Cell-based studies
All reagents were from Sigma, unless stated otherwise. Antibodies used were as follows: mouse anti-opsin [18], rabbit anti-V5 (custom made), rabbit anti-V5 (Abcam), and mouse anti-tubulin (gift from Keith Gull). Fluorophore-conjugated secondary antibodies for microscopy and Western blotting were purchased from Molecular Probes and LI-COR Biosciences, respectively. The plasmids for the expression of V5-tagged PEX19, wild-type SGTA, and the TPR mutant were previously described [12]. The 3x NNP/AAA (NNP positions at 226-228, 239-241 and 255-257) and ΔQ (Δ275-313) variants of SGTA-V5 were generated by multisite-directed mutagenesis and inverse PCR, respectively, and validated by DNA sequencing. HeLa cells were cultured in DMEM supplemented with 10% FCS and 2 mM l-glutamine at 37 °C and 5% CO2. DNA transfections were performed using Lipofectamine 2000 (Invitrogen) and cells analyzed 20–24 h post-transfection. The inducible HeLa cell line expressing OP91 was generated using the Flp-In T-REx system [12] and maintained in complete DMEM supplemented with 100 μg/ml hygromycin B and 4 μg/ml blasticidin S at alternate passages. At 8 h after DNA transfection, T-REx HeLa cells were treated with DMEM containing 1 μg/ml tetracycline for an additional 12–16 h to induce OP91expression. For Western blot analysis, cells were lysed directly into SDS-PAGE sample buffer. Samples were denatured by 1 h incubation at 37 °C with shaking, and then sonicated 3 × 15 s with the Bioruptor (Diagenode). Proteins were separated by SDS-PAGE and analyzed by infrared immunoblotting. The fluorescent bands were visualized and quantified using Image Studio (LI-COR Biosciences). For immunofluorescence microscopy, cells growing on coverslips were fixed with 3% (v/v) formaldehyde, permeabilized with 0.1% (v/v) Triton X-100, and washed with PBS in incubated with primary and secondary antibodies in PBS at room temperature. Coverslips were mounted in ProLong Diamond (Molecular Probes) and analyzed using an Olympus BX60 upright microscope equipped with a MicroMax cooled, slow-scan CCD camera (Roper Scientific) driven by Metaview software (University Imaging Corporation). Images were processed using Adobe Photoshop CS5. Quantification results are expressed as the mean ± s.e.m. from three independent experiments. The statistical significance of the results was assessed by applying a Student’s t test using Prism 7 (GraphPad). *P < 0.05, ** P < 0.01, *** P < 0.001.
DNA cloning and protein production
SGTA gene fragments encoding the following constructs: NT (1-86, including the linker), TPR (85-213), NT-TPR (1-213), FLΔQ (1-274), TPR-CTΔQ (85-274), CTΔQ (213-274), CT (213-313), and the FL SGTA (1-313) were PCR amplified from cDNA (Life Technologies) and inserted into BamHI/XhoI restriction sites of a home-modified pET28 vector, encoding an N-terminal thioredoxin A fusion protein followed by a hexahistidine tag and tobacco etch virus (TEV) protease cleavage site.
Typically, protein expression was carried out in BL21 (DE3) strains after induction with 0.5 mM isopropyl-β-d-thiogalactopyranoside (IPTG) at OD600 ≈ 0.8, followed by overnight incubation at 20 °C. For labeled proteins, growth was carried out in M9 media supplemented with labeled ammonium chloride (> 98% 15N, Sigma-Aldrich), glucose (> 99% U-13C, Sigma-Aldrich), and/or 100% D2O (Sigma-Aldrich).
