University of Groningen Nonribosomal peptide synthetases Zwahlen, Reto Daniel

(1)

University of Groningen

Nonribosomal peptide synthetases

Zwahlen, Reto Daniel

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2018

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Zwahlen, R. D. (2018). Nonribosomal peptide synthetases: Engineering, characterization and biotechnological potential. University of Groningen.

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

CHAPTER V

An engineered

two component

nonribosomal peptide

synthetase (NRPS)

producing a novel

peptide-like compound in

Penicillium chrysogenum

Reto D. Zwahlen,1 Ciprian G. Crismaru,1 Fabiola Polli,2 Catalin M. Bunduc,3

Ulrike Muller,5 Remon Boer,5 Olaf Schouten,5 Roel A.L. Bovenberg,4,5 and Arnold J.M. Driessen1,6 1Molecular Microbiology, Groningen Biomolecular Sciences and

Biotechnology Institute, University of Groningen, Groningen, The Netherlands

2Biochemical Laboratory, Groningen Biomolecular Sciences and Biotechnology Institute, University of Groningen, Groningen, The Netherlands

3Molecular Microbiology, Institute for Molecules, Medicines and Systems, Vrije Universiteit Amsterdam, The Netherlands 4Synthetic Biology and Cell Engineering, Groningen

Biomolecular Sciences and Biotechnology Institute, University of Groningen, Groningen, The Netherlands

5DSM Biotechnology Centre, Delft, The Netherlands 6Kluyver Centre for Genomics of Industrial Fermentations,

Julianalaan 67, 2628BC Delft, The Netherlands manuscript in preparation

(3)

Abstract

Semi-synthetic penicillins such as ampicillin and amoxicillin are major antibiotics used today. For manufacturing, a semi-synthetic production process is employed involving fermentation, enzy-matic and chemical conversion processes. To make the synthesis of semi-synthetic b-lactams more sustainable, a fully fermen-tative production process is desired. As a first step towards the development of such a process, the native nonribosomal peptide synthetase (NRPS) L-d-(a-aminoadipyl)-L-cysteinyl- D-valine syn-thetase (ACVS) involved in b-lactam biosynthesis was engineered into a hybrid two component NRPS system, which uses dedicated inter-NRPS communicating domains (COM). This enabled the re-placement of the first aminoadipate specific module by a heterolo-gous hydroxyphenylglycine (hpg) specific module. The new hybrid NRPS system was expressed in a Penicillium chrysogenum strain lacking the entire penicillin biosynthesis cluster. Analysis of the engineered strains revealed the formation of a new product with the predicted mass of the expected hpgCV tripeptide, but the mol-ecule detected contained a thioester instead of the desired pep-tide bond. The employed two-component system revealed the potential of COM domains in engineering strategies for the devel-opment of hybrid NRPS system.

(4)

An engineer ed tw o c omponent nonribosomal pep tide s ynthetase ( NRPS ) pr

oducing a novel pep

tide-lik e c ompound in P enicillium chry sog enum

5 Introduction

Penicillium chrysogenum represents a valuable source of secondary

metab-olites and is a suitable host for pathway and strain engineering and refine-ment programs due to its metabolic versatility and long history of industrial production of penicillins, cephalosporins and statins [1–4]. Secondary me-tabolites are typically produced by genes which are clustered in the genome and can readily be recognized by the presence of dedicated sequence motifs for multi-modular enzymes [5–8]. This class comprises nonribosomal pep-tide synthetases (NRPS), polykepep-tide synthases (PKS), terpene synthetases (TRP) and fatty acid synthetases (FAS), which all are multi-modular enzymes, consisting of a series of catalytically distinct domains, responsible for mul-tiple reactions that lead to structurally diverse bioactive compounds [9–11]. Engineering efforts based on this multi-modular organization, however, have been cumbersome, despite significant progress in the resolution of partial structures and new approaches for engineering active sites [12–18]. An alter-native approach for NRPS engineering, involves the use of NRPS communi-cating domains (COM). NRPS frequently appear clustered in functionally dif-ferent, but distinct modular compositions, where interactions are facilitated via COM domains [19]. COM domains are short, roughly 20 amino acid long, helix finger-like structures guiding the association of multi-domain NRPS systems, leading to the synthesis of compounds such as teicoplanin, ramo-planin or enduracidin [9;20–21].

In order to develop more sustainable production methods for major classes of b-lactam antibiotics we here explored the engineering potential of a two component NRPS system. The proposed system relies on two distinct parts, consisting of one and two NRPS modules, respectively. Specifically, two NRPS modules specific of the L- and D-hpg from teicoplanin (Tcp) synthesis in Aspergillus teycomiceticus [22–23], serve as initiation components. Through COM domains (COMA/COMD), these modules were combined in Penicillium

chrysogenum with a truncated version of the L-d-(a-aminoadipyl)-L-

cysteinyl-D-valine synthetase (ACVS) in order to synthetize a tripeptide consisting of D-hydroxyphenylglycine (hpg), L-cysteine and D-valine (DLD-hpgCV) which may act as a peptide precursor towards the biosynthesis of amoxicillin. The results show the potential of using COM domains for the engineering of hy-brid NRPS systems, for the production of novel, β-lactam precursors.

(5)

120

5

Materials and methods

Penicillium chrysogenum strains, media, and culture conditions

Penicillium chrysogenum DS58912 (∆Pen-cluster) was kindly provided by

DSM Sinochem Pharmaceuticals Netherlands B.V. To obtain mycelium of

P. chrysogenum for transformation and DNA isolation, fresh spores were

inoculated into YGG medium according to [24]. For the fermentation anal-ysis spores were inoculated in secondary metabolites production medium (PPM), consisting of (in g/liter) glucose, 5.0; lactose, 36; urea 4,5; Na2SO4, 2,9; (NH4)2SO4, 1,1; K2HPO4, 4,8; KH2PO4, 5.2; supplemented with 10 ml of a trace element solution containing (in g/l): FeSO4·7H2O, 24.84; MgSO4·7H2O, 0.0125; EDTA, 31.25; C6H6Na2O7, 43.75; ZnSO4·7H2O, 2.5; CaCl2·2H2O, 1.6; MgSO4·H2O, 3.04; H3BO3, 0.0125; CuSO4·5H2O, 0.625; Na2MoO·2H2O, 0.0125; CoSO4·7H2O, 0.625. The racemic L- and D-hydroxyphenylglycine (L/D-hpg) (Sigma) was added to each culture to a final concentration of 50 µg/ml (300 µM). Solu-tions were adjusted to pH 6.5. The mycelium was grown in a shaking incuba-tor at 200 rpm for 7 days at 25 °C.

