An In Planta-Expressed Polyketide Synthase Produces (R ...€¦ · An In Planta-Expressed Polyketide Synthase Produces (R)-Mellein in the Wheat Pathogen Parastagonospora nodorum Yit-Heng

An In Planta-Expressed Polyketide Synthase Produces (R)-Mellein inthe Wheat Pathogen Parastagonospora nodorum

Yit-Heng Chooi,a Christian Krill,b Russell A. Barrow,c Shasha Chen,c Robert Trengove,b Richard P. Oliver,d Peter S. Solomona

Plant Sciences Division, Research School of Biology, The Australian National University, Canberra, Australian Capital Territory, Australiaa; Separation Science andMetabolomics Laboratory, Murdoch University, Murdoch, Western Australia, Australiab; Research School of Chemistry, The Australian National University, Canberra,Australian Capital Territory, Australiac; Centre for Crop and Disease Management, Department of Environment & Agriculture, Curtin University, Perth, Western Australia,Australiad

Parastagonospora nodorum is a pathogen of wheat that affects yields globally. Previous transcriptional analysis identified a par-tially reducing polyketide synthase (PR-PKS) gene, SNOG_00477 (SN477), in P. nodorum that is highly upregulated during in-fection of wheat leaves. Disruption of the corresponding SN477 gene resulted in the loss of production of two compounds, whichwe identified as (R)-mellein and (R)-O-methylmellein. Using a Saccharomyces cerevisiae yeast heterologous expression system,we successfully demonstrated that SN477 is the only enzyme required for the production of (R)-mellein. This is the first identifi-cation of a fungal PKS that is responsible for the synthesis of (R)-mellein. The P. nodorum �SN477 mutant did not show anysignificant difference from the wild-type strain in its virulence against wheat. However, (R)-mellein at 200 �g/ml inhibited thegermination of wheat (Triticum aestivum) and barrel medic (Medicago truncatula) seeds. Comparative sequence analysis identi-fied the presence of mellein synthase (MLNS) homologues in several Dothideomycetes and two sodariomycete genera. Phyloge-netic analysis suggests that the MLNSs in fungi and bacteria evolved convergently from fungal and bacterial 6-methylsalicylicacid synthases.

The dothideomycete plant pathogens are prolific producers ofsecondary metabolites (1). Many of these secondary metabo-

lites serve as virulence mediators of plant pathogens (2, 3), as wellas mycotoxins that are detrimental to human health (4). Thus,understanding of the secondary metabolites and their functions inplant pathogens is important for advancing our understanding ofplant pathogenesis and may facilitate new prospects for control-ling fungal diseases in the field. The availability of many dothideo-mycete plant pathogen genomes also provides a valuable resourcefor genome mining for novel secondary metabolites that can fuelthe discovery of drugs and agrochemicals.

Parastagonospora nodorum (alternative names, Phaeosphaerianodorum and Stagonospora nodorum) is a model dothideomycetenecrotrophic pathogen and is the causative agent of Septoria nodo-rum blotch (SNB) in wheat (5). Its genome, sequenced in 2007 (6),has revealed a large number of secondary metabolite biosyntheticgene clusters. We recently surveyed the secondary metabolitegenes in the P. nodorum genome and identified 23 polyketide syn-thase (PKS) genes, 14 nonribosomal peptide synthetase (NRPS)genes, 4 terpene synthase genes, and 2 prenyltransferase genes (7).Along with the tailoring biosynthesis genes, they formed a total of38 secondary metabolite gene clusters in P. nodorum. So far, noneof the products of these secondary metabolite gene clusters havebeen identified, and only a few secondary metabolites have beenidentified in this species (7). These metabolites include (R)-mel-lein and derivatives (8, 9), septorines (10, 11), mycophenolic acid(8), alternariol (12), and (�)-4=-methoxy-(2S)-methylbutyro-phenone (13).

Polyketides belongs to a major class of secondary metabolitesthat are assembled from the condensation of acetate and malonateunits by PKS enzymes. They include important phytotoxins (e.g.,host-specific T toxins and the light-activated toxin cercosporin)and mycotoxins (e.g., aflatoxins and fumonisins) (2, 14). To iden-tify the P. nodorum polyketides that may potentially have a role in

pathogen-host interaction and virulence, we mined the transcrip-tomic data from a previous microarray study for PKS genes that aresignificantly upregulated during infection on wheat leaves (15).

SNOG_00477 (herein abbreviated SN477) was identified fromthe transcriptomic data to be the most upregulated PKS gene inplanta. We used a combination of reverse genetics, heterologousexpression, and biological assays to characterize the product andfunction of SN477.

MATERIALS AND METHODSP. nodorum strains and culturing conditions. The wild-type (WT) P.nodorum strain SN15 was obtained from the Department of Agricultureand Food Western Australia (DAFWA). Both the wild-type and mutantstrains generated in this study were maintained on V8-supplemented po-tato dextrose agar (PDA) plates at 20°C under a 12-h dark/12-h near-UVlight condition. For screening for mellein production, the P. nodorumwild-type and mutant strains were grown in 50 ml of defined minimalmedium (30 g sucrose, 2 g NaNO3, 1 g K2HPO4, 0.5 g KCl, 0.5 g MgSO4 ·7H2O, 0.01 g ZnSO4 · 7H2O, 0.01 g FeSO4 · 7H2O, 2.5 mg CuSO4 · 5H2Oin 1 liter, adjusted to pH 6) or modified Fries medium (30 g sucrose, 5 g

Received 21 August 2014 Accepted 14 October 2014

Accepted manuscript posted online 17 October 2014

Citation Chooi Y-H, Krill C, Barrow RA, Chen S, Trengove R, Oliver RP, Solomon PS.2015. An in planta-expressed polyketide synthase produces (R)-mellein in thewheat pathogen Parastagonospora nodorum. Appl Environ Microbiol 81:177–186.doi:10.1128/AEM.02745-14.

Editor: A. A. Brakhage

Address correspondence to Yit-Heng Chooi, [email protected].

Supplemental material for this article may be found at http://dx.doi.org/10.1128/AEM.02745-14.

Copyright © 2015, American Society for Microbiology. All Rights Reserved.

doi:10.1128/AEM.02745-14

January 2015 Volume 81 Number 1 aem.asm.org 177Applied and Environmental Microbiology

on Decem

ber 18, 2014 by Australian N

ational Univ.

http://aem.asm

.org/D

ownloaded from

ammonium tartrate, 1 g NH4NO3, 1 g KH2PO4, 0.5 g MgSO4 · 7H2O, 0.13g CaCl2, 0.1 g NaCl) (16).

Assays for P. nodorum virulence on wheat. Detached leaf assays(DLAs) were performed as described previously (17). Briefly, sections of11-day-old wheat leaves (Triticum aestivum cv. Calingiri) of approxi-mately 5 cm in length were placed on tap water agar containing 75 mg/literbenzimidazole with the ends submerged in the agar with the adaxial sideup. Each leaf was spot inoculated with 5 �l of 106 spores/ml in 0.02%(vol/vol) Tween 20, and the DLA plates were incubated at 25°C in a 12-hlight/12-h dark cycle.

