Structural basis of vilazodone dual binding mode to the serotonin transporter

Kalenderoglou, Iris E.; Nygaard, Andreas; Vogt, Caleb D.; Turaev, Anton; Pape, Tillmann; Adams, Nathan B. P.; Newman, Amy Hauck; Loland, Claus J.

doi:10.1038/s41467-025-65202-3

Download PDF

Article
Open access
Published: 18 November 2025

Structural basis of vilazodone dual binding mode to the serotonin transporter

Nature Communications volume 16, Article number: 10119 (2025) Cite this article

3447 Accesses
2 Citations
6 Altmetric
Metrics details

Subjects

Abstract

The serotonin transporter (SERT) plays a pivotal role in regulating serotonin (5-HT) signaling and is a key target in the treatment of psychiatric disorders. SERT has a binding site (S1) for 5-HT that also serves as a high-affinity binding site for antidepressants. The antidepressant vilazodone has been shown to inhibit SERT by binding to an allosteric site. Here, we present the cryo-EM structure of SERT with vilazodone bound to the S1 site and extending towards the allosteric site. We systematically dissect the vilazodone molecule into fragments and find that the terminal indole ring is the key determinant of its high affinity to SERT. Further, unlike typical Na⁺-dependent SERT-selective antidepressants, vilazodone exhibits a dissociation constant (K_D) for purified SERT in the nanomolar range both in the presence and absence of Na⁺. We substantiate this binding mode by exploring the conformational impact of vilazodone binding to SERT using site-specific insertion of the fluorescent non-canonical amino acid Anap. Our results offer molecular insight into the distinct pharmacological profile of a clinically used polymodal antidepressant.

The antidepressant drug vilazodone is an allosteric inhibitor of the serotonin transporter

Article Open access 20 August 2021

Fluorescent non-canonical amino acid provides insight into the human serotonin transporter

Article Open access 27 October 2024

Ligand coupling mechanism of the human serotonin transporter differentiates substrates from inhibitors

Article Open access 10 January 2024

Introduction

Depression is one of the most prevalent neuropsychiatric disorders, affecting approximately 4.4% of the global population¹. While SSRIs^2,3,4,5,6 have superseded older antidepressants, they are not devoid of side effects⁷, which significantly reduces patients’ adherence to the treatment^5,8,9. Vilazodone (Viibryd^®), characterized as a selective serotonin (5-HT) reuptake inhibitor (SSRI) and a 5-HT_1A receptor partial agonist (SPARI)^10,11, offers a more favorable profile with fewer instances of sexual dysfunction, weight gain, and emotional blunting^{12,13,14,15,16}. It also exhibits a faster onset of action and higher efficacy^{12,13,14,15,16}, making it a compelling antidepressant treatment.

The inhibition of SERT is a common therapeutic strategy for the treatment of neuropsychiatric disorders, including depression, anxiety, and post-traumatic stress disorder. SERT plays a crucial role in regulating the availability of 5-HT, a neurotransmitter involved in numerous brain functions, such as mood, sleep, appetite, and memory¹⁷. Located in the presynaptic neurons, SERT facilitates the reuptake of released 5-HT, thereby terminating neuronal signaling^17,18,19. SERT is a Neurotransmitter Sodium Symporter (NSS) and belongs to the solute carrier (SLC) 6 family²⁰, which is characterized by a common structural fold first found in the bacterial amino acid transporter LeuT²¹. This fold consists of two topological repeats, each of five transmembrane helices (TMs)^19,21 that facilitate substrate transport through an alternating access mechanism^22,23,24. It has been suggested that SERT regulates 5-HT availability through occupation of two substrate binding sites (S1 and S2)²⁵. Specifically, S1 is composed of residues within TM1, TM3, TM6, and TM8, whereas the S2 site is situated in the extracellular vestibule recruiting residues from TM10 and TM12²⁵. The functional role of the S2 site remains unknown. Electrophysiological studies have suggested that only one 5-HT molecule is transported along with one Na⁺ and one Cl^- ion per transport cycle²⁶. After intracellular substrate release, the return-step is then believed to be facilitated by a counter-transport of either K⁺ or H⁺²⁶, providing an overall electroneutral 5-HT transport cycle²⁷.

The S1 site in SERT, accommodating substrates and typical inhibitors, can be organized into three compartments: subsite A (TM1, TM6, and TM8) primarily engages with the positively charged amino group of ligands (D98 for SERT), subsite B (TM3 and TM8) is crucial for hydrophobic coordination of the ligand’s aromatic group, and subsite C (TM6 and TM10) is proximal to the extracellular vestibule²⁸. A key feature in the substrate translocation process is the formation of an electrostatic interaction between R104 and E493 located in the extracellular vestibule. This interaction is thought to be essential for the gating mechanism of SERT, which regulates the entry of solutes towards the S1 site^29,30. Furthermore, close to the S1 site are two Na⁺ binding sites^31,32. Studies have demonstrated that a Na⁺-rich environment promotes outward-open states of NSS proteins, facilitating extracellular ligand binding^{33,34,35,36,37}. In contrast, in the presence of K⁺ ions, which are believed to be counter-transported^27,38, the equilibrium shifts towards an inward-facing conformation^25,39,40. Even though a K⁺ gradient is not required for 5-HT uptake, it will increase the transport rate⁴¹. All investigated SSRIs exhibit Na⁺-dependent binding and seem to stabilize an outward-open SERT conformation^{25,31,42,43,44,45,46}.

The extracellular vestibule of SERT not only serves as an entry pathway for 5-HT and inhibitors binding to the S1 site but also possesses an allosteric binding site^{31,45,47,48,49}. Our studies on vilazodone⁴⁵ showed that it inhibited the dissociation of pre-bound [³H]imipramine with an allosteric potency of ~14 nM. The concomitant cryo-EM structure of the SERT-imipramine-vilazodone complex revealed that SERT was stabilized in the outward-open state with imipramine bound to the S1 site and vilazodone bound to an allosteric site, residing partly in the extracellular vestibule and partly within a cryptic site between TM11 and TM12. This aligned with studies using the substituted cysteine accessibility method⁵⁰, in which vilazodone binding was suggested to stabilize an outward-open and inward-closed conformation⁵¹. The molecular interactions of vilazodone with SERT were also explored through modelling⁵², predicting vilazodone to bind as an orthosteric ligand with the benzofuran ring interacting with the S1 site residues and the indole group extending towards the extracellular vestibule. Interestingly, vilazodone’s inhibition of [³H]5-HT uptake revealed a Hill coefficient of ~1.8⁴⁵, which suggested cooperativity between multiple inhibitor binding sites⁵³. This prompted us to search for additional binding modes, which might be revealed in the absence of other supporting ligands, such as imipramine.

In the present study, we employ molecular pharmacology, cryo-EM, and fluorescence-based techniques to elucidate the molecular mechanism behind vilazodone binding to SERT. The cryo-EM structure shows one molecule of vilazodone bound at the S1 site extending towards the extracellular vestibule, thus revealing a different pose from the one determined previously. Utilizing spectral shift assays of purified and fluorescently labelled SERT, we demonstrate a vilazodone affinity in the low nM range, in the presence of Na⁺, K⁺ or the non-interacting cation N-methyl-D-glucamine (NMDG⁺). To further dissect the interaction dynamics, we systematically deconstructed the vilazodone molecule into its benzofuran and indole fragments. The indole analogues emerged as highly potent 5-HT reuptake inhibitors, whereas the benzofuran ligands exhibited low potency in this assay. Moreover, our engineered SERT construct labelled with the fluorescent non-canonical amino acid (ncAA) Anap⁵⁴, suggested that vilazodone binds in either Na⁺, K⁺ or NMDG⁺ environments and indicated that vilazodone induces distinct SERT conformations depending on the ionic conditions. Overall, this work uncovers the ability of a prescribed antidepressant to adopt distinct high-affinity binding modes in SERT depending on the bound ion or ligand at the S1 site. These findings not only decode the intricate mechanism of SERT inhibition by vilazodone but also provide valuable insights into its unique pharmacological properties.

Results

Mutations in the allosteric site do not decrease the inhibition potency of vilazodone

In our previous work⁴⁵, we solved the cryo-EM structure of SERT in complex with imipramine and vilazodone, revealing the allosteric inhibition mechanism of vilazodone. Notably, the inhibition of [³H]5-HT uptake by vilazodone demonstrated a Hill coefficient of 1.8⁴⁵. Given that this could be the result of cooperative binding of multiple vilazodone molecules, we aimed to further investigate its inhibition mechanism.

We hypothesized that if vilazodone binds to the allosteric site identified in the previous cryo-EM structure, then mutating the interacting residues would reduce its potency to inhibit [³H]5-HT transport. Accordingly, we mutated 9 residues (R104K, Q332A, F335A, E493N, E494Q, Y495A, F556A, P561G, and Y579A) predicted to be important for vilazodone binding (Fig. 1a). We expressed the SERT mutants in COS-7 cells and studied their effect on vilazodone binding potency (Fig. 1b, Supplementary Fig. 1, Supplementary Table 1). The IC₅₀ values measured in homologous competition assays of [³H]5-HT transport were similar to that of SERT WT (Supplementary Fig. 1). Vilazodone inhibited [³H]5-HT transport by SERT WT with an IC₅₀ around 1 nM, which agrees with our previous findings⁴⁵. Surprisingly, all mutants showed no significant change in the IC₅₀ values for vilazodone (p > 0.9999), ranging between 0.56-2.47 nM (Fig. 1b). Thus, these findings did not correlate with our previous study, where the same mutants had a pronounced effect on the allosteric potency of vilazodone⁴⁵. However, these previously reported data were based on the ability of vilazodone to allosterically inhibit the dissociation of S1-bound [³H]imipramine. We reasoned that the presence of [³H]imipramine could potentially prevent vilazodone from assuming multiple binding poses and, therefore, the tested SERT mutants used here were affecting only one of the possible binding modes.