NRPS selection and setup

In order to create hpgCV using a COM domain based engineering approach, an initiation module, an elongation/ termination di-module as well as a pair of communicating NRPS linkers were required. Two targets were selected, Tcp9 and Tcp11 (Figure 1), serving as initiation modules. Both are derived from the teicoplanin biosynthetic cluster and have shown to be specific for D/L- hydroxyphenylglycine (Chapter II/Unpublished data). In addition, the dedicated MbtH-like protein (MLP) Tcp13 associated with Tcp9 and Tcp11 was chosen. Furthermore, two di-modules, activating L-cysteine and L- valine, were selected, representing the two C-terminal modules of the ACVS ( PcbAB) of P. chrysogenum (Pc) and Nocardia lactamdurans (Nl), respectively. Finally, the complementing NRPS communicating linkers, acceptor COMA and donor COMD of the Tcp9 and Tcp11 associated NRPS were chosen to enable specific AHPG – CV association. COMD was fused C-terminally to Tcp9 and Tcp11 (Tcp9-COMD/Tcp11-COMD) and COMA to the N-terminus of the CV di-module (Nl CV-COMA/CV-COMA). The setup is further illustrated in Fig-ure 1. Constructs were subsequently cloned, analyzed and transformed to

(6)

oducing a novel pep

5 NRPS expression, purification and in vitro activity evaluation

E. coli HM 0079 cultures were grown in 2xPY medium, to an OD600 of 0.6,

transferred to 18 °C at 200 rpm for 1 h and subsequently induced using 0.3 mM IPTG and 0.2 % L-arabinose. Harvest followed 18 h after induction by spinning at 3500 g for 15 minutes. After resuspension in lysis buffer (HEPES 50 mM pH7.0, NaCl 300 mM, DTT 2 mM, complete EDTA-free protease in-hibitor; Roche No. 04693159001), cells were disrupted using sonication (6 s/15 s; on/off, 50x, 10 µm amplitude) and cell-free lysate obtained by cen-trifugation at 4 °C, 13000 g for 15 minutes. Purified enzyme was extracted by means of Ni-NTA bead (Qiagen) supported his-tag affinity purification using gravity flow. Wash steps were performed using two column volumes wash buffer (HEPES 50 mM pH7.0, NaCl 300 mM, imidazole 20 mM) and a one-step elution using 5 bed volumes elution buffer (HEPES 50 mM pH7.0, NaCl 300 mM, imidazole 250 mM). Imidazole was removed, while simultaneously concentrating the sample with Ami100 spin filters (Amicon). Final con-centration was determined using A280 measurements or a colorimetric assay (DC protein assay, BioRad).

Isolated enzyme was subjected to in vitro assays, in order to determine product formation properties. Assay conditions included HEPES 50 mM pH 7.0, NaCl 300 mM, ATP 5 mM pH 7.0, CoA 100 µM, sfp (NEB) 0.2 µM, L-cysteine 2 mM, L-valine 2 mM, MgCl2 5 mM, DTT 2 mM, hybrid NRPS 0.3 µM and either L- or D-hydroxyphenylglycine (L/D-hpg) 5 mM. Negative controls included re-actions omitting enzyme or L/D-hpg. Full length N. lactamdurans ACVS served as positive control, using 5 mM L-α-aminoadipic acid instead of L/D-hpg. Reac-tions were run at 30 °C and sampling took place after 0, 60, 120 and 240 min-utes and reactions quenched using 0.1 M NaOH. Samples were subsequently stored at −80 °C and reduced with 10 mM DTT prior to LC/MS analysis.

P. chrysogenum transformation setup and strain selection

P. chrysogenum DS58912 [25] was co-transformed with different combinations

of linear DNA fragments, encoding expression constructs for the MbtH-like protein (MLP) Tcp13, an initiation module (Tcp9/11), a CV di- module (Pc/NlCV), a modified version of the pcbC gene, encoding an evolved isopenicllin- N-synthase variant (Unpublished results) and the selection marker AmdS, ac-cording to Supplementary table 1. Penicillium transformants were selected on regeneration plates containing 0.1 % acetamide as sole nitrogen source. To verify for the presence of the different linear fragments integrated into

(7)

122

5

Materials and methods

the P. chrysogenum genome, genomic DNA (gDNA) was isolated after 48 h of growth in YGG medium using a modified yeast genomic DNA isolation pro-tocol [26]. Fungal mycelium was broken in a FastPrep FP120 system (Qbio-gene, Carlsbad, CA, USA). Diagnostic primers used to check for the presence of

Tcp9/11, Tcp13, amdS and Nl/PcCV are listed in Supplementary table 2.

Small scale fermentation and sampling

The parental P. chrysogenum DS58912 and selected transformants were grown in duplicates using PPM in a volume of 25 ml. Samples were taken as indicated and centrifuged at 14000 g for 5 minutes. Supernatant was fil-tered using 0.2 µm syringe filters with polypropylene Housing (VWR Inter-national Ltd.) and 60 µl of supernatant was transferred to an auto sampler vial. For separation, Accella 1250TM HPLC system coupled in-line to an ES-MS Orbitrap ExactiveTM (Thermo Fisher Scientific, San Jose, CA) was used. Scan range between m/z 80 and m/z 1600 in positive Ion (4.2 kV spray, 87.5 V cap-illary and 120 V of tube lens) mode, with capcap-illary temperature set at 325 °C was used. Separation was performed on a Shim-Pack XR-ODSTM c18 column (3.0 × 75 mm, 2.2 µm) (Shimadzu, Kyoto, Japan) at 40 °C. A linear gradient was used for the elution. A gradient program with miliQ water (A), acetoni-trile (B) and 2 % formic acid (D) was run; 0 min; A 90 %, B 5 %, C 5 %; 4 min, A 90 %, B 5 %, C 5 %; 13 min, A 0 %, B 95 %, C 5 %; 16 min A 0 %, B 95 %, C 5 %; 16 min, A 90 %, B 5 %, C 5 %; 20 min A 90 %, B 5 %, C 5 % at a flow rate of 0.3 ml min-1. Standards were used to identify peaks according to retention time and accurate mass (amoxicillin, hpgCV). A DLD-hpgCV standard curve (0,5; 1; 5; 10; 25; 50 µM) ensured accurate quantification of the produced compound levels (r>0.98).

Compound isolation

Spores of the P. chrysogenum transformant T1C and parental strain DSDS68530 were used to inoculate shake flask fermentations containing 75 ml of PPM medium in a 200 ml flask at 200 rpm and 25 °C. After 48 hours inoculum was then transferred to a 2 liter flask containing 425 ml PPM medium resulting in a total fermentation volume of approximately 500 ml. DL-hydroxyphenyl-glycine (hpg) was then supplemented to a concentration of 0.3 mM. Cultures without DL-hpg added served as control. Sampling of the supernatant was performed immediately before and after addition of DL-hpg, and subsequently

(8)

oducing a novel pep

5

every 24 hours for the following 5 days. For preparative HPLC analysis, the culture supernatants were harvested by centrifugation at 13000 g, 4 °C for 20 min and then passed through a 0.22 mm filter. The supernatant was con-centrated by freezing in liquid nitrogen followed by lyophilization.

