Biosynthesis of Translation Inhibitor Klebsazolicin Proceeds through

Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES

Biosynthesis of Translation Inhibitor Klebsazolicin Proceeds through Heterocyclisation and N-terminal Amidine Formation Catalysed by a Single YcaO Enzyme Dmitrii Y. Travin, Mikhail Metelev, Marina Serebryakova, Ekaterina Komarova, Ilya A Osterman, Dmitry Ghilarov, and Konstantin Severinov J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b02277 • Publication Date (Web): 30 Mar 2018 Downloaded from http://pubs.acs.org on March 30, 2018

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 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 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.

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 21 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

Journal of the American Chemical Society

Biosynthesis of Translation Inhibitor Klebsazolicin Proceeds through Heterocyclisation and N-terminal Amidine Formation Catalysed by a Single YcaO Enzyme Dmitrii Y. Travin1,2*, Mikhail Metelev2,3*‡, Marina Serebryakova2,4, Ekaterina S. Komarova1,2, Ilya A. Osterman4,5, Dmitry Ghilarov2,3,7§, and Konstantin Severinov2,3,6,7

1

Lomonosov Moscow State University, Department of Bioengineering and Bioinformatics, Moscow, 119992, Russia 2

Center for Data-Intensive Biomedicine and Biotechnology, Skolkovo Institute of Science and Technology, Skolkovo, 143025, Russia. 3

Institute of Gene Biology of the Russian Academy of Sciences, Moscow, 119334, Russia

4

Lomonosov Moscow State University, Department of Chemistry and A.N. Belozersky Institute of Physico-Chemical Biology, Moscow, 119992, Russia

5

Center for Translational Biomedicine, Skolkovo Institute of Science and Technology, Skolkovo, 143025, Russia. 6

Waksman Institute for Microbiology, Rutgers, The State University of New Jersey, Piscataway, NJ 08854, USA

7

Correspondence: [email protected] (D.G.), [email protected] (K.S.)

Dmitrii Y. Travin ([email protected]) Mikhail Metelev ([email protected]) Marina Serebryakova ([email protected]) Ilya A. Osterman ([email protected]) Ekaterina S. Komarova ([email protected]) Dmitry Ghilarov ([email protected]) Konstantin Severinov ([email protected])

*These authors contributed equally to the work ‡

Author’s current address: Department of Cell and Molecular Biology, Uppsala University,

Husargatan 3 Box 596, Uppsala, SE-75124, Sweden §

Author’s current address: Malopolska Centre of Biotechnology, Jagiellonian University, 30-387

Cracow, Poland 1 ACS Paragon Plus Environment

Journal of the American Chemical Society 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

Page 2 of 21

ABSTRACT Klebsazolicin (KLB) is a recently discovered Klebsiella pneumonia peptide antibiotic targeting the exit tunnel of bacterial ribosome. KLB contains an N-terminal amidine ring and four azole heterocycles installed into a ribosomally-synthesized precursor by dedicated maturation machinery. Using in vitro system for KLB production, we show that the YcaO-domain KlpD maturation enzyme is a bifunctional cyclodehydratase required for the formation of both the core heterocycles and the N-terminal amidine ring. We further demonstrate that the amidine ring is formed concomitantly with proteolytic cleavage of azole-containing pro-KLB by a cellular protease TldD/E. Members of the YcaO family are diverse enzymes known to activate peptide carbonyls during natural product biosynthesis leading to the formation of azoline, macroamidine, and thioamide moieties. The ability of KlpD to simultaneously perform two distinct types of modifications is unprecedented for known YcaO proteins. The versatility of KlpD opens up possibilities for rational introduction of modifications into various peptide backbones.

KEYWORDS Klebsazolicin, RiPPs, ribosome inhibitor, posttranslational modifications, YcaO-containing enzymes, TOMMs

2 ACS Paragon Plus Environment

Page 3 of 21 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


INTRODUCTION The development of DNA sequencing and annotation techniques in the last decades led to the discovery that multiple natural products originate from genetically encoded peptide precursors. Collectively, they are referred to as RiPPs: ribosomally synthesized and post-translationally modified peptides. RiPPs attract increased attention due to their wide spectrum of biological activities1 and as a potentially rich source of chemical scaffolds onto which sequences of interest can be grafted and/or modified by the action of enzymes involved in RiPPs biosynthesis.2 Klebsazolicin (KLB, 1) is a recently described member of linear azol(in)e-containing peptides (LAPs) – a heterogeneous subclass of RiPPs.3 The mature 23-aminoacid KLB contains three thiazole heterocycles, one oxazole heterocycle, and an N-terminal amidine ring (Figure 1A) introduced during post-translational modification of a ribosomally synthesized KlpA precursor peptide. KLB is a translation inhibitor that obstructs the exit tunnel of bacterial ribosome preventing the passage of the nascent peptide. Although several thiazole/oxazole-modified microcins (TOMMs, a subclass of RiPPs that includes LAPs) that target the ribosome are known, for example thiopeptides4 and bottromycin5, KLB is the only characterized LAP affecting translation, and its mechanism of inhibition is unique among known RiPPs. Three azoles and the N-terminal amidine ring of KLB are essential for activity as they mediate stacking and base-pair-like interactions with the 23S rRNA. Similarly to biosynthetic gene clusters of other RiPPs, the precursor peptide of KLB contains a core sequence, which is the subject for posttranslational modification, and an N-terminal leader sequence, which is recognized by the modification machinery. The KLB biosynthetic gene cluster klpABCDE is located in the chromosome of Klebsiella pneumoniae subspecies ozaenae (ATCC 11296). The precursor peptide is encoded by the klpA gene. Based on amino acid sequence similarity with proteins encoded in a wellstudied microcin B17 (MccB17, a DNA gyrase inhibiting LAP6,7)-producing gene cluster8, KlpC, KlpB and KlpD were annotated as E1-like partner protein, FMN-dependent dehydrogenase, and YcaO-domain-containing cyclodehydratase respectively. The product of the klpE gene is a putative ABC-transporter that exports mature molecule from the bacterial cell (Figure 1B). Genes coding for YcaO-domain (InterPro IPR019938) containing enzymes are found in more than 1500 bacterial genomes and are usually associated with TOMM biosynthetic gene clusters.9 The best-studied function of YcaO-containing enzymes is the formation of thiazoline, oxazoline, and methyloxazoline cycles, which are the result of ATP-dependent cyclodehydration of Cys, Ser, and Thr residues respectively.10 The proposed mechanism of this reaction includes nucleophilic attack of 3 ACS Paragon Plus Environment


