Genome Mining Reveals Endopyrroles from a Nonribosomal Peptide

Subscriber access provided by UNIV OF SOUTHERN INDIANA

Article

Genome Mining Reveals Endopyrroles from a Nonribosomal Peptide Assembly Line Triggered in Fungal-Bacterial Symbiosis Sarah P. Niehs, Benjamin Dose, Kirstin Scherlach, Sacha J. Pidot, Timothy P. Stinear, and Christian Hertweck ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.9b00406 • Publication Date (Web): 08 Jul 2019 Downloaded from pubs.acs.org on July 17, 2019

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 20 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

1 2 3

ACS Chemical Biology

Genome Mining Reveals Endopyrroles from a Nonribosomal Peptide Assembly Line Triggered in Fungal-Bacterial Symbiosis

4 5 6

Sarah P. Niehs,1 Benjamin Dose,1 Kirstin Scherlach,1 Sacha J. Pidot,2 Timothy P. Stinear,2 and Christian Hertweck*1,3

7 8

1Department

of Biomolecular Chemistry, Leibniz Institute for Natural Product Research and Infection Biology, HKI, Beutenbergstr. 11a, 07745 Jena (Germany) 2Department

9 10

of Microbiology and Immunology, Doherty Institute, University of Melbourne, 792 Elizabeth Street, Melbourne, 3000 (Australia)

11

3Faculty

12

[email protected]

of Biological Sciences, Friedrich Schiller University Jena, 07743 Jena (Germany)

1

ACS Paragon Plus Environment

ACS Chemical Biology 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

13

ABSTRACT

14 15 16 17 18 19 20 21 22 23 24 25

The bacterial endosymbiont (Burkholderia rhizoxinica) of the rice seedling blight fungus (Rhizopus microsporus) harbors a large number of cryptic biosynthesis gene clusters. Genome mining and sequence similarity networks based on an encoded non-ribosomal peptide assembly line and the associated pyrrole-forming enzymes in the symbiont indicated that the encoded metabolites are unique among a large number of tentative pyrrole natural products in diverse and unrelated bacterial phyla. By performing comparative metabolic profiling using a mutant generated with an improved pheS Burkholderia counterselection marker, we found that the symbionts' biosynthetic pathway is mainly activated under salt stress, and exclusively in symbiosis with the fungal host. The cryptic metabolites were fully characterized as novel pyrrolesubstituted depsipeptides (endopyrroles). A broader survey showed that endopyrrole production is a hallmark of geographically distant endofungal bacteria, which produce the peptides solely under symbiotic conditions.

2


Page 2 of 20

Page 3 of 20 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


26 27 28 29 30 31 32 33 34 35 36

Microorganisms are an eminent source of natural products that play important roles in the environment and have inspired the development of molecular tools, medicines and agricultural agents. Even so, with the advent of routine whole-genome sequencing of microorganisms, it has become more and more obvious that their metabolic capacity is substantially higher than previously thought.1 Genome mining, the systematic, bioinformatics-based approach to unearth cryptic biosynthetic potential, has been mainly employed to predict potential structures of polyketides and peptides from encoded assembly lines.2, 3 Yet, bacteria and fungi do not reveal their full biosynthetic potential under laboratory conditions. In the absence of unknown environmental triggers or stimuli, the majority of biosynthesis gene clusters remain downregulated or silent.4 Thus, most encoded products remain elusive until specific ways to activate gene expression are found.

37 38 39 40 41 42 43 44 45 46 47 48

We have noted a large discrepancy between the wealth of encoded assembly lines and the number of known metabolites in the genome of bacteria (Burkholderia rhizoxinica HKI0454, syn. Mycetohabitans rhizoxinica)5-8 living in the mycelium of the rice seedling blight fungus (Rhizopus microsporus ATCC 62417).6 In this unusual mutual interaction the bacterial endosymbiont produces the rhizoxin complex, which serves as a virulence factor for the plantpathogenic fungus.9-11 Genomic analyses revealed the gene cluster coding for the rhizoxin assembly line, a hybrid non-ribosomal peptide synthetase (NRPS) and a polyketide synthase (PKS).9 Surprisingly, we found that at least 9% of the B. rhizoxinica genome codes for secondary metabolism. Specifically, we identified fourteen cryptic NRPS gene clusters6 that could not be assigned to any known natural product. Only recently, we identified two NRPS products from the endosymbiont, the lipopeptide holrhizin12 and the depsipeptide heptarhizin,13 still leaving a majority of unidentified natural products.

49 50 51 52 53

Here we present the genomics-inspired discovery of a new family of pyrrole-capped depsipeptides from endofungal Burkholderia spp. We elucidate the molecular basis for their assembly, reveal the broad distribution of related gene clusters in bacteria inhabiting diverse environmental niches, and report an unusual case where a silent gene cluster is activated in bacterial-fungal symbiosis.

