A radiolabeled substrate, typically ³H labeled, is required to perform the assays. Such radioligands are either commercially available (e.g. for the monoamine transporters SLC6A2-4 transporters ³H-norepinephrine, ³H-dopamine, ³H-imipramine, respectively) (Sucic and Bönisch, 2016) or they can be synthesized as demonstrated for the GABA transporter SLC6A12 and the creatine transporter SLC6A8 (Al-Khawaja et al., 2018). In order to measure the amount of radioactive substrate, which was transported inside the cells, the cells are lysed, a scintillation cocktail is added, and the plates are analyzed with a scintillation counter. Performing radioligand uptake assays requires multiple washing steps which results in LTP.

Sucic and Bönisch have described in detail how to perform radioligand uptake assays with special focus on neurotransmitter transporters (Sucic and Bönisch, 2016). On the day prior to the uptake experiment, the cells are plated into multiple well plates, which were precoated with Poly-D-Lysine, to ensure attachment of the cells to the plate. On the day of the uptake assay, the wells are washed multiple times with buffer and are then incubated with the radiolabeled substrate together with different concentrations of inhibitors. Additionally, positive and negative controls are performed by incubating wells with a high concentration of the radioligand to measure maximum inhibition as well as with buffer to measure nonspecific inhibition. The uptake is stopped by multiple washing steps with ice-cold buffer. Finally, the cells are lysed, a scintillation cocktail is added, and the plate is analyzed with a scintillation counter. Data analysis is typically performed by fitting the data to a sigmoidal-dose response model by applying nonlinear-regression in order to obtain IC₅₀ values. K_i values can be calculated for competitive inhibitors according to the Cheng-Prusoff equation (Cheng and Prusoff, 1973).

The amount of radiolabeled substrate, the number of plated cells and the incubation time highly depend on the nature of the transporter and need to be optimized accordingly.

Radioligand uptake assays have been widely employed to study diverse SLC families, including: SLC1 (Garibsingh et al., 2018), SLC2 (Tripp et al., 2017), SLC6 (Borden, 1996; Núñez et al., 2000; Al-Khawaja et al., 2014, 2018; Hofmaier et al., 2014; Richter et al., 2019), SLC7 (Chien et al., 2018), SLC10 (De Bruyn et al., 2011), SLCO (De Bruyn et al., 2011), SLC13 (Colas et al., 2017), or SLC22 (Erdman et al., 2006).

The advantage of radioligand uptake assays is that different transporters as well as mutants can be measured in the same assay set-up under the same conditions. A disadvantage of the assay is that only a functional effect is measured and not the actual binding affinity. For measuring binding affinities other assays such as Surface Plasmon Resonance (SPR), isothermal titration (ITC) (Rajarathnam and Rösgen, 2014) or radioligand binding assays (Sucic and Bönisch, 2016) can be utilized. Another profound shortcoming of the radioligand uptake assay is that it cannot distinguish between inhibitors and substrates. The actual costs for performing the assay depend highly on the cost of the radioligand.

This assay strategy requires a fluorescently labeled substrate and a conventional fluorometric microplate reader, ideally allowing for real time monitoring of fluorescence changes within the cell. For HTP screening, a reader compatible with 384 or 1,536 well plates and with integrated robotic handling systems is advantageous. Alternatively, cells can be analyzed by microscopy on single slides or in multi titer plates by high content imaging to visualize the distribution of the fluorescent substrate within intracellular compartments.

A cell line overexpressing the SLC of interest in an inducible or constitutive manner is generated for comparing SLC-mediated and unspecific substrate uptake in the same cellular background. To run the assay, the growth medium is removed, and cells are incubated with transport buffer. In case of sym- or antiporters, this buffer should contain relevant ions which are co-transported along with the fluorescent substrate. In addition, a cell-impermeable quenching agent can be added to the buffer to eliminate extracellular fluorescence and thus enhance the signal-to-noise ratio (Wemhöner et al., 2006; Zhou et al., 2010). Finally, the fluorescent substrate is added, and its uptake is monitored for several minutes. In presence of a quencher, the assay can be run in a homogenous format and uptake can be monitored continuously. Without addition of a quencher, washing steps with transport buffer need to be included after substrate addition to remove the remaining fluorescent substrate from the extracellular space.

For assay optimization, cell clones are selected based on the signal-to-background ratio of fluorescent substrate uptake and – if available – the observed activity of known potent and selective tool compounds. Further clone selection criteria include qPCR and western blotting to quantitate protein expression levels. To find optimal assay conditions, the K_m of the fluorescent surrogate substrate is estimated (Wittwer et al., 2013), and competition experiments with the unlabeled physiological substrate can be performed.

