5-hydroxytryptamine (5-HT, serotonin) is a monoamine neurotransmitter involved in the modulation of a myriad of functions in the brain and periphery, including mood, appetite, sleep, memory and learning. The magnitude and duration of serotonin signaling is primarily controlled by rapid clearance of extracellular serotonin via the high-affinity serotonin transporter (SERT; SLC6A4) (Blakely et al., 1991; Ramamoorthy et al., 1993; Torres et al., 2003). SERT belongs to a family of twelve transmembrane-spanning domain neurotransmitter::sodium symporters (NSS) that harness the electrochemical gradient of Na⁺ and Cl^– to translocate substrate through a series of structural rearrangements from outward-open to inward-open conformation (Coleman et al., 2016, 2019, 2020; Cheng and Bahar, 2019). Alterations in SERT activity and expression have been implicated in several neuropsychiatric disorders, including anxiety, autism spectrum disorder (ASD), obsessive-compulsive disorder (OCD), attention deficit/hyperactivity disorder (ADHD), and major depressive disorder (MDD) (Kristensen et al., 2011). Accordingly, SERT is an important pharmacological target for widely prescribed selective serotonin reuptake inhibitors (SSRIs), including paroxetine, citalopram, and fluoxetine (Coleman and Gouaux, 2018), which are used to treat MDD, OCD, and anxiety disorders. Moreover, single nucleotide polymorphisms (SNPs) in the SERT gene can affect the expression, membrane trafficking and activity of the transporter and are enriched in families with neuropsychiatric disorders (Hu et al., 2006; Hahn and Blakely, 2007). SERT activity is tightly regulated at the post-translational level by a complex interplay of kinases, phosphatases, and numerous other SERT interacting proteins (Ramamoorthy et al., 1998; Bermingham and Blakely, 2016; Cooper et al., 2019; Quinlan et al., 2020). Consequently, methodologies that can dissect SERT regulation at the single transporter level represent a promising experimental avenue toward understanding the molecular basis of SERT-associated neuropsychiatric disorders.

Single-molecule fluorescence microscopy is an umbrella term for non-invasive imaging modalities that have emerged in the investigation of transmembrane protein trafficking and regulation, as they can access nanometer spatial and sub-second temporal scales (Rosenthal et al., 2011; Kusumi et al., 2014). Typically, an antibody, a peptide or a derivative of a drug or neurotransmitter are conjugated to a fluorophore and utilized as a probe for optical detection of a membrane protein. The fluorophore is usually separated from the targeting functionality via a spacer arm; this is often required to maintain biological activity of the targeting portion of the probe and may also aid in water solubility. Several dye-conjugated antagonists of neuronal receptors and transporters have been reported in the literature (Madras et al., 1990; Bakthavachalam et al., 1991; Jacobson et al., 1995; Rayport and Sulzer, 1995; Cha et al., 2005; Eriksen et al., 2009; She et al., 2020). However, many commonly used fluorescent dyes have low quantum yields and suboptimal photostability, thereby limiting their utility in experiments that require prolonged monitoring of protein membrane dynamics at video rates (Resch-Genger et al., 2008). Biocompatible colloidal semiconductor quantum dots (QDs) are an attractive alternative to dye fluorophores (Rosenthal et al., 2011). These highly fluorescent semiconductor crystals range in diameter from 2 to 10 nm, have a quantum yield that is significantly larger than many commercial dyes and a narrow, size-tunable, Gaussian emission spectrum. Their large absorption cross section is a continuum above the 1st excitation feature, enabling the use of a single excitation source for multicolor experiments. Many studies using antibody- and peptide-conjugated quantum dots in biological labeling of live cells have been reported (Dahan et al., 2003; Bouzigues et al., 2007; Lee et al., 2017; Bailey et al., 2018). The cornerstone of our QD labeling strategy has been the development of agonist- and antagonist-based fluorescent probes. Unlike the antibody-based approach, such ligand probes do not require a suitable extracellular epitope or an engineered tag in the extracellular domain of the target protein for binding; also, their affinity and binding properties may be modified by the incorporation of simple structural changes into the ligand (Tomlinson et al., 2019). To date, we have successfully developed antagonist-based QD probes to label SERT and other membrane-bound neuronal proteins, including structurally related dopamine transporter (DAT) (Kovtun et al., 2011; Chang et al., 2012b; Kovtun et al., 2015; Rosenthal, 2019; Tomlinson et al., 2019).

