We genetically-encode the unnatural amino acid, p-azido-L-phenylalanine at various targeted positions in GPCRs and show the versatility of the azido group in different applications. These include a targeted photocrosslinking technology to identify residues in the ligand-binding pocket of a GPCR, and site-specific bioorthogonal modification of GPCRs with a peptide-epitope tag or fluorescent probe.
To facilitate structural and dynamic studies of G protein-coupled receptor (GPCR) signaling complexes, new approaches are required to introduce informative probes or labels into expressed receptors that do not perturb receptor function. We used amber codon suppression technology to genetically-encode the unnatural amino acid, p-azido-L-phenylalanine (azF) at various targeted positions in GPCRs heterologously expressed in mammalian cells. The versatility of the azido group is illustrated here in different applications to study GPCRs in their native cellular environment or under detergent solubilized conditions. First, we demonstrate a cell-based targeted photocrosslinking technology to identify the residues in the ligand-binding pocket of GPCR where a tritium-labeled small-molecule ligand is crosslinked to a genetically-encoded azido amino acid. We then demonstrate site-specific modification of GPCRs by the bioorthogonal Staudinger-Bertozzi ligation reaction that targets the azido group using phosphine derivatives. We discuss a general strategy for targeted peptide-epitope tagging of expressed membrane proteins in-culture and its detection using a whole-cell-based ELISA approach. Finally, we show that azF-GPCRs can be selectively tagged with fluorescent probes. The methodologies discussed are general, in that they can in principle be applied to any amino acid position in any expressed GPCR to interrogate active signaling complexes.
Heptahelical G protein-coupled receptors (GPCRs) comprise a superfamily of highly dynamic membrane proteins that mediate important and diverse extracellular signals. In the classical paradigm, receptor activation is coupled with ligand-induced conformational changes.1, 2 Recent advancements in structural biology of GPCRs have provided significant insight on the molecular mechanisms of transmembrane signaling.3-5 However, to understand with greater chemical precision the functional mechanism and structural dynamics of GPCR signaling, a toolkit of approaches is required to incorporate informative molecular and chemical probes to interrogate active signaling complexes.
To this end, we adapted a method to site-specifically introduce non- or minimally perturbing probes into expressed receptors based on unnatural amino acid (UAA) mutagenesis using amber codon suppression technology previously pioneered by Schultz and coworkers.6 We optimized the UAA mutagenesis methodology to achieve a high-yield expression and mutagenesis system for proteins such as GPCRs, which are difficult to express in most heterologous systems other than mammalian cells. Using an orthogonal engineered suppressor tRNA and evolved aminoacyl-tRNA synthetase pair for a specific UAA, we site-specifically introduced UAAs in expressed target GPCRs. The successful incorporation of UAAs, p-acetyl-L-phenylalanine (AcF), p-benzoyl-L-phenylalanine (BzF), and p-azido-L-phenylalanine (azF) has been demonstrated in our model GPCRs – rhodopsin and human C-C chemokine receptor, CCR5.7, 8
In principle, an UAA can be genetically encoded at any position within the protein sequence and this property is an invaluable biochemical tool as it enables single-codon scanning of a target GPCR. We focus here specifically on the versatility of the UAA, azF, which has a reactive azido moiety. In addition to serving as a unique infrared (IR) probe,8, 9 azF can also serve as a photoactivatable cross-linker by reacting with neighboring primary amines or aliphatic hydrogens. Additionally, the biologically inert azido group can participate as a selective chemical handle in bioorthogonal labeling reactions. Here we present examples illustrating the useful applications of site-specific incorporation of azF into GPCRs, such as targeted photocrosslinking to trap a receptor-ligand complex, and modification of GPCRs by bioorthogonal epitope tagging and fluorescent labeling strategies.
Photoactivatable reagents have been used to study biological systems since the 1960s.10 In this period, an abundance of receptor-ligand crosslinking experiments have been reported to study GPCR complexes, most of which involved the use of photoaffinity ligands.11, 12 However, these applications are technically limited, as they require synthesis of ligands bearing a crosslinking group.13-15 Moreover, with the crosslinker moiety in the ligand, it is challenging to identify the location of the crosslink on the GPCR. Site-specifically introducing photocrosslinking groups as UAAs into proteins using the amber codon suppression technology is a valuable advancement. 16, 17 We developed a photocrosslinking technique to identify the binding interface on a receptor that is involved in the formation of a receptor-ligand complex in live cells by introducing photolabile groups into GPCRs.18, 19 Here we describe the experimental protocol and method of data analysis for applying this targeted photocrosslinking technology to identify the binding site of a small-molecule ligand, tritiated maraviroc, on CCR5. This method capitalizes on the precise quantification of the radioactive handle on the ligand, in addition to retaining the native chemical structure of the ligand.
