Developing deprotectase biocatalysts for synthesis

Organic synthesis often requires multiple steps where a functional group (FG) is concealed from reaction by a protecting group (PG). Common PGs include N-carbobenzyloxy (Cbz or Z) of amines and tert-butyloxycarbonyl (OtBu) of acids. An essential step is the removal of the PG, but this often requires excess reagents, extensive time and can have low % yield. An overarching goal of biocatalysis is to use “green” or “enzymatic” methods to catalyse chemical transformations. One under-utilised approach is the use of “deprotectase” biocatalysts to selectively remove PGs from various organic substrates. The advantage of this methodology is the exquisite selectivity of the biocatalyst to only act on its target, leaving other FGs and PGs untouched. A number of deprotectase biocatalysts have been reported but they are not commonly used in mainstream synthetic routes. This study describes the construction of a cascade to deprotect doubly-protected amino acids. The well known Bacillus BS2 esterase was used to remove the OtBu PG from various amino acid substrates. The more obscure Sphingomonas Cbz-ase (amidohydrolase) was screened with a range of N-Cbz-modified amino acid substrates. We then combined both the BS2 and Cbz-ase together for a 1 pot, 2 step deprotection of the model substrate CBz-l-Phe OtBu to produce the free l-Phe. We also provide some insight into the residues involved in substrate recognition and catalysis using docked ligands in the crystal structure of BS2. Similarly, a structural model of the Cbz-ase identifies a potential di-metal binding site and reveals conserved active site residues. This new biocatalytic cascade should be further explored for its application in chemical synthesis.


Experimental Materials
All chemicals and solvents were purchased from Sigma Aldrich or Fisher and used as received, without further purification.Plasmids were ordered from Genscript.E. coli competent cells were purchased from New England Biolabs.

NMR spectroscopy
1 H and 13 C NMR spectra were recorded in deuterated chloroform (CDCl3), on a 500 MHz Bruker Spectrometer.The spectra have been referenced with the appropriate residual solvent peaks (CDCl3 7.26 ppm) and the coupling constants are reported to the nearest 0.1 Hz.Chemical shifts of NMR spectra are reported in parts per million (ppm) on the δ scale.Data are reported in the following way for 1 H NMR spectra: chemical shift, integration, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, and m = multiplet) and coupling constant (J) in Hertz (Hz).

Mass spectrometry
Small Molecule Liquid Chromatography-Mass Spectrometry (LC-MS/ESI) was conducted on a Bruker microTOF mass spectrometer with an electrospray ionisation source.
If required, the protein was further purified using size exclusion chromatography prior to freezing using a Superdex HiLoad 16/60 S200 column in storage buffer (sodium phosphate (pH 7.4, 50 mM), NaCl (300 mM), 10% glycerol).Fractions containing the protein were collected, concentrated and frozen at at -80 °C.

Substrate screening Cbzase
Stocks of substrates were prepared in reaction buffer to a final concentration of 20 mM.Those substrates that are less soluble in water were dissolved in 20% DMSO in buffer to a final concentration of 20 mM (Z-Tyr, Z-Trp, Z-Glu(O t Bu)).The Cbz-protected substrate (250 µL, 20 mM) and Cbzase (295 µL, 1.69 mg/mL) were added to an Eppendorf (1.5 mL).Reaction buffer (455 µL, sodium phosphate 50 mM, pH 7.5) was added to reach a final volume of 1 mL.The reactions were left at 37 °C, 250 rpm for 24 hours.A sample (500 µL) of the reaction was taken and trifluoroacetic acid (TFA, 10 µL, 10% in water) was added before centrifuging on a benchtop centrifuge (13000 xg, 10 min).The supernatant (250 µL) was diluted into water (750 µL) and analysed by reverse-phase HPLC using the method described above.

Hydrolysis of L-Phe-O t Bu by BS2
Reactions were prepared in triplicate.H-Phe-O t Bu (200 µL, 100 mM) and BS2 (125 µL, 7.1 mg/mL) were added to an Eppendorf tube (1.5 mL) and the volume was made up to 1 mL with reaction buffer (sodium phosphate 50 mM, pH 7.5).Control reactions were also prepared in the same way without addition of BS2.The reactions were placed at 37 °C, 250 rpm for 24 hours.The reaction was centrifuged on a benchtop centrifuge (10,000 rpm, 10 min) and the supernatant (200 µL) was removed, diluted to 2 mL with HPLC grade water and analysed by reverse-phase HPLC using the method described above.
Anhydrous dichloromethane (DCM, 4 mL) was added and the solution was stirred to dissolve the solid reagents at 0 °C.N, N′-Dicyclohexylcarbodiimide (DCC, 1 g, 4.85 mmol, 2.9 eq.) was added to a dry beaker and dissolved in anhydrous DCM (3 mL).The DCC solution was cooled in an ice bath before being slowly transferred to the round-bottom flask.The mixture was stirred in the ice bath for 1 hour, then continued stirring at room temperature overnight.The mixture was filtered and concentrated under reduced pressure giving a colourless oil.The oil was dissolved in diethyl ether (30 mL) and transferred to a separatory funnel.The solution was washed with (i) HCl (0.5 M, 3 x 10 mL), (ii) NaHCO3 (5% aq., 3 x 10 mL) and (iii) NaCl (30% aq.).The organic layer was dried over MgSO4, filtered and concentrated under reduced pressure.The crude product was purified by column chromatography (hexane/ethyl acetate 9:1).