Page 4 of 21

the side chain of Cys, Ser, or Thr onto the amide carbonyl of the adjacent amino acid residue with subsequent ATP-dependent elimination of the carbonyl-derived oxygen (Figure 1C).11 The order of cyclization and ATP-dependent backbone activation steps has not been yet directly shown.12 Frequently, for heterocyclization to take place YcaO needs to be associated with other proteins of the same pathway; for example, during MccB17 synthesis YcaO-containing McbD forms a complex with partner protein McbB and dehydrogenase McbC to catalyze formation of azole heterocycles.13,14 Several recent studies point to additional enzymatic activities of YcaO proteins. Genetic experiments followed by in vitro reconstruction of bottromycin biosynthetic pathway have shown that one of two YcaO-containing enzymes found in bottromycin biosynthetic gene cluster is required for the formation of the N-terminal macroamidine ring.15–17 Another activity of YcaO-proteins was proposed based on the analysis of thioviridamide biosynthetic gene cluster, where a standalone (having no partner proteins) YcaO protein TvaH is thought to catalyze the formation of peptide backbone thioamide bonds.18 Similarly, an ycaO gene product appeared to be responsible for posttranslational installation of thioglycine into archaeal methyl-coenzyme M reductase.19 Different at first glance, reactions of azoline formation, amidine macrocyclization and thioamide bond formation are believed to share a common mechanism, which includes a nucleophilic attack, activation of peptide backbones and likely proceeds through a hemiorthoamide intermediate10, while the nature of the nucleophile involved in the attack varies (Figure 1C). In the present study we reconstitute the entire pathway of KLB biosynthesis in vitro using recombinant KlpBCD enzyme complex. We demonstrate that KlpD is a bifunctional cyclodehydratase required for the formation of both the core cycles and the N-terminal amidine ring of KLB. Our study broadens the understanding of catalytic versatility of proteins of the YcaO superfamily, which may contribute to in vitro synthesis of new antibacterial compounds.


Page 5 of 21 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


Figure 1. A. Chemical structure of klebsazolicin. Posttranslational modifications are highlighted in color: Nterminal amidine cycle – blue, core azole cycles – red. B. klpABCDE biosynthetic gene cluster organization. Genes are colored in accordance with their predicted functions (listed below). The amino acid sequence of the product of klpA gene (KlpA precursor peptide) is shown with leader and core parts colored in orange and black respectively. Amino acid residues undergoing posttranslational modification are colored in accordance with the color of modifications in KLB chemical structure in panel A. C. Proposed general mechanism of reactions catalyzed by YcaO-containing enzymes. Known post-translational modifications installed by YcaO-enzymes are listed on the right.10