Fluorescence-based techniques support the precise understanding of the structural basis of receptor activation by directly probing the conformational state of the receptor.20, 21 However, techniques possessing the flexibility to introduce fluorescent labels into GPCRs site-specifically are limited. We are interested in employing bioorthogonal chemical modification strategies to facilitate single-molecule detection (SMD) of GPCR signaling complexes.22 The azido group can participate in bioorthogonal chemistries such as Staudinger-Bertozzi ligation,23, 24 copper-free strain-promoted azide-alkyne cycloaddition (SpAAC)25 and copper-catalyzed azide-alkyne cycloaddition (CuAAC).26 We focus here on the Staudinger-Bertozzi ligation reaction that involves the specific reaction between an azide and a phosphine. We demonstrate the use of two different phosphine derivatives, conjugated to a peptide epitope (FLAG peptide) or a fluorescent label (fluorescein) to achieve site-specific modification of GPCRs.
We previously optimized the conditions for site-specific labeling of rhodopsin azF variants using the X-ray crystal structure and dynamic simulations to choose target sites that are solvent exposed. 27, 28 We also illustrated the feasibility to achieve background-free labeling by Staudinger-Bertozzi ligation.29 We demonstrate here the general procedure employed to achieve fluorescent labeling of a detergent-solubilized receptor that is immobilized on an immunoaffinity matrix and subsequently visualized by in-gel fluorescence. Additionally, we demonstrate a useful extension on this labeling strategy to identify positions amenable to labeling on a receptor of unknown structure, CCR5. This is carried out using a targeted peptide-epitope tagging strategy that relies on bioorthogonal modification of azF-GPCR variants in a cell-based semi-high throughput format.30 This method exploits the multi-step detection properties of a cell-surface ELISA to monitor labeling events.
The methodologies we discuss here are general and in principle can be applied to any GPCR incorporated with azF using the amber codon suppression technology. In the protocols presented here we detail the steps involved in mammalian cell expression of receptors incorporated with UAAs such as azF using the unnatural amino acid mutagenesis method and their subsequent applications to facilitate structural and dynamic studies of GPCRs.
1. Site-specific Genetic Incorporation of Unnatural Amino Acids into GPCRs
2. Mapping a Ligand Binding Site Using Targeted Photocrosslinking
3. Targeted Epitope Tagging and Cell Surface Enzyme-linked Immunosorbent Assay (ELISA)
4. Bioorthogonal Fluorescent Labeling of a GPCR
We employed the unnatural amino acid mutagenesis methodology to site-specifically introduce molecular probes into a GPCR using the amber codon suppression technology. The scheme in Figure 1 outlines the salient steps of the methodology and the various applications of incorporating the versatile UAA, p-azido-L-phenylalanine (azF), into GPCRs. Expression of azF-GPCRs in mammalian cells enables targeted photocrosslinking to ligands, and targeted peptide-epitope tagging or fluorescent modifications of a GPCR via bioorthogonal labeling chemistries.
The photocrosslinking technique can be used to investigate the structural aspects of GPCR-ligand complexes in live cells (Figure 1a). Representative data obtained from one such experiment is shown in Figure 2.19 Here we used a radioactive labeled small-molecule ligand, tritiated maraviroc, to identify its binding site on CCR5. In this example, UV irradiation on cells expressing azF-CCR5 variants pre-incubated with tritiated maraviroc results in covalently crosslinked complexes that can be identified using the sensitive radioactive tag on the ligand. Amongst the several positions on CCR5 tested, I28azF and W86azF were able to photocrosslink to tritiated maraviroc as depicted with significantly higher scintillation counts (red bars, bottom panel, Figure 2). The wt receptor showed no crosslinking to the radioactive ligand.
