lable at ScienceDirect

Neuropharmacology 98 (2015) 31e40

Contents lists avai

Neuropharmacology

journal homepage: www.elsevier .com/locate/neuropharm

Invited review

A review of fluorescent ligands for studying 5-HT3 receptors

Martin Lochner a, *, Andrew J. Thompson b, **

a Department of Chemistry and Biochemistry, University of Bern, Freiestrasse 3, CH-3012, Bern, Switzerlandb Department of Pharmacology, Tennis Court Road, Cambridge, CB2 1PD, UK

a r t i c l e i n f o

Article history:Available online 16 April 2015

Keywords:Fluorescent ligand5-HT3 receptorFluorescence polarisationImagingSynthesisFlow cytometry

Abbreviations: 5-HT, 5-hydroxtryptamine; BHK, b4,4-difluoro-4-bora-3a,4a-diaza-s-indacene; Cou, codansyl; Flu, fluorescein; FRET, F€orster resonance enerrescent protein; HEK, human endothelial kidney; NBdiazol-4-yl); Rho, rhodamine; SAR, structureeactivirhodamine; TAMRA, tetramethyl rhodamine; TrCou, t* Corresponding author. Tel.: þ41 31 631 4361.** Corresponding author. Tel.: þ44 1223 334000.

E-mail addresses: martin.lochner@dcb.unibe.ch (M(A.J. Thompson).

http://dx.doi.org/10.1016/j.neuropharm.2015.04.0020028-3908/© 2015 Elsevier Ltd. All rights reserved.

a b s t r a c t

The use of fluorescence is a valuable and increasingly accessible means of probing the pharmacology andphysiology of cells and their receptors. To date, the use of fluorescence-based methods for 5-HT3 receptorresearch has been quite limited and, although a variety of approaches have been described, these arebroadly distributed throughout the literature. In this review we condense these findings into a single,accessible source of reference with the hope of promoting the use of these valuable molecular probes.

This article is part of the Special Issue entitled ‘Fluorescent Tools in Neuropharmacology’.© 2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312. The 5-HT3 receptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333. The chemistry of fluorescent 5-HT3 receptor ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

3.1. Design and synthetic considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333.2. Ondansetron-based probes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343.3. Granisetron-based probes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

4. Biological applications for fluorescent 5-HT3 receptor ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354.1. Fluorescence intensity & fluorescence polarisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354.2. Flow cytometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374.3. High-resolution microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374.4. In vivo imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

5. Future developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

aby hamster kidney; BODIPY,umarin; Cy, cyanine; Dns,gy transfer; GFP, green fluo-D, N-(7-nitrobenz-2-oxa-1,3-ty relationship; SiRho, Sila-riazole coumarine.

. Lochner), ajt44@cam.ac.uk

1. Introduction

Fluorescence can be used to directly label receptors or the li-gands that bind to them. For example, it has become commonpractice to visualise receptors by genetically fusing green fluores-cent protein (GFP) and its engineered homologues (Giepmans et al.,2006; Shaner et al., 2005). This has allowed scientists to visualisereceptor biogenesis, membrane targeting and ligand-mediated re-ceptor internalisation in real-time (Ilegems et al., 2004). Theapplication of these fluorescent proteins in F€orster resonance en-ergy transfer (FRET) has also allowed the study of agonist-induced

M. Lochner, A.J. Thompson / Neuropharmacology 98 (2015) 31e4032

conformational changes, ligand binding and receptor stoichiometry(Ilegems et al., 2005; Miles et al., 2013). However, because of theirsize (e.g. GFP 238 aa, 27 kDa) it can sometimes be challenging toinsert a fluorescent protein without significantly altering theexpression and function of the receptor (Andresen et al., 2004).Small-molecule fluorophores provide an alternative approach thatallows the physical environments of the fluorophores to be moni-tored and the function, pharmacology and physiology of the re-ceptor to be measured (Pantoja et al., 2009; Tairi et al., 1998).

Traditionally high-affinity, radiolabelled ligands were used tocharacterise ligand binding at 5-HT3 receptors, but fluorescent li-gands are providing new opportunities for quantifying bindinginteractions. In contrast to traditional radioligand methods, fluo-rescent approaches can be non-destructive and readily adapted to

Fig. 1. The structure of a 5-HT3 receptor. The 5-HT3 receptor seen from, (A) the extracellulaand competitive antagonists bind to the orthosteric binding sites that are located in the exdomains of only two subunits are shown to highlight the main protein loops that interact wface and loops DeF (blue) are on the complementary face of the binding site. The importancthe mouse 5-HT3 receptor (PDB: 4PIR). The five stabilising nanobodies whose recognition loremoved for clarity. (D) A model for Gran-Flu binding to the 5-HT3 receptor. Gran-Flu (ylabelling and colour code of binding loops is the same as in panel C.

high-throughput methods that provide fast, economical and in-formation rich outputs without the generation of radioactive waste.These direct measurements of quantitative pharmacological pa-rameters rely upon the creation of fluorescent tracers that have asufficiently high affinity for the protein target and give suitablefluorescent signals. 5-HT3 receptors provide excellent opportu-nities for creating these ligands as established high-affinity ligandsalready exist and have synthetically accessible regions that permitthe addition of large substituents. With continued interest in theassociations of 5-HT3 receptors with psychiatric and neurologicaldisorders and recent reports of novel 5-HT3 ligands and allostericmodulators, fluorescent 5-HT3 receptor ligands are likely to findfurther utility in the assessments of in vitro and in vivo pharma-cology (Machu, 2011; Thompson, 2013; Walstab et al., 2010).

r side and, (B) parallel to the plane of the plasma membrane (shown in grey). Agoniststracellular domain at the interface of adjacent subunits (circled). (C) The extracellularith ligands in the orthosteric binding site. Loops AeC (green) are found on the principale of residues within these loops is reviewed in Thompson et al. (2010). The structure isops interact strongly with residues in the receptor orthosteric binding sites have beenellow) was aligned with granisetron (red) in complex with 5HTBP (PDB: 2YME). The

Fig. 2. Chemical structures of indole ligands that bind with nanomolar affinity to 5-HT3 receptors. (A) The clinically used anti-emetic drug, ondansetron. (B) The radio-ligand, [3H]GR65630. (C) The ondansetron isomer, GR67330. (D) Fluorescent 5-HT3receptor ligands reported by Vogel and co-workers, based on the extended derivativeGR119566X. Applications of some these fluorescent probes in cellular studies aredescribed in section 4 of this review.

M. Lochner, A.J. Thompson / Neuropharmacology 98 (2015) 31e40 33

In this review we describe the synthesis of these 5-HT3 receptorligands and the fluorescent approaches that have utilised them.

