Protein–Lipid Interfaces Can Drive the Functions of Membrane-Embedded Protein–Protein Complexes

The roles of surrounding membrane lipids in the functions of transmembrane and peripheral membrane proteins are largely unknown. Herein, we utilize the recently reported structures of the TRPV1 ion channel protein bound to its potent protein agonist, the double-knot toxin (DkTx), as a model system to investigate the roles of toxin–lipid interfaces in TRPV1 activation by characterizing a series of DkTx variants electrophysiologically. Together with membrane partitioning experiments, these studies reveal that toxin–lipid interfaces play an overwhelmingly dominant role in channel activation as compared to lipid-devoid toxin–channel interfaces. Additionally, we find that whereas the membrane interfaces formed by one of the knots of the toxin endow it with its low channel-dissociation rate, those formed by other knot contribute primarily to its potency. These studies establish that protein–lipid interfaces play nuanced yet profound roles in the function of protein–protein complexes within membranes.

No. 1 Detailed protocol for generating the interaction map of the DkTx-lipids-TRPV1 complex 3 2  Table S1:

Detailed protocol for generating the interaction map of the DkTx-lipids-TRPV1 complex
The distances between the atoms of the amino acids of DkTx and those of lipid molecules/amino acids of TRPV1 were calculated by separately analyzing two pdb files of the DkTx-TRPV1 complex-one obtained from a recently published cryo-EM structure 1 and the other from a previously published docking model 2 -by using the UCSF-CHIMERA software. In this software, the command "split" was entered in the command line to split the PDB structure into individual peptide chains. Subsequently, each of the four monomers of the TRPV1 channel was assigned a unique model number (0, 1, 2 or 3), the two DkTx chains were assigned model numbers 4 and 5 respectively, and each of the resiniferatoxin and lipid molecules (which were present in the cryo-EM structure but not in the docking model) were assigned unique model numbers between 6 and 25. The command "distance #a:x@ #b:y@" was used to obtain the distance values between the amino acid residue number "x" of the DkTx chain assigned the chain model number "a" and the amino acid residue number "y" of the TRPV1 chain assigned the chain model number "b". This command yielded the distances between every possible pairs of atoms (one each belonging to the residues "x" and "y" respectively) under analysis. For every DkTx residue, this analysis was performed with respect to every amino acid residue of each of the four monomers of the channel's pore region formed by its S5, S6 and pore helices. For determining the distance between the amino acid residues of a particular DkTx chain and a lipid molecule assigned a model number "z", the command, "distance #a:x@ #z@" was used. This analysis was performed for every DkTx residue with respect to all the lipid molecules in the cryo-EM structure. The distance output obtained from these analyses was saved as a ".txt" file and this data was filtered manually to select for distances within the 4.4 Å cut-off distance to yield a list of TRPV1 residues/lipid molecules proximal to each DkTx residue. The resulting interaction map is depicted in Table S1 (next page).

Table S1. Map of interactions between DkTx and TRPV1/lipids.
Lipids are numbered according to the scheme employed in Figure 6 of the main text of the manuscript. The paper that reported the docking model 2 also reported MD simulations to predict lipid-interacting residues of DkTx-this information is also provided in the  h. Subsequently, dithiothreitol (66 mM) was added and the resulting mixture was incubated at 45 o C for 1 h for performing protein denaturation. This solution was then titrated with conc. HCl to pH 4 and centrifuged to obtain the denatured proteins in the supernatant. Subsequently, the protein solution was dialyzed against water containing 0.1% TFA and then filtered through 0.45 µm syringe filters and lyophilized to yield the linear protein as a white powder.

Refolding of linear DkTx variants:
The lyophilized linear protein was dissolved in 50% acetonitrile containing 0.1% TFA at a concentration of 2 mg/mL. This protein solution was then added to a redox refolding buffer containing Tris-HCl (0.4 M), EDTA (1 mM), Triton X-100 (0.5%), GSH (2.5 mM), GSSG (0.25 mM) at pH 8.0 and incubated at 4 o C for 2 d. Poorly folding DkTx variants, W53A, W53L and G52A required prolonged incubation of 14 d in the refolding buffer to give good yields. The progression of refolding was assessed by subjecting the refolding cocktail to analytical RP-HPLC on a C18 column (300 Å pore size) with a linear acetonitrile-water (0.1% TFA) gradient (5-65% acetonitrile over 30 min). Once refolded optimally, the refolding cocktail was titrated to pH 4 with conc. HCl to quench the refolding reaction. The refolding cocktail was then concentrated in a 3500 MWCO dialysis membrane (Thermo Fischer Inc.) against 400 g/L solution of PEG20. The refolded DkTx variants were then purified from the concentrated refolding cocktail by RP-HPLC on a C18 column (300 Å pore size) with a linear acetonitrilewater (0.1% TFA) gradient (5-65% acetonitrile over 30 min). Purified fractions were quantified by spectrophotometric analysis from the absorbance at 280 nm and then aliquoted in 1 nmol aliquots or 4 nmol aliquots, vacuum-dried and stored at -80 o C until use for electrophysiological activity assays and partitioning experiments, respectively. Figure S1 depicts representative protein production data and Figure S2 depicts the purity evaluation HPLC runs for all variants.
Wild-type and all DkTx variants had an additional glycine residue at their N-termini and three additional residues (HYR) in the linker region. The sequence of DkTx referred to as "wild-type" in the manuscript, therefore, is as follows: GDCAKEGEVCSWGKKCCDLDNFYCPMEFIPHCKKYKPYVPVTTHYRNCAKEGEVCGW GSKCCHGLDCPLAFIPYCEKYR Figure S1. Representative DkTx purification data.
DkTx variants were produced as described above (pg. 19-20). The band for the overexpressed DkTx-KSI fusion protein (depicted by an asterisk in the gel picture below) was expected at 24 kDa. Gel lanes have been labeled as followed: I: induced; UI: uninduced; M: molecular weight markers; P: DkTx protein purified by HPLC after refolding (expected at 9 kDa). The red arrow below denotes the peak for the correctly folded toxin in the HPLC chromatogram. Figure S2. HPLC traces for purity evaluation of all our DkTx variants. The absorbance has been measured at 280 nm.

