Chlorinated Glycopeptide Antibiotic Peptide Precursors Improve

Feb 20, 2017 - The activity of glycopeptide antibiotics (GPAs) depends upon important structural modifications to their precursor heptapeptide backbon...
2 downloads 0 Views 2MB Size
Subscriber access provided by Fudan University

Article

Chlorinated glycopeptide antibiotic peptide precursors improve Cytochrome P450-catalyzed cyclization cascade efficiency Madeleine Peschke, Clara Brieke, Rob Goode, Ralf Schittenhelm, and Max J. Cryle Biochemistry, Just Accepted Manuscript • DOI: 10.1021/acs.biochem.6b01102 • Publication Date (Web): 20 Feb 2017 Downloaded from http://pubs.acs.org on February 21, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Biochemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Chlorinated glycopeptide antibiotic peptide precursors improve Cytochrome P450-catalyzed cyclization cascade efficiency Madeleine Peschke,† Clara Brieke,† Rob J. A. Goode,#,£ Ralf B. Schittenhelm#,£ and Max J. Cryle* †,#,$. †

Department of Biomolecular Mechanisms, Max Planck Institute for Medical Research, Jahnstrasse 29, 69120 Heidelberg, Germany. # The Monash Biomedicine Discovery Institute, Department of Biochemistry and Molecular Biology £ Monash Biomedical Proteomics Facility, Monash University, Clayton, Victoria 3800, Australia $

ARC Centre of Excellence in Advanced Molecular Imaging, Monash University, Clayton, Victoria 3800, Australia. EMBL Australia, Monash University, Clayton, Victoria 3800, Australia. Supporting Information Placeholder ABSTRACT: The activity of glycopeptide antibiotics

depends upon important structural modifications to their precursor heptapeptide backbone: specifically, the Cytochrome P450-catalyzed oxidative crosslinking of aromatic side chains as well as the halogenation of specific residues within the peptide. The timing of halogenation and its effect on the cyclization of the peptide is currently unclear. Our results show that chlorination of peptide precursors improves their processing by P450 enzymes in vitro, which provides support for GPA halogenation occurring prior to peptide cyclization during nonribosomal peptide synthesis. We could also determine that the activity of the second enzyme in the oxidative cyclization cascade, OxyA, remains higher for chlorinated peptide substrates even when the biosynthetic GPA product possesses an altered chlorination pattern, which supports the role of the chlorine atoms in orienting the peptide substrate in the active site of these enzymes.

The glycopeptide antibiotics (GPAs) are an important class of peptide based antibiotics of clinical relevance for the treatment of serious Gram-positive bacterial infections.1 Several GPAs are used in the clinic, including the natural compounds vancomycin and teicoplanin as well as second generation semisynthetic variants: the latter have been developed to aid in the treatment of resistant bacterial infections, which is a major problem facing medicine today.2 GPAs are complex molecules, with the heptapeptide core containing many nonproteinogenic amino acids and displaying a high degree of crosslinking between the aromatic side chains of residues within the peptide (Fig. 1).3, 4 The core of GPAs can be further decorated by many classes of chemical transformations, including glycosylation, acylation, sulfation, methylation and halogenation: most GPA diversity iden-

tified to date stems from alternatively modified forms of a limited number of core peptide structures.1 The crosslinking of GPA peptides into rigid aglycones possessing a defined three-dimensional structure is essential for their activity, which is the inhibition of peptidoglycan biosynthesis.1, 2 This process is mediated via a hydrogen bonding complex of GPAs to the D-Ala-D-Ala dipeptide terminus of the linear, non-crosslinked peptidoglycan precursor.1, 2 Given the importance of the crosslinking of the peptide precursor to the activity of GPAs and that all GPAs in clinical use derive from bacterial fermentation (and thus the activity of the natural biosynthetic machinery), understanding this process is a high priority.3, 5-8 The installation of the peptide crosslinks in GPAs – three crosslinks for vancomycin type GPAs and four crosslinks for teicoplanin type GPAs – has been shown to be catalyzed by members of the Cytochrome P450 superfamily of monooxygenases (Fig. 1a).9 These P450s, known as Oxy enzymes, act in a specific order (OxyB: C-O-D ring, OxyE: F-O-G ring (only in teicoplanin-type GPAs), OxyA: D-O-E ring, OxyC: AB ring) to form the respective fully cyclized aglycone.3 Early results from both in vivo and in vitro studies pointed to close linkage between the activity of peptide synthesis – mediated by a linear non-ribosomal peptide synthetase (NRPS)10-12 – and peptide side chain crosslinking performed by the Oxy enzymes (Fig. 1a).7, 13-16 Recently, it was shown that the substrates for the Oxy enzymes are heptapeptides bound to the final PCP domain of the GPA NRPS and that the Oxy enzymes are recruited to these substrates via binding to the X-domain, which is conserved in the final module of all GPA producing NRPS machineries.3, 5, 17-19 The discovery of the X-domain as the recruitment platform for the Oxy enzymes has enabled the activity of several Oxy homologues to now be reconstituted in vitro, which include examples of bi- and even

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

tricyclization of GPA precursor peptide substrates by a cascade of Oxy enzymes.5, 17-20 Our understanding of the Oxy enzymes has been further enhanced by the recent structural determination of the final member to be characterized – OxyA.20, 21 The structures of OxyA homologues from both the teicoplanin and related A47934 system revealed an overall fold that is similar to the general Oxy/P450 fold.20, 21 Nevertheless, the active site of OxyA contains significantly more polar residues than is the case for OxyB and OxyC. In vitro activity studies have further revealed that OxyA displays more stringent selectivity for the peptide substrate in comparison to OxyB and OxyE.17-20 This is especially true for the stereochemistry of the final amino acid residue in the peptide. Given these differences, we wanted to explore the importance of maintaining potential hydrogen bonding interactions between the peptide substrate and the OxyA active site, and in particular the role of chlorine atoms on the GPA precursor peptide. The chlorination of GPAs plays a very important role in their activity, as the chlorine atoms contribute directly to the binding affinity displayed for GPAs to their target motif (Fig. 2). In spite of the recent advances in understanding GPA biosynthesis, the timing of a further important peptide modification – chlorination, and specifically the timing of chlorination in relation to peptide synthesis and crosslinking – remains unclear. Previous in vitro turnover of PCP-bound peptides by the OxyB homologue from vancomycin biosynthesis had shown that the presence of a chlorine substituent on the Tyr-6 residue of the peptide reduced the processing of hexapeptides significantly, and completely abolished the processing of heptapeptides.22 However, OxyBvan is the only Oxy homologue identified that shows significant activity against peptidyl-PCP substrates without the Xdomain being present, and thus these experiments did not take into account the recruitment of Oxy enzymes to PCP-bound peptide substrates by the unique GPA Xdomain.5, 18, 23, 24 Given the results of the biochemical and structural characterization of OxyA homologues, we thus decided to examine the effect of peptide halogenation on the crosslinking efficiency of both OxyB and OxyA homologues from GPA biosynthesis. Our results demonstrate that both OxyB and OxyA enzymes can accept and process halogenated peptide substrates when present by constructs containing the X-domain, and that the efficiency of this process is even aided by the presence of the chlorine atoms within the peptide. EXPERIMENTAL METHODS Materials

All chemicals and solvents were obtained from commercial suppliers (Sigma-Aldrich; VWR) and used without further purification. Dawson Dbz AM resin (100-200 mesh, 0.49 mmol/g), activators and protected amino

Page 2 of 19

acids not specifically mentioned below were obtained from Merck Novabiochem. (D)- and (L)-4hydroxyphenylglycine were obtained from SigmaAldrich and were converted into their Fmoc-protected forms as previously reported.25, 26 Fmoc-protected, chlorinated (L)- and (D)-tyrosine were obtained from NetChem; Glucose dehydrogenase (274 U/mg) was obtained from Sorachim; NADH was purchased from Gerbu Biotechnik GmbH. Solid phase peptide synthesis was performed on a Tribute UV synthesizer (Protein Technologies) on a 50 µmolar scale. Solid phase extraction purifications were performed using Strata-X SPE columns (Phenomenex). HPLC analysis and purifications were carried out using a High Performance Liquid Chromatograph/Mass Spectrometer LCMS-2020 (ESI, operating both in positive and negative mode) equipped with a SPD-M20A Prominence Photo Diode Array Detector in preparative mode and a SPD-20A Prominence Photo Diode Array Detector in analytical mode, all from Shimadzu. For analytical analyses the solvent delivery module LC-20AD was used; for preparative purifications two LC-20AP units were used. Analytical separations were performed on Waters XBridge BEH300 C18 columns (5 or 10 µm, 4.6 x 250 mm) using a flow rate of 1 mL/min. Preparative separations of peptidyl-CoAs were performed on a Waters XBridge BEH300 preparative C18 column (5 µm, 19 x 150 mm) using a flow rate of 20 mL/min. The solvents used were water + 0.1% formic acid (solvent A) and HPLC-grade acetonitrile + 0.1% formic acid (solvent B). Peptide synthesis

The synthesis of peptidyl-CoA substrates (L-1, D-1, L-2 and D-2, Scheme 1) and their characterization were performed using methods that have been previously described.18, 23, 26 Briefly, peptides were synthesized on Dawson-resin using Fmoc-based solid phase peptide synthesis and displaced from the column with MPAA to afford activated thioesters that, following side chain deprotection, were used to generate the desired peptidylCoAs via thioester exchange. Diastereomers of the peptidyl-CoAs were separated using preparative HPLC (L-1: tR = 22.6 min, D-1: tR = 23.8 min; gradient: 0 - 3 min 95% A, 3 – 33 min up to 30% B, flow rate 20 mL /min) and lyophilized. The purified peptides were subsequently dissolved in MilliQ water (10 mM), aliquoted and stored frozen at –80 °C. Cloning, expression and purification of PCP-Xtei, PCPXdbv, OxyBtei and OxyAtei

The cloning, expression and purification of PCP-Xtei (Tcp12, Uniprot ID Q70AZ6)5, OxyBtei: (Tcp20, Uniprot ID Q70AY8)24 and OxyAtei (Tcp18, Uniprot ID Q6ZZI8)18 was performed as previously described. Briefly, PCP-Xtei was amplified from a codon optimized tcp12 gene (Eurofins Genomics MWG) and cloned into a modified pET vector that allowed the expression of PCP-X with an N-terminal fusion partner (IgG-binding

ACS Paragon Plus Environment

Page 3 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

B1 domain of Streptococcus (GB1)) under the control of a T7-promotor.27 Purification was performed via affinity chromatography using an N-terminal hexahistidine-tag and a C-terminal Strep-tag. Size exclusion chromatography was performed as a final purification step in order to ensure the monomeric state of the protein. PCP-Xdbv (Dbv16, Uniprot ID Q7WZ75, selected sequence range amino acid 967 - 1506) was amplified from a codon optimized dbv16 gene (Eurofins Genomics MWG) using the primer pair (fwd: TATCCATGGCGCCGGATCGG GCCCAAGAATC; rev: ATATCTCGAGCGGGCGTTC GCGCTCGGCATC) thereby generating the restriction sites NcoI (5’) and XhoI (3’). PCP-Xdbv was cloned into the modified pET-GB1 vector (see cloning of PCP-Xtei). Expression and purification was performed as described for PCP-Xtei. The Oxy proteins from the teicoplanin system (OxyBtei and OxyAtei) were amplified from genomic DNA and cloned into a pET151D-TOPO vector (Life Technologies). OxyBtei and OxyAtei were expressed under the control of a T7-promotor equipped with an N-terminal hexahistidine-tag and a V5 epitope followed by TEV protease cleavage site. Purification was performed using metal affinity chromatography, removal of the N-terminal hexahistidine-tag via TEVcleavage, anion exchange chromatography and size exclusion chromatography as previously described for OxyBtei.24 Cloning, expression and purification of Dbv14 (OxyAdbv)

The OxyA protein from the A40926 producing system (Dbv14: Uniprot ID Q7WZ77) was derived from a codon optimized dbv14 gene (Eurofins Genomics MWG) that was synthesized including the desired restriction sites (NdeI and HindIII). The gene was cloned into a pET28a vector and the protein was expressed with an Nterminal hexahistidine-tag. Sequencing of the generated plasmid was performed with the standard T7 promotor and terminator sequencing primers. For expression, KRX E. coli cells were transformed with the pET28a plasmid containing the dbv14 gene. 7.5 L LB containing kanamycin (50 mg/L) were inoculated with an overnight culture of transformed KRX cells (1% v/v) and grown at 37°C. At an OD600 of 0.4 δ-aminolevulinic acid (0.5 mM) was added to the expression culture and the temperature was reduced to 18°C. The expression of Dbv14 was induced at an OD600 of 0.6 through the addition of 0.1 mM IPTG. The culture was grown overnight, the cells were harvested through centrifugation (5,000 x g, 4°C, 15 min) and the resulting cell pellet was resuspended in lysis buffer (50 mM Tris HCl pH 7.4, 50 mM NaCl, 10 mM imidazole, protease inhibitor cocktail (1 tablet for 100 mL lysis buffer, Sigma Aldrich)). Cell lysis was performed via four passes through a microfluidizer (Microfluidics) and the lysate was cleared by centrifugation (20,000 x g, 4°C, 60 min). The supernatant was used for immobilized-metal affinity chromatography (IMAC) in a batch procedure using 3.5 mL NiNTA Agarose (Macherey-Nagel). The elution fraction

was dialyzed overnight in 20 mM Tris, pH 7.4, 50 mM NaCl (buffer A) and further purified via anion exchange chromatography on a ResourceQ column (GE Healthcare) using the gradient: 0–80% buffer B (20 mM Tris, pH 7.4, 50 mM NaCl) over 20 column volumes. Size exclusion chromatography on a Superose12 column (GE Healthcare) was performed as a final polishing step with 50 mM Tris HCl pH 7.4, 100 mM NaCl used as the exchange buffer. The purity of the protein fractions was analyzed by SDS-PAGE and appropriate fractions were pooled, concentrated using Vivaspin ultracentrifugal filters (30,000 MWCO, Sartorius), flash cooled in liquid nitrogen and stored at -80°C. PCP-loading reaction

The PCP-X di-domain proteins were loaded with different peptidyl-CoA substrates (L-1-CoA, D-1-CoA, L-2CoA, D-2-CoA) using an engineered phosphopantetheinyl transferase from B. subtilis (Sfp R4-4).28 PCPX (60 µM) was incubated with 120 µM of the respective peptidyl-CoA substrate and 6 µM Sfp R4-4 in 50 mM Hepes pH 7.0, 50 mM NaCl and 10 mM MgCl2. After incubation for 1h at 30°C the excess of unloaded peptidyl-CoA was removed from the reaction by a concentration dilution procedure (4 x 1:5 dilution) using 0.5 mL ultracentrifugal filters (10,000 MWCO, Merck Millipore) and 50 mM Hepes pH 7.0, 50 mM NaCl as exchange buffer. The produced peptidyl-PCP-X constructs were used immediately as substrates for the P450 activity assay. P450 activity assay

For P450 activity assays 50 µM of the respective peptidyl-PCP-X substrate was incubated with of either OxyBtei alone or a combination of OxyBtei together with one of the OxyA enzymes (OxyAtei or OxyAdbv) using 2 µM for each enzyme in reaction buffer (50 mM Hepes pH 7.0, 50 mM NaCl). Electron transfer to the P450s was achieved by the addition of a functional redox chain comprising 5 µM palustrisredoxin B (PuxB variant A105V) and 1 µM palustrisredoxin reductase (PuR) from Rhodopseudomonas palustris29 as well as 2 mM NADH as the initial electron donor that was used to initiate the reaction. In addition to this, β-D-glucose (0.33% (w/v)) and glucose dehydrogenase (9 U/mL) were added, which allowed the constant regeneration of NADH throughout the turnover reaction. The reactions were incubated at 30°C with gentle shaking for 1 hr. Activity assay analysis

Following completion of activity assays, the reaction was quenched and the peptides cleaved from the PCP-X protein through addition of methylamine (32,000-fold molar excess over peptidyl-PCP-X, incubation for 15 minutes at room temperature). After neutralization of the solution through addition of formic acid (diluted in water) the cleaved peptides were purified by solid phase extraction (Strata-X-33 polymeric reversed phase col-

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

umns, 30 mg/mL, Phenomenex), before being analyzed via HPLC-MS. Specifically, the peptides were separated using the following gradient: 0–4 min 95% A, 4–4.5 min up to 15% B, 4.5–25 min up to 50% B. During the run the masses of the different peptide species were recorded using single ion monitoring (SIM) in negative ion mode. The signals were integrated and the P450 activities were determined based on the amount of cyclized peptide relative to the respective substrate. Due to the presence of chlorine isotopes within the peptides there is always an apparent background signal present in the analysis of turnovers: this corresponds to alternate isotopic peaks of the halogenated, cyclized peptides. However, the lowest mass always corresponds to the peptide in the highest state of crosslinking and with both 35Cl isotopes present and these species were used to determine the retention time of the products of both OxyB (from the reaction with OxyB alone) and OxyA (from a combined reaction of OxyB and OxyA) (Table 1). Table 1. Retention time of the main cyclized peptide products of the chlorinated (L-1/D-1) peptides during LCMS. L-1 peptide D-1 peptide

linear 16.31 min 16.37 min

C-O-D

C-O-D/ D-O-E

16.75 min 17.24 min

17.07 min 17.11 min

HRMS analysis of turnover products

Analyses were performed using a Dionex UltiMate 3000 RSLCnano system equipped with a Dionex UltiMate 3000 RS autosampler. Samples were loaded via an Acclaim PepMap 100 trap column (100 µm x 2 cm, nanoViper, C18, 5 µm, 100å; Thermo Scientific) onto an Acclaim PepMap RSLC analytical column (75 µm x 50 cm, nanoViper, C18, 2 µm, 100å; Thermo Scientific). The peptides were separated by increasing concentrations of 80% acetonitrile/ 0.1% formic acid at a flow of 250 nl/min for 15 min and analyzed with an Orbitrap Fusion Tribrid mass spectrometer (Thermo Scientific). Each cycle consisted of an Orbitrap full MS1 scan (resolution: 500.000; AGC target: 1e6; maximum IT: 118 ms; scan range: 375-1575 m/z) followed by up to 10 Orbitrap MS2 scans (resolution: 60.000; AGC target: 4e5; maximum IT: 118 ms; isolation window: 1.4 m/z; HCD Collision Energy: 32). To ensure high-quality MS2 spectra, the dynamic exclusion was set to 10 sec, which corresponded to the chromatographic peak width (FWHM). Acquired .raw files were analyzed in XCalibur Qual Browser (Thermo Scientific) using in-house generated layouts for extracted ion chromatograms (XICs). RESULTS In order to ascertain the effect of the halogenated peptides on Oxy activity, we initially synthesized a teicoplanin-like heptapeptide bearing chlorinated residues at both Tyr-6 and Tyr-2 residues in the peptide (1)

Page 4 of 19

(Fig. 3a, Scheme 1), which was achieved using our reported Fmoc-based SPPS synthesis route.23, 26 This peptide was synthesized as a CoA-conjugate, allowing it to be transferred onto the PCP-domain of PCP-X didomain proteins via the activity of the promiscuous phosphopantetheinyl transferase from B. subtilis (Sfp). As the stereochemistry of the 7th peptide residue has been shown to affect the activity of the second Oxy enzyme, OxyA, we further separated the diastereomers of the produced peptidyl-CoA (L-1/D-1, Scheme 1).17 We also synthesized the non-chlorinated version of the same peptide (2) in order to be able to test the effect of the chlorine substituents on Oxy activity. This peptide was also separated (as the peptidyl-CoA) to afford the diastereomers resultant from the racemization of peptide residue 7 (L-2/D-2, Scheme 1). Turnover experiments were performed as previously reported, initially with either OxyBtei or a combination of OxyBtei/OxyAtei included along with peptidyl-PCP-X substrate and redox partners to support P450 activity (Fig. 3a). Due to difficulties in obtaining kinetic parameters because to the protein bound nature of the substrates, we instead characterized the percentage activity of the Oxy enzymes. Following cleavage of the peptides from the PCP-domain through the use of methylamine and isolation via solid phase extraction, the activity of the Oxy enzymes was assessed using LCMS (Fig. 3a). During the analysis, chlorine isotope effects (caused by the presence of both 35Cl and 37Cl isotopes and the presence of two such atoms in 1) had to be taken into account: thus, the analysis of turnover was initially performed using a single quadrupole MS set to observe the mass for the lowest mass species formed by the Oxycatalyzed ring cyclization, i.e. the peptide containing two 35Cl isotopes. As the oxidative cyclization reaction generates a product 2 Dalton lower in mass than the substrate peptide, this method allows the product of OxyB (from an isolated reaction) and OxyA (from combined reactions of OxyB/OxyA) to be unambiguously identified and monitored. Nevertheless, the chlorine isotope effect results in the detection of signals corresponding to higher mass appearing at the same retention time. Due to these challenges in analysis, we also utilized high resolution mass spectral measurements, which possessed resolution high enough to differentiate between chlorine isotopes and the loss of 2 hydrogen atoms as a result of the cyclization reaction (Fig. 3a). In this way, we could not only confirm the mass of the products from the OxyB and OxyB/OxyA-catalyzed turnover of 1 but we could also fragment these ions and use these to unambiguously assign the presence of the C-O-D and C-O-D/D-O-E rings within the structure of the cyclized peptide reaction products (as these fragments share similar features to those reported for the cyclized, non-chlorinated peptides). These results confirm the activity of OxyBtei as insertion of the C-O-D ring between residues 4 and 6 of the peptide, and the

ACS Paragon Plus Environment

Page 5 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

activity of OxyAtei as insertion of the D-O-E ring between residues 2 and 4 of the peptide (SI Figs. 1-3). The results of our turnover experiments showed that OxyBtei was able to crosslink both diastereomers of the chlorinated peptide (71% of L-1 and 80% of D-1) (Fig. 3b, Table 2). Compared to the activities obtained for the non-chlorinated peptide 2 (66% of L-2 and 71% of D-2) no significant difference were observed (Table 2). This is in contrast to the results previously reported for OxyBvan where the presence of the chlorine substituent on the residue 6 in the respective substrate heptapeptide completely abolished peptide cyclization.22 However, these studies did not use substrates containing the Xdomain – a unique GPA NRPS domain – which has been shown to significantly improve the oxidative crosslinking of peptide substrates by a number of OxyB enzymes, including OxyBvan.5, 18, 23 Thus these differences suggest that the increase in bulk of chlorinated peptides decrease the efficacy of OxyBvan catalysis in the absence of the Xdomain, but have no effect on catalysis when the Xdomain is present.There was also no diastereoselectivity demonstrated by OxyBtei for the (L)-or (D)-forms of the halogenated heptapeptide 1. This matches what has been demonstrated for OxyBtei-catalyzed turnover of nonhalogenated peptides (including this study).17, 19 However, the effect of halogenation on the activity of OxyAtei was surprising: rather than leading to a diminished level of turnover activity, the presence of chlorine substituents on Tyr-2/Tyr-6 led to a high level of peptide bicyclization (Fig. 3b, Table 2). In addition, the product profile observed was dramatically simpler than that generate by turnover of the equivalent non-halogenated peptides 1 (Fig. 3a, c);5, 17, 18 this is demonstrated by the occurrence of fewer peaks relating to bicyclic peptide product, which now rather elutes as one major product peak. The level of cyclization of the non-natural peptide diastereomer (D-1) was reduced over that of the correct diastereomer (L-1), again matching previous data for the selectivity of this Oxy enzyme (Fig. 3b, Table 2). When compared to the level of conversion of the monocyclic peptide derived from D-2 by OxyAtei, the presence of the chlorine atoms had a negligible effect on the total level of bicyclization.17 Given that the addition of chlorine atoms to the peptide substrate 1 improved the level of in vitro activity of OxyAtei, we were curious as to what would happen to the activity of an OxyA homologue from a GPA system naturally lacking the chlorine substituent on Tyr-2. To this end, we focused on the OxyA homologue (Dbv14, henceforth referred to as OxyAdbv) from the A4092630 (dbv) GPA cluster,31 which is a Type-IV GPA (i.e. related to teicoplanin, Fig. 4a). A40926 contains the same peptide residues as teicoplanin but displays a different chlorination pattern (residues 3 and 6 for A40926 vs 2 and 6 for teicoplanin).4, 31, 32 A sequence comparison of OxyAtei and OxyAdbv indicated high levels of homology of the proteins (78% identity and 92% similarity); further-

more, by comparing the sequences together with the structure available for OxyAtei, it is clear that the active site of these enzymes is highly conserved (SI Fig. 4). Based on this comparison, we hypothesized that the turnover of 1 by OxyAdbv would follow a similar trend as that observed for OxyAtei. In order to test the activity of OxyAdbv against peptides 1 and 2, we first cloned, expressed and purified the enzyme from E. coli, using a comparable approach to that employed for OxyAtei. In order to provide the correct NRPS-protein to present the peptides to OxyAdbv, we also generated a PCP-X construct derived from the A40926 NRPS. Whilst we could express and purify this construct, we observed that this protein was prone to precipitation upon peptide loading. Hence, we used the well-behaved PCP-X construct from teicoplanin biosynthesis that we had used previously for turnover experiments with OxyAtei. Table 2. Oxy-catalyzed activity against chlorinated (L-1/D-1) and non-chlorinated teicoplanin-like peptides (L-2/D-2). OxyB

OxyA

Btei Btei Btei Btei Btei Btei Btei Btei Btei Btei Btei Btei

Atei Atei Atei Atei Adbv Adbv Adbv Adbv

a

peptidylPCP-Xtei L-1 D-1 L-2 D-2 L-1 D-1 L-2 D-2 L-1 D-1 L-2 D-2

C-O-D ring (%)a 71 ± 3 80 ± 5 66 ± 1 71 ± 2 73 ± 4 78 ± 4 90 ± 1 17 88 ± 1 17 73 ± 1 82 ± 1 66 ± 1 69 ± 1

D-O-E ring (%)a 81 ± 2 32 ± 1 64 ± 2 17 26 ± 1 17 58 ± 3 28 ± 1 40 ± 1 19 ± 1

Triplicate experiments including standard deviation.

With the OxyAdbv enzyme in hand, we initially characterized the activity of OxyAdbv for non-halogenated peptides (L-2/D-2) in combination with OxyBtei. We found that this OxyA homologue displayed ~70% of the in vitro activity demonstrated for OxyAtei against L-2 (40% vs 64%) (Fig. 4b, Table 2).17 This level of activity is a significant improvement over that displayed by other OxyA homologues characterized to this point, which include the enzymes StaF (~20%) and OxyAvan (