**Fig. 1: Effect of mutations in the S1 and the allosteric site on vilazodone affinity.**

Molecular dynamics (MD) simulations have been used to assess how vilazodone binds in the absence of an S1 ligand⁵². The results suggested a different binding pose of vilazodone compared to that in the SERT-imipramine-vilazodone structure⁴⁵, with its benzofuran terminus occupying the S1 site and the indole terminus extending into the extracellular vestibule⁵². Hence, we designed a set of S1 mutants (Y95F, I172M, N177V, F341V, and S438T), which were predicted to interact with vilazodone in this alternative binding pose. Of note, several of these residues have been shown to decrease the affinity for a variety of SSRIs^45,55. The 5-HT uptake inhibition of the S1 mutants resulted in potencies for vilazodone (Fig. 1c, Supplementary Fig. 1) that did not provide a clear indication of a specific vilazodone binding pose. The SERT mutants either did not significantly decrease the vilazodone potency or, in some cases, potentiated its inhibition to a significant extent (e.g., I172M, N177M, F341V, and S438T) (Fig. 1c). To further mechanistically explore the binding pose of vilazodone, we generated S1 mutants combined with the allosteric site mutant F556A (Y95F-F556A, I172M-F556A, and S438T-F556A), which were predicted to alter the cation-π interactions with the indole terminus. However, these mutants did not significantly change the inhibition potency of vilazodone either (Fig. 1c). Since our molecular pharmacology approach was unable to support the computationally predicted binding pose of vilazodone, we turned to cryo-EM to capture its interaction at an atomistic level.

Cryo-EM reveals an orthosteric binding mechanism of vilazodone

To obtain the cryo-EM structure of the SERT-vilazodone complex in the absence of imipramine, we expressed the full-length SERT WT in Expi293 mammalian cells using the BacMam transduction system⁵⁶. SERT was solubilized in a glyco-diosgenin (GDN) buffer and purified using a C-terminal twin-strep tag (Supplementary Fig. 2). Notably, GDN-solubilized SERT showed comparable (S)-citalopram binding properties (IC₅₀ = 4.8 [1.6; 12.0] nM, Supplementary Fig. 2) as for SERT expressed in membranes ( ~5 nM⁴⁵). For the cryo-EM preparation, vilazodone (10 to 100 μΜ) was present throughout the purification and grid preparation stages. The 15B8 Fab fragment served as a fiducial for cryo-EM facilitating the precise alignment of micelle-embedded SERT particles (Fig. 2a, b). The 15B8 Fab-bound SERT (Supplementary Fig. 2) can undergo all conformational transitions essential for 5-HT transport^25,40,57,58. Furthermore, vilazodone binding was largely unaffected by 100 nM Fab 15B8, as confirmed by unchanged 5-HT uptake inhibition in whole-cell assays (Supplementary Fig. 2e). Hence, we posited that the presence of 15B8 Fab did not affect the vilazodone-induced SERT conformation and prepared the complex by co-elution (Supplementary Fig. 2). We successfully determined the structure of the SERT-vilazodone assembly at a global resolution of 2.78 Å (PDB ID: 9HCO, Supplementary Fig. 3,4). GDN-solubilized SERT exhibited an outward-open conformation with the extracellular vestibule hydrated and a densely packed intracellular side. The N-terminus (residues 1-76) and C-terminal purification tags are not resolved. The root mean square deviations (RMSDs) of the backbone Cα atoms of the TM helices between this structure and the SERT structures with imipramine-vilazodone, 5-HT outward-open, 5-HT occluded, and ibogaine inward-open structures are 0.64 Å, 0.70 Å, 1.70 Å, and 3.89 Å, respectively. Moreover, TM1 does not adopt the characteristic kink of the inward-facing ibogaine- or 5-HT-bound conformations, indicating that the intracellular gate of SERT is closed (Fig. 2c)^25,40. Furthermore, superimposing the Cα in the TMs with those of the 5-HT-bound outward-open (PDB ID: 7LIA)²⁵ and -occluded (PDB ID: 7MGW)²⁵ structures of SERT suggests that the conformations of the key residues F335, Y107, and W103 correlate with an open extracellular gate. With the above observations, we can conclude that the vilazodone-bound SERT structure is in the outward-open structure (Supplementary Fig. 5).

**Fig. 2: Structure of SERT-Fab with vilazodone bound in the S1 site.**

Within the cryo-EM density, we identified a non-proteinaceous electron density at the S1 site, extending into the vestibule towards, but not reaching, the S2 site (Fig. 2c-e). We attributed this density to a vilazodone molecule with its indole moiety occupying the S1 pocket, in the same manner as S1-bound 5-HT (PDB ID: 7LIA), and with the benzofuran group facing the extracellular vestibule (Fig. 2d, e, Supplementary Fig. 5). Similar to other SSRIs^{31,40,44,45,46} and 5-HT²⁵ in outward-facing SERT structures, the protonated amine in the piperazine group preserves its interaction with the carboxyl group of D98 positioned 4.5 Å apart (Fig. 2d). Furthermore, S336 participates in a hydrogen-bonding network with N368, which in turn forms a hydrogen bond with N101, approximately 4.5 Å from the non-proteinaceous density in the known chloride binding site^{31,40,42,44,45,46,59}. Nestled within subsite B, the indole ring of vilazodone is aligned between TM3 and TM8 through hydrophobic interactions with A169, I172, and Y95, which block the extracellular side right above the indole group. In addition, Y176 forms a stabilizing bridge between TM3 and TM8 through an aromatic hydrogen bond with the backbone carbonyl group of G435. One of the most prominent interactions is the π-π stacking between the indole ring and F341, whose sidechain is lifted upwards, filling the void below vilazodone. With this F341 conformation, vilazodone closely resembles SERT binding of sertraline and fluvoxamine⁴⁶ but not paroxetine or (S)-citalopram³¹. Furthermore, the benzofuran group interacts through π-π stacking with F335 and is stabilized by aromatic hydrogen bonding with the nearby Q332. Right above the benzofuran group, E493 projects towards the extracellular side where it forms a salt-bridge with R104 with a 3.4 Å distance. This putative salt-bridge (R104-E493) has been characterized as an essential intramolecular interaction signaling the opening and the occlusion of SERT. R104 also plays a pivotal role in the accommodation of allosteric ligands⁴⁸. When vilazodone binds at the S1 site its benzofuran group forms potential cation-π interactions with R104, an interaction that is unique among all the investigated SSRIs. Although vilazodone does not reach F556 in the extracellular vestibule, the F556 phenyl faces towards the water bulk, edge-to-edge with F335, similar to the paroxetine-bound conformation³¹. Since vilazodone likely binds with high affinity to both the S1 and allosteric sites, we superimposed the respective SERT structures according to the Cα of the TMs (PDB ID: 7LWD⁴⁵) to compare the two vilazodone poses (Fig. 2f, g). Interestingly, the two vilazodone poses share the interaction with R104, suggesting that this region functions as a binding pocket that can accommodate a bulkier aromatic system, such as a benzofuran or an indole group. This comparison underscores the versatility of the binding pocket beneath R104.

We identified the ion binding sites where the Na2 and Cl ions were well resolved and coordinated as seen previously^31,46,59 (Fig. 3a, b). A water molecule was also identified³¹ at a 3.8 Å distance from the Na⁺ ion in the Na2 site (Fig. 3b), forming a hydrogen bond with D437 that further stabilizes the Na2 cavity. In contrast, the presence and occupancy of the Na1 site are more questionable. We resolved the N101 side chain and encountered a weak density extending towards A96 (Fig. 3a, b). This density could potentially accommodate a Na⁺ ion surrounded by weak stabilizing interactions from the A96 backbone carbonyl (2.0 Å), N101 carbonyl (3.0 Å), D98 carboxyl (3.2 Å), S336 hydroxyl (3.3 Å), N368 carbonyl (4.0 Å), and the water molecule within 4.3 Å distance. The discerned density is within the Na1 binding pocket determined in previous SERT cryo-EM structures^25,32. Between the piperazine moiety and the Na1 site, where the EM density has a resolution below 2.4 Å, two potential water molecules are revealed intricately connected by hydrogen bonds, forming a cooperative network with the F335 and Y95 backbones, as well as the D98 sidechain (Fig. 3c).

**Fig. 3: Structure of binding sites for ions and water.**

The cryo-EM structure allowed us to conduct a deeper structure-guided analysis of the SERT-vilazodone complex with the purpose of creating SERT mutants that would challenge vilazodone inhibition potency while not perturbing 5-HT uptake. In particular, the aliphatic chain connecting the piperazine with the indole group lay in the vicinity of I172 and S438 (Fig. 4a). Thus, upon screening interacting residues, we generated the SERT I172M-S438T double mutant and expressed it in COS-7 cells. We found that vilazodone exhibited a significant 200-fold decrease in the [³H]5-HT uptake inhibition potency relative to that of SERT WT (Fig. 4b, Supplementary Fig. 1e). Furthermore, we observed that the indole group of vilazodone was coordinated by A169 in TM3 (Fig. 4a). To explore this further, we made the SERT A169I mutant, reasoning that the bulkier isoleucine might interfere with the stabilization of the indole group of vilazodone. Notably, A169 mutants have also been demonstrated to interfere with paroxetine⁶⁰ and vortioxetine binding⁶¹. As predicted, vilazodone inhibition of [³H]5-HT uptake was decreased significantly by nearly 700-fold for SERT A169I with an IC₅₀ of 690 [510; 1000] nM (Fig. 4b, Supplementary Fig. 1e).