Page 6 of 21

RESULTS KlpBCD incorporates four azole heterocycles into the KlbA precursor peptide To reconstruct the biosynthesis of KLB in an in vitro system we purified the precursor KlpA peptide (2) by amylose-affinity chromatography of an MBP-KlpA fusion followed by TEV protease cleavage of the affinity tag (Figure S1C). KlpB, KlpC, and KlpD were co-purified from cells harboring a plasmid with klpB, klpC, and klpD genes placed under control of an inducible promoter. KlpC was tagged with an N-terminal hexahistidine tag and affinity purification allowed to obtain all three proteins in apparently stoichiometric amounts (Figure S1A, left), suggesting an existence of a stable complex, an inference which is further supported by an observation of a single peak at the gel filtration profile (Figure S1A, right). The treatment of purified 2 with KlpBCD complex in the presence of ATP produced a compound with four -20 Da (azole) cycles (3) (Figure 2C). In the absence of ATP in the reaction mixture 2 remained unchanged (Figure 2B); similarly, no modification of the substrate was observed when KlpBCDE49A complex containing a mutation in the predicted ATP-binding site (see Figures S1A and S2A) of putative YcaO-dependent cyclodehydratase KlpD was used (Figure 2E). KlpBY196ACD complex carrying a substitution in the active site of putative dehydrogenase KlpB (Figures S1A and S2B) in the presence of ATP partially converted the KlpA precursor into azolinecontaining compounds (Figure 2D). Combining two above mentioned mutant complexes in a single reaction allowed complete conversion of 2 to 3 (Figure 2F), indicating that azoline-containing substrate produced by one complex can be successively modified by the complementing enzyme complex or, alternatively, that KlpBCD complex subunits can be exchanged over the course of the reaction. Overall, our results indicate that KlpD is crucial for ATP-dependent azoline cycle formation (i.e., is a cyclodehydratase) while KlpC acts as a dehydrogenase that oxidizes azolines to azoles. These assignments are consistent with sequence homology-based bioinformatics predictions of the functions of the proteins encoded in the klpABCDE cluster. Thus, the KlpBCD complex behaves as a classical YcaO-dependent oxazole/thiazole synthetase. The incorporation of azole cycles into the structure of TOMMs’ precursor peptides by the action of cyclodehydratases can have N- to C-terminal directionality (e.g., in the case of MccB1720), C- to N-terminal (e.g., in the case of trunkamide21) or have no defined directionality. To study the sequence of cyclodehydratation events during KLB biosynthesis we stopped the in vitro precursor peptide modification reaction before reaching completion and analysed MS-MS spectra of


Page 7 of 21 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


intermediates having 0-4 core cycles. Figure S3 shows that, like MccB17, the KLB precursor peptide is modified from N- to C-terminus.

Figure 2. MALDI-MS analysis of products of in vitro reactions of precursor peptide KlpA (2, m/z = 4989 [M+H]+) with the wild type (panels B and C) and mutant (panels D-F) KlpBCD modification complexes. Peaks corresponding to 2, 2 with 4 azoles (3, m/z = 4909 [M+H]+) and 2 with one to three azolines (m/z = 4971, 4953, 4935 [M+H]+ ) are indicated. Each reaction was incubated for 60 minutes prior for MS analysis.



Page 8 of 21

TldD/E mediated proteolysis of the leader peptide is required for production of functional KLB When introduced in an E. coli host on a plasmid, the klpABCDE cluster is sufficient for the production a leaderless amidine ring-containing bioactive compound.3 In the MccB17 biosynthetic system, TldD/E protease is required for cleavage of the leader peptide after the posttranslational modification of the core peptide has been completed, resulting in the formation of a functional compound, which is exported from the cell.22,23 TldD/E is a heterodimeric Zn2+-dependent metalloprotease encoded in many bacterial and archaeal genomes; although the principal function of this protease remains unclear, in addition to its role on MccB17 production it takes part in degradation of the CcdA toxin23, and maturation of microcin C homologue from Yersinia pseudotuberculosis, a peptide-nucleotide that inhibits an aminoacyl-tRNA synthetase.24 To test whether TldD/E is implicated in leader peptide cleavage during biosynthesis of KLB in heterologous E. coli host, we monitored the production of the compound by ∆tldD and ∆tldE E. coli cells from Keio collection25 using wild-type cells as control. Cells from colonies grown on agar plates as well as the agar surrounding the colonies were analyzed for the presence of KLB by MALDI-MS. A prominent peak corresponding to a KLB precursor with 4 heterocycles was observed in ∆tldD or ∆tldE cells (Figure 3A) whilst only small amounts of the precursor were detected in wild-type cells. In contrast, a strong signal from mature KLB was detected in agar around wild-type cells while no such signal was observed around mutant strains. Thus, TldD/E mediates the removal of leader peptide from the KLB precursor, allowing the export of mature antibiotic outside the producing cell. Apparently, in tld mutants, the cleavage and export do not occur and so a precursor with four heterocycles accumulates inside the cell. When purified recombinant E. coli TldD/E (Figure S1B) was added to reactions containing 2, KlpBCD, and ATP, production of a mass-ion with m/z = 1974.8 [M+H]+ was detected (Figure 3B). High resolution MS (Figure S4A) and MALDI-ToF-MS-MS spectra (Figure S4B) of this compound were identical to mass-spectra of bioactive KLB purified from cultivation medium. The compound produced in vitro inhibited translation of luciferase mRNA in E. coli S30 extract similarly to KLB produced in vivo (Figure 3C) and demonstrated the same effect in an E. coli–based in vivo reporter system developed to screen inhibitors targeting either DNA replication or protein synthesis (Figure 3D).3,26 We conclude that in vitro synthesis of functional KLB from KlpA precursor peptide can be accomplished by KlpBCD modification complex and TldD/E protease in the presence of ATP.


