Understanding Substrate Selectivity of Phoslactomycin Polyketide Synthase by Using Reconstituted in Vitro Systems

Abstract Polyketide synthases (PKSs) use simple extender units to synthesize complex natural products. A fundamental question is how different extender units are site‐specifically incorporated into the growing polyketide. Here we established phoslactomycin (Pn) PKS, which incorporates malonyl‐ and ethylmalonyl‐CoA, as an in vitro model to study substrate specificity. We combined up to six Pn PKS modules with different termination sites for the controlled release of tetra‐, penta‐ and hexaketides, and challenged these systems with up to seven different extender units in competitive assays to test for the specificity of Pn modules. While malonyl‐CoA modules of Pn PKS exclusively accept their natural substrate, the ethylmalonyl‐CoA module PnC tolerates different α‐substituted derivatives, but discriminates against malonyl‐CoA. We show that the ratio of extender transacylation to hydrolysis controls incorporation in PnC, thus explaining site‐specific selectivity and promiscuity in the natural context of Pn PKS.


Figure S3
: Extracted ion chromatograms from assays with PnAV4 and PnB (tetraketide system), PnAV4, PnB, PnC-TEDEBS (pentaketide system) and PnAV4, PnB, PnC and PnD-TEDEBS (hexaketide system). Products of the tetraketide and pentaketide system are three times labelled, upon incorporation of 2-C 13 labelled malonyl-CoA, products of the hexaketide system four times. The tetraketide product (4) Figure S6). For the hexaketide one mass corresponding to the oxidized, dehydrated (6b.5, Figure S6) ion is shown as an example (m/z [M+H] + 303.1955). Figure S4: Mass spectrometric data of pentaketide production. A) Overlaid extracted ion chromatograms of 5a, 5b, 5c, 5d and 5e from single competition assays. B) All assays were run with either malonyl-CoA (red) or 2-C 13 malonyl-CoA (black). This resulted in double labeled pentaketide (PnB module 1 and module 2 incorporate malonyl-CoA). Searched for and shown are the masses of 5a-e. 2 A shift of retention time can be observed for all substrates, except methylmalonyl-PnC_ACP, which elutes at the same retention time as holo-PnC_ACP. Substrates used were malonyl-(2), methylmalonyl-(3a), ethylmalonyl-(3b), butylmalonyl-CoA (3c) 3-methylbutylmalonyl-(3c), hexylmalonyl-(3e) and benzylmalonyl-CoA (3f). Figure S9: HPLC UV-vis analysis for PnC_KS-AT transacylation rate. The three colors represent three technical replicates. Slopes from the linear increase were used for the calculation of kcat. Relative amount of acyl-ACP is given on the y-axis. Substrates used are A) ethylmalonyl-CoA, B) malonyl-CoA, C) butylmalonyl-CoA, D) 3-methylbutylmalonyl-CoA, E) hexylmalonyl-CoA. For ethylmalonyl-ACP and butylmalonyl-ACP a decrease of acyl-ACP can be observed after 25 min and after 150 min, respectively. Note also the maximum conversion reached in the transacylation assay varies between approx. 70% for ethylmalonyl-, butylmalonyl-, 3-methylbutylmalonyl-ACP and approx. 40% for malonyl-ACP. Transacylation rate could not be determined for methylmalonyl-CoA, due to same retention time of holo-PnC_ACP and methylmalonyl-PnC_ACP (see also Figure S7). Figure S10: HPLC based transacylation assay to confirm results of SucC/SucD assay. For the measurement of the kcat of transacylation by PnC_KS-AT the SucC/SucD based assay could not be used due to high background activity, which could also not be reduced by repeated size exclusion chromatography of the used proteins. To assure that the results from the SucC/SucD based assay and the HPLC assay are comparable the control described here was done. Turnover of A) PnB_KS-AT2 (substrate malonyl-CoA) and B) PnC_KS-AT (substrates used were ethylmalonyl-and butylmalonyl-CoA). 500 nM KS-AT and 400 µM cognate ACP were mixed with 600 µM substrate in 200 mM NaH2PO4, pH 7.2 (same protein concentrations in SucC/SucD based assay). The reaction was stopped by adding formic acid and samples were analyzed using HPLC. With the linear slope the turnover was calculated. The kcat of PnB_KS-AT2 with malonyl-CoA was 49.3 per minute and for PnC_KS-AT with ethylmalonyl-CoA 3.4 per minute, with butylmalonyl-CoA 0.31 per minute. With the SucC/SucD based assay a kcat of 74 (PnB_KS-AT2), 4 (PnC_KS-AT with ethylmalonyl-CoA) and 0.6 (PnC_KS-AT with butylmalonyl-CoA) was measured. The determined turnover rates are similar which assures that the measured kinetics displayed in Table S3 and Table 1 can be compared. Scheme S1: Steady state kinetic assay, relying on purified proteins. KS-AT didomains and standalone ACPs are used. Upon covalent binding of the acyl-residue to the AT, CoA-SH is released. With no ACP acceptor available, the acyl-residue is subsequently