Over the past two decades, we have devised several strategies in probe development, including direct attachment of the ligand to a polymer on the QD surface via a covalent bond (Gussin et al., 2006) and ligands that incorporate a biotin moiety for non-covalent binding to streptavidin on the surface of the dot (Kovtun et al., 2011; Tomlinson et al., 2019). We have observed that conjugation of polyethylene glycol (PEG; PEGylation) to the QD surface may be required to overcome non-specific binding to the cell surface (Bentzen et al., 2005) and have incorporated PEG chains into our ligands for this purpose. Currently, we rely on commercially available streptavidin-conjugated QDs that display lower non-specific binding properties. The linker arm of choice in our probes has been a polyethylene glycol (PEG) chain attached to the antagonist via an alkyl spacer, with the biotinylated distal end of the PEG forming the point of attachment to the QD (Tomlinson et al., 2011). Recently, we have observed that the nature of the spacer between the PEG chain and the antagonist can have profound effects on non-specific coverslip binding (Tomlinson et al., 2019). Here, we report synthesis of a novel SERT ligand IDT785 (Figure 1) with an electron-rich phenylethyl spacer between the PEG chain and the antagonist. We study its ability to inhibit the human SERT transiently expressed in HEK-293T cells (hSERT), examine specific labeling of hSERT and SERT fused to yellow fluorescent protein (YFP) at the N-terminus with a combination of biotinylated IDT785 streptavidin-conjugated QDs (SavQD), and demonstrate that hSERT surface dynamics may be tracked with IDT785-QD conjugates.

Over the past two decades, ultrastructural electron microscopy and immunofluorescence approaches have revealed a unique pattern of SERT distribution in the plasma membrane in both brain- and periphery-derived cultures (Zhou et al., 1998; Tao-Cheng and Zhou, 1999; Miner et al., 2000; Steiner et al., 2008; Montgomery et al., 2014). The molecular mechanisms that enable optimal SERT positioning for rapid clearance of extracellular serotonin remain largely unknown, and trafficking-dependent regulation of SERT function has been an area of active investigation. The study of SERT trafficking in live cell cultures has been significantly limited by the lack of suitable antibodies that target the extracellular domain of the transporter. To combat this challenge, the Rosenthal laboratory pioneered the use of small-molecule neurotransmitter or antagonist derivatives to enable real-time monitoring of SERT trafficking at the cell surface (Rosenthal et al., 2002; Tomlinson et al., 2011; Chang et al., 2012b). Our current probe design features an attachment of a biotinylated, PEGylated spacer to a transporter-specific antagonist, which can then be readily detected by streptavidin-conjugated QDs and facilitate single QD tracking of transporter membrane dynamics. In this approach, the point of spacer attachment must be carefully chosen to preserve parent antagonist binding affinity for SERT; moreover, the spacer identity is critically important for stabilizing the antidepressant-transporter interaction as well as minimizing non-specific ligand binding to the plasma membrane or coverslip. In particular, our recent report demonstrated that replacing a flexible alkyl (undecyl) spacer with a more rigid, electron-rich, less hydrophobic phenylethyl spacer in the structure of biotinylated D2 dopamine receptor ligand significantly reduced non-specific ligand binding to the coverslip surface (Tomlinson et al., 2019). Thus, our goal in the current study was to incorporate the phenylethyl spacer into the SERT ligand design and determine its effect on parent drug affinity, ligand binding specificity, and consequently the ability to monitor surface SERT dynamics with ligand-conjugated QDs.