Electrophysiology and data analysis
Activity assays on DkTx variants by Two-Electrode Voltage Clamp recordings: Stage 5 and 6 oocytes from adult female Xenopus laevis (the African clawed frog) were surgically removed from anesthetized frogs in accordance with a protocol approved by the Institutional Animal Ethics Committee (IAEC), IISER Pune. Oocytes were digested with collagenase type 2 enzyme (2 mg/mL) in calcium-free OR2 buffer containing NaCl (82.5 mM), KCl (2.5 mM), MgCl2 (1 mM) and HEPES (5 mM) at pH 7.6 for 1 h at 18 o C on an automated rocker and then de-folliculated with cut pipettes. Rat TRPV1 gene cloned in pGEM-HE vector was used to make TRPV1 cRNA by in vitro transcription using HiScribe T7 ARCA mRNA Kit Dose-response plots: The normalized ratio of toxin-induced current to that of the capsaicin current was plotted against the concentration of the toxin and the Hill equation (below) was fit to the data using Origin 9.0: where Amax is the normalized maximum I(toxin)/I(5 µM Capsaicin) ratio, k is EC50 and n is the Hill coefficient.  Tryptophan fluorescence spectroscopy: DkTx variants were dissolved in 2 mL of a buffer containing HEPES (10 mM) and EDTA (1 mM) at pH 7.0 to a concentration of 2 µM and the resulting solution was excited with 280 nm wavelength light by using a Horiba FluoroMax 4 spectrofluorometer. Emission spectra were recorded between 300 to 450 nm both in the presence and absence of LUVs. Scattering of light from lipid vesicles was corrected following earlier reported methods 4,5 and the data was subjected to a fitting analysis by employing the following equation: where F is the fluorescence intensity at 320 nm for a given lipid concentration, F0 is the fluorescence intensity at 320 nm in absence of lipids, F0 max is fluorescence intensity at 320 nm at a concentration of lipids where partitioning saturates, Our toxin-depletion experiments involved incubation of 100 defolliculated stage 5 or 6 oocytes with 4 nmol of DkTx or its variants in 400 µL (final volume) of 20 mM HEPES buffer at pH 7.4 containing NaCl (50 mM), KCl (50 mM), MgCl2 (1 mM) and BaCl2 (0.3 mM) for 1 h at room temperature. In control experiments, the same concentration and volume of toxin solutions not containing oocytes were used. After incubation, 200 μL of the supernatant from each well was removed, spiked with 2-nitrophenol (2.5 μL of a 2.5 mg/mL solution in MilliQ water) that served as an internal standard, and 100 μL of this solution was subjected to HPLC by employing the same protocol that was used for the purification of DkTx and its variants (described on pg. 19 above). Fractional depletion was calculated as follows: Fraction depletion = (Rcontrol-Roocytes)/Rcontrol where Rcontrol is the averaged ratio of the area under the peak for the toxin and the area under the peak for 2-nitrophenol in the control experiments, and Roocytes is the ratio of the same parameters obtained from the oocyte samples (n = 3-5). Depletion assays on TRPV1-expressing oocytes ( Figure S6b) were performed by using an identical protocol on oocytes injected with the TRPV1 cRNA 4 d earlier.  Table S2. Error bars correspond to SEM (n = 3 or 4).

Structural characterization of DkTx and its variants by Circular Dichroism (CD):
CD spectroscopic measurements were performed on a Jasco J-815 spectropolarimeter on protein solutions (50 μM) in sodium phosphate buffer (10 mM at pH 7.0) in a 0.1 cm cuvette at 25°C. Spectra obtained were reported as molar ellipticity [deg.cm 2 /dmol]. Each representative spectra is an average of the 5 scans smoothened utilizing the Savitzky-Golay algorithm provided in the spectropolarimeter. Figure S8. CD spectra of DkTx and its variants.