Strange India All Strange Things About India and world

Expression and purification of SHOC2

SHOC2 constructs (Homo sapiens residues 2–582, (except SHOC2ΔN, residues 91–582)) were expressed in Spodoptera frugiperda 9 (Sf9) cells. Cells were collected and stored at −80 °C. Cell pellets were resuspended in 100 ml per 1 l pellet of SHOC2 SEC buffer (10% glycerol, 20 mM Tris pH 7.5, 500 mM NaCl, 1 mM TCEP) supplemented with 20 mM imidazole, 1 EDTA-free protease inhibitor tablet, 0.2 mg Benzonase, and lysis detergents 0.3 % Sb3-14 and 0.03% C7BzO. The solution was homogenized and incubated with 1 ml per 1 l pellet of Ni-charged MagBeads (Genscript) for 1 h at 4 °C. Beads were washed with SHOC2 SEC buffer + 20 mM imidazole and protein was eluted with SHOC2 SEC buffer + 500 mM imidazole. SHOC2 was cleaved with TEV, dialysed into SHOC2 SEC buffer overnight at 4 °C, and re-applied to nickel beads. Flow-through containing SHOC2 was concentrated and run on a Superdex S200 16/600 in SHOC2 SEC buffer. Fractions containing purified SHOC2 were pooled, concentrated and snap frozen on liquid nitrogen.

Expression and purification of PP1C

PP1C constructs (H. sapiens residues 1–323) were expressed in Escherichia coli BL21(DE3) Tuner cells in Terrific Broth media supplemented with 1 mM MnCl2. Expression was induced at A600 = 1.0 with 50 µM IPTG and expression occurred overnight at 17 °C. Cells were collected and stored at −80 °C. Cell pellets were resuspended in 100 ml per 1 l pellet of PP1C SEC buffer (20 mM Tris pH 7.5, 700 mM NaCl, 1 mM MnCl2, 1 mM TCEP) supplemented with 20 mM imidazole, 1 EDTA-free protease inhibitor tablet, 0.2 mg benzonase, lysis detergents 0.3% Sb3-14 and 0.03% C7BzO and 20 mg lysozyme. The solution was homogenized and incubated with 2 ml per 1 l pellet of Ni-charged MagBeads (Genscript) for 1 h at 4 °C. Beads were washed with PP1C SEC buffer + 20 mM imidazole and protein was eluted with PP1C SEC buffer + 500 mM imidazole. PP1C was TEV cleaved, dialysed into PP1C SEC buffer overnight at 4 °C, and re-applied to nickel beads. Flow-through containing PP1C was concentrated and run on a Superdex S75 16/600 in PP1C SEC buffer. Fractions containing pure PP1C were pooled, concentrated and snap frozen on liquid nitrogen.

Expression and purification of RAS

H/K/NRAS constructs (H. sapiens residues: HRAS, 2–189; KRAS, 1–188; NRAS, 2–189; KRASΔHVR, 1–169) were expressed in E. coli BL21(DE3) cells in TB autoinduction medium at 17 °C for 48 h. MRAS constructs (H. sapiens residues 1–208) and KRAS–MRAS chimeras were expressed in Sf9 cells. Cells were collected and stored at −80 °C. Cell pellets were resuspended in 100 ml per 1 l pellet of RAS SEC buffer (20 mM Tris pH 7.5, 150 mM NaCl, 1 mM MgCl2, 1 mM TCEP) supplemented with 20 mM imidazole, 1 EDTA-free protease inhibitor tablet, 0.2 mg benzonase, lysis detergents 0.3 % Sb3-14 and 0.03% C7BzO and 20 mg lysozyme (for E. coli-expressed proteins). The solution was homogenized and incubated with 2 ml per 1 l pellet of Ni-charged MagBeads (Genscript). Beads were washed with RAS SEC buffer + 20 mM imidazole and protein was eluted with RAS SEC buffer + 500 mM imidazole. RAS was TEV cleaved, dialysed into RAS SEC buffer + 10 µM GDP overnight at 4 °C, and re-applied to nickel beads. RAS containing flow-through was concentrated and run on a Superdex S75 16/600 in RAS SEC buffer + 10 µM GDP. Fractions containing pure GDP loaded RAS were pooled, concentrated and snap frozen on liquid nitrogen.