2. The 5-HT3 receptor

The 5-HT3 receptor is a ligand-gated ion channel that belongs tothe same family as nicotinic acetylcholine (nAChRs), g-aminobutyric acid (GABAA) and glycine receptors (Thompson et al., 2010).It is located in the cell-surface membranes of central and peripheralsynapses where it is responsible for fast neurotransmission andpre-synaptic regulation of several other neurotransmitters (Rond�eand Nichols, 1998; Tecott et al., 1993). They consist of five sub-units that surround a central ion-conducting pore (Fig.1), with eachsubunit containing three distinct functional regions that arereferred to as the extracellular, transmembrane and intracellulardomains. Agonists and competitive antagonists bind in theorthosteric binding site (that occupied by the endogenous agonist)that is formed at the interface of adjacent extracellular domains bythe convergence of three b-sheets from one subunit (termed loops-A to -C) and three peptide loops from the other (Loops-D to -F).Agonists that bind at this location cause a conformational changethat opens a transmembrane pore, allowing cations to cross theplasma membrane and resulting in a net inward current (underphysiological conditions) that causes depolarisation of the cellmembrane. Competitive antagonists bind at the same location andprevent other ligands from entering the site, but do not activate thereceptor. Several such compounds are marketed as antiemeticdrugs to alleviate symptoms resulting from chemotherapy, radio-therapy and general anaesthesia (Thompson and Lummis, 2007;Walstab et al., 2010), and less frequently for the treatment of irri-table bowel syndrome (Moore et al., 2013). These drugs have highaffinities and receptor specificities and are the most frequentlyused scaffolds for coupling with fluorophores. Many non-competitive ligands bind to the 5-HT3 receptor transmembranedomain, but are usually of lower affinity and lack receptor selec-tivity (Thompson and Lummis, 2013). Although there are somefluorescent non-competitive antagonists such as anaesthetics andsteroids, conjugation of fluorophores to non-competitive ligandswould be expected to significantly alter their binding properties,particularly within the tightly enclosed binding sites of the trans-membrane regions (Butts et al., 2009; Teutsch et al., 1994).

3. The chemistry of fluorescent 5-HT3 receptor ligands

The power of chemical synthesis allows the design and creationof novel fluorescent probes that can be utilised for the biologicalinvestigation of proteins on the surface and the inside of cells.Finding an optimal fluorescent ligand for a receptor target can be alengthy and tedious process. In most cases the three-dimensionalstructure of the receptor target is not known and its binding sitespoorly characterised, which hampers a structure-based designapproach. However, once an ideal fluorescent probe is discovered, itcan be a very rewarding and unique research tool that can deliverimportant real-time information about its receptor target and thebiological processes that the receptor is involved in.

3.1. Design and synthetic considerations

To generate fluorescent ligands, the most common designapproach is to conjugate a fluorescent dye to a known ligand. Aplethora of fluorescent dyes are commercially available fromdifferent vendors and they can often be bought in a reactive form(e.g. as isothiocyanate or as N-hydroxysuccinimide (NHS) ester).These reactive groups ensure that the conjugation reaction with anucleophilic group, such as an amine on the ligand or at the end of a

linker, is selective, very efficient and produces the fluorescent probein high yield. What is more, the resulting thiourea and amidebonds, respectively, are very stable and will not be hydrolysed orcleaved by enzymes in the cellular environment (Hermanson, 2013;Sletten and Bertozzi, 2009).

The attachment of a fluorescent dye to a ligand will inevitablyalter its pharmacological and physicochemical properties. Withoutany detailed structural information for the receptor, in particularwhen ligand-bound, or with suitable structureeactivity relation-ships (SAR) for the ligand, it is virtually impossible to predictwhich positions on the ligand are permissive to changes. It is oftenhelpful to explore the literature for analogues by using appropriatesmall compound search engines (e.g. SciFinder®, Reaxys®) in orderto develop a basic understanding of the ligand SAR, or identifysmall compounds that have been used for ligand-affinity purifi-cation (such as the 5-HT3 receptor antagonist GR119566X (Fig. 2)).These latter types of compounds can be useful scaffolds to attachfluorescent dyes to because suitable points for conjugation havealready been determined (Hovius, 2013). It might also be neces-sary to conduct a short SAR study. For instance, Vernekar et al.(2010) synthesised a small library of substituted granisetron ana-logues and measured their binding affinities at the 5-HT3 receptorto allow the identification of regions on the granisetron core thatwere tolerant to change. Relevant binding and functional assaysare then needed to evaluate the pharmacological profile of theligandefluorophore conjugates. In general, high affinity (i.e. lownanomolar range) for the target receptor is desired as it accom-plishes selectivity and tight binding of the probe (Tairi et al., 1998).For this reason antagonists are often chosen as starting points asthey generally display higher affinity for the receptor and do notinduce conformational changes that could affect ligand binding;indeed it is challenging to determine the exact affinity of anagonist owing to receptor activation (Colquhoun, 1998). On theother hand, fluorescent compounds that act as agonists at a spe-cific receptor or ion channel could be very interesting for studies

M. Lochner, A.J. Thompson / Neuropharmacology 98 (2015) 31e4034

such as voltage-clamp fluorometry, single molecule fluorescenceor for following receptor internalisation upon agonist activation(Barden et al., 2015; Talwar and Lynch, 2015). Developing fluo-rescent agonists, however, is much more challenging. These mol-ecules are often smaller and thus the fluorophore is sometimeslarger than the agonist itself. It is not uncommon that the conju-gation of the fluorescent dye to an agonist can significantly affectits function and the fluorescently-modified agonist acts as anantagonist. For instance, Silva-Lopez et al. (2013) have attached aTAMRA fluorophore to the indole nitrogen of 5-HT via a longaliphatic linker. The Kd value of the 5-HT-TAMRA conjugate is83 nM, as determined by radioligand competition, which is verysimilar to the binding affinity of the parent agonist (5-HT,Ki ¼ 125 nM, Hope et al., 1996). However, this fluorescent 5-HTanalogue was unable to activate human 5-HT3 receptorsexpressed in oocytes (Thompson, unpublished).

The attachment of a fluorophore can not only have an impact onthe pharmacological properties of a ligand, but is also likely tochange physicochemical properties, such as solubility, acidity,overall charge and polarity. Some fluorophores (e.g. BODIPY) arefairly lipophilic (Loudet and Burgess, 2007) and the resultingfluorescent ligands can display non-specific interactions leading tohigh background, or increased plasma membrane incorporationand cell penetration. In this respect, turn-on fluorescent ligandsthat only emit when they are bound to the target receptor are apromising development, but also difficult to rationally design(Lukinavi�cius and Johnsson, 2011).

Linkers are often used to space the fluorescent dye from thebioactive ligand in order to reduce the steric impact on thebioactive compound. However, it should be noted that linkers arenot innocent and their length and nature can affect the propertiesof a ligand in unexpected ways. Long aliphatic linkers, especiallywith six carbon atoms or more, can significantly add lipophiliccharacter to the fluorescent probe and it is advisable to use poly-ethylene glycol (PEG) linkers if longer distances are needed(Hermanson, 2013).

Small-molecule fluorophores with suitable photophysicalproperties can be selected from commercial sources according tothe intended application and the fluorescent properties dictatedby the scientific instruments used for measurements. For FRET-based experiments, for instance, it is crucial that the emissionspectrum of one fluorophore (the FRET donor) has a significantoverlap with the excitation spectrum of the second (the FRETacceptor) in order to achieve efficient energy transfer (Guignetet al., 2004; Vallotton et al., 2001a). Alternatively, common fluo-rophore scaffolds can be customised by following establishedsynthetic routes (Lavis and Raines, 2014). Ideal fluorophores havelarge molar extinction coefficients ε and a high quantum yield FFin aqueous media (brightness ¼ FF � ε in a given solvent). Oneshould also keep in mind that the brightness of a fluorophore canvary depending on the environmental conditions. For example, the

Table 1Photophysical and pharmacological properties of fluorescent 5-HT3 receptor ligandsreported by Vogel and co-workers. Values taken from Ilegems et al. (2005);Wohlandet al. (1999). Chemical structures are shown in Fig. 2.