**Fig. 4: Key SERT residues involved in the inhibition of SERT by vilazodone.**

SERT binds vilazodone Na⁺-independently with nM affinity

Given the unique binding pose of vilazodone and the observation that the Na1 site might be collapsed in our cryo-EM structure, we next investigated whether vilazodone binding was ion-dependent. To this end, we used spectral shift assays⁶² to directly determine the vilazodone affinity for purified SERT in Na⁺-, K⁺- and NMDG⁺-rich environments. Spectral shift relies on fluorescence from a dye coupled to SERT purified in detergent micelles. To confirm that our labeled SERT is sensitive to vilazodone binding, we tested for a ligand-induced spectral shift in the emission spectrum under isothermal conditions, which we measured using a Dianthus instrument⁶². Accordingly, we titrated vilazodone against fluorescently labeled SERT in a buffer containing either 150 mM NaCl, KCl, or NMDG-Cl. The K_D values for vilazodone in Na⁺, K⁺, and NMDG⁺ were measured to be 1.0 [0.4; 2.3] nM, 7.1 [3.5; 14] nM, and 13 [10; 17] nM, respectively (Fig. 5). In comparison, the K_D values for (S)-citalopram and paroxetine in Na⁺ were 0.2 [0.0; 0.7] nM and 1.2 [0.7; 2.0] nM, respectively (Supplementary Fig. 6a, b, Supplementary Table 2), agreeing with previous literature findings³¹. We were unable to detect any binding in K⁺, underscoring the Na⁺-dependency for the binding of these two conventional SSRIs (Supplementary Fig. 6c, Supplementary Table 2). Spectral shift also allowed for a direct determination of the K_D of Fab 15B8 binding to SERT and that the ion-dependency of vilazodone did not change in the presence of the Fab 15B8 (Supplementary Fig. 7).

**Fig. 5: Fluorescence equilibrium binding shows distinct vilazodone affinities in Na⁺, K⁺ and NMDG⁺.**

The observation that vilazodone may bind in the presence of Na⁺, K⁺ and NMDG⁺ suggests either that vilazodone binds to the SERT conformational states induced by these ions, or that vilazodone itself binds in a manner independent of ions. To distinguish between these possibilities, we investigated whether vilazodone binding had any impact on the conformation of SERT by inserting the fluorescent ncAA Anap at position V86 in TM1a (V86Anap) or F556 in TM11 (F556Anap), on the intracellular or extracellular sides, respectively. Detecting changes in Anap fluorescence as a conformational reporter has previously enabled us to show changes in SERT conformational dynamics upon the binding of a set of ions and ligands⁵⁴. Similarly, we used the Anap mutants to investigate structural changes induced by the binding of vilazodone in combination with either Na⁺, K⁺ or NMDG⁺. Notably, the potencies for vilazodone inhibition of [³H]5-HT uptake for these mutants were not significantly different from those of SERT WT (Supplementary Fig. 8a). We measured the fluorescent properties of the inserted Anap, following incubation in the absence or presence of a saturating concentration of vilazodone (200 nM) in either Na⁺ or K⁺ conditions (Fig. 6). Both V86Anap and F556Anap demonstrated different fluorescence patterns in Na⁺ and K⁺, consistent with our previous findings, and likely reflecting the different conformational states adopted with the two ions⁵⁴ (Fig. 6). In the presence of vilazodone, both V86Anap and F556Anap yielded changes in their fluorescence spectra. For V86Anap (Fig. 6a), which captures structural changes in the microenvironment around TM1a, the addition of vilazodone decreased the fluorescence in Na⁺ while increasing the fluorescence in K⁺. These results could indicate that vilazodone binds to SERT in both ionic conditions. For F556Anap, which has Anap inserted into the extracellular vestibule of SERT, the emission spectra with vilazodone and Na⁺ showed a large decrease in intensity accompanied by a blue-shift relative to that with Na⁺ alone (Fig. 6b). In contrast, vilazodone in K⁺ produced a red-shifted and increase in fluorescence compared to that of F556Anap with just K⁺. Given that vilazodone has no effect on the fluorescence of free Anap (Supplementary Fig. 8b), the observation that vilazodone changed the fluorescence of V86Anap and F556Anap in both Na⁺ and K⁺ suggests that vilazodone binds in both ionic conditions. Furthermore, the spectral differences between the vilazodone-bound mutants in Na⁺ and K⁺ indicate that the SERT-vilazodone complexes in Na⁺ and K⁺ are not the same. Lastly, we examined vilazodone-induced conformational changes in NMDG⁺ and observed a decrease in quantum yield and a blue-shift in F556Anap fluorescence. The data support the results from the spectral shift assay, that vilazodone can also bind SERT in its apo-form. The marked change in fluorescence relative to our observations for the Na⁺ and K⁺ condition, indicate a distinct conformational transition and therefore vilazodone binding pose in the NMDG⁺ buffer (Supplementary Fig. 8c).

**Fig. 6: Vilazodone induces distinct conformational dynamics in Na⁺ and K⁺.**

Interestingly, the magnitude of the spectral changes induced by binding of vilazodone for F556Anap in Na⁺ was sufficiently large for us to measure the real-time changes in fluorescence associated with the addition of vilazodone. For this experiment, we applied 200 nM vilazodone to F556Anap pre-equilibrated in Na⁺ and compared the time-dependent changes in fluorescence to those induced by the addition of 200 nM imipramine, shown previously to also decrease the fluorescence⁵⁴ (Supplementary Fig. 8d). For both compounds, we observed time-dependent decreases in fluorescence well modelled by mono-exponential time courses. Although using a single concentration of ligand does not allow us to resolve their kinetic parameters, the traces show that the application of vilazodone, despite its higher SERT affinity, produces a 4.4-fold slower change in fluorescence compared to that of imipramine. This may indicate that vilazodone displays kinetic binding properties distinct from those of classical SERT inhibitors that exclusively bind to S1. We also performed a kinetic experiment where SERT was pre-incubated with 100 nM of imipramine for 30 min, and the association of vilazodone to the allosteric site was measured. The observed τ was 0.8 ± 0.18 min, which is 1.9-fold faster than vilazodone binding in the absence of imipramine, yet 2.4-fold slower than imipramine binding. These results further support our previous findings⁴⁵. The observed differences in vilazodone association kinetics at the two binding sites may reflect variations in affinity and/or conformational dynamics. Furthermore, the signal window observed for vilazodone binding to imipramine-bound SERT suggests that the conformation of SERT in the presence of both ligands differs from that observed with imipramine alone.

Deconstruction of vilazodone reveals moieties determining its affinity

The chemical structure of vilazodone is characterized by its elongated shape, featuring a 2-amido-substituted-benzofuran ring attached to a piperazine, which in turn is linked to an indole through a four-carbon chain. To elucidate the individual contribution of these moieties to the affinity for SERT, we dissected vilazodone into two different series of fragments (Fig. 7a). The first set of compounds retained the benzofuran and piperazine rings found in the parent structure while replacing the indole and four-carbon linker with a methyl (CDV-2-29), n-butyl (CDV-2-26), or n-butylbenzene (CDV-2-32). The second set of compounds preserved the indole and linking chain, removing the benzofuran (CDV-3-1) and deconstructing the piperazine ring (CDV-2-81 and CDV-3-2).

**Fig. 7: Vilazodone fragments reveal structural basis for SERT inhibition potency.**

When screened for their ability to inhibit [³H]5-HT uptake, fragments CDV-2-29, CDV-2-26, and CDV-2-32, which all contained the benzofuran, exhibited IC₅₀ values in the μΜ range: 5300 [4000; 7100] nM, 9500 [6500; 14000] nM, and 3600 [2500; 5400] nM, respectively (Fig. 7b, Supplementary Table 3). Based on our vilazodone-bound structure, the benzofuran preserves its binding site in the pocket adjacent to R104 allowing cation-π interactions, which is essential for other allosteric ligands, such as Lu AF88273⁴⁸. The comparison between the S1 and allosterically bound vilazodone structures (Fig. 2f, g) suggests that there might be a benzofuran binding hot spot below the R104-E493 salt-bridge, essential for the high allosteric potency of vilazodone. However, all three benzofuran derivatives that were no longer tethered to the terminal indole lost affinity for SERT by several orders of magnitude compared to vilazodone. Notably, CDV-2-26 and CDV-2-32 contain a linker attached to the piperazine moiety to mimic the elongated structure of vilazodone, yet these two fragments maintained their IC₅₀ values in the μM range.

In contrast, indole-based fragments CDV-2-81, CDV-3-1 and CDV-3-2 (Fig. 7b), displayed orders of magnitude higher SERT inhibition potencies with IC₅₀ values of 21.6 [19.1;24.5] nM, 2.2 [1.9; 2.6] nM, and 35.2 [32.7; 40.4] nM, respectively. Interestingly, the combination of the indole and piperazine (CDV-3-1) produced a similar IC₅₀ value compared with vilazodone with only a 2.2-fold increase in [³H]5-HT uptake inhibition (Fig. 7b). These results suggest that the protonatable nitrogen of the piperazine ring increases the inhibition potency by at least an order of magnitude when combined with an indole (CDV-3-1) but not a benzofuran (CDV-2-26, CDV-2-29, and CDV-2-32). For each fragment, we calculated the pKa of the piperazine nitrogen using Epik⁶³ to predict their protonation states and further locate the positive charge in reference to D98 that was seen adjacent to vilazodone in the cryo-EM structure. Based on these calculations, the piperazine in CDV-3-1 is protonated at a different nitrogen compared to vilazodone, placing it at the terminal secondary amine. The charge is farther from D98 (6.4 Å vs 4.1 Å). The similar IC₅₀ values for vilazodone and CDV-3-1 could either suggest that the protonated nitrogen has minimal contribution to the affinity of vilazodone or that the changed protonation of CDV-3-2 could largely substitute for that of vilazodone. Notably, CDV-3-1 was the only fragment with a high Hill coefficient of approximately 1.62, not significantly different from that of vilazodone (Supplementary Table 3). Upon removal of the piperazine group (CDV-2-81 and CDV-3-2), the IC₅₀ increases to 21.6 nM and 35.2 nM, respectively, demonstrating the importance of the indole terminus.