Modification of the pyridinyl nitrogen in the parent drug 3-(1,2,3,6-tetrahydropyridin-4-yl)-1H-indole was found to be relatively well-tolerated (Deskus et al., 2007; Tomlinson et al., 2011). The methyl-substituted analog 3-(1-methyl-1,2,3,6-tetrahydropyridin-4-yl)-1H-indole was determined to inhibit [³H]citalopram binding to hSERT in the membrane homogenates of stably transfected HEK-293 cells with the IC₅₀ value of 690 ± 270 nM (Deskus et al., 2007). Our concerns were that the steric constraints of the extracellular SERT vestibule might not support the bulkier phenylethyl spacer extending into the extracellular space and slightly lower hydrophobicity of the aromatic ring compared to an alkyl chain might reduce binding stability. To measure the binding affinity of the biotinylated, PEGylated 3-(1-phenethyl-1,2,3,6-tetrahydropyridin-4-yl)-1H-indole (IDT785), an IDT307 (APP+) fluorescent substrate transport assay was employed. IDT307 is actively transported across the membrane via monoamine neurotransmitter transporters (SERT, DAT, NET) as well as organic cation transporters 1–3 (OCT 1–3) and permits rapid characterization of transporter-inhibitor interaction in a 96-well plate format (Duan et al., 2015). The IC₅₀ values determined by this method show minimal plate-to-plate or day-to-day variability. hSERT-mediated uptake in transiently transfected HEK-293T cells was effectively inhibited by excess (100 μM) IDT785 and 10 μM paroxetine (negative control) (Figure 2), indicating that the phenylethyl spacer did not adversely affect the biological activity of 3-(1,2,3,6-tetrahydropyridin-4-yl)-1H-indole. The IC₅₀ of IDT785 was determined to be 7.2 ± 0.3 μM using IDT307-based inhibition assay, comparable to our previously reported compounds based on 3-(1,2,3,6-tetrahydropyridin-4-yl)-1H-indole featuring the undecyl linker (Tomlinson et al., 2011). Admittedly, the IC₅₀ of the final biotinylated ligand is quite high and an order of magnitude larger than the IC₅₀ of the drug intermediate before the PEGylation step (Supplementary Figure 4). Our next goal is to achieve sub-μM affinity before we proceed with SERT tracking experiments in neuronal cultures or intact tissue. At high ligand doses, one might anticipate several potential undesirable off-target effects, including activation of serotonin receptors [the parent drug of IDT785 is a derivative of RU-24969, a preferential 5-HT_1B agonist, with a K_i of 0.38 nM, which also displays appreciable affinity for the 5-HT_1A receptor (K_i = 2.5 nM) (Aronsen et al., 2014)] or incorporation of the lipophilic moiety of IDT785 into membrane rafts (Erb et al., 2016). These off-target effects can in turn influence SERT surface state by inducing changes in membrane polarization through the activation of membrane ion channels, G protein-mediated stimulation of signaling pathways that control SERT trafficking/phosphorylation/palmitoylation, or local disruption of membrane rafts that are known to stabilize transport-willing SERT conformation. Next, a B4F quenching assay was performed to rule out the possibility (although unlikely) that the PEG chain sterically hindered the availability of the IDT785 biotin end for binding to Sav conjugated to QDs. Quenching of B4F fluorescence upon free Sav or SavQD binding is a well-documented phenomenon, with a nearly 90% B4F fluorescence reduction achieved at B4F:Sav molar ratios ≤ 4:1 (Mittal and Bruchez, 2011; Lippert et al., 2016). Indeed, blocking of QD655Sav with either free biotin or biotinylated IDT785 effectively prevented Sav-mediated B4F quenching, suggesting that the biotin end of IDT785 was readily available for QD655Sav binding (Figure 3).

Molecular mechanisms of SSRI interactions with SERT have been of great interest in the pursuit of drugs with greater specificity and fewer adverse side effects. As a result of extensive efforts, X-ray and cryo-EM structures of SSRI-bound SERT are now publicly available (Coleman and Gouaux, 2018). The general consensus is that high-affinity SSRIs occupy the S1 central-binding site composed of subsites A, B, and C and stabilize the transporter in an outward-facing conformation, although ambiguity remains concerning the precise binding poses of SSRIs within S1 (Coleman et al., 2020). A secondary, allosteric site S2 is formed above S1 in the extracellular vestibule of the transporter in the outward-facing conformation, permitting modulation of S1 ligand dissociation kinetics. Previously, docking of 3-(1,2,3,6-tetrahydropyridin-4-yl)-1H-indole derivative to the vestibular binding pocket of hSERT model based on the bacterial homolog LeuT showed that the indole docked deep into the vestibular pocket and the S2 substrate site, whereas tetrahydropyridine N-substituents extended out to the extracellular environment (Nolan et al., 2011). Surprisingly, molecular docking performed via user-friendly Autodock Vina in UCSF Chimera revealed that the best-scoring poses of the parent drug of IDT785 attached to the phenylethyl spacer and a short PEG₂ chain were oriented in such manner that the binding site of the indole moiety overlapped with the central binding site occupied by paroxetine in the cryo-EM SERT structure and the tetrahydropyridine N-substituent extended into the extracellular space, thus supporting the importance of tetrahydropyridine indole portion for SERT binding and consequently enabling capture of extracellular biotin. A separate cluster of lower-scoring binding poses was located in the outer extracellular vestibule partially formed by EL2 and EL4 of SERT. Extending distal PEG₂ spacer to PEG chain with the number of repeat units n ≥ 3 resulted in a loss of indole access to the central binding site and the best-scoring binding poses (<7 kcal/mol; data not shown) located in the outer extracellular vestibule. This suggests that the introduction of a longer PEG spacer might prevent the indole moiety from accessing deeper into the extracellular vestibule and substrate sites S1 and S2, thus resulting in modest SERT binding affinity as determined in IDT307-based uptake inhibition experiments.

