Human CTP synthase filament structure reveals the active enzyme conformation.

Purification of ecCTPS

Wild-type ecCTPS was purified as described previously. 3 , 14 The ecCTPS CC mutant was purified in the same manner, except 10 mM DTT was included throughout the purification. Purification of hCTPS1 hCTPS1 was expressed in S. cerevisiae strain GHY55, as described by Han et al. 2005. 32 This strain lacks the endogenous S. cerevisiae CTPS genes URA7 and URA8, and contains plasmid pDO105-hCTPS1, which directs expression of 6His-tagged (C-terminal) hCTPS1 from the ADH1 promoter. GHY55 cells were grown in 4× YPD media and harvested by freezing cell pellets in liquid nitrogen. Cell pellets were ground to a powder while frozen, and 15g of cell powder was resuspended in 100 mL of lysis buffer (50 mM Tris-HCl, 200 mM NaCl, 0.3M sucrose, 20 mM imidazole, 0.5 mM PMSF, pH 8.0). Lysates were clarified by centrifugation at 14,000 RPM for 40 minutes at 4°C in a Thermo Scientific Fiberlite F14-14×50cy rotor. Clarified lysates were applied to a 5mL HisTrap FF Crude column (GE) on an ÄKTA Start chromatography system (GE), and the column was washed with 30 column volumes (CV) wash buffer (20 mM Tris-HCl, 0.5M NaCl, 45 mM imidazole, 10% glycerol, pH 7.9). Protein was eluted as 1mL fractions with 5CV elution buffer (20 mM Tris-HCl, 0.5M NaCl, 250 mM imidazole, 10% glycerol, pH 7.9). Fractions containing hCTPS1 were pooled and dialyzed into storage buffer (20 mM Tris-HCl, 0.5M NaCl, 10% glycerol, 7 mM β-mercaptoethanol, pH 7.9) using Snakeskin 3500 MWCO dialysis tubing (Thermo Scientific). Dialyzed protein was concentrated ∼6-fold by centrifugation in a 3 kDa cut-off centrifugal filter unit (Millipore). hCTPS1 prepared for cryoEM was purified in the same manner, except protein was concentrated prior to dialysis, and then dialyzed into 20 mM Tris-HCl, 50 mM NaCl, 7 mM β-mercaptoethanol, pH 7.9. The hCTPS1-H355A mutant was generated by site-directed mutagenesis of plasmid pDO105-hCTPS1, and purified in the same manner as wild-type hCTPS1.

Site-directed mutagenesis

Site-directed mutagenesis of the E. coli and human CTPS expression constructs was performed using the QuickChange (Agilent Technologies) and Q5 (NEB) systems, respectively.

