEV or 10 µM of BRV or 1 µM of UCB7361. For UCB-J, 1 µM UCB-J was added without UCB1244283. SEC fractions of the homogenous peak corresponding to monomeric SV2AΔ64 were further individually analyzed by FSEC.

SEC fractions that were homogenous based on FSEC analysis were pooled and concentrated to 5.3 mg/ml or 8 mg/ml or 7.5 mg/ml using a 100 kDa concentrator before addition of 30 µM UCB1244283 and 1 mM of LEV or BRV or UCB7361, respectively. For the SV2A-UCB-J complex, 1 mM UCB-J was added and concentrated to 5.4 mg/ml. The concentrated sample was ultracentrifuged at 60,000 g prior to cryo-EM grid preparation.

Cryo-EM Sample Preparation and Data Collection.

2.5 µl of SV2A was applied to glow discharged Quantifoil holey carbon grids (1.2/1.3 200 mesh copper). Grids were blotted for 2 to 4 s with a blot force of 2 to 4 (100% humidity, 4 °C) using an FEI Vitrobot MK IV (ThermoFisher). SV2A grids were imaged at the University of Pittsburgh on a Titan Krios G3i operating at 300 kV equipped with a Falcon 4i direct electron detector and a Selectris energy filter set to a slit width of 10 eV. Movies were collected at a pixel size of 0.719 Å/pixel with defocus ranging from −0.5 to −2.0 µm with a total dose of 50 e/Å².

Cryo-EM data processing and model building.

A total of 7,888 or 7182 or 17,362 or 8,433 exposures were selected for SV2A-BRV-UCB1244283 or -LEV-UCB1244283 or -UCB-J or -UCB7361, respectively and CryoSPARC v4.2 (50) was used to perform patch motion correction and contrast transfer function (CTF) estimation. Particles were picked using reference-free blob picker followed by particle extraction at a box size of 330 or 336 pixels. These particles were binned to 128 pixels and classified through five to ten rounds of heterorefinement, using “decoy” classes with random density, empty detergent micelles and low-resolution ab-initio models of SV2A, followed by 2D classification. The selected particles were used to generate an ab-initio model to create projection-based templates. Next, we performed a patch motion correction on a random subset of 100 exposures, which were then used to generate a denoise model. This model was subsequently utilized in denoising the entire set of micrographs, which were used for particle picking using the template picker. Particles from blob pick and template pick were individually classified multiple times (10 to 15×) using heterorefinement as described above followed by several rounds of 2D classification. Duplicate particles were removed. These particles were used for generating three 3D-maps through ab-initio reconstruction. One 3D-reconstruction was chosen for re-extraction at a box size of 256 pixels, which were refined after nonuniform refinement followed by local refinement using the TMD mask. These particles were re-extracted at a box size of 330 or 336 pixels and were subjected to two rounds of Bayesian polishing in RELION v.3.1 (51, 52). Polished particles were subjected to two rounds of 3D-classification in cryoSPARC, which were refined to 3.05 Å for SV2A-BRV-UCB1244282, 3.42 Å for SV2A-UCB-J, and 2.91 Å for SV2A-UCB7361 after local refinement using the TMD mask. SV2A (PDB: 8UO9) model was fit into cryo-EM density maps, which was manually modeled in Coot (53, 54), iteratively real space refined in Phenix (version 1.20.1) (55), and validated by comparing the half maps and refined model.

Thermostability Assays.

SV2AΔ64-mVenus or SV2AΔ64-D670A-mVenus or SV2AΔ64-W300A-mVenus containing tsA201 cell membrane were solubilized in TBS₁₅₀ containing 5 mM LMNG, 1 mM CHS, with and without ligands (apo, 100 µM LEV, 20 µM BRV, 30 µM UCB1244283, LEV/UCB1244283, and BRV/UCB1244283 at the same concentration) for 60 min at 4 °C. These samples were ultracentrifuged at 60,000×g for 30 min to remove cell debris and the supernatant was heated at various temperatures for 15 min, followed by ultracentrifugation for 20 min to remove aggregates. The supernatants were subjected to FSEC (46) on a Superose 6 Increase 5 × 150 column for detection of mVenus fluorescence.

Radioligand Binding Assays.

