Latest surface plasmon resonance advances for G protein-coupled receptors

Abstract

G protein-coupled receptors (GPCRs) are a big family of membrane proteins which represent one of the main classes of drug targets. However, their investigation presents several challenges, among which their instability outside the membrane environment. Different strategies for the drug discovery of this target are available, and surface plasmon resonance (SPR) stands out as one of the most informative and widespread binding assays, with many advantages such as real-time and label-free analyses resulting in the definition of both affinity and kinetic constants. This review covers the applications of SPR in GPCR drug discovery of the last 10 years and classifies the papers based on the immobilization strategy on the SPR sensor chip to maintain receptor stability. In particular, GPCR immobilization can occur in its native membrane by immobilizing whole cells or membrane fragments, using membrane mimetics (such as lipoparticles, lentiviral particles, liposomes, lipoproteins, nanodiscs, or planar lipid membranes) or immobilizing the isolated receptor stabilized by the use of detergents or engineering approaches. Different examples were considered and pros and cons of each strategy were presented.

Graphical abstract

Highlights

• •G protein-coupled receptors (GPCRs) are one of the main classes of drug targets.
• •Surface plasmon resonance (SPR) analysis of GPCRs presents several advantages.
• •GPCRs have an intrinsic instability outside their membrane environment.
• •The main strategies used to maintain GPCRs stability in SPR are discussed.
• •Immobilization on SPR chip promotes GPCRs analysis despite instability challenges.

1. Introduction

G protein-coupled receptors (GPCRs) represent one of the most important classes of drug targets, with around one third of drugs on the market being modulators of their functions. They are indeed the main family of membrane proteins and they are involved in a number of physiological processes, since they are able to transmit extracellular signals deriving from a variety of ligands ranging from small molecules to proteins. They are also associated to different pathological conditions, such as metabolic, cardiovascular and neurological disorders, inflammatory diseases and cancer [[1], [2], [3]].

Despite the fact that many GPCRs are the target of several drugs and the focus of drug development efforts, only a small fraction (25%–40%) of the potentially druggable GPCRs are currently targeted. Furthermore, numerous orphan GPCRs, whose natural ligands and functions are still unidentified, constitute significant gaps in our understanding of this receptor class [1,4].

This is due to the fact that, as transmembrane receptors, their study represents a challenging task. These receptors have a conserved seven-transmembrane helical structure, with three intracellular and three extracellular loops and they are characterized by an intrinsic instability outside their natural lipid environment. In addition, they are expressed at low levels and need an appropriate expression system to produce higher amounts for their study [1,2,5].

In this context, the research on GPCRs is of great interest in medicinal chemistry and much effort is being made for the discovery of new ligands.

2. GPCR drug discovery

Nowadays, the trends on drug discovery for GPCR targets include the use of high-throughput screening (HTS) assays and the research of selective drugs, unlike what usually happens with classical drugs that can bind to different targets (generally from 4 to 10) [1,6].

A possibility to find GPCR potential ligands consists in the in silico screening of virtual libraries, but this structure-based method lacks of accuracy and can generate false-positive results. Therefore, in silico screening is generally used as an auxiliary method to the experimental screening [6].

Different techniques can be applied for the experimental screening and the possible assays can be distinguished in binding-, stability-, and cell signaling-based. The first approaches allow to directly measure the physical interaction of the receptor with single ligands, the second study the stability changes in GPCR receptor when in contact with the potential ligands, while the third assess the effect of ligand binding to GPCR by quantifying second messengers downstream of receptor activation [3].

The classical screening methods for the identification of GPCR ligands are radioligand binding assays. These methods allow to study in detail drug-target interactions, but do not provide information on drug function (agonist or antagonist), the signaling pathways in which it is involved and its potency. In addition, the use of radioactivity is not ideal for a screening assay [1,7].

When performing a binding assay, the screening generally begins by comparing different drugs at a single concentration. This first step is followed by the analysis of different concentrations of the selected drugs to build dose-response curves [6].

Among binding assays, optical biosensors hold a leading position in target-based drug discovery, with surface plasmon resonance (SPR), optical waveguide grating (OWG), and biolayer interferometry (BLI) principally used for screening purposes. These represent powerful techniques because, in addition to a preliminary screening, they can be applied to different stages of the drug discovery and development process, ranging from the validation of targets and hits, to the study of the mode of action, to the determination of binding affinity and kinetics [8,9].

In particular, SPR is considered a gold standard technique to measure biomolecular interactions and has become a precious tool in drug discovery and development. It is based on an optical phenomenon that occurs in thin films made of a conductive material, generally gold, at the interface between two media with different refractive indices. When polarized light strikes on the gold film, it causes an excitation of the conductive electrons which convert to surface plasmons and create an electric field known as evanescent wave. Surface plasmons adsorb light and cause a reduction of the intensity of the reflected light. In these conditions, the detector is very sensitive to the changes in the refractive index of the aqueous solution flowing on the gold chip surface and it monitors interactions occurring between an immobilized molecule and the flowing analytes [9,10].

SPR has several advantages compared to other assays, among which the possibility to perform a label-free and real-time analysis allowing to quantify not only the affinity constants, but also the kinetic parameters of an interaction. In addition, this technique allows a rapid high throughput screening of libraries and requires extremely low amounts of samples. It is also characterized by a high sensitivity and specificity, as well as by its versatility; in fact, it can monitor binding events between molecules of different nature and size and find applications in different fields. Given the many advantages, this technique has gained a crucial role in different steps of the drug discovery and development process [9,10].

SPR technology has undergone huge progress in the past years and the information that can be gained from an analysis has been expanded. The trend is to move towards sensors equipped with multiple channels and parameters, allowing to detect different sites in parallel. Several SPR systems are available on the market, differing mainly in the automation and the optical system; in particular, waveguide-coupling, grating and fiber sensors have also emerged. In addition, to increase the information that can be obtained from SPR, it has been coupled to mass spectrometry, fluorescence and Raman spectroscopy [10].

For all these reasons, several applications of SPR and its technological advancements to GPCRs are present in literature and this review will cover the papers published in the last 10 years, with a specifical focus on peculiar applications.

3. Challenges in the study of GPCRs by SPR

The study of GPCRs by SPR and related techniques presents different challenges. These challenges are related to the structure and nature of GPCRs themselves, but different approaches can be followed to make the SPR analysis feasible.

The first issue is to create a membrane-like environment allowing to maintain the structure and function of the receptor. In fact, GPCRs have a shared architecture with seven transmembrane domains, three intracellular and three extracellular loops, which is important for their stability and activity. A suitable GPCR environment can be obtained maintaining the native receptor membrane, using a membrane mimetic or employing detergents. For this reason, a suitable immobilization strategy should be selected [11,12]. Different strategies to maintain receptor stability will be discussed in the present review.

Another issue concerns the low molecular weight of most GPCR ligands. The SPR signal is influenced by the molecular weight of both the flowing analyte and the immobilized molecule and the detection of small molecules is harder because they produce a lower change in the refractive index at the sensor surface. Therefore, it is important to maintain the stability of the baseline, which is generally challenging when working with lipids or detergents. A further expedient to increase the signal related to the binding is to have a high density of active receptor on the sensor chip [11,13].

However, a high receptor density is difficult to obtain due to the low levels at which most of GPCRs are present on their native tissues. This led to the use of heterologous expression systems in which receptors are overexpressed [13].

When SPR is applied to the screening of multiple ligands, it is also necessary to include suitable washing steps which allow regeneration while avoiding the loss of receptor activity. Attention should be paid in maintaining the stability and activity of GPCR during the experiment; for this reason, it is important to perform stability and activity tests at defined times throughout the screening [11].

Another aspect to consider is the orientation of the immobilized receptor. The ideal situation is when all the receptor molecules are oriented in a way that the binding site is accessible to the ligand, but this could be obtained by capture methods only by possessing information about GPCR amino acid sequence and its structure. In some cases, GPCR response is achieved only if both sides of the molecule are accessible and strategies to obtain this accessibility have been developed [11].

In this review, different approaches to face these challenges will be discussed with particular focus on the applications of the last 10 years.

4. SPR applications to GPCR drug discovery

As discussed, one of the main issues of GPCRs when analyzed by SPR is the maintenance of their stability. The receptor immobilization is crucial for the preservation of its structure and function.

For this reason, the review will classify the applications found in literature based on the immobilization strategy.

4.1. GPCR immobilization from native membranes

4.1.1. Whole-cell immobilization

An interesting approach to maintain GPCR stability is the one adopted by Jin et al. [14], which described the development of a live-cell biosensor.