For GCP-loaded RAS constructs, purified RAS was mixed with a 50-fold molar excess of GCP, 10 U alkaline phosphatase agarose beads (Sigma) and incubated with agitation at 37 °C for 1 h. Protein was buffer-exchanged into 25 mM Tris pH 7.5, 100 mM NaCl, 5 mM MgCl2, 10 µM GCP and snap frozen in liquid nitrogen.

Mass spectrometry

Fifty micrograms of each sample was buffer-exchanged into 50 mM ammonium acetate, pH 7. Samples were directly infused using a TriVersa NanoMate (Advion) and analysed online via nanoelectrospray ionization with a 5 µm nozzle ESI chip (Advion) into a Thermo Exactive Plus EMR Orbitrap mass spectrometer (Thermo Fisher Scientific). Acquired mass spectral data were analysed using UniDec software52.

Analytical SEC

Individual proteins or complexes were run on a size-exclusion column equilibrated in 20 mM Tris pH 7.5, 100 mM NaCl, 1 mM TCEP, 0.5 mM MnCl2, 0.5 mM MgCl2. Fractions were analysed by SDS–PAGE stained with coomassie blue. Columns used were: Fig. 1a and Extended Data Figs. 1b, 2d, 7a,d, 8a and 9c: S200 3.2/300; Extended Data Fig. 7g: S75 3.2/300.

Cryo-EM grid preparation and data collection

To prepare grids for electron microscopy, SEC pure SHOC2–PP1C–RAS complex was diluted to approximately 11 μM in RAS SEC buffer + 10 µM GCP, 4 μl of which was applied to holey gold grids (Ultrafoil R1.2/1.3; Quantifoil) that had been glow-discharged for 20 s using a Solarus plasma cleaner (Gatan). Grids were blotted for 3.5 s using a Vitrobot (Thermo Fisher Scientific) set to 4 °C and 100% relative humidity, and plunged into liquid ethane.

To prepare graphene oxide coated grids, the perpetually hydrated method was used53. Holey gold grids (Ultrafoil R1.2/1.3; Quantifoil) were glow-discharged for 20 s using a Solarus plasma cleaner (Gatan), and 4 μl graphene oxide flakes (Sigma-Aldrich) freshly diluted to 0.2 mg ml−1 in DDI water were applied to the front of the EM grid (side with holey layer). After 50 s incubation, 4 µl of SEC pure SHOC2–PP1C–RAS complex (diluted to approximately 1 μM in RAS SEC buffer + 10 µM GCP) were applied to the back of the EM grid (side with mesh). Without further incubation, grids were blotted for 3.5 s using a Vitrobot (Thermo Fisher Scientific) set to 4 °C and 100% relative humidity, and plunged into liquid ethane.

Data collection was performed using a Titan Krios (Thermo Fisher Scientific) operating at 300 keV, equipped with a BioQuantum energy filter (20 eV slit width) and a K3 (Gatan) direct detection camera. Movies were recorded in super-resolution mode, at a nominal magnification of 105,000× (calibrated pixel size 0.419 Å), 60 frames per movie, 50 ms per frame, and electron fluence per frame of 1.07 e Å−2 (total fluence 64 e Å−2). Data collection was automated using SerialEM54 with a set defocus range of 0.5 to 1.5 μm.

