Phosphatidylserine-dependent structure of synaptogyrin remodels the synaptic vesicle membrane

Molecular cloning, protein expression, and purification

The construct of synaptogyrin 1b (Uniprot: O43759-2) was cloned into the pET16b vector with an N-terminal histidine tag. The synaptogyrin mutants were engineered using site-directed mutagenesis (Thermo Fisher Scientific) and confirmed by DNA sequencing (Supplementary Table 3). Expression was performed in the Escherichia coli strain Lemo21 (DE3) with 250 µM l-rhamnose, following previously established procedures27. Cells were grown in Luria Bertini or minimal medium containing 15NH4Cl and/or [13C6]glucose as the sole nitrogen or carbon source in H2O or D2O, respectively. Single or double amino acid selective labeling was obtained using [15N]Ala, [15N]Leu, [15N]Cys, [15N]Lys, [15N]Phe and [15N]Tyr as precursors, and the other amino acids were unlabeled in the minimal medium to suppress scrambling. Protein methyl-labeled on Ileδ1, LeuproS and ValproS was expressed using 13CH3-methyl specifically labeled precursors (NMR-Bio) in fully deuterated medium. The precursor of 2-(D3) methyl, 2,4-(13C2) acetolactate was added 1 h prior to the deuterated M9 medium. The 2-ketobutyric acid 4-(13C), 3,3 (D2) sodium salt was added 20 min before induction by addition of IPTG.

For synaptogyrin production, the temperature was reduced to 20 °C at an optical density at 600 nm (OD600) of 0.8–1.0 after seeding the preculture. Cells were induced with 0.2 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) after 1 h and were further incubated at 20 °C for 16 h before collection. The cell pellet was resuspended in 50 ml lysis buffer (20 mM NaPi, pH 7.5, 300 mM NaCl, 1 mM DTT, 1 mg ml–1 lysozyme, 0.01 mg ml–1 DNaseI, 1 mM phenylmethylsulfonyl fluoride, and one Complete Protease Inhibitor Cocktail-EDTA free (Roche Applied Science)) and homogenized by stirring for 50 min at 4 °C. The suspension was lysed with an Emulsiflex (Avastin) and centrifuged (10,000g, 10 min, 4 °C) for removal of cell debris. The membrane fraction was collected from the supernatant by ultracentrifugation (Beckman Ti45 rotor) at 100,000g at 4 °C for 50 min. The membranes were resuspended and solubilized by solubilization buffer (20 mM NaPi, pH 7.5, 300 mM NaCl, 1 mM DTT, and one Complete Protease Inhibitor Cocktail-EDTA free, 1 % dodecyl-ß-d-maltoside (DDM)) at 4 °C for 1 h. The solubilized membrane fraction was loaded onto a histidine-trap open column, and the imidazole concentration was increased stepwise to 400 mM. When required, the DDM detergent was exchanged for different detergents, such as 0.005% 2,2-didecylpropane-1,3-bis-β-d-maltopyranoside (NG310) or 0.1% n-undecyl β-maltoside (UDM) in this step. The protein was eluted with 400 mM imidazole, followed by removal of the histidine-tag with TEV protease during dialysis (20 mM sodium phosphate, pH 7.5, 300 mM NaCl, 1 mM DTT, 5% glycerol, 0.03% DDM (0.1% UDM or 0.005% NG310)). After cleavage, the TEV and the histidine tag were removed by loading onto a second histidine-trap column. The pure protein was further purified by gel filtration (HiLoad Superdex 200 column) in 20 mM HEPES, pH 7.5, 150 mM NaCl, 1 mM DTT, 5% glycerol, 0.03% DDM (or 0.1% UDM or 0.005% NG310). The purified protein was verified by SDS polyacrylamide gel electrophoresis. The protein sample buffer was dialyzed against NMR buffer (20 mM sodium phosphate, pH 8.0, 150 mM NaCl, 1 mM TCEP, 0.1 % UDM) for 18 h at 4 °C.

