Crystal Structure Based Mutagenesis of Cattleyene Synthase Leads to the Generation of Rearranged Polycyclic Diterpenes

Abstract The crystal structures of cattleyene synthase (apo‐CyS), and CyS complexed with geranylgeranyl pyrophosphate (GGPP) were solved. The CySC59A variant exhibited an increased production of cattleyene and other diterpenes with diverse skeletons. Its structure showed a widened active site cavity explaining the relaxed selectivity. Isotopic labeling experiments revealed a remarkable cyclization mechanism involving several skeletal rearrangements for one of the novel diterpenes.


Gene expression and protein purification of CyS
The nucleotide sequence encoding the CyS enzyme was codon-optimized and synthesized for expression in E. coli. The DNA fragment of cyS with 15-20 bp homologous arms was amplified by PCR with primer cyS-pET28a-F and cyS-pET28a-R (Table S1). The purified PCR product of cyS was fused to a C-terminal hexahistidine tag in pET28a vector through Gibson Assembly method and transformed into E. coli BL21 Fast-T1 (Vazyme Biotech Co., Ltd., Nanjing, China). The resulting plasmid was confirmed by sequencing.
Overexpression was performed using lysogeny broth (LB) medium with 30 μg/mL kanamycin at 37 °C, 220 rpm until an OD600 of 0.6 was reached, then medium was subsequently cooled to 18

S5
The X-ray diffraction data of apo-CyS and CyS-GGPP-Mg 2+ complex were collected at beamline BL02U1 at the Shanghai Synchrotron Radiation Facility (SSRF) with a wavelength of 0.97918 Å. 1 The diffraction data were processed and scaled with HKL-2000. 2 The crystal structure of apo-CyS was solved by molecular replacement using SvS as the searching model. 3 The initial phases of the CyS-GGPP-Mg 2+ complex structure was determined by molecular replacement, using the apo structure of CyS as a search model. Molecular replacement was performed with Phaser in PHENIX. 4,5 The structure was modified manually with Coot 6 and refined with PHENIX 7 .

Crystallization, data collection and structural elucidation of CyS C59A
The crystallization screening of CyS C59A (20 mg/mL) was performed in 24-well plates with the hangingdrop vapor-diffusion method at 16 °C, and crystals of CyS C59A were observed after 2 days in a screening kit Index 18 (Hampton Research) that contains 0.49 M Sodium phosphate monobasic monohydrate, 0.91 M Potassium phosphate dibasic, pH 6.9.
The X-ray diffraction data of CyS C59A was collected at at 100 K in a liquid nitrogen stream, using a HyPix 6000HE detector at XtaLAB Synergy with a wavelength of 1.54184 Å. The data sets were processed and scaled using CrysAlis Pro software. 8 The crystal structure of CyS C59A was solved by molecular replacement with Phaser using apo-CyS structure as the searching model. Molecular replacement was performed with Phaser in PHENIX. The structure was modified manually with Coot and refined with PHENIX.

Substrate modelling
Substrate modeling was performed using AutoDock Vina 1. 1. 2. 9 Before the docking, water molecules were removed from complex structure CyS-GGPP-Mg 2+ and manually remove the hydrocarbon chain of GGPP. AutoDockTools 1.5.6 was used to prepare the intermediates A-H, macromolecules CyS-GGPP-Mg 2+ complex, and grid map files. The default parameters were used to set the torsion constraints for intermediates A-H, and charges and hydrogen atoms were added to CyS-GGPP-Mg 2+ complex proteins. The search parameters were used with default values.

Construction of engineered E. coli heterologous systems
To improve the production of diterpenes, we reconstructed isopentenol utilization pathway (IUP) which can produce isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP) in E. coli ( Figure   S2). 10 The biosynthetic genes ecTHIM, mjIPK 11 and idi 12 were amplified by PCR and individually cloned into plasmid pCDFDuet-1 with the Gibson assembly method to produce IPP and DMAPP, and the GGPP synthase-encoding gene ptmT4 (Table S3) and cyS were cloned into plasmid pRSFDuet-1 with the same method.
The fermentation was continued at 37 °C, 220 rpm until an OD600 of 0.8 was reached, then the culture was cooled down to 18 °C. At 18 °C, IPTG (final concentration 0.5 mM) was added for induction and S6 isoprenol (final concentration 25 mM) was fed at the same time. The fermentation was continued for additional 72 h at 18 °C and analyzed by GC-MS.

Site-directed mutagenesis
The site-directed mutagenesis experiments of cyS were carried out by PCR using two overlapping primers (Table S1) containing the mutated sequences with pRSFDuet-ptmT4-cyS as the template, to generate a series of mutated plasmids. Subsequently, the parental vector is digested by DpnI and the mixture is transformed into E. coli Fast-T1. The resulting plasmids were confirmed by sequencing. These mutant plasmids were individually co-transformed into E. coli with pCDFDuet-ecTHIM-mjIPK-idi plasmid for subsequent fermentation and detection.

The GC-MS analysis and isolation of diterpenes
The fermentation broths were extracted with an equal volume of hexane, and the hexane extracts were dried under vacuum with a rotary evaporator. For GC-MS analysis, a small part of the hexane extracts was dissolved in hexane and subjected to GC-MS analysis. Samples (1 µL) were injected in splitless mode at 50 °C. After being held at 50 °C for 3 min, the oven temperature was raised to 300 °C with a rate of 14 °C/min, then held for another 3 min.
For isolation of diterpenes, large-scale fermentations were carried out as the same procedures above.
The crude extract was subjected to silica gel flash chromatography eluted with hexane to yield a series of fractions. Fractions containing diterpenes were monitored by TLC and purified by Semi-preparative HPLC eluted with 100% acetonitrile (0 -60 min) at a flow rate of 2.0 mL/min, under the UV detection at 210 nm.  (Table S5).

Structural elucidation of diterpenes
Compared to the 13 C NMR spectrum of 1, the 13 C NMR spectrum of 2 shows significantly changed resonances towards those attributed to C-1-C-8 and C-20 in 1 (Table S6). Those differences above suggest that 2 contains a double bond at C-6/C-7 instead of C-2/C-3 in 1, and contains a methyl group Compound 3 was identified as a variediene analogue that has been reported as the product of a promiscuous terpene synthase FgJ07623, 13 Compound 4 was identified as allokutznerene that has been S7 reported as the product of a diterpene synthase PmS from Allokutzneria albata. 14  (Table S5). Compared to the 13 C NMR spectrum of 4, the 13 C NMR spectrum of 5 shows different double bond chemical shifts (Table S6). Those differences above suggest that 5 only differs from 4 by containing a double bond at C-   Figure S5). The NOESY correlation from H-10 to H-6 shows that they are on the same side of the ring and the absolute configuration of C-2 is R; the NOESY correlations from H-3 to H-18 and H-1 establish the absolute configuration of C-3 is S; and the NOESY correlations from H-19 to H-20 establish the absolute configuration of C-7 is R ( Figure S5). Thus, compound 6 is identified as a new diterpene with a novel carbon skeleton.

Isotopic labeling experiments
The isotopic labeling experiments were performed with the substrates and enzymes as listed in Table S8.