For this assay, the β2-adrenoceptor was selected as a GPCR model. The GPCR was expressed on the plasma membrane of a heterologous system composed of human embryonic kidney (HEK)-293 cells; this created the lipidic environment necessary to maintain the β2-adrenoceptor structure and function. The expression efficiency and the receptor distribution on cell surface were assessed by the binding between the rho-tag sequence at the N-terminus of the receptor and a fluorescent anti-rho-tag antibody, while the receptor function was confirmed by intracellular calcium assays [14]. A culture of HEK-293 cells expressing β2-adrenoceptor was grown on the SPR sensor placed on the bottom of a cell culture chamber. Intracellular changes occurring in close proximity of the cell membrane on the sensor surface induce a change in the refractive index of the solution, which is converted into a SPR signal. Therefore, the intracellular transduction pathway induced by GPCR ligands can be monitored by SPR. This study demonstrated that the ligand isoprenaline generated a SPR signal proportional to its concentration on HEK-293 cells expressing β2-adrenoceptor and not on the negative control cells without expression of the receptor, confirming the validity and the specificity of the approach. In addition, the use of LY294002, an inhibitor of phosphatidylinositol 3-kinase (PI3K), allowed to enhance receptor intracellular signaling and SPR signal, improving the sensitivity by more than three times. In fact, the GPCR signaling cascade involve the activation of phospholipase C, the increase of inositol 1,4,5-triphosphate (IP3) and the intracellular release of Ca²⁺ and it has been shown that PI3K prevents this cascade; therefore, PI3K inhibitors can lead to an increase of intracellular Ca²⁺ concentration and ultimately to an enhancement of SPR signal (Fig. 1) [14]. This application is peculiar because it puts together binding- and signaling-based assays, allowing to obtain both affinity and functional information in a single experiment. Instead, generally it is necessary to perform both assays to define not only the binding to the target receptor, but also the signaling pathways involved [1,6].

Fig. 1.

Representation of the live-cell biosensor developed [14]. PLC: phospholipase C; IP3: inositol 1,4,5-triphosphate; PIP2: phosphatidylinositol 4,5-bisphosphate; IP3R3: inositol 1,4,5-triphosphate receptor type 3; PI3K: phosphatidylinositol 3-kinase. Reprinted from Ref. [14] with permission.

Lu and Li [15] also examined the use of whole cells by plasmonic-based electrochemical impedance microscopy (P-EIM), a technique that combines SPR and electrochemical impedance microscopy (EIM). They applied the developed P-EIM, a label-free technique, to study the intracellular signaling of Ca²⁺ in HeLa cells expressing the H1 receptor. The study revealed that the activation of H1 receptor by histamine ultimately induces movements of the endoplasmic reticulum (ER) due to calcium release, which are detected by SPR because they are associated with changes in the refractive index close to the sensor chip. The authors were also able to perform SPR kinetic analyses which showed differences in the calcium flow in different cells, due to their individual variability. The combined use of EIM, allowing to identify alterations in the local dielectric and conductive characteristics, resulted in the possibility of monitoring the timing of Ca²⁺ release events and of revealing the subcellular distribution and activity of calcium-related ion channels during agonist-induced Ca²⁺ signaling [15]. Therefore, the highly informative P-EIM technique allowed to detect the intracellular Ca²⁺ signaling at the initial phase of GPCR activation, characterizing the different behavior of individual cells by the combination of SPR measurements and EIM images. The same principle could also be applied to different cellular events and signaling pathways [15].

Another interesting study is the one described by Nonobe et al. [16], which performed a two-dimensional (2D) SPR imaging on HEK-293 cells expressing the GPCRs type-1 metabotropic glutamate receptor (mGluR1) and adenosine A₁ receptor (A₁R). These two GPCRs are known to form complexes and are involved in mutual modulation. As in the first example, the GPCR signaling was monitored. In this case, the translocation of protein kinase C (PKC), one of the main GPCR-coupled signaling molecules, was measured by SPR imaging. In fact, it has been shown that the accumulation of proteins close to the cell membrane can induce changes in the SPR curve when cells are grown on the SPR sensor chip. Similarly to the previous work by Jin et al. [14], the HEK-293 cells expressing GPCRs were cultured on the gold surface of the SPR chip [16]. Before applying the method to study the responses mediated by mGluR1 and A₁R, the researchers selected the angle of incident light at which the intensity of the reflection due to intracellular protein translocation was maximized. The treatment of the cells expressing the two GPCRs with agonists of the two receptors and with a PKC inhibitor confirmed that the developed method allowed to monitor PKC translocation; this enabled to study the effect of ligand binding to the receptor mGluR1 and the modulation induced by A₁R. Therefore, this study not only showed that it is possible to monitor GPCR signaling by a SPR system, but also highlighted that even the modulation of the receptor can be investigated.

Lieb et al. [17] also developed a SPR cell-based assay. In this paper, the authors compared different label-free techniques, namely electric cell-substrate impedance sensing (ECIS), resonant waveguide grating (RWG) and SPR. Two GPCRs were investigated: the human histamine H1 receptor (H1R), endogenously expressed by human U-373 MG glioblastoma cells, and the β2-adrenergic receptor (β2-AR), endogenously expressed by bovine aortic endothelial cells (BAEC). Together with human U-373 MG glioblastoma cells and BAEC, the researchers also examined engineered HEK-293T cells expressing H1R and the adhesion protein human macrophage scavenger receptor 1 (hMSR1). In the work by Lieb et al. [17], an innovative ECIS-SPR assay was also developed by using a single sensor chip with both detection systems. ECIS is an electrochemical approach based on the measurement of impedance and characterized by electrodes on which cells grow. It is responsive to the electrode coverage and to cell morphology. The activation of GPCRs can cause cytoskeleton remodeling and these changes are measured by ECIS. For the ECIS-SPR dual sensor platform, the sensor surface included two ECIS small electrodes and one bigger counter electrode used for SPR measurements. In SPR and RWG the evanescent field has a low penetration depth (around 100−200 nm) and so these optical techniques can measure changes close to the sensor surface, while in ECIS the signal deriving from the entire cell body is measured. Therefore, this technique can produce complementary data compared to SPR and RWG. For this reason, the combination of ECIS and SPR can be a strategy to obtain more information on the same cell population. Since SPR and RWG are sensitive to changes occurring close to the sensor surface, the authors investigated a strategy to enhance the signal, showing that cell adhesion is critical to guarantee a good signal. Coherently, cells expressing the adhesion protein hMSR1 gave an enhanced signal compared to cells only expressing H1R [17]. Therefore, cell engineering might be necessary when designing an experiment of this kind.

In 2020, an innovative work exploiting continuous angular-scanning SPR was published [18]. The authors stated that the label-free methods previously reported were not able to distinguish between different GPCRs signaling pathways and proposed a method resulting in a better pathway separation. In fact, different classes of GPCRs exist, each one interacting with different second messengers to produce distinct cellular responses. Compared to traditional functional assays which are generally sensitive to only one signaling pathway, or to other real-time label-free assays which generally require a multi-step analysis to distinguish between different pathways, this continuous angular-scanning SPR method allowed to separate three signaling pathways in a single-step without requiring prior cell treatments, antagonist usage, or the use of compounds that modulate the pathways. The real-time analysis was performed monitoring different SPR parameters at the same time on cells exposed to GPCR agonists. As in the previous examples, the system was sensitive to changes occurring in the cells in close proximity to the sensor surface. Among the investigated parameters, the researcher found that peak angular position (PAP) and peak minimum intensity (PMI) allowed to observe distinctive response profiles and consequently to identify the specific GPCR pathway [18]. This study [18] highlighted that by investigating multiple parameters at the same time, it is possible not only to assess the activation of GPCR signaling, but also to recognize the specific pathway involved.

To sum up, SPR experiments based on the immobilization of whole cells presenting GPCRs on their surfaces present advantages and disadvantages. Table 1 [[14], [15], [16], [17], [18], [19], [20]] summarizes the key parameters of the examined works and their main findings. Cell-based SPR assays allow to obtain not only affinity data, but also functional information. In addition, compared to the classical fluorescence or luminescence-based functional assays, SPR is a label-free technique working on the receptors in their native state and can monitor the binding in real-time. Furthermore, the use of innovative technologies such as P-EIM, SPR imaging, ECIS-SPR and continuous angular-scanning SPR can generate a high information content analysis. However, for this kind of assays, a difficult and time-consuming method development is needed, it is often desirable to increase SPR signal by different strategies and specific equipment should be applied. Attention should be given to the distance between cells and the sensor surface, since SPR is able to monitor events occurring in close proximity to the sensor.

Table 1.

Key parameters of the works concerning whole-cell or membrane fragment immobilization and their main findings.