CTPS activity assays

CTPS activity was determined at 37°C by measuring Δotimes;Abs 291nm over time using a NanoDrop 2000c spectrophotometer (Thermo Scientific). Absolute CTP production was calculated using the change in extinction coefficient between UTP and CTP at 291 nm (1338 M -1 cm -1 ). 33 Kcat values were determined from the initial linear region of the Abs 291nm versus time measurements. For ecCTPS, assays were performed with 0.5-2.5 μM ecCTPS or ecCTPS CC in a standard reaction buffer containing 50 mM Na-HEPES (pH 8.0) and 10 mM MgCl 2 , with or without 10 mM DTT as appropriate. Final substrate concentrations were 0.6 mM UTP, 1.5 mM ATP, 0.2 mM GTP and 10 mM glutamine or ammonia. Assays were carried out using the “annealing” method described by Habrian, et al 34 : ecCTPS was incubated at 21°C for 3 minutes, and then combined with pre-warmed (37°C) nucleotides. The enzyme-nucleotide mix was then incubated for a further 4 minutes at 37°C, after which pre-warmed glutamine or ammonia was added. The complete reaction mixture was then immediately transferred to a pre-warmed UV cuvette and Δotimes;Abs 291nm over time measured at 37°C. For hCTPS1, assays were performed with 8 μM hCTPS1 in a standard reaction buffer containing 20 mM Tris-HCl (pH 7.9) and 10 mM MgCl 2 , with final substrate concentrations of 2 mM UTP, 2 mM ATP, 0.2 mM GTP, and 10 mM glutamine. The annealing method described above was also used for hCTPS1 activity assays, except the enzyme-nucleotide mix was incubated for 1 hour at 37°C to allow for hCTPS1 polymerization. Negative stain EM Samples for negative stain EM were prepared by applying CTPS to carbon-coated grids and staining with 0.7% uranyl formate. For ecCTPS CC , protein was dialyzed overnight into 50 mM Na-HEPES (pH 8.0) with or without 10 mM DTT before being coated onto grids at a final concentration of 2.5 μM. For hCTPS1 and hCTPS1-H355A, samples were prepared with 2.5 μM protein in 20 mM Tris-HCl (pH 7.9) and 10 mM MgCl 2 , supplemented with appropriate nucleotides at the following concentrations: UTP (2 mM), ATP (2 mM), AMP-PNP (2mM), GTP (0.2 mM), ADP (2 mM), and CTP (2 mM). hCTPS1 was incubated with nucleotides for 1 hour at 37°C before being coated onto grids. For depolymerization experiments, glutamine (10 mM) was added following incubation with nucleotides, and negative stain samples were prepared 30 and 120 minutes following glutamine addition. Electron microscopy was performed on a Tecnai G2 Spirit (FEI co.) operating at 120 kV, and images were acquired at 52,000× magnification on a US4000 4k × 4k CCD camera (Gatan, Inc.). CryoEM Samples for cryoEM were prepared by applying protein to glow-discharged C-Flat holey-carbon grids (Protochips Inc.), blotting with a Vitrobot (FEI co.), and rapidly plunging into liquid ethane. For ecCTPS CC , protein was dialyzed into non-reducing buffer (50 mM Na-HEPES pH 8.0) to promote filament assembly. Samples for cryoEM were then prepared with 2.5 μM ecCTPS CC in the absence of nucleotides or following 1 hour incubation of filaments with 10 mM MgCl 2 in addition to UTP (0.6 mM) and ATP (1.5 mM) or CTP (0.6 mM) and ADP (1.5 mM). hCTPS1 (5 μM) was incubated for 1 hour with 10 mM MgCl 2 , 2 mM UTP, 2 mM ATP, and 0.2 mM GTP before preparing cryoEM samples. For ecCTPS tetramers, 5 μM ecCTPS was incubated with 10 mM MgCl 2 , 0.6 mM UTP, 1.5 mM AMP-PNP, 0.2 mM GTP and 10 mM glutamine before preparing cryoEM samples. CryoEM data for the wildtype ecCTPS filament were acquired in a Tecnai Polara operating at 300 kV and recorded on a K2 Summit Direct Detect camera with a total dose of ∼34 e - /Å 2 with 36 frames per exposure. For all other samples, cryoEM data was collected on a Tecnai G2 F20 (FEI co.) operating at 200 kV with a K2 Summit Direct Detect camera (Gatan Inc.) with a pixel size of 1.26 Å/pixel. Movies were acquired in counting mode with 36 frames and a total dose of ∼45 e - /Å 2 (filaments) or ∼68 e - /Å 2 (ecCTPS tetramers), and with a defocus range between 0.8 and 2.5 μm (filaments) or 1.0 and 4.5 μm (ecCTPS tetramers). Leginon software 35 was used to automate data collection.

Image processing and 3D reconstruction

Movie frames were aligned using DOSEFGPU DRIFTCORR 36 and CTF parameters were estimated using CTFFIND3 37 . For helical samples, lengths of helices were manually defined using Appion 38 manual picker, and overlapping segments were extracted along the length of each helix. Particle stacks were CTF corrected by phase flipping in SPIDER 39 , using parameters determined by CTFFIND3. 3D reconstruction of helices was performed by iterative helical real space reconstruction (IHRSR) 40 , 41 in SPIDER, using hsearch_lorentz 42 to refine helical symmetry parameters. Cylinders were used as starting models, and the D2 point group symmetry of the CTPS tetramer was enforced. Iterative gold-standard refinement was performed with increasingly smaller angular increments (minimum 1.5°). For single particle (tetramer) samples, particles were picked automatically using DoG Picker 43 and then extracted using Appion. Relion 44 was used for CTF correction (using CTFFIND3 parameters) and subsequent classification and refinement. We performed reference-free 2D classification, and selected 7700 and 39000 particles for 3D classification of the hCTPS1-H355A and ecCTPS tetramers, respectively. For 3D classification, initial reference volumes were generated from the homology model of the hCTPS1 tetramer or crystal structure of the product-bound ecCTPS tetramer. Both reference volumes were low-pass filtered to 60Å, and D2 symmetry was enforced during 3D classification. Following 3D classification, 4400 and 6400 particles were selected for hCTPS1-H355A and ecCTPS, respectively, for gold-standard 3D refinement. Again, reference volumes were low-pass filtered to 60Å and D2 symmetry was enforced during 3D refinement. The gold-standard FSC=0.143 criterion was used for estimating resolution. Raw volumes were amplitude corrected and low and high-pass filtered using SPIDER. Details of each 3D reconstruction are summarized in Table 1 .