hydrolyzed by nucleophilic attack of water and the AT can bind a new acyl-residue. If holo-ACP is available, the acyl-residue is transferred and the AT can bind a new acyl-residue. Free CoA-SH is converted to succinyl-CoA by E. coli SucC/SucD. In this reaction ADP is produced, which can be used by pyruvate kinase/lactic dehydrogenase (PK/LDH) under NADH consumption. For Michaelis Menten plots with data collected with this assay refer to Figure S6. Table S1: phosphopantetheinyl coverage, selfacylation and transacylation of Pn_ACPs. Extent of phosphopantetheinlyation of Pn_ACP purified from E. coli BAP1. Acyl-ACP buildup upon 2 h incubation of 100 µM holo-ACP with 300 µM extender unit at 25°C ("selfacylation") and acyl-ACP buildup upon 2 h incubation of 5 µM KS-AT, 300 µM CoA thioester, 100 µM ACP at 25°C. The percentage of apo-, holo-and acyl-ACP was determined by mass spectrometric measurements. Substrates used were malonyl-(2), methylmalonyl-(3a), ethylmalonyl-(3b), butylmalonyl-CoA (3c) 3-methylbutylmalonyl-(3c), hexylmalonyl-(3e) and benzylmalonyl-CoA (3f). Malonyl-, methylmalonyl-and 3-methylmalonyl-ACP can be detected without the presence of an AT domain. This is likely due to transesterification of the acyl-residue from CoA to the thiolgroup of the ACP. In case of malonyl-and methylmalonyl-CoA this is also partly due to sfp mediated loading of the ACP during expression in E. coli BAP1. Subtracting already acylated ACPs ("selfacylation") from the transacylated ACP produced by the AT-mediated transacylation shows that only malonyl-CoA 2 is transferred in the case of PnB_AT1, PnB_AT2 and PnD_AT and that all substrates (BenzM-ACP under detection limit) are transferred in the case of PnC_AT.

Experimental section
Phoslactomycin PKS in vitro assays. The production of phoslactomycin polyketide derivatives was initiated using PnAV4 and CHC-CoA (1), or PnB and (2Z)-cyclohexanepropenyl-SNAC (7). The assay was run at a volume of 50 µL and contained the PKS proteins at 9 µM concentration, Npt at 3.5 µM and all substrates at 1 mM concentration (in the case of strategy 3 competition assays, the substrate concentration of α-substituted acyl-CoA derivatives was reduced to 0.5 mM to prevent inhibition by high substrate concentrations). The assays additionally contained 5 mM NADPH, 0.1 mM CoA, 5 mM MgCl2 and 100 mM NaH2PO4, pH 7.2. The assays were prepared on ice and 10 µL sample was mixed with 1:1 MeOH and stored at -80 °C as the negative sample. The assay was run at 25°C overnight, stopped with 1:1 MeOH and stored at -80 °C until analysis. Epimerase, to yield in the production of (2S)-acyl-malonyl-CoA derivatives, was added but did not have an impact on product spectrum and was therefore left out. UPLC-high resolution MS analysis was carried out using an Agilent 6550 iFunnel Q-TOF LC-MS system equipped with an electrospray ionization source set to positive ionization mode. The analyte was separated on a RP-18 column (50 mm x 2.1 mm, particle size 1.7 µm, Kinetex EVO C18, Phenomenex) using a mobile phase system comprised 0.1% formic acid in water (Solvent A) and acetonitrile (Solvent B). Chromatographic separation was carried out using the following gradient condition at a flow rate of 250 µL/min: 0 min 5% B; 1 min 5% B, 6 min 95% B; 6.5 min 95% B; 7 min 5% B. The column oven was set to 40 °C and auto sampler was maintained at 8 °C. Standard injection volume was 10 µL. Capillary voltage was set at 3.5 kV and nitrogen gas was used as nebulizing (20 psig), drying (13 L/min, 225 °C) and sheath gas (12 L/min, 40 °C). MS data were acquired with a scan range of 50-1200 m/z. LC-MS data were analyzed using MassHunter Qualitative Analysis software (Agilent). Mass spectrometric measurements of standalone ACPs. Loading of holo-and apo-ACPs by KS-AT and Npt was done as described above. All reactions were run at 25 °C for 2 h and placed on 4 °C until analysis by mass spectrometry. 2 µL of the buffered protein solutions were desalted online using a Waters ACQUITY H-Class HPLC-system equipped with a MassPrep column (Waters). Desalted proteins were eluted into the ESI source of a Synapt G2Si mass spectrometer (Waters) by the following gradient of buffer A (water with 0.05% formic acid) and buffer B (acetonitrile with 0.045% formic acid) at a column temperature of 60 °C and a flow rate of 0.1 mL/min: Isocratic elution with 5% A for two minutes, followed by a linear gradient to 95% B within 8 minutes and holding 95% B for additional 4 minutes. Positive ions within the mass range of 500-5000 m/z were detected. Glu-Fibrinopeptide B was measured every 45 s for automatic mass drift correction. Averaged spectra were deconvoluted after baseline subtraction and eventually smoothing using MassLynx instrument software with MaxEnt1 extension.