GPCR	Cells/membranes	Technique	Findings	Refs.
β2-adrenoceptor	HEK-293	SPR	Development of a method able to study also the intracellular signaling and use of an inhibitor of PI3K to improve the sensitivity.	[14]
H1 receptor	HeLa	P-EIM (combination of SPR and EIM)	Development of an informative P-EIM method to study intracellular calcium signaling.	[15]
mGluR1 and A1R	HEK-293	2D SPR imaging	Study of mGluR1 modulation induced by A1R.	[16]
H1R and β2-AR	Human U-373 MG glioblastoma cells, BAEC, and HEK-293T	ECIS, RWG, and SPR	Use of the adhesion protein hMSR1 to enhance the signal and development of a highly informative ECIS-SPR assay.	[17]
Bradykinin B2, β2-adrenoceptors, histamine H1, H2, H3, and H4 receptors	CHO–K1, A431, and HeLa cells	Continuous angular-scanning SPR	Development of a method able to distinguish between different GPCRs signaling pathways.	[18]
OR6M1	MCF-7 and HEK293T/17	SPR	The immobilization of membrane fragments enhanced the sensitivity compared to intact cells.	[19]
NOP receptor	Commercial membrane preparations from a recombinant cell line	SPR	The developed immobilization procedure and the use of membrane preparations resulted in good SPR signals.	[20]

GPCR: G protein-coupled receptor; HEK: human embryonic kidney; SPR: surface plasmon resonance; PI3K: phosphatidylinositol 3-kinase; P-EIM: plasmonic-based electrochemical impedance microscopy; mGluR1: type-1 metabotropic glutamate receptor; A₁R: adenosine A₁ receptor; 2D: two-dimensional; H1R: histamine H1 receptor; β2-AR: β2-adrenergic receptor; BAEC: bovine aortic endothelial cells; ECIS: electric cell-substrate impedance sensing; RWG: resonant waveguide grating; hMSR1: human macrophage scavenger receptor 1; CHO–K1: chinese hamster ovary cell line K1; OR6M1: olfactory receptor 6M1; MCF-7: Michigan cancer foundation-7; NOP: nociceptin/orphanin FQ peptide.

4.1.2. Membrane fragment immobilization

One of the drawbacks of immobilizing whole cells on SPR chips is that the measurement is characterized by a low sensitivity, as seen in Section 4.1.1.

Choi et al. [19] highlighted and tried to solve this aspects in their paper. They performed a preliminary screening on the whole cells to maintain structure and function of the receptors, followed by a more sensitive secondary screening on membrane fragments. In this way, they were able to discover new ligands of the human orphan olfactory receptor (OR)6M1. In the experimental setup, both oxidized intact cells and membrane fragments were covalently immobilized on a classical carboxymethyl dextran sensor chip by the binding between aldehyde group on the cell or membrane surfaces and carbohydrazide on the SPR sensor chip. Therefore, differently from the whole cell immobilization strategies seen so far, the cells were not simply grown on the gold surface of the sensor chip, but they were injected into the instrument and covalently bound to the chip (Fig. 2) [19]. At first, the authors immobilized cells expressing OR6M1 and found that, among 108 compounds screened, only one showed a signal related to its interaction with the immobilized cells. The binding between a ligand and an immobilized cell can be detected when the interaction occurs in proximity of the sensor chip and when it causes the activation of a signaling cascade, but the detection suffers from some drawbacks such as low sensitivity, possibility of non-specific reactions and high noise. To address these issues, the authors decided to decrease the size of the material fixed on the chip immobilizing membrane fragments. To do so, the membranes were fragmented using an appropriate kit for membrane extraction, characterized by scanning electron microscopy (SEM) and dynamic light scattering (DLS) and covalently immobilized on the sensor chip. Using this configuration, it was possible to identify another ligand of the orphan receptor and to confirm the ligand already identified using the whole cells [19]. The researchers demonstrated that the use of membrane fragments allowed to maintain receptor stability, enhance the sensitivity because the size of the immobilized material was included in the SPR detection range and decrease the non-specific interactions compared to the intact cells. In addition, no particular equipment was necessary for these experiments, as a classical SPR instrument with a typical configuration was employed [19]. A summary of this work is reported in Table 1 [[14], [15], [16], [17], [18], [19], [20]].

Fig. 2.

Intact cell and membrane fragment covalent immobilization on the surface plasmon resonance (SPR) sensor chip [19]. EDC: 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride; NHS: N-hydroxysuccinimide. Reprinted from Ref. [19] with permission.

Another possibility to exploit membrane fragments is to use commercial receptor membrane preparations, when available. This strategy was adopted by Murray et al. [20], which investigated the binding to the nociceptin/orphanin FQ peptide (NOP) receptor (also known as ORL1) using a NOP/ORL1 Opioid Receptor Membrane Preparation (Table 1) [[14], [15], [16], [17], [18], [19], [20]]. This consists in a crude membrane preparation deriving from a stable recombinant cell line with high GPCR surface expression, recommended for high throughput screening of NOP ligands [21]. In this case, an unusual immobilization strategy was followed. The procedure included the immersion of the gold chip in ethanol and its irradiation with UV light before its coating with a fibronectin-gelatin mixture. After chip incubation at 4 °C for 24 h, it was put in contact with the membrane preparation and incubated again for 2 h at 4 °C. It was then equilibrated with the buffer and inserted into the SPR. Using this procedure, good signals were obtained for the interaction between neuropeptide phytochemical conjugates and the NOP receptor, demonstrating the efficacy of the immobilization and the suitability of the membrane preparations for this kind of study [20].

Hence, the use of membrane fragments can solve some of the issues typical of whole-cell immobilization, while preserving the GPCR natural lipidic environment. However, with this approach it is not possible to derive information about intracellular signaling.

4.2. Use of membrane mimetics

Besides the use of native membranes, several approaches for the maintenance of GPCR stability via membrane mimetics have been described. These include lipoparticles, lentiviral particles, liposomes, lipoproteins, nanodiscs and planar lipid membranes.

A paper by Heym et al. [22] describes the use of lipoparticles expressing the GPCR chemokine (C−X–C motif) receptor 4 (CXCR4) for SPR and surface acoustic wave (SAW) measurements. Both the immobilization of whole cells and the use of purified receptors present different challenges for SPR studies of GPCRs. Therefore, in this study the authors immobilized lipoparticles (also known as virus-like particles), which present several advantages, such as the stability, the high level of expression of the receptor, the small size compared to intact cells, the uniformity, and the fixed receptor orientation. The lipoparticles used in this study were produced in HEK-293 cells and included proteins from murine leukemia virus. The investigated lipoparticles expressed the human CXCR4 presenting a His-tag at the C-terminus; both biotinylated and non-biotinylated lipoparticles were considered and lipoparticles not expressing CXCR4 were used as controls. In the paper by Heym et al. [22], different immobilization approaches on different sensor chips have been tested. In fact, a high level of immobilization of lipoparticles with active receptors is necessary to obtain a suitable signal. It was observed that higher immobilization levels and interaction signals were obtained using matrix-free C1 chips compared to chips presenting a dextran matrix, probably due to the higher distance of the lipoparticles from the sensor surface when immobilized on the dextran matrix. The best results were obtained when capturing the lipoparticles on a matrix-free carboxymethylated gold chip coated with wheat germ agglutinin (WGA). Under the same conditions, better signals were revealed by SAW than by SPR, due to the different detection principle. While SPR detects changes in refractive index deriving from the binding between the flowing analyte and the immobilized ligand, SAW can detect changes in mass and in the viscoelasticity (produced for example by conformational changes in the immobilized protein) on the sensor surface which cause phase shifts or variations in the amplitude of the acoustic wave, respectively. Probably, the higher signal detected by SAW is due to a conformational change of CXCR4 upon ligand binding. The best immobilization and detection conditions were applied to the kinetic monitoring of a small-molecule ligand of CXCR4 and this work offers the first demonstration of label-free kinetic analysis of a GPCR within a membrane environment [22]. The use of lipoparticles expressing GPCRs at high levels and the detection by SAW proved therefore to be efficient for GPCR ligand analyses.

Another study describing the use of CXCR4 lipoparticles is the one by Griffiths et al. [23], which used lipoparticles produced by an external supplier. The lipoparticles presented the CXCR4 on their surface and were biotinylated to allow the capture on streptavidin-coated chips. Streptavidin was covalently immobilized on the sensor chip by amine coupling and the CXCR4 lipoparticles were subsequently captured by the binding between immobilized streptavidin and lipoparticle biotin. The obtained SPR sensor chip was successfully applied to the investigation of the interaction between CXCR4 and a library of i-bodies, consisting in single-domain antibody-like scaffolds. Interestingly, an SPR assay was also developed to assess the competition between a known CXCR4 ligand and the i-body. Together with the previous example, this paper confirms the potential of GPCR-expressing lipoparticles for SPR measurements.