Fluorescent ligand Absmax [nm] Emmax [nm] Brightnessa Kd [nM]b

GR-NBD 465 535 2400e6000 0.5 ± 0.2GR-Flu 494 518 15,200e59,200 0.32 ± 0.19GR-Rho 6G 528 551 98,000 0.8 ± 0.2GR-Rho B 543 565 32,900 1.2 ± 0.5GR-Cy5 649 670 >70,000 18.0 ± 2.0

a Brightness¼ FF � ε, where FF is the quantum yield and ε is the molar extinctioncoefficient.

b Determined by radioligand binding assay.

frequently used fluorescein dye is known to have a significantlyreduced quantum yield at low pH and is prone to photo-bleaching,especially when excited with the intense laser beam that iscommonly used in fluorescence microscopes (see also section 4.3.of this review).

3.2. Ondansetron-based probes

The first fluorescent 5-HT3 receptor ligand, GR-Flu (Fig. 2D,Table 1), was reported by Vogel and co-workers (Schmid et al.,1998; Tairi et al., 1998). It was originally synthesised by theformer Geneva Biomedical Research Institute (Glaxo Wellcome) byreacting the primary amine of GR119566X (Kd¼ 65 pM (Boess et al.,1992)) with fluorescein isothiocyanate (for a protocol see Hovius,2013). As can be seen from the chemical structures in Fig. 2,GR119566X is similar to the 5-HT3 antagonists ondansetron(Ki ¼ 4.9 nM, Hope et al., 1996), GR65630 (Kd ¼ 2.5 nM, Lummiset al., 1990) and GR67330 (Kd ¼ 110 pM, Kilpatrick et al., 1990).Despite the significant structural modification and rather shortlinker, GR-Flu retained its high binding affinity for the 5-HT3 re-ceptor (Kd ¼ 320 pM) and acted as an antagonist with a presumed1:1 complex with the receptor (Tairi et al., 1998; Wohland et al.,1999). GR-Flu emerged as a very powerful fluorescent tool inmany subsequent cellular studies published by the Vogel group(see section 4 of this review). Using the same synthetic strategyVogel and co-workers later coupled other reactive fluorescent dyesto GR119566X which yielded fluorescent 5-HT3 receptor ligandsGR-Rho 6G, GR-Rho B, GR-NBD and GR-Cy5 (Fig. 2D, Table 1), all ofwhich exhibited low nanomolar binding affinity at the 5-HT3 re-ceptor (Pick et al., 2003; Wohland et al., 1999). For instance, thespectrally distinguishable properties of these probes proved to beuseful for time-resolved pulse label experiments (Pick et al., 2003)and as suitable FRET acceptors (Ilegems et al., 2005). However, it isworth mentioning that fluorescence correlation spectroscopy dataindicates that GR-Rho 6G and GR-NBD bind non-specifically andcause receptor aggregation (Wohland et al., 1999).

3.3. Granisetron-based probes

The clinically used anti-emetic drug granisetron (Fig. 3A) bindswith high affinity and specificity to 5-HT3 receptors and its isotope-labelled version (Fig. 3B) is a frequently used 5-HT3 receptor radi-oligand (Kd¼ 1.35 nM, Hope et al., 1996). Due to its high affinity andselectivity granisetron was chosen as the scaffold to develop fluo-rescent ligands that target 5-HT3 receptors by Lochner and co-workers. SARs on granisetron were very scarce in the publicdomain and therefore synthetically possible positions on granise-tron were substituted with differently sized functional groupsyielding a small library of congeners (Vernekar et al., 2010). Bindingaffinities were measured by competition with [3H]granisetron andrevealed that large substituents at the N-1 and C-7 positions of theheterocycle and the N-9 of the bicyclic moiety were permitted(Fig. 3D), and a general synthetic route to access N1-substitutedgranisetron-derivatives was developed (Fig. 3C). This allowed thegeneration of a small library of fluorescent granisetron conjugates(Jack et al., 2015; Simonin et al., 2012) featuring different linkersand spectroscopic properties (Fig. 3D, Table 2). It was found thatattaching the fluorescent dyes via short propyl or butyl linkers (asin Gran-Dns, Gran-BODIPYFL, Gran-Flu, Gran-TAMRA and Gran-SiRho) did not deteriorate the high binding affinity of the probes,whereas longer polyethylene glycol linkers reduced it to highnanomolar values (Gran-Rho B). Fluorescent ligands that have thedye appended to the C-7 position of granisetron are syntheticallymore demanding as it requires a 4-step synthesis of an appropri-ately substituted granisetron core precursor (Simonin et al., 2012).

Fig. 3. Chemical structures of indazole carboxamide ligands that exhibit nanomolar binding affinity at 5-HT3 receptors. (A) The clinically used anti-emetic drug, granisetron. (B) Theradioligand, [3H]BRL43694. (C) The synthetic strategy leading to fluorescent granisetron probes. (D) A selection of fluorescent granisetron conjugates developed by Lochner and co-workers. Applications of some of these fluorescent ligands in cellular studies are described in section 4 of this review.

M. Lochner, A.J. Thompson / Neuropharmacology 98 (2015) 31e40 35

Nevertheless, isolated fluorescent ligands Gran-Cou and Gran-TrCou exhibited high binding affinities, confirming the originalSAR. In the latter compound the coumarin dye was attached via aCu-catalysed alkyne-azide cycloaddition reaction (Sletten andBertozzi, 2009) yielding a heterocycle-extended fluorophore witha larger Stokes' shift (Table 2). Overall, it is fair to say that the

Table 2Photophysical and pharmacological properties of a selection of fluorescent 5-HT3receptor ligands developed by Lochner and co-workers. Values taken from Jack et al.(2015); Simonin et al. (2012). Chemical structures are shown in Fig. 3.

Fluorescent ligand Absmax [nm] Emmax [nm] Brightnessa Ki [nM]b

Gran-Dns 350c 524c 3,300c 1.2 ± 0.6Gran-TrCou 414d 500d 250d 11.6 ± 0.6Gran-Cou 430d 479d 800d 1.9 ± 0.9Gran-Flu 493d 519d 12,800d 1.1 ± 0.2Gran-BODIPYFL 504c 511c 54,400c 2.8 ± 0.7Gran-TAMRA 552d 579d 28,400d 13.9 ± 1.2Gran-Rho B 557d 584d 12,700d 384 ± 68Gran-SiRho 650d 666d 43,600d 2.4 ± 0.8

a Brightness¼ FF � ε, where FF is the quantum yield and ε is the molar extinctioncoefficient.

b Measured using radioligand competition with [3H]granisetron.c In MeOH.d In PBS buffer pH 7.

developed synthetic routes enabled flexibility in terms of fluores-cent ligand design, as individual modules (differently substitutedligand cores, linkers and fluorescent dyes) could be combined inorder to meet experimental requirements.

4. Biological applications for fluorescent 5-HT3 receptorligands

Measuring ligand binding properties is a fundamental part ofpharmacological research andmethods such as radioligand bindingand electrophysiology have beenwidely used to define the kineticsof drug-receptor interactions. Fluorescence is an alternativemethod of performing the same analyses. Notably, fluorescencedetection is far more sensitive than radioactive decay detection (ca.106 times; Hovius, 2013). Moreover, the non-invasive character offluorescent methods allows high-resolution real-time study ofprocesses in live cells. Several approaches have been used in 5-HT3receptor research and are discussed below.