Compound enrichment and analysis using preparative HPLC,

micro-scale NMR and MSX

Obtained lyophilized samples of culture broths were dissolved in milliQ wa-ter and the amount of the thioeswa-ter compound present was measured by LC-MS analysis according to the protocol described in the previous section. After confirmation of the presence of the target compound, preparative HPLC analysis was initiated for enrichment of the compound from the larger vol-ume penicillium culture supernatant.

Separation was performed by direct injection of 100 ml of concentrated supernatant on a XTerra PREP MS C18 column (D × L 7.8 × 150 mm, ID 10 mm) (Waters) kept at 40 °C. Elution was done with water containing 0.1 % (v/v) trifluoroacetic acid (TFA, Merck) (solvent A) and acetonitrile containing 0.1 % TFA (solvent B) at a flow rate of 1 ml/min. Solvent A and solvent B were used with a gradient program as follows: 0–3 min, 90:10; 3–30.5 min, from 90:10 to 47.5:52.5; 30.5–37 min, from 47.5:52.5 to 5:95; 37–42 min, 5:95; 42–45 min from 5:95 to 90:10; from 45–65 min re-equilibration at 90:10. Detection was performed using an SPD-20A Shimadzu UV/VIS detector at 272 nm. Enriched fractions were collected using a fraction collector. Fractions containing the compound of interest were further lyophilized, dissolved in an appropriate solvent analyzed by means of LC-MS(X) and NMR. Similar analysis were per-formed using the DLD-hpgCV standard.

Results

In order to explore the feasibility of a novel, more sustainable biosynthetic pathway for the production of amoxicillin, a novel NRPS is required. There-fore, we chose two D-/L-hydroxyphenylglyine recruiting modules, Tcp9 and Tcp11 and their associated MLP Tcp13, from the teicoplanin biosynthetic cluster, in combination with two L-cysteine and L-valine specific di-modules, PcCV and NlCV of P. chrysogenum and N. lactamdurans, respectively. Two complementing NRPS communicating domains, the donor COMD and accep-tor COMA were integrated C-terminally in Tcp9 and Tcp11 (Tcp9/11-COMD)

(9)

124

5

Results

E

A

T

CO M

A

CO M

D

A

Te

C

A

T E

D

A S O O _CH 3 C H3 S O S O O H N H2 O

T

C

D

A

T

E

A

T

A

Te

C

A

T E

D

A S O O H N H2 O

T

S O _NH O CH3 CH3 NH OH NH2 O O SH S O _NH OH NH2 O O SH N H O O H C H3 C H3 N H O H N H2 O O SH N O O H S C H3 N H O H N H2 O O C H3 IPNS*

C

D

A N H2 O _SH N H2

(10)

oducing a novel pep

5

constructs and N-terminally in CV- constructs (Nl/PcCV- COMA), respectively. Different combinations of Tcp and CV parts, together with a DLD-hpgCV spe-cific amoxicillin synthase (unpublished results) were included in the experi-mental design (Figure 1). The different transformations resulted in 22 strains containing all desired DNA expression constructs. A selection of strains was sequentially subjected to small scale fermentations and the culture super-natant was screened for pathway relevant products, DLD-/LLD-hpgCV and DL-amoxicillin, respectively. None of the screened transformant strains nor any of the controls led to the production of amoxicillin. However, in a series of the transformants we were able to detect significant amounts of a com-pound resembling DLD- hpgCV in retention time on HPLC and by accurate mass on LC/MS (Compound X). The strains producing compound X showed significant differences in their metabolite profile (Figure 2A). Other observed metabolites appear to contain cysteine and valine according to an accurate mass reference search via METLIN [27]. The presence of a reduced com-pound in addition to the oxidized, bis-form of each hypothetical peptide, fur-ther points towards cysteine containing compounds, linked via a side-chain di-sulfide bond (Figure 2B).

Compound X levels were further quantified in all available T1–7 strains and peak production levels were compared (Figure 3). Product levels were recorded in a range of 0.07 µM up to 13.94 µM in the top producer strain T1C.

NlCV containing strains T2 and T4, however, did not produce compound X.

Furthermore, no effect was observed upon the genomic presence of Tcp13, the MLP for Tcp9 and Tcp11, as shown for strains T5 and T7. Only the com-bination of either Tcp9 or Tcp11 with PcCV results in average compound X levels of 5.8–6.7 µM. A distinct inter-strain variation resulted from the non- targeted genome insertion strategy, leading to different genomic integration sites in addition to variations in copy numbers of the transformed genes. This was further reflected in the diverse compound production profiles of the different compound X producer strains (Figure 4A), peaking at days

Figure 1 — NRPS parts and pathway setup used in this study.

(A) A schematic representation of Tcp9, Tcp11, Tcp13 and the NlCV and PcCV di-modules. The NRPS is displayed in an active state, having the amino acids loaded onto the

phospho-pantetheine arm. (B) Proposed pathway, including the subsequent condensation reactions between the three substrates hpg, cysteine and valine and the ultimate cyclization into amoxicillin performed by an amoxicillin synthase (IPNS*). A = adenylation domain; T = thio-lation domain/peptidyl-carrier-protein; C = condensation domain; E = epimerization domain; Te = thioesterase domain.

(11)

126

5

(12)

oducing a novel pep

5

Figure 2 — Overview of total ion chromatograms of different producing and non-produc-ing strains.

(A) Selection of different metabolic profiles of producer, T1C; T5C; T7X, and non-producer strains, T2B and wild type DS58912. Additionally, a 100 µM DLD-hpgCV standard is shown. The most prominent peaks are additionally annotated and displayed in (B). The list includes

the retention time, the protonated Monoisotopic mass of potential monomers and dimers, the molecular composition and the METLIN compound prediction. nd = not determined. Figure 3 — Overview of analyzed strains, and production levels of the 370.14 M/Z [H]+ compound.

Average compound levels at the peak day of production are shown here, grouped by po-tential producer strains (T1–T4) and controls (T5–T21). The precise genetic setup is shown under the respective strains. n = number of different purified and analyzed transformants. 1 NlCV appeared inactive and has not been further tested. 2 PcA, the first module of the ACVS, substitutes Tcp9 and Tcp11. X -X -X -2 -2 X -1 -X -1 -TCP 9 TCP 11 PcCV NlCV TCP 13 -X X -X X -X X -X X -1 -X -X X TCP 11 TCP 9 Control X -X -1 -TCP 13 -TCP PcA 0 2 4 6 8 10 12 T1 T2 T3 T4 T5 T7 T11 T21 μM n = 4 2 2 2 2 3 4 3

(13)

128

5

Results

Figure 4 — Production profiles of selected strains and hpg feed dependency.