Page 9 of 21 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


Figure 3. A. MALDI-MS analysis of whole-cells and surrounding agar (top and bottom spectra, respectively) of wild-type BW25113 (shown in red) and in ∆tldD and ∆tldE cells (shown in black) cells transformed with KLB production plasmid. Peaks of mature KLB (m/z = 1975 [M+H]+) and KLB precursor with leader peptide and four core azole cycles (m/z = 4545 [M+H]+) are indicated. Mass difference with 3 (m/z = 4909 [M+H]+) is the result of cloning-derived additional residues at the N-terminus of the leader peptide in 3. B. MALDI-MS spectra of KlpA precursor peptide (m/z = 4985.9 [M+H]+) and mature KLB (m/z = 1974.8 [M+H]+) emerging after KlpA precursor in vitro modification with KlpBCD complex in the presence of ATP and TldD/E protease. C. Kinetic curves showing luciferase mRNA translation inhibition in vitro by KLB obtained in vitro compared to KLB purified from the cultivation medium in three different concentrations (0.5 µM, 5µM, and 50µM). D. KLB obtained in vitro and KLB from the cultivation medium display similar size of inhibition zones and both induce the expression of Katushka2S protein, which indicates the inhibition of translation in E.coli cells transformed with the pDualrep2 plasmid.26 Erythromicin (Ery) and levofloxacin (Lev) are used as controls for translation inhibition and DNA damage respectively.



Page 10 of 21

YcaO-containing KlpD is essential for the formation of the N-terminal amidine ring To dissect the reaction of N-terminal amidine cycle formation independently from the incorporation of azole cycles into the core peptide, we purified intermediate 3 (Figure 2C) from an in vitro reaction and used it as a substrate for KlpBCD. Incubation of 3 with KlpBCD in the presence of ATP did not lead to additional modifications (Figure 4A). However, after the addition of TldD/E, we observed signals from mature KLB (1, m/z = 1974.8 [M+H]+) and a complementary leader peptide fragment (m/z = 2933 [M+H]+) (Figure 4C). When KlpBCD (Figure 4B) or ATP (Figure 4D) was omitted, KLB without the N-terminal amidine cycle 4, (m/z = 1992.8 [M+H]+) and the leader peptide peak were detected instead. The same result was obtained with KlpBCDE49A complex in the presence of ATP (Figure 4F). In contrast, dehydrogenase activity-deficient KlpBY196AСD afforded the production of mature KLB (Figure 4E). These results indicate that the formation of the amidine cycle requires ATP-dependent activity of the YcaO cyclodehydratase KlpD. Unlike the situation previously observed for MccB17 and microcin C homolog from Y. pseudotuberculosis, where the role of TldD/E is limited to removal of the leader peptide, in the case of KLB, TldD/E is apparently also required for the formation of the amidine cycle essential for bioactivity. To understand the role of TldD/E in amidine cycle formation, we substituted TldD/E for chymotrypsin. Chymotrypsin preferentially cleaves peptide bonds after large hydrophobic amino acids and therefore was expected to cleave C-terminal to a tyrosine residue at the KlpA leader-core junction (Figure 1B). As can be seen from Figure 5B, mature KLB was produced in an in vitro reaction containing intermediate 3, KlpBCD and ATP, while 4 was observed when ATP was omitted from the reaction (Figure 5C). Thus, TldD/E does not play a specific role in amidine cycle formation and can be substituted by other proteases capable of cleaving azole-containing KlpA peptide (3) at the leader-core junction. To test whether the N-terminal amidine cycle can be installed into the azole-containing precursor after TldD/E cleavage, we combined 4, HPLC-purified from the reaction displayed in the Figure 4B, with KlpBCD in the presence of ATP. No amidine cycle formation was observed. The addition of the leader peptide in trans also did not promote amidine cycle formation. Apparently, KlpA binding to the synthetase is mediated by the leader, which implies very high efficiency of amidine formation as after dissociation the substrate cannot rebind to the enzyme. It was suggested that leader peptide binding in many RiPPs is mediated by a conserved RiPP peptide recognition element (RRE)27 found, for example, in cyanobactin heterocyclase LynD28 and microcin C synthetase MccB29. We have investigated whether RREs can be identified within the 10 ACS Paragon Plus Environment

Page 11 of 21 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


KlpBCD synthetase subunits. HHpred analysis30 identified high (>95% probability) similarity between KlpD (residues 54-429) and YcaO domain of LynD (residues 243-775) as well as E. coli YcaO11; however, KlpD lacks the LynD N-terminal wHTH domain (domain 1) containing RRE. KlpB also does not contain sequences similar to known RREs. In contrast, HHpred analysis of KlpC reveals clear similarity (>95% probability score) to domains 1 and 2 (residues 1-270) of LynD and a segment of MccB that contains RRE. Thus, though direct biochemical data is absent, we conclude that KlpC is likely responsible for leader peptide binding by KlpBCD during both heterocyclisation and amidine synthesis.

Figure 4. A-F. Formation of amidine cycle in 3 (m/z = 4909 [M+H]+) in the presence of KlpBCD complex, TldD/E protease, and ATP. Left panels show MS spectra obtained in linear mode, right panels show parts of MS spectra of same compounds in m/z = 1970-2010 range obtained in reflection mode for better resolution. Peak with m/z = 1974.8 corresponds to mature KLB with an amidine cycle (1), peak m/z = 1992.8 corresponds to KLB without N-terminal modifications (4) and peak with m/z = 2933 is the complement peak of the leader peptide removed by TldD/E.



Page 12 of 21

Figure 5. MALDI-MS spectra of HPLC-purified intermediate 3 (panel A, m/z = 4906.9 [M+H]+) and the products of its in vitro treatment with KlpBCD modification complex in the presence of chymotrypsin substituting TldD/E protease (panels B and C). In the presence of ATP a peak of mature KLB with the N-terminal amidine cycle (1, m/z = 1974.8 [M+H]+) is observed, while without ATP there is no amidine modification detected (4, m/z = 1992.8 [M+H]+).


