The Product of Yersinia pseudotuberculosis mcc ... - ACS Publications

Oct 18, 2017 - Institute of Gene Biology, Russian Academy of Science, 34/5 Vavilov str., 119334 Moscow, Russia. ‡. Center for Data-Intensive Biomedi...
0 downloads 3 Views 5MB Size
Download PDF
Article Cite This: J. Am. Chem. Soc. 2017, 139, 16178-16187

pubs.acs.org/JACS

The Product of Yersinia pseudotuberculosis mcc Operon Is a PeptideCytidine Antibiotic Activated Inside Producing Cells by the TldD/E Protease Darya Tsibulskaya,†,‡ Olga Mokina,‡,† Alexey Kulikovsky,‡,†,# Julia Piskunova,‡,† Konstantin Severinov,*,‡,†,§ Marina Serebryakova,†,∥ and Svetlana Dubiley‡,† †

Institute of Gene Biology, Russian Academy of Science, 34/5 Vavilov str., 119334 Moscow, Russia Center for Data-Intensive Biomedicine and Biotechnology, Skolkovo Institute of Science and Technology, 3 Nobel str., 143026 Moscow, Russia § Waksman Institute for Microbiology, 190 Frelinghuysen Road, Piscataway, New Jersey 08854-8020, United States ∥ A.N. Belozersky Institute of Physico-Chemical Biology, Lomonosov Moscow State University, Leninskie Gory 1, Bldg. 40, Moscow 119991, Russia # Department of Biochemistry, University of Illinois at Urbana-Champaign, 600 S. Mathews Ave., Urbana, Illinois 61801, United States ‡

S Supporting Information *

ABSTRACT: Microcin C is a heptapeptide-adenylate antibiotic produced by some strains of Escherichia coli. Its peptide part is responsible for facilitated transport inside sensitive cells where it is proteolyzed with release of a toxic warheada nonhydrolyzable aspartamidyl-adenylate, which inhibits aspartyl-tRNA synthetase. Recently, a microcin C homologue from Bacillus amyloliquefaciens containing a longer peptide part modified with carboxymethylcytosine instead of adenosine was described, but no biological activity of this compound was revealed. Here, we characterize modified peptide-cytidylate from Yersinia pseudotuberculosis. As reported for B. amyloliquefaciens homologue, the initially synthesized compound contains a long peptide that is biologically inactive. This compound is subjected to endoproteolytic processing inside producing cells by the evolutionary conserved TldD/E protease. As a result, an 11-amino acid long peptide with C-terminal modified cytosine residue is produced. This compound is exported outside the producing cell and is bioactive, inhibiting sensitive cells in the same way as E. coli microcin C. Proteolytic processing inside producing cells is a novel strategy of peptide−nucleotide antibiotics biosynthesis that may help control production levels and avoid toxicity to the producer.

■

INTRODUCTION

McC is a founding member of a family of antibiotics encoded by numerous mcc-like operons in diverse Gram-negative and Gram-positive bacteria.6 Most of validated mcc-like gene clusters contain just three genes, mccABC, which is sufficient to produce a toxic adenylated peptide and export it outside the producing cells.7 The E. coli mcc gene cluster contains 6 genes (Figure 1). The products of the mccDEco and mccEEco genes are jointly responsible for decoration of the phosphate of adenylated MccAEco with aminopropyl group,8 increasing the bioactivity.9 MccEEco is a bifunctional protein, which participates in aminopropylation with its N-terminal decarboxylase domain8 and provides self-immunity by acetylating processed McC with its C-terminal domain.10 The product of mccFEco contributes to self-immunity by cleaving the phosphoramidate bond connecting the terminal aspartamide and AMP.11

Microcin C (McC) is a peptide-adenylate produced by some strains of Escherichia coli and other enteric bacteria carrying the mcc gene cluster.1 The ribosomally synthesized McC precursor heptapeptide MRTGNAN, the product of mccAEco gene, is adenylated by MccBEco, a THIF-type adenylyltransferase. The reaction proceeds in two steps: ATP-dependent formation of peptide-succinimide intermediate followed by hydrolysis of succinimide ring and the attachment of AMP to resulting aspartamide through a nonhydrolyzable phosphoramidate bond.2 The peptide part of McC is responsible for export from the producing cell through the MccCEco pump and import into sensitive cells through the YejABEF transporter.3 Inside the sensitive cells the peptide part is degraded by aminopeptidases releasing “processed McC”a nonhydrolyzable aspartamidyl-adenylate4 that binds aspartyl-tRNA synthetase (Asp-RS) and inhibits it, leading to cessation of protein synthesis.5 © 2017 American Chemical Society

Received: July 8, 2017 Published: October 18, 2017 16178

DOI: 10.1021/jacs.7b07118 J. Am. Chem. Soc. 2017, 139, 16178−16187

Article

Journal of the American Chemical Society

residue of cognate MccABam.12 MccSBam is a carboxy-Sadenosyl-L-methionine (cxSAM) synthetase. The cxSAM is used by MccBCTDBam as a donor of carboxymethyl group for additional modification of MccABam-CMP. Though no biological activity for McCBam was revealed, modified aspartamidylcytidylate that is released after processing of chimeric compound based on E. coli MccA peptide inhibits Asp-RS. The presence of carboxymethyl modification increases the toxicity and inactivates the self-immunity/resistance mechanism mediated by MccEEco.12 Here, we have investigated McC Yps , a product of Y. pseudotuberculosis IP 32953 mcc operon. We show that upon modification with cytidine and propylamine, the 42 amino acid long peptide part of pro-McCYps is processed inside the producing cell releasing bioactive McCYps whose peptide part is just 11 amino acids long. The processing is carried out by an evolutionary conserved TldD/E protease. Upon processing, McCYps leaves the producing cell, is actively taken up by the YejABEF transporter in sensitive cells and demonstrates toxicity comparable with that of McCEco. Proteolytic processing inside producing cells is a novel strategy of peptide−nucleotide antibiotics biosynthesis that may help control production levels and avoid toxicity to the producer.

