Detergent-free Lipodisq Nanoparticles Facilitate High-Resolution Mass Spectrometry of Folded Integral Membrane Proteins

Integral membrane proteins pose considerable challenges to mass spectrometry (MS) owing to the complexity and diversity of the components in their native environment. Here, we use native MS to study the post-translational maturation of bacteriorhodopsin (bR) and archaerhodopsin-3 (AR3), using both octyl-glucoside detergent micelles and lipid-based nanoparticles. A lower collision energy was required to obtain well-resolved spectra for proteins in styrene-maleic acid copolymer (SMA) Lipodisqs than in membrane scaffold protein (MSP) Nanodiscs. By comparing spectra of membrane proteins prepared using the different membrane mimetics, we found that SMA may favor selective solubilization of correctly folded proteins and better preserve native lipid interactions than other membrane mimetics. Our spectra reveal the correlation between the post-translation modifications (PTMs), lipid-interactions, and protein-folding states of bR, providing insights into the process of maturation of the photoreceptor proteins.

which was pelleted by centrifugation (2000 ×g, 5 min, RT). The pellet was resolubilised in water and was subjected to multiple washing and centrifugation steps (2000 ×g, 5 min). Once washed, ddH 2 O was added to the solution and dialysis was performed overnight in order to remove the excess of salt and to adjust the pH. The resulting pale yellow, transparent SMA solution was lyophilized. The solid white powder was redissolved in ddH 2 O or buffer to a final concentration of 125 mg/ml and the solution adjusted to pH 8. Solubilizaton of bR in OG and reconstitution of bR into MSP Nanodiscs. Proteins were delipidated following published protocols 1,2 . Briefly, the protein-containing membranes were pelleted by centrifugation (70,000 ×g, 30 min, 4 °C), and resuspended in 6 mL of 25 mM NaH 2 PO 4 , pH 6.9. Detergent (n-octyl-β-D-glucoside (OG), Glycon) was added to the sample (2 mL of 10% w/v OG in ddH 2 O pH 7) and the preparation was sonicated for 1 min in a bath sonicator at room temperature. The sample was incubated at 22 °C overnight without stirring. The solution was adjusted to pH 5.5, and centrifuged (100,000 ×g, 45 min, 15 °C) to remove aggregates and any non-solubilized material. The supernatant was applied to a preparative gel column (Hi-Load 16/600, Superdex 200 pg) pre-equilibrated in 25 mM NaH 2 PO 4 , 1.2% w/v OG at pH 5.5. Fractions were collected from chromatogram peaks and further analyzed. Colored fractions were further concentrated using a VivaSpin concentrator to 9 mg/mL and stored at 4 °C.
OG-solublized bR was reconstituted following an established protocol 3 . Briefly, purifed bR, MSP, and DMPC solubilized in cholate, were mixed at 1:2:80 ratio and incubated at room temperature for 2 h. After addition of amberlite XAD-2 hydrophobic beads, the mixture was incubated overnight. MSP Nanodiscs containg bR were purified on a Superose 6 increase 10/300 column, equilibrated in 0.2 M ammonium acetate, pH 7.
Incorporation into Lipodisq nanoparticles. bR from purple membrane and AR3 from claret membranes, were incorporated separately into Lipodisqs as described previously 4 with some changes outlined below. Briefly, 4 mg of protein was pelleted (45,000 ×g, 30 min, 4°C) and resuspended in 1 mL of 50 mM sodium phosphate and 200 mM NaCl buffer containing 20 mg of 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) small unilamellar vesicles (SUVs) produced by the freeze/thaw and extrusion technique in a final protein/lipid molar ratio of 1:172. The solution was sonicated (30 min) in a bath sonicator with gentle heating (30-35 °C). Styrene-maleic acid copolymer (25 mg, added as a solution in ddH 2 O) was added to the vesicle membrane sample, and the solution incubated (42 °C, 1 h). Samples were centrifuged (100,000 ×g, 30 min, 15 °C) in order to remove non-solubilized membrane and the supernatant collected.
Reconstitution of detergent-bR into Lipodisqs using proteoliposomes. The protocol was adapted from 5,6 . Briefly, DMPC was purchased from Avanti Polar Lipids. The appropriate amount of lipid was dissolved in chloroform:methanol (50:50) to 10-25 mg/mL in a round-bottom flask. The lipid solution was dried down to a lipid film either under nitrogen. The film was dried further overnight in a desiccator under vacuum and stored at -20 °C until needed or used immediately.
The lipid film was suspended in liposome buffer (50 mM sodium phosphate and 200 mM NaCl, pH 8) to give a final concentration of 5 mg/mL, and sonicated (3×1 min) using a bath sonicator, followed by 10 freeze-thaw cycles using liquid nitrogen and a 37 °C water bath. The lipid solution was then extruded through a 100 nm polycarbonate filter using a miniextruder, for at least 11 passes to obtain a homogeneous distribution of liposomes of 100 nm in diameter. OG was added to the lipid suspension at a final concentration of 25.6 mM and the lipids were gently stirred for 0.5-1h (RT). The detergentliposome mix was then added to bR (or bleached bR) at the desired lipid-to-protein ratio (172:1), and incubated for 30 minutes at room temperature.