Fluorescent surrogate substrate uptake assays are used broadly and have successfully been applied to various SLC families such as SLC6 (Zwartsen et al., 2017), SLC10 (Mita et al., 2006), SLC18 (Hu et al., 2013), SLC27 (Sandoval et al., 2010; Zhou et al., 2010), SLCO and SLC22 (Fardel et al., 2015), and SLC47 (Yasujima et al., 2010; Fardel et al., 2015)

Fluorescent surrogate substrates offer the advantage of performing a rapid, simple and homogenous assay without washing steps, if performed in presence of a quenching agent. Thus, fluorescent surrogate substrate assays are amenable for HTP screening aiming for rapid characterization of lead compounds and can replace more laborious and cost intensive approaches like using radiolabeled substrates or isolation and fractionation of natural substrates. Furthermore, the activity of SLCs can be monitored in real time and cellular process such as trafficking, sequestration or compartmentalization of fluorescent solutes can be visualized. The limitations of this strategy include the need to identify a fluorescent surrogate substrate, which is likely not feasible for every SLC. Also, compounds which are autofluorescent or fluorescence quenchers can interfere with the readout. The costs of the assay largely depend on the costs of the surrogate substrate.

A cell line co-expressing the SLC of interest together with a biosensor for the transported substrate is required. The change in fluorescence can be detected by microscopy, flow cytometry or using a plate reader. Ideally, instruments are equipped with perfusion or injection modules to enable a time-resolved study of transport. Plate-based measurements of biosensors are generally applicable to HTP screening by the use of plate readers such as FLIPR or Hamamatsu FDSS that can accommodate 384-well plates format.

For a successful assay, the cell line should be optimized for homogenous and stable expression levels of both SLC and biosensor, as both will influence the dynamic range. The gene coding for the biosensor is always introduced exogenously and its expression should be examined for correct subcellular targeting and absence of overexpression or folding artifacts. Before the experiment, cells can be starved or treated with drugs to deplete intracellular levels of transporter substrate. Next, cells are incubated in an appropriate assay buffer containing test compounds (e.g. drug candidates). In the case of intensiometric biosensors, a first measurement needs to be performed for normalization. Then, the substrate is added, and the resulting fluorescence change is either recorded immediately to measure kinetics of the transport reaction or with a time-delay to measure the steady-state level. Fitting the concentration of the externally supplied substrate against the fluorescence change results in an apparent K_0.5, or IC₅₀, representing the combination of biosensor affinity, transporter properties and metabolic conversion of the substrate. Alternatively, the substrate concentration can be held constant while varying the test compound concentration for IC₅₀ determinations. As an additional benefit of ratiometric biosensors, the fluorescence change can be converted into absolute intracellular concentrations with the requirements of careful calibration and ratio-processing (Hou et al., 2011; Pomorski et al., 2013).

Biosensors can be widely applied to study SLCs which transport ions or metabolites detectable by a biosensor (Sanford and Palmer, 2017). Biosensors were successfully employed in a number of assays for glucose transporters from families SLC2 and SLC5 (Takanaga and Frommer, 2010; Keller et al., 2019), chloride transport mediated by the SLC26 family (Galietta et al., 2001; Zhong et al., 2014), iodide transported by the SLC12 family (Valdez-Flores et al., 2016), glutamine transported by the SLC1 family (Gruenwald et al., 2012), pyruvate transported by the SLC54 family (Arce-Molina et al., 2020), pyruvate and lactate transported by the SLC16 family (Contreras-Baeza et al., 2019) and ammonium ions transported by yeast homologues of SLC42 family (Ast et al., 2017).

Biosensors can directly measure the concentration of the substrate, offer temporal resolution and can be targeted to cellular compartments. Biosensors overcome the need of cell loading with chemical organic dyes, potentially affecting cell physiology. Biosensor-based assays are inexpensive and essentially have the costs of running a cell culture. A practical disadvantage can be the requirement of the expression of two genes, transporter and biosensor. The main limitation is the availability of biosensors, most of which have been developed to study signaling events and not for transport measurements. However, these can be optimized or repurposed for transporter assays. For instance, a popular class of biosensors is expressed on the plasma membrane for measuring release of neurotransmitters, such as glutamate (Marvin et al., 2013), GABA (Marvin et al., 2019), glycine (Zhang et al., 2018), dopamine (Patriarchi et al., 2018; Sun et al., 2018), norepinephrine (Feng et al., 2019), acetylcholine (Jing et al., 2018), and serotonin (Wan et al., 2020). These biosensors could be co-expressed with SLCs transporting their ligand, similar to coupling of SLC transport to G protein coupled receptor (GPCR) downstream signaling (Vlachodimou et al., 2019). After ligand application and SLC-dependent transport into cells, the reduction of extracellular ligand concentration will decrease the biosensor’s apparent affinity. In fact, pharmacological inhibition of SLCs involved in the clearance of glutamate were measurable with biosensors and support this assay strategy (Armbruster et al., 2016, 2019; Pinky et al., 2018; Wan et al., 2020).