Figure 1. Comparison of mcc gene clusters from E. coli, B. amyloliquefaciens DSM7, and Y. pseudotuberculosis IP 32953. Genes are indicated by arrows (letters indicate gene names, i.e., “B” stands for “mccB”), same colors show homologous genes. MccE of E. coli is a bifunctional protein and is encoded by two adjacent genes in Y. pseudotuberculosis IP 32953 clusters. For E. coli and Y. pseudotuberculosis IP 32953 clusters, positions of transcription terminators are schematically shown as hairpin. Known (E. coli, B. amyloliquefaciens DSM7) or predicted (Y. pseudotuberculosis IP 32953) mcc genes products functions are listed at the bottom. RRE, RiPP precursor peptide recognition element; MFS, Major Facilitator Superfamily; cxSAM, carboxy-SAM.

■

RESULTS Y. pseudotuberculosis mcc Operon Encodes a Compound with Antibacterial Activity. A putative mcc-like biosynthesis/self-immunity gene cluster present in Yersinia pseudotuberculosis IP32953 combines characteristics of both E. coli and B. amyloliquefaciens mcc operons (Figure 1). The Y. pseudotuberculosis MccA peptide (MccAYps) is 42 amino acids long and contains a terminal asparagine. The mccAYps gene is located upstream of mccCYps, which encodes a predicted export pump of the Major Facilitator Superfamily (YPTB1884). The next gene is mccBYps (YPTB1885). Recombinant MccBNTDYps was shown to adenylate MccAYps in vitro.7 Similar to MccBBam

The mccD, mccE, and mccF homologues are present in only a handful of non-E. coli mcc clusters. On the other hand, some mcc-like operons contain additional genes that are absent from the E. coli operon.7 A group of mcc-like clusters from Bacillus amyloliquefaciens DSM7, Streptococcus bovis JB1, and various strains of Yersinia sp. and Serratia sp. encode MccB homologues extended with a C-terminal methyltransferase domain and also contain an additional gene mccS. In a recent study of B. amyloliquefaciens McC-like compound we have shown that the N-terminal part of B. amyloliquefaciens MccB, MccBNTDBam, performs cytidylation instead of adenylation of the terminal Asn

Figure 2. E. coli cells carrying plasmid-borne mcc operon from Y. pseudotuberculosis IP 32953 produce an antibacterial compound(s) in the absence of the hns gene. (A) Growth inhibition by Y. pseudotuberculosis #3526 and wild-type or Δhns E. coli BW25113 cells harboring pDT1-Ymcc or control pDT1 plasmids on E. coli B cell lawns. Results of overnight growth at 30 °C are shown. A growth inhibition zone is only visible around the patch of Δhns E. coli BW25113 transformed with pDT1-Ymcc. (B) Expression levels of individual mccYps genes in Y. pseudotuberculosis #3526 and wild-type or Δhns E. coli BW25113 cells harboring pDT1-Ymcc. Results of RT-PCR with primer pairs specific for indicated genes are presented. Expression levels were normalized to that of kanamycin resistance aph gene of pDT1-Ymcc. 16179

DOI: 10.1021/jacs.7b07118 J. Am. Chem. Soc. 2017, 139, 16178−16187

Article

Journal of the American Chemical Society

Figure 3. Identification of products of full-sized and partial Y. pseudotuberculosis IP 32953 mcc operons accumulating inside E. coli Δhns cells. (A) MALDI-MS spectra of BW25113 Δhns cells harboring control pDT1 vector (upper panel) and pDT1-Ymcc (lower panel) plasmids and the corresponding growth inhibition zones formed on lawns of E. coli B tester cells. Mass-peaks that are common to both control and experimental cells match E. coli ribosomal protein RS22 (m/zcalc 5097) and double-charged ion of acidic stress chaperones HdeA (m/zcalc 4870) and HdeB (m/zcalc 4532). The m/z values of difference mass-peaks of MccAYps peptide and its modified forms are highlighted with red-color font (see text for details). (B) Mass-spectra of Δhns E. coli cells transformed with various pDT1-Ymcc plasmid derivatives harboring indicated truncations of the Y. pseudotuberculosis IP 32953 mccACBXDE1E2 operon and growth inhibition zones on lawns of E. coli B tester cells formed around patches of same cells are shown. (C) The biosynthesis pathway of the product of Y. pseudotuberculosis IP 32953 mcc operon. The part of the pathway that has been reconstituted in vitro is highlighted by green shading. The last two steps without shading are analogous to previously characterized enzymatic maturation steps of E. coli microcin C8 and are carried out by homologous Y. pseudotuberculosis gene products. R = MHQSEIKLTKRLKIKRVDVNKVKEQQKKVLECGAATCGGGS.

and unlike MccBEco, MccBYps contains, in addition to a nucleotydyltransferase part, a C-terminal SAM-dependent methyltransferase domain. The mccBYps gene is followed by mccSYps (YPTB1886), a homologue of mccSBam. The following gene, mccXYps, encodes a protein of unknown function (YPTB1887). MccD Yps (YPTB1888) is homologous to MccDEco, which uses SAM to transfer carboxylated propylamine onto the phosphate of heptapeptide adenylate produced by MccBEco. The Y. pseudotuberculosis MccE homologue is split into two separate polypeptides encoded by adjacent genes. MccE1Yps (YPTB1889), the homologue of MccENTDEco, is likely a decarboxylase that removes the carboxyl from the carboxylated propylamine transferred by MccD.8 MccE2Yps