Harvested cells were resuspended in lysis buffer (20 mM potassium phosphate, pH 8.0, 300 mM NaCl, 10 mM Imidazole, 250 μM TCEP), supplemented with protease inhibitors (0.3 μM Aprotinin, 10 μM Leupeptin, and 1 μM Pepstatin A), and 1 mM PMSF and lysed by sonication or cell disruptor (Constant Systems Ltd.). Cell debris was removed by centrifugation, and the soluble fractions were purified using nickel affinity chromatography (HisTrapTM HP 5 ml, GE Healthcare) and eluted with buffer containing 300 mM imidazole, followed by dialysis into cleavage buffer (20 mM potassium phosphate, pH 8.0 and 300 mM NaCl) and digestion with homemade TEV protease (≈ 100 μg/ml) at 4 °C overnight. After removing the fusion protein and the histidine tag by nickel affinity chromatography, the target protein was recovered in the flow through and gel filtration steps were carried out using HiLoad 16/60 Superdex 200 column (GE Healthcare), previously equilibrated in buffer containing 10 mM potassium phosphate pH 6.0, 100 mM NaCl and 250 μM TCEP. Proteins were concentrated using Vivaspin concentrators (Sartorius Stedin) and the sample purity and homogeneity was assessed by SDS-PAGE and NMR.
Biochemical characterization
Analytical size-exclusion chromatography (SEC) was performed using Superdex 200 10/300 GL column (GE Healthcare) pre-equilibrated with 10 mM KPi, 100 mM NaCl, pH 6.0 buffer. Molecular mass was estimated based on the migration of protein standards on the SEC column (aprotinin – 6.5 kDa, ribonuclease A – 13.7 kDa, carbonic anhydrase – 29.0 kDa, ovalbumin – 44.0 kDa, conalbumin – 75.0 kDa, and aldolase – 158.0 kDa; GE).
Circular dichroism (CD) spectra of CT, CT 3x NNP/AAA and CTΔQ SGTA constructs were acquired using an Aviv Circular Dichroism Spectrophotometer, Model 410 (Biomedical Inc., Lakewood, NJ, USA). Protein samples were adjusted to 0.5 mg/ml in 10 mM KPi, 100 mM NaCl, and pH 6.0 buffer, and the experiments were recorded using a rectangular demountable Suprasil quartz cell of 0.1 mm pathlength (Hellma Analytics). Each sample was scanned three times from 260 to 195 nm, at 1-nm intervals with an averaging time of 0.5 s. After the background subtraction for all CD spectra, data were converted to mean residue molar ellipticity and deconvoluted using SELCON3 [38].
Dynamic light scattering (DLS) was performed using a Nanosizer S diffraction particle sizer (Malvern Instruments, UK) with a 5003 multi-digital correlator. The light source was a 2 mW He-Ne laser, linearly polarized, with λ = 633 nm, and scattering angle θ = 173°. Samples were prepared at 0.5 mg/ml in 10 mM KPi, 100 mM NaCl, and pH 6.0 buffer and loaded into 0.5 ml volume disposable cuvettes (Sigma, Poole, UK). The experiments were measured at room temperature in triplicate.
NMR
For NMR experiments, protein samples were prepared at concentrations between 200 and 500 μM containing 10% D2O (Sigma-Aldrich) in 10 mM potassium phosphate pH 6.0, 100 mM NaCl, and 250 μM TCEP buffer (supplemented with 10 μM DSS for proton chemical shift referencing). All experiments were acquired in 5 mm NMR tubes at 25 °C using Bruker Avance spectrometers at 500, 700, or 950 MHz equipped with cryoprobes and controlled by the TopSpin 3.1 software package. Backbone assignments were carried out using 3D experiments (HNCO, HN(CA)CO, CBCA(CO)NH, and CBCANH) [39] for CT, TPR-CTΔQ, and NT constructs (respective BMRB accession numbers: 27272, 27275, and 27276); assignment of the other constructs was compiled using the same information. NMRPipe [40] and CcpNMR Analysis [41] were used for spectral processing and analysis.