The assay requires incubation of cells or organelles in an appropriate medium for a short period of time. Comparison of the uptake of compounds by genetically engineered cell lines with SLC knock-out or overexpression (Gründemann et al., 2005), as well as the manipulation of uptake medium conditions (e.g. pH, addition of inhibitors or other compounds for competition (Dickens et al., 2018)) can facilitate the identification of transporter substrates. Subsequently samples must be appropriately prepared for mass spectrometric analysis (Dunn et al., 2011; Vuckovic, 2012).

Cajka and Fiehn provide an excellent review of the various MS technologies available for metabolomics, along with advantages and limitations (Cajka and Fiehn, 2016). Untargeted MS following methodologies and guidelines in (Brown et al., 2005; Broadhurst and Kell, 2007; Dunn et al., 2011; Mullard et al., 2015; Broadhurst et al., 2018) enable the measurement of differences in the uptake of a broad range and number (thousands) of compounds by cell lines (Wright Muelas et al., 2020). Targeted MS can alternatively be used, limiting the number of compounds measured (typically <200) but at the same time enabling absolute quantification. Throughput depends on the approach and instrumentation used.

The following steps describe the preparation of intra- and extracellular samples for MS analysis over a time course (Wright Muelas et al., 2020). Following incubation of cells in uptake medium, spent medium is collected after centrifugation, followed by extraction using methanol. The remaining cell pellet is washed, followed by quenching and extraction of intracellular metabolites using methanol. The spent medium and intracellular extracts are subsequently lyophilised, and reconstituted in water ready for analysis by LC-HRMS/MS.

Mass spectrometry analysis of transporter substrates is applicable to most SLC families.

Advantages of the assay are that sampling over a period of time enables transport kinetics to be measured. Untargeted metabolomics allows measurement of relative changes in a wide range of compounds, known and unknown, potentially leading to novel substrate identification. Whilst fewer compounds can be reliably measured using targeted techniques, these enable quantification of the changes in specific compounds to be measured. A disadvantage with both methods is the requirement for expensive instrumentation with specialist knowledge and skills required to run and maintain, along with complex data processing and analysis. However, these disadvantages are outweighed by the wealth of information provided by these assays. This experimental set up is particularly well suited to SLCs expressed at the plasma membrane. Similar approaches have been reported for SLCs localized in cellular organelles such as lysosomes and mitochondria, but require additional cell line engineering to enable the pulldown approach ensuring a quick and efficient organelle isolation (Abu-Remaileh et al., 2017; Chen et al., 2017).

Ionomics assays use mammalian cells overexpressing or bearing knock-out genes encoding particular transporters to quantify the change of inorganic ions by ICP-MS upon cell lysis. ICP-MS-based ionomics is rather a LTP assay, as a significant volume of sample is required.

To analyze the amount of intracellular ions present in the sample (i.e. ²³Na, ²⁴Mg, ³¹P, ³²S, ³⁹K, ⁴⁴Ca, ⁵¹V, ⁵²Cr, ⁵⁵Mn, ⁵⁶Fe, ⁵⁹Co, ⁶⁰Ni, ⁶³Cu, ⁶⁶Zn, ⁹⁷Mo), HEK293 cells stably expressing SLC transporters under the control of a doxycycline inducible promoter are grown in standard cell culture medium, naturally containing a selection of ions and metals. Transporter expression is induced by overnight addition of doxycycline. The next day cells are thoroughly washed with an isotonic Tris/choline-chloride based wash buffer, to completely remove all extracellular ions and subsequently lysed with a Tris/choline-chloride/Triton X-100 based lysis buffer, not containing any of the measured ions. The sample is then ionized by the inductively coupled plasma and the ions are transferred to the mass spectrometer, where they are separated based on their mass-to-charge ratio (m/z). The detector receives a signal proportional to the quantity of ions present in the sample. The ion intensities are normalized to either cell lysate protein concentration or intensity of ³¹P, which were shown to change linearly with the number of cells harvested. To evaluate the contribution of a particular SLCs in the transport of inorganic ions, the normalized intensity ratios of the ions are compared between HEK293 cells with or without induced overexpression of a particular SLC.

Due to the sensitivity of ICP-MS, an exhaustive optimization of washing steps, cell lysis, cell count normalization, ion detection and statistical analysis is required to precisely detect intracellular ion levels (Malinouski et al., 2014; Konz et al., 2017).

This assay format was shown to be suitable for both efflux and influx transporters of metal ions and metalloids including, but not limited to, the aforementioned ions (Malinouski et al., 2014; Konz et al., 2017). Examples of these transporters are from families SLC11, SLC30, SLC31, SLC39, and SLC40. Furthermore, we speculate that SLCs for which the main substrate transport is driven by metal ions detectable with ICP-MS may also be amenable to this assay technology.