Whole-plant spray assays were performed on 11-day-old leaves as de-scribed previously (18). Briefly, the plants were sprayed with 12 ml 106

spores/ml suspended in 0.02% Tween 20 or 0.02% Tween 20 withoutspores (negative control). The plants were covered and kept in the dark at20°C for 2 days, followed by a further 5 days under a 12-h dark/12-h lightregime. Disease symptoms were assessed visually and scored on a scalefrom 0 (uninfected) to 9 (widespread necrosis).

qRT-PCR. Quantitative reverse transcriptase PCR (qRT-PCR) wasperformed on total RNA samples extracted from P. nodorum-infectedwheat leaves from the DLAs at 3, 5, 7, and 10 days postinoculation (p.i.).Four leaves were collected at each time point and immediately snap-fro-zen. Isolation of total RNA was performed using the TRIzol reagent (LifeTechnologies, CA, USA). Possible DNA contamination was removed us-ing a DNA-free reagent (Life Technologies) according to the manufactur-er’s instructions.

cDNA was synthesized from extracted RNA using an iScript kit(Bio-Rad, CA, USA), according to the manufacturer’s instructions, inan Eppendorf Mastercycler thermal cycler (Eppendorf, Germany).Quantitative PCR (qPCR) was performed using iO SYBR green Super-mix (Bio-Rad, CA, USA) with the 00477RT-f/r primers (see Table S1 inthe supplemental material) in a Rotor Gene (version 6) apparatus(Corbett Research, Australia). The thermal cycling program was asfollow: 95°C for 3 min (initial denaturation), followed by 40 cycles of94°C for 10 s (denaturation), 57°C for 20 s (annealing), and 72°C for 30s (extension). Sample fluorescence was detected using a 470-nm exci-tation wavelength and a 510-nm detection wavelength. The proprie-tary Rotor Gene (version 6) software (Corbett Life Science) was used toprocess the data (comparative quantitation feature). All reactions werecarried out in duplicate.

Transformation and screening of P. nodorum mutants. The SN477-knockout (KO) cassette was synthesized using a fusion PCR approach asdescribed elsewhere (19). Upstream and downstream regions flankingSN477 were amplified with the primer sets 00477KO5-f/r and 00477KO3-f/r (see Table S1 in the supplemental material), and the phleomycin resis-tance cassette was amplified with the primers pan8-f/r from plasmidpAN8-1 (20). Transformation of P. nodorum with the SN477-knockoutcassette was achieved by the polyethylene glycol-mediated protoplasttransformation protocol described previously (21).

P. nodorum �SN477 mutants were screened by diagnostic PCR usingthe 00477Scr-f/r primers (see Table S1 in the supplemental material). Theforward primers were designed to anneal outside the 5= flanking region ofthe KO constructs, while the reverse primers anneal within the resistancecassette. Gene deletion results in a 1-kb amplicon, while ectopic mutantsare not amplified. The genomic DNA used as the template for diagnosticPCR was extracted using a Retsch MM301 ball mill and a Qiagen Bio-Sprint kit according to the manufacturers’ instructions. The PCR wascarried out using TaKaRa Ex Taq polymerase (TaKaRa Bio Inc., Japan) inan Eppendorf Mastercycler ep thermocycler. The number of integrationsin the �SN477 mutants was determined using a qPCR method describedpreviously (22). The qPCR was performed in a Rotor Gene (version 6)apparatus using the PhleoqPCRf/r primers (see Table S1 in the supple-mental material), as follows: 95°C for 3 min (initial denaturation), fol-lowed by 40 cycles of 95°C for 10 s (denaturation), 57°C for 10 s (anneal-ing), and 72°C for 20 s (extension). The proprietary Rotor Gene (version6) software was used to preprocess the data (comparative quantitation

feature). The results were imported into a spreadsheet to calculate theratio of phleomycin to actin per sample. A ratio of 1 is diagnostic ofsingle-copy integration. �SN477 mutants with single-copy integrationwere chosen for further experiments.

Heterologous expression of SN477 in Saccharomyces cerevisiae. Theintronless SN477 gene was amplified directly from the genomic DNA of P.nodorum using primers xw-SN477-F_new and xw-SN477-R. The openreading frame is based on a corrected SN477 coding sequence (depositedin GenBank under accession number KM365454), which contains an ad-ditional 65 amino acids at the N terminus compared to the original anno-tation in the NCBI GenBank and JGI databases. The forward and reverseprimers contain a 40-bp overhang for direct cloning of SN477 into theSaccharomyces cerevisiae yeast expression plasmid YEplac-ADH2p yeast-Escherichia coli shuttle vector by in vivo yeast recombination. YEplac-ADH2p contains an autoinducible adh2 promoter for protein expressionin S. cerevisiae (23). The resulting plasmid, YEplac-SN477n, was used totransform the engineered yeast strain S. cerevisiae BJ5464-NpgA, which isdeficient in vacuolar proteases and harbors an integrated copy of Asper-gillus nidulans phosphopantetheinyl transferase gene npgA (24, 25).

The BJ5464-NpgA yeast harboring plasmid YEplac-SN477n wasgrown in 50 ml yeast extract-peptone-dextrose (YPD) broth in a 250-mlshake flask along with the control culture (an empty YEplac-ADH2p vec-tor) at 28°C and 220 rpm. Five milliliters of the yeast cultures was sampledat 48 and 72 h and extracted with a mixed organic solvent containing ethylacetate, methanol, and acetic acid (89:10:1). The organic layer was dried invacuo and redissolved in 500 �l methanol for analysis by liquid chroma-tography (LC)-diode array detection (DAD)-mass spectrometry (MS).

LC-MS analysis. LC-MS metabolomics of the P. nodorum wild typeand mutants were performed on an Agilent 1200 LC system (Agilent,Santa Clara, CA, USA) coupled to an Agilent 6520 quadrupole time offlight (QTOF) system with a Jetstream electrospray ionization (ESI)source. The wild type and �SN477 mutants (SN477-KO26 and SN477-KO28) were grown in triplicate on various media and extracted with theethyl acetate-methanol-acetic acid (89:10:1) solvent mixture. The crudeextracts were dried in vacuo and redissolved in methanol for LC-MS anal-ysis. Chromatographic separation was performed at 40°C using a ZorbaxEclipse RRHD C18 column (particle size, 1.8 �m; 2.1 mm [inside diameter{i.d.} by 150 mm; Agilent) and an in-line filter. The mobile phase con-sisted of a linear gradient of 98% eluent A (0.1% [vol/vol] formic acid indeionized water) to 70% eluent B (0.1% [vol/vol] formic acid in 90%acetonitrile) over 30 min at a flow rate of 200 �l min�1. The gradient wasfollowed by a 20-min hold at 70% eluent B and was then back to 98%eluent A over 1 min, and a 14-min reequilibration was then performed.The data were collected in the m/z range from 100 to 1,000 atomic massunits in the positive mode. Data processing was performed with MassProfiler Professional software (Agilent).

For the yeast cultures expressing the SN477 gene, the analysis of themetabolite profiles was performed on an Agilent 1200 LC system coupledto a diode array detector and an Agilent 6120 quadrupole MS with an ESIsource. Chromatographic separation was performed at 30°C using a Kine-tex C18 column (particle size, 2.6 �m; 2.1 mm [i.d.] by 100 mm; Phe-nomenex, Torrance, CA, USA). The compositions of eluents A and B usedin the mobile phase were the same as those described above. The gradientconsisted of a quick increment of eluent B from 5% to 30% over 5 minbefore it was gradually increased to 100% eluent B over 25 min at a flowrate of 200 �l min�1. This was followed by a 10-min hold at 100% eluentB and was then back to 5% eluent B over 1 min, and a 15-min reequili-bration was then performed. The metabolite profile of the yeast culturewas compared with the profiles for purified (R)-mellein from P. nodorum(below) and an authentic (R)-mellein standard (Cayman Chemical, MI,USA).

Compound isolation and structural characterization. For produc-tion of compounds, P. nodorum SN15 was grown in modified Fries liquidmedium as described previously (8, 16). Briefly, about 108 spores wereinoculated in six 2-liter flasks containing 500 ml liquid medium and in-

Chooi et al.