Optimization
Since two stretches of peaks were missing from our SGTA_CT spectra, we tested numerous conditions to optimize the experiments and potentially reveal the missing peaks. We produced a variety of different SGTA_CT constructs, removing potentially aggregating regions (CTΔQ) and adding the contiguous stable domain, TPR (TPR_CT and TPR_CTΔQ). For each of these, we ran 1H-15N HSQC experiments at a range of temperatures (Additional file 4: Figure S4), pH (6.0–8.9), protein concentrations (10–500 μM), and with the addition of detergents at various concentrations (DDM or OG at 0.05–0.2%).
Relaxation
NMR relaxation experiments were performed for NT, TPR, NT-TPR, TPR-CTΔQ, and FL constructs at concentrations between 200 and 400 μM. 15N - {1H} heteronuclear measurements were obtained from the ratio of crosspeak volumes between two experiments recorded with 4 s of interscan delay (equilibrium) or 4 s of proton saturation (saturated). A spectrum series with 30.8, 61.6, 123.2, 246.4, 369.6, 554.4, 739.2, 985.5, 1232, 1386, and 1540 ms of inversion-and-recovery delays and 16.96, 33.92, 50.88, 67.84, 84.8, 118.72, 152.64, 186.56, 220.48, and 254.4 ms of CPMG echo delays was recorded for T1 and T2 measurements, respectively. 15N T1 and T2 relaxation times were computed using standard methods analogous to previous approach [42], from the single exponential decay fitting of the peak intensities for each amide signal. Correlation times (τc) have been estimated from the T1/T2 averaged values for each domain (NT comprising residues from 5 to 65 and TPR from 87 to 206) using the following equation:
$$ {\tau}_c\approx \frac{1}{4\pi {\upsilon}_N}\sqrt{6\frac{T_1}{T_2}-7} $$
Values for n were as follows: NT = residues 5–65: in construct NT, n = 57; in construct NT-TPR, n = 45; and in construct FL, n = 37 and TPR = residues 87–205: in construct TPR, n = 97; in construct NT-TPR, n = 81; in construct TPR-CTΔQ, n = 92; and in construct FL n = 40.
Native mass spectrometry
Mass spectra of SGTA samples (NT-TPR, FLΔQ, FL, CT, and CT 3x NNP/AAA mutant) were recorded on a Synapt HD mass spectrometer (Waters) modified for studying high masses. Protein samples were exchanged into 0.20–0.75 M ammonium acetate (pH 7.0) solution using Micro Bio–Spin 6 chromatography columns (Bio-Rad) and diluted to a final concentration of 5–10 μM before analysis. 2.5 μL of protein solution was electrosprayed from a borosilicate emitter (Thermo Scientific) for sampling. Typical conditions were capillary voltage 1.8–2.5 kV, cone voltage 60–120 V, collision voltage 10–30 V, with backing pressure 3–4 mbar, and source temperature of 20 °C. Spectra were calibrated externally using cesium iodide. Data acquisition and processing were performed using MassLynx 4.1.
SAXS
Small-angle X-ray scattering data were collected at the EMBL beamline P12 at PETRA 3 storage ring (DESY, Hamburg). All measurements were carried in 10 mM KPi, 100 mM NaCl, pH 6.0, buffer at 25 °C with protein solutions at concentrations ranging from 0.5 to 8.8 mg/ml (for NT-TPR and FL SGTA). The experiments were recorded using a PILATUS 2-M detector (DECTRIS, Switzerland) at a sample-detector distance of 3.1 m and a wavelength of λ = 0.12 nm, covering the range of momentum transfer 0.07 < s < 4.80 nm−1 (s = 4π sinθ/λ, where 2θ is the scattering angle). The measurements were taken in an in-vacuum capillary; no measurable radiation damage was detected by comparison of 20 successive frames with 50-ms exposures. The experimental scattering profiles from all solutes were corrected for the solvent scattering, normalized against transmitted intensity and sample concentration, and processed using standard protocols [43]. Extrapolation to infinite dilution and merging of different data sets were performed with PRIMUS [43]. The overall parameters such as radius of gyration (Rg), the maximum particle dimension (Dmax), and the Porod volume (Vp) were evaluated using standard procedures [43]. The program GNOM [44] was used to calculate the distance distribution function. The molecular weight (MW) was estimated by comparing the forward scattering with that of a standard protein (bovine serum albumin).
