Electronic Supporting Information

Sensitive proton-detected solid-state NMR spectroscopy of large proteins with selective CH3 labelling: application to the 50S ribosome subunit Vilius Kurauskas, a,b,c Elodie Crublet,*,a,b,c Pavel Macek,a,b,c Rime Kerfah,d Diego F. Gauto,a,b,c Jérôme Boisbouviera,b,c and Paul Schanda*,a,b,c

Table of Contents Protein production, purification and sample preparation..................................................................... 2 Materials.......................................................................................................................................... 2 Production and purification of TET2...............................................................................................2 Culture of Thermus thermophilus.................................................................................................... 2 Isolation of 50s ribosomes............................................................................................................... 3 Preparation of rotors for MAS ssNMR............................................................................................3 NMR spectroscopy and data analysis...................................................................................................4 NMR setup and experiments............................................................................................................4 Processing and analysis of NMR data............................................................................................. 6 Comparison of solution and solid-state NMR spectra of TET2........................................................... 6 Additional discussion........................................................................................................................... 7 Proton longitudinal relaxation and recycle delay............................................................................ 7 CP vs INEPT transfer.......................................................................................................................7 References............................................................................................................................................ 9

Protein production, purification and sample preparation Production and purification of TET2 Expression of TET2 in D2O-based minimal medium was performed following established protocols for specific labelling of isoleucine-δ1 1,2 or alanine3 methyl groups. Briefly, bacteria were adapted to D2O-based media with successive precultures in LB (H 2O), M9 (H2O), M9 (H2O/D2O), M9 (D2O). The culture medium (M9 medium with 1 g/L 15NH4 and 2 g/L 2H,12C-glucose as sole nitrogen and carbon sources) was supplied with the corresponding precursors, as described, 2,3 one hour before induction. Purification was performed using heat shock (85º C), anion exchange chromatography and size exclusion chromatography, as described earlier. 4 Typical protein yields were of the order of 20-25 mg/L of culture.

Culture of Thermus thermophilus Strain HB8 Thermus thermophilus (ATCC 27634) was stored at –80°C, in a protonated synthetic medium described by Yoshida et al.5 containing 30% glycerol. Deuterated synthetic medium for growth of Thermus thermophilus strain in D2O was adapted from the reported protonated medium.5 Deuteration of the inorganic materials was achieved by 2 repeated cycles of dissolving in D 2O followed by lyophilization. Cells were adapted to fully deuterated medium in a three–step manner. The inoculum volume was usually no less than A600= 0,4 to avoid lag phase. First, culture was started from one glycerol stock (A600=2) inoculated in 50 mL of H 2O synthetic medium for 24 h at 75°C. This first preculture was transferred in a larger volume of synthetic (200 mL) medium prepared in 50% D 2O at A600=0,4 and grown for 24 h (Abs=2-2,5) and finally a last preculture (800 mL) was grown in alanine 13CH3-enriched 100% D2O medium in the same conditions (A600=0,4, 24 h, 75°C). This last culture was centrifuged and used as a starting culture for ribosome production. Culture (3.5 L) was inoculated at A600=0,4 in alanine 13CH3-enriched 100% D2O medium and cells were grown at 75°C to an A600=1,6 (found to be half of the saturation value determined in enriched D2O synthetic medium). The cells were then cooled down to 4°C to avoid run-off ribosomes and harvested by centrifugation.

Isolation of 50s ribosomes T. thermophilus ribosomes were isolated according to the slightly modified procedure for E. coli ribosomes, described by Fechter et al.6 Briefly, 1L cultured cells were collected in mid-log phase and resuspended in 25 mL of buffer AE (20 mM Tris pH 7,5, 200 mM NH 4Cl, 20 mM MgCl2, 0,1 mM

EDTA, 6 mM β-mercatpoethanol). Cells were disrupted by one passage in a cell-disrupter, and cell debris were removed by centrifugation for 20 min at 24 000 rpm at 4°C (Ti 45). Ribosomes were collected by centrifugation at 150 000 g in a Ti45 rotor for 4h at 4°C. Salt-washed ribosomes were obtained from a centrifugation on a 42% sucrose cushion in buffer 1E (20 mM Tris pH 7,5, 500 mM NH4Cl, 20 mM MgCl2, 0,1 mM EDTA, 6 mM β-mercaptoethanol) for 16h at 38 000 rpm (Ti 70 rotor). Pelleted ribosomes were resuspended in 0,5 to 1 mL of buffer BE (20 mM Tris pH 7,5, 50 mM NH4Cl, 10 mM MgCl2, 0,1 mM EDTA, 6 mM β-mercaptoethanol) and extensively dialyzed in buffer DE (20 mM Tris pH 7,5, 300 mM NH4Cl, 1 mM MgCl2, 0,15 mM EDTA, 2 mM DTT) at 4°C. The sample was then layered on top of a 5-20% sucrose gradient prepared in DE buffer, with a gradient maker, in SW32 rotor at 19 000 rpm for 16h (20-50 AU/ tube). Sucrose gradients were fractionated and OD 260 nm was monitored. Fractions containing pure 50s ribosomes were pooled, buffer exchanged against buffer CE (20 mM Tris pH 7,5, 50 mM KCl, 10 mM MgCl2, 0,1 mM EDTA, 1 mM DTT) and concentrated on a Amicon 100 kDa centrifugal device for NMR analysis.