Cryo-EM data processing and model building

Initial steps of cryo-EM data were processed using CryoSPARC Live55. A total 10,780 of movies were motion corrected and CTF corrected. Micrographs with a CTF fit poorer than 7.0 Å were immediately discarded, leaving 5,060 micrographs. Particles were picked using the blob picker tool and 2D classified. 2D classes representing protein complex were used to re-pick, resulting in a stack of 3,996,056 particles, which were exported to CryoSPARC, where all further processing was undertaken. Particles were subjected to 2D classification to remove junk particles, leaving 1,261,442 particles. An ab initio Reconstruction was performed with 6 3D classes, revealing 4 classes of trimeric and 1 class of hexameric SHOC2–PP1C–MRAS volumes, along with 1 class of poorly resolved junk. A total of 551,091 particles encompassing the 2 best trimeric 3D classes were combined and used for non-uniform refinement56, with the best ab initio 3D volume as the starting model. Micrographs with a CTF fit poorer than 4.0 Å were further discarded, leaving 3,847 micrographs, and particles with a separation distance of less than 20 Å were also discarded. Remaining particles were subject to a further round of 2D classification and poorly resolved 2D classes were discarded, leaving 323,910 particles, which were re-extracted using a box size of 256 pixels. These particles were subjected to a final non-uniform refinement, including refinement of per-particle defocus and per-group CTF parameters (tilt and trefoil). This resulted in a final model with a gold standard Fourier shell correlation resolution of 2.95 Å.

Our crystal structure of SHOC2 (PDB: 7DS1), and existing structures of MRAS (PDB: 1X1S) and PP1C (PDB: 4MOV) were initially placed into the density map with ChimeraX57. The refinement and building were undertaken with Phenix58 Real Space Refine and COOT59 respectively. Molecular visualizations were created with ChimeraX and PyMol.

SHOC2 crystal structure determination

Crystals of SHOC2 were obtained using hanging-drop vapour diffusion with 1 µl of 6.0 mg ml−1 protein in SHOC2 SEC buffer mixed with 1 µl mother liquor (100 mM Tris pH 8.5, 200 mM MgCl2, 14%(w/v) PEG4000) over a reservoir of mother liquor at 16 °C. X-ray diffraction data was collected at the Stanford Synchrotron Radiation Lightsource beamline 12-2 with an X-ray wavelength of 0.97946 Å at 100 K. Data were integrated and scaled with XDS60. To obtain phases, molecular replacement was attempted with various homologous LRR proteins, where PDBL 4U06 (Leptospira interrogans LRR protein LIC10831, 23% sequence identity) resulted in a solution. The SHOC2 chain was initially placed with Phenix Autobuild, then iteratively built in COOT59 and refined in Phenix58.

Towards the end of this process, the predicted structure of SHOC2 became available in the AlphaFold Protein Structure Database61 (, and the LRR portion of the predicted structure was used as a molecular replacement solution. After building and refinement, Rwork/Rfree improved markedly to 0.209/0.238 vs 0.269/0.300 for the manually built structure. Therefore, the AlphaFold based molecular replacement solution with manual rebuilding and refinement against our experimental data resulted in the final model. Final Ramachandran statistics were 96.1% favoured, 3.9% allowed, 0% outliers.

SEC–multi-angle light scattering–SAXS

Data were collected at the ALS beamline 12.3.1 LBNL Berkeley, California62,63 with an X-ray wavelength of 1.127 Å. All experiments were performed at 20 °C and data was processed as previously described64. In brief, a SAXS flow cell was directly coupled with an HPLC system. A 55 µl volume of each sample was run through SEC and 3-s X-ray exposures were collected continuously during a 30-min elution. The SAXS frames recorded prior to the protein elution peak were used to subtract all other frames. The subtracted frames were investigated by radius of gyration (Rg) derived by the Guinier approximation \(I(q)\approx I(0){{\rm{e}}}^{-{q}^{2}{R}_{{\rm{g}}}^{2}/3}\) with the limits65qRg < 1.5, where q is the scattering vector. The elution peak was mapped by comparing the integral of ratios to background and Rg relative to the recorded frame using the program SCÅTTER. Final merged SAXS profiles, derived by integrating multiple frames at the elution peak, were used for further analysis. The program SCÅTTER was used to compute the P(r) function.

SAXS solution structure modelling

The model for full-length SHOC2 was built based on our x-ray crystal structure with the addition of missing N- and C-terminal regions in MODELLER66. BILBOMD67 was used to model conformational flexibility of the N-terminal region. The experimental SAXS profiles of SHOC2 were then compared to theoretical scattering curves of the atomistic models generated by BILBOMD using FOXS68,69, followed by multi-state model selection by MultiFoXS70,71.