To assess whether these fragments, like vilazodone, could shift the conformation of SERT independently of the ions present, we recorded emission spectra of V86Anap and F556Anap. This was performed after incubation with CDV-3-1 and CDV-3-2 (which lacks the piperazine) in the presence of Na⁺ or K⁺ (Supplementary Fig. 9). In V86Anap experiments, CDV-3-1 shifted fluorescence in both ion environments similar to vilazodone, suggesting that CDV-3-1 binds in both Na⁺ or K⁺ (Supplementary Fig. 9a). The changes in fluorescence observed for F556Anap with CDV-3-1, on the other hand, were less pronounced (Supplementary Fig. 9b). Interestingly, the addition of CDV-3-2 gave rise to changes in fluorescence for both Anap SERT constructs in Na⁺ but not in K⁺ (Supplementary Fig. 9c, d), suggesting either that CDV-3-2 binding is Na⁺-dependent or that the state induced by its binding is very similar to the K⁺-bound SERT conformation. Notably, the smaller fluorescence changes induced by these two fragments compared to those of vilazodone could be attributed to the fact that its benzofuran ring is in close proximity to the F556Anap residue and, thereby, may affect the local chemical environment of Anap through direct interactions.

Discussion

Vilazodone is classified as an SSRI with its therapeutic properties primarily attributed to interactions with SERT and 5-HT_1A receptors. Multiple mechanistic factors likely contribute to the unique pharmacological profile of vilazodone. In our previous work⁴⁵, we demonstrated that vilazodone acts as a non-competitive inhibitor of SERT and binds with high affinity to a pocket in the extracellular vestibule, supporting an allosteric mechanism of action. Notably, vilazodone exhibited a Hill coefficient of ~1.8, suggesting positive cooperativity for inhibiting SERT despite only a single density for vilazodone seen in the EM map⁴⁵. This discrepancy, along with its distinct pharmacological profile, made us hypothesize that potentially the bound imipramine biased the vilazodone binding pose in the cryo-EM structure. Accordingly, it inspired us to further investigate the structure-activity relationships of vilazodone for SERT inhibition and understand the molecular basis of this drug’s possible multimodal effect.

Compared to our previous findings, we report herein a cryo-EM structure of SERT, showing that in the absence of imipramine, vilazodone itself acts as an orthosteric inhibitor of SERT. We did not observe any additional EM densities that could be attributed to allosteric binding. The observed orthosteric binding could explain the unchanged vilazodone affinity when placing mutations in the allosteric site. However, if the S1 site was the only site available for vilazodone, we would have expected an impact on affinity from the S1 mutations.

Our current cryo-EM structure reveals that, in the absence of orthosteric ligands, vilazodone binds at the S1 site. Yet, it is important to integrate these findings with our previous observations of vilazodone allosteric binding⁴⁵. Previously, the addition of imipramine revealed a vilazodone binding site with an allosteric potency of 14 nM, suggesting that imipramine may stabilize SERT in a conformation favorable for high-affinity allosteric binding of vilazodone. We hypothesize that the same effect is plausible if mutations are performed which destabilize vilazodone binding within the S1 site. In these cases, vilazodone binding would be biased towards the previously described allosteric site⁴⁵. Furthermore, our dose-response curves for vilazodone yield a Hill coefficient >1, but the cryo-EM map resolves only a single ligand-binding cavity that overlaps with the allosteric pose⁴⁵, ruling out intramolecular cooperativity. In addition to this, our cryo-EM structure was captured in the absence of 5-HT. Therefore, the dynamic nature of 5-HT transport and its S1-binding could potentially enable vilazodone to bind both the S1 and the allosteric site, interchangeably. This is further corroborated by the apparent non-competitive inhibition of 5-HT transport by vilazodone⁴⁵. Accordingly, the steep Hill slope likely reflects other alternative mechanisms—such as an unresolved third high-affinity site or inter-protomer communication—and is interpreted here as an empirical descriptor. While minor methodological differences between our previous study⁴⁵ and the results herein may contribute to a variation in Hill coefficients, both datasets support the conclusion that vilazodone inhibits [³H]5-HT uptake with a Hill coefficient different from 1. Several factors could contribute to an altered Hill coefficient. One possibility is that vilazodone binding exhibits a component of a positive cooperative mechanism.

In addition to this, we find that vilazodone can inhibit functional SERT mutants such as Y95F, I172M, S438T, and F556A, as well as double mutants involving both S1 and allosteric sites, with a potency similar to WT. This may be attributable to either its bimodal binding mode or to its elongated structure, which enables interactions with multiple residues across both sites. It could also be a combination of both. Although vilazodone cannot occupy the S1 and allosteric sites simultaneously, due to steric clashes, its capacity to interact with either site may preserve function even when one site is mutated. This is consistent with the observed loss of cooperativity (Hill coefficient ~1) upon disruption of one site. However, two S1 SERT mutants (A169I and I172M–S438T) reduced vilazodone’s inhibitory potency, providing in vitro confirmation that the S1 site is the primary binding site in the absence of other orthosteric ligands.

When we investigated the ion-dependent nature of vilazodone binding, we found that vilazodone binds with similar affinities in the presence of Na⁺, K⁺ and NMDG⁺. The experiments with SERT encoding Anap suggested that vilazodone may bind to SERT conformations promoted by either Na⁺, K⁺ or NMDG⁺. These findings align with previous attempts using SERT-expressing membranes (5.8 nM and 36.7 nM, in Na⁺ and K⁺, respectively)⁶⁴. Indeed, the reported pharmaco-chaperoning effects of vilazodone^64,65, were proposed to stabilize inward-facing states, thereby substantiating our Na⁺-independent binding results.

To better understand the structure-activity relationships underlying its high affinity we deconstructed vilazodone into fragments. Using this approach, we demonstrated that the indole derivatives of vilazodone are potent SERT inhibitors. Addition of the piperazine group enhanced the affinity of the indole derivatives. In contrast, all the benzofuran-piperazine derivatives exhibited a low inhibition potency for [³H]5-HT. We previously showed that the benzofuran moiety is important when vilazodone occupies the allosteric site⁴⁵, suggesting a preference for these derivatives towards this site. This is in agreement with previous studies, in which similar benzofuran ligands showed low affinity for SERT⁶⁶. Overall, only CDV-3-1 (indole + piperazine) showed a high affinity (2.2 nM), a high Hill coefficient and a Na⁺ and K⁺ conformational response similar to that of vilazodone in the SERT-Anap experiments. Therefore, the high affinity of vilazodone for SERT, its potential Na⁺-independence for binding and its high Hill coefficient may be attributed to the combination of the terminal indole attached to the piperazine moiety. Further investigation is necessary to fully understand the ion-independence of these different fragments. Designing multimodal ligands like vilazodone may offer improved pharmacological profiles and the potential use in treatment-resistant patients⁶⁷.

Methods

Site-directed mutagenesis

The gene for full-length hSERT was encoded with a C-terminal thrombin cleavage site followed by a Twin-Strep tag and a His12-tag, each separated by GGS linkers. The construct was synthesized (Eurofins Genomics). hSERT for uptake experiments was cloned into the pUbi1z vector using the NotI and XbaI. The primers (EurofinsGenomics) were designed as follows: Y95F: gcccaggtccactgc; R104K: cctgggcaatgtctggaaattcccctacatatgttaccag; A169I: ggttatgccatctgcatcattatattttacattgcttcctactacaac; I172M: gcatcattgccttttacatggcttcctactacaac; N177V: gccttttacattgcttcctactacgtgacc; Q332A: gatagatgcagccgctgcaatcttcttctctcttggtccg; F335A: cagccgctcagatcttcgcgtctcttggtccgggc; F341V: ctctcttggtccgggcgtgggggtcctgct; S438T: caagcctgcssscgttgtgtccaagcc; E493N: ccccgtggcatactcgttcagcagcttcac; E494Q: ggtgaagctgctggagcaatatgccacggggc; Y495A: gaagctgctggaggaggcagccacggggcc; F556A: cctgttcatcatttgcagtgcgctgatgagcccgcc; Y579A: ctggagyayacatcttgggtgcgtgcataggaacctcatctttc. Cloning procedures for V86Anap and F556Anap can be found at Nygaard et al. (2024)⁵⁴. Gene sequences were verified by DNA sequencing (Eurofins Genomics).