SV2AΔ64-mVenus wild-type or single residue mutant plasmids were transfected into tsA201 cells in suspension culture (200 ml) using polyethylenimine 40K (PEIMAX). After 48 to 60 h post transfection, cells were harvested, and membranes enriched with wild-type or each mutant were prepared as described above. Protein expression of each construct was analyzed by FSEC, and membranes containing equal amounts of proteins were used in the filter binding assays. We prepared a working radioligand ³H-LEV by diluting concentrated ³H-LEV (200 µM, 5 Ci/mmol) using unlabeled LEV to a final concentration of 2.5 µM and 100 µM, respectively.

For saturation-binding studies, SV2AΔ64-mVenus enriched membranes were incubated for 2 h in presence of 30 µM UCB1244283 and a twofold serially diluted ³H-LEV (10 µM to 0.0781 µM) in triplicate. Specific binding of ³H-LEV to SV2AΔ64-mVenus or mutant membranes was calculated by subtracting nonspecific binding from the total binding. Nonspecific binding was measured by residual binding observed in the presence of 10 mM LEV.

For EC₅₀ calculations, SV2AΔ64-mVenus or single residue mutant enriched membranes were incubated for 45 min on ice with ³H-LEV at 3.3 µM and a series of UCB1244283 ligand concentrations at a final concentration of 30 µM, 15 µM, 1 µM, 100 nM, 10 nM, and 1 nM, respectively. Specific binding was calculated by subtracting nonspecific binding from the total binding, which was measured by residual binding observed in the presence of unlabeled 1 mM PSL. For all experiments, incubated samples were filtered on glass microfiber filters (Cytiva, Cat No. 1822-6580, 25mm) presoaked in 1% polyethylenimine (PEI), followed by a very rapid washing with 20 ml ice-cold TBS_150. Filters were dried and were transferred into a 24-well plate (PerkinElmer, Cat No. 1450-402) followed by the addition of 500 µl of scintillant to each well for measuring radioligand binding. Counts were measured using a Microbeta-2 scintillation counter (PerkinElmer). The K_d and EC₅₀ values were calculated by GraphPad Prism 10 software using nonlinear fitting for one site-specific binding and log (agonist) vs response (three parameters) equation, respectively.

Supplementary Material

Appendix 01 (PDF)

pnas.2510239122.sapp.pdf^{(10.8MB, pdf)}

Acknowledgments

We thank James Conway and the Pittsburgh Center for Cryo-EM for their support. This work was supported by a Young Investigator Grant from the Brain and Behavior Research Foundation (Grant no. 30153) and by the University of Pittsburgh Competitive Medical Research Fund and Startup Funding to J.A.C. Cryo-EM at the University of Pittsburgh was supported by NIH Grants no. S10 OD025009 and no. S10 OD019995.

Author contributions

A.M., P.S.H., and J.A.C. designed research; A.M., M.F.M., M.L., C.W., M.G., P.S.H., and J.A.C. performed research; L.P., A.H., and P.S.H. contributed new reagents/analytic tools; A.M., M.F.M., M.L., P.S.H., and J.A.C. analyzed data; and A.M., M.F.M., L.P., A.H., M.L., C.W., M.G., P.S.H., and J.A.C. wrote the paper.

Competing interests

L.P., A.H., M.L., C.W., M.G., P.S.H. are employees of UCB Pharma. L.P., A.H., C.W., M.G., P.S.H. hold shares in UCB Pharma. The other authors declare no competing interests.

Footnotes

This article is a PNAS Direct Submission.

Contributor Information

Peter S. Horanyi, Email: Peter.Horanyi@ucb.com.

Jonathan A. Coleman, Email: coleman1@pitt.edu.

Data, Materials, and Software Availability

The cryo-EM maps and models has been deposited under the following accession code: SV2A-BRV-UCB1244283: PDB 9NTC (56), EMD-49758 (57); SV2A-LEV-UCB1244283: EMD-71502 (58); SV2A-UCB-J: PDB 9PCB (59), EMD-71501 (60); SV2A-UCB7361: PDB 9PC9 (61), EMD-71500 (62).