178 aem.asm.org January 2015 Volume 81 Number 1Applied and Environmental Microbiology

on Decem

ber 18, 2014 by Australian N

ational Univ.

http://aem.asm

.org/D

ownloaded from

cubated at 22°C for 48 h with shaking, followed by stationary incubationin the dark at 22°C for another 2 weeks. The culture was extracted with amixed organic solvent containing ethyl acetate, methanol, and acetic acid(89:10:1), and the organic layer was dried in vacuo. The crude extract wasthen fractionated on a Reveleris flash chromatography system (Grace,MD, USA) using a hexane-ethyl acetate-methanol gradient. Eluting com-pounds were monitored with a UV detector (245 nm and 310 nm) and anevaporative light scattering detector (ELSD) coupled to the flash chroma-tography system. Fractions corresponding to single peaks were sampledand analyzed by LC-DAD-MS. Two of the fractions, identified to containa single peak with m/z 179 and m/z 193, respectively, were dried in vacuoand redissolved in deuterated chloroform (CDCl3) for nuclear magneticresonance (NMR) analysis on an Inova 500 MHz NMR system (Varian,CA, USA). The specific optical rotations of the two compounds wererecorded on a model 343 polarimeter (PerkinElmer, MA, USA) at 589 nm(sodium D line). The specific optical rotation of (R)-mellein was alsocompared to that of an authentic standard (Cayman Chemical).

Phytotoxicity and antigerminative activity assays. The phytotoxicityof (R)-mellein and (R)-O-methylmellein was assayed on the leaves of11-day-old wheat seedlings that were grown in 10-cm planting pots at20°C under a 12-h light/12-h dark cycle regime. Briefly, approximately 80�l of each compound in 2% methanol solution was infiltrated on theadaxial face of the leaves at various concentrations (25, 50, 100, 200 �g/ml) using a 1-ml-volume syringe. Experiments were performed in tripli-cate, and the infiltrated leaves were examined for the presence of necrosisor chlorosis after 24 and 48 h.

To test the ability of the compounds to inhibit seed germination,grains of wheat seed (Triticum aestivum cv. Lincoln) were first surfacesterilized briefly in a solution containing 10% ethanol and 1% hydrogenperoxide. A single grain of the surface-sterilized wheat seed was thenplaced on agar slants consisting of 1.5 ml of tap water agar supplementedwith either (R)-mellein or (R)-O-methylmellein at a 200-�g/ml final con-centration (agar containing 2% methanol was used as a control). Barrelmedic (Medicago truncatula cv. Jemalong A17) seeds were also tested us-ing the same protocol to determine the host specificity of the compounds.The assays were performed in triplicate, and the progress of seed germi-nations was observed and recorded on a daily basis for 1 week.

Phylogenetic analysis. The phylogeny of the PKSs was inferred usingconserved �-ketosynthase (KS) domains of the PKS protein sequences(corresponding to amino acids 67 to 463 of the corrected SN477 proteinsequence). The PKSs used in the phylogenetic analysis included four char-acterized fungal 6-methylsalicylic acid synthases (6MSASs) and all NCBIBLASTp hits (queried with the SN477 KS domain) from the GenBankdatabase with a percent identity and score that were above those for thecharacterized 6MSAS (ATX) from Aspergillus terreus (see Table S3 in thesupplemental material). To investigate the distribution of SN477 homo-logues among Dothideomycetes, all BLASTp hits (queried with the SN477KS domain) from the JGI Dothideomycetes genome database (http://genome.jgi-psf.org/dothideomycetes/dothideomycetes.info.html) withan E value of �10�100 were included in the phylogenetic analysis. Sixcharacterized bacterial iterative PKSs known to produce 6-methylsalicylicacid (6-MSA), orsellinic acid, naphthoic acids, and (R)-mellein were in-cluded for comparison, while seven selected fungal highly reducing (HR)PKSs and PKS-nonribosomal peptide synthetase (NRPS) hybrids wereincluded for rooting of the trees (see Table S3 in the supplemental mate-rial). A total of 50 KS domain sequences were aligned using the MUSCLEalignment program embedded in Geneious (version 7.17) software (Bio-matters Ltd., Auckland, New Zealand) (26). The resulting multiple-se-quence alignment was used for phylogenetic analysis.

The KS domain phylogenetic tree in Fig. 4 was constructed using theGeneious Tree Builder program embedded in Geneious (version 7.17)software with the neighbor-joining method (27). The tree was con-structed with 1,000 bootstrap replicates, and branches corresponding topartitions that were reproduced in less than 50% of bootstrap replicateswere collapsed. For comparison, a maximum likelihood tree was con-

structed using the RAxML (version 7.2.8) plug-in (28) in Geneious (ver-sion 7.17) software (see Fig. S2 in the supplemental material).

Nucleotide sequence accession number. The corrected SN477 codingsequence has been deposited in GenBank under accession numberKM365454.

RESULTSSNOG_00477 is the most highly expressed PKS gene in planta.Ipcho et al. previously used a custom microarray to obtain theglobal transcriptomic profile of 16,586 nuclear gene models of P.nodorum (15). The transcriptomic profiles of the genes expressedduring infection of detached wheat leaves at time points spanningfrom early infection to sporulation (in planta) and during growthon defined minimal medium (in vitro) were compared. The studyshowed that 2,882 genes were expressed at a higher level in plantaand 3,630 were expressed more highly in vitro. Using these tran-scriptome data, we extracted the gene expression profiles of all the23 PKS genes in P. nodorum (see Table S2 in the supplementalmaterial). The analysis showed that SNOG_00477 (abbreviatedSN477) was the most upregulated PKS gene in planta compared toits level of expression in vitro. Although SN477 was expressed bothin planta and in vitro, its expression in planta was, on average,3-fold higher than that in vitro. The expression of SN477 was thehighest at 3 days postinoculation (p.i.) on wheat leaves, prior todecreasing to approximately the same level as that in vitro on day 10 inplanta (see Fig. S1A and Table S2 in the supplemental material).

The transcription of SN477 in planta, using the same detachedleaf assay (DLA) used in the previous microarray study, was vali-dated by quantitative reverse transcriptase PCR (qRT-PCR). Thetranscription of two additional PKS genes (SNOG_05791 andSNOG_011272) was included in the qRT-PCR experiments forcomparison. The qRT-PCR results showed expression profilessimilar to those observed in the microarray data for SNOG_05791and SNOG_011272, but SN477 was expressed at the highest levelat 5 days p.i. (see Fig. S1B in the supplemental material). The highlevel of SN477 expression in planta at the early infection stagesuggests a possible role of the polyketide product in establishinginfection and prompted us to further investigate the product andfunction of this PKS gene.

Analysis of the SN477 sequence showed that it encodes a typicalpartially reducing (PR) PKS with �-ketosynthase (KS), acyltrans-ferase (AT), dehydratase (DH), ketoreductase (KR), and acyl ca-reer protein (ACP) domains (7, 29–31). Among the characterizedfungal PKSs, SN477 was the most similar to the partially reducingPKS ATX from Aspergillus terreus, which synthesizes 6-methylsali-cylic acid (6-MSA) (32). It shares 54% amino acid sequence iden-tity with A. terreus ATX and has a domain architecture identical tothat of A. terreus ATX. Thus, it was considered that SN477 likelyencodes the biosynthesis of a partially reduced polyketide com-pound similar to 6-MSA.