(YPTB1890) is a GNAT-type acetyltransferase expected to be involved in self-immunity. To check if any McC-like compounds are produced by Y. pseudotuberculosis IP 32953 carrying the mccYps operon in its genome, we monitored the ability of these cells to inhibit growth of McC-sensitive E. coli or Y. pseudotuberculosis #3526 strain that does not carry the mccYps operon in its genome. No growth inhibition zones were observed and we therefore cloned the entire mccYps cluster along with the upstream region that should contain a promoter into a multicopy pDT1 plasmid vector and introduced the resulting pDT1-Ymcc plasmid in Y. pseudotuberculosis #3526 or wild-type E. coli. Massspectrometric analysis of pDT1-Ymcc-containing Y. pseudotu16180

DOI: 10.1021/jacs.7b07118 J. Am. Chem. Soc. 2017, 139, 16178−16187

Article

Journal of the American Chemical Society berculosis #3526 cells did not reveal any additional peaks compared to control strain transformed with the pDT1 vector. In contrast, in the mass-spectrum of E. coli cells harboring the pDT1-Ymcc plasmid, two mass-peaks that were absent from cells transformed with pDT1 and that corresponded to unmodified full-sized MccAYps peptide and its N-terminal 31amino acid fragment were clearly seen (Figure S1). Despite the presence of MccAYps product in E. coli cells transformed with pDT1-Ymcc, no inhibition of microcin C-sensitive E. coli tester cells growth was observed (Figure 2A). Considering the length of the mccYps operon, we wondered if all of its genes were expressed in the heterologous E. coli host. To this end, real-time RT-PCR reactions with primer pairs specific for every gene of plasmid-borne mccYps operon were performed for RNA prepared from wild-type E. coli. Expression levels were normalized to that of pDT1-Ymcckanamycin-resistance gene aph. As can be seen from Figure 2B, there was strong expression of the mccAYps gene. However, expression of downstream mccCYps and mccBYps genes was very low. Since mccBYps encodes a nucleotydyltransferase responsible for MccAYps modification,7 the result may explain the presence of unmodified MccAYps in E. coli cells transformed with pDT1Ymcc as revealed by mass-spectrometry. Analysis of RNA prepared from Y. pseudotuberculosis #3526 transformed with pDT1-Ymcc revealed much more uniform mccYps genes expression. However, the level of mccAYps expression was significantly lower than in E. coli, which may explain the lack of formation of growth inhibition zones around these cells (Figure 2A). We tested an E. coli Δhns, the mutant lacking a DNA-binding protein known to be involved in global negative control of expression of horizontally transferred genes as a production host.13 Though expression of mccAYps was decreased several fold compared to levels seen in wild-type E. coli in this host (Figure 2B), expression levels of mccCYps and mccBYps were strongly increased. Moreover, clear growth inhibition zones on lawns of microcin C-sensitive E. coli B tester cells were observed around patches of E. coli Δhns cells carrying pDT1-Ymcc (Figure 2A). Comparison of mass spectra of E. coli Δhns cells with and without plasmid-borne mccYps operon revealed several difference mass-peaks that were present only in cells transformed with pDT1-Ymcc (Figure 3A). In addition to m/z 4656 peak matching unmodified MccAYps peptide, there was a m/z 4638 peak matching its succinimide intermediate, and an additional m/z 5076 peak (Figure 3A). MS/MS analysis of 5076 peak (Figure S2) revealed a major fragmentation mass-ion with an m/z of 4656 matching unmodified MccAYps peptide. Thus, the compound of the 5076 peak consists of the MccAYps peptide, likely modified by the products of the mccYps operon, and this modification(s) adds 420 Da. The Biosynthetic Pathway of Y. pseudotuberculosis mcc Operon Product and Identification of Genes Involved in Production of Toxic Compound. To define the roles of individual mccYps genes in MccAYps modification, derivatives of pDT1-Ymcc were constructed by sequentially deleting mccYps operon genes. The E. coli Δhns cells harboring these plasmids were tested for biological activity and subjected to mass-spectrometric analysis (Figure 3B). Cells harboring a plasmid with the largest deletion that left only the minimal three-gene arrangement (mccACB) did not inhibit growth of McC-sensitive E. coli. In addition to unmodified MccAYps these cells accumulated a m/z 4638 and m/z 4961 compounds corresponding to MccAYps-succinimide and cytidylated MccAYps

(MccAYps-CMP) (Figure 3C). The identity of the later compound was confirmed by MS/MS analysis (Figure S2). Thus, while earlier in vitro work showed that MccBYps attaches AMP to its cognate peptide,7 in vivo, CMP is attached. A similar situation was observed for MccB encoded by Bacillus amiloliquefaciens.12 A peak with m/z 5019 appeared in cells harboring a four-gene plasmid (mccACBS). This compound corresponds to carboxymethylated form MccAYps-cxCMP (Figure S2), an expected product based on the analysis of Bacillus amiloliquefaciens mcc operon.12 No growth inhibition zones were observed around patches of these cells. The compound of the same mass was synthesized when five genesmccACBSXwere present on the plasmid, suggesting that MccXYps does not participate in modifications that lead to a mass-shift of the product MccB/MccS action. There was no biological activity evident for cells transformed with mccACBSX plasmid. When the plasmid-borne cluster contained mccDYps but lacked the downstream genes, the biological activity was still lacking. A new peak with average m/z of 5120 was observed. The E. coli MccD protein attaches carboxylated propylamine to cognate peptide adenylate, adding extra 101 Da.8 The mass difference between MccAYps-cxCMP (m/z 5019) and the m/z 5120 compound also equals 101 Da, suggesting that carboxylated propylamine was attached to the phosphate group of MccAYps-cxCMP by MccDYps. This interpretation is supported by MS/MS analysis presented in Figure S2. When the plasmid-borne mccYps contained the ACBSXDE1 genes, growth inhibition of E. coli tester cells was finally observed and a mass-ion with an average m/z of 5076 was detected in massspectra of cells. The same mass-ion is detected in cells with pDT1-Ymcc harboring complete mccYps cluster (Figure 3A). In E. coli, the N-terminal domain of bifunctional MccE protein decarboxylates the carboxypropylamine group, resulting in a loss of 44 Da.8 The mass difference between the 5120 compound present in cells carrying ACBSXD genes and the m/ z 5076 compound is the same. Overall, the results of our analysis are consistent with a biosynthesis pathway involving the initial attachment of CMP to the MccAYps peptide followed by carboxymethylation and aminopropylation to produce a fully modified, biologically active compound (Figure 3C). We attempted to reconstruct the entire biosynthesis pathway in vitro using chemically synthesized MccAYps and recombinant modification enzymes (Figure S3). MccBYps selectively attached CMP to MccAYps when supplied with an equimolar mixture of NTPs, producing MccAYps-CMP (m/z 4961). In the presence of MccBYps, MccSYps, SAM, and prephenate, MccAYps-cxCMP (m/z 5019) was produced. The same product was obtained in the presence of MccBYps only when reactions were supplemented with cxSAM (Figure S3). Thus, MccS Yps role in the MccAYps modification pathway is limited to the production of carboxySAM from SAM and prephenate. Further in vitro modification of MccAYps-cxCMP impossible since recombinant MccDYps protein was insoluble. The steps of modification of MccAYps that have been reconstituted biochemically are highlighted by green shading in the scheme of Figure 3C. The last two steps are proposed based on analogy with the E. coli microcin C maturation pathway8 and are supported by analysis of intermediates produced in vivo (Figure 3B) and homology between MccD and MccE enzymes of E. coli and Y. pseudotuberculosis. The Bioactive Product of Y. pseudotuberculosis mcclike Operon Is Different from the Compound That 16181