Martínez-Muñoz et al. [24] described the use of lentiviral particles to maintain the native GPCR membrane environment for SPR analyses. The advantages of this method are that it does not require receptor manipulation and that it allows to retain the different conformations that the receptor can assume. Lentiviral particles were obtained by the transfection of HEK-293T cells with the selected GPCR, CXCR4, and with different lentiviral plasmids (Fig. 3A) [24]. Then, they were characterized by enzyme-linked immunosorbent assay (ELISA), flow cytometry, and electron microscopy to confirm the presence of CXCR4 on the surface of lentiviral particles. Researchers immobilized the lentiviral particles on two different SPR sensor chips, the first homemade and the second consisting of a carboxymethylated dextran commercial chip [24]. The immobilization strategies aimed at preserving receptor activity while ensuring an optimal density of lentiviral particles on the chip surface and preventing non-specific interactions. The homemade chip was composed of a gold sensor coated by a self-assembled monolayer of mercaptoundecanoic acid (Fig. 3B) [24], since it is known that this kind of coating enables an easy, oriented and reproducible immobilization with limited non-specific interactions. Lentiviral particles carrying CXCR4 were immobilized both on the homemade chip and on the commercial one by a classical amine coupling procedure. The proper immobilization of lentiviral particles on the chip surface can be verified by injecting anti-CXCR4 antibodies or CXCR4 known ligands. Advantages of using lentiviral particles [24] include their easy coupling to the SPR chips by classical immobilization approaches and the fact that it is not necessary to solubilize, stabilize, purify and reconstitute GPCRs. In addition, the native membrane environment of GPCRs is preserved, allowing their long-term stability.

Fig. 3.

Use of lentiviral particles for surface plasmon resonance (SPR) analyses of G protein-coupled receptors (GPCRs). (A) Production of lentiviral particles expressing the chemokine (C−X–C motif) receptor 4 (CXCR4) GPCR. (B) Representation of the immobilization of lentiviral particles on a SPR chip [24]. pLVTHM, psPAX2, and pMD2G are the names of the plasmids. GFP: green fluorescent protein; VSV-G: vesicular stomatitis virus envelope glycoprotein; pGPCR: pasmid containing the G protein-coupled receptor of interest; HEK: human embryonic kidney; LVP: lentiviral particles; SAM: self-assembled monolayers. Reprinted from Ref. [24] with permission.

In a work by Oh et al. [25], the use of the human olfactory receptor human olfactory receptor 3A1 (hOR3A1) in liposomes has been evaluated. The difficulties in the culture of olfactory sensory neurons and the low expression efficacy of olfactory receptors in different heterologous expression systems such as mammalian cells and yeast prompted the authors to express hOR3A1 in Escherichia coli (E. coli). This system allows for high expression efficiency, but produces olfactory receptors mainly as inclusion bodies. Therefore, the authors employed lipid/detergent mixed micelles to obtain hOR3A1 in liposomes. They studied different conditions to obtain the liposomes containing the receptor and they analyzed these structures by SPR. To form hOR3A1-containing liposomes, while maintaining hOR3A1 function during the reconstitution process, the researchers applied the Chapso detergent, which is known to be appropriate for GPCR solubilization. Lipid/detergent mixed micelles were formed and mixed with solubilized hOR3A1 to obtain the liposomes (Fig. 4) [25]. The authors confirmed the presence of the receptor inside the liposomes by fluorescence microscopy, since hOR3A1 was conjugated to a fluorescent dye, and Western blotting. Then, they selected liposomes of a 40–50 nm size for an efficient detection in SPR. The SPR technique was applied to confirm the maintenance of hOR3A1 function in the liposomes. The liposomes were captured on the sensor chip coated with poly d-lysine thanks to the interaction between the positive charges of the sensor surface and the negative charges of the liposomes. By comparing the response of hOR3A1-cointaining liposomes with the one of negative control liposomes (without hOR3A1), the authors demonstrated that the receptor maintained its structure, function and selectivity and was able to bind to a known ligand. This protocol can be applied to different GPCRs and when an expression system such as E. coli is necessary to obtain high receptor expression levels. The reconstitution of GPCRs in liposomes can preserve their function and lead to good SPR responses [25]. However, compared to other strategies, the process results laborious.

Fig. 4.

Scheme of the process of liposome formation and surface plasmon resonance (SPR) analysis [25]. OR: olfactory receptor. Reprinted from Ref. [25] with permission.

Segala et al. [26] used high-density lipoprotein (HDL) particles to reconstitute the adenosine A_2A receptor. Reconstituted HDLs consist in self-assembled, disc-shaped fragments of nanolipid bilayers, which are kept stable in solution by amphipathic helical scaffold proteins. These membrane mimetics are both stable and highly soluble, with a lipid composition that can be precisely controlled. The A_2A receptor used in this study [26] was engineered to increase its stability, was isolated from the cell culture after cell disruption and membrane solubilization, purified by affinity chromatography and reconstituted into HDLs. For the reconstitution, the A_2A receptor was incubated with phospholipids and the Zap1 scaffold protein and the self-assembly was started upon removal of the detergent. The desired reconstituted HDLs were selected by gel filtration. The HDLs including A_2A receptor were captured on a nitrilotriacetic acid (NTA) sensor chip thanks to the A_2A His-tag. The binding of nine antagonists with different molecular weights was then tested and confirmed the correct immobilization of the receptor-HDL complexes and the maintenance of receptor activity. By comparing the SPR measurements on chips obtained by the immobilization of A_2A-HDL complexes or of the receptor in detergent micelles, the authors concluded that both approaches are valuable and can provide reliable kinetic and affinity data. While detergent micelles are characterized by an easier production and by a higher sensitivity in SPR analyses, HDLs provide a more natural environment, prevent issues related to ligands partitioning into micelles and facilitates the study of ligand binding to GPCRs in the presence of other elements, like G proteins [26].

Another tool that can mimic the membrane environment consists of nanodiscs, which are lipid bilayer structures with a hydrophobic core surrounded by a ring composed of a dimer of membrane scaffold protein (MSP) derived from the human Apolipoprotein 1A. These structures can encapsulate membrane proteins, stabilizing them and allowing for SPR measurements after the immobilization of the complex, generally through an engineered MSP. The paper by Bocquet et al. [27] described for the first time the SPR kinetic evaluation of small molecules interacting with a human GPCR in nanodiscs. The authors incorporated the human A_2A receptor into nanodiscs comprising a modified MSP with a C9 tag at the C-terminus. In this way, the GPCR/nanodisc complex can be immobilized on the SPR sensor chip by the 1D4 antibody recognizing the C9 epitope. The A_2A/nanodisc complexes used in this work [27] were obtained by mixing the receptor, the modified MSP and the lipids in a buffer containing detergents; the three components self-assembled after detergent removal and the nanodiscs not including the receptor were discarded by a step of affinity chromatography (Fig. 5) [27]. The best conditions to obtain the complexes were studied in order to obtain homogeneous samples. The A_2A receptor included in the nanodiscs carried some mutations compared to the wild type protein to enhance its stability; in addition, it was fused at the C-terminus with the green fluorescent protein (GFP) to monitor the nanodiscs by fluorescence and with a 10×His tag used for the immobilization of the nanodiscs. Two immobilization strategies were tested: the affinity capture of the A_2A His-tag on a nickel NTA sensor chip followed by cross-linking and the capture of the C9 tag of MSP on a 1D4 antibody covalently immobilized on a CM5 or CM7 sensor chip (Fig. 5) [27]. Empty nanodiscs were immobilized on the reference channel of the sensor chip. The researchers demonstrated that the kinetic parameters of A_2A-small molecules interactions were not affected by the different immobilization strategies. Importantly, the chips obtained after nanodisc immobilization were proved to provide reproducible results and to have an improved stability compared to detergent solubilized A_2A receptors, maintaining their activity for several weeks instead of a few days. A higher affinity was observed compared to the detergent solubilized receptor, probably due to the better partitioning of the hydrophobic ligands inside the lipid bilayer of the nanodiscs mimicking the cell membrane. Overall, Bocquet et al. [27] demonstrated that nanodiscs provide a simplified lipid environment that allows for SPR analyses, maintaining receptor stability and leaving both the intracellular and the extracellular sides of the receptor available for ligand binding.

Fig. 5.

Procedure for the formation of nanodiscs and their immobilization on two different surface plasmon resonance (SPR) chips [27]. GFP: green fluorescent protein; eGFP: enhanced green fluorescent protein; CHS: cholesterol hemisuccinate; DDM: n-dodecyl-β-d-maltopyranoside; His: histidine; TEV: tobacco etch virus; IMAC: immobilized metal affinity chromatography; NTA: nitrilotriacetic acid; CM5: carboxymethylated dextran matrix. Reprinted from Ref. [27] with permission.