Generation and characterization of P. nodorum �SN477mutants. P. nodorum �SN477 mutants were generated by poly-ethylene glycol-mediated transformation with the SN477-knock-out cassette described above. Diagnostic PCR identified five pos-itive transformants where SN477 had been deleted by doublehomologous crossover recombination and replaced with a phleo-mycin resistance marker (ble). Two of the positive mutants,SN477-KO26 and SN477-KO28, were confirmed to contain only asingle-copy integration by qPCR. No differences in the growthrate between the two mutants and the wild-type (WT) P. nodorum

Mellein Synthase Gene from Parastagonospora nodorum

January 2015 Volume 81 Number 1 aem.asm.org 179Applied and Environmental Microbiology

on Decem

ber 18, 2014 by Australian N

ational Univ.

http://aem.asm

.org/D

ownloaded from

strains were observed. The growth rate and colony morphology onV8-PDA and minimal medium plates were indistinguishable fromthose of the WT. There was also no significant difference in spo-rulation. Thus, the two �SN477 mutants did not demonstrate anyobservable growth defect.

Metabolic profiling revealed that SN477 is involved in thesynthesis of (R)-mellein. On the basis of the previous microarraydata, the SN477 gene showed modest expression on defined min-imal medium agar at 4 and 16 days postinoculation. Since inplanta metabolite analysis would be complicated by the complexmetabolite background in the wheat leaves, we first compared themetabolic profiles between the wild-type and the mutants grownon defined minimal medium in triplicate using LC-QTOF-MS in

positive mode. Using subtractive metabolomics, we were able toidentify two peaks (m/z 179.07 and m/z 193.08) in the P. nodorumwild-type culture that were consistently absent in the �SN477mutants. However, these two peaks were present in only moderateto low abundance in the WT culture on defined minimal medium(data not shown).

Other media and culture conditions were tested to improve theproduction of the two compounds. It was found that these twopeaks appeared as major metabolites in the P. nodorum WT cul-ture when grown in modified Fries liquid medium under station-ary culture conditions, similar to the findings described previously(8, 16). As expected, the two peaks were absent in the correspond-ing cultures of the �SN477 mutants (Fig. 1A and B).

FIG 1 Comparative metabolomics analysis of P. nodorum strains by LC-MS. (A, B) Base peak chromatogram (BPC) and extract ion chromatogram (EIC) ofculture extracts from the P. nodorum wild type (A) and �SN477 mutant (B); (C) structural and molecular formulas of (R)-mellein and (R)-O-methylmellein.

Chooi et al.

180 aem.asm.org January 2015 Volume 81 Number 1Applied and Environmental Microbiology

on Decem

ber 18, 2014 by Australian N

ational Univ.

http://aem.asm

.org/D

ownloaded from

To confirm the identities of the two compounds, the P. nodo-rum WT culture was scaled up to 3 liters and extracted after 21days incubation. Using flash chromatography with a silica col-umn, approximately 18 mg and 25 mg of the two compoundscorresponding to m/z 179.07 and m/z 193.08, respectively, wereisolated. NMR spectrum analysis established that the compoundwith m/z 179.07 [M � H]� is mellein, while the compound withm/z 193.08 [M � H]� corresponds to O-methylmellein (see Ta-bles S4 and S5 and Fig. S3 and S4 in the supplemental material) (8,33, 34). Optical rotation analysis showed that both compoundsare the negative (�) enantiomers which have an R configurationat C-3 (8, 33). Together, the data established that the two com-pounds are (R)-(�)-mellein [(3R)-(�)-3,4-dihydro-8-hydroxy-3-methylisocoumarin] and (R)-O-methylmellein [(3R)-(�)-3,4-dihydro-8-methoxy-3-methylisocoumarin] (Fig. 1C). Both ofthese two compounds have been identified previously from P.nodorum (8). Hence, these data have confirmed that SN477 is thePKS gene involved in the biosynthesis of (R)-mellein and (R)-O-methylmellein.

Heterologous expression of SN477 in yeast resulted in theproduction of (R)-mellein. The reverse genetics and metabolo-mics approaches described above confirmed that SN477 has a rolein the synthesis of (R)-mellein. To determine if SN477 is solelyresponsible, we cloned the intron-less SN477 gene from P. nodo-rum into the YEplac-ADH2p vector for expression in the S. cerevi-siae BJ5464-NpgA strain. This S. cerevisiae BJ5464-NpgA strainhas been successfully used for the heterologous expression of largemegasynthases from fungi (25, 29). Our initial attempt, based onthe SN477 annotation in the NCBI and JGI databases, failed toproduce (R)-mellein in the yeast strain (data not shown). Align-ment of the protein sequence with other fungal 6MSASs showedthat the annotated SN477 protein sequence has a slightly shorter Nterminus. A more detailed analysis identified another start codon195 bp upstream of the originally annotated 5= end, which corre-sponded to an additional 65 amino acids supported by homologyto other 6MSASs (the corrected sequence is deposited in GenBankunder accession number KM365454). Recloning of the SN477coding sequence with the inclusion of the missing 195 bp resultedin the plasmid YEplac-SN477n. Transformation of S. cerevisiaeBJ5464-NpgA with YEplac-SN477n resulted in the production of

a new peak at 48 and 72 h upon inoculation of the seed culture inYPD liquid medium (Fig. 2). The new peak corresponded to theretention time, UV spectrum, and m/z of the purified (R)-melleinfrom P. nodorum and an authentic (R)-mellein standard. Thisresult conclusively demonstrated that SN477 is a mellein synthase(MLNS) and is the only enzyme required for the production of(R)-mellein.

Do (R)-mellein and (R)-O-methylmellein play a role in viru-lence or affect plant development? The ability of the �SN477mutant strains to cause disease was assessed using the whole-plantspray assay. No significant difference in disease symptoms be-tween wheat leaves inoculated with �SN477 mutants and wheatleaves inoculated with the controls (wild-type and ectopic integra-tion strains) was observed. Subsequent reanalysis using a detachedleaf pathogenicity assay confirmed that SN477 does not have a rolein virulence.

The phytotoxic activity of the purified (R)-mellein and (R)-O-methylmellein on wheat was then assayed. No necrosis was ob-served on wheat leaves treated with up to 200 �g/ml of (R)-mel-lein and O-methylmellein by leaf infiltration. The capacity of (R)-mellein and O-methylmellein to affect seed germination was alsoassessed. Complete inhibition of wheat seed germination was ob-served for (R)-mellein at 200 �g/ml (Fig. 3). However, (R)-O-methylmellein at the same concentration exhibited only moderateto low antigerminative activity. Likewise, 200 �g/ml of (R)-mel-lein completely inhibited the germination of barrel medic seeds,while (R)-O-methylmellein exhibited only moderate inhibition.These data indicate that the inhibitory activity is neither host spe-cific nor restricted to monocots or dicots.

Phylogeny of (R)-mellein synthase and its distributionamong Dothideomycetes. Phylogenetic analysis of fungal PR-PKSs and bacterial iterative PKS was inferred by either the neigh-

FIG 2 Heterologous production of (R)-mellein in S. cerevisiae BJ5464-NpgA.The high-pressure liquid chromatography–DAD profile of the culture extractfrom S. cerevisiae BJ5465-NpgA expressing SN477 (YEp-SN477n) comparedto the profiles of the culture extracts from the empty vector control (YEplac-ADH2p) and the (R)-mellein standard is shown. mAU, milli-absorbanceunits.

FIG 3 Effects of (R)-mellein and (R)-O-methylmellein on seed germination.(A) Wheat (Triticum aestivum) on day 7; (B) barrel medic (Medicago trunca-tula) on day 5. Both (R)-mellein and 3-O-methylmellein were present at a finalconcentration of 200 �g/ml. Controls contained 4% methanol.