Since biological activity and availability for binding of both ends of IDT785 was verified, the next step was to assess whether IDT785 facilitated specific QD labeling of surface SERT proteins. To this end, wild-type hSERT and YFP-tagged hSERT constructs were transiently expressed in HEK-293T cells and then labeled with biotinylated IDT785 (10 μM or ∼1.4 × its IC₅₀ value) and 0.1 nM QD655Sav stepwise; subnanomolar QD concentration was chosen as non-specific binding of commercially available streptavidin-conjugated QDs is well-controlled below 1 nM (Bentzen et al., 2005; Rosenthal et al., 2011) and is thus preferred for single QD tracking assays. Imaging of hSERT-expressing cells revealed sparse QD labeling associated with the plasma membrane that was considerably blocked in the presence of paroxetine (Figure 5). This observation was confirmed by imaging of HEK-293T cells transiently expressing YFP-tagged hSERT and incubated with IDT785 and QD655Sav at identical conditions (Figure 6). YFP-SERT signal was particularly useful for identifying a fraction of QDs non-specifically adsorbed to the coverslip. Minimal non-specific binding of IDT785-QDs to the coverslip and plasma membrane in paroxetine-blocked cells was observed, likely due to the synergistic effect of the reduced surface area, length and hydrophobicity (logP) of the phenethyl spacer compared to the undecyl linker in our earlier ligands. Interestingly, YFP-SERT was not uniformly distributed in the plasma membrane at the coverslip interface when imaged either via spinning disk confocal microscope (Figure 6) or TIRF microscope (Figure 7). This phenomenon allowed us to generate binary maps of YFP-SERT cell surface distribution and quantitate the extent of QD labeling specificity by determining individual QD localization to the YFP-SERT-rich membrane regions. As a result, the total number and fraction of detected QDs that were colocalized with YFP-SERT were significantly greater for transfected HEK-293T cells in the absence of paroxetine compared to paroxetine-blocked cells. Quantitative confirmation of labeling specificity at these conditions (10 μM IDT785 for 10 min; 0.1 nM QD655Sav) allowed us to proceed toward single QD tracking experiments.

Single-molecule fluorescence microscopy has revealed that lateral diffusion is a fundamental feature of post-translational regulation of various transmembrane receptors, transporters, and ion channels (Rosenthal et al., 2011; Kusumi et al., 2014; Kovtun et al., 2018). Our previous QD-based studies of SERT surface dynamics demonstrated that endogenous SERT was preferentially compartmentalized to cholesterol-rich microdomains and was readily mobilized in response to cholesterol depletion and phosphorylation of its intracellular domain in immortalized serotonergic rat neurons and cultured rat dorsal raphe neurons (Chang et al., 2012b; Bailey et al., 2018). QD tracking in stably transfected CHO cells showed that the ASD-associated, hyperphosphorylated G56A hSERT variant exhibited ∼40% faster diffusion rate at the cell surface (Kovtun et al., 2018). To demonstrate that IDT785 enables QD tracking of YFP-SERT, HEK-293T expressing YFP-tagged transporters were labeled with a combination of IDT785 (500 nM) and QD655Sav (0.05 nM). QD movement was monitored for 600 frames at the rate of 10 Hz, and single QD trajectories were reconstructed from the acquired time-lapse image series (Figure 8). A wide range of diffusive behavior of YFP-SERT was readily evident from the shape of reconstructed QD trajectories, and the RD analysis of MSD-Δt curves in comparison to simulated Brownian motion trajectories showed that 85% of QD trajectories were mobile versus 15% of immobilized QDs, although it was not possible to distinguish QDs non-specifically adsorbed to the coverslip and QDs bound to immobile YFP-SERT (Dahan et al., 2003). The slight majority of mobile QD trajectories exhibited simple Brownian diffusion (45%; median D = 0.04 μm²/s), whereas 40% of mobile tracks were determined to undergo restricted diffusion (40%; median D = 0.02 μm²/s), consistent with our previous observation that a large pool of endogenous SERT in the neuronal membrane is confined to cholesterol-rich membrane microdomains (Chang et al., 2012b; Bailey et al., 2018). Naturally, an important question arises - how accurately do QDs measure the rate of lateral diffusion of membrane SERT proteins? In fact, suitability of commercially available QDs for tracking surface dynamics of neuronal receptors and transporters has recently been under scrutiny, as improvements in organic fluorophore photophysics, instrumentation, and stochastic activation-based imaging modalities have simplified implementation of dye-based tracking (Lavis, 2017; Shen et al., 