The flexibility of the different constructs was compared using the program EOM 2.0 [45]: EOM 2.0 is a program that fits the averaged theoretical scattering intensity from an ensemble of conformations into the experimental SAXS data. A pool of n-independent models based upon sequence and structural information was first generated. Then, a genetic algorithm was performed for the selection of the ensemble of conformations that best fit the data. High-resolution structures for individual subunits, if available, were used as constraints for the generation of the pool. Data were deposited in the SASBDB [46] under accession codes: SASDDB6 and SASDDC6.
TPR mutant design, spin labelling, and EPR sample preparation
SGTA contains four cysteine residues (C38, C129, C148, and C153), and to prepare the single cysteine mutants in NT-TPR and FL SGTA constructs we first changed all wild-type cysteine amino acids into serine residues using site-directed mutagenesis. We confirmed that the folding of the resultant cysteine-free protein is conserved using NMR. Then, four positions were selected (S88, S136, C153, and S197) to contain the solvent-exposed cysteine for spin labeling. These different mutants were created as well using site-directed mutagenesis (except in the case of the wild-type C153). Proteins were prepared in the same way as the wild-type versions.
Two milliliters of 0.25 mM SGTA (NT-TRP and FL mutants) in 20 mM potassium phosphate and 300 mM NaCl buffer at pH 8.0 were incubated with 150 μl of 37.8 mM (1-oxyl-2,2,5,5-tetramethylpyrroline-3-methyl) methanethiosulfonate (MTSL, Santa Cruz Biotechnology) spin label (~ 20-fold excess) overnight at 4 °C in the dark. The spin label was removed by size-exclusion chromatography using HiLoad 16/60 Superdex 200 column (GE Healthcare), previously equilibrated in buffer containing 20 mM potassium phosphate pH 7.0 and 100 mM NaCl. Continuous-wave EPR spectra were acquired on samples in the elution buffer, whereas PELDOR samples were exchanged into D2O-containing buffer (Sigma-Aldrich) before dilution with 50% d8-glycerol (Sigma-Aldrich). The elution volume of each labeled mutant following size-exclusion chromatography was similar for all labeled proteins, indicating similar stability of the mutants and the wild-type SGTA. To make EPR samples (final concentration 300–500 μM), ~ 200 μL of a given sample were transferred to a borosilicate glass tube (O.D. 5 mm, Wilmad 500 MHz precision). Labeling efficiencies were above 70%, as determined by continuous-wave EPR [47].
EPR spectroscopy
Continuous-wave EPR measurements were conducted at room temperature on a Bruker E-scan bench top spectrometer, with 1-mW microwave power, 0.1-mT modulation amplitude and 20-ms conversion time. After flash freezing the samples in liquid nitrogen, DEER measurements were performed at 50 K on an ELEXSYS E500 spectrometer (Bruker) operating at 9.6 GHz equipped with an ER 4118 X-MD5 resonator, a cryogen-free close-circuit cryostat (Cryogenics Ltd.) and a Lakeshore temperature controller. The four-pulse double electron-electron resonance sequence [48] used was π/2(νobs)-τl-π(νobs)-t-π(νpump)-(τl + τ2-t)-π(νobs)-τ2-echo, where the observer pulse lengths were 16 and 32 ns for the π/2 and π pulses, respectively. The pump pulse length (π(νpump)) was 12 ns and τ2 was 7 or 8 μs. All other parameters, namely τ1 = 400 ns and Δτ1 = 56 ns for nuclear modulation averaging, were selected as described earlier [48]. Data points were collected in 16 ns time steps. The acquisition time for each DEER spectrum was between 3 and 12 h. Time-domain spectra were analyzed using the program DeerAnalysis2016 [49]. A homogeneous three-dimensional fit was used as background correction and the distance distributions computed by Tikhonov regularization.