DOI: 10.1021/jacs.7b07118 J. Am. Chem. Soc. 2017, 139, 16178−16187

Article

Journal of the American Chemical Society

Figure 4. Identification of bioactive Y. pseudotuberculosis IP 32953 mcc operon product. (A) HPLC traces of cultured media of cells harboring pDT1Ymcc (red trace) and control cells with empty pDT1 vector (blue trace). Results of bioactivity testing of indicated HPLC fractions on lawns of E. coli B cells are shown below. “Gm” shows the results of growth inhibition zones formed around drops of gentamicin. HPLC peak with antibiotic activity is labeled “McCYps”. (B) MALDI-MS spectra of bioactive HPLC fraction 5 from panel A. The bottom spectrum was obtained after preincubation with 2 mM DTT. (C) Fragmentation spectra of m/z 1315.4 and m/z 1331.4 mass-ions from the upper spectrum in panel B and a structure of a compound whose fragmentation pattern is consistent with the top spectrum. “D-NH2”, aspartamide; “cxC”, carboxymethylated cytosine; “Rib.”, ribose; “ap.”, aminopropyl group; “HPO3+ap.”, aminopropylated phosphate; “cxC+O”, oxidized form of carboxymethylated cytosine. On the structure, sites of fragmentation that generate experimentally observed mass-peaks are indicated.

Accumulates in Producing Cells. Mass-spectrometric analysis of conditioned media after growth of E. coli Δhns cells harboring pDT1-Ymcc failed to reveal a mass-peak of m/z 5076 corresponding to fully modified MccAYps. The cultured medium was therefore fractionated by reverse phase HPLC and antibiotic activity in the fractions was followed. A single fraction

that inhibited the growth of wild-type E. coli was obtained (Figure 4A). Two prominent mass-peaks with m/z of 1315 and 1331 were present in this fraction (Figure 4B, upper panel). Treatment with dithiothreitol led to a mass-shift of 2 Da (to m/ z of 1317 and 1333) corresponding to disulfide bond reduction 16182

DOI: 10.1021/jacs.7b07118 J. Am. Chem. Soc. 2017, 139, 16178−16187

Article

Journal of the American Chemical Society

Figure 5. tld genes are required for McCYps production. (A) MALDI-MS spectra of E. coli Δhns cells harboring empty pDT1 vector (top) or the pDT1-Ymcc plasmid (middle), and Δhns E. coli ΔtldD ΔtldE derivative harboring pDT1-Ymcc (bottom). The peaks of full-length MccAYps modified peptide (m/z 5076) and its N-terminal proteolytic fragments 1−31(m/z 3777) and 1−32 (m/z 3880) are marked with red-color font. Mass-peaks present in control and experimental cells match E. coli ribosomal proteins RL36 (m/zcalc 4364) and RS22 (m/zcalc 5097). (B) Growth inhibition zones on the lawn of E. coli B cells formed around patches of Tld+ (E. coli Δhns) and Tld− (E. coli Δhns ΔtldE ΔtldE) strains harboring pDT1 or pDT1-Ymcc plasmids. (C) An HPLC trace of cellular extract of E. coli Δhns ΔtldD ΔtldE cells harboring pDT1-Ymcc. HPLC peak labeled “proMcCYps” contains a major mass ion with m/z 5076 corresponding to fully modified MccAYps peptide. (D) 5-μL drops of 40 μM solutions of McCYps and pro-McCYps) purified from Tld− E. coli were deposited on a fresh lawn of E. coli B cells. Growth inhibition zones appearing after 16 h incubation at 30 °C are shown. A 5-μL drop of 0.5 mg/mL gentamycin (Gm) was used as a positive control.