Supporting Information

References

1.Chang B. S., Lowenstein D. H., Epilepsy. N. Engl. J. Med. 349, 1257–1266 (2003). [DOI] [PubMed] [Google Scholar]
2.Tian N., et al. , Mortality and mortality disparities among people with epilepsy in the United States, 2011–2021. Epilepsy Behav. 155, 109770 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Klein P., Diaz A., Gasalla T., Whitesides J., A review of the pharmacology and clinical efficacy of brivaracetam. Clin. Pharmacol. 10, 1–22 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Lynch B. A., et al. , The synaptic vesicle protein SV2A is the binding site for the antiepileptic drug levetiracetam. Proc. Natl. Acad. Sci. U.S.A. 101, 9861–9866 (2004). [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Contreras-Garcia I. J., et al. , Levetiracetam mechanisms of action: From molecules to systems. Pharmaceuticals (Basel) 15, 475 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Loscher W., Gillard M., Sands Z. A., Kaminski R. M., Klitgaard H., Synaptic vesicle glycoprotein 2A ligands in the treatment of epilepsy and beyond. CNS Drugs 30, 1055–1077 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Niespodziany I., et al. , Brivaracetam does not modulate ionotropic channels activated by glutamate, gamma-aminobutyric acid, and glycine in hippocampal neurons. Epilepsia 58, e157–e161 (2017). [DOI] [PubMed] [Google Scholar]
8.Klitgaard H., et al. , Brivaracetam: Rationale for discovery and preclinical profile of a selective SV2A ligand for epilepsy treatment. Epilepsia 57, 538–548 (2016). [DOI] [PubMed] [Google Scholar]
9.Bajjalieh S. M., Peterson K., Shinghal R., Scheller R. H., SV2, a brain synaptic vesicle protein homologous to bacterial transporters. Science 257, 1271–1273 (1992). [DOI] [PubMed] [Google Scholar]
10.Feany M. B., Lee S., Edwards R. H., Buckley K. M., The synaptic vesicle protein SV2 is a novel type of transmembrane transporter. Cell 70, 861–867 (1992). [DOI] [PubMed] [Google Scholar]
11.Gingrich J. A., et al. , Identification, characterization, and molecular cloning of a novel transporter-like protein localized to the central nervous system. FEBS Lett. 312, 115–122 (1992). [DOI] [PubMed] [Google Scholar]
12.Buckley K., Kelly R. B., Identification of a transmembrane glycoprotein specific for secretory vesicles of neural and endocrine cells. J. Cell Biol. 100, 1284–1294 (1985). [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Bajjalieh S. M., Peterson K., Linial M., Scheller R. H., Brain contains two forms of synaptic vesicle protein 2. Proc. Natl. Acad. Sci. U.S.A. 90, 2150–2154 (1993). [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Wu P. P., Cao B. R., Tian F. Y., Gao Z. B., Development of SV2A ligands for epilepsy treatment: A review of levetiracetam, brivaracetam, and padsevonil. Neurosci. Bull. 40, 594–608 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Wood M., et al. , Pharmacological profile of the novel antiepileptic drug candidate Padsevonil: Interactions with synaptic vesicle 2 proteins and the GABA(A) receptor. J. Pharmacol. Exp. Ther. 372, 1–10 (2020). [DOI] [PubMed] [Google Scholar]
16.Rademacher M., et al. , Efficacy and safety of adjunctive padsevonil in adults with drug-resistant focal epilepsy: Results from two double-blind, randomized, placebo-controlled trials. Epilepsia Open 7, 758–770 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Dong M., et al. , SV2 is the protein receptor for botulinum neurotoxin A. Science 312, 592–596 (2006). [DOI] [PubMed] [Google Scholar]
18.Yeh F. L., et al. , SV2 mediates entry of tetanus neurotoxin into central neurons. PLoS Pathog. 6, e1001207 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Liu Z., et al. , Structural basis for botulinum neurotoxin E recognition of synaptic vesicle protein 2. Nat. Commun. 14, 2338 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Biello A., Zhu B., “Botulinum toxin treatment of the upper face” in StatPearls (Treasure Island (FL), 2025). [PubMed] [Google Scholar]
21.Mittal A., et al. , Structures of synaptic vesicle protein 2A and 2B bound to anticonvulsants. Nat. Struct. Mol. Biol. 31, 1964–1974 (2024). [DOI] [PubMed] [Google Scholar]
22.Yamagata A., et al. , Structural basis for antiepileptic drugs and botulinum neurotoxin recognition of SV2A. Nat. Commun. 15, 3027 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Liu S., et al. , Recognition of antiepileptic brivaracetam by synaptic vesicle protein 2A. Cell Discov. 10, 56 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Khanppnavar B., Leka O., Pal S. K., Korkhov V. M., Kammerer R. A., Cryo-EM structure of the botulinum neurotoxin A/SV2B complex and its implications for translocation. Nat. Commun. 16, 1224 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Bradberry M. M., Chapman E. R., Structural insights into SV2A and the mechanism of racetam anticonvulsants. Nat. Struct. Mol. Biol. 31, 1818–1820 (2024). [DOI] [PubMed] [Google Scholar]
26.Schenck S., Laeremans T., Steyaert J., Brunner J. D., Structures of native SV2A reveal the binding mode for tetanus neurotoxin and anti-epileptic racetams. Nat. Commun. 16, 4172 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Nowack A., Yao J., Custer K. L., Bajjalieh S. M., SV2 regulates neurotransmitter release via multiple mechanisms. Am. J. Physiol. Cell Physiol. 299, C960–967 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Daniels V., Wood M., Leclercq K., Kaminski R. M., Gillard M., Modulation of the conformational state of the SV2A protein by an allosteric mechanism as evidenced by ligand binding assays. Br. J. Pharmacol. 169, 1091–1101 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Wood M. D., Gillard M., Evidence for a differential interaction of brivaracetam and levetiracetam with the synaptic vesicle 2A protein. Epilepsia 58, 255–262 (2017). [DOI] [PubMed] [Google Scholar]
30.Wood M. D., Sands Z. A., Vandenplas C., Gillard M., Further evidence for a differential interaction of brivaracetam and levetiracetam with the synaptic vesicle 2A protein. Epilepsia 59, e147–e151 (2018). [DOI] [PubMed] [Google Scholar]