Bioorthogonal labeling chemistries can be used to site-specifically modify azF-GPCR variants. We employed the Staudinger-Bertozzi ligation chemistry to label two GPCRs: CCR5 and rhodopsin using phosphine derivatives (Scheme 1). One strategy involves the targeted epitope tagging of a GPCR in live cells using a peptide-conjugated label (Figure 1b). We demonstrate one such experiment, in which a FLAG peptide-triarylphosphine labeling reagent is used to site-specifically and selectively modify azF-CCR5 variants. In Figure 3 we display representative data from a sensitive and multi-step ELISA detection strategy used to identify positions that are tagged with the FLAG peptide.30 Amongst the positions on CCR5 investigated, all azF-CCR5 variants were expressed at the cell surface (red bars, Figure 3) and approximately half of them successfully underwent targeted epitope tagging (blue bars, Figure 3). wt CCR5 served as a negative control for epitope tagging, exhibited by the lack of anti-FLAG signal. In an alternative strategy, we labeled detergent-solubilized azF-GPCRs immobilized on an immunoaffinity matrix, such as anti-1D4 mAb conjugated to sepharose 2B resin, with a fluorophore-conjugated label (Figure 1c). The in-gel fluorescence image shown in Figure 4 demonstrates the successful fluorescent labeling of the GPCR, rhodopsin, at three different positions incorporated with azF by the Staudinger-Bertozzi ligation using a phosphine derivative of fluorescein.29 As expected, the wt rhodopsin showed no background labeling as seen by the absence of a fluorescent band at the expected molecular weight of the receptor.
Scheme 1. The Staudinger-Bertozzi ligation reaction between a triarylphosphine and an azide to form a stable amide-linked adduct.
Figure 1. Scheme illustrating the site-specific incorporation of an UAA (azF) and its applications to study GPCRs. Site-specific UAA mutagenesis is achieved by co-transfection of three separate plasmids into mammalian cells that carry: an evolved orthogonal amino acyl-tRNA synthetase specific for the UAA, an orthogonal suppressor tRNA that recognizes an amber stop codon (UAG), and the gene encoding the GPCR containing a TAG mutation at a desired position. Cells grown in the presence of the UAA express a full-length GPCR with a site-specifically introduced azF. These cells can then be (a) incubated with ligand and activated with UV-light to identify crosslinks between a GPCR and its ligand, (b) used to site-specifically epitope-tag a GPCR in culture with a peptide-conjugated chemical label, or (c) solubilized to purify GPCRs, which are then bioorthogonally modified with a fluorescent label under conditions where the receptor is immobilized on an antibody-conjugated sepharose resin.
Click here to view larger image
Figure 2. Targeted photocrosslinking of a GPCR to a tritiated ligand. Displayed here are the representative results from a crosslinking experiment between azF-CCR5 variants and a tritium-labeled small molecule ligand, maraviroc. Top panel: The Western immunoblot generated after a typical experiment, showing the expression of the tested azF-CCR5 variants. The colored segments highlight where the PVDF membrane was cut for scintillation counting and correspond to the detected scintillation counts plotted in the bar graph in the bottom panel. The red segment represents the expected molecular mass (around 40 kDa) of the covalent crosslinked complex between the receptor and ligand. Therefore, we would expect to detect tritium in those segments only when azF is introduced at a position in the receptor that is in close proximity to the ligand-binding pocket and can form a covalent crosslink with the ligand. Reprinted (adapted) with permission from Grunbeck, A., et al. Genetically encoded photo-cross-linkers map the binding site of an allosteric drug on a G protein-coupled receptor. ACS Chemical Biology 7, 967-972 (2012). Copyright (2012) American Chemical Society.
Click here to view larger image
Figure 3. ELISA detection of targeted epitope-tagging of a GPCR. Top panel: Cell surface expression of HEK 293T cells expressing azF-CCR5 variants and wt and mock transfected controls probed with anti-CCR5 mAb 2D7 using the whole cell-based ELISA (red bars). Bottom panel: Cells from the same transfection were treated with 0.05 mM FLAG-triarylphosphine for 1 hr in live cells and then quantified by subsequent ELISA using anti-FLAG M2 mAb to determine the extent of peptide-epitope tagging at each position (blue bars). Error bars represent the standard error of the mean for two or more replicate data sets. Reprinted (adapted) with permission from Naganathan, S., et al. Site-specific epitope tagging of GPCRs by bioorthogonal modification of a genetically encoded unnatural amino acid. Biochemistry DOI: 10.1021/bi301292h (2013). Copyright (2013) American Chemical Society.