Another paper describing the use of nanodiscs is the one by Adamson and Watts [28]. The nanodiscs were used to reconstitute the neurotensin receptor type 1 (NTS1). In this case, the nanodiscs were both immobilized on the sensor chip and injected in the buffer as analytes. The authors mixed lipids, MSP and the receptor in a ratio allowing to obtain nanodiscs including mainly NTS1 monomers, then isolated the receptor-containing nanodiscs by affinity chromatography thanks to a tag present on the receptor and obtained homogeneous nanodiscs after a step of size exclusion chromatography. After their characterization, nanodiscs were applied to SPR analyses. In the experiment configuration in which nanodiscs were immobilized on the sensor chip, they were captured on an L1 chip specific for the direct attachment of vesicles and liposomes. Instead, when nanodiscs were used as analytes, the α subunits of G_s or G_i1 proteins were immobilized by a classical amine coupling on a CM5 chip characterized by a matrix of carboxymethylated dextran. In both cases, it was possible to define the kinetics of the interaction between the receptor in nanodiscs and G proteins and this was the first time in which GPCR nanodiscs were used as analytes, allowing to avoid drift from the chip [28].

Overall, the formation of nanodiscs requires different steps but it provides several advantages compared to the free receptor, among which an excellent stability of the complex.

Gessesse et al. [29] investigated by SPR two GPCRs in lipid nanodiscs and in micelles. They produced C−X₃−C motif chemokine receptor 1 (CX₃CR1) and C–C chemokine receptor type 5 (CCR5) in a cell-free system and integrated them into the lipidic or micellar structures. To form the nanodiscs, a mixture of lipids was solubilized in a detergent-containing buffer and was incubated with MSP for 1 h at 4 °C, followed by detergent removal to allow nanodisc formation. Another incubation step was performed, followed by two centrifugation steps and gel filtration for nanodisc purification. Nanodiscs were added to the Protein synthesis Using Recombinant Elements system (PURE system) to incorporate the produced GPCRs. For micelle formation, brain polar lipids were mixed with a detergent and the mixture was sonicated. The micelles were then incubated with the PURE system to integrate the receptors. Purification of both GPCR nanodiscs and micelles was carried out by centrifugation and immobilized metal affinity chromatography. For SPR analyses, two chips were prepared. For nanodiscs, an anti-His antibody was covalently immobilized by amine coupling on a CM5 chip and was used to capture the His-tagged receptor-nanodisc complex, while for micelles the CX₃CL1 ligand with a Strep-tag II was captured thanks to a strepMAB-Immo antibody. To assess the binding, the CX₃CL1 ligand was injected at different concentrations on the nanodisc chip, while CX₃CR1 micelles were injected on the second chip with the captured ligand. Data confirmed that both nanodiscs and micelles enable to preserve receptor binding activity [29]. Gessesse et al. [29] further confirmed the potential of nanodiscs as membrane mimetics to maintain GPCR structure and function, even when produced in a cell-free system, and showed that also the use of micelles can be a valuable approach.

A more recent publication [30] describes the use of nanodiscs to stabilize the A_2A receptor, which was captured on the SPR sensor chip thanks to a monoclonal antibody specific for nanodiscs. The monoclonal antibody specifically recognized the MSP of nanodiscs and was covalently immobilized on a CM5 sensor chip by amine coupling. Thanks to its high affinity for the MSP, the antibody was able to capture the nanodiscs containing the A_2A receptor. On the reference channel, nanodiscs without the receptor were captured by the antibody. In fact, when working with complex systems, it is essential to obtain proper reference channels with the same features of the channel in which the receptor is present (except for the receptor) to evaluate possible non-specific interactions. The entities that are immobilized on the sensor chip surface with the target receptor can significantly impact the goodness of the results if a suitable reference channel is not prepared. The authors of this paper also evaluated the dissociation of nanodiscs from the antibody, which consisted in 10–15 RU/h. This is another reason why it is important to capture nanodiscs also on the reference channel, in order to have comparable sensor surfaces on the two channels. The resulting chip was successfully used for kinetic analyses of small molecule ligands towards the A_2A receptor. Interestingly, since the nanodiscs are not covalently immobilized, it is possible to regenerate the chip injecting a solution of 100 mM glycine HCl, pH 2.6. In this way, the nanodiscs dissociate from the immobilized antibody, which is free to bind new nanodiscs. In the paper by Nakagawa et al. [30], the use of nanodiscs in SPR is further explored and the application of an anti-MSP antibody is interesting because it allows to capture nanodiscs carrying different receptors, leaving them free to interact with their ligands.

A further possibility to reconstitute GPCRs is the use of planar lipid membranes. In a work published in 2016, Calmet et al. [31] used this strategy to obtain stable GPCRs for analyses in plasmon waveguide resonance (PWR), a variant of SPR. The GPCR analyzed in this study is again CCR5, produced as a stabilized construct. The planar lipid bilayer was formed directly in the PWR instrument by using a mixture of lipids and the formation was followed by PWR measurements. The inclusion of the detergent-solubilized CCR5 in the planar membrane was performed by the detergent-dilution method. Specifically, the receptor was solubilized in decylmaltoside (DM) and cholesteryl hemisuccinate (CHS), diluted into a milder detergent and added to the cell sample including the membrane, leading to the spontaneous incorporation of CCR5 inside the membrane. This is the first report in which the kinetics of such a process was monitored. This was possible because PWR is able to measure changes in mass resulting from the receptor inclusion, changes in the thickness of the bilayer and anisotropy changes in oriented systems such as the one of the lipid bilayer. The receptor reconstitution inside the membranes took about 1 or 2 h and it was considered completed when no further changes in resonance spectra were observed. To confirm the maintenance of the receptor activity after reconstitution in the lipid bilayer, its ability to bind to the allosteric inhibitor maraviroc was investigated. Analyses showed that the ligand induced changes in resonance position related to changes in receptor conformation upon maraviroc binding, since no changes were observed when the same experiment was performed using a lipid bilayer not including the receptor. The presented method also allowed for the study of the effect of membrane lipid composition on the kinetics of receptor reconstitution and on its binding to maraviroc. Differently from most of the methods employed in this field, the PWR method developed in this work [31] allowed both to directly monitor the reconstitution of CCR5 receptor in the planar lipid membrane due to the detection of changes in mass density and to assess the receptor activity by measuring its binding to a ligand. The advantages are that both aspects are evaluated in real-time, with a single label-free experiment and with a very high sensitivity.

This publication [31] also emphasizes that all the reported immobilization methods can be applied to other techniques besides SPR, such as PWR, SAW, BLI and grating-coupled interferometry (GCI). In general, all the techniques involving the immobilization of one of the interaction partners on a sensor chip can exploit this kind of immobilization methods.

Table 2 [[22], [23], [24], [25], [26], [27], [28], [29], [30], [31]] summarizes the main findings of the works using membrane mimetics to stabilize GPCRs. Additional works for further reference can be found at [[32], [33], [34], [35], [36]].

Table 2.

Key parameters of the works concerning membrane mimetic immobilization and their main findings.

GPCR	Membrane mimetics	Technique	Findings	Refs.
CXCR4	Lipoparticles	SPR and SAW	Lipoparticles present different advantages compared to whole cells and purified receptors; use of SAW as an alternative label-free technique	[22]
CXCR4	Lipoparticles	SPR	Biotinylated lipoparticles were successfully captured on a streptavidin-coated SPR chip and used for kinetic analyses.	[23]
CXCR4	Lentiviral particles	SPR	Lentiviral particles can be immobilized on SPR chips by standard approaches and they present several advantages compared to isolated receptors.	[24]
hOR3A1	Liposomes	SPR	Use of liposomes to reconstruct the receptor expressed in E. coli as inclusion bodies.	[25]
Human A2A receptor	HDLs	SPR	HDLs provide a native-like environment for GPCRs allowing to reliably define binding kinetics of small ligands.	[26]
Human A2A receptor	Nanodiscs	SPR	Nanodiscs allow for a stable membrane-like environment promoting hydrophobic ligand binding compared to detergent-solubilized receptors.	[27]
NTS1	Nanodiscs	SPR	Nanodiscs can be used also as analytes in SPR measurements.	[28]
CX3CR1 and CCR5	Nanodiscs and micelles	SPR	Cell-free systems can be used to produce GPCRs, which could be stabilized by nanodiscs and micelles.	[29]
A2A receptor	Nanodiscs	SPR	An anti-MSP monoclonal antibody was used to capture the nanodiscs on the SPR chip surface; the chip can be regenerated	[30]
CCR5	Planar lipid membranes	PWR	PWR can be used to directly monitor the reconstitution of GPCRs in planar lipid membranes.	[31]

GPCR: G protein-coupled receptor; CXCR4: chemokine (C−X–C motif) receptor 4; SPR: surface plasmon resonance; SAW: surface acoustic wave; hOR3A1: human olfactory receptor 3A1; A_2A receptor: adenosine 2A receptor; HDLs: high-density lipoproteins; NTS1: neurotensin receptor type 1; CX₃CR1: C−X₃−C motif chemokine receptor 1; CCR5: C–C chemokine receptor type 5; PWR: plasmon waveguide resonance.