Circular dichroism

Far-ultraviolet CD measurements were performed at 40 °C using 10 μM synaptogyrin in 20 mM sodium phosphate, pH 7.5, 1 mM DTT, 0.1% UDM on a Chirascan spectrometer (Applied Photophysics). The path length of the cuvette was 0.2 mm. CD spectra were recorded from 190 to 280 nm with an integration time of 0.5 s. The measurement was repeated three times. The final spectrum was obtained by baseline subtraction using the buffer-only measurement.

Liposome morphology

DLS measurements were performed at 25 °C using a DynaPro NanoStar instrument with a detector positioned at a 90° angle concerning the incident light. The monochromatic laser with a wavelength of 662 nm (auto-attenuation turned on) irradiated the liposome sample in disposable COC cuvettes. Liposome samples (0.4 mM lipid) containing increasing synaptogyrin concentrations were incubated for 18 h at 25 °C in 20 mM HEPES, pH 7.5, and 1 mM TCEP prior to the measurement. DLS data were acquired with 20 cycles and 5 s of acquisition time for three repeated measurements and were analyzed with DYNAMICS v7.10.0.23. Error bars represent the s.d. of three measurements. Microsoft Excel v16.43 was used to fit data, as well as for statistical analysis.

In parallel, samples from the same preparation were adsorbed onto 400-mesh carbon-coated copper grids, and the buffer was removed using filter paper. Samples were then stained by adding 1% uranyl acetate solution, which was dried with a filter paper. The grids were imaged using a FEI Tecnai Spirit electron microscope with a TVIPS F416 4K camera.

NMR spectroscopy

NMR spectra were recorded at 40 °C on Bruker 600, 700, 800, 900, and 950 MHz Bruker NMR spectrometers using cryogenic probes. 1H spin-spin echo experiments with echo times from 0.1 to 8.1 ms were recorded on 0.1–0.3 mM synaptogyrin in 20 mM sodium phosphate, pH 8.0, 150 mM NaCl, 1 mM TCEP, 0.1% UDM (or 0.005% NG310 or 0.03% DDM) at 40 °C. Relaxation times (T2) were calculated using peak intensities of NMR signals between 9.5 ppm and 8.5 ppm.

The backbone resonances were assigned by TROSY-based 3D triple resonance through bond scalar correlation HNCO, HN(CA)CO, HNCA, HN(CO)CA, and HN(CA)CB experiments47,48,49 using 2H-13C-15N-labeled synaptogyrin (0.8 mM synaptogyrin in 20 mM sodium phosphate, pH 8.0, 150 mM NaCl, 1 mM TCEP, 0.1 % UDM). Backbone resonance assignment was further supported by a 3D 15N-edited NOESY-HSQC (NOE mixing time of 150 ms) experiment recorded on 0.8 mM 2H-15N-labeled synaptogyrin in 20 mM sodium phosphate, pH 8.0, 150 mM NaCl, 1 mM TCEP, 0.1% UDM, as well as 1H-15N TROSY-HSQC experiments recorded on [15N]Ala/[15N]Leu-, [15N]Leu/[15N]Tyr]-, [15N]Cys/[15N]Phe-, [15N]Thr-, and [15N]Lys-labeled synaptogyrin (0.8–1.0 mM). Side chain assignments were determined using a 3D (H)CCH-TOCSY (mixing time 10 ms), two 3D 13C-edited NOESY-HSQC (80 ms and 300 ms mixing time; 1.0 mM 13C-15N-labeled synaptogyrin in 20 mM sodium phosphate, pH 8.0, 150 mM NaCl, 1 mM TCEP, 0.03 % d25-DDM, 100% D2O), and two 3D 15N-edited NOESY-HSQC (80 ms and 300 ms mixing time; 0.9 mM 15N-labeled synaptogyrin in 20 mM sodium phosphate, pH 8.0, 150 mM NaCl, 1 mM TCEP, 0.03 % d25-DDM, 10 % D2O) experiments. For [U-2H-15N; Iδ1 /LVproS 13CH3]-labeled synaptogyrin, 2D 1H-13C HSQC and 13C-edited NOESY-HSQC (200 ms mixing time) experiments were recorded and used for the assignment of the methyl groups of the Ile, Leu and Val residues. Long-range NOE restraints between HN backbone resonances and side chain methyl protons were extracted from 3D 15N-resolved NOESY-HSQC experiments. recorded for 0.4 mM [U-2H-15N; Iδ1 /LVproS 13CH3]-labeled synaptogyrin in 20 mM sodium phosphate, pH 8.0, 150 mM NaCl, 1 mM TCEP, 0.03% d25-DDM, 10% D2O.