(A) Different production profiles of independent strains have been analyzed daily for a time frame of 5 days, revealing different compound production profiles. (B) Peak production lev-els observed in strains grown with and without the addition of 300 µM L-/D-hpg. All values represent averages of at least two biological and two technical replicates. Error bars indicate the standard deviation.

(14)

oducing a novel pep

5

3, 4 or 5 after DL-hpg addition, respectively. Subsequent experiments omit-ting the racemic addition of 300 µM DL-hpg to the culture at day 0, revealed a significant increase in compound X levels by up to nearly 2-fold (Figure 4B). A 10-fold increase in DL-hpg addition to 3 mM did not show any difference in compound X formation (Data not shown). In addition to the transformant strains T1–7, further control strains T11 and T21, both lacking Tcp9 and Tcp11, showed traceable amounts of compound X formation in their supernatant. In order to further elucidate the origin and chemical structure of compound X, focus was subsequently laid on its purification and structural characteriza-tion analysis.

Compound X characterization indicates a thioester-bond peptide

Compound X was isolated using the supernatant of the transformant strain T1C. Supernatant was collected, processed and subsequently analyzed using HPLC and LC/MS(x) in combination with a mass fragmentation analysis. Under preparative HPLC conditions the standard DLD-hpgCV tripeptide eluted at 17.5 min and had a mass of 370.1428 M/Z. Preparative-HPLC fractions were subsequently collected between elution times 15.5 min and 18.2 min (Fig-ure 5A). The LC-MS analysis performed on the collected fractions revealed that fractions 6 to 10 contained a compound, with matching M/Z (370.1428) and elution time (17.5 min) values as the DLD-hpgCV standard.

For further analysis fractions 8 to 10 were subsequently lyophilized (Fig-ure 5B), dissolved in an appropriate amount of solvent and subjected to MS2 and MS3 analysis. The DLD-hpgCV tripeptide standard was analyzed under the same conditions. The obtained MS2 and MS3 data of the DLD-hpgCV standard and compound X were used for the possible structure elucidation of compound X. The MS2 spectrum of DLD-hpgCV standard clearly differs from that of compound X suggesting that these are different compounds (Figure 5C). The MS2 spectrum of DLD-hpgCV is difficult to elucidate since cyclisation seems to occur during MS2 analysis (Figure 5D). On the other hand, the MS2 spectrum of compound X shows two fragments, with M/Z values of 120.0542 and 221.0946. Subsequent fragment assignment of com-pound X and DLD-hpgCV ultimately lead to the conclusion that the observed compound is composed of hpg, cysteine and valine in which hpg and cysteine are coupled via a thioester bond, while cysteine and valine are connected via a peptide bond. Subsequent µ-scale NMR results were inconclusive due to the intrinsic instability and low abundance of the purified compound X.

(15)

130

5

Results

Figure 5 — LC/MS2 fragmentation of DLD-hpgCV and the compound isolated from T1C supernatant.

(A) Chromatogram derived from preparative HPLC of strain T1C supernatant. Displayed is the time of 15.25–18.25 minutes of the 60 minutes method. Fractions are additionally indi-cated as green or blue peak areas including an assigned number (1–10). (B) Fractions dis-played as total retrieved volume (ml) per fraction (y-axis, right) and compound X content is overlaid in the figure as blue bars (y-axis, left). (C) MS2 fragmentation of compound X and DLD-hpgCV, showing clearly diverging patterns. (D) Expected tripeptide structure of DLD-hpgCV with indicated peptide bond (red) versus probable structure of compound X, highlighting the thioester bond (green).

(16)

oducing a novel pep

5 Enzymatic nature of compound X

To further investigate the involvement of the hpg activating initiation module as well as the PcCV di-module, a series of mutations were introduced into functional sequence motifs of the Tcp11 T domain as well as the first PcCV C domain. First, Tcp11 was mutated at the conserved serine (S497) in the se-quence LGGDSISSM, serving as anchor for the phosphopantetheine (ppant) group attachment, thus rendering the domain, and hence the module inactive.

Figure 6 — Production of compound X by P.chrysogenum strains containing the S497A or H130V mutation in the thiolation domain of Tcp11 and condensation domain of the cyste-ine module of PcCV, respectively.

Two mutations were introduced into Tcp11 and PcCV, which result in reduced product levels (S497A) and a virtually complete interruption of compound X production (H130V).

(17)

132

5

Discussion

In addition, two downstream serine residues, LGGDSISSM at position 499 and 500 were targeted for mutation, to exclude for their functional replacement of the conserved serine residue at position 497. The different single, double and triple serine mutants were subsequently expressed in E. coli BL21 (DE3), purified and tested in vitro for phosphopantetheine attachment, using fluo-rescent CoA647 as ppant-donor (Supplementary figure 2). Only in the S497A mutants, we recorded no ppant attachment. Thus, the downstream serine residue 499 and 500 could not substitute for the mutation in the functional motif. Tcp11-COMD S497A was further transformed to P. chrysogenum, where sequential fermentations revealed a 6.9-fold decrease in compound X levels, relative to strain T1C, or in average a decline from 9.85 µM to 1.42 µM, re-spectively (Figure 6). The second functional motif we targeted is at the ac-tive site of the first PcCV condensation domain and hypothetically enabled the formation of a peptide bond between the upstream and downstream activated substrates. The conserved histidine at position 130 in the conden-sation domain sequence SCHHAILDGWSL was mutated (H130V) and the re-sulting construct transformed and analyzed in P. chrysogenum. Compound X level quantification indicated a near to complete abolition of production with H130V strains. Compound X levels exhibited a small peak at a concentration of 0.03 µM, essentially leaving the PcCV di-module inactivated (Figure 6). These data demonstrate that the thioester bonded hpgCV peptide results

from enzymatic catalysis involving the two-component hybrid NRPS.