4.3. Immobilization of the isolated receptor

Besides immobilization of GPCRs in a lipid membrane environment, it is also possible to use the receptor isolated from the membranes. This approach has been used by several research groups, following different strategies to stabilize the GPCR.

Chu et al. [37] studied the human CXCR5 receptor without introducing stabilizing mutations. The receptor carried a histidine and HPC4 tandem tag for its purification and capture. The receptor was expressed in insect Sf9 cells, which were disrupted by homogenization for its isolation. The membranes deriving from the process were purified and solubilized by the addition of a detergent mixture and the receptor was subsequently purified by affinity chromatography. After purification, CXCR5 was directly captured on a NTA SPR chip thanks to its 6× His tag and was stabilized by an optimized cross-linking step. The stability of the immobilized receptor was verified by analyzing the binding of the CXCL13 ligand and the anti-CXCR5 monoclonal antibody to CXCR5; high-quality data were obtained, confirming the stability and the activity of the receptor. The cross-linking step introduced by the authors stabilized the receptor in a way that the sensor chip was resistant to harsh conditions such as the ones used for soluble proteins (i.e. use of surfactants and low pH regeneration). Therefore, this approach allowed to obtain stable receptors for SPR analyses without the need for time-consuming protein engineering.

Shepherd et al. [38] also describe the use of wild type, non-mutated GPCRs. In their study, the human A_2A receptor was considered. The native receptor including also an His-tag was expressed in HEK cells, which were resuspended in a buffer containing detergents to stabilize membrane proteins, sonicated and centrifuged. A_2A receptor was purified using an affinity resin. Then, the capture of the purified receptor on a nickel-NTA SPR chip was carried out thanks to the receptor His-tag. The running buffer included three different detergents to create a suitable environment for A_2A receptor stability. Other strategies used to increase receptor stability included its cross-linking on the chip and the use of a low temperature (10 °C) for the analyses. The activity of the captured and cross-linked A_2A receptor was confirmed by analyzing known ligands, for which the obtained affinity constants were in agreement with literature. The authors applied the A_2A receptor chip to fragment screening and subsequently analyzed the selected fragments also on chips obtained by the capture and cross-linking of other adenosine receptors (A₁, A_2B and A₃). All the prepared sensor chips proved to maintain receptor activity and stability, allowing to monitor the kinetics of ligand interactions, confirming the validity of the developed procedure for SPR analyses of GPCR. In addition, the proposed assay confirms the potential of SPR in screening of GPCR ligands.

In a paper by Boonen et al. [39], the chemokine receptor CXCR4 was captured on a NTA sensor chip thanks to its 10× His-tag. The receptor was expressed in Sf9 insect cells, which were disrupted and homogenized to separate the membrane fraction. CXCR4 was isolated after the addition of a detergent-containing buffer to the membranes and centrifugation to collect the supernatant. As in the previous examples, no engineering of the receptor was performed to increase its stability. The membrane fraction including the receptor was injected on an activated NTA chip for the capture of CXCR4 thanks to its His-tag. The running buffer for SPR analyses comprised detergents for receptor stabilization, since they mimic the receptor natural environment. Since the capture of His-tagged proteins on NTA chips is not characterized by a strong interaction, there is the possibility of a release of the captured protein. For this reason, the authors included a 30-min equilibration step in which the running buffer was flowed over the chip surface. By investigating the binding of a CXCR4 natural ligand and of two anti-CXCR4 antibodies to the immobilized CXCR4, the authors demonstrated that the receptor maintained its functionality when analyzed in a buffer containing a detergent mix, while it was not stable in a denaturing buffer including the surfactant Triton X-100 instead of the detergent mix. The chip with captured CXCR4 was then used to study the kinetics of the interaction with different nanobodies, confirming its binding activity. The developed method does not need a purification of the receptor from the membrane fraction, allowing to save money and time. To verify the absence of non-specific interactions, the authors carried out a negative control experiment on cells not expressing the His-tagged CXCR4 and observed that, even if some components of the membrane fraction were captured on the NTA chip, no significant binding response was observed upon injection of nanobody ligands. This confirmed that the nanobodies bound only to CXCR4. The chips obtained by CXCR4 capture were found to be stable for at least 12 h. The strategy used in this paper allowed to avoid the time-consuming purification, which might increase protein instability; the addition of a detergent mix to the running buffer enabled to stabilize the receptor and to preserve its function and binding capacity for SPR kinetic analyses.

A similar approach was used by Capelli et al. [40] to study the GPR17 receptor. Also in this case, the purification step was avoided and the receptor was captured on the SPR chip from solubilized membrane extracts. In this paper, the receptor was engineered for stabilization and two different variants were produced in Sf9 insect cells. Both variants carried a small soluble protein, the T4 lysozyme, in one of the intracellular loops of GPR17 in order to stabilize it. As reported in the previous example [39], after isolation of the membrane fraction, the receptor was obtained by the addition of a buffer containing a mixture of detergents (dodecyl maltoside (DDM) and CHS). After centrifugation, the receptor was captured on the sensor chip directly from the solubilized membrane fraction by the interaction between an His-tag on GPR17 and an anti-His₆ antibody previously immobilized on the chip by amine coupling. The maintenance of receptor stability and activity was confirmed by the analysis of a receptor antagonist and an agonist. The researchers demonstrated that the developed protocol allowed to retain receptor binding capacity for over 24 h. In addition, the use of the same chip for different experiments was possible after regeneration allowing to separate the captured receptor from the immobilized anti-His₆ antibody.

Another paper describing the use of stabilizing mutations in GPCR was published by Huber et al. [41], which studied the NTS1. They produced the stabilized variant NTS1-H4 by introduction of some truncations and replacement of free cysteine residues to facilitate purification, a source of potential instability, and the following experiments. Differently from the previous examples, the receptor was purified following a multi-step protocol. In particular, a buffer containing a mixture of DDM and CHS detergents was added to the cells, which were lysed and centrifuged. The supernatant underwent purification by immobilized metal affinity chromatography, ion exchange chromatography and size exclusion chromatography. All the buffers used in the different steps included detergents to stabilize the transmembrane receptor. The purified receptor was then captured on streptavidin-coated chips thanks to its avi-tag to which biotin can be covalently bound, allowing the binding to the streptavidin on the chip. The tag was placed distant from the binding site to promote its accessibility. The SPR running buffer contained lauryl maltose neopentyl glycol (LMNG), a detergent to stabilize membrane proteins. The maintenance of receptor binding capacity was confirmed by the SPR analysis of different peptides derived from neurotensin, an endogenous NTS1 agonist, and an antagonist consisting in a small molecule. These results derived from molecules covering a wide range of affinities and suggested that no denaturation occurred during the purification process and during SPR analyses, that receptor binding sites were accessible to the ligands and that the receptor maintained its active conformation. The developed procedure resulted in a long-term stability of NTS1-H4, revealing only a 2% decrease of the SPR signal within 24 h. For this reason, it was possible to successfully apply the SPR assay to a fragment screening of more than 6000 compounds. Interestingly, a method to differentiate between the specific interaction of ligands with the orthosteric binding site and the binding to other sites was also developed. Globally, this work further highlights that purification can cause membrane protein instability and therefore the use of engineering to stabilize the receptor is often necessary; the additional use of detergents for SPR analyses enables the maintenance of receptor stability for kinetic and affinity measurements.

Another work in which the authors used protein engineering and a buffer added with a mixture of detergents to stabilize the GPCR was published by Lu et al. [42] in 2019. The researchers investigated the human A_2A receptor and engineered it so that the fusion protein of apocytochrome b562RIL (BRIL) substituted the third intracellular loop of the GPCR and a truncation of the C-terminus was introduced. The produced receptor also included a FLAG tag and a His-tag. To obtain the purified receptor, the production was performed in Sf9 cells, which were disrupted to collect the membrane fraction. Membranes were resuspended in a buffer including the DDM and CHS detergents and the receptor was purified by immobilized metal affinity chromatography. For SPR analyses, the authors performed a screening of detergents, studied different buffer compositions and identified a proper reagent (theophylline, an A_2A receptor antagonist) to regenerate the receptor after ligand binding. The final SPR running buffer contained CHS and LMNG detergents to stabilize membrane proteins. In addition, SPR experiments were carried out at 10 °C to enhance protein stability. The receptor was captured on an NTA sensor chip by its His-tag and was applied to the validation of potential ligands identified by affinity mass spectrometry. Also in this case, the combination of protein engineering and stabilizing conditions in SPR allowed to obtain affinity and kinetic data of the tested ligands.