Determination of the intensities of the monoisotopic ion ensures specificity and allows direct measurement of inorganic SLC substrates. The method also allows normalization based on either protein concentration of cell lysates or amount of ³¹P with the limitation in case of studies with phosphate transporters. Furthermore, the assay was demonstrated to be suited for SLCs located on the plasma membrane or endoplasmic reticulum. The use of this approach for smaller subcellular compartments (mitochondria, lysosomes, vesicles, etc.) has not been evaluated systematically so far, and it may require isolation of organelles to obtain ion intensities above the limit of detection. The described ionomics assay is applicable in a LTP mode (6-well plate). Another limitation of the assay is that the ICP-MS is a relatively expensive equipment and therefore not available in the vast majority of labs.

For performing a cellular TSA experiment, a source of heating (such as thermoblock or PCR cycler) and a readout discriminating native from aggregated protein are necessary. The choice of readout determines the assay throughput. The most widely used method with Western blotting (WB) as a readout (Martinez Molina et al., 2013) can test only limited compound-target combinations but technologies such as AlphaScreen (Almqvist et al., 2016) or split reporters can enable screening in HTP (reviewed in Henderson et al., 2019). Using MS as a readout, the method can be applied to study target engagement of a single compound on proteome level (Huber et al., 2015; Reinhard et al., 2015).

Since the method was originally introduced for cytoplasmic proteins (Jafari et al., 2014), modifications were necessary for membrane proteins (Hashimoto et al., 2018). Typically, compounds are incubated either with intact cells or cell lysates. The use of intact cells accounts for membrane crossing or metabolic modifications of the compound. Next, samples are aliquoted and exposed to heating. Lysing the cells prior to a heating step could facilitate easier aggregation of membrane proteins after melting, however some studies with membrane protein lysed the cells only after the heating step (Huber et al., 2015; McMillan et al., 2018; Kawatkar et al., 2019). To avoid resolubilization of aggregates, a mild detergent should be used (Reinhard et al., 2015). Finally, the remaining protein in native conformation is quantified in each sample. The most commonly used technique is to remove aggregates with centrifugation and quantify the native protein by WB, but readouts employing reporters can specifically distinguish native protein, and thus the removal of aggregates is not necessary (Martinez et al., 2018). Dose dependency can be confirmed via an isothermal dose-response fingerprint (ITDRF_CETSA) experiment where the sample is treated with several ligand concentrations at constant temperature (Martinez Molina et al., 2013).

Length of compound incubation, compound concentration, sample volume, cell density, heating duration, and the efficiency of native–aggregated protein discrimination should all be optimized first. If available, a potent and selective ligand, such as a specific inhibitor, can be used to determine the possible degree of Tm shift. However, similarly potent inhibitors targeting the same protein can have a different degree of Tm shift (Kawatkar et al., 2019). Although the most widely used heating duration is 3 min, longer heating duration could result in a better Z’ factor and thus be beneficial for HTP screening (Martinez et al., 2018).

A proof-of-principle study showed thermal stabilization of members of SLC1 and SLC16 families upon treatment with available inhibitors, and in the case of SLC16 also stabilization with substrate (Hashimoto et al., 2018). Although to our knowledge only few other studies use the cellular TSA to target SLCs, namely SLC2 family (McMillan et al., 2018; Reckzeh et al., 2019), a number of studies have applied the cellular TSA for transmembrane proteins (Huber et al., 2015; Reinhard et al., 2015; Kawatkar et al., 2019) and SLCs were also detected in TPP studies (Reinhard et al., 2015), demonstrating the potential to apply the method more broadly.

Advantages are that the TSA is probing the direct interaction of a target with a compound, its versatility and that it can use lysate, intact cells, and even whole tissue (Martinez Molina et al., 2013). However, the major limitation is that not all compounds binding the protein will shift the Tm for reasons like insufficient stabilization of the structure or loss of the interaction between protein and compound due to high, non-physiological temperatures. Thus, while the method is relatively resistant to false-positive results, false-negative results can occur. The Drug Affinity Responsive Target Stability (DARTS) assay represents an alternative assay technology similar to the TSA that can assess drug-target engagement for SLC inhibitors (Lomenick et al., 2009; Benjamin et al., 2018). The costs of running a LTP cellular TSA are basically equal to the costs of running a WB experiment, while costs of a HTP TSA depend on the readout and thus on reagents.

Detection of fluorescent indicator dyes is achieved by means of a fluorescence plate reader (such as Fluorescent Imaging Plate Reader (FLIPR) or Hamamatsu FDSS) able to excite the probe and to read its emission. This instrumentation together with the characteristics of the dyes enable the HTP of the assay. Alternatively, fluorescence can be assessed by microscopy or flow cytometry.