4.1. Fluorescence intensity & fluorescence polarisation

Fluorescence intensity measures the light emitted from anexcited fluorophore. Excitation and emission maxima are atdifferent wavelengths (Stokes' shift) and with appropriate filters

M. Lochner, A.J. Thompson / Neuropharmacology 98 (2015) 31e4036

excitation light can be eliminated to allow measurements of theemitted light alone. This can vary according to the physical sur-roundings of the fluorophore such as changes in pH, the polarity ofthe surrounding medium or environment the fluorophore is boundin, and intramolecular quenching mechanisms, such as photoin-duced electron transfer (PET). In contrast to these simple intensitymeasurements, fluorescence polarisation measures the lightemitted from a fluorescent ligand in horizontal and vertical planesafter excitation with plane-polarised light in one of the planes.Because polarisation is a general property of fluorescent moleculesand does not depend on the fluorescence emission intensity,polarisation-based readouts are less affected by fluorophore con-centrations, the nature of the dye and environmental interferences,such as pH changes, than assays based on fluorescence intensitymeasurements. The precision of fluorescence polarisation signaldetermination is crucial, however, as it significantly influences theaccuracy of the methods. A movement of the ligand between thetime it absorbs light to the time the light is emitted will scatter thelight relative to the plane of excitation. If a molecule is immobileduring the lifetime of the excited state 60% of the emitted light willremain in the same plane as the excitation light. If the moleculetumbles during its fluorescence lifetime, less than 60% of theemitted light will be aligned with the excitation source. The speedat which the fluorophore rotates relative to the lifetime of the in-terval between absorbing and emitting a photon therefore de-termines whether rotation of a molecule can be detected. If it can,the intensities of the light detected in the parallel and perpendic-ular planes are measured. Unlike radioligand methods fluorescence

Fig. 4. Affinities of classical 5-HT3 receptor ligands measured using fluorescence polarisatiomethods give comparable results. The curves on the upper left are representative fluoresc(Gran-Flu) and competition (bottom) of the same ligand with 5-HT and granisetron. On the(bottom) studies using flow cytometry. Geomean denotes the specific shift in the geometricin Jack et al. (2015).

polarisation does not require separation of bound and free ligandwhich allows ligand binding to be quantified without alteringequilibria. In particular, this makes it ideal for measuring low-affinity interactions and repetitive measurements of the samesample under different conditions such as varying temperatures.What is more, Allen et al. (2000) showed that fluorescence polar-isation assays may be run with high precision at high throughput(96-well and 384-well format) for a range of cell surface receptors.Receptors activated by peptides and small molecules, such as the 5-HT3 receptor (using GR-Flu as fluorescent ligand), were included inthis study which demonstrated that fluorescence polarisation of-fers an alternative to radioligand binding assay. However, fluores-cence polarisation depends upon sufficient quantities of receptorthat have been accurately quantified and a fluorophore that hasbeen conjugated to a suitable ligand without overly affecting itsbinding properties; see Rossi and Taylor (2013) for a discussion ofthe advantages and disadvantages.

For 5-HT3 receptors the challenges of purifying the receptorsand developing fluorescent ligands have been overcome and fluo-rescence has been used to accurately determine affinities forlabelled and unlabelled ligands, the underlying rate constants (konand koff), and the contributions of enthalpy (DH0) and entropy (DS0)(Jack et al., 2015; Tairi et al., 1998) (Fig. 4). These studies used eitheran N-terminal FLAG-tagged 5-HT3 receptor purified from HEK293Tcells stably transfected using a Lentiviral system, or a C-terminalHis6-tagged 5-HT3 receptor transiently expressed in BHK cells usingthe Semliki Forest virus. In both studies, fluorescence was notadversely affected by using detergent purified receptors, giving

n (FP), flow cytometry (FC) and radioligand binding (RB). It can be seen that all threeence polarisation data for saturation binding (top) of a granisetron-fluorescein probeupper right the same ligands are examined in saturation binding (top) and competitionmean as measured by flow cytometry. The original data within this figure can be found

M. Lochner, A.J. Thompson / Neuropharmacology 98 (2015) 31e40 37

comparable results as receptors in crude cell homogenates (radio-ligand binding) or on live cell surfaces (flow cytometry, electro-physiology). Such considerations can be important for fluorescentstudies as the physical environment of the fluorophore can havesignificant impacts upon the fluorescent signal; for example GR-NBD has no detectable fluorescent signal in the absence of deter-gent (Wohland et al., 1999).

Tairi et al. (1998) found that fluorescence intensity decreasedwhen GR-Flu (Fig. 2D) was specifically bound to the 5-HT3 receptorand the change was concentration-dependent (Kd ¼ 0.24 nM);quenching was also seenwith an NBD derivative of the same ligand(GR-NBD) and with 5-HT-TAMRA, while GR-Cy5 was unaffectedand the fluorescence intensity of GR-Rho 6G increased (Silva-Lopezet al., 2013; Wohland et al., 1999). Association of GR-Flu wasconsistent with a bimolecular reaction with a kon of1.1 � 106 M�1 s�1. However, the increase in fluorescence intensitythat accompanied dissociation was biphasic, with a fast rate(255 � 10�6 s�1) that was consistent with the product of Kd and kon(koff ¼ Kd � kon), and a slower rate (43 � 10�6 s�1) that was harderto reconcile but has been reported for the parent ligand granisetronin radioligand binding studies (Steward et al., 1995). Jack et al.(2015) used a structurally related ligand (Gran-Flu, Fig. 3D) toperform fluorescence polarisation and found a comparable affinity(Kd ¼ 1.1 nM), but with slightly elevated values for kon(1.4 � 107 M�1 s�1) and koff (10�3 s1) that were more comparable tothe monophasic rate constants that are reported for the parentmolecule granisetron in radioligand binding studies ((Hope et al.,1996), kon ¼ 7.9 � 106 M�1 s�1, koff ¼ 0.003 s�1; (Wong et al.,1995), kon ¼ 2.8 � 106 M�1 s�1, koff ¼ 0.003 s�1). The similaritiesof the fluorophore-conjugated ligand and the parent moleculesuggests that the incorporation of the bulky fluorophore at the N-1position does not significantly affect the binding properties at 5-HT3 receptors (Hope et al., 1996; Jack et al., 2015; Steward et al.,1995; Wong et al., 1995). It is likely that this is because the fluo-rophore protrudes from the binding site, as indicated by the effectsthat Cys substitutions in this region of the receptor have on Gran-Flu binding, and by homology modelling and docking that showsthe N-1 position lies in a sterically unrestricted region of theorthosteric binding (Jack et al., 2015) (Fig. 1D). These studies alsoreported the affinities of a comprehensive range of competing non-labelled ligands and showed that fluorescence intensity, fluores-cence polarisation, radioligand binding and electrophysiology gavecomparable rank orders and affinities for both agonists and an-tagonists (also see Fig. 4).