The initial atomistic models of SHOC2–PP1C, SHOC2–PP1C–MRAS and SHOC2–PP1C–KRAS were built based on our cryo-EM structure. Minimal molecular dynamics simulations were performed on flexible regions in the models by the rigid body modelling strategy BILBOMD in order to optimize conformational space of SHOC2 N-terminal region. The selection of multi-state models was performed as described above for SHOC2.

Phosphatase assays

For peptide dephosphorylation assays, PP1C was mixed with varying concentrations BRAF pS365 peptide, SHOC2 and/or RAS at concentrations specified in each figure. At 2 min and 4 min time points, the reaction was stopped by the addition of 20 µl of reaction mix to 80 µl malachite green solution and incubated 30 min at room temperature for colour development. A standard curve of 0–100 µM inorganic phosphate was also generated. Absorbance was read at 640 nm and the enzymatic turnover for each reaction was calculated by reference to the standard curve.

For PNPP dephosphorylation assay, 10 nM PP1C, 1 µM SHOC2 and 1 µM KRAS were mixed with varying concentrations of PNPP. Absorbance was read at 405 nm at 1 min intervals. Turnover was calculated by reference to the extinction coefficient of NPP (dephosphorylated PNPP A405 = 18,000 M−1 cm−1) and was averaged across 5, 10, 15 and 20 min time points. Graphs were generated with GraphPad Prims8 for phosphatase assays and other biochemical assays.

Surface plasmon resonance

Avi-tagged KRAS was expressed and purified as described above. KRAS was biotinylated with a BirA biotin-protein ligase kit (Avidity) according to the manufacturer’s instructions. All SPR steps were performed on a Biacore S200 Instrument (Cytiva) at a temperature of 20 °C. In PBS a C1 chip (Cytiva) was functionalized with ~1,000–1,200 RU Neutravidin by activation with EDC/NHS followed by injection of 100 μg ml−1 Neutravidin (prepared in 10 mM sodium acetate, pH 4.6), and capping using 1 M ethanolamine. The surface was then conditioned with 1M NaCl/50 mM NaOH. After priming the system into data collection buffer (50 mM HEPES, pH 7.4, 100 mM NaCl, 500 µM TCEP, 0.01% Tween-20, 10 mM MgCl2, 500 nM GDP or GNP), biotinylated Avi-tagged KRAS pre-loaded with either GDP or GNP was then captured to ~40 RU using a brief injection of 10 s at 2 μl min−1, and remaining biotin-binding sites were blocked with 200 nM biotin for 60 s at 100 μl min−1. A twofold dilution series for each sample was injected sequentially from low to high concentration over the KRAS- coupled surfaces in multi-cycle kinetics mode, at a flow rate of 50 μl min−1 for 45 s, monitoring dissociation for 300 s.

All analysis was performed using the S200 Evaluation software after applying standard double-referencing. Theoretical Rmax  (the maximal feasible SPR signal generated by an interaction between a ligand–analyte pair; presented in response units (RU)) was determined as the (molecular mass of the protein in solution)/(molecular mass of the immobilized target) × (amount of immobilized target captured), and the per cent surface activity was determined as the (experimental Rmax)/(theoretical Rmax).

TR-FRET binding assay

C terminally SNAP-tagged PP1C and N-terminally SNAP-tagged SHOC2 were expressed and purified as described above. PP1C was labelled with SNAP-Lumi4-Tb labelling reagent (Cisbio) and SHOC2 was labelled with SNAP-Red labelling reagent (Cisbio) as per the manufacturer’s protocols. Sets of 20 µl solutions were made with final concentrations of 2 nM Tb–PP1C, 200 nM Red–SHOC2 and variable RAS concentrations, each component having been diluted with TR-FRET Assay buffer (25 mM Tris pH 7.5, 200 mM NaCl, 2 mM MgCl2, 2 mM MnCl2, 1 mM TCEP, 0.2 mg ml−1 BSA) and incubated at room temperature for 30 min. TR-FRET was read with excitation at 340 nm and emission at 620 nm and 665 nm. A second reading of each plate was taken after a 10 min delay to confirm that binding had reached equilibrium. The ratio of 665 nm emission was divided by 620 nm emission at each time point to give TR-FRET ratio.