Deprotection cascade of Z-L-Phe-O t Bu
A 10 mM stock solution of Z-L-Phe-O t Bu was prepared in acetonitrile.A solution of purified BS2 (900 µL, 2.8 mg/mL) in sodium phosphate buffer (50 mM, pH 7.5) was prepared.To this, Z-L-Phe-O t Bu (100 µL, 10 mM) was added in 4 x 25 µL volumes over 4 hours at 37 °C.This was done to prevent precipitation of Z-Phe-O t Bu.A control reaction was also prepared in the same way but without the addition of BS2.The reaction was placed at 37 °C, 500 rpm for 24 hours.A sample of this reaction was taken (200 µL), diluted with acetonitrile (100 µL) to ensure dissolution of reaction components, centrifuged on a benchtop centrifuge (10,000 rpm, 5 min) and analysed by RP-HPLC.Cbz-ase (400 µL, 1 mg/ml) was added to the remainder of the reaction (800 µL) and left to react at 37 °C, 500 rpm for 6 hours and analysed in the same way.

Design of Phylogenetic Trees BS2:
A small database of 200 proteins with sequence identity to BS2 ranging from 39-99% for PDB/Uniprot, RefSeq and NR database with Blastp was generated.Sequence alignment was performed via Mega7 using ClustalW algorythim (Pairwise alignment: Gap opening penalty = 10, Gap extension penalty = 0.10; Multiple Gap opening penalty = 10, Gap extension penalty = 0.20).A phylogenetic tree of aligned proteins was generated with Mega7 using Neighbor-joining tree with bootstrap method.Phyla of the 3 bacteria species expressing the 134 proteins were affirmed via DSMZ: BacDive database.The generated phylogenic tree was designed with ITOL v6 (Fig. S8A).

Cbz-ase:
A small database of 200 proteins with sequence identity to Cbz-ase ranging from 25-40% for PDB/Uniprot databases and >85% for NR database with Blastp was generated.Sequence alignment was performed via Mega7 using ClustalW algorythim (Pairwise alignment: Gap opening penalty = 10, Gap extension penalty = 0.10; Multiple Gap opening penalty = 10, Gap extension penalty = 0.20).A phylogenetic tree of 36 aligned proteins was generated with Mega7 using Neighbor-joining tree with bootstrap method.Phyla of the 5 bacteria species expressing the 36 proteins were affirmed via DSMZ: BacDive database.The generated phylogenic tree was further designed with ITOL v6 to achieve the final figure (Fig. S8B).

HPLC calibration curves
Figure S16.HPLC calibration curve for L-Phe (9.6 min) using HPLC method described on page 2.

Figure S8B .
Figure S8B.Phylogenetic analysis of CBZase representing a small variety of bacteria endogenously expressing CBZase-alike enzymes.CBZase from (insert actual strain) shows highest of sequence identity of amidohydrolases from Sphingomonadaceae.

Figure S9 .
Figure S9.The predicted binding pocket of BS2 esterase, with a computed surface area and volume of 830.4 Å 2 and 695.8 Å 3 , respectively.Topological analysis was performed using the CASTp 3.0 server.11

Figure
Figure S12.A) A composite homotetrameric model of Cbz-ase, showing the relative organisation of the satellite and catalytic domains of each protomer.B) The Cbz-ase homotetramer superimposed on the crystal structure of an amidohydrolase homologue from Staphylococcus aureus (PDB:4EWT).12

Figure S13 .
Figure S13.A graphical summary of the Zn 2+ -docked Cbz-ase MDS.A) The radius of gyration (Rg) of the Zn 2+ -docked Cbz-ase over time.B) A 1D RMSD plot (relative to t = 0 ns) of the Zn 2+ -docked Cbzase over time.C) A pairwise (2D) RMSD map computed between Cbz-ase C atoms at every time point of the MDS.D) A pairwise (2D) RMSD map computed between the Zn 2+ ions at every time point of the MDS.E) Cbz-ase morphology over time; frames were extracted at t = 0, 2.5, 5, 7.5 and 10 ns.The docked Zn 2+ ions are depicted as blue-grey spheres.

Figure S29 .Figure S30 .Figure S31 .
Figure S29.HPLC chromatogram for the second step of the deprotection cascade in which Cbz-ase hydrolyses the Cbz group of Z-Phe to yield the free L-Phe and benzyl alcohol.(Note: dilution occurs between steps due to addition of Cbz-ase).Retention times; L-Phe = 9.6 min, benzyl alcohol = 14.2 min, Z-Phe = 20.1 min, Z-Phe-O t Bu = 24.4min.

Figure S32 .
Figure S32.Size exclusion chromatogram for the Cbz-ase purification using a Superdex HiLoad 16/60 S200 column.Peak elution of 67 mL estimates the structure to be a tetrameric protein.