Discussion

The feasibility of a new pathway for the production of the β-lactam antibiotic amoxicillin was evaluated. The new process aims at incorporating hydroxy-phenylglycine, cysteine and valine into the amoxicillin precursor tripeptide DLD-hpgCV, and thus requires a novel nonribosomal peptide synthetase ac-tivity. Therefore, we created four genes encoding two hpg activating modules (Tcp9/11-COMD) and two cys-val activating di-modules (Nl/PcCV- COMA), in-cluding two complementing communicating domains (COMA/D), projected to facilitate the association of the two protein products. The ultimate conver-sion of DLD-hpgCV into amoxicillin requires an amoxicillin synthase, which was developed in a separate study through the mutagenesis of the native isopenicillin N synthetase rendering it capable of converting DLD-hpgCV into amoxicillin (unpublished results). The corresponding genes were subse-quently transformed to P. chrysogenum and genetically characterized. The fer-mentation of the strains and the analysis of the extracellular broth revealed

(18)

oducing a novel pep

5

the presence of a novel compound with the exact accurate mass as well as the retention time of the DLD-hpgCV standard. However, in none of the sam-ples L- amoxicillin nor D-amoxicillin was detected. This was not entirely un-expected, due to the low activity of the amoxicillin synthase, at <2 % of the native IPNS enzyme (unpublished results). Furthermore, the low abundance of the potential substrate, might interfere with amoxicillin formation. Intrigu-ingly, increased novel compound levels were observed when omitting the DL-hpg precursor from the fermentation. Although P. chrysogenum is not known to contain a hpg biosynthesis pathway [28], the unequivocal detection of hpg as a constituent of compound X implies the presence of such a pathway. Further analytical characterization revealed that the novel product does not corre-spond to DLD- hpgCV but rather to a unique thioester bonded hpgCV structure. To gain additional information on the structure of compound X, a preparative enrichment and purification was conducted. During the purification process, compound X appeared temperature sensitive (Data not shown) and readily de-composed. MS fragmentation patterns of compound X pointed towards a weak bond, leading to essentially two fragments which by accurate mass could be identified as hpg and cys-val. NMR analysis was not conclusive due to the low abundance and lability of the compound. Overall, it appeared that compound X is the product of an incomplete peptide bond formation, which may leave the thiol group of the cysteine tethered with the carboxyl group of the hpg. This formation remains reactive and is probably an intrinsic intermediate in the peptide bond formation process of an NRPS condensation domain. Due to the nature of the NRPS setup or composition, however, the geometry of our altered NRPS modules and in particular the interaction with the first conden-sation domain of the CV di-module may prohibit this process from completion. The truncation of the ACVS into the CV di-module, potentially interfered with crucial inter-domain interactions of the native ACVS, leaving the cysteine side chain in a position more favorable for peptide bond formation, rather than the amino group. Because of the unusual chemical characteristics of compound X, its enzymatic origin was further elucidated through additional transformants. Those data showed that the PcCV di-module is strictly required to allow for any

compound X production while the addition of either Tcp9 or Tcp11 is crucial for the high product levels. No activity was observed with the NlCV module. Remarkably, the activity was independent of Tcp13, the MLP of Tcp9 and Tcp11, despite the beneficial effects on Tcp9 and Tcp11 adenylation activity in in vivo and in vitro studies (Chapter II/Unpublished data).

To unravel further details on the precise role of Tcp11 and PcCV, two muta-tions were introduced, targeting conserved residues in the functional motifs of the thiolation domain of Tcp11 (S497A) and the N-terminal condensation

(19)

134

5

Discussion

domain of the PcCV di-module (H130V), respectively. Analysis of the S497A mutants revealed drastically reduced amounts of compound X, approximately to the levels observed with T11 strains, containing the PcCV di-module only. Further mutations of downstream serine residues in the core T domain se-quence Tcp11 — FFALGGDSISSMQ, at position 499 and 500, demonstrated a non-complementing role of those two positions for the conserved S497 residue. In vitro experiments using fluorescent CoA 647 as a phosphopan-tetheine donor (Supplementary figure 2), supported the in vivo data and was successfully used as a new tool for the in vitro analysis of NRPS apo- and holo- states [29] (This study).

The second mutation, PcCV H130V, was expected to render PcCV inac-tive, which was confirmed using P. chrysogenum H130V transformant strains. Compound X production was essentially abolished, which ultimately demon-strates that the catalytic activity of PcCVs' first condensation domain is es-sential for the formation of the observed thioester containing compound X and that the hpg activating module leads to increased compound X levels. However, the low, but significant levels of compound X seen in penicillium strains containing only PcCV remained elusive. Two possible explanations could be considered: (i) an inefficient chemical conversion of NRPS released cys-val leading to compound X formation or, (ii) the PcCV or COMA depen-dent recruitment of an intrinsic Penicillium “Orphan” NRPS module [30]. With respect to the first hypothesis, chemical conversion is difficult to ex-amine in a complex cellular setup, especially where the resulting product, the thioester bonded compound X is chemically unstable. The high amount of potentially cys-val containing compounds observed in the culture broth may suggest that a substantial level of compound X is produced that due to instability decomposes into the CV product. With respect to the second hypothesis, NRPS module or domain recruitment is a phenomenon which has just recently been observed in several studies [31–32]. Inspection of all annotated single NRPS modules in P. chrysogenum, revealed one interesting target, Pc20g12670 (Supplementary figure 1), which was predicted to accept a bulky substrate. However, in order to further analyze and potentially proof the origin and mechanism of compound X formation and accomplish the de-sired DLD-hpgCV formation, a redesign of the COM linker based NRPS genes, including a subsequent targeted gene integration in Penicillium, and a sys-tematic compound characterization needs to be conducted. Based on this work we present here clear evidence for the production of a novel compound, guided by a two component NRPS system, which effectively associates using NRPS inter-module communicating domains, representing the first example of successful in vivo implementation in a heterologous host.

(20)

oducing a novel pep

5 Conclusion and perspectives

Here we have employed a COM linker strategy to generate hybrid NRPS en-zymes in vivo, yielding a novel NRPS product, a thioester bonded hpgCV pep-tide. In a wider context, NRPS COM domains as laid out in this research, have the potential to complement extensive site, region or domain directed engi-neering approaches, in strategies to create NRPS with novel activities. This is not only crucial for the production and discovery of new compounds, but may also lead to the replacement of semi-synthetic or synthetic production routes in favor of more sustainable, fully fermentative ones.

Acknowledgements

The authors would like to give their gratitude to DSM Sinochem Pharmaceu-ticals BV, for the construction of plasmids, transfer of strains and further materials as well as for analytical and intellectual support. The research has been financially supported by the research program of the biobased ecologi-cally balanced sustainable industrial chemistry (BE-BASIC).