2017). For instance, Lee et al. (2017) discovered that over 90% of AMPA glutamate receptors bound to commercially available QDs (diameter > 20 nm) were extrasynaptic and highly mobile, whereas ∼90% of AMPA receptors labeled with smaller dyes and compact QDs were diffusing in a confined manner in nanodomains within the post-synaptic density (PSD). As another example, Abraham et al. (2017) reported that although QDs could resolve large differences in B cell receptor mobility, larger QDs compared to a small cyanine dye (Cy3) sterically hindered receptor mobility at the cell-coverslip interface, altered the frequency of transitions between fast and slow diffusion states, and reduced the size of transient confinement zones. Recently, ensemble fluorescence recovery after photobleaching (FRAP) measurements performed on monomeric GFP fused to the amino terminus of hSERT heterologously expressed in HEK-293 cells yielded a mobile fraction of 82 ± 8%, with a population diffusion coefficient of 0.151 ± 0.003 μm²/s estimated via direct tracking of individual mGFP-SERT (Anderluh et al., 2017). Single-molecule tracking of mGFP-SERT in transiently transfected CHO cells revealed a major mobile fraction of transporters (85%) characterized by D₁ = 0.30 μm²/s and a minor pool (15%) of mGFP-SERT immobilized during the observation period with D₂ = 0.05 μm²/s (Anderluh et al., 2014). To perform direct comparison, we recorded lateral movement of individual dim YFP-SERT spots in the TIRF mode in HEK-293T cells 6 h post-transfection to ensure low expression density suitable for single molecule tracking. The majority of tracks (63% of 7802 tracks) were determined to be mobile as defined by the previously used D_cutoff = 5 × 10^–4 μm²/s with a lateral diffusion rate (median D) of 0.07 μm²/s (Δt of 50 ms; Supplementary Figure 5). The diffusion rate of the mobile YFP-SERT pool 6 h post-transfection was comparable to that of the mobile QD-tagged YFP-SERT pool imaged ≥ 24 h post-transfection in our study (D_{median, YFP} = 0.066 μm²/s [0.017–0.16 IQR]; D_{median, Qdot} = 0.030 μm²/s [0.015–0.051 IQR]; ^∗∗∗p < 0.001, Mann-Whitney U test). It should be noted that the diffusion coefficient values in our studies were nearly an order of magnitude lower that previously reported average diffusion rates for mGFP-SERT (Anderluh et al., 2014, 2017). Several factors might contribute to the variability in the diffusion rate measurement. Higher photon output per frame at lower acquisition frame rates improves the localization accuracy of slowly moving molecules, resulting in significantly smaller reported diffusion coefficients for slower molecules (Lee et al., 2019). It is also possible that antagonist-conjugated QDs target a membrane pool of YFP-SERT that is “transport-willing” and distinct from tracked YFP-SERT or mGFP-SERT species. In addition to the probe potentially altering the transporter conformational equilibrium, a ligand-based approach may also lead to changes in constitutive transport trafficking to and from the plasma membrane. In fact, SERT was reported to undergo constitutive endocytosis in transfected CAD and HEK-293 cells, with a modest surface loss under basal conditions (up to 10% over 1 h at 37°C) when assessed via internalization assays based on the antibody feeding, surface ELISA, and fluorescent cocaine analog labeling (Rahbek-Clemmensen et al., 2014). However, Kittler et al. demonstrated that antagonists and substrates differentially regulated SERT endocytosis in serotonergic neurons over the course of 4 h (Kittler et al., 2010). In particular, fluoxetine led to a dose-dependent internalization of SERT (1 μM: to 63.8 ± 2.5% of control; 10 μM: to 33.9 ± 1.9% of control); paroxetine induced a dose-independent downregulation of SERT (500 nM: 75.5 ± 5.9; 10 μM: 66.0 ± 6.7% of control); sertraline induced SERT internalization (1 μM: 75.8 ± 2.2%; 5 μM: 61.9 ± 1.8%), while exposure to 50 μM cocaine significantly enhanced SERT surface expression (123 ± 3.9%) compared to control levels. Application of substrates led to SERT internalization as follows: 5-HT, 1 μM: 64.7 ± 2.9% and 10 μM: 57.7 ± 1.9%; MDMA, 43.4 ± 1.6%. Additionally, Jørgensen et al. (2014) demonstrated 5-HT-induced reduction in SERT surface expression in both stably transfected HEK-293 cells and cultured raphe serotonergic neurons. When we assessed SERT-bound Qdot localization with respect to the plasma membrane labeled with Cell Mask Deep Red (Supplementary Figure 6) during the typical time course of a labeling/tracking experiment, we observed minimal intracellular Qdot accumulation. Thus, minimizing time elapsed between the final wash steps post-labeling and time-lapse image series acquisition likely allows us to capture the surface trafficking events.