Detergent removal was acheived in two steps. First, wet Bio-Beads (BioRad) (80 mg/mL) were added directly to the bRlipid-detergent suspension. The mixture was incubated with light stirring at room temperature for 3 hr. A second portion of slightly wet beads was then added and mixed overnight with a small shaker to remove residual detergent. Bio-Beads were then removed, and proteoliposomes were harvested by centrifugation (100,000g, 1 h, 15 °C). The supernatant was removed and the pelleted proteoliposomes resuspended in buffer. To form Lipodisqs, SMA in a ratio of 1:1.5 (lipid-to-polymer w/w ratio) was added, and the solution incubated (37°C, 30 mins). Samples were centrifuged (100,000 ×g, 30 min, 15°C) in order to remove non-solubilized membrane and the supernatant collected. SMA was removed either by size-exclusion or by Vivaspin concentrator (Merck, MWCO 100 kDa).
Denaturing MS. Denaturing MS was performed following a well-established protocol with minor modifications 7 . Briefly, intact membrane protein was analysed on an UltiMate 3000 UHPLC system (Thermo Fisher Scientific) coupled to an Exactive Plus EMR mass spectrometer (Thermo Fisher Scientific). Membrane protein samples were loaded and separated on a ProSwift RP-4H capillary monolithic column (250 mm, Thermo Fisher Scientific) with a binary buffer system. Buffer A was 0.1% formic acid in 100% H 2 O and buffer B was 0.1% formic acid in 5% H 2 O, 45% acetonitrile, and 50% isopropanol. A linear gradient from 0% to 100% Buffer B was applied to elute membrane proteins. The mass spectrometer was operated in full MS scan mode with a mass range of 300 to 2,000 m/z. Spectra were deconvoluted by UniDec to obtained the mass distribution 8 .
Proteomics. The amount of protein in the solution before and after Lipodisq reconstitution was quantified by gel-based proteomics. Briefly, equal amount of samples (15 µL) were separated by SDS-PAGE. The target gel bands were sliced into small pieces and digested with trypsin as described previously 9 . The digested membrane proteins were analysed on a Dionex 3000 UHPLC coupled to a LTQ-Orbitrap XL mass spectrometer (Thermo Fisher Scientific) as described previously 10 . The peptides were separated on a 75 µm × 15 cm Pepmap C18 column (Thermo Fisher Scientific). LTQ-Orbitrap XL was operated in data-dependent acquisition mode with one full MS scan followed by 5 MS/MS scans with collision-induced dissociation. Proteomics data was analysed using Maxquant (version 1.6.7.0) 11 . Protein quantification is based on LFQ intensities. The means and errors were determined from three biological replicates.
Size-exclusion chromatography Lipodisq samples were applied to a Superdex Increase 200 10/300 GL column connected to an Akta Pure (GE Healthcare). The column was equilibrated with buffer for 2 column volumes and sample was applied at a flow rate of 0.5 mL/min. Protein was detected using a UV-detector at a wavelength of 280 nm and fractions were collected in 15 mL falcon tubes (Greiner). Purple fractions were collected and concentrated using a Vivaspin concentrator (Merck) and colourless fractions were discarded.
UV-Vis spectroscopy Spectra were recorded on a Jasco V-630 instrument at different wavelengths in the visible range, using a 1 cm quartz cuvette and were blanked against buffer.
Dynamic Light Scattering (DLS) Particle size was measured using a Malvern Zetasizer Nano S instrument in a disposable cuvette at 633 nm. The data was processed using Malvern Zetasizer software.
Circular dichroism (CD) CD spectra were acquired with a Jasco J-815 Spectropolarimeter using a 1 mm cuvette. A wavelength range (185-240 nm) was selected to monitor secondary structure of the proteins at ambient temperature. Concentration range used for measurements was adjusted between 0.01-0.1 mg/mL and the protein was diluted in low salt buffer in order to perform experiments. The data was analysed, and baseline corrected using the Spectra Manager software from Jasco. α, β and random coil content was calculated from the Dichroweb website 12 using the CDSSTR analysis programme 13 and the reference set SMP180 for membrane protein 14 . Figure S1. (A) Dynamic light scattering data for bR in proteoliposomes formed from the purple membrane by osmotic shock (purple) and in Lipodisq (black) with a measured size of 569 ± 245 nm and 9.5 ± 2.5 nm. (B) Absorption spectra of bR in purple membrane with A max at 565 nm (purple), in OG detergent (green) and in Lipodisq with A max at 555 nm (black), show a hypsochromic shift, characteristic of the change in oligomeric state from purple membrane (trimer) to Lipodisq (monomer) 4 . (C) UV CD spectra showing that the protein is folded both in the purple membrane (purple) and in Lipodisq nanoparticles (black).