31.Nabulsi N. B., et al. , Synthesis and preclinical evaluation of 11C-UCB-J as a PET tracer for imaging the synaptic vesicle glycoprotein 2A in the brain. J. Nucl. Med. 57, 777–784 (2016). [DOI] [PubMed] [Google Scholar]
32.Lee J., et al. , Exploring the interaction of SV2A with racetams using homology modelling, molecular dynamics and site-directed mutagenesis. PLoS One 10, e0116589 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Shi J., et al. , Combining modelling and mutagenesis studies of synaptic vesicle protein 2A to identify a series of residues involved in racetam binding. Biochem. Soc. Trans. 39, 1341–1347 (2011). [DOI] [PubMed] [Google Scholar]
34.Jumper J., et al. , Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Drew D., North R. A., Nagarathinam K., Tanabe M., Structures and general transport mechanisms by the major facilitator superfamily (MFS). Chem. Rev. 121, 5289–5335 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Coleman J. A., Green E. M., Gouaux E., X-ray structures and mechanism of the human serotonin transporter. Nature 532, 334–339 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Bartholome O., et al. , Puzzling out synaptic vesicle 2 family members functions. Front. Mol. Neurosci. 10, 148 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Madeo M., Kovacs A. D., Pearce D. A., The human synaptic vesicle protein, SV2A, functions as a galactose transporter in Saccharomyces cerevisiae. J. Biol. Chem. 289, 33066–33071 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Rossi R., Arjmand S., Baerentzen S. L., Gjedde A., Landau A. M., Synaptic vesicle glycoprotein 2A: Features and functions. Front. Neurosci. 16, 864514 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Ciruelas K., Marcotulli D., Bajjalieh S. M., Synaptic vesicle protein 2: A multi-faceted regulator of secretion. Semin. Cell Dev. Biol. 95, 130–141 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Qian M., Sun Z., Chen X., Calenbergh S., Study of G protein-coupled receptors dimerization: From bivalent ligands to drug-like small molecules. Bioorg. Chem. 140, 106809 (2023). [DOI] [PubMed] [Google Scholar]
42.Andersen J., et al. , Interrogating the molecular basis for substrate recognition in serotonin and dopamine transporters with high-affinity substrate-based bivalent ligands. ACS Chem. Neurosci. 7, 1406–1417 (2016). [DOI] [PubMed] [Google Scholar]
43.Lane J. R., Sexton P. M., Christopoulos A., Bridging the gap: Bitopic ligands of G-protein-coupled receptors. Trends Pharmacol. Sci. 34, 59–66 (2013). [DOI] [PubMed] [Google Scholar]
44.Valant C., Robert Lane J., Sexton P. M., Christopoulos A., The best of both worlds? Bitopic orthosteric/allosteric ligands of G protein-coupled receptors. Annu. Rev. Pharmacol. Toxicol. 52, 153–178 (2012). [DOI] [PubMed] [Google Scholar]
45.Arroyo-Urea S., et al. , A bitopic agonist bound to the dopamine 3 receptor reveals a selectivity site. Nat. Commun. 15, 7759 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Goehring A., et al. , Screening and large-scale expression of membrane proteins in mammalian cells for structural studies. Nat. Protoc. 9, 2574–2585 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Hattori M., Hibbs R. E., Gouaux E., A fluorescence-detection size-exclusion chromatography-based thermostability assay for membrane protein precrystallization screening. Structure 20, 1293–1299 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Kawate T., Gouaux E., Fluorescence-detection size-exclusion chromatography for precrystallization screening of integral membrane proteins. Structure 14, 673–681 (2006). [DOI] [PubMed] [Google Scholar]
49.Schimek C., et al. , Extraction of recombinant periplasmic proteins under industrially relevant process conditions: Selectivity and yield strongly depend on protein titer and methodology. Biotechnol. Prog. 36, e2999 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Punjani A., Rubinstein J. L., Fleet D. J., Brubaker M. A., CryoSPARC: Algorithms for rapid unsupervised cryo-EM structure determination. Nat. Methods 14, 290–296 (2017). [DOI] [PubMed] [Google Scholar]
51.Zivanov J., et al. , New tools for automated high-resolution cryo-EM structure determination in RELION-3. eLife 7, e42166 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Scheres S. H., RELION: Implementation of a Bayesian approach to cryo-EM structure determination. J. Struct. Biol. 180, 519–530 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Casanal A., Lohkamp B., Emsley P., Current developments in Coot for macromolecular model building of electron cryo-microscopy and crystallographic data. Protein Sci. 29, 1069–1078 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Emsley P., Lohkamp B., Scott W. G., Cowtan K., Features and development of coot. Acta Crystallogr. D Biol. Crystallogr. 66, 486–501 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Liebschner D., et al. , Macromolecular structure determination using X-rays, neutrons and electrons: Recent developments in Phenix. Acta Crystallogr. D Struct. Biol. 75, 861–877 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with BRV and UCB1244283. Protein Data Bank. https://www.rcsb.org/structure/9NTC. Deposited 19 March 2025.
57.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with BRV and UCB1244283. Electron Microscopy Data Bank. https://www.ebi.ac.uk/emdb/EMD-49758. Deposited 19 March 2025.
58.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with LEV and UCB1244283. Electron Microscopy Data Bank. https://www.ebi.ac.uk/emdb/EMD-71502. Deposited 30 June 2025.
59.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with UCB-J. Protein Data Bank. https://www.rcsb.org/structure/9PCB. Deposited 30 June 2025.
60.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with UCB-J. Electron Microscopy Data Bank. https://www.ebi.ac.uk/emdb/EMD-71501. Deposited 30 June 2025.
61.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with UCB7361. Protein Data Bank. https://www.rcsb.org/structure/9PC9. Deposited 30 June 2025.
62.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with UCB7361. Electron Microscopy Data Bank. https://www.ebi.ac.uk/emdb/EMD-71500. Deposited 30 June 2025.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix 01 (PDF)