Preparation of rotors for MAS ssNMR TET2 protein in buffer (20 mM Tris, 20 mM NaCl, pH 8, H 2O-based) was concentrated to 10 mg/ml and precipitated/nanocrystallized by addition of 2,4-methyl-pentane-diol (MPD) at a ratio of 50%/50% (volume). The resulting white precipitate was filled into 1.3 mm MAS ssNMR rotors (Bruker Biospin) using a centrifugal device, adapted for a Beckman SW32 centrifuge rotor. Rotors were typically filled with TET2 samples in an ultracentrifuge at 20000 rpm (approximately 50000 g) during 2 hours. We also performed experiments of TET2 Ala-CH 3 samples with a protein D2Obased buffer. The line widths are essentially the same as in H 2O buffer, showing that the addition of amide protons does not lead to significant line broadening. Ribosome samples (in D2O) were directly sedimented into 1.3 mm MAS rotors using the ultracentrifuge device, without addition of any precipitant, at 20000 rpm (SW32) for 16 hours at 4ºC.

NMR spectroscopy and data analysis NMR setup and experiments All NMR experiments were performed on a Bruker 600 MHz Avance2 spectrometer, equipped with a 1.3 mm MAS probe tuned to 1H, 13C, 15N and 2H frequencies. MAS frequencies of 20-57 kHz were employed, as specified in the text. In all cases, the sample temperature was kept constant, by adjusting the VT gas temperature. The effective sample temperature, determined by the bulk water frequency and the MPD signal at 4.1 ppm as a reference (in TET2 samples), was set to 30 °C (TET2) and 22 °C (ribosome), respectively. Pulse sequences used in this study are shown in Figure S1. Typical RF field strengths for hard pulses were 100 kHz (1H), 92 kHz (13C) and 75 kHz ( 15N). Cross-polarization steps applied a linear ramp on 1H (with typical mean RF field strengths of about 90 kHz at MAS frequencies of 40-57 kHz, ramped ± 5%) and the 13C RF field strength was optimized to the n=1 Hartmann-Hahn condition (e.g. about 40 kHz at 55 kHz MAS). Low-power CP conditions resulted in similar transfer efficiencies. For the 2D spectra shown in Figure 1 on the protein TET2, the chemical-shift evolution times were 49 ms (1H) and 74 ms (13C). The INEPT-based correlation spectrum of ribosome 50s shown in Figure 3 (black) was acquired with evolution times of 49 ms ( 1H) and 68 ms (13C), using 88 scans per increment and 256 increments (spectral width 25 ppm), resulting in a total experimental time of 12.5 hours. As a comparison an additional experiment was performed (based on CP transfer), which was intended to boost sensitivity, sacrificing some resolution. To this end, the 13C evolution time was reduced to 8.5 ms (8 times less than in the spectrum shown in black), and the number of scans was increased to 512 scans, resulting in a total experimental time of 9 hours. The relaxation rate constants shown in Figure 2 were determined from a series of 1D spectra with varying delay Δ, using typically 8 to 15 different time points. The experiments for measuring relaxation rate constants are shown in Figure 1. Except for the measurement of multiple-quantum versus single-quantum rate constants, these experiments used CP-based transfers, as shown in Figures S1 (B), (C), (D). MQ and SQ rate constants were measured with sequences shown in Figure S1 (G), (H). The comparison of absolute intensities of CH3 vs CHD2 samples shown in Figure 1g relies on a comparable absolute quantity of sample in the rotor. Different amounts of sample as well as different labelling efficiencies would bias the result. We performed two control experiments: first, a one-dimensional direct-detected 13C experiment was recorded with each of the samples, using a 12 seconds long recycle delay (to ensure that differences in longitudinal

13

C relaxation do not have

effects), at a 20 kHz MAS frequency, and 100 kHz TPPM 1H decoupling during 13C detection. In such an experiment, the different proton multiplicity on the methyls in CH3 and CHD2 samples is

not expected to alter the signal amplitude, such that these experiments reflect the sample amount. The relative signal integrals in these experiments matched to within about 7 %. As a second test, we performed a 15N-filtered hNH 1D experiment with the two samples. Again, the intensity observed in this spectrum is not dependent on the methyl proton multiplicity, such that the relative integrals of these spectra, obtained on the two samples, reflects only the sample amounts. Again, the signal integrals closely matched (