The fluorescent GR-Flu was also used to characterise mutantreceptors (Schreiter et al., 2003). Specifically, the role of threeconserved acidic residues (Glu-97, Glu-224 and Glu-235) in theorthosteric binding site was investigated. The presence offunctional mutant and wild type receptors on the plasmamembrane of live cells was probed by fluorescent labelling withGR-Flu. In addition, GR-Flu was used in binding studies withdetergent-solubilised wild type and mutant receptors,measuring fluorescence intensity and anisotropy. The photo-physical properties of the fluorescein dye of GR-Flu are pH-dependent and this was exploited to sense the acidic micro-environment of the binding site. By comparison of fluores-cence intensity and anisotropy of GR-Flu upon binding to wildtype and mutant 5-HT3 receptors (E97D, E97Q, E224D, E224Q,E235D and E235Q) Schreiter et al. (2003) concluded that Glu-235 is in close proximity (within 10 Å) of receptor-bound GR-Flu. The same group has used the environmental sensitivity ofGR-Flu in time-resolved fluorescence studies previously todemonstrate that the binding site has a wide entrance and isslightly acidic compared to the bulk solution (Tairi et al., 1998;Vallotton et al., 2001b).

4.2. Flow cytometry

For examining ligand-receptor interaction of cell-surface 5-HT3receptors, flow cytometry can be a truly high-throughput methodthat can be performed on live cells. Purified receptors are notneeded which makes preparation easier than for other techniquesthat are used to probe ligand-receptor interactions such as surfaceplasmon resonance, isothermal calorimetry, nuclear magneticresonance, X-ray crystallography or fluorescence polarisation, but ahigh-affinity fluorescent ligand is essential. Similar to radioligandbinding, fluorescence intensity and fluorescence polarisationmeasurements, it is possible to evaluate affinities of labelled andunlabelled compounds and their rate constants. Because ligand-receptor interactions can be assessed on live cells, flow cytometrycan also be a very practical solution for assessing the ligand bindingproperties of mutant receptors (Jack et al., 2015).

Flow cytometry relies upon a stream of fluid containing singlecells passing a light source and the measurement of the scatter thatresult. Scattered light and fluorescence emission correlate with thesize of the cells and their complexity, and a measurement of thediffraction of light in a flat angle (forward scatter or FSC) dependson the volume of the cell, while diffraction of light at a right angle(sideward scatter or SSC) depends upon features such as cell size,granularity, the structure of its nucleus, and the amount of vesiclesinside the cells. As FSC and SSC are unique for each particle typepassing through the light beam, the combination of the two canhelp identify healthy cells of different types. Excitation and emis-sion of fluorophores can be simultaneously measured to detectmolecular interactions. Consequently this method can measurequantitative and qualitative data from thousands of cells a second.

The use of flow cytometry for directly measuring 5-HT3 receptorligand interactions has been limited, but has provided accuratemeasurements of ligand affinities for both fluorophore-labelled andunlabelled ligands (Jack et al., 2015) (Fig. 4). Kinetic measurementsusing flow cytometry were also found to be comparable to thosefrom radioligand binding and fluorescence polarisation experi-ments, showing that flow cytometry can be a suitable method forproviding quantitative measurements of binding that are consis-tent with these more experimentally demanding techniques.Binding of Gran-Flu to 5-HT3 receptors containing substitutions inthe orthosteric binding site also revealed effects on ligand affinitythat were similar to those measure using radioligand binding withthe radiolabelled parent molecule granisetron. This suggests thatthe addition of a bulky fluorophore did not significantly alter thebinding orientation of the parent ligand, consistent with observa-tions made using the fluorescent methods described in the previ-ous section of this review.

4.3. High-resolution microscopy

In addition to more complex pharmacological measurements,fluorescent ligands also provide opportunities for visualising thelocations of 5-HT3 receptors in live or fixed cells. In the most basicof experiments this can simply be a comparison of labelling in thepresence and absence of a competing ligand to determine specificand non-specific binding (Simonin et al., 2012; Vallotton et al.,2001b; Vernekar et al., 2010) (Fig. 5). This type of study hasbeen extended by using time-lapse confocal microscopy tomonitor the temporal and spatial changes in 5-HT3 receptor traf-ficking in live HEK293 cells following transient transfection.Ilegems et al. (2004) used fluorophore-tagged 5-HT3 receptors thatwere pulsed with GR-Cy5 to show that intracellular 5-HT3 re-ceptors were localised in the endoplasmic reticulum or Golgiapparatus. The finding that these receptors bind GR-Cy5 showsthat the population of intracellular receptors has fully formed

Fig. 5. Fluorescent labelling of human homomeric 5-HT3A receptors in live COS-7 cells. Cells were either transfected with human 5-HT3A subunit cDNA (left and middle panels) ormock transfected (right panels). 24 h after transfection, cells were incubated with 100 nM of fluorescent ligand for 1 h in the dark. The cells were imaged using a confocal mi-croscope with appropriate excitation and emission wavelengths (Table 2). Some cells (middle panels) were co-incubated with the non-fluorescent competitor ligand ondansetron(ond, 10 mM). Both fluorescent ligands Gran-Flu and Gran-TAMRA show specific labelling of receptors on the cell surface. Gran-TAMRA is more photostable than Gran-Flu under theconfocal microscope excitation conditions and is therefore more convenient for time-lapse imaging studies in live cells. Gran-Flu images are from Simonin et al. (2012).

M. Lochner, A.J. Thompson / Neuropharmacology 98 (2015) 31e4038

binding sites, consistent with results from radioligand binding onintracellular populations of 5-HT3 receptors. Application of GR-Rho B to live cells and GR-Cy5 to the same cells following theirfixation and permeabilisation allowed both extracellular andintracellular populations to be distinguished, with 57% of re-ceptors being found within the cell; in the same study, radioligandmethods identified that between 40 and 60 % of receptors wereintracellular, depending upon the cell type studied. Migration of 5-HT3 receptor-containing vesicles from these locations to the cell-surface was tubulin dependent with the first appearance of cell-surface clusters 4e5 h post-transfection. A similar fluorescence-based approach was also used to observe 5-HT3 receptorsfollowing their emergence on the cell surface (Pick et al., 2003).Using GR-Flu and GL-Rho B, 5-HT3 receptors were seen to emergeonto the cell surface in the basolateral region 5e6 h after trans-fection. During this time the fluorescence intensity increased,consistent with the increasing number of receptors, and the re-ceptors formed clusters that migrated towards the top of the cell,with no further changes in the quantities after 24 h; this clusteringof 5-HT3 is also commonly seen with 5-HT3 receptor immuno-labelling (Mukerji et al., 1996; Thompson et al., 2006). Tomonitor mobility within the clusters, photobleaching and fluo-rescence correlation spectroscopy was used, and with both ap-proaches clustered 5-HT3 receptors were seen to be relativelystatic during a 15 min period. It is possible that this clustering maybe responsible for the slow receptor internalisation of 5-HT3 re-ceptors on resting cells as compared to other Cys-loop receptors(Jacob et al., 2008; Kamerbeek et al., 2013; Mammen et al., 1997),but agonist-induced internalisation of these receptors results inthe loss of clusters and a more diffuse pattern of localisation onthe cell surface (Ilegems et al., 2005). It was also reported that thedensities of 5-HT3 receptors observed in these cell-surface clusterswas up to 12,000/mm2, a density that is suitable for determiningthe three-dimensional structures of membrane proteins. By com-parison Torpedo marmorata neuromuscular junctions contain10,000/mm2 and have been used for cryo-electron microscopy(Fairclough et al., 1983; Unwin, 2005). In 2014, the X-ray structuraldetermination of the mouse 5-HT3 receptor, in complex withstabilising nanobodies, was achieved by crystallisation of over-expressed receptors in the same HEK293 cells (Hassaine et al.,2014, 2013). The fluorescent antagonist GR-Cy5 has also beenemployed in high-resolution single-molecule fluorescence

experiments to study the diffusion of the 5-HT3 receptor in cellplasma membranes of live cells (Guignet et al., 2007). These ex-periments revealed a complex and heterogeneous diffusionpattern that was also seen when an NTA-Atto-647 probe wasinstead used to repetitively and reversibly label C-terminallyHis10-tagged 5-HT3 receptors. Also, because the NTA-Atto-647probe coupled to a region outside the binding site the effects ofserotonin could also be assessed, showing the agonist causes asubstantial decrease in the diffusion of the receptors.