Page 13 of 21 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


Point mutations reveal KlpA residues essential for the formation of the amidine cycle KLB is a founding member of a family of LAPs which share a conserved N-terminal amino acid sequence in the core part3 (Figure S5). As an amidine ring is essential for ribosome binding and inhibition and the KLB homologs likely have common mechanism of action, we hypothesized that specific features of this sequence are required for installment of the amidine. To better understand sequence requirements of amidine cycle synthesis in KLB, we determined the effect of amino acid substitutions of the first four residues of KlpA precursor peptide core part (Ser1, Gln2, Ser3 and Pro4). Plasmids expressing KlpA mutants (Table S2) were introduced into cells carrying a compatible klpBCDE plasmid and production of KLB variants was monitored around the patches of the E. coli host cells by MALDI-MS analysis of surrounding agar. Some mutants (KlpAS1A, KlpAQ2A, KlpAS3A, KlpAS3T, KlpAS3C, and KlpAP4A) were also produced as MBP-fusions and their modification by KlpBCD complex was studied in the in vitro system (Figure 6). Substitutions of Ser1 and Gln2, two N-terminal amino acids that become part of the amidine cycle in the wild-type KLB (Figure 6H), had no impact on amidine formation (Figure 6B and 6C) despite the fact that their side chains are required for biological activity.3 In contrast, substitution of Ser3, which is not part of the amidine cycle, to amino acids other than Thr and Cys resulted in the loss of modification (Figure 6D, Table S2). The KlpAS3T mutant contained an amidine (-18 Da, m/z=1988.8 [M+H]+, Figure 6E) while KlpAS3C contained a thiazole (-20 Da, m/z=1988.8 [M+H]+, Figure 6F, Figure S6) formed at the site of the Gln2Cys3 dipeptide and an unmodified N-terminal amino acid. This thiazole heterocycle was formed prior to the leader peptide cleavage, as can be seen from an experiment where mutant precursor peptide was treated with KlpBCD complex in the absence of TldD/E protease (Figure S7B). The substitution of Pro4 also influenced the formation of the amidine cycle: the KlpAP4A mutant modification yielded a mixture of -18 Da (m/z=1948.8 [M+H]+) and -20 Da (m/z=1946.8 [M+H]+) products with the prevalence of amidine-containing compound (Figure 6G). In the absence of TldD/E, modification of KlpAP4A resulted in the formation of an extra azole in this position, similarly to KlpAS3C mutant, (Figure S7C), which was not observed for wild-type KlpA (Figure S7A) or in mutants other than KlpAS3C. The mechanistic implications of the results of KlpA mutagenesis are considered below in the Discussion section.



Page 14 of 21

Figure 6. A-G. MS analysis of the products of in vitro modification reactions containing mutant KlpA precursor peptides. Sequences of the first four N-terminal amino acid residues of core peptide used in modification reactions are shown in the left with substituted residues highlighted in red. Mass-spectra are aligned to match the type of N-terminal modification (azole, amidine or no terminal heterocycle) in reaction products. H. N-terminal region of mature KLB containing a formed amidine cycle, four first amino acid residues (subjected for site-specific mutagenesis) are represented in different colors showing the origin of amidine-forming atoms (from Ser1 and Gln2).


Page 15 of 21 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


DISCUSSION In this work, we reconstituted the entire biosynthetic pathway of klebsazolicin in vitro. The minimal KLB synthesis system consists of the precursor peptide KlpA, the KlpBCD synthetase complex, and the TldD/E protease. The first stage of KLB synthesis is the introduction of three thiazoles and one oxazole cycle in the core part of the KlpA peptide by the combined action of the YcaO-containing cyclodehydratase KlpD and dehydrogenase KlpB, as is typical for other TOMMs.10 In addition to its role in azole cycle formation, KlpD is also responsible for the biosynthesis of an Nterminal amidine cycle unique for KLB and required for its biological activity. While proteolytic cleavage of the leader peptide by TldD/E protease is needed for formation of fully-modified mature KLB both in vivo (in E. coli host) and in vitro, our data show that the role of TldD/E is limited to the generation of the primary amine and not to the catalysis of amidine formation per se, since an unrelated protease, chymotrypsin, can substitute for TldD/E and support the synthesis of mature KLB. One can assume that KlpD is a bifunctional YcaO enzyme capable of using different nucleophiles: sulfhydryl and hydroxyl groups of Cys and Ser side chains for biosynthesis of core azole cycles as well as the primary amino group for the formation of the N-terminal amidine. This scenario envisions direct mechanism of amidine formation (Figure 7A): a nucleophilic attack of the primary amino group on Ser1, generated after cleavage of 3 intermediate by TldD/E, onto the amide carbonyl of Gln2. A similar reaction was recently shown to be catalyzed by a YcaO protein during bottromycin biosynthesis.16,17 However, while two different YcaO enzymes – an azoline-forming and amidine-forming – participate in bottromycin synthesis, KlpD is solely responsible for both modifications in KLB. An attempt to dissect amidine biosynthesis by a site-specific mutagenesis of KlpA precursor (Figure 6) unexpectedly revealed the importance of the hydroxyl group on the side chain of Ser3, an amino acid adjacent to but not constituting a part of the amidine cycle. At first sight, this observation argues against the “amine nucleophile” mechanism presented in Figure 7A which does not anticipate any special role for this side chain. An alternative mechanism that is consistent with mutagenesis results is presented in Figure 7B: it includes an attack of the side chain of Ser3 on Gln2 carbonyl as in generally accepted31 mechanism of YcaO-dependent oxazoline formation (Figure 7B (i)) immediately followed by an attack carried by the N-terminal amino group (Figure 7B (ii)). This causes a rearrangement (Figure 7B (iii)) leading to the amidine cycle. In this scenario, N-terminal amine group happens to be in good attacking position with respect to an activated hemiorthoamide 15 ACS Paragon Plus Environment