DepMap data analysis

DepMap release Public 21Q4 datasets containing cell line information (sample_info.csv), chronos scores (CRISPR_gene_effect.csv) and mutational status (CCLE_mutations.csv), were downloaded from We assembled a dataset containing chronos scores for SHOC2, MRAS, KRAS and NRAS in 1,061 cell lines and annotated each cell line with the corresponding hotspot mutational status at the position G12, G13 and Q61 for KRAS, NRAS and HRAS. Overall dependency of each cell line on H/K/NRAS was calculated as the minimal value among H/K/NRAS chronos scores. The Pearson correlation coefficient and two-sided t-test P-value between SHOC2 and MRAS, HRAS, KRAS or NRAS chronos scores, gated by mutational status on G12, G13 and Q61 positions, were calculated across the 1,061 cell lines using the cor.test function in R. Data analysis was performed using R custom scripts.

Molecular dynamics

The complex structures were parametrized using FF19SB force field for proteins72 and parameters for phosphorylated amino acids73, fully solvated in OPC74 water boxes extending 12 Å from protein edges. The GCP ligand was parameterized using the Hartree-Fock/6-31 G* basis set with Gaussian09 to calculate the restraint electrostatic potential (RESP) charges, and antechamber to RESP fit the calculated potentials to generate the force field files. Na+ counter ions were added to neutralize the system. The GPU implementation of Amber 2018 with the SPFP precision model75 was used for the molecular dynamics simulation. First, the structure was relaxed with 2,000 steps of conjugate-gradient energy minimization in which the solutes and solvent were kept fixed. Then, the solvent molecules were allowed to move using the NPT ensemble with a temperature of 310 K. Another step of conjugate-gradient energy minimization was performed with 2,000 steps while removing all the restrains. Next, the pressure was maintained at 1 atm and the thermostat temperature increased to 310 K over the course of 500 ps, while Harmonic positional restraints of strength 10 kcal mol−1 Å−2 was applied to the solute. The system was then equilibrated for 1 ns with a restraint force constant of 1 kcal mol−1 Å−2. All restraints were removed for the production stage. The hydrogen mass repartition was used allowing a time step76 of 4 fs. A 10 Å cut-off radius was used for range limited interactions, with particle mesh Ewald electrostatics for long-range interactions. The production simulation was carried out using NPT conditions. Langevin dynamics77 was used to maintain the temperature at 310 K with a collision frequency of 3 ps−1. During dynamics the SHAKE algorithm78 was applied. Default values were used for all other simulation parameters. The production stage of the conventional molecular dynamics simulation was performed for 50 ns.

Each one of the four independent conventional molecular dynamics runs were followed by GaMD simulation45,79 module in AMBER 2018, which included 200-ps short cMD simulation used to collect potential statistics, 2-ns equilibration after adding the boost potential, and finally, 500-ns GaMD production runs. All GaMD simulations were run at the dual-boost level by setting the reference energy to the lower bound. The average and s.d. of the system potential energies were calculated in every 50,000 (100 ps). The upper limit of the boost potential s.d., σ0 was set to 6.0 kcal mol−1 for both the dihedral and total potential energetic terms. The simulation frames were saved every 20 ps for analysis.

CPPTRAJ80 was used to postprocess the ensembles from the combined GaMD trajectories. Contacts were determined by a simple distance cut-off (5 Å) between any two atoms within each residue. PCA was performed on the dihedral angles of the full structures. Each frame was RMS-fit to the first frame to remove global translational and rotational motion. The eigenvectors and eigenvalues were then obtained from diagonalization of the combined covariance matrix, after which coordinates from each independent trajectory were projected along eigenvectors of interest to obtain projection values for given modes. The isolated first principal components, Mode1 and Mode2, showing the largest variation in the data, were used to generate the 2D PCA density plots.

Reporting summary

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

Source link


Leave a Reply

Your email address will not be published. Required fields are marked *