GR-Flu has also been used in total internal reflection fluores-cence (TIRF) microscopy. This technique uses an evanescent waveto excite fluorophores in a restricted region of the specimenimmediately adjacent to a glassewater interface, such as the cellmembranes of cells growing as monolayers on a glass coverslip.The evanescent field decays exponentially from the glassewaterinterface, and consequently only penetrates to a depth of ~100 nminto the sample, allowing regions such as the basal plasmamembrane to be visualised. Using this method Schmid et al.(1998) were able to observe real-time binding of ligands toimmobilised, purified and detergent solubilised 5-HT3 receptorsand evaluate the affinities of labelled and unlabelled ligands. Ki-netic parameters were presented in the same study but kon andkoff were slower by two orders and one order of magnitude,respectively, than those reported using radioligand binding,fluorescence intensity, fluorescence polarisation or flow cytom-etry (Jack et al., 2015). This highlights the difficulty of using TIRFfor measuring kinetic parameters when exchange times arepartially restricted by the close bearing of cell-membranes andthe glass coverslip; the use of purified receptors reconstitutedinto cushioned lipid membranes provides a possible alternative(Barden et al., 2015). Elsewhere TIRF has offered extraordinaryresolution where it has been used to observe the fluorescence ofsingle molecules. For example Pantoja et al. (2009) detected thefluorescent unnatural amino acids incorporated into functionalnACh receptors expressed in Xenopus oocytes, but to date therehave not been any similar studies conducted on 5-HT3 receptors.

4.4. In vivo imaging

In addition to cell-labelling, fluorescent ligands also provideopportunities for in vivo imaging. In particular, ligands that emit inthe near infra-red region of the spectrum (ca. 650 nm and above)

M. Lochner, A.J. Thompson / Neuropharmacology 98 (2015) 31e40 39

can be excited at longer wavelengths (lower energy) and conse-quently have reduced phototoxicity. What is more, the longerwavelength excitation and emission allows deeper tissue penetra-tion and reduces auto-fluorescence that is often seen in biologicaltissues (Lukinavi�cius et al., 2013). Jack et al. (2015) used Gran-SiRho, granisetron conjugated to Sila-rhodamine (Fig. 3D), andexamined 5-HT3 receptors in live mice under sedation. Thisrevealed conspicuous fluorescent labelling of gut and salivaryglands. For gut this was expected as 5-HT3 receptors are wellknown to be expressed in these tissues. Labelling in salivary glandsis less reported, but Western blots confirmed the presence of 5-HT3receptors at this location (Jack et al., 2015; Perren et al., 1995). 5-HT3 receptors are also found in the brain, but as the fluorescentsignal is unable to penetrate the skull it was not possible to visu-alise these in the live animal studies (Jack et al., 2015). These resultsshow that appropriately labelled fluorescent ligands are suitable forin vivo imaging where they can be used as an alternative to anti-bodies that are not able to penetrate tissues.

5. Future developments

A considerable range of fluorophore-conjugated ligands havebeen developed for the 5-HT3 receptor with properties that extendover the whole fluorescent spectrum. In this review we havedescribed a wide range of fluorescent approaches, but to date theexperiments that have been conducted with these are somewhatlimited in number, and the ligands that have been modified withfluorophores have remained limited to relatively few of the avail-able range (Thompson, 2013). Other opportunities still exist asfluorophores such as acridone and TAMRA have been conjugated togranisetron to create high-affinity ligands (Jack et al., 2015). Bothfluorophores are highly photostable and have been used elsewherefor high resolution techniques such as two-photon excitation andphotoactivated localisation microscopy (PALM) (Banala et al., 2012;Puliti et al., 2011; Reymond et al., 1996). Thiazole orange is knownto self-quench owing to stacking, but emits light when the stackingis disrupted by binding (Lukinavi�cius and Johnsson, 2011; Volkovaet al., 2008). This results in turn-on fluorescent probes with verylow levels of background fluorescence that are ideal for in vivo andcell labelling in general. Extensivewashing steps in order to removeexcess fluorescent ligand, which is not feasible for most in vivoapplications, can thus be omitted when using turn-on fluorescentprobes.

It is also apparent that in vivo 5-HT3 receptor physiology iscomplex, due to the existence of at least five subunits (5-HT3A e 5-HT3E) that can assemble into heteromeric receptors (Niesler et al.,2003). With this in mind, fluorescent ligands that selectively bindto these heteromeric receptors would greatly enhance our under-standing of their distribution and physiological role. However,compounds with adequate subtype-selectivity are needed beforefluorescent dyes can be attached, and these would ideally bind tosites incorporating subunits 5-HT3B, 5-HT3C, 5-HT3D or 5-HT3E,rather than the AþA-interface that is thought to form the orthos-teric binding site in both homomeric and heteromeric receptors(Thompson and Lummis, 2013).

With the identification of suitable ligands (high-affinity, recep-tor specificity), that have regions permissive to the addition offluorophores (without substantially altering ligand properties), themethods described here are also widely applicable to other re-ceptor types. Combined with the rapid and substantial de-velopments in the technologies needed to visualise fluorescentsignals we anticipate that the types of fluorescent ligands wedescribed will be increasingly used in probing the pharmacology,function and physiology of ligand-gated ion channels.

Acknowledgements

ML thanks the Swiss National Science Foundation for financialsupport (SNSF-professorship PP00P2_123536 and PP00P2_146321/1). AJT thanks the British Heart Foundation for financial support(PG/13/39/30293).

References

Allen, M., Reeves, J., Mellor, G., 2000. High throughput fluorescence polarisation: ahomogeneous alternative to radioligand binding for cell surface receptors.J. Biomol. Screen 5, 63e69.

Andresen, M., Schmitz-Salue, R., Jakobs, S., 2004. Short tetracysteine tags to beta-tubulin demonstrate the significance of small labels for live cell imaging. Mol.Biol. Cell. 15, 5616e5622.

Banala, S., Maurel, D., Manley, S., Johnsson, K., 2012. A caged, localizable rhodaminederivative for superresolution microscopy. ACS Chem. Biol. 7, 289e293.

Barden, A.O., Goler, A.S., Humphreys, S.C., Tabatabaei, S., Lochner, M., Ruepp, M.-D.,Thompson, A.J., Jones, J.P., Brozik, J.A., 2015. Tracking individual membraneproteins and their biochemistry: the power of direct observation. Neuro-pharmcology 98, 22e30.

Boess, F.G., Lummis, S.C., Martin, I.L., 1992. Molecular properties of 5-hydroxytryptamine3 receptor-type binding sites purified from NG108-15 cells.J. Neurochem. 59, 1692e1701.