[³H]5-HT uptake experiments

Uptake experiments were performed at room temperature (RT) using 5-[1,2-³H] hydroxytryptamine ([³H]5-HT, 43,1 Ci/mmol, Perkin Elmer) on COS-7 cells. COS-7 cells were transiently transfected with SERT WT or mutants, using Lipo2000 transfection protocol (Invitrogen): 0.8 μg SERT plasmid and 2.4 μL Lipofectamine were mixed each with 200 μL Opti-Mem^R (1X) and incubated for 20 minutes (min) for complex formation. The mixture was added to 13 mL DMEM 1885 medium (in house) containing 2 million COS-7 cells and seeded in a 24-well plate. To ensure an insignificant degree of radioligand depletion within the assay, the seeded cell number was adjusted within in each SERT construct (WT or mutant) to achieve an uptake level of maximally 10% of the total added [³H]5-HT. After 5 h, 600 μL of DMEM1885 with penicillin, streptomycin and L-glutamine was added. Transfections with SERT encoding Anap were performed as described previously⁵⁴. Briefly, pre-seeded HEK293 cells were transfected at ~50% confluency with 442 ng DNA and 1.33 µg polyethyleneimine per well in 24-well plates. A 2:2:1 plasmid ratio of SERT (pcDNA3), Anap-specific tRNA and aaRS (pANAP) and eukaryotic release factor E55D (pCAG) was used. The medium was supplemented with 15 µM Anap during and following transfection. The uptake assays were carried out 2 days after transfection. Just prior to the experiment, the cells were washed once in 400 µL uptake buffer (UB) (25 mM HEPES, 130 mM NaCl, 5.4 mM KCl, 1.2 mM CaCl₂, 1.2 mM MgSO₄, 1 mM L-ascorbic acid, 5 mM D-glucose, pH 7.4) RT. To assess the preincubation time needed to establish equilibrium conditions for vilazodone binding a time interval between 0 and 60 min were tested (Supplementary Fig. 10). Accordingly, 50 µL of vilazodone was added to cells in the indicated concentrations, 30 min prior to the addition of 50 µL 12.25 nM [³H]5-HT. Nonspecific binding was determined with 1 µM paroxetine (Sigma-Aldrich). After 3 min of incubation (2-4 min for the Anap mutants), the uptake reaction was stopped by washing twice with 500 µL ice-cold UB. Cells were lysed in 250 µl 1% SDS and incubated for 60 min at 37 °C. All samples were transferred to 24-well counting plates (Perkin Elmer, Waltham, MA), 500 µL of Opti-phase Hi Safe 3 scintillation fluid (Perkin Elmer) was added, and the plates were counted in a Wallac Tri-Lux β-scintillation counter (Perkin Elmer). Data were normalized to 100% of control (WT in absence of inhibitor) - equal to approximately 22720 ± 1591 fmol/min/10⁵ cells. Non-specific background activity was ~20 fmol/min/10⁵ cells in all experiments. All experiments were carried out in triplicate.

SERT expression and membrane preparation

The human SERT construct used for the cryo-EM studies was the full-length WT transporter tagged with a twin-strep tag followed by a His12-tag, into a pEG BacMam vector. SERT was expressed in mammalian cells using the Bac-to-Bac baculovirus expression system. Baculovirus for the expression of SERT was produced using Sf9 cells (Expression Systems). The virus was used for the infection of mammalian Expi293F cells (Gibco)⁶⁸. The Expi293F cells were incubated at 37 °C, 5% CO2, 70% humidity, 130 rpm, in Expi293 Expression Medium (Gibco) until they reached optimal density for the expression of SERT. The cells were infected with 1.25% of P2 virus and 2 mM of valproic acid at a cell density 3 ×106 million cells/mL of medium and incubated for 72 hours (h) at 37 °C, 5% CO2, 70% humidity, 130 rpm, before harvesting. For SERT with Anap, cells were transiently transfected as described previously⁵⁴. In short, 6 µg polyethyleneimine and 2 µg DNA, pre-mixed in Opti-MEM, were added per mL of culture together with 15 µM Anap. A 2:2:1 ratio of the following plasmids was used: pcDNA3 encoding SERT, a plasmid containing the Anap-specific tRNA and aaRS (pANAP), and a plasmid encoding a eukaryotic release factor E55D. The cells were harvested at 6200 g, washed with cold PBS (in-house), harvested at 5000 g, snap-frozen and stored at −80 °C. On the day of membrane preparation, cells were resuspended into 20 mL of lysis buffer (30 mM NaHEPES pH 8, 30 mM NaCl, 5 mM KCl, 10% sucrose, 10 μg/ml benzamidine and 10 μg/mL leupeptin, protease inhibitor cocktail (Sigma Aldrich), 7 mM MgCl₂, 2 μg/mL DNase, 2 μg/mL RNase) per 10 g of cell pellet. Next, cells were lysed on ice with ultrasound probe (Branson Sonifier 250 set at 50% duty cycle and output control 5). The sample was centrifuged at 1000 × g for 5 min at 4 °C to remove non-lysed cells. Supernatant was collected and the pellet was resuspended in 20 mL of fresh lysis buffer like previously. Sonication and harvest were repeated to collect the supernatant. The membranes from the resultant supernatant were harvested for 1 h at 100,000 × g using a fixed angle rotor T-647.5 at 4 °C. Membranes were washed twice with wash buffer (30 mM NaHEPES pH 8, 30 mM NaCl, 1 M NaCl, 10 mM DTT), centrifuged for 1 h at 100,000 × g using a fixed angle rotor T-647.5 at 4 °C, resuspended in resuspension buffer (30 mM NaHEPES pH 8, 30 mM NaCl, 5 mM KCl, 10% sucrose), snap-frozen in liquid N₂ and stored at −80 °C.

SERT purification

The membranes were subsequently resuspended in 4x resuspension buffer (80 mM Tris, pH 8.0, 600 mM NaCl, 20% glycerol, 10 μg/mL benzamidine and 10 μg/mL leupeptin, protease inhibitor cocktail (Sigma Aldrich), 0.5 mM TCEP), 9 mL per 1 g membrane. Cold solubilization buffer (200 mM Tris, pH 8.0, 200 mM n-dodecyl-β-D-maltoside (DDM) (10%), 40 mM cholesteryl hemisuccinate (CHS) (2%)) was added (1 mL / 1 g of membrane) and incubated on gentle rotation at 4 °C for 1.5 h. The pellet was harvested at 120,000 × g using a fixed angle rotor T-647.5 for 1 h and was discarded. The supernatant was filtered through 0.2 μm filters and diluted 2x in resuspension buffer (20 mM Tris, pH 8.0, 150 mM NaCl, 5% glycerol, 10 μg/mL benzamidine and 10 μg/mL leupeptin, protease inhibitor cocktail (Sigma Aldrich), 0.5 mM TCEP). The solubilized protein was purified by batch purification with single step nickel affinity purification and eluted with 20 mM Tris pH 8.0, 300 mM NaCl, 0.86 mM GDN (glycol-diosgenin), 0.5 mM TCEP, 300 mM imidazole, 5% glycerol, 10 µg/ml benzamidine, and 10 µg/mL leupeptin. The eluted sample was injected onto a Superdex® 200 Increase 10/300 GL gel filtration column (GE Healthcare) equilibrated in SEC buffer (20 mM Tris pH 8, 300 mM NaCl, 0.43 mM GDN) and fractions containing SERT (Supplementary Fig. 2) were pooled and concentrated using a Vivaspin 500 PES 50 kDa cut off spin filter (Sartorius, Cat#VS.0131). For SERT with Anap incorporated, the solubilized protein was immobilized to a HisTrap column and eluted with a linear imidazole gradient in buffer containing 20 mM TrisCl pH 8.0, 300 mM NaCl, 10% (v/v) glycerol, 1 mM DDM, 0.2 mM CHS, and 24 µM lipids (1-palmitoyl-2-oleoyl-glycero-3-phosphocholine (POPC), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (POPE), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoglycerol (POPG)) in a 1:1:1 ratio. The same buffer was used for the subsequent size-exclusion⁵⁴.

Scintillation proximity assay

The activity of hSERT was verified by [³H](S)-citalopram competition binding using scintillation proximity assay (SPA), and each assay was performed in white, clear-, flat-bottomed 96-well plates (Corning). SERT at a final concentration of 5 nM was added to a solution of 5% Copper HIS-Tag yttrium silicate (YSi) SPA beads (PerkinElmer), and 10 nM [³H](S)-citalopram (81 Ci/mmol) in sample buffer (20 mM Tris pH 8.0, 100 mM NaCl, 0.43 mM GDN). [³H](S)-citalopram binding was competed with (S)-citalopram in the indicated concentrations. All data points were performed in triplicate. Nonspecific binding was determined using 10 µM paroxetine. The binding experiment was incubated at room temperature for 1 h followed by overnight incubation at 4 °C to obtain equilibrium conditions. Binding activity was quantified on a MicroBeta scintillation counter (PerkinElmer). The data was plotted in GraphPad Prism 10 using the equation for a monophasic inhibition.

Fab 15B8 expression and purification

The Fab 15B8 sequence cloned in a pFastBac vector was acquired from Eric Gouaux and included a His8 tag. A recombinant baculovirus was generated based on the Bac-to-Bac expression system standard protocols for protein expression in SF9 insect cells. The Sf9 cells were infected with 2.5% v/v recombinant baculovirus at a cell density of 2 million cells/mL of medium at 27 °C. The culture supernatant was collected 96 h after infection by centrifugation at 5000 rpm for 20 min using a JLA 8.1000 rotor at 4 °C. 50 mM of pH 8 phosphate buffer was added, and precipitation was harvested at 5000 rpm for 20 min using a JLA 8.1000 rotor at 4 °C. The 15B8 Fab was purified from Sf9 supernatant by Histrap HP 5 mL pre-packed column equilibrated in 50 mM phosphate pH 8, 300 mM NaCl, and protease inhibitor cocktail (Sigma-Aldrich). 15B8 fractions were collected after elution with 50 mM phosphate, pH 8, 300 mM NaCl, and 500 mM imidazole. The fractions were concentrated Vivaspin 20 PES 10 kDa cut off spin filter (Cytiva) to inject and buffer exchange using a Superdex ® 200 Increase 10/300 GL gel filtration column (GE Healthcare) equilibrated 20 mM Tris, pH 8, 300 mM NaCl, and fractions containing 15B8 were pooled and concentrated using a Vivaspin 20 PES 10 kDa cut off spin filter (Cytiva). Samples were snap-frozen in liquid N₂ and stored at −80 °C.