54 55

RESULTS AND DISCUSSION

56 57 58 59 60 61 62

Bioinformatic analyses of the B. rhizoxinica genome revealed a cryptic gene locus (epyA–E; Genbank accession number MN081797) that stands out because it not only encodes a large, multimodular NRPS but also contains a three-gene composition (epyABC) for a free-standing dehydrogenase (DH, EpyA), an adenylating enzyme (A, EpyB), and a carrier protein (T, EpyC). The products of related three-gene compositions have been shown to provide pyrrole carboxylate building blocks by modification of proline in various biosynthetic pathways (Figure 1a).14

63 64

To assess the uniqueness of the cryptic gene cluster, we chose a homology-based genome mining approach with the DH (EpyA) amino acid sequence as the seed query. To increase the 3



65 66 67 68 69 70 71 72 73 74 75 76 77 78 79

specificity of our search, we took the physical association of the three-gene assembly into account (DH, A, and T; Figure 1b). Therefore, we submitted the EpyA amino acid sequence to the Enzyme Function Initiative Enzyme Similarity Tool (EFI-EST)15 followed by the EFI-Genome Neighborhood Network Tool (EFI-GNN) and created sequence similarity networks (SSNs). Furthermore, we generated genome neighborhood networks (GNN)16 using the UniProt, Pfam and InterPro databases (Figures 1b and 1c). Since EFI-EST is restricted to one query sequence, we also considered SSN creation using EpyB and EpyC as query sequences (not shown), but too many false-positive hits were obtained. In the SSN, protein sequences are represented as nodes that are connected with an edge if the BLAST pairwise similarity scores are equal or above a user-defined threshold. The alignment score was empirically determined and adjusted to obtain mainly groups of orthologous gene clusters. The Pfam family hub-nodes GNN displays the Pfam family co-occurrence frequencies of the neighboring sequences found in the SSN. For every Pfam family identified in the SSN cluster a central hub-node is present. The hub-nodes are in turn connected to spoke-nodes that represent clusters containing a gene of the respective Pfam family.

80 81 82 83 84 85

The network analysis shows that genes for pyrrole-forming DH domains are often associated with neighboring genes encoding AMP-binding enzymes (A domains) and PP-binding (T) domains (orange and green nodes, Figure 1c). Since an intact A-DH-T gene assembly is obligatory for pyrrole biosynthesis, we colored all nodes gray that represent gene clusters lacking either the A or T domain genes. Thus, the SSN identified 20 known biosynthesis gene clusters for pyrrole-containing natural products (orange nodes, Figure 1c).17-32

86 87 88 89 90 91 92 93 94

In addition, we unveiled a large number of orphan gene clusters putatively coding for the biosynthesis of metabolites bearing pyrrole rings (green nodes, Figure 1c). Specifically, we identified putative producers of pyrrole compounds in 11 bacterial phyla, including both aerobes and anaerobes, Planctomycetes, Cyanobacteria, Chloroflexi, Firmicutes, and all classes of Proteobacteria. In the genomes of these bacteria, the three-gene compositions for pyrrole biosynthesis are frequently associated with genes encoding NRPS or PKS assembly lines, tailoring proteins (e.g. halogenases, methyltransferases, glycosyltransferases, cytochrome P450), transport-related genes as well as regulatory genes. This analysis illustrates that there is a large, untapped potential of diverse, pyrrole-containing bacterial natural products.

95 96 97 98 99

Notably, we detected orthologous genes in all currently available (draft) sequences of genomes of bacteria residing in Rhizopus spp. (Bacterial strains B1, B. rhizoxinica HKI0454; B5, B. endofungorum; B4, Burkholderia sp. b13; B7, Burkholderia sp. b14) (Figure 1c). Apart from strain B5, for which only incomplete genome data are available, the association of the pyrroleforming gene assembly with NRPS-encoding genes was corroborated for all endosymbionts.

100 101 102 103 104

It is particularly noteworthy that the cryptic gene clusters of endofungal bacteria fall into a protein network composed mainly of bacteria that live in symbiotic relationships. These include marine strains associated with a range of animal and plant hosts (Pseudoalteromonas spp.), endophytes (Azoarcus sp.), cyanobacteria (Leptolyngbya sp.), and endosymbiotic bacteria of entomopathogenic nematodes (Xenorhabdus and Photorhabdus species) (Figures 1c and 4


Page 4 of 20

Page 5 of 20 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


105 106 107 108 109

Figure S4). As yet, however, none of the encoded natural products have been elucidated. Taken together, our analyses reveal that a) the potential for the production of pyrrole-containing natural products is even larger than previously thought, b) the type of cryptic pyrrole-containing peptides encoded on the genomes of endofungal bacteria is unprecedented, and c) the encoded metabolites may play particular roles in symbioses.

110 111 112 113 114 115 116 117 118

To further predict the structure of the pyrrole-containing peptide in endofungal bacteria, we performed in silico analyses of the respective NPRS encoded by two large NRPS genes (epyD, 15.6 kb, and epyE, 10.6 kb) (Figure 2a). AntiSMASH 4.2.033 and the NRPS/PKS Analysis program34 suggested that the eight modules of EpyD and EpyE assemble a peptide with the sequence L-Val L-X L-Val D-Val L-Thr D-Ser L-Tyr L-Tyr. Only the second amino acid incorporated by EpyD (L-X) could not be envisaged on the basis of the 8-letter amino acid code (DIWYISLL). The presence of a starter condensation (CS) domain in EpyD suggests that the peptide assembly line would be primed with the pyrrole-2-carboxylate building block, thus yielding a nonapeptide (Figure 2a).

