This article describes specific methods to obtain biochemical quantities of detergent-solubilized TRPV1 for spectroscopic analysis. The combined protocols provide biochemical and biophysical tools that can be adapted to facilitate structural and functional studies for mammalian ion channels in a membrane-controlled environment.
Polymodal ion channels transduce multiple stimuli of different natures into allosteric changes; these dynamic conformations are challenging to determine and remain largely unknown. With recent advances in single-particle cryo-electron microscopy (cryo-EM) shedding light on the structural features of agonist binding sites and the activation mechanism of several ion channels, the stage is set for an in-depth dynamic analysis of their gating mechanisms using spectroscopic approaches. Spectroscopic techniques such as electron paramagnetic resonance (EPR) and double electron-electron resonance (DEER) have been mainly restricted to the study of prokaryotic ion channels that can be purified in large quantities. The requirement for large amounts of functional and stable membrane proteins has hampered the study of mammalian ion channels using these approaches. EPR and DEER offer many advantages, including determination of the structure and dynamic changes of mobile protein regions, albeit at low resolution, that might be difficult to obtain by X-ray crystallography or cryo-EM, and monitoring reversible gating transition (i.e., closed, open, sensitized, and desensitized). Here, we provide protocols for obtaining milligrams of functional detergent-solubilized transient receptor potential cation channel subfamily V member 1 (TRPV1) that can be labeled for EPR and DEER spectroscopy.
With recent advances in single-particle cryo-electron microscopy (cryo-EM), mammalian ion channel structures have been obtained at an extraordinary rate. Particularly, structural studies of polymodal ion channels, such as the transient receptor potential vanilloid 1 (TRPV1), have provided further understanding of its activation mechanisms1,2,3,4,5. However, dynamic information about ion channels embedded in a membrane environment is required to understand their polymodal gating and drug-binding mechanisms.
Electron paramagnetic resonance (EPR) and double electron-electron resonance (DEER) spectroscopies have provided some of the most definitive mechanistic models for ion channels6,7,8,9,10,11,12,13. These approaches have been mainly restricted to the examination of prokaryotic and archeal ion channels that yield a large amount of detergent-purified proteins when overexpressed in bacteria. With the development of eukaryotic membrane proteins production in insect and mammalian cells for functional and structural characterization14,15,16, it is now possible to obtain biochemical amounts of detergent-purified proteins for spectroscopic studies.
The EPR and DEER signals arise from a paramagneticspin label (SL) (i.e., methanethiosulfonate) attached to a single-cysteine residue in the protein. The spin-labels report three types of structural information: motion, accessibilities, and distances. This information allows determining whether residues are buried within the protein or are exposed to the membrane or aqueous environment in the apo and ligand-bound states13,17,18,19. In the context of a high-resolution structure (when available), the EPR and DEER data provide a collection of constraints for deriving dynamic models in their native environment while monitoring reversible gating transition (i.e., closed, open, sensitized, and desensitized). Moreover, flexible regions that might be difficult to determine by X-ray crystallography or cryo-EM could be obtained by using these environmental data sets to assign secondary structures as well as location within the protein20. Cryo-EM structures obtained in lipid nanodiscs provided valuable information about the gating of ion channels3,21,22,23,24,25; however, spectroscopic approaches could provide dynamic information from conformational states (e.g., thermal changes) that might be difficult to determine using cryo-EM.
Many difficulties must be overcome to implement EPR and DEER, including lack of protein function when removing all cysteine residues (especially abundant in mammalian channels), low protein yield, protein instability during purification and after spin labeling, and protein aggregation in detergent or liposomes. Here, we have designed protocols to overcome these critical barriers and have obtained DEER and EPR spectra information for a mammalian sensory receptor. The purpose here is to describe methodologies for the expression, purification, labeling, and reconstitution of a functional minimal cysteine-less rat TRPV1 (eTRPV1) construct for spectroscopic analyses. This methodology is appropriate for those membrane proteins that keep their function despite the removal of cysteine residues or that contain cysteine forming disulfide-bonds. This collection of protocols could be adapted for the spectroscopic analysis of other mammalian ion channels.
1. TRPV1 Mutagenesis
Note: A minimal TRPV1 construct for spectroscopic analysis26 was built from the full-length cysteine-less channel TRPV127 using the polymerase chain reaction (PCR) method (Figure 1). This cysteine-less minimal TRPV1 construct (referred to as eTRPV1 hereafter) consists of residues 110 – 603 and 627 – 764. eTRPV1 was cloned in pMO (a pcDNA3.1-based vector) for functional analysis and in a recombinant donor vector28 containing 8x histidine-maltose-binding protein (MBP)-tobacco etch virus (TEV) for expression and purification (Figure 2A). eTRPV1 single–cysteine mutants were generated using site-directed mutagenesis. Vector and channel sequences are specified in the Supplementary File 1: Supplementary Methods.
2. Functional Analysis of eTRPV1 and Single-cysteine Mutants
3. Generating the Recombinant Bacmid and Baculovirus for Protein Expression
4. eTRPV1 and Single-cysteine Mutant Purification
5. eTRPV1 Single-cysteine Mutant Site-Directed Spin Labeling
6. eTRPV1 Spin-labeled Single Cysteine Mutant Reconstitution
7. DEER and EPR Spectroscopies
Functional Characterization of the Minimal Cysteine-less TRPV1 Construct (eTRPV1) and Single-cysteine Mutants
The first step toward spectroscopic studies is to engineer and characterize cysteine-less protein constructs (Figure 2A) that are functional and yield biochemical amounts of proteins. eTRPV1 is functional as determined by Ca2+ imaging and TEVC (Figure 2B-C). Moreover, eTRPV1 provides sufficient amounts of detergent-purified protein for EPR and DEER experiments (0.5 – 1 mg per liter of Sf9 cells)26. Introducing single cysteines in particular protein regions might affect their function. Figure 2B shows some examples of single-cysteine mutants (E651C and A702C) on eTRPV1 that behave like wild type (WT), since their fluorescence intensity increases when TRPV1 agonist (e.g., capsaicin) is added to eTRPV1-containing HEK293 cells, as shown by Ca2+ imaging26. On the other hand, A680C is the typical example of a non-functional cysteine mutant, as the fluorescence is indistinguishable from the background. The A680C mutant was excluded for further analysis. After Ca2+ imaging experiments, single-cysteine mutants' function is tested using TEVC or patch clamp to evaluate their biophysical properties. Figure 2C shows that mutants recapitulate the outward rectification characteristic of TRPV1 when challenged with pH 5, as determined by TEVC26. Functional analysis of single-cysteine mutants is the first checkpoint before undertaking expression and purification protocols.
Biochemical Characterization of the eTRPV1 and Single-cysteine Mutants
The purification protocol described above yields detergent-solubilized protein that can be labeled via cysteine covalent modification (e.g., fluorophores, spin-labeled [SL] methyl-methanethiolsulfonate) along the TRPV1 sequence for spectroscopic analysis. Minimal TRPV1 channel-containing cysteine residues migrate as stable and monodisperse species (~ 13.6 mL), as determined by size-exclusion chromatography (Figure 3). eTRPV1 and single-cysteine spin-labeled mutants (E651C-SL and A702C-SL) recapitulate the elution profile and stability featured by the minimal TRPV1 construct (Figure 3)26. The elution profile could vary among different mutants, as single cysteines might affect protein stability. In some instances, mutants form aggregates; and a fraction of the sample could elute at the void volume of the column (8 – 9.5 mL), reducing the amount of protein that migrates as a tetramer (Figure 3, bottom red arrow). In this case, cell culture volumes can be scaled up to compensate for the protein lost in the aggregated fraction, or one can exclude this mutant from further analysis and proceed to test the neighboring residue. Other examples include a broadening of the peak that corresponds to the tetramer. Main peaks wider than 3 mL are not recommended for spectroscopic analysis, as they contain multi-disperse species. Biochemical characterization of spin-labeled single-cysteine mutants is the second checkpoint before undertaking reconstitution and spectroscopic analysis.
Functional Characterization of Reconstituted Spin-labeled eTRPV1 Single-cysteine Mutants
Because the EPR and DEER signals rely on the attachment of a paramagnetic spin-label to a cysteine residue, it is important to verify whether the location of the spin-label alters protein function. There are several ways to test for function after reconstituting the spin-labeled protein, including planar lipid bilayer experiments, patch clamp, and TEVC. TEVC allows the evaluation of a large number of spin-labeled channels while recording macroscopic currents. To assess the functionality of the spin-labeled single-cysteine mutants, channels are reconstituted in pre-formed asolectin liposomes. During this step, it is important to exclude the 10% glycerol that is present throughout purification, since glycerol decreases the efficiency of protein reconstitution into liposomes. Figure 4 shows a representative result of A702C-SL TRPV1 reconstituted in asolectin liposomes and microinjected into Xenopus oocytes. As expected, pH 5 elicits robust outwardly rectifying currents37 that are blocked by co-application of the TRPV1 antagonist capsazepine (CPZ, blue trace)26. Mutants that do not retain functionality after spin labeling and reconstitution should be excluded from further analysis. Functional characterization of reconstituted spin-labeled single-cysteine mutants is the third checkpoint before undertaking spectroscopic analysis.
Structural Dynamics of Spin-labeled eTRPV1 Mutants Monitored by CW-EPR and DEER
Nitroxide spin labels are sensitive to the surrounding environment and the flexibility of the protein backbone to which the label is attached17. Hence, CW-EPR allows determination of the dynamic regimen of a given spin-labeled cysteine; namely, positions exposed to the aqueous media or the membrane are more dynamic than those restricted to protein-protein interactions17. Figure 5A shows positions Glu651, Ile679, and Ala702 in TRPV1 chosen to monitor mobility parameters. To calculate the mobility parameter of the spin label probe, one calculates the inverse of the central line width of the first derivative absorption spectra (Figure 5B, black line ΔHo−1). For instance, E651C-SL (Figure 5B) displays a spectral line shape and mobility value (0.24) commonly found on dynamic positions at the membrane interface26. On the other hand, I679C-SL and A702C-SL exhibit broadening of the spectra (see dotted lines) and a decrease in the mobility values (0.16 and 0.18, respectively; Figure 5B)26 that are consistent with their constrained surrounding environment (protein-protein), according to the cryo-EM structure1.
Figure 5C shows the spectra of spin-labeled TRPV1 mutants after reconstitution in asolectin liposomes. As expected, the dynamic features of the spectra for E651C-SL and A702C-SL did not change after reconstitution, as their shape and mobility values are the same in solution as in a membrane environment26. On the other hand, I679C-SL illustrates a noisy spectrum; consequently, the lower (blue arrow) and higher (red arrow) field components become less obvious. Noisy spectra are not usually desired, as they tend to underestimate the mobility of the spin-labeled position. This spectrum type could be the product of inefficient protein reconstitution rather than under-labeling, since the I679C-SL signal in solution is robust.
DEER data are a sum of oscillating, sinusoidal signal decays containing information about the distance distribution of interacting spin-labels38,39. The period and complexity of the decay directly reflects the underlying distance distribution. The distance distribution is comprised of the number of structural components present in the sample and their disorder. Hence, the time-domain signal derives from the dipolar coupling between non-fixed, distant spin labels in solution which usually cause an exponential background decay, andrigidly coupled spins within the same protein19,40. Specifically, DEER allows the determination of intra-protein long-range distances (20 – 70 Å) between spin-labeled residues19. Figure 6 shows the signal decay and the corresponding distance distribution of E651C-SL. The distribution of Glu651 shows three peaks corresponding to 24, 36, and 58 Å26. The two shorter distances are consistent with the closed TRPV1 structure (23 and 32 Å for the Cβ-Cβ distances, respectively)1, whereas the third 58 Å peak might correspond to protein aggregation.
Figure 1. Experimental outline. Diagram of the experimental outline required to express, purify, and reconstitute eTRPV1 for EPR and DEER. Please click here to view a larger version of this figure.
Figure 2. Functional characterization of eTRPV1 and single-cysteine mutants. (A) Schematic representation of the TRPV1 constructs used for spectroscopic analysis (eTRPV1: cysteine-less TRPV1 110-603/627-764). (B) HEK293 cells expressing WT TRPV1, eTRPV1, and mutants (loaded with Ca2+-sensitive Fluo-4-AM) were analyzed for capsaicin (10 µM)-evoked responses using fluorescence Ca2+ imaging. Color bar indicates relative changes in fluorescence intensity, with blue and red denoting the lowest and highest cytoplasmic Ca2+, respectively. White bar represents 100 µm. (C) Left, one subunit (S5, pore helix, S6 and TRP domain) of TRPV1 tetramer structure highlighting single-cysteine residues (yellow spheres) introduced along the channel sequence. Right, current-voltage relationships determined by TEVC recordings from Xenopus oocytes expressing WT TRPV1, eTRPV1, and single-cysteine mutants challenged with pH 5. Background currents (bkgrd). Modified from the original figure26. Please click here to view a larger version of this figure.
Figure 3. Biochemical characterization of eTRPV1 and single-cysteine mutants.
Size-exclusion chromatography profile of DDM-solubilized eTRPV1 and single-cysteine spin-labeled mutants after expression and purification from Sf9 cells. Inset: SDS-PAGE gel showing the monomeric eTRPV1-MBP fusion protein (modified from the original figure26). Please click here to view a larger version of this figure.
Figure 4. Functional characterization of a spin-labeled eTRPV1 mutant.
Current-voltage relationships determined by TEVC from Xenopus oocytes microinjected with proteoliposomes containing spin-labeled A702C challenged with pH 5 (red) and blocked by capsazepine (CPZ, blue: 40 µM). Background currents (bkgrd). Modified from the original figure26. Please click here to view a larger version of this figure.
Figure 5. Mobilities of spin-labeled eTRPV1 mutants determined by CW-EPR.
(A) Two subunits (S5, pore helix, S6 and TRP domain) of TRPV1 tetramer structure highlighting the amino acid residues (yellow spheres) probed through site-directed spin-labeling spectroscopies. (B) First derivative of CW EPR spectra of spin-labeled cysteine mutants in DDM-solution. (C) First derivative of CW EPR spectra of spin-labeled cysteine mutants reconstituted in asolectin liposomes. Spectra were obtained at pH 7.4 (closed state). ΔHo−1 denotes the magnitude of the mobility parameter. The black dotted line highlights the broadening of the spectra. Blue and red arrows denote the low and high field components of the spectra, respectively. EPR spectra were normalized to the total number of spin labels. Modified from the original figure26. Please click here to view a larger version of this figure.
Figure 6. Distance distribution of an eTRPV1 mutant in detergent.
DEER echo (A) and distance distribution (B) of spin-labeled mutant E651C; a sum of Gaussians was fitted to the DEER data. P(r) denotes distance distribution of the spin pairs41. Spectra were obtained at pH 7.4 (closed state). Modified from the original figure26. Please click here to view a larger version of this figure.
Current technologies for expression and purification of mammalian membrane proteins have made it possible to obtain sufficient amounts of protein for spectroscopic studies14,15,16,42. Here, we have adapted these technologies to express, purify, reconstitute, and perform spectroscopic analyses in TRPV1.
Among the critical steps in the protocol, below are the ones that we have troubleshot for TRPV1 and that might be adjusted for other proteins. Modify the DNA sequence to generate a template that yields 0.5 – 1 mg/mL of detergent-purified protein; templates that yield less than 0.5 mg/mL of protein are challenging, since growing more than 6 liters of insect cells cultures would be required per mutant. Protein must be concentrated, no less than 2 mg/mL, without TCEP to increase spin-labeling efficiency. Labeling at low protein concentrations and/or in the presence of TCEP traces will generate noisy EPR spectra. Although the proteins are kept at 4 °C during purification, it is essential to perform spin labeling at RT to improve efficiency. During purification and labeling, glycerol improves protein stability; however, it must be completely removed before reconstitution, as it decreases the amount of protein in liposomes and generates noisy EPR spectra. Decreasing the detergent concentration below the CMC is a common practice during reconstitution; however, TRPV1 tends to precipitate before incorporating into the liposomes. To solve this problem, TRPV1 was incubated overnight with pre-formed liposomes exactly at the CMC to avoid protein precipitation and increase the quantity of channels in the proteoliposome preparation. TRPV1 is fully functional in asolectin liposomes28; hence, it was the preferred lipid mixture used in this protocol. However, it is essential to determine the lipid composition in which a particular protein is functional (e.g., cholesterol) before proceeding with spectroscopic analyses.
Spectroscopic approaches such as EPR and DEER present several limitations, including: changes in protein template sequence that improve biochemical stability might have an impact in function26; introduction of non-native single cysteines, as well as the spin label, might not be well tolerated in certain regions of the protein; obtaining large amounts of labeled proteins; and limited conformational changes when determining distances with DEER in detergent micelles. However, the latter limitation could be overcome by collecting the spectra, at Q-band frequency, of TRPV1 mutants reconstituted in nanodiscs19,43. Nonetheless, the advantage of these approaches mainly comes from the pattern of global dynamics, accessibilities, and distances more than from the absolute value of a given position.
Temperature sensitivity is one of the most fascinating and less understood gating mechanisms; hence, using spectroscopic approaches, it would be possible to determine how membrane proteins translate thermal energy into protein motion in a membrane environment. Importantly, it would be challenging to determine structural changes during thermal gating using X-ray crystallography or cryo-EM, as these techniques are performed at freezing temperatures. Future experiments will be directed towards determining TRPV1 conformational changes compatible with thermal-dependent gating using EPR, DEER, or fluorescence. EPR spectroscopy has provided detailed mechanistic models for prokaryotic ion channels, so we expect that by using the protocols described above, we will gain insight into the mechanisms of activation and drug-binding of mammalian ion channels using spectroscopic approaches.
The authors have nothing to disclose.
We are very grateful to Dr. H. Mchaourab for providing access to the EPR and DEER spectrometers and Dr. T. Rosenbaum for providing the full-length cysteine-less TRPV1 plasmid.
QuikChange Lightning Site-Directed Mutagenesis Kit | Agilent Technologies | 210519-5 | |
2-Propanol (Isopropanol) | Fisher Scientific | A416 | |
Albumin Bovine Serum (BSA) | GoldBio.com | A-420-10 | |
Amylose resin | NEB | E8021L | |
Aprotinin | GoldBio.com | A-655-25 | |
Asolectin from Soybean | Sigma | 11145 | |
Bac-to-Bac Baculovirus Expression System | Invitrogen Life Technologies | 10359016 | |
Biobeads SM-2 Adsorbents | Bio-Rad | 152-3920 | |
Borosilicate glass pipettes (3.5'') (oocyte inyection) | Drummond Scientific | 3-000-203 G/X | |
Borosilicate glass pipettes (oocyte recordings) | Sutter Instrument | B150-110-10HP | |
CaCl2 2H2O | Fisher Scientific | C79 | |
Carbenicillin (Disodium) | GoldBio.com | C-103-5 | |
Cellfectin Reagent | Invitrogen Life Technologies | 10362-010 | |
cellSens | Olympus | ||
Chloroform | Fisher Scientific | C606SK | |
Collagenase Type 1 | Worthington-Biochem | LS004196 | |
Critiseal | VWR | 18000-299 | |
D-(+)-Glucose | Sigma | G8270 | |
D-(+)-Maltose Monohydrate | Fisher Scientific | BP684 | |
DDM (n-Docecyl-B-D-Maltopyranoside) | Anatrace | D310S | |
High glucose medium (Dulbecco’s Modified Eagle’s Medium) | Sigma | D0572 | |
Disposable PD-10 Desalting Columns | GE Healthcare | 45-000-148 | |
EGTA | Fisher Scientific | O2783 | |
Fetal Bovine Serum | Invitrogen Life Technologies | 10082-147 | |
Fluo-4 AM | Life Technologies | F-14201 | |
GenCatch Plus Plasmid DNA Mini-Prep Kit | Epoch Life Science, Inc | 2160250 | |
GenCatch PCR Cleanup Kit | Epoch Life Science, Inc | 2360050 | |
Gentamicin Sulfate | Lonza | 17-518Z | |
Glass capillary (25 µl) | VWR | 53432-761 | |
Glass Flask 2800 mL | Pyrex USA | 4423-2XL | |
Glycerol | Fisher BioReagents | BP229 | |
HEK293S GnTl- | ATCC | CRL-3022 | |
HEPES | Sigma | H4034 | |
IPTG (isopropyl-thio-B-galactoside) | GoldBio.com | I2481C25 | |
Kanamycin Sulfate | Fisher Scientific | BP906-5 | |
KCl | Fisher Chemical | P217 | |
LB Broth, Miller | Fisher bioReagents | BP1426 | |
Leupeptin Hemisulfate | GoldBio.com | L-010-5 | |
Lipofectamine 2000 | Invitrogen Life Technologies | 11668-019 | |
MgCl2 6H2O | Fisher Scientific | BP214 | |
MgSO4 7H2O | Fisher Scientific | BP213 | |
mMESSAGE mMACHINE T7 Kit | Ambion | AM1344 | |
MOPS | Fisher bioReagents | BP2936 | |
MTSL (1-Oxyl-2,2,5,5-tetramethylpyrrolidin-3-yl) Methyl Methanethiosulfonate | Toronto Research Chemicals, Inc | O873900 | |
NaCl | Fisher Chemical | S271 | |
Opti-MEM | Life Technologies | 31985-062 | |
Pepstatin A | GoldBio.com | P-020-5 | |
Pluronic Acid F-127 (20%) | PromoKine | CA707-59004 | |
PMSF | GoldBio.com | P4170 | |
Poly-L-lysine Solution | Sigma-Aldrich | P4707 | |
Rneasy Mini Kit | Qiagen | 74104 | |
Sealed capillary | VitroCom | special order | |
SF-900 II SFM (insect cell medium) | Gibco, Life Technologies | 10902-088 | |
Sf9 Cells (SFM Adapted) | Invitrogen Life Technologies | 11496-015 | |
Soybean Polar Lipid Extract | Avanti Polar Lipids, Inc | 541602C | |
Sucrose | Fisher Scientific | S25590 | |
Superose 6 Increase 10/300 GL | GE Healthcare | 29091596 | |
TCEP HCl | GoldBio.com | TCEP1 | |
Tetracyclin Hydrochloride | Fisher Scientific | BP912-100 | |
Tris Base | Fisher BioReagents | BP152 | |
Tryptone | Difco | 0123-01 | |
X-gal | GoldBio.com | X4281C | |
Xenopus oocytes | Nasco | LM00935M | |
XL1 – Blue Competent Cells | Agilent Technologies, Inc | 200249 | |
Yeast Extract | Difco | 0127-01-7 | |
Econo-Pack chromatography column | Bio-Rad | 7321010 | |
Mini-PROTEAN TGX Stain-Free Precast Gels | Bio-Rad | 17000436 | |
pFastBac1 Expression Vector | Invitrogen Life Technologies | 10360-014 | |
DH10Bac Competent Cells | Invitrogen Life Technologies | 10361-012 | |
Critiseal capillary tube sealant | Leica Microsystems | 02-676-20 | |
ABI Model 3130XL Genetic Analyzers | Applied Biosystems | 4359571 | |
Transfer pipete | Fishebrand | 13-711-9AM | |
Nanoject II | Drummond Scientific | 3-000-204 |