Butts, C.A., Xi, J., Brannigan, G., Saad, A.A., Venkatachalan, S.P., Pearce, R.A., Klein,M.L.,Eckenhoff, R.G., Dmochowski, I.J., 2009. Identification of a fluorescent generalanesthetic, 1-aminoanthracene. Proc. Natl. Acad. Sci. U. S. A. 106, 6501e6506.

Colquhoun, D., 1998. Binding, gating, affinity and efficacy: the interpretation ofstructureeactivity relationships for agonists and of the effects of mutating re-ceptors. Br. J. Pharmacol. 125, 923e947.

Fairclough, R.H., Finer-Moore, J., Love, R.A., Kristofferson, D., Desmeules, P.J.,Stroud, R.M., 1983. Subunit organization and structure of an acetylcholine re-ceptor. Cold Spring Harb. Symp. Quant. Biol. 48, 9e20.

Giepmans, B.N.G., Adams, S.R., Ellisman, M.H., Tsien, R.Y., 2006. The fluorescenttoolbox for assessing protein location and function. Science 312, 217e224.

Guignet, E.G., Hovius, R., Vogel, H., 2004. Reversible site-selective labeling ofmembrane proteins in live cells. Nat. Biotechnol. 22, 440e444.

Guignet, E.G., Segura, J.-M., Hovius, R., Vogel, H., 2007. Repetitive reversible labelingof proteins at polyhistidine sequences for single-molecule imaging in live cells.ChemPhysChem 8, 1221e1227.

Hassaine, G., Deluz, C., Grasso, L., Wyss, R., Tol, M.B., Hovius, R., Graff, A.,Stahlberg, H., Tomizaki, T., Desmyter, A., Moreau, C., Li, X.-D., Poitevin, F.,Vogel, H., Nury, H., 2014. X-ray structure of the mouse serotonin 5-HT3 receptor.Nature 512, 276e281.

Hassaine, G., Deluz, C., Li, X.-D., Graff, A., Vogel, H., Nury, H., 2013. Large scaleexpression and purification of the mouse 5-HT3 receptor. Biochim. Biophys.Acta 1828, 2544e2552.

Hermanson, G.T., 2013. Bioconjugate Techniques, third ed. Elsevier, Amsterdam.Hope, A.G., Peters, J.A., Brown, A.M., Lambert, J.J., Blackburn, T.P., 1996. Character-

ization of a human 5-hydroxytryptamine3 receptor type A (h5-HT3R-AS) sub-unit stably expressed in HEK 293 cells. Br. J. Pharmacol. 118, 1237e1245.

Hovius, R., 2013. Characterization and validation of fluorescent receptor ligands: acase study of the ionotropic serotonin receptor. In: Banghart, M.R. (Ed.),Chemical Neurobiology: Methods and Protocols, Methods in Molecular Biology,vol. 995. Springer ScienceþBusiness Media, New York, pp. 161e178.

Ilegems, E., Pick, H., Deluz, C., Kellenberger, S., Vogel, H., 2005. Ligand bindingtransmits conformational changes across the membrane-spanning region tothe intracellular side of the 5-HT3 serotonin receptor. Chembiochem 6,2180e2185.

Ilegems, E., Pick, H.M., Deluz, C., Kellenberger, S., Vogel, H., 2004. Noninvasive im-aging of 5-HT3 receptor trafficking in live cells: from biosynthesis to endocy-tosis. J. Biol. Chem. 279, 53346e53352.

Jack, T., Simonin, J., Ruepp, M.-D., Thompson, A.J., Gertsch, J., Lochner, M., 2015.Characterizing new fluorescent tools for studying 5-HT3 receptor pharma-cology. Neuropharmacology 90, 63e73.

Jacob, T.C., Moss, S.J., Jurd, R., 2008. GABAA receptor trafficking and its role in thedynamic modulation of neuronal inhibition. Nat. Rev. Neurosci. 9, 331e343.

Kamerbeek, C.B., Borroni, V., Pediconi, M.F., Sato, S.B., Kobayashi, T., Barrantes, F.J.,2013. Antibody-induced acetylcholine receptor clusters inhabit liquid-orderedand liquid-disordered domains. Biophys. J. 105, 1601e1611.

Kilpatrick, G.J., Butler, A., Hagan, R.M., Jones, B.J., Tyers, M.B., 1990. [3H]GR67330, avery high affinity ligand for 5-HT3 receptors. Naunyn Schmiedeb. Arch. Phar-macol. 342, 22e30.

Lavis, L.D., Raines, T.R., 2014. Bright building blocks for chemical biology. ACS Chem.Biol. 9, 855e866.

Loudet, A., Burgess, K., 2007. BODIPY dyes and their derivatives: syntheses andspectroscopic properties. Chem. Rev. 107, 4891e4932.

Lukinavi�cius, G., Johnsson, K., 2011. Switchable fluorophores for protein labeling inliving cells. Curr. Opin. Chem. Biol. 15, 768e774.

Lukinavi�cius, G., Umezawa, K., Olivier, N., Honigmann, A., Yang, G., Plass, T.,Mueller, V., Reymond, L., Correa Jr., I.R., Luo, Z.-G., Schultz, C., Lemke, E.A.,Heppenstall, P., Eggeling, C., Manley, S., Johnsson, K., 2013. A near-infrared

M. Lochner, A.J. Thompson / Neuropharmacology 98 (2015) 31e4040

fluorophore for live-cell super-resolution microscopy of cellular proteins. Nat.Chem. 5, 132e139.

Lummis, S.C., Kilpatrick, G.J., Martin, I.L., 1990. Characterization of 5-HT3 receptorsin intact N1E-115 neuroblastoma cells. Eur. J. Pharmacol. 189, 223e227.

Machu, T.K., 2011. Therapeutics of 5-HT3 receptor antagonists: current uses andfuture directions. Pharmacol. Ther. 130, 338e347.

Mammen, A.L., Huganir, R.L., O'Brien, R.J., 1997. Redistribution and stabilization ofcell surface glutamate receptors during synapse formation. J. Neurosci. 17,7351e7358.

Miles, T.F., Dougherty, D.A., Lester, H.A., 2013. The 5-HT3AB receptor shows an A3B2stoichiometry at the plasma membrane. Biophys. J. 105, 887e898.

Moore, N.A., Sargent, B.J., Manning, D.D., Guzzo, P.R., 2013. Partial agonism of 5-HT3receptors: a novel approach to the symptomatic treatment of IBS-D. ACS Chem.Neurosci. 4, 43e47.

Mukerji, J., Haghighi, A., Seguela, P., 1996. Immunological characterization andtransmembrane topology of 5-hydroxytryptamine3 receptors by functionalepitope tagging. J. Neurochem. 66, 1027e1032.

Niesler, B., Frank, B., Kapeller, J., Rappold, G.A., 2003. Cloning, physical mapping andexpression analysis of the human 5-HT3 serotonin receptor-like genes HTR3C,HTR3D and HTR3E. Gene 310, 101e111.

Pantoja, R., Rodriguez, E.A., Dibas, M.I., Dougherty, D.A., Lester, H.A., 2009. Single-molecule imaging of a fluorescent unnatural amino acid incorporated intonicotinic receptors. Biophys. J. 96, 226e237.