119 120 121 122 123 124

To monitor metabolite production we freshly isolated the endosymbiont B. rhizoxinica from its fungal host (R. microsporus ATCC62417) and varied culture conditions physically (temperature, oxygen supply) and/or chemically (media compositions, addition of growth stimulators) (Figure 2b). LC/MS analyses of the culture extracts revealed the presence of the rhizoxin complex,11 the lipopeptide holrhizin A,12 and the cyclopeptide heptarhizin.13 Yet, no mass was detected that would be a match for the predicted peptide.

125 126 127 128 129 130 131 132 133 134 135

Since B. rhizoxinica naturally resides in the rice seedling blight fungus R. microsporus, we decided to investigate the holobiont (H1). Culture methods were modified in order to match the fungal (and not the bacterial) requirements for growth, such as media composition and the use of agar plates instead of shaking culture vessels. Eventually, we found one condition in particular (complex medium with a high salt content) allowed the detection of a compound with a mass (1001.5 Da) that fits with the predicted structure (1; Figure 2b). From HRESI/MS measurements we determined its molecular formula as C50H68N9O13 (calc. m/z 1002.4931 [M+H]+, found m/z 1002.4930 [M+H]+). The MS/MS fragmentation pattern of 1 provided further evidence for a nonapeptide backbone (Figure S7). In addition, we detected a less abundant congener with m/z 857.4407 Da [M+H]+ (2; C41H61N8O12, calc. m/z 857.4403 [M+H]+) and a similar peptide fragmentation pattern as the main product (Figure S8).

136 137 138 139 140

To verify that the detected peptide was composed of one or more proline-derived residues we performed isotope-labeling experiments with 13C5-proline. MS/MS fragmentation of the isotopic peaks of m/z 1002.5 Da [M+H]+ showed the incorporation of two 13C5-proline units at two different positions (single: m/z 1007.5 Da [M+H]+, double: m/z 1012.5 Da [M+H]+; Figures 2c and S9–11).

141 142 143

To link the biosynthesis of 1 and 2 to the epy gene cluster we constructed a targeted knockout mutant of B. rhizoxinica (Figures 2d and 2e). We employed the double-selection plasmid pGL42a containing the pheS gene35 for a modified version of a phenylalanyl-tRNA synthetase 5



144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159

(Figure 2d). If grown on double selection agar, the phenylalanyl-tRNA synthetase leads to the incorporation of 4-L-chlorophenylalanine into proteins, thereby hampering translation and causing cell death.36 Even though genetic manipulation of B. rhizoxinica using this system has been reported before,9, 12 the process of creating mutants remained laborious and timeconsuming. Thus, we improved the pheS-Burkholderia counterselection marker by introducing a point mutation resulting in pGL42a_T251A as has been described for pheS from Escherichia coli (Figure S3).37 The reported point mutation increases the susceptibility of the phenylalanyl-tRNA synthetase to 4-L-chlorophenylalanine and thereby the selection pressure. Furthermore, we identified sequence regions within epyD coding for the fifth A domain that have limited homology to the rest of the genome. These DNA sequences (~1 kb) served as the homologous regions flanking a kanamycin resistance cassette. A restriction and ligation plasmid construction approach using the plasmid pGL42a_T251A resulted in pBD61 (Figure S2). This procedure facilitated the construction of the B. rhizoxinica Pepy mutant, which was verified by PCR (Figure 2e) and re-introduced into apo-symbiotic R. microsporus. Metabolic profiling of the fungus hosting the Pepy mutant showed that the production of 1 and 2 was completely abrogated (Figure 2f).

160 161 162 163 164 165

To obtain sufficient amounts of 1 and 2 for a full structure elucidation we scaled up the fermentation of the wild type under producing conditions on high-saline agar (14 L BMSW agar). The organic agar extract was subjected to size-exclusion chromatography using Sephadex LH20, followed by preparative HPLC, yielding pure endopyrrole A (1; 4 mg) and endopyrrole B (2; 0.5 mg). Their structures were elucidated by a combination of 1D- and 2D-NMR analyses, HRESI/MS fragmentation, and chemical derivatization.

166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181

Comparison of the 13C and DEPT135 spectra of 1 indicated the presence of 23 methines, six methylenes, seven methyl groups, and 14 quaternary carbons from which nine were assigned to carbonyl groups. NH functions accounted for seven doublets (7.5–9 ppm) in the 1H spectrum. In addition, 1H-13C HMBC couplings of these signals to those of putative carbonyl groups (160– 172 ppm) supported our in silico structure prediction (Figure 2a). NOE couplings of the second and first valine-NH groups to the C of proline and pyrrole completed the peptide backbone. The ester bond formation between the threonine and tyrosine moieties was verified using the MS/MS fragmentation pattern in combination with informatics (MassFrontier version 7.0; Figure S7). The N-capped pyrrole carboxylate unit was assigned based on 1H-1H COSY and 1H-13C HMBC couplings (Figures S18 and S21). Thus, the 2D structure of endopyrrole A (1) was established (Figure 3a). The absolute configuration of endopyrrole A was elucidated by Marfey analysis of the amino acid fragments (Supporting Information). The deduced molecular formula of 2 indicated that endopyrrole B (2) lacks one amino acid unit, compared to the main product 1. Comparison of the NMR spectra of 1 and 2 showed that in 2 five methines, one methylene, and three quaternary carbons are missing. Thus, we concluded that 2 is a linear congener of 1 lacking the terminal tyrosine residue (Figures 3a and S23–30).