The cell line with the expression of the SLC of interest together with an adequate control cell line (mock control that does not express the target, not-induced if the target is overexpressed with an inducible system or knock-out cell line) is loaded with the fluorescent dye in a suitable buffer for a period that typically ranges from 10 min to 1 h. Then, the solution is exchanged to an appropriate assay buffer containing test compounds (e.g. drug candidates). Next the buffer containing a transporter substrate is added and changes in fluorescence are measured. By fitting the concentration of the tested compound against the changes in fluorescence, the EC₅₀ value is obtained.

The membrane potential dye has been used to study electrogenic transporters from families SLC1 (Jensen and Bräuner-Osborne, 2004) and SLC6 (Benjamin et al., 2005; Danthi et al., 2019). BCECF dye has been used to study SLCs from SLC9 (Windler et al., 2018), SLC16 (Contreras-Baeza et al., 2019) and SLC12 families (Reynolds et al., 2007). Potassium transport was measured with the thallium dye (as a surrogate) for SLC12 family (Carmosino et al., 2013) and with the sodium indicator Asante Natrium Green for SLC4 and SLC12 families (Noor et al., 2018).

The main limitation in the use of dyes is that a cell loading step is required, with the consequent risk of affecting cell physiology. Nevertheless, most of the dyes are very easy to load and require a single incubation step without washing the cell monolayer (Wolff et al., 2003), which results in a rapid and HTP assay. In addition, these assays are flexible and have low temporal resolution, given the dyes' fast responses. Finally, the use of probes brings along with it some elevated costs.

Patch clamp requires one skilled person to run a so-called “electrophysiology setup”. The basic version is composed of an inverted microscope, an operational amplifier and a digital-analogic transducer coupled with a computer for data collection and analysis. Usually, a Faraday cage is included for electrical isolation and a fluidic perfusion system is in place to apply compounds diluted in physiological saline solutions (Rubaiy, 2017). Since the technique is versatile, any laboratory currently equipped with the setup can extend this approach to study SLCs without major changes to the protocols already in place. The high informativity is given at the cost of the intrinsic LTP.

The day before the experiment SLC-expressing cells are seeded as single isolated cells on coated glass coverslips. On the day of experiment the coverslips with cells are placed in the recording chamber of the inverted microscope equipped with the headstage of the operational amplifier to run the patch clamp experiment. A glass micropipette is filled with the solution mimicking the cytosolic environment, and firmly stabilized on the headstage of a micromanipulator. The tip of the micropipette is carefully attached to the cell membrane. To form an electrical seal between the micropipette and cell membrane, a constant negative pressure is applied. The membrane is ruptured by a sudden pulse of negative pressure or by brief applications of currents. Afterwards, the voltage clamp configuration is used to modulate the cell membrane potential by applying voltage waveforms specifically designed to favor the activation of the SLC under investigation.

Typically, a perfusion system is integrated in the manual patch clamp setup to apply inhibitors, activators or substrates directly on the patched cell. If such a system is not available, compounds can be directly applied by pipetting small amounts of solution in the recording chamber.

The changes in the membrane currents upon application of substrates/inhibitors are recorded in real-time and analyzed offline, with the major advantage being the internal normalization control for each application of a given compound since transmembrane current is measured before and after.

The SLC of interest needs to be electrogenic (Supplementary Table S1) and expressed in sufficient amount at the plasma membrane. Patch clamp was used for example to validate the effects of two molecules (SEA0400 and KB-R9743) now recognized as reference SLC8A1 inhibitors (Elias et al., 2001; Lee et al., 2004).

The main advantage is the high time resolution and the accuracy of the readout signal, allowing direct monitoring of electrogenic protein activity and their modulation by compounds in real-time. Single cell analysis of a stable clone provides information about homogeneity of the cell line, i.e. the percentage of cells functionally expressing the SLC of interest. A disadvantage for its use for SLCs may be that the net charge caused by electrogenic transport is much lower compared to the opening of an ion channel. The amplitude of recorded signals may not be high enough to allow a dose-response experiment and different techniques are required for a full pharmacological characterization, unless the SLC is expressed at very high levels.

SSM-based electrophysiology requires a SURFE ² R instrumentation, which is available as a SURFE ² R N1 for LTP assays, or as a SURFE ² R 96 SE enabling HTP and automatization in a 96-well plate like format (96 sensors in parallel) (Nanion Technologies Munich, 2021). Each membrane containing the protein of interest is suitable for sample preparation and for measurements using SSM-based electrophysiology, but nevertheless a high protein density and purity can compensate for low turnover and low electrogenicity (Nanion Technologies Munich, 2021; Bazzone et al., 2013, 2017). Normally, the transporter of interest is either recombinantly overexpressed or used from native tissue including different organisms. Commonly used expression systems range from bacteria to eukaryotic cell lines. Also, cell-free expression systems have been used to assay transporter function with SSM-based electrophysiology, where protein is purified, followed by reconstitution into liposomes of ∼100 nm in diameter at high protein densities (Barthmes et al., 2016; Bazzone et al., 2017). An advantage of reconstituted samples is the possibility to vary the membrane composition of the sample which can affect the protein function or ion gradient stabilities (Bazzone et al., 2013, 2017). Due to its mechanical robustness, SSM-based electrophysiology has a high potential for screening applications, allowing determination of the dose dependence of 100 compounds in less than 30 min (Nanion Technologies Munich, 2021; Bazzone et al., 2013, 2017). A SURFE ² R 96SE system (Nanion Technologies Munich, 2021) allows the recording of 96 wells in parallel and the automatization of experimental workflows including sensor preparation, data analysis and export, and practically results in the measurement of six 96-well plates per day (Nanion Technologies Munich, 2021).