Another factor that confounds interpretation of method-to-method differences in SERT diffusion rate is the irregular pattern of YFP-SERT distribution in the plasma membrane of HEK-293T cells observed in our spinning disk confocal and TIRF images. Analysis of SIM images of YFP-SERT distribution corroborated our preliminary observation that YFP-SERT density was markedly increased in the membrane edges and protrusions (Figure 9). This phenomenon is similar to conformation-dependent preferential accumulation of dopamine transporter in membrane protrusions (filopodia) that was previously observed in several neuronal and non-neuronal heterologous expression hosts (Caltagarone et al., 2015; Kovtun et al., 2019). QD tracking of YFP-SERT localized to membrane protrusions revealed a simple Brownian, unrestricted pattern of diffusion that resembles surface dynamics of the extrasynaptic pool of neuronal receptors (Maynard and Triller, 2019). Physiological significance of SERT preferential targeting to protrusions and its unhindered lateral mobility within these regions is unclear. Ultrastructural studies demonstrated that SERT primarily resided in the extrasynaptic compartments along the axonal projections in the vicinity of asymmetric synapses (Zhou et al., 1998; Tao-Cheng and Zhou, 1999; Miner et al., 2000); thus, to draw parallels with our state-of-the-art knowledge of neuronal receptor surface dynamics, lateral mobility might represent a yet unappreciated dynamic mechanism of SERT delivery and stabilization in the vicinity of synaptic junctions in vivo, where SERT is optimally positioned to clear synaptically released serotonin.

In conclusion, this work presents synthesis, extensive characterization, and QD-based imaging application of a novel biotinylated, PEGylated 3-(1-phenylethyl-1,2,3,6-tetrahydropyridin-4-yl)-1H-indole (IDT785) ligand. Biological activity of IDT785 was demonstrated by its ability to inhibit hSERT-mediated IDT307 uptake with IC₅₀ = 7.2 ± 0.3 μM and through the efficient capture of its biotin terminus by streptavidin-conjugated QDs in solution. As a result, IDT785 enabled specific QD labeling of surface WT and YFP-tagged SERT proteins heterologously expressed in HEK-293T cells. IDT785-based single QD tracking of YFP-SERT membrane dynamics allowed classification of the membrane pool of transporters into distinct fractions undergoing simple Brownian, restricted, or immobilized diffusion. Moreover, IDT785-enabled QD tracking resulted in characterization of membrane diffusion of YFP-SERT proteins preferentially localized to the membrane edges and protrusions in HEK-293T cells. Our next goals are (1) to determine whether IDT785 structure can be optimized to achieve sub-μM SERT binding affinity and (2) to evaluate if IDT785 enables specific QD labeling and tracking of endogenous SERT proteins. Overall, our work provides a generalizable blueprint for developing and rigorous characterization of biotinylated ligands that enable QD-based detection of transmembrane neuronal proteins. Our ultimate goal is to implement this approach in the native neuronal context, and we anticipate further synthetic optimizations toward the development of a high-affinity, specific, compact Qdot probe that would allow us to detect SERT and to monitor its surface dynamics in situ. We envision that tracking SERT in intact tissue would shed light on intrinsic differences in spatially segregated compartments of raphe 5-HT neurons, which display highly branched axonal processes with numerous axonal-like varicosities. We are particularly interested in SERT diffusion with respect to the axon initial segment that may serve as a physical barrier to separate somatodendritic and axonal SERT pools. We also aim to answer the question whether stimulation is capable of inducing transient SERT stabilization in the vicinity of presynaptic boutons, ultimately leading to altered 5-HT clearance rates.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

IT, OK, RT, LB, and TJ performed the experiments and analyzed data. IT, OK, RT, LB, and SR conceived the study, designed the experiments, and wrote the manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.