Click here to view larger image
Figure 4. Fluorescent labeling of rhodopsin azF incorporated at various positions. Wt and azF-rhodopsin mutants immobilized on an immunoaffinity matrix labeled with fluorescein-phosphine. Purified samples were separated by 4-12% SDS-PAGE and the in-gel fluorescence image shown was taken with a confocal 488-nm laser fluorescence scanner using an emission filter set optimized for fluorescein detection. Figure adapted from Huber, T., et al. Bioorthogonal labeling of functional G protein-coupled receptors at genetically encoded azido groups. submitted (2013).
Click here to view larger image
We describe here a robust methodology for site-specific incorporation of a reactive probe, azF, into GPCRs and demonstrate three useful applications of this tool to study the structure and dynamics of GPCRs. Our method to site-specifically incorporate UAAs circumvents a fundamental problem with an alternative strategy based on chemically labeling32, 33 or attaching photocrosslinkers34 to a single accessible cysteine mutant. Although, chemistries that target cysteine thiol groups have been used to attach fluorescent probes, such a strategy necessitates the generation of a cysteine-free protein background. This can be limiting since all GPCRs possess more than one cysteine, many of which may be essential to maintain proper receptor function. Therefore, the feasibility to introduce a UAA that is amenable to chemical modification or can be activated by UV light is extremely attractive.
The covalent reaction of a photoactivatable crosslinker with a target molecule is dependent on distance and neighboring side chains. Therefore, these probes can be utilized in identifying interactions that are within a specific distance from the crosslinker, for example, benzophenone-based crosslinkers are suggested to have a distance dependency of 3 Å.35 Previously, we introduced both azF and BzF, a UAA containing a benzophenone reactive group, to identify residues in a GPCR that are within a defined distance from a fluorescein or tritium labeled ligand.18, 19 The general applicability of this technology depends only on the availability of such labeled ligands and the generation of amber mutants of a target GPCR. Additionally, in theory this targeted photocrosslinking technique can be applied to identify interactions between GPCRs and their other binding partners, such as a G protein or β-arrestin. The data gained from these crosslinking studies can be used to create more accurate molecular models of GPCR signaling complexes.
Several bioorthogonal chemical reactions have been described in the literature that allow for the site-specific modification of UAA residues. The UAA, AcF, for example, possesses a keto group that can partake in hydrazone or oxime ligation reactions.36, 37 We previously compared the modification of keto groups and azido groups with biotinylated and fluorescein conjugated reagents.29 Our observations indicated that the azido group of azF is truly more bioorthogonal than its keto counterpart. Hence, we subsequently focused on the use of the Staudinger-Bertozzi ligation to modify two GPCRs with a peptide epitope tag or a fluorescent label. It is noteworthy, however, that the Staudinger-Bertozzi ligation exhibits sub-stoichiometric38 and poor kinetic properties of labeling. Moreover, phosphines are highly susceptible to air oxidation.25 We have observed that the Staudinger-Bertozzi ligation only achieves 30% labeling after incubation for 12 hr at room temperature.29 For these reasons, we are currently evaluating alternative labeling chemistries such as CuAAC and SpAAC to site-specifically label azF-GPCRs.
Albeit the drawbacks of Staudinger-Bertozzi ligation reactions, we demonstrate here the use of a live-cell peptide-epitope tagging strategy to identify positions on a target GPCR that are accessible for future fluorescent labeling. Additionally, this epitope tagging strategy has numerous other useful applications, such as, the introduction of an epitope tag at a single modified residue in contrast to inserting the entire epitope by replacing native sequences. This feature is of specific importance with GPCRs since modifying the length and sequence of highly conserved loop segments could dramatically alter function. Finally, we also illustrated here the proof-of-concept experiments to label GPCRs with fluorophores, making them suitable for single-molecule detection experiments.
In conclusion, azF is a versatile tool for the study of the structure and function of GPCRs. We have shown that azF is site-specifically incorporated into GPCRs expressed in mammalian cells using the amber codon suppression system. This allows for the modification of a GPCR at only one single position. The structure and function of the labeled receptor can then be evaluated either in the native plasma membrane environment or in a detergent solution. We are now poised to investigate GPCR signaling complexes with the flexibility of the azF probe.
The authors have nothing to disclose.
We thank the generous support of several foundations and philanthropic donors (see SakmarLab.org).