Residual 1DNH dipolar couplings were obtained by taking the difference in the J splitting values measured in oriented (7% acrylamide gel alignment medium) and isotropic sample conditions using interleaved 2D 1H-15N TROSY-HSQC/2D 1H-15N HSQC spectra recorded on 0.8 mM 2H-15N-labeled synaptogyrin in 20 mM sodium phosphate, pH 8.0, 300 mM NaCl, 1 mM TCEP, 0.1% UDM. NMR spectra were processed with Topspin 3.6.1 (Bruker) and analyzed using the software CcpNmr (Analysis 2.4.2)50. Residue-specific secondary structure scores and random coil indices (RCI S2) were determined from the assigned NMR chemical shifts of synaptogyrin using TALOS+ (ref. 51). The RDC values were analyzed using the PALES program52.

For the titration experiments of wild-type and mutant synaptogyrin with DMPS, 0.2 mM 2H-15N-labeled synaptogyrin was solubilized in isotropic bicelles without or with 20 % DMPS ((DMPC:DMPS = 8:2)/DHPC, q = 0.3) corresponding to a protein:DMPS ratio of 1:40. The buffer was 20 mM sodium phosphate, pH 8.0, 150 mM NaCl, 1 mM TCEP. 1H and 15N normalized weighted average chemical shift perturbations were calculated using ΔδNH = [((ΔH)2 + (ΔN)2 / 5) / 2]1/2, where \(\Delta H\) and \(\Delta N\) correspond to the 1H and 15N chemical shift differences between two states (for example, without and with DMPS).

To measure T2, we used a 90°–180° pulse sequence. Fitting the intensity decay curve to I0 (–t / T2) with I0 being the starting intensity gives the characteristic time constant T2. The correlation time τc was subsequently calculated according to:

$$\frac=\frac\left[3_+\frac_}_^_}+\frac_}_^_}\right]$$

$$K=\frac_^}^}\frac^^}^}.$$

The relaxation times are approximately associated with the correlation time by the equations according to Bloembergen–Purcell–Pound (BPP) theory, where ω0 is the rotational frequency of the signal, µ0 is the magnetic permeability of free space (4π × 10−7 H m−1), ħ is the reduced Planck constant 1.054571726 × 10−34 J s–1, γ is the gyromagnetic ratio of the nucleus (for 1H, it is 2.67513 × 108 rad s−1 T−1) and r is the distance between magnetically active spin-½ nuclei. The molecular weight of synaptogyrin in different micelle conditions was then calculated according to:

η is viscosity, v is the partial specific volume, h is the degree of hydration, R is the ideal gas constant and T is temperature.

Paramagnetic relaxation enhancement

For the spin labeling on Cys124 of synaptogyrin with MTSL (S-(1-oxyl-2,2,5,5-tetramethyl-2,5-dihydro-1H-pyrrol-3-yl)methyl methanesulfonothioate), the three other native cysteines in synaptogyrin were mutated to either serine or alanine (C59A C68A C82S). To remove residual reducing agent, synaptogyrin was passed through a desalting column (HiPrep 26/10 Desalting; GE Healthcare). The protein was collected into a solution containing tenfold excess of MTSL and incubated at 4 °C for 18 h. Using the desalting column, we removed the excess MTSL, then exchanged the protein into NMR buffer containing 20 mM NaPi, pH 8.0, 150 mM NaCl, and 0.1% UDM. To determine the paramagnetic relaxation enhanced-broadening of the NMR signals, a 1H-15N TROSY-HSQC experiment was recorded for 1.0 mM 15N-labeled mutant synaptogyrin (C59A C68A C82S) in 20 mM sodium phosphate, pH 8.0, 150 mM NaCl, 1 mM TCEP, 0.1 % UDM.