As is shown in Figure 2A, Y. pseudotuberculosis strain #3526 with the pDT1-Ymcc plasmid did not cause growth inhibition of E. coli tester cells. However, since there was detectable level of mccYps gene expression (Figure 2B) we fractionated medium conditioned by growth of these cells and obtained a fraction with biological activity (Figure S5). MALDI-MS analysis revealed that both active fractions contained m/z 1315 and m/z 1331 compounds (Figure S5). MS/MS analysis of m/z 1315 ion confirmed its identity to McCYps produced in E. coli. We take this result as a strong indication that the product of Y. pseudotuberculosis mcc operon is endoproteolytically processed in the natural host. The TldD/E Protease Is Responsible for Endoproteolytic Processing of Pro-McCYps Inside Producing Cells. The finding that active McCYps is distinct from the compound that accumulates in the producing cells raises a question about the mechanism by which active compound with shorter peptide part is generated. Inspection of mass spectra of Δhns cells carrying pDT1-Ymcc and its derivatives revealed a prominent peak with an average m/z 3777 that corresponded to Nterminal 31-amino acid fragment of MccAYps (Figure 5A, middle panel). The same fragment is detected in wild-type E. coli transformed with pDT1-Ymcc (Figure S1). This fragment must arise from a proteolytic event that generates McCYps. To identify a protease(s) involved in MccAYps cleavage, a set of mutants in known and putative E. coli protease genes present in the Keio collection of single-gene E. coli mutants14 was transformed with pDT1-based plasmid harboring the mccACB genes and screened for the presence of m/z 3777 mass-ion. Cells lacking either the tldD or tldE genes coding for

for both peaks (Figure 4B, lower panel). We were unable to separate these peaks by additional chromatographic steps. Material from bioactive fraction was subjected to MS/MS analysis. A daughter ion with m/z 897 of the 1315 mass-ion, matched the C-terminal part of the MccAYps peptide, CGAATCGGGSN (Figure 4C). This assignment was supported by signals resulting from fragmentation of the peptide part. Daughter ions with m/z 1034 and 1148 matched CGAATCGGGSN-aminopropyl-phosphate and CGAATCGGGSN-aminopropyl-phosphate-ribose fragments. The loss of 169 Da during fragmentation corresponds to removal of carboxymethyl-cytidine nucleobase. Fragmentation of the 1331 mass-ion resulted in accumulation of identical daughter ions (Figure 4C). Thus, the 16 Da difference between 1315 and 1331 compounds resides on the cytidine base. Overall, the results demonstrate that the compound with m/z 1315 is a truncated form of m/z 5076 compound detected in producing cells. It corresponds to the last 11 amino acids of MccAYps, with C-terminally attached CMP, aminopropyl on the phosphate group, and carboxymethyl on the base. We name this compound McCYps (microcin C from Yersinia pseudotuberculosis). High-resolution MS measurements of material from active fraction resulted in m/z values of 1315.4087 and 1331.4042, which is in 0.5 ppm accordance with brutto formulas of [M-monoiso-H+] C45H72N16O24S2P1 with the calculated value of 1315.4082. The shift of 15.9955 Da that results in m/z of 1331.4042 is consistent with one oxygen atom addition (Figure S4). Thus, the 1331 compound is an oxidized form of McCYps. 16183

DOI: 10.1021/jacs.7b07118 J. Am. Chem. Soc. 2017, 139, 16178−16187

Article

Journal of the American Chemical Society subunits of TldD/E protease15 did not contain the N-terminal MccAYps fragment, suggesting that TldD/E is involved in McCYps maturation. Indeed, when cells carrying deletions of tld genes and hns were transformed with pDT1-Ymcc and tested for bioactivity, no growth inhibition zones were formed (Figure 5B). Mass spectrometric analysis of intracellular contents of ΔtldD ΔtldE Δhns triple mutant transformed with pDT1-Ymcc revealed, compared to Δhns mutant cells, a strongly increased m/z 5076 signal corresponding to full-sized MccAYps with Cterminal modifications. The m/z 3777 peak corresponding to N-terminal product of cleavage that generates bioactive McCYps was absent from spectra of ΔtldD ΔtldE Δhns triple mutant cells. We therefore conclude that the TldD/E protease is an essential component of McCYps production. A minor m/z 3880 peak was seen in Δhns cells with or without tld mutations. This peak matches the N-terminal fragment of MccAYps cleaved at position 32 (Figure 5A, lower panel). The intensity of this peak increased in tld mutants suggesting that there exist additional proteases capable of off-target cleavage that does not result in biologically active compound. Accumulation of unprocessed microcin C-like compound in E. coli Δhns ΔtldD ΔtldE strain harboring pDT1-Ymcc allowed us to purify pro-McCYpsfull length modified MccAYpsand test its activity. The compound was found to be inactive at concentrations when McCYps produced robust growth inhibition zones on lawns of wildtype E. coli (Figure 5C−D). The TldD/E proteins are highly conserved in eubacteria and Y. pseudotuberculosis TldD and TldE are 85 and 83% identical to E. coli counterparts. Recombinant Y. pseudotuberculosis TldD and TldE, when present together and combined with MccAYpsCMP caused accumulation of a m/z 3777 mass-ion which corresponds to N-terminal fragment present in E. coli cells with intact tld genes (Figure S6). No corresponding C-terminal peptide−nucleotide fragment was detected, suggesting that it was rapidly degraded. An identical result was obtained with recombinant E. coli TldD/E (Figure S6). The results may imply that in vivo McCYps is removed from producing cells immediately after processing to prevent further degradation. Be it as it may, we take these results as evidence that the TldD/ E protease is responsible for proteolytic maturation of proMcCYps in the natural host Y. pseudotuberculosis. In principle, two mechanisms of TldD/E involvement in McCYps biosynthesis are possible. Unmodified MccAYps may be proteolytically processed first followed by enzymatic modification of shorter C-terminal peptide. Alternatively, the modified full-length MccAYps may be proteolyzed by TldD/E with release of active compound. We tested synthetic peptides corresponding to the full-sized MccAYps and the 11-amino acid long peptide found in McCYps for ability to be modified by MccBYps in vitro (Figure S7). The results showed that MccBYps modified only the 42-amino acid peptide. It is therefore likely that TldD/E role in McCYps production is to cleave pro-McCYps rather than unmodified MccAYps. We considered a possibility that activation of McCYps production in E. coli hns mutants is due to increased expression of tld genes. Indeed, RT-PCR analysis (Figure S8) showed that there were 10 and 2-fold increases in levels of respectively, tldD and tldE transcripts in Δhns cells compared to the wild-type. Thus, in addition to direct effects on mccYps genes transcription (Figure 2B) the absence of H-NS may also stimulate processing of modified MccAYps and formation of bioactive product. Comparing Y. pseudotuberculosis and E. coli Microcins Action. McC Yps was active against both E. coli and