Mellein Synthase Gene from Parastagonospora nodorum

January 2015 Volume 81 Number 1 aem.asm.org 181Applied and Environmental Microbiology

on Decem

ber 18, 2014 by Australian N

ational Univ.

http://aem.asm

.org/D

ownloaded from

bor-joining (27) or the randomized accelerated maximum likeli-hood (RAxML) (28) method and resulted in consensus trees withsimilar topologies (Fig. 4; see also Fig. S2 in the supplementalmaterial). The separation of a clade containing bacterial iterativePKSs and fungal PR-PKSs from fungal highly reducing (HR) PKSswas well supported. The fungal PR-PKSs were divided into twolarge clades, one with SN477 and the other with the characterized6MSASs. Four fungal PR-PKS genes have so far been matched tothe polyketide products. All four of them, Penicillium griseofulvumMSAS (35), A. terreus ATX (32), Glarea lozoyensis PKS2 (36), andAspergillus niger PKS48 (37), have been shown to encode theproduction of 6-MSA. Thus, SN477 is the first fungal PR-PKSshown to produce a polyketide compound other than 6-MSAthat is instead linked to mellein. It is likely that most of theother fungi in the mellein clade may also produce this com-pound.

A previous phylogenetic study has inferred that the fungi ac-quired the 6MSAS-type PR-PKSs by horizontal gene transfer(HGT) from bacteria (38). Here, our analysis showed that SN477and its homologues are more closely related to fungal 6MSASs

than Saccharopolyspora erythraea SACE_5532, which synthesizesthe same polyketide product, (R)-mellein (33). This suggests aPKS convergent evolution after the initial HGT, where the fungalMLNSs evolved from an ancestral fungal 6MSAS independentlyfrom the bacterial MLNS SACE_5532.

A homologous gene search and phylogenetic analysis indicatethat SN477 homologues are widely but not uniformly distributedamong the Dothideomycetes class of fungi. Thirteen of the 63Dothideomycetes in the JGI Dothideomycetes genome databaseand 3 nondothideomyceteous (sordariomycete) species in theGenBank database contain a copy of the SN477 homologue (Fig.4; see also Table S3 in the supplemental material). SN477 and itsclosest homologues (9 from Dothideomycetes and 3 from Sordar-iomycetes) formed a distinct clade, and an additional 4 dothideo-mycete homologues formed a closely related sister clade separatedfrom the fungal 6-MSASs. Neofusicoccum parvum has previouslybeen shown to produce (3R,4R)-(�)-4-hydroxymellein and(3R,4S)-(�)-4-hydroxymellein (39). Out of the three PR-PKSs inN. parvum, UCRNP2_6207, which grouped with SN477, is mostlikely to be the PR-PKS responsible for production of the mellein

FIG 4 Phylogeny and distribution of (R)-mellein synthase homologues in fungi in relation to other fungal PR-PKSs and the bacterial counterparts. The KSdomain consensus tree was constructed using the neighbor-joining method with 1,000 bootstrap replicates (numbers at the nodes indicate bootstrap values).Branches corresponding to partitions that were reproduced in less than 50% of bootstrap replicates are collapsed. †, SNOG_00470 is (R)-mellein synthasecharacterized in this study; *, fungal 6-MSASs characterized in other studies; d, homologues from dothideomycete fungi (see Table S3 in the supplementalmaterial). The tree is rooted with fungal HR-PKSs and hybrid PKS-NRPSs. A maximum likelihood tree with a similar topology is presented in Fig. S2 in thesupplemental material.

Chooi et al.

182 aem.asm.org January 2015 Volume 81 Number 1Applied and Environmental Microbiology

on Decem

ber 18, 2014 by Australian N

ational Univ.

http://aem.asm

.org/D

ownloaded from

precursor. This observation, in turn, supports the postulation thatthe 12 PR-PKSs (9 from Dothideomycetes and 3 from Sordario-mycetes) in the same clade as SN477 are functional homologues ofSN477 and likely synthesize mellein (Fig. 4).

DISCUSSION

In this study, we have functionally characterized SN477, the PKSgene most highly expressed in planta. SN477 was shown to encodea PR-PKS that synthesizes (R)-mellein and, as such, is the first(R)-mellein synthase identified in fungi. SN477 harbors a typicalPR-PKS domain architecture (KS-AT-thiohydrolase [TH]-KR-ACP) similar to that of 6MSASs (29–31). The biosynthesis of (R)-mellein is highly parallel to that of 6-MSA but requires additionalchain elongation and keto reduction steps (Fig. 5). The nascentpentaketide intermediate then undergoes an aldol cyclization andis aromatized via dehydration. The stereospecific lactonizationand release of the matured polyketide product are likely catalyzedby the TH domain in MLNS, similar to that identified in 6MSASs(33, 40). The (R)-O-methylmellein isolated from P. nodorum inthis study is most likely to be derived from (R)-mellein via anadditional methylation at the hydroxyl group. Interestingly, noO-methyltransferase gene is encoded in the vicinity of SN477 onthe chromosome. Thus, the O-methylation is likely to be catalyzedby an endogenous O-methyltransferase encoded elsewhere in thegenome of P. nodorum.

Since the discovery of the first 6MSAS from Penicillium patu-lum (35), subsequent homologues of PR-PKSs characterized infungi have been shown to produce 6-MSA (32, 36, 37). Thus,SN477 is the first PR-PKS that has been demonstrated to producea polyketide compound other than 6-MSA. The discovery andcharacterization of SN477 as an MLNS hint that there may beremaining undiscovered chemical diversity in this group of fungalPKSs. Our phylogenetic analysis inferred that the fungal and bac-terial MLNSs (sharing 37% amino acid sequence identity) haveevolved independently from ancestral fungal and bacterial6MSASs, respectively. A similar convergent evolution of bacterialand fungal type I PKSs has previously been observed in orsellinicacid biosynthesis. While the fungal orsellinic acid synthases aretypical fungal nonreducing PKSs (41–43), the bacterial orsellinic

acid synthases are related to bacterial iterative type I PKSs (homol-ogous to fungal PR-PKSs) and are likely to have evolved fromancestral bacterial 6MSASs via mutations in the KR domain (44–46). From a protein engineering perspective, the convergent evo-lution of MLNSs implies that there is more than one way in natureto modify the chain length specificity and ketoreduction regiose-lectivity of iterative PKSs.

(R)-Mellein (also known as ochracin) and its dihydroisocou-marin derivatives are widespread in fungi, particularly among theDothideomycetes. (R)-Mellein was first isolated from Aspergillusmelleus in 1933 (47), and its structure and stereochemistry werelater determined in 1955 and 1968, respectively (48, 49). Subse-quently, mellein and its derivatives have been isolated from a va-riety of fungi, including Septoria nodorum (synonym Parastagono-spora nodorum) (8), Cercospora taiwanensis (50), Botryosphaeriaspp. (51–53), Phoma tracheiphila (54), Xylaria longiana (55), Mi-crosphaeropsis sp. (56), Sphaeropsis sapinea (57), a Nigrospora sp.(58), Apiospora montagnei (59), and Pezicula livida, a Plectopho-mella sp., and Cryptosporiopsis spp. (60). Despite the widespreadoccurrence of mellein in fungi, this is the first time that a fungalPKS gene responsible for the synthesis of mellein has been identi-fied. Outside of the fungal kingdom, mellein has been found inactinomycetes (33) and insects (61–63).