To probe the hydrophobic burial and solvent exposure of synaptogyrin residues, we used the detergent-soluble 16-doxylstearic acid (16-DSA) and the water soluble [5,8-bis-(carboxymethyl)-11-[2-(methylamino)-2-oxoethyl]-3-oxo-2,5,8,11-tetraazatridecan-13-oato(3-)]-gadolinium (gadodiamide), respectively. The paramagnetic reagents were stepwise added to 0.8 mM U-[2H-15N]-synaptogyrin solubilized in UDM micelles. 1H-15N TROSY-HSQC spectra were then recorded at 40 °C using Bruker 900 and 800 MHz spectrometers.

Structure calculation

The recorded backbone chemical shifts (HN, N, C, Cα and Cβ) of residues 16–176 of synaptogyrin were used in RASREC CS-Rosetta to select 200 fragments of 3 and 9 residues in length. The N- and C-terminal tails of synaptogyrin were not included into the structure determination process, because they were predicted to be disordered by TALOS+ (ref. 51) on the basis of the experimental NMR chemical shifts. In addition, the loop regions formed by residues 47–65, 91–98 and 132–143, which had a TALOS+-predicted order parameter of less than 0.7, were removed from the Rosetta score computation, but were part of the structure calculation. NOE restraints from 3D 15N-edited NOESY-HSQC experiments were manually assigned. In the later stages of the structure calculation, NOE restraints determined automatically by CYANA were also included. The NOE-based distances (d0) were restrained with a ROSETTA flat-bottom potential with fixed lower bound (lb) = 1.5, upper bound (ub) = d0 + 0.15 Å, and inverse curvature (c) = 0.3 Å. In addition, 68 long-range PRE-based distances were restrained between the Cβ atom of the MTSL-labeled residue (Cys124) and the corresponding backbone amide protons with error bounds of ±4 Å in order to account for the size and flexibility of the Cys-MTSL53. NOE, PRE and RDC data maintained weights of 5 and 0.1 for scoring in the low-resolution sampling stage and the all-atom sampling stage, respectively. A weight equal to 5 was used for the NOE, PRE and RDC data for the centroid and all-atom structures to select a pool of iterated structures. The standard Rosetta all-atom energy function was used with the experimental weight set mpframework_smooth54. The energy function was adjusted for TM proteins to model the hydrophobic TM region55. TM regions were determined from the amino acid sequence using the OCTOPUS method56. Five thousand structures were calculated, and the ten structures with the lowest Rosetta all-atom energy score were selected for further analysis. Structure figures were prepared using the PyMOL Molecular Graphics System (Version 1.8.2.1).

Liposome assays

Liposomes were formed from L-α-phosphatidylcholine (PC), L-α-phosphatidylethanolamine (PE), L-α-phosphatidylserine (PS), L-α-phosphatidylinositol (PI), and cholesterol (Avanti Polar Lipids) in selected wt/wt ratios. After addition of each phospholipid to the glass tube, the phospholipid mixture was evaporated off chloroform-methanol (1:1) under a low stream of nitrogen. The organic solvent was lyophilized to obtain a thoroughly dried lipid film. The dry lipid film was hydrated with 1 mL of liposome buffer (20 mM NaPi HEPES, pH 7.5, 1 mM TCEP). To produce unilamellar liposomes, the rehydrated lipid was sonicated three times for 2 min using an ultrasonic water bath sonicator. Next, the lipid film was extruded through a filter with a 200-nm pore size and thus liposomes of more uniform size were obtained. In order to reconstitute DDM-solubilized synaptogyrin into liposomes, DDM was then removed using Bio-Beads SM-2 Resin (Bio-rad). After DDM removal, the synaptogyrin sample was directly added to the liposomes. The sample was incubated for 18 h at 25 °C in 20 mM HEPES, pH 7.5, 1 mM TCEP. For NMR experiments, the sample was further concentrated using 30-kDa cut-off vivaspin (Satorious). NMR spectroscopy was used to confirm removal of DDM.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

留言 (0)

沒有登入
gif