References

1. Cantwell C, Beckmann R, Whiteman P, Queener SW, Abraham EP. Isolation of deacetoxycephalosporin C from fermen-tation broths of Penicillium chrysoge-num transformants: construction of a new fungal biosynthetic pathway. Proc Biol Sci. 1992;248: 283–9.

doi:10.1098/rspb.1992.0073

2. Harris DM, Westerlaken I, Schipper D, van der Krogt Z a, Gombert AK, Sutherland J, et al. Engineering of Penicillium chrysoge-num for fermentative production of a novel carbamoylated cephem antibiotic precursor. Metab Eng. 2009;11: 125–37. doi:10.1016/j.ymben.2008.12.003 3. McLean KJ, Hans M, Meijrink B, van

Scheppingen WB, Vollebregt A, Tee KL,

et al. Single-step fermentative produc-tion of the cholesterol-lowering drug pravastatin via reprogramming of Peni-cillium chrysogenum. Proc Natl Acad Sci U S A. 2015;112: 2847–52.

doi:10.1073/pnas.1419028112

4. Veiga T, Nijland JG, Driessen AJM, Boven-berg RAL, Touw H, van den Berg MA, et al. Impact of velvet complex on tran-scriptome and penicillin G production in glucose-limited chemostat cultures of a β-lactam high-producing Penicillium chrysogenum strain. OMICS. 2012;16: 320–33. doi:10.1089/omi.2011.0153 5. Weber T, Blin K, Duddela S, Krug D, Kim

HU, Bruccoleri R, et al. antiSMASH 3.0 — a comprehensive resource for the

(21)

136

5

Ref er enc es

genome mining of biosynthetic gene clusters. Nucleic Acids Res. 2015;43: W237–43. doi:10.1093/nar/gkv437 6. Blin K, Medema MH, Kottmann R, Lee

SY, Weber T. The antiSMASH database, a comprehensive database of microbial sec-ondary metabolite biosynthetic gene clus-ters. Nucleic Acids Res. England; 2017;45: D555–D559. doi:10.1093/nar/gkw960 7. Umemura M, Koike H, Machida M.

Mo-tif-independent de novo detection of sec-ondary metabolite gene clusters-toward identification from filamentous fungi. Front Microbiol. Switzerland; 2015;6: 371. doi:10.3389/fmicb.2015.00371

8. Khaldi N, Seifuddin FT, Turner G, Haft D, Nierman WC, Wolfe KH, et al. SMURF: Genomic mapping of fungal secondary metabolite clusters. Fungal Genet Biol. United States; 2010;47: 736–741. doi:10.1016/j.fgb.2010.06.003

9. Hoertz AJ, Hamburger JB, Gooden DM, Bednar MM, McCafferty DG. Studies on the biosynthesis of the lipodepsipeptide antibiotic ramoplanin A2. Bioorganic Med Chem. Elsevier Ltd; 2012;20: 859– 865. doi:10.1016/j.bmc.2011.11.062 10. Finking R, Marahiel MA. Biosynthesis of

nonribosomal peptides. Annu Rev Micro-biol. Annual Reviews; 2004;58: 453–488. doi:10.1146/annurev.micro.58.030603. 123615

11. Cane DE, Walsh CT. The parallel and convergent universes of polyketide thases and nonribosomal peptide syn-thetases. Chem Biol. 1999;6: R319–R325. doi:10.1016/S1074-5521(00)80001-0 12. Bloudoff K, Fage CD, Marahiel MA,

Schmeing TM. Structural and muta-tional analysis of the nonribosomal

peptide synthetase heterocyclization domain provides insight into catalysis. Proc Natl Acad Sci U S A. United States; 2016; doi:10.1073/pnas.1614191114 13. Winn M, Fyans JK, Zhuo Y, Micklefield

J. Recent advances in engineering non-ribosomal peptide assembly lines. Nat Prod Rep. The Royal Society of Chemis-try; 2016;33: 317–347.

doi:10.1039/C5NP00099H

14. Nielsen ML, Isbrandt T, Petersen LM, Mortensen UH, Andersen MR, Hoof JB, et al. Linker Flexibility Facilitates Mod-ule Exchange in Fungal Hybrid PKS-NRPS Engineering. PLoS One. United States; 2016;11: e0161199.

doi:10.1371/journal.pone.0161199 15. Doekel S, Marahiel MA. Dipeptide

for-mation on engineered hybrid peptide synthetases. Chem Biol. 2000;7: 373–84. Available: http://www.ncbi.nlm.nih.gov/ pubmed/10873839

16. Eppelmann K, Stachelhaus T, Marahiel MA. Exploitation of the selectivity-con-ferring code of nonribosomal peptide synthetases for the rational design of novel peptide antibiotics. Biochemistry. 2002;41: 9718–26.

17. Stachelhaus T, Schneider A, Marahiel MA. Rational design of peptide antibiotics by targeted replacement of bacterial and fungal domains. Science. 1995;269: 69–72. 18. Mootz HD, Schwarzer D, Marahiel MA. Construction of hybrid peptide synthe-tases by module and domain fusions. Proc Natl Acad Sci U S A. National Acad-emy of Sciences; 2000;97: 5848–53. doi:10.1073/pnas.100075897

19. Hahn M, Stachelhaus T. Harnessing the potential of communication-mediating

(22)

oducing a novel pep

5

domains for the biocombinatorial syn-thesis of nonribosomal peptides. Proc Natl Acad Sci U S A. National Academy of Sciences; 2006;103: 275–80. doi:10.1073/pnas.0508409103 20. Yin X, Zabriskie TM. The enduracidin

bio-synthetic gene cluster from Streptomy-ces fungicidicus. Microbiology. 2006;152: 2969–2983. doi:10.1099/mic.0.29043-0 21. Wang H, Fewer DP, Holm L, Rouhiainen

L, Sivonen K. Atlas of nonribosomal peptide and polyketide biosynthetic pathways reveals common occurrence of nonmodular enzymes. Proc Natl Acad Sci. 2014;111: 9259–9264. doi:10.1073/ pnas.1401734111

22. Pootoolal J, Thomas MG, Marshall CG, Neu JM, Hubbard BK, Walsh CT, et al. Assem-bling the glycopeptide antibiotic scaffold: The biosynthesis of A47934 from Strepto-myces toyocaensis NRRL15009. Proc Natl Acad Sci U S A. 2002;99: 8962–8967. 23. Sosio M, Kloosterman H, Bianchi A, de

Vreugd P, Dijkhuizen L, Donadio S. Orga-nization of the teicoplanin gene cluster in Actinoplanes teichomyceticus. Micro-biology. Microbiology Society; 2004;150: 95–102. doi:10.1099/mic.0.26507-0 24. Polli F, Meijrink B, Bovenberg RAL,

Driessen AJM. New promoters for strain engineering of Penicillium chrysogenum. Fungal Genet Biol. 2015;

doi:10.1016/j.fgb.2015.12.003

25. Kovalchuk A, Weber SS, Nijland JG, Bovenberg RAL, Driessen AJM. Fungal ABC transporter deletion and localization analysis. Methods Mol Biol. 2012;835: 1–16. doi:10.1007/978-1-61779-501-5_1 26. Harju S, Fedosyuk H, Peterson KR. Rapid

isolation of yeast genomic DNA: Bust

n’ Grab. BMC Biotechnology. London; 2004. p. 8. doi:10.1186/1472-6750-4-8 27. Smith CA, O’Maille G, Want EJ, Qin C, Trauger

SA, Brandon TR, et al. METLIN: a metabolite mass spectral database. Ther Drug Monit. United States; 2005;27: 747–751.