Y. pseudotuberculosis. The minimal inhibitory concentration (MIC), at which inhibition zones on lawns of tester cells were visible, was 3 μM for E. coli B, which is 8 times higher than the corresponding value for McCEco. Both compounds showed similar level of activity against Y. pseudotuberculosis (Figure S9). To compare the mechanism of action of McCYps with that of well-characterized E. coli McC, their ability to inhibit various derivatives of sensitive E. coli strains was tested (Figure 6). Cells

Figure 6. Sensitivity of E. coli mutants or cells producing autoimmunity proteins to McCYps and McCEco. 5-μL drops of 25 μM McCYps, 2.5 μM McCEco, and 0.5 mg/mL of gentamicin solutions were deposited on the surface of freshly prepared top agar seeded with indicated E. coli cultures. Growth inhibition zones appearing after 16 h incubation at 30 °C are shown.

lacking the yejA gene that encodes a component of McCEco transporter YejABEF, were resistant to McCYps. Cells lacking aminopeptidases A, B, and N involved in degradation of McCEco peptide part4 were fully resistant to McCYps. Cells overproducing Asp-RS, the target of McCEco, were also resistant to both McCYps and McCEco while no resistance of cells overproducing prolyl-tRNA synthetase (Pro-RS) was observed. 12.5 μM of McCEco or McCYps was added to E. coli S30 extracts and, after incubation for various times, aminoacylation reaction of tRNAAsp was carried out (Figure S10). Aminoacylation levels remained unchanged after the first 5 min of incubation compared to control extracts (no microcin addition), then decreased by 10 min of incubation and remained stable 16184

DOI: 10.1021/jacs.7b07118 J. Am. Chem. Soc. 2017, 139, 16178−16187

Article

Journal of the American Chemical Society

Be it as it may, the fact that mccYps operons are widely distributed in Y. pseudotuberculosis suggests that its products offer advantages to bacteria that carry them at least in their particular ecological niche. The processing of modified MccAYps inside the producing cell is carried by TldD/E. The exact function of this highly evolutionary conserved endoprotease16,17 remains obscure. Our recent structural work indicates that it may be a generalpurpose enzyme capable of degrading unstructured polypeptides.15 In E. coli, TldD/E is essential for production of microcin B (McB), a thiazole-oxazole containing gyrase inhibitor that is unrelated to McC. During the McB synthesis TldD/E removes the leader part of pro-McB after posttranslational modifications required for activity have been introduced by the McB synthase.18 There is evidence that TldD/E cleavage of pro-McB is associated with its export outside the producing cell.18 The situation may be similar in the case of McCYps, since we are unable to detect mature McCYps inside the producing cells and during in vitro digestion only the N-terminal MccAYps fragment is detected, while the McCYps is degraded. Apparently, the MccCYps pump is unable to act on pro-McCYps but is able to recognize the shorter peptide of McCYps and export it outside the cell. The presence of stable Nterminal 31-amino acid MccAYps TldD/E cleavage fragment inside the producing cells suggests that this peptide is structured and may have a function of its own, for example in regulating the mccYps expression. The additional proteolytic processing step during the biosynthesis of active compound encoded by the Y. pseudotuberculosis mcc operon may help solve the problem of selfintoxication of producing cells that inevitably arises due to processing by aminopeptidases of peptide−nucleotides accumulating inside the producing cells. In the case of McCEco producers, such toxicity was recently demonstrated directly.19 Relying on a much longer, possibly structured MccA peptide should increase intracellular stability of post-translationally modified MccA Yps and prevent accumulation of toxic aspartamidyl-cytidylate that will inhibit protein synthesis in the producer. Though nothing is known about the regulation of TldD/E activity, one can imagine that its activation at specific conditions may allow Y. pseudotuberculosis carrying the mcc operon to produce massive amounts of McCYps from a depot of previously synthesized inactive pro-McCYps. The layout of mcc genes suggests that they form an operon with a single upstream promoter. However, it appears that activity of this promoter is not sufficient for McCYps production. This follows from the observation that McCYps was only produced when the hns gene was deleted. While H-NS is a wellknown factor that silences transcription initiation20 its role must be different in the case of McCYps production. This follows from the fact that MccAYps accumulates inside the cells even in the presence of functional hns, indicating that the main promoter of the operon is active in the presence of H-NS. Therefore, H-NS is either essential for expression of downstream mcc genes or plays a direct role in silencing the enzymes of the pathway. For example, H-NS may serve as a roadblock preventing RNA polymerase transcription elongation through the operon.21 We detected activation of tld genes expression in hns mutants. This effect is likely unrelated to McC Yps production, since accumulation of N-terminal MccAYps fragment, presumably generated by TldD/E cleavage, occurs in wild-type cells carrying plasmid-borne mcc operon. However,

afterward. The kinetics of inhibition suggests that processing of peptide parts of both microcins proceeds at similar rates. However, while reactions containing McCEco retained only 5% of aminoacylation activity after processing was complete, reactions containing McCYps were inhibited only by 50%. Thus, as expected, the McCYps warhead, an aspartamide attached to aminopropylated carboxymethyl CMP is a worse mimic of aspartyl adenylate compared to aminopropylated aspartamidyl-AMP produced after processing of McCEco. Overall, we conclude that both McCYps and McCEc enter sensitive cells through YejABEF transporter and, after intracellular processing of the peptide part by aminopeptidases, inhibit Asp-RS. Cells expressing the MccECTD acetyltransferase from E. coli, while resistant to McCEco, remained sensitive to McCYps, in agreement with earlier observation that carboxymethyl modification of cytidine base renders the self-immunity mechanism mediated by MccEEco inactive.12 Conversely, cells expressing MccE2Yps were resistant to McCYps but sensitive to McCEco. These results indicate that MccE proteins, which provide self-immunity by acetylating processed McC, are adapted for their cognate warheads.