pnas.2510239122.sapp.pdf^{(10.8MB, pdf)}

Data Availability Statement

[r1] 1.Chang B. S., Lowenstein D. H., Epilepsy. N. Engl. J. Med. 349, 1257–1266 (2003). [DOI] [PubMed] [Google Scholar]

[r2] 2.Tian N., et al. , Mortality and mortality disparities among people with epilepsy in the United States, 2011–2021. Epilepsy Behav. 155, 109770 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r3] 3.Klein P., Diaz A., Gasalla T., Whitesides J., A review of the pharmacology and clinical efficacy of brivaracetam. Clin. Pharmacol. 10, 1–22 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r4] 4.Lynch B. A., et al. , The synaptic vesicle protein SV2A is the binding site for the antiepileptic drug levetiracetam. Proc. Natl. Acad. Sci. U.S.A. 101, 9861–9866 (2004). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r5] 5.Contreras-Garcia I. J., et al. , Levetiracetam mechanisms of action: From molecules to systems. Pharmaceuticals (Basel) 15, 475 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r6] 6.Loscher W., Gillard M., Sands Z. A., Kaminski R. M., Klitgaard H., Synaptic vesicle glycoprotein 2A ligands in the treatment of epilepsy and beyond. CNS Drugs 30, 1055–1077 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r7] 7.Niespodziany I., et al. , Brivaracetam does not modulate ionotropic channels activated by glutamate, gamma-aminobutyric acid, and glycine in hippocampal neurons. Epilepsia 58, e157–e161 (2017). [DOI] [PubMed] [Google Scholar]

[r8] 8.Klitgaard H., et al. , Brivaracetam: Rationale for discovery and preclinical profile of a selective SV2A ligand for epilepsy treatment. Epilepsia 57, 538–548 (2016). [DOI] [PubMed] [Google Scholar]