The same approach was used by Heine P. and coworkers, which used an engineered NTS1 and a running buffer enriched by detergents for SPR analyses [43]. The variations included in the engineered receptor were N- and C-terminal truncations, removal of some amino acid residues in the third intracellular loop (a flexible region) and mutations of two cysteines to alanine residues to improve the stability. In addition, an Avi-tag was added to the construct. For receptor purification, cell lysis was performed in a buffer including a mixture of DDM and CHS detergents, followed by isolation of the membrane fraction, purification by immobilized metal affinity chromatography, fraction isolation by size exclusion chromatography and concentration by ultracentrifugation using filters with an appropriate cutoff. SPR experiments were carried out on a carboxymethyldextran chip on which neutravidin was immobilized in order to capture the receptor presenting a biotinylated Avi-tag. Measurements were performed in a running buffer including 0.1% (w/v) DDM. The preservation of receptor integrity was assessed after its capture on the chip and at different times during the subsequent screening; to this aim, a known peptide ligand was used. The SPR chip was then applied to the validation of potential ligands identified by a fluorescence polarization assay. Interestingly, the SPR technology was exploited also to define ligand specificity for the receptor orthosteric binding site by competition binding assays. This paper further demonstrates the potential of mutational approaches and use of detergents in GPCR analyses by SPR.

In a work by Avsar et al. [44], the GPCR jumping spider rhodopsin-1 (JSR1) was used as the analyte in SPR analyses, while its ligand arrestin-3 protein was immobilized on the sensor chip. Since the wild-type form of JSR1 is highly thermostable, it was not necessary to introduce stabilizing mutations. Therefore, the wild type receptor was produced in HEK293S GnTI cells, which were subsequently homogenized and added with a buffer containing the DDM detergent. Separation of the receptor from the membranes was performed by ultracentrifugation. Then, purification of JSR1 was achieved by affinity chromatography thanks to a tag linked to the receptor. A buffer containing DDM detergent was used for storage of the purified JSR1. For SPR measurements, an HPA chip characterized by a hydrophobic surface was used for the immobilization of arrestin-3 protein. Vescicles functionalized with Ni-NTA were fused to the chip to create a supported lipid monolayer, on which arrestin-3 (presenting an His-tag) was captured. The GPCR injected on the sensor chip was stabilized thanks to the formation of detergent micelles. Different concentrations of JSR1 were injected on the chip surface and after each injection a regeneration using 200 mM NaOH was carried out. The SPR assay allowed to measure kinetic and equilibrium constants, confirming the binding between JSR1 and arrestin-3. Interestingly, in this paper two other label-free techniques were applied to the same samples, namely BLI and quartz crystal microbalance with dissipation monitoring (QCM-D). All the techniques allowed to monitor the interactions between arrestin-3 and JSR1. Therefore, GPCRs can be used as analytes in SPR and other label-free biosensor techniques if properly stabilized with detergents. This consents to avoid some difficulties associated with the use of immobilized GPCRs on biosensors, such as their isolation, reconstitution and oriented immobilization.

Another paper describing the strategy of using detergents as stabilizing agents is the one by Koretz et al. [45], which investigated the effects of the truncation of the A_2A receptor on its function. The wild type and truncated A_2A receptors were produced in Saccharomyces cerevisiae. The buffer for membrane resuspension after cell lysis contained a mixture of detergents to maintain receptor stability, as already seen in other examples. The receptors were then purified by affinity chromatography; all the buffers used included the detergent mix. As described in the previous example [44], the purified GPCRs were used as the analytes in SPR analyses and were not immobilized on the sensor chip. Instead, the Gα_s protein was captured on a Ni-NTA chip thanks to its His-tag and was investigated as a partner of the interaction with A_2A receptors. Denatured Gα_s protein was used as a negative control. The running buffer for SPR analyses included DDM, CHS and 3-[(3-cholamidopropyl)-dimethylammonio]-1-propane sulfonate (CHAPS) detergents above the critical micelle concentration to guarantee micelle generation and therefore the maintenance of receptor folding when it was injected as an analyte on the sensor chip. Specific binding of A_2A receptor to Gα_s protein was successfully measured and results showed that the introduction of truncations in the receptor decreased the affinity towards Gα_s protein. These two works [44,45] demonstrated that the formation of detergent micelles can provide the proper environment to use GPCRs as analytes in SPR.

A recent work by Yu et al. [46] describes the use of GCI, an evolution of SPR technology, to study the μ-opioid receptor (μOR). As reported in the previous examples, also in this study the GPCR was used as the analyte. A 4PCP chip (functionalized with a thin polycarboxylate layer) was used to immobilize neutravidin, through which a biotinylated nanobody able to bind μOR was captured. The analyte consisted of purified μOR and was injected on the surface of the sensor chip to investigate the interaction, using a running buffer containing the LMNG and CHS detergents. This work further confirmed that the addition of proper detergents to the running buffer allows to use GPCRs as analytes in biosensor analyses.

Talking about the use of detergents, a paper by Navratilova et al. [47] describes a systematic study to find the best solubilization conditions for two model GPCRs, CXCR4 and CCR5. The selection of these receptors was guided by the fact that they share only around 30% identity, therefore the data deriving from the present study could be applicable to many GPCRs. A large-scale screen was performed on 950 combinations of solubilization conditions and the outcome was followed by SPR thanks to the use of anti-CXCR4 or CCR5 antibodies specific for their active conformation. An advantage of using antibodies as analytes in SPR is that, due to their high molecular weight, their binding to the immobilized receptors results in a high signal. As already reported by previous examples, in this paper non-mutated receptors were employed and they were not purified since they were captured on the sensor chips only after membrane fraction solubilization. To compare the different solubilization conditions, the authors considered two parameters: the receptor capture level as an indicator of the efficacy of its removal from membranes and the antibody binding signal as an indicator of the maintenance of receptor correct conformation and folding. After the screening using antibodies, the best conditions were tested also in the binding of small molecule ligands to CXCR4 and CCR5 to confirm the obtained data. This systematic study highlighted that important parameters to test when investigating GPCR solubilization conditions are the type of salt in the buffer, the use of PEG for purification optimization and the use of detergent mixtures. In particular, the combination of different detergents can result optimal even if the single detergent was not effective when used alone. Therefore, a screening of different detergent mixtures is suggested to establish the working conditions for a new GPCR.

Liu et al. [48] described an approach in which the GPCR μOR was stabilized through engineering to obtain a water-soluble protein. To do so, a multi-step protocol was adopted in which the 3D structure of the receptor was simulated for the identification of hydrophobic residues on the surface of the transmembrane area of the protein, followed by computational design, structural analysis and molecular dynamics simulation with the aim to design a receptor stable in aqueous solutions without significantly affecting the structure. After finding the suitable sequence to obtain a stable receptor, the protein was expressed in E. coli and purified. Being water-soluble, the engineered receptor does not require a lipidic environment for its stability, therefore it can be directly used for biosensor analyses. The authors proved the feasibility of these studies by a graphene enabled biosensor and SPR, monitoring the binding of morphine to μOR. In this example, much effort was spent in the engineering of the receptor, which was made stable in a water environment thanks to the introduced mutations and did not need any other stabilizing strategies.

Overall, the use of isolated GPCRs is still the main approach because it is easier and less time-consuming compared to the other strategies seen so far, especially when the receptor is not purified. However, the stability of the receptor is generally lower when using wild type GPCRs, often requiring the use of engineering for further protein stabilization (Table 3) [[37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47], [48]].

Table 3.

Key parameters of the works concerning isolated receptor immobilization and their main findings.

GPCR	Stabilization strategy	Technique	Findings	Refs.
CXCR5	Cross-linking of the captured receptor	SPR	Cross-linking allowed to obtain stable receptors and high-quality data without the need for protein engineering.	[37]
Human A1, A2A, A2B, and A3 receptor	Cross-linking of the captured receptor, low temperatures and use of detergents	SPR	The wild type purified receptor was stabilized by its cross-linking on the chip surface, the use of low temperatures for the analyses and the addition of detergents in the running buffer.	[38]
CXCR4	Use of detergents	SPR	The use of a detergent mix in the SPR running buffer stabilized the receptor. No purification from the membranes was needed before receptor immobilization on the SPR chip.	[39]
GPR17	Engineering of the receptor and use of detergents	SPR	The receptor was stabilized by engineering and the addition of detergents in the running buffer, without the need for purification. It was possible to regenerate the chip and use it for multiple experiments.	[40]
NTS1	Engineering of the receptor and use of detergents	SPR	The receptor was stabilized by mutations and truncations and after purification the use of detergents maintained its stability for SPR analyses.	[41]
A2A receptor	Engineering of the receptor and use of detergents	SPR	The thermostabilized engineered receptor was successfully studied by SPR after screening of detergents, definition of the optimal buffer composition and regeneration conditions; a low temperature was used for SPR analyses.	[42]
NTS1	Engineering of the receptor and use of detergents	SPR	The mutational approach enabled to obtain a stable receptor which was studied by SPR using a running buffer enriched with detergents.	[43]
JSR1	Use of detergents	SPR, BLI, and QCM-D	The stabilization by detergent micelles allowed to inject the receptor as an analyte in SPR analyses.	[44]
A2A receptor	Use of detergents	SPR	The use of a running buffer including detergents above the critical micelle concentration ensured the maintenance of receptor folding when used as an analyte.	[45]
μOR	Use of detergents	GCI	The addition of proper detergents to the running buffer allowed the use of GPCRs as analytes in biosensor analyses.	[46]
CXCR4 and CCR5	Solubilization conditions	SPR	Detergent mixtures can exert a significant influence on receptor extraction from membranes and on the maintenance of its activity.	[47]
μOR	Engineering of the receptor	Graphene enabled biosensor and SPR	Engineering allowed to obtain a stable water-soluble receptor that did not need any further stabilization for biosensor analyses.	[48]

CXCR5: chemokine (C−X–C motif) receptor 5; SPR: surface plasmon resonance; human A₁, A_2A, A_2B, and A₃ receptor: human adenosine 1, 2A, 2B and 3 receptor; GPR17: uracil nucleotide/cysteinyl leukotriene receptor; NTS1: neurotensin receptor type 1; JSR1: jumping spider rhodopsin-1; BLI: biolayer interferometry; QCM-D: quartz crystal microbalance with dissipation monitoring; μOR: μ-opioid receptor; GCI: grating-coupled interferometry; CCR5: chemokine receptor type 5.