Synthesis of acyl-CoAs. Acyl-CoA synthesis was done like previously described. CHC-CoA was synthesized by chemical CDI coupling of the free acid, analogous as previously explained 6 .
Construction of plasmids used in this study. The coding sequence for the phoslactomycin polyketide synthase genes from Streptomyces platensis was synthesized and codon optimized by the Joint Genome Institute. Inserts were amplified by PCR. Oligonucleotides with overhangs for the restriction enzymes NdeI (CATATG) and HindIII (AAGCTT) were used and the agarose gel purified fragments were ligated into the cloning vector pJET1.2. From there the insert was cut out with NdeI and HindIII and inserted into expression vectors. SucC (Accession number WP_113400154) and SucD (Accession number WP_096861694) coding sequence was amplified by colony-PCR from E.coli strain Top10 and cloned following the procedure above. Npt coding sequence was codon optimized for E.coli and synthesized. All cloning was done using the E.coli strain Top10. Catalytic knock out mutants were constructed using mismatching oligonucleotides, containing the DNA triplet for the desired amino acid. The plasmid was divided into two or three parts and amplified by PCR using Q5® High-Fidelity DNA polymerase. The purified fragments were assembled using the Gibson Assembly ® Master Mix purchased from New England Biolabs. Anion exchange. The eluate from the Nickel beads purification was diluted to 120 mL with Anion A (50 mM NaH2PO4, pH 7.5) and loaded onto a 5 mL HiTrap Q HP anion exchange 5 mL chromatography column, purchased from GE Healthcare Life Sciences, with a flow of 3 mL/min. A gradient to 100% Anion B (50 mM NaH2PO4, 500 mM NaCl pH 7.5) with a flow of 4 mL/min in 30 min was run and protein containing fractions were collected and concentrated using Amicon® Ultra Centrifugal Filters, purchased from Merck Millipore. Proteins were used immediately for assays or stored with 30% v/v glycerol in -80 C after shock freezing in liquid nitrogen.

Synthesis of (2Z)-cylcohexanepropenyl-SNAC
General Information. All non-aqueous reactions were carried out using flame-dried glassware under argon atmosphere. All solvents were distilled by rotary evaporation. Solvents for non-aqueous reactions were dried as follows prior to use: Tetrahydrofuran (THF) was dried with potassium hydroxide and subsequently distilled from Solvona®. Dichloromethane and triethylamine were distilled from calcium hydride. N-acetylcysteamin (HSNAC) was synthesised according to literature 7 . All other commercially obtained reagents were used as received. Reactions were monitored by thin layer chromatography (TLC) using Merck Silica Gel 60 F254-plates and visualised by fluorescence quenching under UV-light. In addition, TLC-plates were stained using a potassium permanganate stain. Chromatographic purification of products was performed on Merck Silica Gel 60 (230-400 mesh) unless otherwise noted using a forced flow of eluents. Concentration in vacuo was performed by rotary evaporation at 40 °C and appropriate pressure and by exposing to fine vacuum at room temperature if necessary. NMR spectra were recorded on a Bruker AV 300 MHz, AV III 500 MHz, AV III HD 500 MHz spectrometer at room temperature. Chemical shifts are reported in ppm with the solvent resonance as internal standard. Data are reported as follows: s = singlet, d = doublet, t = triplet, q = quartet, quint = quintet, m = multiplet. Mass spectra were recorded by the mass service department of the Philipps-Universität Marburg. HR-ESI mass spectra were acquired with an LTQ-FT mass spectrometer (Thermo Fischer Scientific). The resolution was set to 100 000. IR spectra were recorded on a Bruker IFS 200 spectrometer. The absorption bands are given in wave numbers (cm -1 ), intensities are reported as follows: s = strong, m = medium, w = weak, br = broad band. Melting points were determined on a Mettler Toledo MP70 using one end closed capillary tubes. Optical rotations were determined at 20 o with a Krüss P8000-T polarimeter.