28. van den Berg MA, Albang R, Albermann K, Badger JH, Daran JM, Driessen AJM, et al. Genome sequencing and analysis of the filamentous fungus Penicillium chrysogenum. Nat Biotechnol. 2008;26: 1161–1168. doi:10.1038/nbt.1498 29. Beld J, Sonnenschein EC, Vickery CR,

Noel JP, Burkart MD. The phosphopan-tetheinyl transferases: Catalysis of a posttranslational modification crucial for Life. Nat Prod Rep. 2014;31: 61–108. doi:10.1039/c3np70054b

30. Marta M. Samol, Oleksandr Salo, Peter Lankhorst, Roel A.L. Bovenberg, Ar-nold J.M. Driessen. 9. Secondary me-tabolite formation by the filamentous fungus Penicillium chrysogenum in the post-genomic era. Aspergillus Penicil-lium Post-genomic Era. 2016; 145–172. doi:doi.org/10.21775/9781910190395.08 31. Haslinger K, Peschke M, Brieke C,

Maxi-mowitsch E, Cryle MJ. X-domain of pep-tide synthetases recruits oxygenases crucial for glycopeptide biosynthesis. Nature. 2015;521: 105–9.

doi:10.1038/nature14141

32. Ulrich V, Peschke M, Brieke C, Cryle MJ. More than just recruitment: the X-do-main influences catalysis of the first phe-nolic coupling reaction in A47934 biosyn-thesis by Cytochrome P450 StaH. Mol Biosyst. England; 2016;12: 2992–3004. doi:10.1039/c6mb00373g

(23)

138

5

Supplementary material

Supplementary table 1 — DNA used to set up different P. chrysogenum COM-linker path-way transformations.

Genetic setup of analyzed strains in this study. Indicated are the utilized genetic fragments and the respective amounts used for transformation purposes to P. chrysogenum.

hpg module CV module IPNS MbtH-like Marker

part amount part amount part amount part amount part amount

T₁ Tcp11-1-ComD 3 µg ComA-PcCV 5 µg AnI101 2 µg Tcp13 0.8 µg Amds 0.25 µg

T₂ Tcp11-1-ComD 3 µg ComA-NlCV 5 µg AnI101 2 µg Tcp13 0.8 µg Amds 0.25 µg

T₃ Tcp9AT-ComD 3 µg ComA-PcCV 5 µg AnI101 2 µg Tcp13 0.8 µg Amds 0.25 µg

T₄ Tcp9AT-ComD 3 µg ComA-NlCV 5 µg AnI101 2 µg Tcp13 0.8 µg Amds 0.25 µg

T₅ Tcp11-1-ComD 3 µg ComA-PcCV 5 µg AnI101 2 µg - - Amds 0.25 µg

T₆ Tcp11-1-ComD 3 µg ComA-NlCV 5 µg AnI101 2 µg - - Amds 0.25 µg

T₇ Tcp9AT-ComD 3 µg ComA-PcCV 5 µg AnI101 2 µg - - Amds 0.25 µg

T₈ Tcp9AT-ComD 3 µg ComA-NlCV 5 µg AnI101 2 µg - - Amds 0.25 µg

T₉ Tcp11-1-ComD 3 µg - - AnI101 2 µg Tcp13 0.8 µg Amds 0.25 µg

T₁₀ Tcp9AT-ComD 3 µg - - AnI101 2 µg Tcp13 0.8 µg Amds 0.25 µg

T₁₁ - - ComA-PcCV 5 µg AnI101 2 µg Tcp13 0.8 µg Amds 0.25 µg

T₁₂ - - ComA-NlCV 5 µg AnI101 2 µg Tcp13 0.8 µg Amds 0.25 µg

T₁₃ Tcp11-1-ComD 3 µg ComA-PcCV 5 µg - - Tcp13 0.8 µg Amds 0.25 µg

T₁₄ Tcp11-1-ComD 3 µg ComA-NlCV 5 µg - - Tcp13 0.8 µg Amds 0.25 µg

T₁₅ Tcp9AT-ComD 3 µg ComA-PcCV 5 µg - - Tcp13 0.8 µg Amds 0.25 µg

T₁₆ Tcp9AT-ComD 3 µg ComA-NlCV 5 µg - - Tcp13 0.8 µg Amds 0.25 µg

T₂₁ PcA-COM 3 µg Coma-PcCV 5 µg - - Tcp13 0.8 µg Amds 0.25 µg

T₂₂ NlA-Com 3 µg Coma-NlCV 5 µg - - Tcp13 0.8 µg Amds 0.25 µg

T₂₃ PcA-COM 3 µg Coma-PcCV 5 µg - - - - Amds 0.25 µg

(24)

oducing a novel pep

5

Supplementary figure 1 — Orphan NRPS modules overview.

List of functionally unassigned NRPS modules of P. chrysogenum [30]. Protein size and pro-posed adenylation domain placement are indicated. Furthermore, adenylation domain se-quences were used for in silico specificity determination, including an indication to the reli-ability of the prediction (-; +/-; +; ++).

Gene Accession Class Length AA Domains A range Prediction A spec Pred rel

Pc20g12670 CAP86596.1 NRPS-like 1241 [A]-[CNAD] 235-719 Phenylalanine-(Acetate) +

Pc12g13170 XP_002558125 NRPS-like 1056 [A]-T-[CNAD] 1-545 Proline +

Pc13g12570 XP_002559673.1 NRPS-like 1000 A-T-[CNAD] 1-507 Serine

+/-Pc21g22530 CAP97150.1 NRPS-like 878 [A]-T 265-774 Aminoadipic acid +

Pc22g09430 CAP98231.1 NRPS-like 1030 [A]-T-[CNAD] 1-538 Alanine

+/-Pc14g01790 XP_002560173.1 NRPS-like 1038 [A]-T-[CNAD] 1-537 Leucine +

Pc06g01540 CAP79147 NRPS-like 1025 A-T-[CNAD] 1-530 Alanine +

Pc20g02590 CAP85588.1 NRPS-like 1011 A-T-[CNAD] 1-516 Alanine ++

Pc20g02260 CAP85555.1 NRPS-like 1080 [A]-T-[CNAD] 1-575 Alanine +

Pc22g09430 CAP98231.1 NRPS-like 1030 [A]-T-[CNAD] 1-538 Alanine

+/-Pc12g09980 XP_002557823 NRPS-like 1619 [A]-T-[Te] 93-620 Alanine +

Pc16g09930 XP_002561303.1 NRPS-like 967 [A]-T-[Te/X] 1-573 Alanine +

Pc18g00380 XP_002561885.1 NRPS-like 1276 A-T-[CNAD]-[X] 1-525 Alanine +

Pc20g09690 CAP86298.1 PKS-like 1038 [A]-T-[CNAD] 1-535 Valine

(25)

140

5

Supplementary material

Supplementary table 2 — Primer used in this study.