Additional works for further reference can be found at [49], a book chapter reviewing SPR analyses of wild-type and thermostabilized GPCRs, with a special focus on the study of orthosteric and allosteric GPCR ligands.

5. Further considerations

Besides the immobilization strategy, another important aspect to take into account when working with GPCRs is to prepare a proper reference channel to evaluate non-specific interactions. In fact, the presence on the sensor chip of other components besides the target receptor (such as whole cells, membrane fragments or membrane mimetics) can lead to non-specific binding and impact on the goodness of the results. The ideal reference channel should possess the same features of the channel in which the receptor is immobilized, except for the receptor. Different approaches have been applied to obtain a suitable reference channel. When using membrane mimetics to immobilize the receptor, the reference channel can be prepared by immobilizing empty membrane mimetics, not including the receptor; as an example, publications [27,30] describe the immobilization of empty nanodiscs on the channel used as a reference, while in paper [22] the authors used empty lipoparticles. Other strategies are used when the isolated receptor is studied. In particular, it is possible to immobilize the receptor also on the reference channel, blocking its binding site with a strong ligand or an irreversible covalent inhibitor [36,41]. In this way, only the specific interaction with the GPCR binding site is measured and the signal of non-specific interactions is subtracted. Another approach consists in the production of a mutant of the receptor with a blocked binding site, which can be immobilized on the reference channel as a negative control [41]. Alternatively, it is possible to prepare a reference channel by immobilizing a receptor type that is related to the target GPCRs. In this case, it is important that the stability and binding activity of the two receptors is maintained in the same conditions [41]. In the work by Capelli et al. [40], the authors used an anti-His₆ antibody to capture the GPR17 receptor; for this reason, they immobilized the same antibody also on the reference channel. The aim also in this case was to have comparable surfaces between the channel in which the receptor is immobilized and the channel considered as a reference.

A further aspect to consider in GPCR studies is the establishment of a suitable regeneration protocol, which should remove any trace of ligand from the receptor while leaving the GPCR intact and active. This is particularly important due to the low stability of these receptors. In addition, the regeneration step should preserve the other components immobilized on the sensor chip together with the receptor. Therefore, a fine balance between the strength of the regeneration solution to remove the bound analyte and its mildness to maintain the integrity of the immobilized entities should be found. A possibility to obtain a mild regeneration is to inject a high concentration of a low-affinity or rapidly-dissociating ligand to displace the analyte bound to the receptor; an example is the use of a 5 mM solution of the antagonist theophylline to regenerate the A_2A receptor [42]. Boonen et al. [39] found that injecting sequentially 350 mM ethylenediaminetetraacetic acid (EDTA), 1 M imidazole and 50 mM NaOH allowed to regenerate the immobilized CXCR4 without affecting its stability. In a study describing the SPR analysis of CXCR4 on lentiviral particles [33], the authors proved that the regeneration by 5 mM HCl allowed to maintain a constant response (with less than 10% variation) for over 20 cycles and they attributed the success of this regeneration protocol to the covalent immobilization used to obtain the sensor chip. In the paper by Chu et al. [37], the authors found that the cross-linking of the isolated CXCR5 allowed the receptor to maintain its stability also after a low pH regeneration step using 50 mM HCl. Therefore, even harsh regeneration conditions could be used when the receptor is particularly stable. The publication [40] describes another type of regeneration, to remove the captured GPR17 from the immobilized anti-His₆ antibody once its stability was lost and to replace it with fresh active receptor. The protocol for the regeneration included some short injections (30 s) of a 50 mM NaOH solution at a flow rate of 50 μL/min. This step was useful since the receptor was stable for around 24 h and by regeneration it was not necessary to prepare a new chip for every experiment. The same concept was exploited by Nakagawa et al. [30], which regenerated their anti-nanodisc antibodies by injecting 100 mM glycine HCl pH 2.6 to obtain a chip surface able to bind new nanodiscs containing the target receptor.

Due to the stability issues of GPCRs, it is also important to ensure the maintenance of their binding activity for multiple cycles. This aspect has been considered in several studies; some examples are reported hereafter. As already mentioned, in publication [33] the authors found that CXCR4 on the surface of lentiviral particles was stable for over 20 cycles in the investigated conditions, with a degree of variation between analyses lower than 10%. In a screening of solubilization conditions for isolated receptors CCR5 and CXCR4 [47], researchers studied the stability by injecting control samples at fixed times during the screening. They observed that the CCR5 receptor gave comparable results for multiple cycles, proving its stability, while CXCR4 showed a decline over time. Therefore, results obtained for CXCR4 were always compared to control samples and in this way the activity loss was considered. The authors of this paper also compared two different instruments for the screening, Biacore 3000 and Biacore 4000. Results showed that around 100 solubilization conditions can be tested on the Biacore 4000, while in Biacore 3000 the stability was lower. In publication [37], the cross-linked CXCR5 exhibited a very high stability and showed comparable results for up to 2000 cycles, even using harsh regeneration conditions (50 mM HCl). Instead, Capelli et al. [40] observed that the isolated GPR17 preserved its binding activity for over 24 h. In general, the use of whole cells, membrane fragments or membrane mimetics increases stability compared to the immobilization of isolated wild type GPCRs; this was demonstrated by Bocquet et al. [27] who showed that the A_2A receptor inside nanodiscs maintained its binding activity for two to three weeks, while the isolated receptor lost its activity in 80 h.

Therefore, different aspects should be considered when working with GPCRs and multiple strategies can be adopted to obtain reliable results in SPR.

6. Conclusions

SPR has emerged as a powerful technique in drug discovery and its use for GPCR tragets is now widespread. However, to face the intrinsic instability of these targets, several strategies have been investigated, especially for the immobilization of GPCRs on the SPR sensor chip.

Target immobilization can be faced following different strategies: by immobilizing the receptor in its native membrane, using a membrane mimetic, employing detergents or applying engineering approaches.

Compared to the immobilization of whole cells, the use of membrane fragments allows to enhance sensitivity since the binding occurs closer to the SPR sensor chip. In addition, a reduction of non-specific interactions is observed when using this immobilization strategy. The same occurs with the use of membrane mimetics where high stability of the receptors can be reached. However, sample preparation and method development are often challenging and time-consuming.

Looking at the recent literature, the most common approach for GPCR study by SPR is the immobilization of the isolated receptor. This results in a more simplified system compared to the use of receptors in native membranes or membrane mimetics. However, this method is characterized by a lower stability of GPCRs and for this reason different strategies for their stabilization have been applied, including a target cross-linking step after immobilization, the use of detergents and low temperatures, or the receptor engineering with the introduction of stabilizing mutations.

Hence, each strategy has its pros and cons, which should be carefully evaluated when approaching the study of a new GPCR. When high sensitivity is required, moving towards smaller immobilized entities (such as the isolated receptor) is mandatory; instead, if the aim is to obtain a high information content analysis, the best choice is to immobilize the receptor in conditions mimicking its natural environment.

Other aspects to take into consideration when working with GPCRs in SPR are the preparation of a suitable reference channel to subtract the contribution of non-specific binding, the use of proper regeneration solutions to avoid receptor loss of activity, and the evaluation of receptor stability over time for multiple cycles.

CRediT authorship contribution statement

Giulia De Soricellis: Writing – original draft. Enrica Calleri: Writing – review & editing, Resources, Supervision. Sofia Salerno: Writing – review & editing. Gloria Brusotti: Writing – review & editing. Sara Tengattini: Writing – review & editing. Caterina Temporini: Writing – review & editing. Gabriella Massolini: Writing – review & editing. Francesca Rinaldi: Writing – review & editing, Conceptualization, Supervision, Writing – original draft.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.