For the laboratory setup, a detailed protocol performing experiments was published by Bazzone et al. in 2013 (Bazzone et al., 2013). Sensor preparation includes the thiolization of the sensor surface, the assembly of the lipid layer, and finally the application of membranes to the sensor (Nanion Technologies Munich, 2021). The protein containing membrane and the SSM will form a capacitively coupled membrane system and therefore, charge translocation at the protein containing membrane can be detected by capacitive coupling via the SSM. Upon substrate addition, transient currents are recorded, whereas the peak current represents the maximum speed of the transport. Due to the high stability of the SSM, up to one hundred sequential measurements can be performed on the same sample and allow the determination of parameters such as EC₅₀ or IC₅₀ (Nanion Technologies Munich, 2021). SSM-based electrophysiological experiments only require 0.1 – 1 µg protein per sensor (Nanion Technologies Munich, 2021; Bazzone et al., 2013, 2017).

The method is suitable for the detection of any kind of reaction associated with a charge displacement or with the change in water accessibility close to a charge (electrogenic transporters listed in Supplementary Table S1). The SLC families which were measured using this method are for example SLC1, SLC6, SLC8, SLC15 or SLCO (Geibel et al., 2006; Bazzone et al., 2017; Gerbeth-Kreul et al., 2021; Nanion Technologies Munich, 2021). In addition, since SSM-based electrophysiology has been used to assess the function of mitochondrial proteins, this technique may be applied to intracellular SLCs (Watzke et al., 2010).

The technology allows real-time data acquisition with a high time resolution and a high signal amplification compared to conventional patch-clamp (Nanion Technologies Munich, 2021). SSM-based electrophysiology additionally allows to resolve fast binding kinetics and EC₅₀ or IC₅₀ determination in a HTP manner (Nanion Technologies Munich, 2021).

In contrast to patch clamp and voltage clamp techniques, SSM-based electrophysiology cannot be used to apply a membrane potential (Bazzone et al., 2013). Transporter characterization is therefore restricted to transport modes which do not rely on a membrane potential (Bazzone et al., 2013). In general, SSM-based electrophysiology has no limitations concerning the type of the transporter, but voltage clamp or patch clamp methods can have advantages, if intracellular components like binding proteins are required for protein functionality (Bazzone et al., 2013, 2017). Limitations can arise, if solution exchange creates large artifact currents which happens when the substrate interacts strongly with the SSM like in the case of lipophilic compounds (Bazzone et al., 2013).

Costs per data point are dependent on the assay protocol, especially how many activations, concentrations of compounds or conditions are tested in one well/sensor.

Overall, SSM-based electrophysiology is an ideal methodology in cases where conventional electrophysiology cannot be applied and is also attractive for screening applications in drug discovery especially because of its robustness and its potential for automation (Geibel et al., 2006; Nanion Technologies Munich, 2021).

The hardware requirements for an SLC-GPCR coupling assay are a fluorescence or luminescence microtiter plate reader which ideally allows for kinetic measurements. GPCR activation typically leads to a change in the intracellular calcium or cyclic adenosine monophosphate (cAMP) concentrations, or to altered gene expression. These events can be easily detected by using a typical “GPCR toolbox” consisting mainly of fluorescent dyes and genetically encoded luminescent biosensors (Thomsen et al., 2005; Zhang and Xie, 2012). In brief, calcium responsive fluorescent dyes (e.g. Fluo-8) or photoproteins (e.g. aequorin) can be used to detect increases in intracellular calcium (Ma et al., 2017). Changes in cAMP levels can be detected e.g. via a cAMP-responsive luciferase (Fan et al., 2008), or proximity-based homogenous assays relying on cAMP-antibodies (Williams, 2004). Alterations in gene expression are typically monitored by inserting a reporter, e.g. a luciferase, into the respective genomic locus. All these assays can be run in 384-well-plates which allows for HTP.