Name of Reagent/Material | Company | Catalog Number | Comments |
Plasmid pSVB. Yam | (Ye et al., 2008) | ||
Plasmid pcDNA.RS for azF | (Ye et al., 2009) | ||
Plasmids pMT4.Rho and pcDNA 3.1.CCR5 | Optionally contain the amber stop codon (TAG) at a desired position | ||
HEK 293T cells | Adherent cells | ||
Dulbecco's Modified Eagle's Medium (DMEM) | Gibco | 10566 | |
Phosphate Buffered Saline | Gibco | 14200 | |
Fetal Bovine Serum | Gemini Bio-products | 100-106 | |
Lipofectamine Plus | Invitrogen | Lipofectamine: 18324-012 Plus: 11514-015 |
|
p-azido-L-phenylalanine | Chem-Impex International | 6162 | |
Table 1. Site-specific genetic incorporation of unnatural amino acids into GPCRs materials | |||
[header] | |||
Maxima ML-3500S UV-A lamp | Spectronics Corporation | azF is activated by 365-nm light | |
Hank's Buffered Salt Solution (HBSS) | Gibco | 14065 | |
HEPES | Irvine Scientific | 9319 | |
Bovine serum albumin | Roche | 3117405001 | |
Tritium-labeled ligand | From collaborator (Grunbeck et al., 2012) | ||
1% (w/v) n-dodecyl-β-D-maltoside | Anatrace | D310LA | |
1D4-sepharose resin | 1D4 mAb immobilized on CNBr-activated sepharose 2B resin | ||
1% (w/v) sodium dodecyl sulfate (SDS) | Fisher Scientific | BP166 | |
NuPAGE Novex 4 – 12% SDS gels | Invitrogen | NP0322BOX | SDS-PAGE performed on NuPAGE apparatus |
Trans-Blot SD apparatus | Biorad | Apparatus for semi-dry transfer | |
Immobilon polyvinylidene difluoride (PVDF) membrane | Millipore | IPVH00010 | |
Non-fat powered milk | Fisher Scientific | NC9934262 | |
Tween-20 | Aldrich | 274348 | |
Ecoscint A | National Diagnostics | LS-273 | Scintillation fluid |
Scintillation vials | Fisher Scientific | 333726 | |
LKB Wallac 1209 Rackbeta Liquid Scintillation Counter | Perkin Elmer | Beta-Scintillation counter | |
Anti-rhodopsin 1D4 mAb | National Cell Culture Center | custom | |
Horseradish peroxidase (HRP)-conjugated anti-mouse IgG | KPL, Inc. | 474-1806 | |
SuperSignal West Pico Chemiluminescent Substrate | Thermo Scientific | 34080 | |
Hyblot CL AR film | Denville | E3018 | |
Table 2. Targeted photocrosslinking materials | |||
[header] | |||
0.25% Trypsin | Invitrogen | 15050065 | |
FLAG-triarylphosphine | Sigma | GPHOS1 | |
M2 FLAG mAb | Sigma | F1804 | |
anti-CCR5 2D7 mAb | BD Biosciences | 555990 | |
Poly-D-lysine | Sigma | P6407 | |
96-well plate | Costar | 3601 | Clear bottom, high binding EIA/RIA |
Phosphate Buffered Saline (Calcium, Magnesium) | Gibco | 14040 | |
16% Paraformaldehyde | EMS | 28908 | |
HRP-conjugated | KPL, Inc. | 474-1516 | |
anti-rabbit IgG | KPL, Inc. | 474-1516 | |
Amplex Red | Invitrogen | A12222 | |
Hydrogen peroxide | DE Healthcare Products | 97-93399 | |
CytoFluor II fluorescence multi-well plate reader | Perseptive Biosystems | ||
Table 3. Targeted peptide-epitope tagging and cell surface ELISA materials | |||
[header] | |||
Fluorescein-phosphine | (Huber et al., submitted 2012) | ||
Nonapeptide (C9 peptide) | AnaSpec | 62190 | Peptide mimicking the 1D4-epitope NH2-TETSQVAPA-COOH |
1% (w/v) n-dodecyl-β-D-maltoside | Anatrace | D310LA | |
1D4-sepharose resin | 1D4 mAb immobilized on CNBr-activated sepharose 2B resin | ||
NuPAGE Novex 4 – 12% SDS gels | Invitrogen | NP0322BOX | SDS-PAGE performed on NuPAGE apparatus |
Confocal Typhoon 9400 fluorescence scanner | GE Healthcare | discontinued | Scanner with 488-nm wavelength laser |
Table 4. Fluorescent labeling materials |