Cryo-EM sample preparation

SERT was incubated in 100 μM of vilazodone for 30 min at RT. SERT and Fab 15B8 were added in a 1:2 molar ratio and incubated for 30 min at RT. The SERT-Fab-vilazodone complex was co-eluted using a Superose^TM 6 Increase 10/300 GL gel filtration column (GE Healthcare) equilibrated 20 mM Tris pH 8, 100 mM NaCl, and 10 μM vilazodone. Fractions containing both SERT and Fab 15B8 were pooled and concentrated using a Vivaspin 20 PES 100 kDa cut off spin filter (Cytiva). A further 100 μM vilazodone was added in the concentrated sample. 3 μL of the sample was applied to a glow discharged for 15 mA for 30 s (Leica EM ACE200) UltrAufoil grid (R1.2/1.3 300 mesh, Quantifoil) (Jena Bioscience, Cat.-No.: X-201-Au300), and a Vitrobot Mark IV (FEI) was used to blot away excess sample plunge the grids into liquid ethane for sample vitrification at 100% humidity and 4 °C in an environmental chamber. Grids were stored in liquid N₂ until data collection.

Cryo-EM data acquisition and processing

Movies were acquired using a Titan Krios G2 (FEI) fitted with a Falcon 3EC detector operated in counting mode at 165,000×magnification (300 keV), yielding a raw pixel size of 0,832 Å per pixel, using EPU 3.2 software (Thermo Fisher). Movies were collected with a total doe of 42 e/Å⁻² at a dose rate of 1.1 e/px/s and a defocus value range of −0.6 to −2.0 μm. The final reconstructions were obtained using CryoSPARC⁶⁹ v.3.3.2 to v.4.4.1 as shown in Supplementary Fig. 3.

Model building and refinement

The cryo-EM structure (PDB ID: 7LIA)²⁵ of the human serotonin transporter was used as an initial model after removing residues 1-68 that did not correspond to electron density. The structure was initially docked in ChimeraX v1.8⁷⁰ and next, manual adjustment and initial refinements were made in Coot v0.9.8.6⁷¹. Real-space refinement was carried out in Phenix v1.20.1-4487⁷². Vilazodone was generated in Phenix and optimized through the eLBOW package⁷³ using the semi-empirical quantum mechanical method Austin Model 1 (AM1)⁷³. An iterative process of refinement and model manual correction was performed until optimal stereochemistry and geometry were calculated by MolProbity⁷⁴. The FSC curve between the refined model and half maps was computed and compared to avoid overfitting. Molecular graphics images were produced using the UCSF Chimera package from the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco (supported by NIH P41 RR-01081).

Fluorescence spectral shift binding experiments

Covalent labeling of lysine residues of SERT was performed using the Protein Labeling Kit RED-NHS 2^nd Generation (cat# MO-L011; NanoTemper Technologies GmbH, Munich, Germany). In brief, frozen samples of SERT were thawed and buffer-exchanged from 20 mM Tris, 300 mM NaCl, 0.05% GDN, pH 8 to 20 mM HEPES, 150 mM KCl, 0.05% GDN, pH 7.5, using the A-column from the labeling kit (NanoTemper Technologies GmbH, Munich, Germany). Next, SERT samples were incubated at RT for 2 h or ON at 4 °C. Approximate SERT concentration was 5 μΜ. SERT was labeled with a 5-fold molar excess of dye for 30 min at RT. Unreacted dye was removed from labeled SERT by performing SEC using Superdex 200 SEC column on an AKTA Pure FPLC system (Cytiva), with 20 mM HEPES, 200 mM NaCl, and 0.005% GDN as an elution buffer. Fractions were pooled based on the 280 nm absorbance profile from SEC. The degree of labeling (DOL), indicating the number of dye molecules attached to SERT, was calculated using the sample’s absorbance at 280 nm and 650 nm - equal to 3.7. Eluted SERT was stored at 4 °C until further use. For the equilibrium experiments, 2 nM of SERT was incubated in 20 mM HEPES, 0.05% GDN, 0.4% DMSO, and 150 mM NaCl, KCl or NMDGCl and the tested ligand (vilazodone, (S)-citalopram and paroxetine) in 16 serial-dilution concentrations from 0.012 nM to 400 nM. Samples were equilibrated for 2 h at RT to reach equilibrium and measurements were recorded using Dianthus (NanoTemper Technologies GmbH, Munich, Germany). Dianthus records fluorescence emission intensities at 670 nm and 650 nm upon excitation at 590 nm. Spectral Shift is registered as a ratio of these intensities.

The K_D is estimated by fitting the Eq. (1):

$$f\left(c\right)={{{\rm{Unbound}}}}+\left({{{\rm{Bound}}}}-{{{\rm{Unbound}}}}\right)\,x \frac{C+C{{{\rm{target}}}}+{Kd}\sqrt{({C+C{{{\rm{target}}}}+{Kd}})^{2}-4C\,C{{{\rm{target}}}}}}{2C{{{\rm{target}}}}}$$

(1)

Where $f\left(c\right)$ is the signal at a given ligand concentration $C$, ${{{\rm{Unbound}}}}$ is the ratio signal of the target alone, ${{{\rm{Bound}}}}$ is the ratio signal from the complex, ${Kd}$ is the dissociation constant or binding affinity and $C{{{\rm{target}}}}$ is the final concentration of target in the experiment.

Anap fluorescence-based experiments

For the equilibrium measurements, fluorescence emission spectra for V86Anap and F556Anap were recorded as described previously⁵⁴. Briefly, SERT was centrifuged for 10 min at 4 °C and diluted to 10 or 4 nM of V86Anap or F556Anap, respectively, in 20 mM TrisCl, pH 8, 10 % v/v glycerol, 500 µM TCEP, 1 mM DDM, 0.2 mM CHS, 24 µM lipids (POPC, POPE, POPG; molar stoichiometry 1:1:1), supplemented with 200 mM NaCl or KCl and with or without 200 nM vilazodone, CDV-3-1, or CDV-3-2. Samples were incubated for 30 min at 19 °C shielded from light before fluorescence emission spectra (excitation at 360 nm; emission at 375–550 nm) were recorded on a FluoroMax−4 spectrofluorometer (HORIBA Scientific). Spectra obtained in the absence of protein were subtracted from those obtained in the presence of protein. Of note, DMSO from the ligand stocks was diluted to ≤0.02%. Control spectra for 5 nM free Anap, incubated under the same conditions as those for the SERT mutants, were measured in parallel. Experiments were repeated with protein from ≥2 transfections and purifications.

The real-time fluorescence measurements were recorded on a FluoroMax-4 spectrofluorometer (HORIBA Scientific) at 19 °C using a 5 × 5 mm quartz cuvette (Hellma). Samples were excited at 360 nm and emission was measured at 450 nm, using 9 nm excitation- and emission slit widths. Fluorescence intensities were measured as reference-corrected signals (S1c/R1c) with 0.5 sec time increments and 0.2 sec integration time. SERT F556Anap was centrifuged for 10 min at 4 °C and diluted to 15 nM in fluorescence buffer supplemented with 200 mM NaCl. Following equilibration in NaCl, samples were mixed 1:1 with buffer with or without 400 nM vilazodone or imipramine (200 nM final concentration) upon initiation of the fluorescence measurements. To test vilazodone binding to the allosteric site, we pre-incubated 100 nM of imipramine with SERT for 30 min at RT and initiated vilazodone association by the addition of 400 nM vilazodone (200 nM final concentration. To correct for photobleaching, the fluorescence traces obtained without ligand were subtracted from those obtained with vilazodone or imipramine. These data were normalized to the initial fluorescence (at time 0) and fitted by a one-phase exponential model. Experiments were repeated with protein from ≥2 transfections and purifications.

Chemistry

Vilazodone fragments (CDV-2-29, CDV-2-26, CDV-2-32, CDV-2-81, CDV-3-2, and CDV-3-1) were prepared by adapting procedures found in the literature^75,76,77. Experimental details and characterization data for all compounds, including intermediates, are provided in the Supplementary Methods. pKa calculations for vilazodone and the fragments were performed using the Maestro software tool (Schrödinger Release 2020-3: Schrödinger, LLC, New York, NY, 2020), pKa was calculated using Epik⁶³.

Statistical analysis

All data were plotted and analyzed using GraphPad Prism 10.1. Affinities derived from logarithmic analyses are expressed as mean [S.E.M. interval]. To account for the logarithmic nature of IC₅₀, statistical analyses were conducted on log-transformed values, ensuring normalized distribution and robustness in comparison. The statistical analysis involving comparisons of more than two means were performed using one-way ANOVA with Dunnett or Tukey multiple comparison as the post-hoc test. All experiments were repeated at least three times using cells or protein samples from at least two independent preparations.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The cryo-EM reconstruction has been deposited in the Protein Data Bank under the accession code 9HCO and the corresponding EM map has been deposited in the Electron Microscopy Data Bank under the accession code EMD-52050. All data generated in this study are provided in the Supplementary Information and Source Data. All data, as well as the associated metadata, that support the findings in this study are also available on request from the corresponding author. Source data are provided with this paper.