List of diagnostic primers used for the detection of genomically integrated expression cassettes.

Name Sequence 5’ → 3’ Purpose

FW Pipns CTTATACTGGGCCTGCTG Promoter IPNS integration check

RV Pipns TGATATCCTGTCTTCAGTCTTA Promoter IPNS integration check

FW Ipns ATGGGTTCAGTCAGCAAAG IPNS integration check

RV Ipns CTAGGTCTGGCCGTTCTTG IPNS integration check

FW Tcp13 ATGACGAACCCATTTGACAAC Tcp13 integration check

RV Tcp13 ATCGCTAATCTGGGAGATCAGG Tcp13 integration check

FW NlA ATGACGTCAGCACGACAC Nocardia lactamdurans AT module _{integration check}

RV NlA TTACTGCTCGATTTCCCACTC Nocardia lactamdurans AT module _{integration check}

FW PcA ATGACTCAACTGAAGCCACC P.chrysogenum AT module integration check

RV PcA TTACTGCTCGATCTCCCATTC P.chrysogenum AT module integration check

FW Tcp9 ATGAATTCCGCAGCGCAG Tcp9 integration check

RV Tcp9 TTACTGCTCGATCTCCCATTC Tcp9 integration check

FW Tcp11 ATGGCGGAACGGCTC Tcp11 integration check

RV Tcp11 TTACTGCTCGATCTCCCATTC Tcp11 integration check

FW-1 NlCV ATGTCGCAGCAGCCGACC Nocardia lactamdurans CV module _{integration check}

FW-2 NlCV TCGAGCAGCTGGTCAAGGAAC Nocardia lactamdurans CV module _{integration check}

RV-1 NlCV ACGTCGCCCGAGGTCAG Nocardia lactamdurans CV module _{integration check}

RV-2 NlCV TTAGTCGCTCCCTAGGCTCGTC Nocardia lactamdurans CV module _{integration check}

FW-1 PcCV ATGAGCCAGCAGCCCACC P.chrysogenum CV module integration check

FW-2 PcCV CTGAGCAAGAAGGAAACGGAGAAC P.chrysogenum CV module integration check

RV-1 PcCV CTGGTTCAACGACGCTACTTTCTG P.chrysogenum CV module integration check

RV-2 PcCV TTAATAGCGAGCGAGGTGTTCC P.chrysogenum CV module integration check

FW Amds ATGTTAGACCTCCGCCTC Amds integration check

(26)

oducing a novel pep

5

Supplementary figur e 2 — C oA 647 labelling o f T cp11 mutants. Inc orpor ation o f C oA647 b y sfp into Tcp11 c onstructs. Sho wn ar e mutants tar

geting the functional

T domain mo tif GGD S ( S497) as w ell as tw o additional do wnstr eam serine r esidues ( S499 / S500 ). The f our panels sho w the wildtype (T cp11 WT), single mutant S497

A, double mutant S499A/S500A and the triple

mutant S497

A/S499A/S500A. P

ictur

es on the left displa

y DIA light e

xposur

e o

f the g

el and on the right fluor

esc ent imag es at e xcitation o f 647nm. F urthermor e no-sfp c ontr ols ( sfp-), no pr otein c ontr ol (-hpg) and no C oA c ontr ol (-C oA) w er e c onducted as w ell as diff er ent r atios o f C oA647:C oA (2:1; 1:1; 1:2) w er e tested.

(27)

142

5

Supplementary material pI-Tcp9AT-COMD-AT-zeo 5096 bps 1000 2000 3000 4000 5000 HindIII EcoRI ApaI PspOMI ScaI RsrII NdeI EcoRI SmaI XmaI ApaI PspOMI AatII ZraI MluI SfiI BsiWI Bsp1407I BsrGI BstAUI SspBI BsiWI MluI SmaI XmaI SmaI XmaI MluI SmaI XmaI SgrAI BclI FbaI Ksp22I SmaI XmaI PasI Bsu36I BsiWI MluI FseI SgrAISmaI XmaIAatII ZraISfiI AclI SspIPsiI Pipns Tcp9AT-ComD 'Tat pUC ori Zeo 'F1 ori pI-tcp11-1-COMD-AT-zeo 6353 bps 1000 2000 3000 4000 5000 6000 MluI MluI MluI Pipns tcp11-1ATE-COMD 'Tat pUC ori Zeo 'F1 ori pI-COMA-PcCV-AT-zeo 11812 bps 2000 4000 6000 8000 10000 XbaI XbaI XbaI Pipns COMA-PcCV 'Tat pUC ori Zeo 'F1 ori pI-COMA-NlCV-AT-zeo 11640 bps 2000 4000 6000 8000 10000 HindIII EcoRI NdeI EcoRI NotI NotI NotI NotI NotINsiI NotI NotI Pipns COMA-NlCV 'Tat pUC ori Zeo 'F1 ori

(28)

oducing a novel pep

5

Supplementary figure 3 — Plasmids used in this study.

pIAT-AnI101-zeo 4278 bps 1000 2000 3000 4000 HindIII EcoRI PshAI AccIII ApaI PspOMI ScaI TatI RsrII Ecl136II SacI NdeI ClaI BsmBI EcoO109I PpuMI XmnI BamHI SbfI PstI BfrBI NsiI BsiWI MluI DraI FseI SexAI BmgBISgrAI AatIIZraI BssHIISfiI AclI SspIPsiI Pipns AnI101 'Tat pUC ori Zeo 'F1 ori pIAT-tcp13-zeo 3492 bps 500 1000 1500 2000 2500 3000 HindIII EcoRI PshAI AccIII ApaI PspOMI ScaI TatI RsrII Ecl136II SacI BsiWI MluI DraI FseI SexAI BmgBISgrAI AatIIZraI BssHIISfiI AclI SspIPsiI Pipns tcp13 'Tat pUC ori Zeo 'F1 ori pDONR221-AMDS 5710 bps 1000 2000 3000 4000 5000 NheI NheI BspTI ApaI SmaI PstI EcoRI SphI SphI MunI SalI PstI BglII NheI NcoI NdeI PvuI EcoRV NcoI EcoRV NcoI XbaI NcoI EcoRV NruI PvuI PciI attL1 PgpdA amdS AT attL2 KanR ori

(29)