Page 16 of 21

and does not require a specific recognition and activation by KlpD. The efficiency of proposed rearrangement appears to be controlled by the local sequence context. Indeed, substitution of Ser3 with cysteine, whose side chain contains a stronger nucleophile, results in predominant formation of a thiazole cycle prior to the leader peptide cleavage. Subsequent treatment of a KLB precursor containing five azoles with TldD/E leads to the cleavage in the expected place and release of a compound with an intact N-terminus. Substitution of flanking Pro4, which is conserved in precursor peptides of predicted KLB homologs from Pantoea ananatis strain PA4 and Paracoccus sp. 228, leads to a mixture of products containing the amidine or an oxazole formed by the side chain of Ser3. It thus appears that conformationally constrained proline located immediately downstream of the modification site decreases the ability of the KlpBCD complex to incorporate the oxazole cycle during KlpA modification, which was also observed for other TOMMs32,33. We propose that the presence of proline prevents the formation of an initial hemiorthoamide, which can be partially rescued by the presence of a more reactive cysteine residue (similar to what was observed previously for plantazolicin). Upon endoproteolytic processing at the leader-core junction site and resulting change in the conformation of the peptide chain the inhibition is released. However, an alternative explanation for importance of Ser3 might be due to specific recognition of side chains of Ser, Thr and Cys by the KlpBCD complex. According to this scenario, KlpBCD bypasses Ser3 due to steric constraints caused by the presence of a proline, unidirectionally incorporates core cycles, and then backtracks to take part in the amidine formation (see TOC scheme). The defined location of a peptide inside KlpD catalytic center is essential for the activation of N-terminal amine once it becomes available. Thus, here Ser3 is important for the repositioning of KLB biosynthetic machinery but not for the catalysis itself. This is reminiscent of the MccB17 system in which the formation of the ninth heterocycle is inefficient, requires a backtracking of enzyme and in vivo competes with the alternative reaction product (an ester bond).34 Be that as it may, the ability of KlpD to use two different nucleophiles, which have to be positioned differently in the catalytic center, is puzzling. Structural studies of KLB modification complex ongoing in our laboratory will shed light on the YcaO-containing KlpD active center, help further understand the process of the formation of amidine modification, and allow to harness the KLB biosynthetic machinery for introduction of desired modifications into various peptides.


Page 17 of 21 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


Figure 7. Possible mechanisms of YcaO-dependent N-terminal amidine cycle formation. A. A one-step bottromycin-like mechanism with nucleophilic attack of the N-terminal amino group. B. A two-step mechanism including an YcaO-dependent attack of Ser3 sidechain (i) prior to the attack of the Nterminal amino group (ii) and subsequent rearrangement (iii) resulting in the amidine cycle.

ACKNOWLEDGMENTS The work on KLB biosynthesis was supported by the RFBR-Royal Society International Exchanges scheme (KO165410043/IE160246) which allowed acquisition of high-resolution MS spectra of KLB by Gerhard Saalbach (John Innes Centre, Norwich, UK). Work in Russia was supported by the Russian Science Foundation (grant 18-44-04005) to I.A.O. (used to in vitro and in vivo activity tests) and Dynasty Foundation Fellowship (to M.M.). Study of TldD/E role in KLB maturation was supported by Russian Science Foundation RSF 16-14-10356 to Svetlana Dubiley. Access to MALDI-MS facilities was through the framework of the Moscow State University Development Program PNG 5.13. 17 ACS Paragon Plus Environment


Page 18 of 21

SUPPORTING INFORMATION Figures S1-S7, Tables S1 and S2, detailed description of all experimental procedures. This information may be found free of charge on the ACS website, http://pubs.acs.org.

REFERENCES (1)

Ortega, M. A.; van der Donk, W. A. New Insights into the Biosynthetic Logic of Ribosomally

Synthesized and Post-Translationally Modified Peptide Natural Products. Cell Chem. Biol. 2016, 23 (1), 31–44. (2)

van der Donk, W. A. Introduction: Unusual Enzymology in Natural Product Synthesis. Chem.