(R)-Mellein and its derivatives have been shown to display amyriad of bioactivities. It is weak to moderately active against arange of bacteria and fungi (56, 58, 60). (R)-Mellein has also beenshown to exhibit antiviral activity against hepatitis C virus (64),antiparasitic activity against Schistosoma mansoni (59), andzootoxicity against Artemia salina larvae (54). It has been reportedto be a trail pheromone of ants (61, 63), while the larvae of theparasitoid wasp Ampulex compressa apparently impregnate thecockroach host with (R)-mellein as one of the antimicrobials fordefense against entomopathogenic microbes (62). Above all, themost commonly reported bioactivity of mellein and its derivativesis their non-host-specific toxicity against plants. (R)-Mellein andhydroxymelleins have been isolated as phytotoxins of the appleand grapevine pathogen Botryosphaeria obtusa (53, 65), the pinepathogen S. sapinea (57), the citrus pathogen Phoma tracheiphila(54), and the grapevine canker agent N. parvum (39). It has alsobeen shown to induce phytotoxic symptoms on tomato cuttingswhen it is used at 100 �g/ml (54) and causes necrosis on detachedgrape (Vitis vinifera) leaf at concentrations as low as 3 �g/ml (65).On the other hand, a recent study showed that up to 500 �g/ml of(R)-mellein is required to cause necrosis on the calli of V. viniferaand induces defense gene expression (66). Production of O-meth-ylmellein was induced in an agar medium coculture with Botryo-sphaeria obtusa and Eutypa lata, the two wood-decaying fungiinvolved in esca disease of grapevine (34). The O-methylmelleinisolated in the study showed a strong antigerminative effectagainst garden cress at a 0.001% (10 �g/ml) concentration. Forwheat, Keller et al. reported that (R)-mellein inhibits the growth ofwheat embryo culture at 50 �g/ml (67).

Here, we showed that the P. nodorum �SN477 mutants did nothave observable differences in virulence against wheat comparedwith that of the P. nodorum WT. The phytotoxic effects of (R)-mellein and O-methylmellein were also tested by infiltration of thecompounds on wheat leaves on whole plants, but neither com-pound induced any necrotic or chlorotic symptoms at concentra-tions up to 200 �g/ml. Thus, the results together rule out thepossibility that (R)-mellein and O-methylmellein play any impor-

FIG 5 Biosynthesis of (R)-mellein in comparison to that of 6-MSA. CoA,coenzyme A.

Mellein Synthase Gene from Parastagonospora nodorum

January 2015 Volume 81 Number 1 aem.asm.org 183Applied and Environmental Microbiology

on Decem

ber 18, 2014 by Australian N

ational Univ.

http://aem.asm

.org/D

ownloaded from

tant role in lesion development on wheat. Curiously, when testedfor antigerminative activity, (R)-mellein at 200 �g/ml completelyinhibited the germination of wheat and barrel medic seeds. Theresults suggest that (R)-mellein at 200 �g/ml may interfere withthe cellular pathway involved in germination or hormone signal-ing in plants. This result is consistent with the previous observa-tion that (R)-mellein suppressed the elongation of roots and co-leoptiles of wheat seedlings at concentrations ranging from 50 to200 �g/ml (16). In addition, it has been observed that melleinslows the cell cycle, with the cell having an extended mitotic phase(68). Interestingly, the biological sources of (R)-mellein are over-represented in fungal plant pathogens (39, 53, 54, 57, 65) andendophytes (51, 52, 58, 59). Whether mellein plays a role in thisplant-fungus interaction and why SN477 is strongly upregulatedin planta are questions that warrant future investigation.

The characterization of SN477 encoding a MLNS prompted usto survey its distribution in other fungal genomes, in particular, inthose of the Dothideomycetes, many members of which are plantpathogens. The survey showed that about one-fifth of thedothideomycete genomes in the JGI database encode a close ho-molog of SN477. In contrast, among the nondothideomycete fun-gal genomes, SN477 homologues can be found in only two sor-dariomycete genera (Pestalotiopsis fici [PFICI] and Stachybotrysspp. [S40285 and S7711]). Interestingly, the four Botryosphaeri-aceae species (N. parvum [UCNRP2], Aplosporella prunicola[Aplpr1], Botryosphaeria dothidea [Botdo1], Macrophominaphaseolina [MPH]) in the phylogenetic tree contain both MLNSand 6MSAS homologues. This suggests that the fungal MLNSsmay have originated from a typical gene duplication and func-tional divergence from an ancestral Dothideomycetes fungus andboth orthologues were retained in some of the Botryosphaeriaceaespecies. It also suggests that the PKS gene has frequently been lostfrom many taxa.

ACKNOWLEDGMENTS

Y.-H.C. is supported by an Australian Research Council Discovery (ARC)Early Career Researcher Award (DECRA) fellowship (DE130101350).P.S.S. is an ARC Future Fellow. R.P.O. and P.S.S. gratefully acknowledgethe funding support from the Australian Grains Research and Develop-ment Corporation (GRDC).

LC-QTOF-MS analysis was performed at The Australian NationalUniversity Research School of Biology Mass Spectrometry Facility.

We thank Nancy Da Silva and Yi Tang for the S. cerevisiae BJ5464-NpgA strain and YEplac-ADH2p plasmid. Y.-H.C. acknowledges M. JordiMuria-Gonzalez for assistance with MS analysis and helpful discussions.

REFERENCES1. Muria-Gonzalez MJ, Chooi YH, Breen S, Solomon PS. 29 May 2014. The

past, present and future of secondary metabolite research in the Dothideo-mycetes. Mol Plant Pathol http://dx.doi.org/10.1111/mpp.12162.

2. Mobius N, Hertweck C. 2009. Fungal phytotoxins as mediators of viru-lence. Curr Opin Plant Biol 12:390 –398. http://dx.doi.org/10.1016/j.pbi.2009.06.004.

3. Collemare J, Lebrun M-H. 2011. Fungal secondary metabolites: ancienttoxins and novel effectors in plant-microbe interactions, p 379 – 402. InMartin F, Kamoun S (ed), Effectors in plant-microbe interactions. Wiley-Blackwell, Oxford, United Kingdom.

4. Bohnert M, Wackler B, Hoffmeister D. 2010. Spotlights on advances inmycotoxin research. Appl Microbiol Biotechnol 87:1–7. http://dx.doi.org/10.1007/s00253-010-2565-8.

5. Oliver RP, Friesen TL, Faris JD, Solomon PS. 2012. Stagonospora nodorum:from pathology to genomics and host resistance. Annu Rev Phytopathol 50:23–43. http://dx.doi.org/10.1146/annurev-phyto-081211-173019.

6. Hane JK, Lowe RG, Solomon PS, Tan KC, Schoch CL, Spatafora JW,

Crous PW, Kodira C, Birren BW, Galagan JE, Torriani SF, McDonaldBA, Oliver RP. 2007. Dothideomycete plant interactions illuminated bygenome sequencing and EST analysis of the wheat pathogen Stagonosporanodorum. Plant Cell 19:3347–3368. http://dx.doi.org/10.1105/tpc.107.052829.

7. Chooi YH, Muria-Gonzalez MJ, Solomon PS. 2014. A genome-widesurvey of the secondary metabolite biosynthesis genes in the wheat patho-gen Parastagonospora nodorum. Mycology 5:192–206. http://dx.doi.org/10.1080/21501203.2014.928386.

8. Devys M, Bousquet JF, Kollmann A, Barbier M. 1980. Dihydroisocou-marins and mycophenolic acid of culture medium of a phytopathogenicfungus, Septoria nodorum. Phytochemistry 19:2221–2222. http://dx.doi.org/10.1016/S0031-9422(00)82234-7.