[r9] 9.Bajjalieh S. M., Peterson K., Shinghal R., Scheller R. H., SV2, a brain synaptic vesicle protein homologous to bacterial transporters. Science 257, 1271–1273 (1992). [DOI] [PubMed] [Google Scholar]

[r10] 10.Feany M. B., Lee S., Edwards R. H., Buckley K. M., The synaptic vesicle protein SV2 is a novel type of transmembrane transporter. Cell 70, 861–867 (1992). [DOI] [PubMed] [Google Scholar]

[r11] 11.Gingrich J. A., et al. , Identification, characterization, and molecular cloning of a novel transporter-like protein localized to the central nervous system. FEBS Lett. 312, 115–122 (1992). [DOI] [PubMed] [Google Scholar]

[r12] 12.Buckley K., Kelly R. B., Identification of a transmembrane glycoprotein specific for secretory vesicles of neural and endocrine cells. J. Cell Biol. 100, 1284–1294 (1985). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r13] 13.Bajjalieh S. M., Peterson K., Linial M., Scheller R. H., Brain contains two forms of synaptic vesicle protein 2. Proc. Natl. Acad. Sci. U.S.A. 90, 2150–2154 (1993). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r14] 14.Wu P. P., Cao B. R., Tian F. Y., Gao Z. B., Development of SV2A ligands for epilepsy treatment: A review of levetiracetam, brivaracetam, and padsevonil. Neurosci. Bull. 40, 594–608 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r15] 15.Wood M., et al. , Pharmacological profile of the novel antiepileptic drug candidate Padsevonil: Interactions with synaptic vesicle 2 proteins and the GABA(A) receptor. J. Pharmacol. Exp. Ther. 372, 1–10 (2020). [DOI] [PubMed] [Google Scholar]

[r16] 16.Rademacher M., et al. , Efficacy and safety of adjunctive padsevonil in adults with drug-resistant focal epilepsy: Results from two double-blind, randomized, placebo-controlled trials. Epilepsia Open 7, 758–770 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r17] 17.Dong M., et al. , SV2 is the protein receptor for botulinum neurotoxin A. Science 312, 592–596 (2006). [DOI] [PubMed] [Google Scholar]

[r18] 18.Yeh F. L., et al. , SV2 mediates entry of tetanus neurotoxin into central neurons. PLoS Pathog. 6, e1001207 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r19] 19.Liu Z., et al. , Structural basis for botulinum neurotoxin E recognition of synaptic vesicle protein 2. Nat. Commun. 14, 2338 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r20] 20.Biello A., Zhu B., “Botulinum toxin treatment of the upper face” in StatPearls (Treasure Island (FL), 2025). [PubMed] [Google Scholar]

[r21] 21.Mittal A., et al. , Structures of synaptic vesicle protein 2A and 2B bound to anticonvulsants. Nat. Struct. Mol. Biol. 31, 1964–1974 (2024). [DOI] [PubMed] [Google Scholar]

[r22] 22.Yamagata A., et al. , Structural basis for antiepileptic drugs and botulinum neurotoxin recognition of SV2A. Nat. Commun. 15, 3027 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r23] 23.Liu S., et al. , Recognition of antiepileptic brivaracetam by synaptic vesicle protein 2A. Cell Discov. 10, 56 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r24] 24.Khanppnavar B., Leka O., Pal S. K., Korkhov V. M., Kammerer R. A., Cryo-EM structure of the botulinum neurotoxin A/SV2B complex and its implications for translocation. Nat. Commun. 16, 1224 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r25] 25.Bradberry M. M., Chapman E. R., Structural insights into SV2A and the mechanism of racetam anticonvulsants. Nat. Struct. Mol. Biol. 31, 1818–1820 (2024). [DOI] [PubMed] [Google Scholar]

[r26] 26.Schenck S., Laeremans T., Steyaert J., Brunner J. D., Structures of native SV2A reveal the binding mode for tetanus neurotoxin and anti-epileptic racetams. Nat. Commun. 16, 4172 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r27] 27.Nowack A., Yao J., Custer K. L., Bajjalieh S. M., SV2 regulates neurotransmitter release via multiple mechanisms. Am. J. Physiol. Cell Physiol. 299, C960–967 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r28] 28.Daniels V., Wood M., Leclercq K., Kaminski R. M., Gillard M., Modulation of the conformational state of the SV2A protein by an allosteric mechanism as evidenced by ligand binding assays. Br. J. Pharmacol. 169, 1091–1101 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r29] 29.Wood M. D., Gillard M., Evidence for a differential interaction of brivaracetam and levetiracetam with the synaptic vesicle 2A protein. Epilepsia 58, 255–262 (2017). [DOI] [PubMed] [Google Scholar]