Building and fitting of atomic models

An ecCTPS monomer (PDB 2AD5) was initially fit as a single rigid body into the cryo-EM structure of the filament. For the human filament an initial homology model of the full-length hCTPS monomer was generated from the crystal structures of the individual domains (PDB 2VO1 and 2VKT) aligned to ecCTPS (PDB 2AD5) using MODELLER 45 , then fit to cryoEM density as three rigid bodies: GAT domain, linker domain, and ALase domain. For both structures atomic models were refined with the Rosetta Relax application 46 , using the cryoEM density as a constraint. Adjacent protomers within the tetramer and at filament contacts sites were included during the Relax procedure, to allow for refinement at subunit interfaces. X-ray structure of an alternate CTP-inhibited ecCTPS conformation N-terminal His-tagged ecCTPS mutant C268A was expressed and purified using metal chelate chromatography as previously described. 16 ecCTPS(C268A) had identical k cat and CTP inhibition as wildtype. Protein for crystallization was stored in 10 mM Tris-Cl, 0.5 mM TCEP, pH 8.0 at -80°C at 10-15 mg/mL. P2 1 2 1 2 crystals isomorphous to the published apo- and CTP/ADP-liganded structures (1S1M and 2AD5) were grown using vapor diffusion from 7.5 mg/mol Ec CTPS-C268A, 0.8-1.2 M ammonium sulfate, 0.1 M Tris-Cl, pH 8.0 as previously described 14 with drops supplemented with 5 mM CTP, 5 mM magnesium sulfate and 10 mM glutamine. For cryoprotection, crystals were briefly rinsed in 1:1 mixture of 50% MPD:mother liquor (25% MPD final), wicked and flash-cooled in liquid nitrogen. Reflection data were acquired at Stanford Synchrotron Radiation Lightsource Beamline 1-5 at 100°K at a wavelength of 0.979610 Å and 1° oscillation/image, using a ADSC Quantum 315 detector. Eighty eight frames were processed using DENZO/Scalepack (HKL2000 package). Phases were derived from the water-free published CTP and ADP complex structure (PDB 2AD5). Maximum-likelihood structure factor refinement was carried out on non-hydrogen atoms using Refmac 5.6 with Babinet scaling with an anisotropic B-factor correction (BSOL fixed at 125.00 and final anisotropic scaling parameters B11 = -1.82 B22 = 2.83 B33 = -1.01 B12 = 0.00 B13 = 0.00 B23 = 0.00). Quaternary changes were accounted for by initial rigid body refinement for individual monomers and then for three separate domains of each monomer (1-266, 267-287, 288-544). Cycles of manual model building with Coot 0.6.2 and combined positional and B-factor refinement led to the final model. The distribution of favored, allowed and outlier Ramachandran angles were 94%, 5% and 1%, respectively (51 st percentile for structures at this resolution). The data collection and final model statistics are given in Table 2 .

Sequence comparison

CTPS sequences were obtained from BLAST searches 47 , and multiple sequence alignments of several hundred sequences was performed with MAFFT 48 .

Data availability

Coordinates and structure factors for the CTP-inhibited ecCTPS crystal structure have been deposited in the Protein Data Bank under accession code PDB 5TKV. EM structures and associated atomic models have been deposited in the Electron Microscopy Data Bank and Protein Data Bank with the following accession codes: ecCTPS filament (EMD-8504; PDB 5U3C), hCTPS1 filament (EMD-8474; PDB 5U03), hCTPS1-H355A tetramer (EMD-8476), ecCTPS tetramer (EMD-8475; PDB 5U05), ecCTPS-CC (apo) (EMD-8490), ecCTPS-CC (substrates) (EMD-8491), ecCTPS-CC (products) (EMD-8513; PDB 5U6R). Other data supporting this study are available from the corresponding authors upon reasonable request.

Supplementary Material 1