PMP–diketopiperazine adducts form at the active site of a PLP dependent enzyme involved in formycin biosynthesis

ForI is a PLP-dependent enzyme from the biosynthetic pathway of the C-nucleoside antibiotic formycin.


Supplementary figures
, the reaction was carried out in D 2 O.The resulting mass spectrum shows the presence of an ion (m/z = 431) corresponding to the anion of the diketopiperazine but no incorporation of deuterium was observed, although the experiment was set up in D 2 O.

General methods
Unless specified, all chemicals were purchased from Sigma-Aldrich or Fisher Scientific.
Oligonucleotides were purchased from ThermoFisher Scientific. Restriction enzymes were purchased from New England Biolab. dNTP and DNA polymerase were purchased as part of the EMD Millipore Novagen KOD Hot Start DNA Polymerase kit. DNA sequencing was performed by GATC.

Cloning and expression
Streptomyces kaniharaensis Shomura and Niida (ATCC® 21070™) was purchased from ATCC (Middlesex, UK). The freeze-dried cell sample was directly dissolved in autoclaved Tryptic Soy Broth media containing 25% glycerol in biological safety cabinet. 10 l of the cell culture was plated on the TSB agar and incubate at 26 °C for 15 days. Four 0.5-inch diameter mycelia appeared on the plate, and one of the mycelia was picked for inoculating 50 mL TSB liquid media. The cell culture was incubated at 26 °C for 5 days before harvesting by centrifugation at 3500 rpm at 4 °C. The cell pellets were washed by deionized water twice before store in -80 °C overnight. The genomic DNA extraction was performed using PureLink™ Genomic DNA Mini Kit (ThermoFisher, UK) and the manufacture protocol for Gram-positive bacteria. To increase the purity, the yield genomic DNA was purified again using the same procedure. This yielded 17 g genomic DNA (260nm/280nm = 1.9, 260nm/230nm = 2.0) for sequencing. Preparation of PacBio DNA library from the genomic DNA sample, data generation on RSII SMRT cell, and genome assembly using HGAP were performed at University of Liverpool Centre for Genomic Research. The final assembly of the genome of S. kaniharaensis contained 19 contigs. The genome was annotated through the PATRIC software pipeline (https://www.patricbrc.org/). 1 The final assembly of the genome of S. kaniharaensis contained 19 contigs. The genome sequence has been deposited with accession number is: SAMN12859417. We tentatively identified the gene cluster encoding by searching for the homologs of the known formycin synthesis protein ForH (PDB id: 6NKO) using the internal PATRIC BlastP tool and then analyzing the surrounding CDSs for homologs of previously identified forA-X genes. 2  The plasmid has been deposited with ADDGENE. This construct has an N terminal His 6 tag followed by a tobacco etch virus cleavage site before the start of the protein. The resulting protein was expressed in Escherichia coli BL21(DE3) cells grown in the autoinduction media described by Studier 4 for 48 h at 20 °C. and purity of ForI was confirmed by SDS gel electrophoresis ( Figure S7) and MS. 50 μM external PLP is added in the protein before it was fast frozen by liquid nitrogen.

UV -vis absorbance spectroscopy of ForI.
All UV absorbance spectra were recorded on a SpectraMax 2e microplate reader (Molecular Devices) and analysed using Graphpad Prism 6. An additional 1 μM PLP was added to enzyme to aid protein stability. For UV -vis absorbance assays, the concentration of recombinant protein was 50 μM. The plate reader was blanked with 100 mM HEPES (pH 7.5) and spectra were collected from 300 nm to 500 nm. The screening experiments used a total sample volume of 100 μl. The instrument was set to record spectra at a wavelength step of 2 nm. Changes in the spectrum were monitored after addition of 0.5 mM L-glutamate or 5 mM cycloserine (D/L). Index detector. The protein elution peak was characterised by the differential refractive index (dRI).

Crystallography
The enzyme was screened for suitable crystallisation conditions. 1 mM PLP was added into protein The data sets were collected at two different beam lines (I04 and I24) and all data processing used the Diamond online automated software XIA2 6 DIALS 7 . The resolution limit of data is determined where CC 1/2 ≥ 0.5 8 and intensity fall off. Structures were determined using molecular replacement in PHASER 9 (the native ForI used the GSAAT as search model, subsequent and refined in REFMAC5 10 with anisotropic B-factors as implemented in CCP4. 11 Ligands were introduced into density during refinement when the Fo-Fc map was judged unambiguous in COOT. 12 The maps that we introduced the ligands into are the same omit maps that we showed in the paper. Thus the phases for these Fo-Fc maps were derived from models that never had the ligand present. The final refined coordinates of the ligand were then shown in this Fo-Fc map. This method of calculating the omit maps is unbiased. The geometry of the diketopiperazine was restrained guided by parameters from the PRODRG server. 13 Data and structures have been deposited with the RCSB. spectrometer was operated in electrospray ionisation positive or negative mode. Negative ion mass spectrometric analysis of both ForI/cycloserine reaction mixtures clearly showed ions with m/z = 431 corresponding to the anion of the corresponding diketopiperazine distereoisomer ( Figure S8).

Mass spectrometric detection of diketopiperazines
This peak was absent in the corresponding mass spectra of the control samples of protein nor was any peak corresponding to diketopiperazine observed.
For deuterium experiment, ForI was exchanged into 100 mM NH 4 OAc, pH 8.0 in D 2 O. 2 mM Lcycloserine (in D 2 O) and 100M ForI dissolved in 100 mM NH 4 OAc, pH 8.0, (50 L total volume) were then incubated overnight at 4 °C before the addition of MeOH (50 L) to precipitate the enzyme. Centrifugation then gave a supernatant that was analysed by mass spectrometry.