[r30] 30.Wood M. D., Sands Z. A., Vandenplas C., Gillard M., Further evidence for a differential interaction of brivaracetam and levetiracetam with the synaptic vesicle 2A protein. Epilepsia 59, e147–e151 (2018). [DOI] [PubMed] [Google Scholar]

[r31] 31.Nabulsi N. B., et al. , Synthesis and preclinical evaluation of 11C-UCB-J as a PET tracer for imaging the synaptic vesicle glycoprotein 2A in the brain. J. Nucl. Med. 57, 777–784 (2016). [DOI] [PubMed] [Google Scholar]

[r32] 32.Lee J., et al. , Exploring the interaction of SV2A with racetams using homology modelling, molecular dynamics and site-directed mutagenesis. PLoS One 10, e0116589 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r33] 33.Shi J., et al. , Combining modelling and mutagenesis studies of synaptic vesicle protein 2A to identify a series of residues involved in racetam binding. Biochem. Soc. Trans. 39, 1341–1347 (2011). [DOI] [PubMed] [Google Scholar]

[r34] 34.Jumper J., et al. , Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r35] 35.Drew D., North R. A., Nagarathinam K., Tanabe M., Structures and general transport mechanisms by the major facilitator superfamily (MFS). Chem. Rev. 121, 5289–5335 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r36] 36.Coleman J. A., Green E. M., Gouaux E., X-ray structures and mechanism of the human serotonin transporter. Nature 532, 334–339 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r37] 37.Bartholome O., et al. , Puzzling out synaptic vesicle 2 family members functions. Front. Mol. Neurosci. 10, 148 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r38] 38.Madeo M., Kovacs A. D., Pearce D. A., The human synaptic vesicle protein, SV2A, functions as a galactose transporter in Saccharomyces cerevisiae. J. Biol. Chem. 289, 33066–33071 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r39] 39.Rossi R., Arjmand S., Baerentzen S. L., Gjedde A., Landau A. M., Synaptic vesicle glycoprotein 2A: Features and functions. Front. Neurosci. 16, 864514 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r40] 40.Ciruelas K., Marcotulli D., Bajjalieh S. M., Synaptic vesicle protein 2: A multi-faceted regulator of secretion. Semin. Cell Dev. Biol. 95, 130–141 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r41] 41.Qian M., Sun Z., Chen X., Calenbergh S., Study of G protein-coupled receptors dimerization: From bivalent ligands to drug-like small molecules. Bioorg. Chem. 140, 106809 (2023). [DOI] [PubMed] [Google Scholar]

[r42] 42.Andersen J., et al. , Interrogating the molecular basis for substrate recognition in serotonin and dopamine transporters with high-affinity substrate-based bivalent ligands. ACS Chem. Neurosci. 7, 1406–1417 (2016). [DOI] [PubMed] [Google Scholar]

[r43] 43.Lane J. R., Sexton P. M., Christopoulos A., Bridging the gap: Bitopic ligands of G-protein-coupled receptors. Trends Pharmacol. Sci. 34, 59–66 (2013). [DOI] [PubMed] [Google Scholar]

[r44] 44.Valant C., Robert Lane J., Sexton P. M., Christopoulos A., The best of both worlds? Bitopic orthosteric/allosteric ligands of G protein-coupled receptors. Annu. Rev. Pharmacol. Toxicol. 52, 153–178 (2012). [DOI] [PubMed] [Google Scholar]

[r45] 45.Arroyo-Urea S., et al. , A bitopic agonist bound to the dopamine 3 receptor reveals a selectivity site. Nat. Commun. 15, 7759 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r46] 46.Goehring A., et al. , Screening and large-scale expression of membrane proteins in mammalian cells for structural studies. Nat. Protoc. 9, 2574–2585 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r47] 47.Hattori M., Hibbs R. E., Gouaux E., A fluorescence-detection size-exclusion chromatography-based thermostability assay for membrane protein precrystallization screening. Structure 20, 1293–1299 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r48] 48.Kawate T., Gouaux E., Fluorescence-detection size-exclusion chromatography for precrystallization screening of integral membrane proteins. Structure 14, 673–681 (2006). [DOI] [PubMed] [Google Scholar]