■

DISCUSSION In this work, we characterized McCYps, a product of Y. pseudotuberculosis mcc operon, that inhibits susceptible cells growth and determined the role of individual Yersinia sp. mcc genes in its production and immunity. The microcin encoded by Y. pseudotuberculosis mcc operon is an interesting case of a hybrid molecule: it contains both the terminal carboxymethylated cytosine previously observed in B. amyloliquefaciens microcin C-like compound and propylamine at the phosphate group that is characteristic for E. coli McC. The structure of carboxymethylated cytidine remains to be resolved, although one can speculate, based on analogy with post-transcriptionally modified nucleotides in tRNAs, that C5 is the position for alkyl group attachment. This modification would leave the N3 atom as a likely target for subsequent oxidation that we observe in McCYps. Further study of McCYps structure and how its modified nucleobase mimics adenine and fits the active site of Asp-RS is required. Uniquely for peptide−nucleotide antibiotics, McCYps undergoes an additional modification in the production host that is strictly required for biological activity. The 42-amino acid long MccAYps peptide part is proteolytically processed inside the producing cell releasing mature McCYps whose peptide part is only 11 amino acids long. Upon processing, the compound leaves the producing cell, presumably due to the action of the MccCYps export pump, and is actively taken up by the YejABEF transporter in sensitive cells. The peptide parts of both McCYps and McCEco are processed inside sensitive cells by aminopeptidases A, B, and N, with release of toxic warheads that inhibit Asp-RS. Aminopropylated aspartamidyl-carboxymethylcytidylate that is produced after processing of McCYps appears to be a less potent inhibitor of Asp-RS than aminopropylated aspartamidyl-adenylate released after processing of McCEco. We can assume that aminopropyl decoration that is known to increase affinity of aspartamidyl-adenylates to E. coli Asp-RS also increases antibiotic potency of modified aspartamidylcytidylate. In addition, the presence of the carboxymethyl group on cytidine nucleobase strengthens the interaction with a AspRS,12 allowing it to compete more efficiently for the binding site normally occupied by the aspartyl-adenylate intermediate. 16185

DOI: 10.1021/jacs.7b07118 J. Am. Chem. Soc. 2017, 139, 16178−16187

Article

Journal of the American Chemical Society the finding suggests that the physiological role of TldD/E is connected with stresses that derepress H-NS inhibited genes. Microcins are a class of small ribosomally synthesized antibacterial peptides produced by Enterobacteriaceae and active against closely related bacterial species.22 With the accumulation of sequenced bacterial genomes it was recognized that E. coli microcin B and microcin J are representatives of vast groups of, respectively, linear azole-containing peptides and lasso peptides,23 while microcin C is a member of peptide− nucleotide group antibiotics. Linear azole-containing peptides, lasso peptides, and peptide−nucleotides are in turn subfamilies of ribosomally synthesized post-translationally modified peptides (RIPPs). A common theme in RIPP biosynthesis pathways is the presence in the ribosomally synthesized peptide precursor of a leader part to which modification enzymes bind, and a core part onto which modifications are introduced.24 Once modifications are completed, the leader part is removed releasing mature RIPP. Up to now, peptide−nucleotides appeared to deviate from this scheme. However, the McCYps fully conforms to the RIPP synthesis paradigm: it contains a removable leader peptide that is required for nucleotide attachment by MccBYps to the core part. The N-terminal domain of MccBYps nucleotydyltransferase contains structural similarity to a divergent winged helix-turn-helix RiPP precursor peptide recognition domain (RRE) present in leader binding domains of modification enzymes of RiPPs of other classes.25 We hypothesize that in MccBYps this domain also recognizes the leader peptide. Interestingly MccBEco also contains an RRE, which, however, makes only limited contacts with the short, 7 amino acid-long MccAEco peptide.26 These observations suggest that mcc operons with longer peptides are ancestral: they interacted with their post-translational machinery in a “standard” way but eventually became shorter, as is now seen in most mcc operons, and acquired a new mode of interactions with cognate modification enzymes.

■

Foundation RSF 16-14-10356 to SD and an NIH grant R01 AI117210 (NIAID) to Satish A. Nair and KS. MALDI MS facility became available to us in the framework of the Moscow State University Development Program PNG 5.13.