Perren, M.J., Rogers, H., Mason, G.S., Bull, D.R., Kilpatrick, G.J., 1995.A pharmacological comparison of [3H]-granisetron binding sites in brain andperipheral tissues of the mouse. Naunyn Schmiedeb. Arch. Pharmacol. 351,221e228.

Pick, H., Preuss, A.K., Mayer, M., Wohland, T., Hovius, R., Vogel, H., 2003. Monitoringexpression and clustering of the ionotropic 5HT3 receptor in plasma mem-branes of live biological cells. Biochemistry 42, 877e884.

Puliti, D., Warther, D., Orange, C., Specht, A., Goeldner, M., 2011. Small photo-activatable molecules for controlled fluorescence activation in living cells.Bioorg. Med. Chem. 19, 1023e1029.

Reymond, J.-L., Koch, T., Schr€oer, J., Tierney, E., 1996. A general assay for antibodycatalysis using acridone as a fluorescent tag. Proc. Natl. Acad. Sci. U. S. A. 93,4251e4256.

Rond�e, P., Nichols, R.A., 1998. High calcium permeability of serotonin 5-HT3 re-ceptors on presynaptic nerve terminals from rat striatum. J. Neurochem. 70,1094e1103.

Rossi, A.M., Taylor, C.W., 2013. High-throughput fluorescence polarisation assayof ligand binding to IP3 receptors. Cold Spring Harb. Protoc. 2013,938e946.

Schmid, E.L., Tairi, A.-P., Hovius, R., Vogel, H., 1998. Screening ligands for membraneprotein receptors by total internal reflection fluorescence: the 5-HT3 serotoninreceptor. Anal. Chem. 70, 1331e1338.

Schreiter, C., Hovius, R., Costioli, M., Pick, H., Kellenberger, S., Schild, L., Vogel, H.,2003. Characterization of the ligand-binding site of the serotonin 5-HT3 re-ceptor: the role of glutamate residues 97, 224 and 235. J. Biol. Chem. 278,22709e22716.

Shaner, N.C., Steinbach, P.A., Tsien, R.Y., 2005. A guide to choosing fluorescentproteins. Nat. Methods 2, 905e909.

Silva-Lopez, E.I., Barden, A.O., Brozik, J.A., 2013. Near native binding of a fluorescentserotonin conjugate to serotonin type 3 receptors. Bioorg. Med. Chem. Lett. 23,773e775.

Simonin, J., Vernekar, S.K.V., Thompson, A.J., Hothersall, J.D., Connolly, C.N.,Lummis, S.C.R., Lochner, M., 2012. High-affinity fluorescent ligands for the 5-HT3 receptor. Bioorg. Med. Chem. Lett. 22, 1151e1155.

Sletten, E.M., Bertozzi, C.R., 2009. Bioorthogonal chemistry: fishing for selectivity ina sea of functionality. Angew. Chem. Int. Ed. 48, 6974e6998.

Steward, L.J., Ge, J., Bentley, K.R., Barber, P.C., Hope, A.G., Lambert, J.J., Peters, J.A.,Blackburn, T.P., Barnes, N.M., 1995. Evidence that the atypical 5-HT3 receptorligand, [3H]-BRL46470, labels additional 5-HT3 binding sites compared to [3H]-granisetron. Br. J. Pharmacol. 116, 1781e1788.

Tairi, A.-P., Hovius, R., Pick, H., Blasey, H., Bernard, A., Surprenant, A., Lundstrom, K.,Vogel, H., 1998. Ligand binding to the serotonin 5HT3 receptor studied with anovel fluorescent ligand. Biochemistry 37, 15850e15864.

Talwar, S., Lynch, J.W., 2015. Investigating ion channel conformational changesusing voltage clamp fluorometry. Neuropharmacology 98, 3e12. http://dx.doi.org/10.1016/j.neuropharm.2015.03.018.

Tecott, L.H., Maricq, A.V., Julius, D., 1993. Nervous system distribution of theserotonin 5-HT3 receptor mRNA. Proc. Natl. Acad. Sci. U. S. A. 90,1430e1434.

Teutsch, G., Klich, M., Bouchoux, F., Cerede, E., Philibert, D., 1994. Synthesis of afluorescent steroid derivative with high affinities for the glucocorticoid andprogesterone receptors. Steroids 59, 22e26.

Thompson, A.J., 2013. Recent developments in 5-HT3 receptor pharmacology.Trends Pharmacol. Sci. 34, 100e109.

Thompson, A.J., Lester, H.A., Lummis, S.C., 2010. The structural basis of function inCys-loop receptors. Q. Rev. Biophys. 43, 449e499.

Thompson, A.J., Lummis, S.C., 2013. Discriminating between 5-HT3A and 5-HT3ABreceptors. Br. J. Pharmacol. 169, 736e747.

Thompson, A.J., Lummis, S.C.R., 2007. The 5-HT3 receptor as a therapeutic target.Expert Opin. Ther. Targ. 11, 527e540.

Thompson, A.J., Padgett, C.L., Lummis, S.C., 2006. Mutagenesis and molecularmodeling reveal the importance of the 5-HT3 receptor F-loop. J. Biol. Chem. 281,16576e16582.

Unwin, N., 2005. Refined structure of the nicotinic acetylcholine receptor at 4Åresolution. J. Mol. Biol. 346, 967e989.

Vallotton, P., Hovius, R., Pick, H., Vogel, H., 2001b. In vitro and in vivo ligand bindingto the 5HT3 serotonin receptor characterised by time-resolved fluorescencespectroscopy. Chembiochem 2, 205e211.

Vallotton, P., Tairi, A.-P., Wohland, T., Friedrich-B�enet, K., Pick, H., Hovius, R.,Vogel, H., 2001a. Mapping the antagonist binding site of the serotonin type 3receptor by fluorescence resonance energy transfer. Biochemistry 40,12237e12242.

Vernekar, S.K., Hallaq, H.Y., Clarkson, G., Thompson, A.J., Silvestri, L., Lummis, S.C.,Lochner, M., 2010. Toward biophysical probes for the 5-HT3 receptor: struc-tureeactivity relationship study of granisetron derivatives. J. Med. Chem. 53,2324e2328.

Volkova, K.D., Kovalska, V.B., Balanda, A.O., Losytskyy, M.Y., Golub, A.G., Vermeij, R.J.,Subramaniam, V., Tolmachev, O.I., Yarmoluk, S.M., 2008. Specific fluorescentdetection of fibrillar a-synuclein using mono- and trimethine cyanine dyes.Bioorg. Med. Chem. 16, 1452e1459.

Walstab, J., Rappold, G., Niesler, B., 2010. 5-HT3 receptors: role in disease and targetof drugs. Pharmacol. Ther. 128, 146e169.

Wohland, T., Friedrich, K., Hovius, R., Vogel, H., 1999. Study of ligand-receptor in-teractions by fluorescence correlation spectroscopy with different fluo-rophores: evidence that the homopentameric 5-hydroxytryptamine type 3Asreceptor binds only one ligand. Biochemistry 38, 8671e8681.

Wong, E.H., Clark, R., Leung, E., Loury, D., Bonhaus, D.W., Jakeman, L., Parnes, H.,Whiting, R.L., Eglen, R.M., 1995. The interaction of RS 25259-197, a potent andselective antagonist, with 5-HT3 receptors, in vitro. Br. J. Pharmacol. 114,851e859.