Increase of enzyme activity through specific covalent modification with fragments

Structure-guided tethering of a fragment activator significantly increases enzyme activity.


Modelling of the linking strategy.
Tether models were constructed using Discovery Studio 3.5 [1]. Coordinates of 4-ethoxyquinazoline (PDB: BK9) and the residue chosen for the tether were copied from 4UR9. The tether was then built from this by adding and removing atoms and manually rotating bonds. The resulting models were typed in CHARMm. Fixed atom constraints were added to N, CA, C and O of the amino acid, as well as the ring atoms and oxygen atom of BK9 (including hydrogens bonded to these atoms). Three rounds of a Standard Dynamics Cascade followed by a Minimisation were performed using default parameters to produce the final tether model. [1] Discovery Studio, version 3.5. Accelrys Inc., San Diego, CA, USA: 2012.
The BtGH84 gene was expressed and purified as previously described [1]. Mutants were cloned using site-directed mutagenesis and expressed and purified in a manner analogous to wild-type BtGH84. Site-directed mutagenesis was performed with the Gibson Assembly method (New England Biolabs). Primers containing the appropriate mutations were used to prepare linear DNA amplicons with 16-22 bp overlaps. These fragments were mixed with linearised plasmid DNA with complementary overlaps and combined using the Gibson Assembly master mix. This allowed the introduction of multiple mutations in a single step. The resulting clones were fully sequenced to confirm that the correct mutations had occurred.
Primer sequences for BtGH84_TM -C420S, Y550C, C654S (5' → 3')  Overlaps:  F1+R5,F2+R1,F3+R2,F4+R3,F5+R4  PCR pairs:  F1+R1,F2+R2,F3+R3,F4+R4,F5+R5  F1  AACTTTAAGAAGGAGATATACCATGGGCAG  F2  AACTTGAAAGCTTTGCAATGCACAATTCCGA  F3  CAGAATCCTTGCCAACCGGGAGTAAA  F4  GGATGCTCCTAGCACATGGGGACGCC  F5  CGCGCCTTCTCCTCACTGTTCCA  ---------R1  CATTGCAAAGCTTTCAAGTTCTTCCGCT  R2  GGTTGGCAAGGATTCTGATTACTCGTC  R3  ATGTGCTAGGAGCATCCTTACCAAAGTTTATC  R4  GTGAGGAGAAGGCGCGTTATTTCTTCT  R5  GGTATATCTCCTTCTTAAAGTTAAACAAAATTATTTCT For Y137F mutants the BtGH84 and BtGH84_TM plasmids were modified using a similar approach to previously using the following primers: BtGH84_TM crystals were grown using the sitting-drop vapour diffusion method. 1 μl of 12 mg/ml BtGH84 was mixed with 1 μl of reservoir solution (125 mM imidazole pH 8.0, 4% PEG8000 (w/v), 3% trimethylamine N-oxide-dehydrate, 15% ethylene glycol) and the crystals were allowed to grow at room temperature for at least 2 days. Crystal soaking was performed by gradually adding well solution containing 10 mM PUGNac or Thiamet G. Crystals were picked with a nylon fibre loop (Hampton Research) and plunged into liquid nitrogen. Data were collected at Diamond beamline I02 or I03 using a Pilatus 6M-F or Pilatus3 6M detector respectively (Dectris) at 100 K. An oscillation range of 0.1° degree was chosen and 180° were collected. Data were autoprocessed in the Diamond xia2 pipeline, data reduction was carried out using Aimless (CCP4i2) and molecular replacement performed using MolRep (CCP4i2) from PDB 2CHO (A chain only). The model was improved and adjusted by alternating cycles of manual rebuilding and real-space refinement in COOT [2] and reciprocal-space refinement in Refmac [3] and Phenix [4]. Coordinate files and libraries for the ligand were created using Make Ligand (CCP4i2) or eLBOW (Phenix) from a SMILES string of the ligand. The quality of the final model was evaluated using the MolProbity-  °C. For all reactions the resulting product was separated from unreacted labelling compound using Zeba spin desalting columns, 7 kDa MWCO (Thermo). This clean-up step enabled simultaneously transfer of the product into appropriate buffer for the subsequent application.
Finally, all protein concentration values were determined via absorbance at 280 nm using extinction coefficients calculated using the ExPASy ProtParam tool.

Protein mass spectrometry
Electrospray Mass Spectrometry was conducted on a Waters LCT Premier XE system with MassLynx 4.1 software. The system was calibrated with sodium formate solution and calibration verified/corrected with horse heart myoglobin (16951.5 ± 1.5 Da).
Samples were purified from salt containing buffer solutions into ammonium acetate buffer using Zeba Spin Desalting Columns (7K MWCO, 0.5 mL) according to the manufacturer's protocol.
Samples in ammonium acetate buffer were diluted directly into 1:1 acetonitrile-water containing 0.1% formic acid.
Sample was infused into the MS source using a syringe pump. When a stable signal was achieved, data was collected for 3 minutes over a scan range of 200 -2000 m/z. The scans were combined and baseline subtracted. For samples of unknown mass, the spectrum was smoothed and centroided, then peak series identified by eye and measured using the Manual Find Component procedure in MassLynx. When the components present were known or identified, the data in the m/z range containing the peak series was then processed using the Maxent1 procedure over an appropriate mass range with suitable peak half-width settings determined from the raw data.
The resultant deconvoluted data is shown in the Supplementary Data section below with corrected masses noted for the major species. The secondary peak seen at +178.6 Da ± 2.49 Da in each spectra is likely to be α-N-6-phosphogluconoylation of the His-tag as previously seen in recombinantly expressed proteins in E. coli.