[r49] 49.Schimek C., et al. , Extraction of recombinant periplasmic proteins under industrially relevant process conditions: Selectivity and yield strongly depend on protein titer and methodology. Biotechnol. Prog. 36, e2999 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r50] 50.Punjani A., Rubinstein J. L., Fleet D. J., Brubaker M. A., CryoSPARC: Algorithms for rapid unsupervised cryo-EM structure determination. Nat. Methods 14, 290–296 (2017). [DOI] [PubMed] [Google Scholar]

[r51] 51.Zivanov J., et al. , New tools for automated high-resolution cryo-EM structure determination in RELION-3. eLife 7, e42166 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r52] 52.Scheres S. H., RELION: Implementation of a Bayesian approach to cryo-EM structure determination. J. Struct. Biol. 180, 519–530 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r53] 53.Casanal A., Lohkamp B., Emsley P., Current developments in Coot for macromolecular model building of electron cryo-microscopy and crystallographic data. Protein Sci. 29, 1069–1078 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r54] 54.Emsley P., Lohkamp B., Scott W. G., Cowtan K., Features and development of coot. Acta Crystallogr. D Biol. Crystallogr. 66, 486–501 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r55] 55.Liebschner D., et al. , Macromolecular structure determination using X-rays, neutrons and electrons: Recent developments in Phenix. Acta Crystallogr. D Struct. Biol. 75, 861–877 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r56] 56.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with BRV and UCB1244283. Protein Data Bank. https://www.rcsb.org/structure/9NTC. Deposited 19 March 2025.

[r57] 57.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with BRV and UCB1244283. Electron Microscopy Data Bank. https://www.ebi.ac.uk/emdb/EMD-49758. Deposited 19 March 2025.

[r58] 58.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with LEV and UCB1244283. Electron Microscopy Data Bank. https://www.ebi.ac.uk/emdb/EMD-71502. Deposited 30 June 2025.

[r59] 59.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with UCB-J. Protein Data Bank. https://www.rcsb.org/structure/9PCB. Deposited 30 June 2025.

[r60] 60.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with UCB-J. Electron Microscopy Data Bank. https://www.ebi.ac.uk/emdb/EMD-71501. Deposited 30 June 2025.

[r61] 61.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with UCB7361. Protein Data Bank. https://www.rcsb.org/structure/9PC9. Deposited 30 June 2025.

[r62] 62.Mittal A., Martin M. F., Ledecq M., Horanyi P. S., Coleman J. A., cryo-EM structure of SV2A bound with UCB7361. Electron Microscopy Data Bank. https://www.ebi.ac.uk/emdb/EMD-71500. Deposited 30 June 2025.

PERMALINK

Mechanisms underlying allosteric modulation of antiseizure medication binding to synaptic vesicle protein 2A (SV2A)

Anshumali Mittal

Matthew F Martin

Laurent Provins

Adrian Hall

Marie Ledecq

Christian Wolff

Michel Gillard

Peter S Horanyi

Jonathan A Coleman

Significance

Abstract

Fig. 1.

Results

UCB1244283 Modulates the SV2A-LEV Complex Through an Allosteric Mechanism.

Cryo-EM of the SV2A-BRV-UCB1244283 and SV2A-LEV-UCB1244283 Complexes.

Fig. 2.

Molecular Architecture of the UCB1244283 and BRV Binding Sites.

Analysis of the Allosteric Site Among SV2 Family Members.

Fig. 3.

Molecular Basis of Binding and Selectivity of UCB1244283.

SV2A-UCB-J and SV2A-UCB7631 Complexes and Analysis of the Molecular Architecture of Binding Sites.

Fig. 4.

Ligand-Induced Remodeling of the SV2A TMD Conformation.

Table 1.

Fig. 5.

Discussion

Fig. 6.

Materials and Methods

Small-Molecule Ligands.

Construct Design and Cloning.

Fluorescence-Detection Size-Exclusion Chromatography.

Expression and Purification of 8783 and 8667-Nbs.

Expression and Purification of SV2A-LEV-UCB1244283-8783-Nb, -BRV-UCB1244283-8783-Nb, -UCB-J-8667-Nb, -UCB7361-8783-Nb Complexes.

Cryo-EM Sample Preparation and Data Collection.

Cryo-EM data processing and model building.

Thermostability Assays.

Radioligand Binding Assays.

Supplementary Material

Acknowledgments

Author contributions

Competing interests

Footnotes

Contributor Information

Data, Materials, and Software Availability

Supporting Information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases