Sample Preparation of the Human TRPA1 Ion Channel for Cryo-EM Studies

Abstract

The transient receptor potential ankyrin 1 (TRPA1) ion channel is a member of the TRP channel family that is involved in sensing noxious stimuli that elicit pain and inflammation. Because of its critical physiological role and therapeutic importance, great efforts have been made to understand the structure and mechanism of TRPA1. Several human TRPA1 structures have been reported using single particle cryo-electron microscopy (cryo-EM) over the last six years. Here, we present a protocol for the heterologous expression, large-scale purification, and nanodisc reconstitution of the human TRPA1 channel for cryo-EM and biochemical studies.

Introduction

Transient receptor potential (TRP) ion channels comprise a superfamily of cation channels which impart the myriad sensations of pain and itch, warm or cold, as well as the detection of pungent compounds. As a member of the diverse TRP channel family, mammalian TRPA1 primarily functions as a sensor for structurally diverse, noxious environmental stimuli such as allyl isothiocyanate (AITC), cannabinoids, and endogenous molecules involved in inflammatory stress such as 4-hydroxynoneal (Bandell et al. 2004, Hinman et al. 2006, Macpherson et al. 2007, Trevisani et al. 2007, Taylor-Clark et al. 2009). There is also accumulating evidence that TRPA1 is sensitive to either hot or cold, displaying significant tunability in temperature sensation (Laursen et al. 2015). Thus, TRPA1 has been established as a promising drug target for suppressing pain and inflammation (Han et al. 2018, Koivisto et al. 2018, Xie and Hu 2018).

Despite the great interest and research effort, only two crystal structures of TRP channels have been reported to date (Saotome et al. 2016, Zubcevic, Le, et al. 2018). This is largely due to their inherent flexibility, which makes TRP channels challenging targets for crystallization. Since the dawn of the cryo-EM “resolution revolution” in 2012–2013, we have witnessed an exponential growth in reported TRP channel structures (Madej and Ziegler 2018, Yuan 2019, Cao 2020, Zubcevic 2020, Zubcevic and Lee 2019), owing mainly to the development of direct electron detectors and advances in data processing strategies (Cheng et al. 2015). Currently, hundreds of TRP channel structures (both cryo-EM and X-ray) are available in the Protein Data Bank (Berman et al. 2000). For TRPA1, multiple structures reconstituted into various systems (detergents, amphipol, and nanodiscs) have been reported since 2015 (Paulsen et al. 2015, Liu et al. 2021, Suo et al. 2020, Zhao et al. 2020). It is noteworthy that only one study, so far, has been reported for human TRPA1 reconstituted in lipid nanodiscs (Suo et al. 2020).

Compared to X-ray crystallography, which requires samples with very high homogeneity and stability, requirements for sample quality in cryo-EM studies are less demanding, allowing for structural determination of proteins with high instability and flexibility. In our experience, however, in order to obtain a sub-3Å reconstruction the biochemical behavior of the sample must be optimized. Although sub-optimal sample quality can often be offset by simply collecting more data, the current scarcity and price in accessing high-end electron microscopes often discourages prolonged data collection. Thus, preparing sample to the highest purity and homogeneity should still be considered the top priority in single particle 3D cryo-EM reconstruction. Batch-to-batch consistency is necessary in order to optimize cryo-EM sample freezing conditions so that structures in complex with different ligands, binding partners, etc. may be obtained. (Carragher et al. 2019). It is critical to maintain consistency throughout different preparations to allow proper interpretation among these data.

As with many membrane proteins, TRP channels form both specific and non-specific interactions with membrane lipids. The specific interactions between various TRP channels and lipids and lipid-like molecules have been well-studied in recent publications (Gao et al. 2016, Zubcevic, Herzik, et al. 2018, Zubcevic et al. 2019, Yin et al. 2019, Yin and Lee 2020). During detergent solubilization and subsequent size-exclusion chromatography, most of the protein-bound lipids are stripped away. Reconstituting membrane proteins into lipid nanodiscs provides a membrane-like environment for the protein and thus, enables the study of protein-lipid interactions in detail. Moreover, certain toxins or ligands require a lipid membrane to bind to ion channels (Bae et al. 2016, Henriques et al. 2016), thereby making TRP-nanodisc systems ideal for these types of studies.

Here, we present a protocol for human TRPA1 protein expression, purification and nanodisc reconstitution for cryo-EM structural studies (Suo et al. 2020).

Step-by Step Method Details

Generating baculovirus for TRPA1 expression

Timing: 16–17 days

1. The codon optimized full-length Homo sapiens TRPA1 gene was synthesized (BioBasic) then cloned into the pEG-BacMam vector (Goehring et al. 2014), in-frame with a tandem C-terminal FLAG-tag and 10×His-tag. Baculovirus is generated in Sf9 cells according to the manufacturer’s protocol (Bac-to-Bac, Invitrogen).

2. Baculovirus encoding the TRPA1 gene is amplified by passaging to P3. Freshly harvested virus-containing supernatant is used to infect HEK293S GnTI⁻ cells within two weeks.

Cell culture and TRPA1 expression in HEK293S GnTI⁻ cells

Timing: 3 days

1. HEK293S GnTI⁻ cells are purchased from ATCC. A frozen aliquot is thawed and passaged according to the manufacturer’s protocol. The cells are used for protein expression only between the 4^th and 22^nd passages. Declining protein yield was observed after the 22^nd passage, necessitating replacement with freshly thawed cells.

2. Cells are maintained in FreeStyle293 Media (Life Technologies) supplemented with 2% (v/v) Fetal Bovine Serum (FBS, Gibco) and 0.5% (v/v) penicillin and streptomycin (Gibco), in an orbital shaker incubator set at 130 rpm, 37°C and an atmosphere of 8% CO₂ and 80–100% humidity. Typically, 1.5–2 L cultures are inoculated from a master stock and grown to a density of 3×10⁶ mL⁻¹. The culture is then infected with 8% (v/v) P3 baculovirus together with supplementation of 3μM ruthenium red [diluted from a 6 mM stock] to reduce channel toxicity. Cultures are incubated for 20–22 hours at 37°C in the presence of 8% CO₂ before 10 mM sodium butyrate [diluted from a 1 M stock] is added to boost protein expression. The cultures are then transferred to an incubator set at 30°C, 8% CO₂ and 130 rpm shaking. Cells are then harvested after 40–44 hours.

TRPA1 Protein purification

Timing: 5–8 hours

1. Prepare size-exclusion chromatography (SEC) buffers. Two different SEC buffers are used in this protocol. SEC buffer (detergent) contains 20 mM Tris-HCl pH 8, 150 mM NaCl, 1 mM inositol hexaphosphate (InsP₆), 3 mM tris(2-carboxyethyl)phosphine (TCEP), and 0.07% (w/v) digitonin. Digitonin powder is directly added to buffer, and SEC buffer (detergent) is separately prepared with 5% digitonin stock mentioned in step 2. The buffer is filtered twice to remove insoluble digitonin through a 0.45 μm filter membrane. SEC buffer (detergent-free) contains 20 mM Tris-HCl pH 8, 150 mM NaCl, 1 mM InsP₆, and 3 mM TCEP. This buffer is filtered once before use.

   Note: Both buffers must be prepared fresh to avoid oxidization of TCEP.

2. Prepare 5% (w/v) digitonin stock for protein solubilization. Weigh out the desired amount of digitonin powder (typically 700–750 mg for each TRPA1 sample preparation) and gently resuspend in TBS (20 mM Tris pH8, 150 mM NaCl) in a 50-ml Falcon tube. Place the tube in a 95 – 100 °C water bath and incubate for 10 minutes, with occasional inversion every 3–4 minutes. All the solid should dissolve within 2–3 minutes. Leave the solution to slowly cool to room temperature before use.

   Note: It is important to slowly cool down the 5% digitonin stock after boiling since a rapid cooling can cause precipitation of the digitonin. Typically, the cooling process takes 30 minutes.

3. Prior to harvesting the 1.5–2 L culture, the cells are inspected for morphology and viability under an optical microscope. The cells are then pelleted by centrifugation at 550 × g for 12 min and the cell pellets are weighed.

   Note: Cell pellets should appear pink due to the added ruthenium red, while the clarified media should appear nearly colorless.

4. Prepare lysis buffer (20 mM Tris pH 8, 150 mM NaCl, 12 μg mL⁻¹ leupeptin [from 1 mg mL⁻¹ stock], 12 μg mL⁻¹ pepstatin [from 1 mg mL⁻¹ stock], 12 μg mL⁻¹ aprotinin [from 0.5 mg mL⁻¹ stock], 2 μg ml ⁻¹ DNase I, 1 mM phenylmethylsulphonyl fluoride [PMSF, from 100 mM stock], 1 mM inositol hexaphosphate [InsP₆], 5 mM TCEP-HCl).

5. Resuspend the cell pellets in approximately 2.2 ml of lysis buffer per gram of cell pellet and gently pipet up and down to break down cell clumps. Add the freshly prepared 5% digitonin dropwise to a final concentration of 1%, making up the final solubilization volume to 4 ml per gram of cell pellet. For example, 10 g (~10 ml in volume) of harvested cells would need to be resuspended 22 ml of lysis buffer, followed by addition of 8 ml of 5% digitonin stock, then making the final volume up to 40 ml. Solubilize membranes with gentle agitation or stirring for 1 hour at 4°C.

   Note: It is normal to notice small clumps (<1 mm in size) form within the first 10–15 minutes of solubilization. If larger clumps form, break them up using a serological pipet.

6. Clarify the solubilization mix by centrifuging at 13000 x g for 30 minutes, 4°C. Take a small aliquot of both the supernatant and pellet for SDS-PAGE analysis.

7. Transfer the supernatant to a clean 50 mL Falcon tube and add another 1 mM PMSF, followed by 0.5 ml (resin volume) of anti-FLAG M2 resin per liter of cell culture. Rotate at 4°C for 1 hour to allow protein binding to the FLAG resin.

8. Meanwhile, prepare the gravity-flow columns, FLAG-wash buffer (same composition as SEC buffer (detergent)) and FLAG-elution buffer (same composition as SEC buffer (detergent) and supplemented with 0.1 mg mL⁻¹ FLAG peptide).

9. When binding is complete, spin down the FLAG resin by centrifuging at 700 × g for 10 minutes, 4°C. Pour off most (but not all) of the supernatant and load the rest into a gravity-flow column.

10. Wash resin with 10 column volumes (CV) of FLAG-wash buffer.

11. Elute protein by five repeated additions of 1 CV FLAG-elution buffer; incubating 5 minutes then draining and collecting the flow through with each addition. Concentrate the eluted protein to 500 μl using an Amicon Ultra Centrifugal Filter (Millipore, 100 kDa cutoff) and inject the protein sample onto a Superose 6 10/300 size-exclusion column (GE Life Sciences) that has been pre-equilibrated with SEC buffer (detergent). Analyze the size-exclusion chromatogram and perform SDS-PAGE analysis to determine protein yield and homogeneity. Collect and pool peak fractions containing TRPA1 and concentrate, as before, to 400–500 μl for nanodisc reconstitution.

Preparation of POPC:POPE:POPG (3:1:1) mixture for nanodiscs reconstitution

Timing: 1 hour

1. Preparation of lipids for nanodiscs reconstitution. For TRPA1, we used a 3:1:1 (wt:wt:wt) mixture of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (POPE), 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-(1’-rac-glycerol) (POPG), from Avanti Polar Lipids. Lipids are formulated as 10 mg mL⁻¹ solutions in chloroform.

   Note: The choice of lipids should be optimized and tested for individual proteins for optimal reconstitution. The inclusion of signaling lipids, such as phosphatidylinositol 4,5-bisphosphate, into the lipid mix may be beneficial.

2. The desired quantities of chloroform-solubilized lipids (see below) are transferred to a 10 mL glass tube with screw cap (VWR). The chloroform is evaporated using a gentle argon stream while continuously rotating the tube so that the lipids form a thin, semi-transparent film around the bottom of the tube. After the last traces of chloroform have evaporated, leave the tube under a stream of argon for an additional 5 minutes to ensure complete removal of chloroform.

3. Re-dissolve the lipids in half the initial volume’s-worth of pentane (reagent grade, Sigma-Aldrich) and repeat the above procedure in step 2 to remove any residual organic solvent.

   Note: It is important that the lipids are dried completely as residual organic solvent will interfere with the nanodisc reconstitution.

4. Hydrate the lipid film by adding the desired volume of TBS to the glass tube to make the final lipid concentration 10 mg mL⁻¹. Vortex vigorously for 5 minutes to completely resuspend all of the lipid film, which should result in an opaque, cloudy solution.

5. To form the small unilamellar vesicles required for nanodisc reconstitution, the lipid suspension is sonicated using a bath sonicator. To avoid over-heating, alternately sonicate for 1 min followed by placing on ice for 1 min. After 10–15 rounds of sonicating, the lipid suspension should appear semi-transparent. To further solubilize the lipids, transfer the cleared lipid solution to a 1.5 mL Eppendorf tube and add digitonin to a final concentration of 0.1% from a 5% stock solution; rotate the tube at 4°C for 2–3 hours, or until ready to use.

TRPA1-MSP2N2 nanodisc reconstitution

Timing: 16–20 hours

1. Membrane scaffold protein, MSP2N2, was expressed and purified according to a previously published protocol (Ritchie et al. 2009). Purified MSP2N2 was concentrated to 4 mg mL ⁻¹, frozen as 100 μL aliquots and stored at −80 °C.

2. Measure the protein concentration of the pooled and concentrated SEC fractions by OD₂₈₀ and/or other methods. Determine the amounts of lipid and MSP needed for reconstitution according to a 1:3:200 molar ratio of TRPA1 (tetramer): MSP2N2: lipids. For example, given 400 μL of TRPA1 at 2 mg mL⁻¹, prepare the following:

   TRPA1 (2 mg mL−1) MW 500,000 Da	400 μL
   POPC:POPE:POPG 3:1:1 (10 mg mL−1) MW ~750Da	23.8 μL
   MSP2N2 (4 mg mL−1) MW 42,400 Da	51.0 μL

3. TRPA1 protein is first equilibrated with the lipids by gently rotating the mixture for 30 min at 4°C, then MSP2N2 is added and the mixture is rotated for an additional 30 min.

4. While TRPA1 protein is incubating with MSP2N2 and lipids in step 4, prepare two equal aliquots of Bio-Beads (Bio-Rad). For 400 μL of TRPA1/MSP2N2/lipids mix made in steps 4, prepare two aliquots of 40 mg Bio-Beads (100 mg mL⁻¹). Weigh out desired amount of Bio-Beads to 1.5 mL Eppendorf tubes. Wash the Bio-Beads with TBS then aspirate all remaining liquid. Keep Bio-Beats at 4°C until ready to use in steps 5 and 6.

5. Initiate detergent removal by transferring the reconstitution mix to one of the freshly washed Bio-bead aliquots, place on a tube inverter at 4°C for 2 hours.

6. Pellet the Bio-beads by spinning the reconstitution mixture at 1000 rpm for 1 minute, then transfer the solution to the second aliquot Bio-beads and rotating at 4°C overnight.

   Note: The minimum time tested for the second Bio-beads incubation is 12 hours. It is recommended to incubate 12–16 hours for a complete reconstitution.

7. Separate Bio-Beads from solution by aspirating all liquid using a 1 ml pipette tip. Take caution not to pipet any Bio-beads into the tip and transfer the solution into a clean 1.5 mL Eppendorf tube. Inject the TRPA1/MSP2N2/lipids reconstitution mix into a Superose 6 10/300 column pre-equilibrated with SEC buffer (detergent-free) for SEC analysis. Analyze the chromatogram and conduct SDS-PAGE on the fractions to assess reconstitution efficiency. If the sample behavior is acceptable, collect peak fractions containing nanodisc-reconstituted TRPA1 and prepare cryo-EM grids.

Summary

Obtaining a homogenous, pure protein sample is the key to not only structural studies, but also for a number of functional studies such as activity assays. Outlined above is a method to purify and reconstitute human TRPA1 into lipid nanodiscs, suitable for cryo-EM structural studies. The reproducibility and versality has allowed us to determine several high-resolution structures of TRPA1 in the apo state as well as with covalent ligands bound. Moreover, because the sample is reconstituted into lipid nanodiscs, structures of nanodisc-reconstituted TRPA1 in complex with different ligands has highlighted the potential roles of lipids in TRPA1 gating and TRPA1-related human disease (Suo et al. 2020). We have successfully applied the above methods to similarly produce highly homogenous and well-behaved TRP channels for cryo-EM studies (Yoo et al. 2018, Zubcevic et al. 2019).

Figure 1.

Flowchart of the complete protocol.

Figure 2.

Example results from TRPA1 sample preparation, nanodisc reconstitution and a representative cryo-EM micrograph. A, SEC trace of a TRPA1 sample reconstituted into lipid nanodiscs. The black bar indicates the range of SEC fractions sampled for SDS-PAGE. B, SDS-PAGE of the reconstitution mix prior to SEC, and SEC fractions. The contaminant band appeared in SDS-PAGE is likely to be a certain type of heat shock protein, as we often observe it in other membrane protein expressed in HEK293 cells. C, Representative cryo-EM micrograph of TRPA1-nanodiscs. Sample is imaged at 81,000 X magnification with −5 μm defocus. Representative TRPA1 particles are circled.

Key Resources Table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Bacterial and Virus Strains
XL1-Blue	Agilent	200249
DH10Bac	Thermo Fisher	10361012
Chemicals, Peptides, and Recombinant Proteins
Freestyle 293 Expression Medium	Gibco	12338–026
Fetal Bovine Serum	Gibco	16000044
Ruthenium Red	Sigma-Aldrich	R2751
Sodium butyrate	Sigma-Aldrich	B5887
ANTI-FLAG M2 Affinity Gel	Sigma-Aldrich	A2220
TCEP-HCl	Thermo Fisher	PG82089
Digitonin	Sigma-Aldrich	D141
InsP6	Sigma-Aldrich	P8810
POPC	Avanti Polar Lipids	850457
POPE	Avanti Polar Lipids	850757
POPG	Avanti Polar Lipids	840457
Bio-Beads SM-2	Bio-Rad	1523920
Critical Commercial Assays
Superose 6 10/300 GL	GE Healthcare	17–5172-01
Experimental Models: Cell Lines
HEK293S GnTI-	ATCC	CRL-3022
SF9	ATCC	CRL-1711
Recombinant DNA
pMSP2N2	Addgene	29520
hTRPA1 (codon optimized)	Bio Basic	N/A