For assay development, two cell lines are generated: one expressing the SLC and the other cell line expressing the corresponding GPCR. In a first step, the GPCR assay is independently optimized using the GPCR ligand (i.e. the SLC substrate) from an external source, as well as known inhibitors of the GPCR to verify sensitive detection of GPCR activation, inhibition, and desensitization. In a second step, the supernatant of the SLC-expressing cell line can be used for GPCR activation. At this stage, different SLC stimulation parameters can be evaluated, and clonal selection can be performed. It is also conceivable to express both the SLC and the GPCR in a single cell line. Specifically, for assaying SLC59A2 (MFSD2B) and SLC63A2 (Spinster2, SPNS2), we generated cell lines which stably express SLC59A2 or SLC63A2 + SPHK1 (sphingosine kinase 1) and feed these cells with (unlabeled) sphingosine for several hours. Intracellular sphingosine kinases phosphorylate the pre-substrate to S1P, which is exported by the SLC. Transferring the SLC cells’ supernatant to cells co-expressing S1P3 (a Gq coupled S1P receptor) as well as the calcium photoprotein obelin results in a luminescent calcium response.

In order to make the assay more amenable to HTP, we adapted it to a desensitized format: Continuous agonist stimulation eventually leads to GPCR desensitization, inactivation or internalization (Siehler and Guerini, 2006; Rajagopal and Shenoy, 2018). Thus, instead of detecting the SLC substrate (S1P) as an agonist, it can also be detected as a functional antagonist when applying a second agonist challenge using the substrate (S1P) from an external source. A previously desensitized GPCR will remain “silent”, while not previously stimulated GPCRs will show an agonistic signal. Once optimized, we adapted the desensitized format to a co-culture set-up where S1P3 and SLC cells are seeded into the same well, to eliminate the need for supernatant transfer.

We have successfully applied this strategy to set up assays for the two sphingosine-1-phosphate (S1P) transporters SLC63A2 and SLC59A2. Previously described assays for these transporters were low in throughput because they involved laborious sample preparation followed by TLC (thin layer chromatography) or HPLC (high performance liquid chromatography) detection of radiolabeled or fluorescent S1P (Kawahara et al., 2009; Hisano et al., 2012; Vu et al., 2017; Kobayashi et al., 2018). This approach is applicable to any other SLC which can export a GPCR ligand into the extracellular medium.

SLC-GPCR coupling involves several steps which need optimization and pose confounding factors that may lead to false positive/negative hits. For HTP screening, this approach is therefore most suitable for SLCs which cannot be assayed with more straightforward options such as membrane potential or pH-sensitive dyes, a fluorescent surrogate substrate, or a substrate-specific biosensor. On the other hand, GPCR coupling poses a highly specific and sensitive tool for substrate detection. Therefore, SLC-GPCR coupling can be a valuable downstream or orthogonal assay for hit validation. The cost of an SLC-GPCR coupling assay heavily depends on the GPCR readout strategy. In the S1P transporter assay described above, the cost is low with coelenterazine (obelin’s luminophore) being the most expensive reagent per well.

Requirements to run xCELLigence experiments are a general cell culture facility, temperature and CO₂ controlled environment (e.g. cell culture incubator), an xCELLigence real-time cell analysis (RTCA) instrument and E-plates, which are the main consumable of this application. The xCELLigence RTCA system has multiple plate configurations ranging from 16 to 96 wells up to 384 wells which are amenable to HTP screening (Halai and Cooper, 2012). To allow detection of SLC activity, a (preferably adherent) cell line is required with heterologous or endogenous expression of both the SLC and a concomitant GPCR.

Reports on label-free cell-based transporter assays have been limited (Wong et al., 2012). Recently, a label-free impedance-based assay was developed using xCELLigence to assess functional activity of SLC29A1 (equilibrative nucleoside transporter 1 or ENT1) in living cells (Vlachodimou et al., 2019). Here, activation of adenosine receptors (ARs) by adenosine, a SLC29A1 substrate/AR agonist, is used as a readout. The assay is based on the hypothesis that active SLC29A1 mediates influx of adenosine when extracellular concentrations are higher than cytosolic adenosine, thereby controlling the tone and magnitude of adenosine-mediated signaling events. Upon addition of exogenous adenosine to cells that endogenously express both SLC29A1 and ARs, adenosine is partially taken up via SLC29A1, while the remaining extracellular adenosine is able to activate ARs expressed on the cell membrane. When SLC29A1 transport is blocked by an inhibitor, the extracellular concentration of exogenous adenosine is increased which leads to augmented activation of ARs resulting in an enhanced cell response. This provides an assay window which has been used to characterize inhibitory potency (Vlachodimou et al., 2019) and binding kinetics (Vlachodimou et al., 2020) of SLC29A1 inhibitors. More recently, this assay principle was validated for the human dopamine transporter (DAT, SLC6A3) in two cell lines with heterologous expression of DAT (Figure 9) (Sijben et al., 2021).