182 183

Consistent with the in silico prediction, endopyrroles A and B are cyclodepsipeptides, natural products that have previously been unknown. Pure NRPS products capped with pyrrole 6


Page 6 of 20

Page 7 of 20 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


184 185 186 187 188 189 190

carboxylates are rare. The only described molecules are the bacterial hormone hormaomycin from Streptomyces * 2 3 38, the cyclopeptides tataricins isolated from a Chinese medical plant (Aster tataricus),39 and the cochinmicins from a Microbispora species.40, 41 The bacterial hormones evaded our bioinformatic survey because of low sequence similarity between the hormaomycin-DH domain and the query sequence (Supporting Information). Notably, the genes for the cochinmicin and tataricin biosynthetic pathways are also unknown. Even so, the structures of these compounds differ substantially from those of the endopyrroles.

191 192 193 194 195 196 197 198 199

To gain insight into the potential function of endopyrrole A (1), we subjected the pure compound to a panel of whole-cell cytotoxicity and antimicrobial assays. Yet, no inhibitory effects were observed. To investigate the role of the endopyrroles in the holobiont, we scrutinized the reinfection rate of R. microsporus with producing and non-producing strains of the endosymbiont. Since the presence of the endosymbiont is obligatory for asexual reproduction,42 regaining sporulation ability of R. microsporus with B. rhizoxinica Pepy in comparison to the axenic fungus demonstrates the successful reestablishment of symbiosis in both cases. These results show that endopyrroles are not essential for the reconstitution and propagation of the symbiosis, and obvious changes in the phenotypes could not be observed (data not shown).

200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222

As indicated by the SSN, however, putative endopyrrole biosynthesis genes can be found in all sequenced genomes of endofungal bacteria (Figure 3b). To learn more about the distribution and diversity of endopyrroles, we investigated eight different endofungal Burkholderia strains (B1–8) either as free-living bacteria or in symbiosis (holobionts H1–8) with their respective Rhizopus hosts. The eight holobionts were shown to constitute four taxonomically and geographically distant clades.7 To test whether these strains are able to produce the predicted pyrrole-containing natural product, the endosymbionts were subjected to the metabolic profiling that has been successful for strain B1 in symbiosis with the fungal host (H1). HPLC/ESI-HRMS analyses of the holobiont cultures revealed that all members of the (so-called) Pacific clade (symbioses containing strains B1, B2 and B6; mostly from rice pathogens) produce endopyrroles A and B. Members of the Eurasian clade (symbioses containing strains B4, Burkholderia sp. b13; B7, Burkholderia sp. b14; and B3, Burkholderia sp. HKI04557) produce congeners of endopyrrole A in symbiosis with their natural hosts (Figure 3c). Here we detected a metabolite with m/z 1018.4872 Da [M+H]+ (deduced molecular formula of C50H68N9O14, calc. m/z 1018.4880 [M+H]+), named endopyrrole C (3). We isolated pure 3 (0.8 mg from 6 L agar) and found, by comparison of MS/MS fragmentation patterns and NMR data, that 3 differs from 1 by two units of threonine in lieu of the valine and serine moieties (Pyr-Thr-Pro-Val-Val-Thr-Thr-Tyr-Tyr, Figures 3a, 3c, and S12). Furthermore, strain B4 produces the putative congener endopyrrole D (4) in trace amounts (m/z 920.3810 Da [M+H]+; C48H54N7O12; Figure S13). Only symbioses of Rhizopus with B. endofungorum (H5), the producer of the hepatotoxic rhizonins and sole member of the African clade,43, 44 and symbioses with Burkholderia sp. HKI04047 (H8), the only member of the Australian clade, did not produce endopyrroles under the conditions tested (Figure 3c).

7



223 224 225 226 227 228 229 230 231 232 233 234 235 236

All six endopyrrole-positive strains of the endobacteria produce the peptides exclusively in symbiosis with their fungal hosts. This finding is remarkable in light of various recent studies in fungal-bacterial co-cultivations that aimed at inducing otherwise silent biosynthetic pathways.45 For example, competitive co-cultivations of Aspergillus and Streptomyces spp. have been shown to trigger the production of secondary metabolites e.g. orsellinic acid, formyl-xanthocillin analogues, cyclodipeptides, and fumicycline A.46-50 Another example is the interaction of Burkholderia gladioli pv. cocovenenans and R. microsporus var. oligosporus, which is frequently observed in the fermentative production of tempeh bongkrek.51, 52 In the presence of the fungus, bacterial growth is positively Q& and bongkrekic acid production is dramatically increased, which in return inhibits fungal growth.53 Herein we present a rare case where mutualism rather than competition between bacteria and fungi is vital for secondary metabolite production. Even though endopyrroles are not essential for maintaining the fungal-bacterial symbiosis, the widespread distribution of the gene cluster suggests a relevant but yet unknown function in the symbiotic interaction.

237 238 239 240 241 242 243 244 245 246 247 248

In conclusion, we have shown that a protein network based on the proline-modifying dehydrogenase sequence can be used to identify biosynthesis gene clusters for pyrrole metabolites independent of the orientation and order of the individual genes, and to predict the uniqueness of the encoded pathways. As a proof of concept, we report the genomics-inspired discovery of previously unknown pyrrole-capped depsipeptides (endopyrroles) from endofungal bacteria. Although the bacteria can live independently from the fungal host, endopyrrole production is exclusively observed under symbiotic conditions and never in cultures of free-living bacteria. This represents one of the rare cases where an endosymbiotic environment is required to trigger an otherwise silent gene cluster. The unearthing of the endopyrroles expands our knowledge about chemical mediators involved in the unusual relationship between the ricepathogenic fungus R. microsporus and its endosymbiont, B. rhizoxinica.