Rev. 2017, 117 (8), 5223–5225. (3)

Metelev, M.; Osterman, I. A.; Ghilarov, D.; Khabibullina, N. F.; Yakimov, A.; Shabalin, K.;

Utkina, I.; Travin, D. Y.; Komarova, E. S.; Serebryakova, M.; Artamonova, T.; Khodorkovskii, M.; Konevega, A. L.; Sergiev, P. V.; Severinov, K.; Polikanov, Y.S. Klebsazolicin Inhibits 70S Ribosome by Obstructing the Peptide Exit Tunnel. Nat. Chem. Biol. 2017, 13 (10), 1129–1136. (4)

Just-Baringo, X.; Albericio, F.; Álvarez, M. Thiopeptide Antibiotics: Retrospective and

Recent Advances. Mar. Drugs 2014, 12 (1), 317–351. (5)

Otaka, T.; Kaji, A. Mode of Action of Bottromycin A2: Effect of Bottromycin A2 on

Polysomes. FEBS Lett. 1983, 153 (1), 53–59. (6)

Parks, W. M.; Bottrill, A. R.; Pierrat, O. A.; Durrant, M. C.; Maxwell, A. The Action of the

Bacterial Toxin, Microcin B17, on DNA Gyrase. Biochimie 2007, 89 (4), 500–507. (7)

Vizán, J. L.; Hernández-Chico, C.; del Castillo, I.; Moreno, F. The Peptide Antibiotic

Microcin B17 Induces Double-Strand Cleavage of DNA Mediated by E. Coli DNA Gyrase. EMBO J. 1991, 10 (2), 467–476. (8)

Genilloud, O.; Moreno, F.; Kolter, R. DNA Sequence, Products, and Transcriptional Pattern of

the Genes Involved in Production of the DNA Replication Inhibitor Microcin B17. J. Bacteriol. 1989, 171 (2), 1126–1135. (9)

Cox, C. L.; Doroghazi, J. R.; Mitchell, D. A. The Genomic Landscape of Ribosomal Peptides

Containing Thiazole and Oxazole Heterocycles. BMC Genomics 2015, 16, 778. 18 ACS Paragon Plus Environment

Page 19 of 21 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


(10)

Burkhart, B. J.; Schwalen, C. J.; Mann, G.; Naismith, J. H.; Mitchell, D. A. YcaO-Dependent

Posttranslational Amide Activation: Biosynthesis, Structure, and Function. Chem. Rev. 2017, 117 (8), 5389–5456. (11)

Dunbar, K. L.; Chekan, J. R.; Cox, C. L.; Burkhart, B. J.; Nair, S. K.; Mitchell, D. A.

Discovery of a New ATP-Binding Motif Involved in Peptidic Azoline Biosynthesis. Nat. Chem. Biol. 2014, 10 (10), 823–829. (12)

Truman, A. W. Cyclisation Mechanisms in the Biosynthesis of Ribosomally Synthesised and

Post-Translationally Modified Peptides. Beilstein J. Org. Chem. 2016, 12 (Figure 1), 1250–1268. (13)

Milne, J. C.; Roy, R. S.; Eliot, A. C.; Kelleher, N. L.; Wokhlu, A.; Nickels, B.; Walsh, C. T.

Cofactor Requirements and Reconstitution of Microcin B17 Synthetase: A Multienzyme Complex That Catalyzes the Formation of Oxazoles and Thiazoles in the Antibiotic Microcin B17. Biochemistry 1999, 38 (15), 4768–4781. (14)

Li, Y. M.; Milne, J. C.; Madison, L. L.; Kolter, R.; Walsh, C. T. From Peptide Precursors to

Oxazole and Thiazole-Containing Peptide Antibiotics: Microcin B17 Synthase. Science 1996, 274 (5290), 1188–1193. (15)

Crone, W. J. K.; Vior, N. M.; Santos-Aberturas, J.; Schmitz, L. G.; Leeper, F. J.; Truman, A.

W. Dissecting Bottromycin Biosynthesis Using Comparative Untargeted Metabolomics. Angew. Chemie - Int. Ed. 2016, 55 (33), 9639–9643. (16)

Franz, L.; Adam, S.; Santos-Aberturas, J.; Truman, A. W.; Koehnke, J. Macroamidine

Formation in Bottromycins Is Catalyzed by a Divergent YcaO Enzyme. J. Am. Chem. Soc. 2017, 139 (50), 18158–18161. (17)

Schwalen, C. J.; Hudson, G. A.; Kosol, S.; Mahanta, N.; Challis, G. L.; Mitchell, D. A. In

Vitro Biosynthetic Studies of Bottromycin Expand the Enzymatic Capabilities of the YcaO Superfamily. J. Am. Chem. Soc. 2017, 139 (50), 18154–18157. (18)

Izawa, M.; Kawasaki, T.; Hayakawa, Y. Cloning and Heterologous Expression of the

Thioviridamide Biosynthesis Gene Cluster from Streptomyces Olivoviridis. Appl. Environ. Microbiol. 2013, 79 (22), 7110–7113.