9. Devys M, Barbier M, Bousquet JF, Kollmann A. 1992. Isolation of thenew (�)-(3R,4S)-4-hydroxymellein from the fungus Septoria nodorumBerk. Z Naturforsch C 47:779 –781.

10. Devys M, Barbier M, Kollmann A, Bousquet JF. 1982. Septorine andN-methoxy septorine, substituted pyrazines from the fungus Septorianodorum Berk. Tetrahedron Lett 23:5409 –5412. http://dx.doi.org/10.1016/0040-4039(82)80143-3.

11. Devys M, Bousquet JF, Kollmann A, Barbier M. 1978. Isolation of anew pyrazine, septorine, from culture medium of phytopathogen fun-gus Septoria nodorum Berk. C R Hebd Seances Acad Sci C 286:457– 458.

12. Tan K-C, Trengove RD, Maker GL, Oliver RP, Solomon PS. 2009.Metabolite profiling identifies the mycotoxin alternariol in the pathogenStagonospora nodorum. Metabolomics 5:330 –335. http://dx.doi.org/10.1007/s11306-009-0158-2.

13. Yang XL, Awakawa T, Wakimoto T, Abe I. 2013. Induced biosynthesesof a novel butyrophenone and two aromatic polyketides in the plantpathogen Stagonospora nodorum. Nat Prod Bioprospect 3:141–144. http://dx.doi.org/10.1007/s13659-013-0055-2.

14. Huffman J, Gerber R, Du L. 2010. Recent advancements in the biosyn-thetic mechanisms for polyketide-derived mycotoxins. Biopolymers 93:764 –776. http://dx.doi.org/10.1002/bip.21483.

15. Ipcho SV, Hane JK, Antoni EA, Ahren D, Henrissat B, Friesen TL,Solomon PS, Oliver RP. 2012. Transcriptome analysis of Stagonosporanodorum: gene models, effectors, metabolism and pantothenate dispens-ability. Mol Plant Pathol 13:531–545. http://dx.doi.org/10.1111/j.1364-3703.2011.00770.x.

16. Bousquet JF, Skajennikoff M. 1974. Isolement et mode d’action d’unephytotoxine produite en culture par Septoria nodorum Berk. J Phytopathol80:355–360. http://dx.doi.org/10.1111/j.1439-0434.1974.tb02768.x.

17. Benedikz PW, Mappledoram CJ, Scott PR. 1981. A laboratory techniquefor screening cereals for resistance to Septoria nodorum using detachedseedling leaves. Trans Br Mycol Soc 77:667– 669. http://dx.doi.org/10.1016/S0007-1536(81)80126-X.

18. Solomon PS, Waters ODC, Simmonds J, Cooper RM, Oliver RP. 2005.The Mak2 MAP kinase signal transduction pathway is required for patho-genicity in Stagonospora nodorum. Curr Genet 48:60 – 68. http://dx.doi.org/10.1007/s00294-005-0588-y.

19. Solomon PS, Rybak K, Trengove RD, Oliver RP. 2006. Investigating therole of calcium/calmodulin-dependent protein kinases in Stagonosporanodorum. Mol Microbiol 62:367–381. http://dx.doi.org/10.1111/j.1365-2958.2006.05380.x.

20. Mattern I, Punt P, Van den Hondel C. 1988. A vector of Aspergillustransformation conferring phleomycin resistance. Fungal Genet Newsl35:25.

21. Solomon PS, Tan KC, Sanchez P, Cooper RM, Oliver RP. 2004. Thedisruption of a Galpha subunit sheds new light on the pathogenicity ofStagonospora nodorum on wheat. Mol Plant Microbe Interact 17:456 – 466.http://dx.doi.org/10.1094/MPMI.2004.17.5.456.

22. Solomon PS, Ipcho SV, Hane JK, Tan KC, Oliver RP. 2008. A quanti-tative PCR approach to determine gene copy number. Fungal Genet Rep55:5– 8.

23. Lee KM, DaSilva NA. 2005. Evaluation of the Saccharomyces cerevisiaeADH2 promoter for protein synthesis. Yeast 22:431– 440. http://dx.doi.org/10.1002/yea.1221.

24. Lee KK, Da Silva NA, Kealey JT. 2009. Determination of the extent ofphosphopantetheinylation of polyketide synthases expressed in Esche-richia coli and Saccharomyces cerevisiae. Anal Biochem 394:75– 80. http://dx.doi.org/10.1016/j.ab.2009.07.010.

25. Ma SM, Li JW, Choi JW, Zhou H, Lee KK, Moorthie VA, Xie X, KealeyJT, Da Silva NA, Vederas JC, Tang Y. 2009. Complete reconstitution of

Chooi et al.

184 aem.asm.org January 2015 Volume 81 Number 1Applied and Environmental Microbiology

on Decem

ber 18, 2014 by Australian N

ational Univ.

http://aem.asm

.org/D

ownloaded from

a highly reducing iterative polyketide synthase. Science 326:589 –592.http://dx.doi.org/10.1126/science.1175602.

26. Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S,Buxton S, Cooper A, Markowitz S, Duran C, Thierer T, Ashton B,Meintjes P, Drummond A. 2012. Geneious Basic: an integrated andextendable desktop software platform for the organization and analysis ofsequence data. Bioinformatics 28:1647–1649. http://dx.doi.org/10.1093/bioinformatics/bts199.

27. Saitou N, Nei M. 1987. The neighbor-joining method: a new method forreconstructing phylogenetic trees. Mol Biol Evol 4:406 – 425.

28. Stamatakis A. 2006. RAxML-VI-HPC: maximum likelihood-based phy-logenetic analyses with thousands of taxa and mixed models. Bioinformat-ics 22:2688 –2690. http://dx.doi.org/10.1093/bioinformatics/btl446.

29. Chooi YH, Tang Y. 2012. Navigating the fungal polyketide chemicalspace: from genes to molecules. J Org Chem 77:9933–9953. http://dx.doi.org/10.1021/jo301592k.

30. Cox RJ. 2007. Polyketides, proteins and genes in fungi: programmednano-machines begin to reveal their secrets. Org Biomol Chem 5:2010 –2026. http://dx.doi.org/10.1039/b704420h.

31. Schuemann J, Hertweck C. 2009. Biosynthesis of fungal polyketides, p331–351. In Anke T, Weber D (ed), The Mycota: physiology and genetics,vol XV. Springer-Verlag, Berlin, Germany.

32. Fujii I, Ono Y, Tada H, Gomi K, Ebizuka Y, Sankawa U. 1996.Cloning of the polyketide synthase gene atX from Aspergillus terreus andits identification as the 6-methylsalicylic acid synthase gene by heter-ologous expression. Mol Gen Genet 253:1–10. http://dx.doi.org/10.1007/s004380050289.

33. Sun H, Ho CL, Ding F, Soehano I, Liu XW, Liang ZX. 2012. Synthesisof (R)-mellein by a partially reducing iterative polyketide synthase. J AmChem Soc 134:11924 –11927. http://dx.doi.org/10.1021/ja304905e.

34. Glauser G, Gindro K, Fringeli J, De Joffrey JP, Rudaz S, Wolfender JL.2009. Differential analysis of mycoalexins in confrontation zones of grape-vine fungal pathogens by ultrahigh pressure liquid chromatography/time-of-flight mass spectrometry and capillary nuclear magnetic resonance. JAgric Food Chem 57:1127–1134. http://dx.doi.org/10.1021/jf8033539.

35. Beck J, Ripka S, Siegner A, Schiltz E, Schweizer E. 1990. The multifunc-tional 6-methylsalicylic acid synthase gene of Penicillium patulum. Its genestructure relative to that of other polyketide synthases. Eur J Biochem192:487– 498.