249 250 251 252 253 254

8


Page 8 of 20

Page 9 of 20 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




258 259 260 261 262 263 264 265 266 267 268

Figure 1 Identification of putative producers of pyrrole-containing natural products using an SSN. a) Depiction of the dehydrogenase (DH, yellow), adenylation (A, light red) and thiolation (T, black) gene composition and respective proteins responsible for pyrrole formation starting from proline. b) Section of the Pfam family hub-nodes genome neighborhood network showing the co-occurrence frequencies of neighboring sequences found in the sequence similarity network (SSN). A central hub-node (hexagon) is present for the DH, A and T domain identified in the SSN cluster. A spoke-node is present for each SSN cluster that contains a member of the respective Pfam family. The size of the spoke-nodes corresponds to the cooccurrence frequencies. Color code as in depicted in c). c) Section of the SSN generated using the EFI-EST web tool with the DH domain amino acid sequence as query. Nodes of gene cluster containing an incomplete three-gene assembly have been omitted. Known pyrrole-containing natural products are depicted next to the respective nodes. *Occurs in two networks. Further information can be found in the Supporting Information.

10


Page 10 of 20

Page 11 of 20 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




271 272 273 274 275 276 277 278 279 280 281 282 283

Figure 2 Endopyrrole biosynthesis, dereplication of culturing conditions for silent gene cluster activation, and null mutant construction. a) epy gene cluster; yellow – dehydrogenase (DH); light red – adenylation (A); black – thiolation (T); dark blue – non-ribosomal peptide synthetase (NRPS, CS – starter, LCL – L to L amino acid condensation, Dual – epimerization and condensation). b) Extracted ion chromatograms (EICs) of endopyrrole A (1002.5 Da) in positive mode under different cultivation conditions; H1 – Rhizopus microsporus ATCC62417 (holobiont); B1 – Burkholderia rhizoxinica HKI0454; abbreviations for different media compositions and culture conditions can be found in the Supporting Information. c) Isotopic pattern and 13C-incorporation into endopyrrole A after addition of 13C5-labeled proline. Asterisks indicate labelling. d) Construction of a targeted endopyrrole-deficient mutant through homologous recombination using the double selection plasmid pBD61; pheS – phenylalanine tRNA synthetase gene; kanR – kanamycin resistance cassette; arrows depict primer binding sites. e) PCR-based confirmation of the B. rhizoxinica endopyrrole mutant 1Pepy); kilobase pairs (kb); marker (M). f) Metabolic profile of wild type fungus R. microsporus (H1) and re-infected fungus with endopyrrole-deficient mutant 19 JP $ 3 as EIC m/z 1002.5 [M+H]+.

OH

a

OH

O

O N H HN HO NH O O O O NH HN H N

HN O

O

OH O OH HN O

H N

HN NH

O

O

OH

2 O

O

O N H HN HO NH O O O O NH HN H N O

O

O

N

1H,13C

HMBC COSY 1H,1H NOESY

c

H N

HN

1H,1H

b

O N H HN HO NH O O O O NH HN

OH

O NH

O

N

1

HN

O

O NH

O

OH

N

1

O

O NH

O

O

N OH

2

3 43

B1, (1)-BGC H1 B4, putative (4)-BGC B7, putative (3)-BGC B1 B5

#

#

H2

284

Pacific branch

H6 H3 H4

Eurasian branch

H7

B7

10 kb

O

HN

OH

O

NH OH

NH

Relative abundance

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 12 of 20

Xenorhabdus spp. Burkholderia spp. Query sequence 15

14

14

Time (min)

12


H8

Australian branch

H5

African branch

OH

Page 13 of 20 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

285 286 287 288 289 290 291


Figure 3 Endopyrroles A–C and their distribution. a) Structures of endopyrroles A (1), B (2), and C (3) with 2D-NMR couplings for endopyrrole A. b) Extracted SSN highlighting the similarities of the proline-modifying DH sequences (black circle – query sequence) in Burkholderia and Xenorhabdus spp. with selected gene clusters shown. Incomplete genome sequences are indicated by brackets. B1 – B. rhizoxinica; B5 – B. endofungorum; B7 – Burkholderia spp. b14. Color code according to Figure 2a. c) Distribution of endopyrrole production among endofungal Burkholderia strains as extracted ion chromatograms in positive mode; endopyrrole A as m/z 1002.5, B as 857.5, C as 1018.5 and putative D (4) as 920.5 Da; H1–8 – Rhizopus sp. hosts, see Supporting Information for strain assignment; # – trace amounts.

13



292

ASSOCIATED CONTENT

293

Supporting Information

294 295 296

Supporting Information Available: The supporting information including experimental details, compounds characterization, and further information concerning the sequence similarity network is available free of charge via the Internet.