References

World Health Organization. Depression and Other Common Mental Disorders: Global Health Estimates. World Health Organization (2017).
Garnock-Jones, K. P. & McCormack, P. L. Escitalopram: A review of its use in the management of major depressive disorder in adults. CNS Drugs 24, 769–796 (2010).
Emslie, G. J., Mayes, T. L. & Ruberu, M. Continuation and maintenance therapy of early-onset major depressive disorder. Pediatr. Drugs 7, 203–217 (2005).
Pundiak, T. M., Case, B. G., Peselow, E. D. & Mulcare, L. Discontinuation of maintenance selective serotonin reuptake inhibitor monotherapy after 5 years of stable response: a naturalistic study. J. Clin. Psychiatry 69, 1811–1817 (2008).
Article PubMed Google Scholar
Edinoff, A. N. et al. Selective serotonin reuptake inhibitors and adverse effects: a narrative review. Neurol. Int. 13, 387–401 (2021).
Locher, C. et al. Efficacy and safety of selective serotonin reuptake inhibitors, serotonin-norepinephrine reuptake inhibitors, and placebo for common psychiatric disorders among children and adolescents: a systematic review and meta-analysis. JAMA Psychiatry 74, 1011–1020 (2017).
Article PubMed PubMed Central Google Scholar
Forns, J. et al. Antidepressant use in Denmark, Germany, Spain, and Sweden between 2009 and 2014: Incidence and comorbidities of antidepressant initiators. J. Affect Disord. 249, 242–252 (2019).
Article PubMed Google Scholar
Niarchou, E., Roberts, L. H. & Naughton, B. D. What is the impact of antidepressant side effects on medication adherence among adult patients diagnosed with depressive disorder: a systematic review. J. Psychopharmacol. 38, 127–136 (2024).
Kennedy, S. H., Eisfeld, B. S., Dickens, S. E., Bacchiochi, J. R. & Bagby, R. M. Antidepressant-induced sexual dysfunction during treatment with moclobemide, paroxetine, sertraline, and venlafaxine. J. Clin. Psychiatry 61, 276–281 (2000).
Article PubMed CAS Google Scholar
Starr, K. R. et al. SB-649915-B, a novel 5-HT1A/B autoreceptor antagonist and serotonin reuptake inhibitor, is anxiolytic and displays fast onset activity in the rat high light social interaction test. Neuropsychopharmacology 32, 2163–2172 (2007).
Article PubMed CAS Google Scholar
Atkinson, P. J. et al. 3,4-Dihydro-2H-benzoxazinones are 5-HT1A receptor antagonists with potent 5-HT reuptake inhibitory activity. Bioorg. Med Chem. Lett. 15, 737–741 (2005).
Article PubMed CAS Google Scholar
Rickels, K. et al. Evidence for efficacy and tolerability of vilazodone in the treatment of major depressive disorder: A randomized, double-blind, placebo-controlled trial. J. Clin. Psychiatry 70, 326–333 (2009).
Article PubMed CAS Google Scholar
Mathews, M., Gommoll, C., Chen, D., Nunez, R. & Khan, A. Efficacy and safety of vilazodone 20 and 40mg in major depressive disorder: A randomized, double-blind, placebo-controlled trial. Int Clin. Psychopharmacol. 30, 67–74 (2015).
Article PubMed PubMed Central Google Scholar
Bathla, M., Anjum, S., Singh, M., Panchal, S. & Singh, G. A 12-week comparative prospective open-label randomized controlled study in depression patients treated with vilazodone and escitalopram in a tertiary care hospital in North India. Indian J. Psychol. Med. 40, 80–85 (2018).
Article PubMed PubMed Central Google Scholar
Robinson, D. S. et al. A 1-year, open-label study assessing the safety and tolerability of vilazodone in patients with major depressive disorder. J. Clin. Psychopharmacol. 31, 643–646 (2011).
Article PubMed CAS Google Scholar
Hughes, S., Lacasse, J., Fuller, R. R. & Spaulding-Givens, J. Adverse effects and treatment satisfaction among online users of four antidepressants. Psychiatry Res. 255, 78–86 (2017).
Article PubMed CAS Google Scholar
Berger, M., Gray, J. A. & Roth, B. L. The expanded biology of serotonin. Annu. Rev. Med. 60, 355–366 (2009).
Gether, U., Andersen, P. H., Larsson, O. M. & Schousboe, A. Neurotransmitter transporters: molecular function of important drug targets. Trends Pharmacol. Sci. 27, 375–383 (2006).
Kristensen, A. S. et al. SLC6 neurotransmitter transporters: Structure, function, and regulation. Pharm. Rev. 63, 585–640 (2011).
Article PubMed CAS Google Scholar
Saier, M. H., Tran, C. V. & Barabote, R. D. TCDB: the transporter classification database for membrane transport protein analyses and information. Nucleic Acids Res. 34, D181–6 (2006).
Yamashita, A., Singh, S. K., Kawate, T., Jin, Y. & Gouaux, E. Crystal structure of a bacterial homologue of Na+/Cl –dependent neurotransmitter transporters. Nature 437, 215–223 (2005).
Article ADS PubMed CAS Google Scholar
Drew, D. & Boudker, O. Shared molecular mechanisms of membrane transporters. Annu Rev. Biochem 85, 543–572 (2016).
Article PubMed CAS Google Scholar
Mitchell, P. A general theory of membrane transport from studies of bacteria. Nature 180, 134–136 (1957).
Article ADS PubMed CAS Google Scholar
Jardetzky, O. Simple allosteric model for membrane pumps. Nature 211, 969–970 (1966).
Yang, D. & Gouaux, E. Illumination of serotonin transporter mechanism and role of the allosteric site. Sci. Adv. 7, 3857 (2021).
Article ADS Google Scholar
Rudnick, G. & Sandtner, W. Serotonin transport in the 21st century. J. Gen. Physiol. 151, 1248–1264 (2019).
Rudnick, G. & Nelson, P. J. Platelet 5-hydroxytryptamine transport, an electroneutral mechanism coupled to potassium. Biochemistry 17, 4739–4742 (1978).
Article PubMed CAS Google Scholar
Sørensen, L. et al. Interaction of antidepressants with the serotonin and norepinephrine transporters: Mutational studies of the S1 substrate binding pocket. J. Biol. Chem. 287, 43694–43707 (2012).
Article PubMed PubMed Central Google Scholar
Pedersen, A. V., Andreassen, T. F. & Loland, C. J. A conserved salt bridge between transmembrane segments 1 and 10 constitutes an extracellular gate in the dopamine transporter. J. Biol. Chem. 289, 35003-14 (2014).
Szöllősi, D. & Stockner, T. Investigating the mechanism of sodium binding to SERT using direct simulations. Front Cell Neurosci. 15, 673782 (2021).
Coleman, J. A., Green, E. M. & Gouaux, E. X-ray structures and mechanism of the human serotonin transporter. Nature 532, 334–339 (2016).
Article ADS PubMed PubMed Central CAS Google Scholar
Felts, B. et al. The two Na+ sites in the human serotonin transporter play distinct roles in the ion coupling and electrogenicity of transport. J. Biol. Chem. 289, 1825–1840 (2014).
Article PubMed CAS Google Scholar
Claxton, D. P. et al. Ion/substrate-dependent conformational dynamics of a bacterial homolog of neurotransmitter:sodium symporters. Nat. Struct. Mol. Biol. 17, 822–829 (2010).
Article PubMed PubMed Central CAS Google Scholar
Tavoulari, S. et al. Two Na+ sites control conformational change in a neurotransmitter transporter homolog. J. Biol. Chem. 291, 1456–1471 (2016).
Article PubMed CAS Google Scholar
Zhao, C. & Noskov, S. Y. The role of local hydration and hydrogen-bonding dynamics in ion and solute release from ion-coupled secondary transporters. Biochemistry 50, 1848–1856 (2011).
Article PubMed CAS Google Scholar
Li, J., Zhao, Z. & Tajkhorshid, E. Locking two rigid-body bundles in an outward-facing conformation: the ion-coupling mechanism in a leut-fold transporter. Sci. Rep. 9, 19479 (2019).
Stolzenberg, S. et al. Mechanism of the association between Na+ binding and conformations at the intracellular gate in neurotransmitter:sodium symporters. J. Biol. Chem. 290, 13992–14003 (2015).
Article PubMed PubMed Central CAS Google Scholar
Nelson, P. J. & Rudnick, G. Coupling between platelet 5-hydroxytryptamine and potassium transport. J. Biol. Chem. 254, 10084–10089 (1979).
Article PubMed CAS Google Scholar
Schicker, K. et al. Unifying concept of serotonin transporter-associated currents. J. Biol. Chem. 287, 438–445 (2012).
Article PubMed CAS Google Scholar
Coleman, J. A. et al. Serotonin transporter–ibogaine complexes illuminate mechanisms of inhibition and transport. Nature 569, 141–145 (2019).
Article ADS PubMed PubMed Central CAS Google Scholar
Hasenhuetl, P. S., Freissmuth, M. & Sandtner, W. Electrogenic Binding of Intracellular Cations Defines a Kinetic Decision Point in the Transport Cycle of the Human Serotonin Transporter. J. Biol. Chem. 291, 25864–25876 (2016).
Article PubMed PubMed Central CAS Google Scholar
Tavoulari, S., Forrest, L. R. & Rudnick, G. Fluoxetine (Prozac) Binding to Serotonin Transporter Is Modulated by Chloride and Conformational Changes. J. Neurosci. 29, 9635 (2009).
Article PubMed PubMed Central CAS Google Scholar
Wang, H. et al. Structural basis for action by diverse antidepressants on biogenic amine transporters. Nature 503, 141–145 (2013).
Article ADS PubMed PubMed Central CAS Google Scholar
Coleman, J. A. et al. Chemical and structural investigation of the paroxetine-human serotonin transporter complex. Elife 9, 1–38 (2020).
Article Google Scholar
Plenge, P. et al. The antidepressant drug vilazodone is an allosteric inhibitor of the serotonin transporter. Nat. Commun. 12, 1–12 (2021).
Article Google Scholar
Coleman, J. A. & Gouaux, E. Structural basis for recognition of diverse antidepressants by the human serotonin transporter. Nat. Struct. Mol. Biol. 25, 170–175 (2018).
Plenge, P. et al. Steric hindrance mutagenesis in the conserved extracellular vestibule impedes allosteric binding of antidepressants to the serotonin transporter. J. Biol. Chem. 287, 39316–39326 (2012).
Article PubMed PubMed Central CAS Google Scholar
Salomon, K. et al. Dynamic extracellular vestibule of human SERT: Unveiling druggable potential with high-affinity allosteric inhibitors. Proc. Natl. Acad. Sci. USA 120, e2304089120 (2023).
Article PubMed PubMed Central CAS Google Scholar
Niello, M., Gradisch, R., Loland, C. J., Stockner, T. & Sitte, H. H. Allosteric Modulation of Neurotransmitter Transporters as a Therapeutic Strategy. Trends Pharmacol. Sci. 41, 446–463 (2020).
Javitch, J. A. et al. Probing structure of neurotransmitter transporters by substituted- cysteine accessibility method. Methods Enzymol. 296, 331–46 (1998).
Li, M., Chen, Q. & Zhang, Y. W. Determining ligand and ion-induced conformational changes in serotonin transporter with its fluorescent substrates. Int. J. Mol. Sci. 23, 10919 (2022).
Zhang, Y. et al. The binding mode of vilazodone in the human serotonin transporter elucidated by ligand docking and molecular dynamics simulations. Phys. Chem. Chem. Phys. 22, 5132–5144 (2020).
Article PubMed CAS Google Scholar
Weiss, J. N. et al. The Hill equation revisited: uses and misuses. FASEB J. 11, 835–841 (1997).
Nygaard, A., Zachariassen, L. G., Larsen, K. S., Kristensen, A. S. & Loland, C. J. Fluorescent non-canonical amino acid provides insight into the human serotonin transporter. Nat. Commun. 15, 1–14 (2024).
Article ADS Google Scholar
Andersen, J. et al. Location of the antidepressant binding site in the serotonin transporter: Importance of Ser-438 in recognition of citalopram and tricyclic antidepressants. J. Biol. Chem. 284, 10276–10284 (2009).
Article PubMed PubMed Central CAS Google Scholar
Hofmann, C. et al. Efficient gene transfer into human hepatocytes by baculovirus vectors. Proc. Natl. Acad. Sci. USA 92, 10099-103 (1995).
Esendir, E. et al. Extracellular loops of the serotonin transporter act as a selectivity filter for drug binding. J. Biol. Chem. 297, 100863 (2021).
Article PubMed PubMed Central CAS Google Scholar
Yang, D., Zhao, Z., Tajkhorshid, E. & Gouaux, E. Structures and membrane interactions of native serotonin transporter in complexes with psychostimulants. Proc. Natl. Acad. Sci. USA 120, e2304602120 (2023).
Forrest, L. R., Tavoulari, S., Zhang, Y. W., Rudnick, G. & Honig, B. Identification of a chloride ion binding site in Na+/Cl –dependent transporters. Proc. Natl. Acad. Sci. USA 104, 12761–12766 (2007).
Article ADS PubMed PubMed Central CAS Google Scholar
Davis, B. A., Nagarajan, A., Forrest, L. R. & Singh, S. K. Mechanism of Paroxetine (Paxil) Inhibition of the Serotonin Transporter. Sci. Rep. 6, 1–13 (2016).
Article Google Scholar
Andersen, J. et al. Binding of the multimodal antidepressant drug vortioxetine to the human serotonin transporter. ACS Chem. Neurosci. 6, 1892–900 (2015).
Langer, A. et al. A new spectral shift-based method to characterize molecular interactions. Assay. Drug Dev. Technol. 20, 83–94 (2022).
Article PubMed PubMed Central CAS Google Scholar
Shelley, J. C. et al. Epik: A software program for pKa prediction and protonation state generation for drug-like molecules. J. Comput Aided Mol. Des. 21, 681–691 (2007).
Article ADS PubMed CAS Google Scholar
El-Kasaby, A. et al. Allosteric inhibition and pharmacochaperoning of the serotonin transporter by the antidepressant drugs trazodone and nefazodone. Mol. Pharmacol. 106, 56–70 (2024).
Freissmuth, M., Stockner, T. & Sucic, S. Slc6 Transporter Folding Diseases And Pharmacochaperoning. in Handbook of Experimental Pharmacology 245, 249–270 (Springer New York LLC, 2018).
Shimazu, S. et al. Transporter-mediated actions of R-(-)-1-(benzofuran-2-yl)-2- propylaminopentane. Eur. J. Pharm. 482, 9–16 (2003).
Article CAS Google Scholar
Støier, J. F. et al. Disruptive mutations in the serotonin transporter associate serotonin dysfunction with treatment-resistant affective disorder. medRxiv https://doi.org/10.1101/2023.08.29.23294386 (2023).
Article PubMed PubMed Central Google Scholar
Goehring, A. et al. Screening and large-scale expression of membrane proteins in mammalian cells for structural studies. Nat. Protoc. 9, 2574–2585 (2014).
Article PubMed PubMed Central CAS Google Scholar
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).
Article PubMed CAS Google Scholar
Pettersen, E. F. et al. UCSF Chimera - a visualization system for exploratory research and analysis. J. Comput Chem. 25, 1605–1612 (2004).
Article ADS PubMed CAS Google Scholar
Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and development of Coot. Acta Crystallogr D. Biol. Crystallogr 66, 486–501 (2010).
Article ADS PubMed PubMed Central CAS Google Scholar
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).
Article ADS PubMed PubMed Central CAS Google Scholar
Moriarty, N. W., Grosse-Kunstleve, R. W. & Adams, P. D. Electronic ligand builder and optimization workbench (eLBOW): A tool for ligand coordinate and restraint generation. Acta Crystallogr D. Biol. Crystallogr 65, 1074–1080 (2009).
Article ADS PubMed PubMed Central CAS Google Scholar
Chen, V. B. et al. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr D. Biol. Crystallogr 66, 12–21 (2010).
Article ADS PubMed CAS Google Scholar
Che, Z. et al. Synthesis and quantitative structure-activity relationship (QSAR) study of novel N-arylsulfonyl-3-acylindole arylcarbonyl hydrazone derivatives as nematicidal agents. J. Agric. Food Chem. 61, 5696–705 (2013).
Hu, B., Song, Q. & Xu, Y. Scale-up synthesis of antidepressant drug vilazodone. Org. Process Res. Dev. 16, 1552–1557 (2012).
Abdel-Magid, A. F., Carson, K. G., Harris, B. D., Maryanoff, C. A. & Shah, R. D. Reductive amination of aldehydes and ketones with sodium triacetoxyborohydride. Studies on direct and indirect reductive amination procedures. J. Org. Chem. 61, 3849–3862 (1996).

Download references

Acknowledgements

We thank L. Rosenquist for technical assistance. The Fab15B8 was kindly provided by HHMI Eric Gouaux. The financial support for this work was provided by the Novo Nordisk Foundation (NNF22OC0079091 to C.J.L.); the Independent Research Fund Denmark (1030-00036B to C.J.L.), the Carlsberg Foundation (CF23-1034 to C.J.L.) and NIDA intramural Research (Z1A DA000389 to AHN). I.E.K. is the recipient of a H2020 Marie Sklodowska-Curie training network (Program number 860954). Cryo-EM data collection was performed at the Danish Cryo-EM Facility at the Core Facility for Integrated Microscopy (CFIM), University of Copenhagen supported by the Novo Nordisk Foundation (grants NNF17SA0024386 and NNF22OC0075808).

Author information

Authors and Affiliations

Department of Neuroscience, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark
Iris E. Kalenderoglou, Andreas Nygaard & Claus J. Loland
Medicinal Chemistry Section, Molecular Targets and Medications Discovery Branch, National Institute on Drug Abuse – Intramural Research Program, National Institutes of Health, 333 Cassell Drive, Baltimore, MD, US
Caleb D. Vogt & Amy Hauck Newman
NanoTemper Technologies, Floessegasse 4, Munich, Germany
Anton Turaev & Nathan B. P. Adams
Structural Molecular Biology Group, Protein Structure & Function Program, Novo Nordisk Foundation Center for Protein Research, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark
Tillmann Pape
Core Facility for Integrated Microscopy (CFIM), Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark
Tillmann Pape

Authors

Iris E. Kalenderoglou
View author publications
Search author on:PubMed Google Scholar
Andreas Nygaard
View author publications
Search author on:PubMed Google Scholar
Caleb D. Vogt
View author publications
Search author on:PubMed Google Scholar
Anton Turaev
View author publications
Search author on:PubMed Google Scholar
Tillmann Pape
View author publications
Search author on:PubMed Google Scholar
Nathan B. P. Adams
View author publications
Search author on:PubMed Google Scholar
Amy Hauck Newman
View author publications
Search author on:PubMed Google Scholar
Claus J. Loland
View author publications
Search author on:PubMed Google Scholar

Contributions

The study was conceptualized and designed by I.E.K., A.H.N. and C.J.L. I.E.K., A.N., and A.T. performed experiments with assistance from N.B.P.A. T.P. assisted with cryo-EM grid preparation and performed the cryo-EM data collection with I.E.K. I.E.K. performed the cryo-EM computational analysis with initial support from T.P. Data were analyzed and interpreted by I.E.K., A.N. and C.J.L. with support from A.T. and N.B.P.A. C.D.V. and A.H.N. provided vilazodone and the designed and deconstructed analogues. The manuscript was written by I.E.K. and C.J.L. with important contributions from A.N., C.D.V. and A.H.N. All authors read and commented on the manuscript.

Corresponding author

Correspondence to Claus J. Loland.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Michael Freissmuth and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Reporting Summary

Transparent Peer Review file

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Kalenderoglou, I.E., Nygaard, A., Vogt, C.D. et al. Structural basis of vilazodone dual binding mode to the serotonin transporter. Nat Commun 16, 10119 (2025). https://doi.org/10.1038/s41467-025-65202-3

Download citation

Received: 27 December 2024
Accepted: 07 October 2025
Published: 18 November 2025
Version of record: 18 November 2025
DOI: https://doi.org/10.1038/s41467-025-65202-3