SLCs that are suitable for assessment with a TRACT assay should have a known substrate ascribed to them, meaning that orphan transporters are not amenable. So far, TRACT assays have been developed for SLC29A1 and SLC6A3 using the xCELLigence technology. In theory, any SLC that transports a substrate which at the same time is an agonist for a membrane-bound GPCR is admissible for this assay. Examples of these are SLCs for monoamine neurotransmitters (e.g. SLC6A2, SLC6A4), glutamate (SLC1) (Doornbos et al., 2018), carboxylic acids (SLC16), and fatty acids (SLC27). To further widen the scope, any SLC that is involved in or influences a process that leads to detectable changes in cell morphology could potentially be assessed (Wang W. W. et al., 2019). However, this remains to be demonstrated experimentally.

Label-free cell-based assays detect whole-cell responses that are essentially an accumulation of all intracellular signaling events resulting from a perturbation. This allows the researcher to capture comprehensive information in real-time without the use of non-physiological labels and invasive methods. The cumulative signal may also be perceived as a disadvantage of this approach as it produces a “black box” readout, which warrants thorough signal validation during assay development. Additionally, compounds inducing off-target effects that mask or amplify observed cell responses can result in false positives or false negatives, which are mitigated with appropriate counter screens. Main costs for running xCELLigence assays come down to the E-plate consumables. Some protocols describe the reuse of E-plates which provides perspective to reduce overall assay costs (Stefanowicz-Hajduk et al., 2016).

The assay needs to be performed in a cell culture lab and requires a suitable microtiter plate reader for luminescence or absorbance measurements. Both assay systems, the protocol using the full-length nuclear receptor as well as the GAL4 fusion method can be adapted to a HTP format. The GAL4 fusion protocol shows in general higher signal windows.

In order to establish a transporter assay cell line, the expression constructs for the SLC, the corresponding nuclear hormone receptor and its reporter plasmid are transfected into a suitable mammalian host cell line. An inducible expression of the target SLC will enable the analysis of the contribution from the endogenous SLCs in the host cell line to the total hormone uptake. The assay starts by incubating the test compounds with the cells, then the SLC substrate is added for a given time, followed by its removal and an additional incubation time for the cells to express the reporter. Optimization of SLC substrate concentration and incubation time can improve assay performance.

In principle all SLCs for which nuclear receptor agonists or their precursors are substrates can be assayed in this format based on steroid sulfates and corresponding nuclear hormone receptors, such as SLC10A6, members of the SLCO and SLC22A families. Thyroid hormone transporters like SLC16A2 and SLC16A10 might also be amenable for this assay type (Visser et al., 2011).

An advantage of this assay method is the use of unmodified, natural SLC substrates and its independence of electrogenicity or cotransport of certain ions. An important limitation for some substrates is redundancy, especially when respective SLCs are endogenously expressed in the host cell line. The costs of running the assay are low.

The assay requires the ability to genetically engineer a cell line of choice and the necessary tools are vectors allowing to express Cas9 and transporter-specific guides or chemical probes for the SLCs under investigation. If there is no prior knowledge on synthetic lethality to other SLCs, a prior genetic screen may be required. A cell-based viability assay is amenable to HTP screening.

Key requirement is establishing the unambiguous SLC-phenotype relationship. This includes the ability to engineer human cell lines to depend on a particular SLC gene and at least a second control cell line that is isogenic.

The first round of chemical screening is started with the engineered cell line dependent on the SLC of interest. All compounds which exhibit toxicity in the first round are screened in the secondary chemical screen, which includes additional isogenic control cell lines (e.g. WT parental cell line, and cell line dependent on the reciprocal SLC). Afterwards, all compounds which reduce viability only in the cell line dependent on the targeted SLCs are considered as a hit. Since this screening campaign enriches compounds targeting all genes with a synthetic lethal relationship to the targeted SLC, hits should be followed up for example with a binding assay to further filter compounds which interact physically with the targeted SLC.

The great advantage of this approach is that it is amenable to all SLC families, irrespective of their subcellular localization, topology or abundance, provided there is evidence of a genetic interaction between the SLC of interest and another protein and both are expressed in the cell line of choice. For example, strong negative genetic interactions were confirmed between SLC16A1 (MCT1) and SLC16A3 (MCT4), or between SLC25A28 (Mitoferrin-2) and SLC25A37 (Mitoferrin-1) (Girardi et al., 2020b). Using the parental cell line and the SLC16A1 KO as controls, differential activity of compounds against the SLC16A1 KO cell line yielded bona fide SLC16A3 inhibitors (Dvorak and Superti-Furga, unpublished).

The main advantage of this assay is the straightforward readout based on cell viability, which is suitable for HTP screening. Since many of the genetic interactions are of reciprocal character, it is possible to counter screen for specificity by screening hits from the primary screen in reciprocal knock-out cells and WT cells. This approach should filter all compounds which affect viability of cells independently of the targeted SLC, and thus provide a fast path to compound specificity. The main limitation is the requirement of a strong genotype – phenotype association, which may require a prior dedicated genetic screen. Phenotypic assays can be applied also for screening for transporters of cytotoxic compounds (Girardi et al., 2020a).