297 298

AUTHOR INFORMATION

299

Corresponding authors

300

*[email protected]

301

Author contributions

302 303 304 305 306

S.P.N. performed isolation of compounds and chemical characterization. K.S. and S.P.N. elucidated the structures. B.D. constructed protein networks and performed genetic manipulations. S.J.P. and T.P.S. performed additional bioinformatics analyses. K.S. and C.H. designed and guided all aspects of this work. All authors contributed to manuscript preparation. C.H. wrote the Y version.

307

Conflict of interest

308

The authors declare no conflict of interest.

309

ACKNOWLEDGEMENT

310 311 312 313

We thank H.-M. Dahse and C. Weigel for bioactivity testing, A. Perner for MS/MS and H. Heinecke for NMR measurements. We also gratefully acknowledge S. Richter for assistance in compound isolation and Dr. K. Dunbar for helpful discussions. This work was financially supported by the DFG (SFB 1127, ChemBioSys, and Leibniz Award to C.H.).

314

315

14


Page 14 of 20

Page 15 of 20 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

317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364


REFERENCES [1] Katz, L., and Baltz, R. H. (2016) Natural product discovery: past, present, and future, J. Ind. Microbiol. Biotechnol. 43, 155-176. [2] Baltz, R. H. (2017) Gifted microbes for genome mining and natural product discovery, J. Ind. Microbiol. Biotechnol. 44, 573-588. [3] Mohimani, H., and Pevzner, P. A. (2016) Dereplication, sequencing and identification of peptidic natural products: from genome mining to peptidogenomics to spectral networks, Nat. Prod. Rep. 33, 73-86. [4] Scherlach, K., and Hertweck, C. (2009) Triggering cryptic natural product biosynthesis in microorganisms, Org. Biomol. Chem. 7, 1753-1760. [5] Partida-Martinez, L. P., and Hertweck, C. (2005) Pathogenic fungus harbours endosymbiotic bacteria for toxin production, Nature 437, 884. [6] Lackner, G., Moebius, N., Partida-Martinez, L. P., Boland, S., and Hertweck, C. (2011) Evolution of an endofungal lifestyle: deductions from the Burkholderia rhizoxinica genome, BMC Genomics 12, 210. [7] Lackner, G., Möbius, N., Scherlach, K., Partida-Martinez, L. P., Winkler, R., Schmitt, I., and Hertweck, C. (2009) Global distribution and evolution of a toxinogenic Burkholderia-Rhizopus symbiosis, Appl. Environ. Microbiol. 75, 2982-2986. [8] Estrada-de los Santos, P., Palmer, M., Chávez-Ramírez, B., Beukes, C., Steenkamp, E., Briscoe, L., Khan, N., Maluk, M., Lafos, M., and Humm, E. (2018) Whole genome analyses suggests that Burkholderia sensu lato contains two additional novel genera (Mycetohabitans gen. nov., and Trinickia gen. nov.): implications for the evolution of diazotrophy and nodulation in the Burkholderiaceae, Genes 9, 389. [9] 3 ! % ,1 ! % ( L. P., and Hertweck, C. (2007) A gene cluster encoding rhizoxin biosynthesis in “Burkholderia rhizoxina”, the bacterial endosymbiont of the fungus Rhizopus microsporus, ChemBioChem 8, 41-45. [10] Scherlach, K., Partida-Martinez, L. P., Dahse, H.-M., and Hertweck, C. (2006) Antimitotic rhizoxin derivatives from a cultured bacterial endosymbiont of the rice pathogenic fungus Rhizopus microsporus, J. Am. Chem. Soc. 128, 11529-11536. [11] Iwasaki, S., Namikoshi, M., Kobayashi, H., Furukawa, J., Okuda, S., Itai, A., Kasuya, A., Iitaka, Y., and Sato, Z. (1986) Studies on macrocyclic lactone antibiotics. VIII. Absolute structures of rhizoxin and a related compound, J. Antibiot. 39, 424-429. [12] Niehs, S. P., Scherlach, K., and Hertweck, C. (2018) Genomics-driven discovery of a linear lipopeptide promoting host colonization by endofungal bacteria, Org. Biomol. Chem. 16, 6967-6974. [13] Niehs, S. P., Dose, B., Scherlach, K., Roth, M., and Hertweck, C. (2018) -( #.%$&, !%'( discovery of a &)./%# ,&"($%+%$ cyclopeptide from bacteria residing in the rice seedling blight fungus, ChemBioChem 19, 2167-2172. [14] Walsh, C. T., Garneau-Tsodikova, S., and Howard-Jones, A. R. (2006) Biological formation of pyrroles: nature's logic and enzymatic machinery, Nat. Prod. Rep. 23, 517-531. [15] Gerlt, J. A., Bouvier, J. T., Davidson, D. B., Imker, H. J., Sadkhin, B., Slater, D. R., and Whalen, K. L. (2015) Enzyme function initiative-enzyme similarity tool (EFI-EST): A web tool for generating protein sequence similarity networks, Biochim. Biophys. Acta. Proteins Proteom. 1854, 10191037. [16] Zallot, R., Oberg, N. O., and Gerlt, J. A. (2018) ‘Democratized’genomic enzymology web tools for functional assignment, Curr. Opin. Chem. Biol. 47, 77-85. [17] Cane, D. E., Tan, W., and Ott, W. R. (1993) Nargenicin biosynthesis. Incorporation of polyketide chain elongation intermediates and support for a proposed intramolecular Diels-Alder cyclization, J. Am. Chem. Soc. 115, 527-535.

15



365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413

[18] Ravindran, A., Sunderrajan, S., and Pennathur, G. (2019) Phylogenetic studies on the prodigiosin biosynthetic operon, Curr. Microbiol. 76, 597-606. [19] Thomas, M. G., Burkart, M. D., and Walsh, C. T. (2002) Conversion of L-proline to pyrrolyl-2-carboxylS-PCP during undecylprodigiosin and pyoluteorin biosynthesis, Chem. Biol. 9, 171-184. [20] Wu, Q., Liang, J., Lin, S., Zhou, X., Bai, L., Deng, Z., and Wang, Z. (2011) Characterization of the biosynthesis gene cluster for the pyrrole polyether antibiotic calcimycin (A23187) in Streptomyces chartreusis NRRL 3882, Antimicrob. Agents Chemother. 55, 974-982. [21] Li, C., Roege, K. E., and Kelly, W. L. (2009) Analysis of the indanomycin biosynthetic gene cluster from Streptomyces antibioticus NRRL 8167, ChemBioChem 10, 1064-1072. [22] Flatt, P. M., Wu, X., Perry, S., and Mahmud, T. (2013) Genetic insights into pyralomicin biosynthesis in Nonomuraea spiralis IMC A-0156, J. Nat. Prod. 76, 939-946. [23] Siebenberg, S., Burkard, N., Knuplesch, A., Gust, B., Grond, S., and Heide, L. (2011) Two pathways for pyrrole formation in coumermycin A1 biosynthesis: the central pyrrole moiety is formed from Lthreonine, ChemBioChem 12, 2677-2685. [24] Nowak-Thompson, B., Chaney, N., Wing, J. S., Gould, S. J., and Loper, J. E. (1999) Characterization of the pyoluteorin biosynthetic gene cluster of Pseudomonas fluorescens Pf-5, J. Bacteriol. 181, 2166-2174. [25] Kopp, M., Irschik, H., Gemperlein, K., Buntin, K., Meiser, P., Weissman, K. J., Bode, H. B., and Müller, R. (2011) Insights into the complex biosynthesis of the leupyrrins in Sorangium cellulosum So ce690, Mol. BioSys. 7, 1549-1563. [26] Meiser, P., Weissman, K. J., Bode, H. B., Krug, D., Dickschat, J. S., Sandmann, A., and Müller, R. (2008) DKxanthene biosynthesis—Understanding the basis for diversity-oriented synthesis in myxobacterial secondary metabolism, Chem. Biol. 15, 771-781. [27] Méjean, A., Paci, G., Gautier, V., and Ploux, O. (2014) Biosynthesis of anatoxin-a and analogues (anatoxins) in cyanobacteria, Toxicon 91, 15-22. [28] Yamanaka, K., Ryan, K. S., Gulder, T. A. M., Hughes, C. C., and Moore, B. S. (2012) Flavoenzymecatalyzed atropo-selective N,C-bipyrrole homocoupling in marinopyrrole biosynthesis, J. Am. Chem. Soc. 134, 12434-12437. [29] Mantovani, S. M., and Moore, B. S. (2013) Flavin-linked oxidase catalyzes pyrrolizine formation of dichloropyrrole-containing polyketide extender unit in chlorizidine A, J. Am. Chem. Soc. 135, 18032-18035. [30] Witte, S. N. R., Hug, J. J., Géraldy, M. N. E., Müller, R., and Kalesse, M. (2017) Biosynthesis and total synthesis of pyrronazol B: a secondary metabolite from Nannocystis pusilla, Chem. Eur. J. 23, 15917-15921. [31] Ma, H.-M., Zhou, Q., Tang, Y.-M., Zhang, Z., Chen, Y.-S., He, H.-Y., Pan, H.-X., Tang, M.-C., Gao, J.-F., and Zhao, S.-Y. (2013) Unconventional origin and hybrid system for construction of pyrrolopyrrole moiety in kosinostatin biosynthesis, Chem. Biol. 20, 796-805. [32] El Gamal, A., Agarwal, V., Diethelm, S., Rahman, I., Schorn, M. A., Sneed, J. M., Louie, G. V., Whalen, K. E., Mincer, T. J., and Noel, J. P. (2016) Biosynthesis of coral settlement cue tetrabromopyrrole in marine bacteria by a uniquely adapted brominase–thioesterase enzyme pair, Proc. Natl. Acad. Sci. U. S. A. 113, 3797-3802. [33] Blin, K., Wolf, T., Chevrette, M. G., Lu, X., Schwalen, C. J., Kautsar, S. A., Suarez Duran, H. G., De Los Santos, E. L., Kim, H. U., and Nave, M. (2017) antiSMASH 4.0—improvements in chemistry prediction and gene cluster boundary identification, Nucleic Acids Res. 45, W36-W41. [34] Bachmann, B. O., and Ravel, J. (2009) Methods for in silico prediction of microbial polyketide and nonribosomal peptide biosynthetic pathways from DNA sequence data, Methods Enzymol. 458, 181-217. [35] Lackner, G., Moebius, N., and Hertweck, C. (2011) Endofungal bacterium controls its host by an hrp type III secretion system, ISME J. 5, 252. 16


Page 16 of 20

Page 17 of 20 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

414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461


[36] Barrett, A. R., Kang, Y., Inamasu, K. S., Son, M. S., Vukovich, J. M., and Hoang, T. T. (2008) Genetic tools for allelic replacement in Burkholderia species, Appl. Environ. Microbiol. 74, 4498-4508. [37] Miyazaki, K. (2015) Molecular engineering of a PheS counterselection marker for improved operating efficiency in Escherichia coli, BioTechniques 58, 86-88. [38] Höfer, I., Crüsemann, M., Radzom, M., Geers, B., Flachshaar, D., Cai, X., Zeeck, A., and Piel, J. (2011) Insights into the biosynthesis of hormaomycin, an exceptionally complex bacterial signaling metabolite, Chem. Biol. 18, 381-391. [39] Xu, H.-M., Yi, H., Zhou, W.-B., He, W.-J., Zeng, G.-Z., Xu, W.-Y., and Tan, N.-H. (2013) Tataricins A and B, two novel cyclotetrapeptides from Aster tataricus, and their absolute configuration assignment, Tetrahedron Lett. 54, 1380-1383. [40] Lam, Y. T., Williams, D. L., Sigmund, J. M., Sanchez, M., Genilloud, O., Kong, Y. L., Stevens-Miles, S., Huang, L., and Garrity, G. M. (1992) Cochinmicins, novel and potent cyclodepsipeptide endothelin antagonists from a Microbispora sp., J. Antibiot. 45, 1709-1716. [41] Zink, D., Hensens, O. D., Lam, Y. T., Reamer, R., and Liesch, J. M. (1992) Cochinmicins, novel and potent cyclodepsipeptide endothelin antagonists from a Microbispora sp., J. Antibiot. 45, 17171722. [42] Moebius, N., Üzüm, Z., Dijksterhuis, J., Lackner, G., and Hertweck, C. (2014) Active invasion of bacteria into living fungal cells, eLife 3, e03007. [43] Partida-Martinez, L. P., de Looß, C. F., Ishida, K., Ishida, M., Roth, M., Buder, K., and Hertweck, C. (2007) Rhizonin, the first mycotoxin isolated from the zygomycota, is not a fungal metabolite but is produced by bacterial endosymbionts, Appl. Environ. Microbiol. 73, 793-797. [44] Steyn, P. S., Tuinman, A. A., van Heerden, F. R., van Rooyen, P. H., Wessels, P. L., and Rabie, C. J. (1983) The isolation, structure, and absolute configuration of the mycotoxin, rhizonin A, a novel cyclic heptapeptide containing N-methyl-3-(3-furyl) alanine, produced by Rhizopus microsporus, J. Chem. Soc. Chem. Commun., 47-49. [45] Scherlach, K., Graupner, K., and Hertweck, C. (2013) Molecular bacteria-fungi interactions: effects on environment, food, and medicine, Annu. Rev. Microbiol. 67, 375-397. [46] Sakuda, S., Ono, M., Furihata, K., Nakayama, J., Suzuki, A., and Isogai, A. (1996) Aflastatin A, a novel inhibitor of aflatoxin production of Aspergillus parasiticus, from Streptomyces, J. Am. Chem. Soc. 118, 7855-7856. [47] Wu, C., Zacchetti, B., Ram, A. F., Van Wezel, G. P., Claessen, D., and Choi, Y. H. (2015) Expanding the chemical space for natural products by Aspergillus-Streptomyces co-cultivation and biotransformation, Sci. Rep. 5, 10868. [48] Schroeckh, V., Scherlach, K., Nützmann, H.-W., Shelest, E., Schmidt-Heck, W., Schuemann, J., Martin, K., Hertweck, C., and Brakhage, A. A. (2009) Intimate bacterial–fungal interaction triggers biosynthesis of archetypal polyketides in Aspergillus nidulans, Proc. Natl. Acad. Sci. U. S. A. 106, 14558-14563. [49] Zuck, K. M., Shipley, S., and Newman, D. J. (2011) Induced production of N-formyl alkaloids from Aspergillus fumigatus by co-culture with Streptomyces peucetius, J. Nat. Prod. 74, 1653-1657. [50] König, C. C., Scherlach, K., Schroeckh, V., Horn, F., Nietzsche, S., Brakhage, A. A., and Hertweck, C. (2013) Bacterium induces cryptic meroterpenoid pathway in the pathogenic fungus Aspergillus fumigatus, ChemBioChem 14, 938-942. [51] Moebius, N., Ross, C., Scherlach, K., Rohm, B., Roth, M., and Hertweck, C. (2012) Biosynthesis of the respiratory toxin bongkrekic acid in the pathogenic bacterium Burkholderia gladioli, Chem. Biol. 19, 1164-1674. [52] Rohm, B., Scherlach, K., and Hertweck, C. (2010) Biosynthesis of the mitochondrial adenine nucleotide translocase (ATPase) inhibitor bongkrekic acid in Burkholderia gladioli, Org. Biomol. Chem. 8, 1520-1522.

17



462 463

[53] Ross, C., Opel, V., Scherlach, K., and Hertweck, C. (2014) Biosynthesis of antifungal and antibacterial polyketides by Burkholderia gladioli in coculture with Rhizopus microsporus, Mycoses 57, 48-55.

464

18


Page 18 of 20

Page 19 of 20 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




Graphical abstract 1418x1086mm (55 x 55 DPI)


Page 20 of 20

Genome Mining Reveals Endopyrroles from a Nonribosomal Peptide

Recommend Documents