(19)

Page 20 of 21

Nayak, D. D.; Mahanta, N.; Mitchell, D. A.; Metcalf, W. W. Post-Translational Thioamidation

of Methyl-Coenzyme M Reductase, a Key Enzyme in Methanogenic and Methanotrophic Archaea. Elife 2017, 6 (I), 1–18. (20)

Kelleher, N. L.; Hendrickson, C. L.; Walsh, C. T. Posttranslational Heterocyclization of

Cysteine and Serine Residues in the Antibiotic Microcin B17: Distributivity and Directionality. Biochemistry 1999, 38 (47), 15623–15630. (21)

Koehnke, J.; Bent, A. F.; Zollman, D.; Smith, K.; Houssen, W. E.; Zhu, X.; Mann, G.; Lebl,

T.; Scharff, R.; Shirran, S.; Botting, C. H.; Jaspars, M.; Schwarz-Linek, U.; Naismith, J. H. The Cyanobactin Heterocyclase Enzyme: A Processive Adenylase That Operates with a Defined Order of Reaction. Angew. Chemie - Int. Ed. 2013, 52 (52), 13991–13996. (22)

Ghilarov, D.; Serebryakova, M.; Stevenson, C. E. M.; Hearnshaw, S. J.; Volkov, D. S.;

Maxwell, A.; Lawson, D. M.; Severinov, K. The Origins of Specificity in the Microcin-Processing Protease TldD/E. Structure 2017, 25 (10), 1549–1561.e5. (23)

Allali, N.; Afif, H.; Couturier, M.; Van Melderen, L. The Highly Conserved TldD and TldE

Proteins of Escherichia Coli Are Involved in Microcin B17 Processing and in CcdA Degradation. J. Bacteriol. 2002, 184 (12), 3224–3231. (24)

Tsibulskaya, D.; Mokina, O.; Kulikovsky, A.; Piskunova, J.; Severinov, K.; Serebryakova, M.;

Dubiley, S. The Product of Yersinia Pseudotuberculosis Mcc Operon Is a Peptide-Cytidine Antibiotic Activated Inside Producing Cells by the TldD/E Protease. J. Am. Chem. Soc. 2017, 139 (45), 16178– 16187. (25)

Baba, T.; Ara, T.; Hasegawa, M.; Takai, Y.; Okumura, Y.; Baba, M.; Datsenko, K. a; Tomita,

M.; Wanner, B. L.; Mori, H. Construction of Escherichia Coli K-12 in-Frame, Single-Gene Knockout Mutants: The Keio Collection. Mol. Syst. Biol. 2006, 2, 2006.0008. (26)

Osterman, I. A.; Komarova, E. S.; Shiryaev, D. I.; Korniltsev, I. A.; Khven, I. M.; Lukyanov,

D. A.; Tashlitsky, V. N.; Serebryakova, M. V; Efremenkova, O. V; Ivanenkov, Y. A.; Bogdanov, A. A.; Sergiev, P. V.; Dontsova, O. A. Sorting Out Antibiotics’ Mechanisms of Action: A Double Fluorescent Protein Reporter for High-Throughput Screening of Ribosome and DNA Biosynthesis Inhibitors. Antimicrob. Agents Chemother. 2016, 60 (12), 7481–7489.


Page 21 of 21 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


(27)

Burkhart, B. J.; Hudson, G. A.; Dunbar, K. L.; Mitchell, D. A. A Prevalent Peptide-Binding

Domain Guides Ribosomal Natural Product Biosynthesis. Nat. Chem. Biol. 2015, 11 (8), 564–570. (28)

Koehnke, J.; Mann, G.; Bent, A. F.; Ludewig, H.; Shirran, S.; Botting, C. H.; Lebl, T.;

Houssen, W. E.; Jaspars, M.; Naismith, J. H. Structural Analysis of Leader Peptide Binding Enables Leader-Free Cyanobactin Processing. Nat. Chem. Biol. 2015, 11 (8), 558–563. (29)

Regni, C. A.; Roush, R. F.; Miller, D. J.; Nourse, A.; Walsh, C. T.; Schulman, B. A. How the

MccB Bacterial Ancestor of Ubiquitin E1 Initiates Biosynthesis of the Microcin C7 Antibiotic. EMBO J. 2009, 28 (13), 1953–1964. (30)

Söding, J.; Biegert, A.; Lupas, A. N. The HHpred Interactive Server for Protein Homology

Detection and Structure Prediction. Nucleic Acids Res. 2005, 33 (Web Server issue), W244-8. (31)

Dunbar, K. L.; Melby, J. O.; Mitchell, D. A. YcaO Domains Use ATP to Activate Amide

Backbones during Peptide Cyclodehydrations. Nat. Chem. Biol. 2012, 8 (6), 569–575. (32)

Deane, C. D.; Burkhart, B. J.; Blair, P. M.; Tietz, J. I.; Lin, A.; Mitchell, D. A. In Vitro

Biosynthesis and Substrate Tolerance of the Plantazolicin Family of Natural Products. ACS Chem. Biol. 2016, 11 (8), 2232–2243. (33)

Melby, J. O.; Dunbar, K. L.; Trinh, N. Q.; Mitchell, D. A. Selectivity, Directionality, and

Promiscuity in Peptide Processing from a Bacillus Sp. Al Hakam Cyclodehydratase. J. Am. Chem. Soc. 2012, 134 (11), 5309–5316. (34)

Ghilarov, D.; Serebryakova, M.; Shkundina, I.; Severinov, K. A Major Portion of DNA

Gyrase Inhibitor Microcin B17 Undergoes an N,O-Peptidyl Shift during Synthesis. J. Biol. Chem. 2011, 286 (30), 26308–26318.

TOC graphic.


Biosynthesis of Translation Inhibitor Klebsazolicin Proceeds through

Recommend Documents