■

(1) Guijarro, J. I.; González-Pastor, J. E.; Baleux, F.; San Millán, J. L.; Castilla, M. A.; Rico, M.; Moreno, F.; Delepierre, M. J. Biol. Chem. 1995, 270, 23520−23532. (2) Roush, R. F.; Nolan, E. M.; Löhr, F.; Walsh, C. T. J. Am. Chem. Soc. 2008, 130, 3603−3609. (3) Novikova, M.; Metlitskaya, A.; Datsenko, K.; Kazakov, T.; Kazakov, A.; Wanner, B.; Severinov, K. J. Bacteriol. 2007, 189, 8361− 8365. (4) Kazakov, T.; Vondenhoff, G. H.; Datsenko, K. A.; Novikova, M.; Metlitskaya, A.; Wanner, B. L.; Severinov, K. J. Bacteriol. 2008, 190, 2607−2610. (5) Metlitskaya, A.; Kazakov, T.; Kommer, A.; Pavlova, O.; Praetorius-Ibba, M.; Ibba, M.; Krasheninnikov, I.; Kolb, V.; Khmel, I.; Severinov, K. J. Biol. Chem. 2006, 281, 18033−18042. (6) Severinov, K.; Semenova, E.; Kazakov, A.; Kazakov, T.; Gelfand, M. S. Mol. Microbiol. 2007, 65, 1380−1394. (7) Bantysh, O.; Serebryakova, M.; Makarova, K. S.; Dubiley, S.; Datsenko, K. A.; Severinov, K. mBio 2014, 5, e01059−14. (8) Kulikovsky, A.; Serebryakova, M.; Bantysh, O.; Metlitskaya, A.; Borukhov, S.; Severinov, K.; Dubiley, S. J. Am. Chem. Soc. 2014, 136, 11168−11175. (9) Metlitskaya, A.; Kazakov, T.; Vondenhoff, G. H.; Novikova, M.; Shashkov, A.; Zatsepin, T.; Semenova, E.; Zaitseva, N.; Ramensky, V.; Van Aerschot, A.; Severinov, K. J. Bacteriol. 2009, 191, 2380−2387. (10) Novikova, M.; Kazakov, T.; Vondenhoff, G. H.; Semenova, E.; Rozenski, J.; Metlytskaya, A.; Zukher, I.; Tikhonov, A.; Van Aerschot, A.; Severinov, K. J. Biol. Chem. 2010, 285, 12662−12669. (11) Tikhonov, A.; Kazakov, T.; Semenova, E.; Serebryakova, M.; Vondenhoff, G.; Van Aerschot, A.; Reader, J. S.; Govorun, V. M.; Severinov, K. J. Biol. Chem. 2010, 285, 37944−37952. (12) Serebryakova, M.; Tsibulskaya, D.; Mokina, O.; Kulikovsky, A.; Nautiyal, M.; Van Aerschot, A.; Severinov, K.; Dubiley, S. J. Am. Chem. Soc. 2016, 138, 15690−15698. (13) Navarre, W. W. Adv. Microb. Physiol. 2016, 69, 157−186. (14) Baba, T.; Ara, T.; Hasegawa, M.; Takai, Y.; Okumura, Y.; Baba, M.; Datsenko, K. A.; Tomita, M.; Wanner, B. L.; Mori, H. Mol. Syst. Biol. 2006, 2, 2006.0008. (15) Ghilarov, D.; Serebryakova, M.; Stevenson, C. E. M.; Hearnshaw, S. J.; Volkov, D.; Maxwell, A.; Lawson, D. M.; Severinov, K. Structure 2017, 25, 1549−1561. (16) Hu, Y.; Peng, N.; Han, W.; Mei, Y.; Chen, Z.; Feng, X.; Liang, Y. X.; She, Q. Biosci. Rep. 2012, 32, 609−618. (17) Rife, C.; Schwarzenbacher, R.; McMullan, D.; Abdubek, P.; Ambing, E.; Axelrod, H.; Biorac, T.; Canaves, J. M.; Chiu, H. J.; Deacon, A. M.; DiDonato, M.; Elsliger, M. A.; Godzik, A.; Grittini, C.; Grzechnik, S. K.; Hale, J.; Hampton, E.; Han, G. W.; Haugen, J.; Hornsby, M.; Jaroszewski, L.; Klock, H. E.; Koesema, E.; Kreusch, A.; Kuhn, P.; Lesley, S. A.; Miller, M. D.; Moy, K.; Nigoghossian, E.; Paulsen, J.; Quijano, K.; Reyes, R.; Sims, E.; Spraggon, G.; Stevens, R. C.; van den Bedem, H.; Velasquez, J.; Vincent, J.; White, A.; Wolf, G.; Xu, Q.; Hodgson, K. O.; Wooley, J.; Wilson, I. A. Proteins: Struct., Funct., Genet. 2005, 61, 444−448. (18) Allali, N.; Afif, H.; Couturier, M.; Van Melderen, L. J. Bacteriol. 2002, 184, 3224−3231. (19) Piskunova, J.; Maisonneuve, E.; Germain, E.; Gerdes, K.; Severinov, K. Mol. Microbiol. 2017, 104, 463−471. (20) Schröder, O.; Wagner, R. J. Mol. Biol. 2000, 298, 737−748. (21) Winardhi, R. S.; Yan, J.; Kenney, L. J. Biophys. J. 2015, 109, 1321−1329. (22) Severinov, K.; Nair, S. K. Future Microbiol. 2012, 7, 281−289. (23) Yang, X.; van der Donk, W. A. Chem. - Eur. J. 2013, 19, 7662− 7677.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b07118. Figures S1−S10 and detailed experimental procedures (PDF)

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*[email protected] ORCID

Svetlana Dubiley: 0000-0001-9225-5534 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are grateful to Professor M. Skurnik (University of Helsinki, Finland) for extensive help with Yersinia strains handling and valuable advises. We thank Dr. M. Metelev (Uppsala University, Sweden) and Prof. Sergey Borukhov (Rowan University, NJ) for critical reading of this manuscript and Dr. Tatyana Artamonova from Research Center of Nanobiotechnologies, Peter the Great St. Petersburg Polytechnic University, St. Petersburg, Russia for invaluable help with high-resolution mass spectroscopy. The work was supported Russian Science 16186

DOI: 10.1021/jacs.7b07118 J. Am. Chem. Soc. 2017, 139, 16178−16187

Article

Journal of the American Chemical Society (24) Ortega, M. A.; van der Donk, W. A. Cell. Chem. Biol. 2016, 23, 31−44. (25) Burkhart, B. J.; Hudson, G. A.; Dunbar, K. L.; Mitchell, D. A. Nat. Chem. Biol. 2015, 11, 564−570. (26) Regni, C. A.; Roush, R. F.; Miller, D. J.; Nourse, A.; Walsh, C. T.; Schulman, B. A. EMBO J. 2009, 28, 1953−1964.

16187

DOI: 10.1021/jacs.7b07118 J. Am. Chem. Soc. 2017, 139, 16178−16187