Ellman's reagent assay
Estimation of thiol concentration in samples of unmodified and modified BtGH84_TM and BtGH84_QM was carried out using Ellman's reagent (5,5'-dithiobis (2-nitrobenzoic acid) or DTNB) which reacts with free thiols to release the coloured 2-nitro-5-thiobenzoic acid (TNB) anion [1]. This quantification was performed by combining 40 µl protein sample (150 µM), with 33 µl DTNB (10 mM in 0.1 M Na 2 HPO 4 , pH 8.0) and 927 µl of denaturing buffer (6 M guanidinium chloride in 0.1 M Na 2 HPO 4 , pH 8.0). Absorbance at 412 nm was recorded using a Cary 100 UV-Vis spectrophotometer. Values were collected for the denaturing buffer alone, following the addition of DTNB and finally after addition of the protein sample. The concentration of thiol present was calculated using the molar absorbance of the TNB anion (E 412 TNB 2-= 1.37 x 10 4 M -1 cm -1 ) from the background corrected values for absorbance. This was compared to the protein concentration calculated using absorbance at 280 nm to determine the ratio of labelled protein. Unlabelled denatured protein contains two thiols per protein molecule whereas fully labelled denatured protein contains a single thiol per protein molecule.

Activity assay with 4MU-GlcNAc
Measurements of enzyme activity were performed using a fluorescent kinetic assay. Assays were carried out in a 96-well plate (Black Nunc F96 MicroWell Plates) at 25°C in a total volume of 100 μl per well. Each well contained 90 µl buffer (50 mM MES pH 6.5, 200 mM NaCl, 0.01%  and substrate 4-methylumbelliferyl-β-D-GlcNAc (4MU-GlcNAc) across a 2-fold serial dilution range from 4 mM -3.91 µM. Reactions were initiated by addition of 10 μl of protein, to a final concentration of 50 nM, followed by 5 s of shaking. Fluorescent 4-methylumbelliferyl release was recorded continuously at 355 nm excitation and 460 nm emission for 10 s using a BMG PolarStar Optima plate reader. Initial velocity rates were calculated using Sigmaplot as the slope of the curve of the collected data. Corrections were made to the detected gradients to adjust for the inner filter Additional experiments (results in Supplementary Figure 10) were performed to demonstrate that the enzyme kinetics is not dependent on protein concentration.

ITC methods
Isothermal titration calorimetry (ITC) measurements were carried out using a MicroCal Auto-iTC200 (Malvern Instruments) at 25°C. BtGH84 was prepared in filtered and degassed 50 mM MES, pH 6.5, 200 mM NaCl, and concentrated to an appropriate range. PUGNAc and Thiamet G were prepared in the same buffer. Samples were centrifuged and degassed prior to use. Using Microcal PEAQ-ITC Analysis software (v1.1.0.1262) the enthalpy for each injection was plotted against the molar ratio and fitted to a bimolecular model. The stoichiometry (N), enthalpy (ΔH), and association constant (Ka) were obtained from this fitting, and were used to calculate the Gibbs free energy (ΔG) and entropy (TΔS).

Tables
Supplementary Table 1: Data collection and refinement statistics (molecular replacement) for structures of   Note that the date in Supplementary Table 4 are the mean and standard deviation of 3 repeats. This gives slightly different values to those shown in Figure 2b of the main text which is for a single titration (a) Representative Michaelis-Menten plots for BtGH84_TM-fragment conjugates obtained using the 4MU-GlcNAc substrate as described in Methods. Final kinetic parameters can be found in Table 1.
(b) Representative Michaelis-Menten plots for BtGH84_QM-fragment conjugates obtained using the 4MU-GlcNAc substrate as described in Methods. Final kinetic parameters can be found in Supplementary Table 3.
(a-g) Representative ITC data for PUGNAc binding to BtGH84_TM-fragment conjugates, panels a-g correspond to modifications 4-10 respectively. (h-n) Representative ITC data for thiamet-G binding to BtGH84_TM-fragment conjugates, panels h-n correspond to modifications 4-10 respectively.

Supplementary Figure 10: Activity is not protein concentration dependent
(a) Michaelis-Menten plots normalised to protein concentration for wild-type BtGH84 at 12.5, 25, 50, and 100 nM shown with green, orange, purple and pink circles respectively. (b) Fold activation of two BtGH84_TM activator conjugates with enzyme concentration at 12.5, 25, 50, and 100 nM shown in green, orange, purple and pink respectively.

Intact protein mass spectrometry of BtGH84_TM and covalently modified BtGH84_TM
See supplementary methods for information on sample preparation and data collection.

Synthesis and compound characterisation
Description of synthetic strategy -Schemes S1 -S5 To generate our covalent activators, we developed a synthetic approach using the Curtius rearrangement reaction to construct the 2-amino-quinazoline structure. The synthetic route started from the commercially available isatoic anhydride, which was first amidated to the corresponding amide S1 with NH 3 in MeOH and then cyclised with diethyl oxalate to give quinazolin-4-one S2.
Then, the mixture was diluted with hexane (10 mL

4-Morpholinothieno[2,3-d]pyrimidin-2-amine S9
TFA (2 mL) was added dropwise to a stirred solution of carbamate S8 (134 mg, 0.4 mmol, 1.0 eq.) in CH 2 Cl 2 (4 mL) at 0 °C. The resulting solution was stirred at rt for 1 h and the solvent was removed under reduced pressure. NH 4 OH (aq) (5 mL) was added and the mixture was extracted with CH 2 Cl 2 (3 × 10 mL). The combined organic layers were dried (MgSO 4 ) and evaporated under reduced pressure to give the crude product. Purification by flash column chromatography on silica with 100