36. Lu P, Zhang A, Dennis LM, Dahl-Roshak AM, Xia YQ, Arison B, An Z,Tkacz JS. 2005. A gene (pks2) encoding a putative 6-methylsalicylic acidsynthase from Glarea lozoyensis. Mol Genet Genomics 273:207–216. http://dx.doi.org/10.1007/s00438-005-1132-y.

37. Holm DK, Petersen LM, Klitgaard A, Knudsen PB, Jarczynska ZD,Nielsen KF, Gotfredsen CH, Larsen TO, Mortensen UH. 2014. Molec-ular and chemical characterization of the biosynthesis of the 6-MSA-derived meroterpenoid yanuthone D in Aspergillus niger. Chem Biol 21:519 –529. http://dx.doi.org/10.1016/j.chembiol.2014.01.013.

38. Schmitt I, Lumbsch HT. 2009. Ancient horizontal gene transfer frombacteria enhances biosynthetic capabilities of fungi. PLoS One 4:e4437.http://dx.doi.org/10.1371/journal.pone.0004437.

39. Evidente A, Punzo B, Andolfi A, Cimmino A, Melck D, Luque J. 2010.Lipophilic phytotoxins produced by Neofusicoccum parvum, a grapevinecanker agent. Phytopathol Mediterr 49:74 –79.

40. Moriguchi T, Kezuka Y, Nonaka T, Ebizuka Y, Fujii I. 2010. Hiddenfunction of catalytic domain in 6-methylsalicylic acid synthase for productrelease. J Biol Chem 285:15637–15643. http://dx.doi.org/10.1074/jbc.M110.107391.

41. Ishiuchi K, Nakazawa T, Ookuma T, Sugimoto S, Sato M, TsunematsuY, Ishikawa N, Noguchi H, Hotta K, Moriya H, Watanabe K. 2012.Establishing a new methodology for genome mining and biosynthesis ofpolyketides and peptides through yeast molecular genetics. Chembi-ochem 13:846 – 854. http://dx.doi.org/10.1002/cbic.201100798.

42. Lackner G, Bohnert M, Wick J, Hoffmeister D. 2013. Assembly ofmelleolide antibiotics involves a polyketide synthase with cross-couplingactivity. Chem Biol 20:1101–1106. http://dx.doi.org/10.1016/j.chembiol.2013.07.009.

43. Schroeckh V, Scherlach K, Nutzmann HW, Shelest E, Schmidt-Heck W,Schuemann J, Martin K, Hertweck C, Brakhage AA. 2009. Intimatebacterial-fungal interaction triggers biosynthesis of archetypal polyketidesin Aspergillus nidulans. Proc Natl Acad Sci U S A 106:14558 –14563. http://dx.doi.org/10.1073/pnas.0901870106.

44. Gaisser S, Trefzer A, Stockert S, Kirschning A, Bechthold A. 1997.

Cloning of an avilamycin biosynthetic gene cluster from Streptomyces viri-dochromogenes Tu57. J Bacteriol 179:6271– 6278.

45. Ahlert J, Shepard E, Lomovskaya N, Zazopoulos E, Staffa A, BachmannBO, Huang K, Fonstein L, Czisny A, Whitwam RE, Farnet CM, Thor-son JS. 2002. The calicheamicin gene cluster and its iterative type Ienediyne PKS. Science 297:1173–1176. http://dx.doi.org/10.1126/science.1072105.

46. Ding W, Lei C, He Q, Zhang Q, Bi Y, Liu W. 2010. Insights into bacterial6-methylsalicylic acid synthase and its engineering to orsellinic acid syn-thase for spirotetronate generation. Chem Biol 17:495–503. http://dx.doi.org/10.1016/j.chembiol.2010.04.009.

47. Nishikawa H. 1933. Biochemistry of filamentous fungi. II. A metabolicproduct of Aspergillus melleus Yukawa. Part I. J Agric Chem Soc Jpn 9:107–109.

48. Blair J, Newbold G. 1955. Lactones. Part II. The structure of mellein. JChem Soc. 2871–2875.

49. Arakawa H. 1968. Absolute configuration of mellein. Bull Chem Soc Jpn41:2541. http://dx.doi.org/10.1246/bcsj.41.2541.

50. Camarda L, Merlini L, Nasini G. 1976. Metabolites of Cercospora. Tai-wapyrone, an alpha-pyrone of unusual structure from Cercospora taiwan-ensis Phytochemistry 15:537–539.

51. Pongcharoen W, Rukachaisirikul V, Phongpaichit S, Sakayaroj J. 2007.A new dihydrobenzofuran derivative from the endophytic fungus Botryo-sphaeria mamane PSU-M76. Chem Pharm Bull (Tokyo) 55:1404 –1405.http://dx.doi.org/10.1248/cpb.55.1404.

52. Rukachaisirikul V, Arunpanichlert J, Sukpondma Y, Phongpaichit S,Sakayaroj J. 2009. Metabolites from the endophytic fungi Botryosphaeriarhodina PSU-M35 and PSU-M114. Tetrahedron 65:10590 –10595. http://dx.doi.org/10.1016/j.tet.2009.10.084.

53. Venkatasubbaiah P, Sutton TB, Chilton WS. 1991. Effect of phytotoxinsproduced by Botryosphaeria obtusa, the cause of black rot of apple fruit andfrogeye leaf-spot. Phytopathology 81:243–247. http://dx.doi.org/10.1094/Phyto-81-243.

54. Parisi A, Piattelli M, Tringali C, Di San Lio GM. 1993. Identification of thephytotoxin mellein in culture fluids of Phoma tracheiphila. Phytochemistry32:865–867. http://dx.doi.org/10.1016/0031-9422(93)85221-C.

55. Edwards RL, Maitland DJ, Oliver CL, Pacey MS, Shields L, WhalleyAJS. 1999. Metabolites of the higher fungi. Part 31. Longianone, aC7H6O4 spiro bicyclic lactone from the fungus Xylaria longiana (Rehm).J Chem Soc Perkin Trans 1:715–719.

56. Holler U, Konig GM, Wright AD. 1999. Three new metabolites frommarine-derived fungi of the genera Coniothyrium and Microsphaeropsis. JNat Prod 62:114 –118. http://dx.doi.org/10.1021/np980341e.

57. Cabras A, Mannoni MA, Serra S, Andolfi A, Fiore M, Evidente A. 2006.Occurrence, isolation and biological activity of phytotoxic metabolitesproduced in vitro by Sphaeropsis sapinea, pathogenic fungus of Pinus ra-diata. Eur J Plant Pathol 115:187–193. http://dx.doi.org/10.1007/s10658-006-9006-7.

58. Zhao JH, Zhang YL, Wang LW, Wang JY, Zhang CL. 2012. Bioactivesecondary metabolites from Nigrospora sp. LLGLM003, an endophyticfungus of the medicinal plant Moringa oleifera Lam. World J MicrobiolBiotechnol 28:2107–2112. http://dx.doi.org/10.1007/s11274-012-1015-4.

59. Ramos HP, Simao MR, de Souza JM, Magalhaes LG, Rodrigues V,Ambrosio SR, Said S. 2013. Evaluation of dihydroisocoumarins pro-duced by the endophytic fungus Arthrinium state of Apiospora montagneiagainst Schistosoma mansoni. Nat Prod Res 27:2240 –2243. http://dx.doi.org/10.1080/14786419.2013.811659.

60. Krohn K, Bahramsari R, Florke U, Ludewig K, Kliche-Spory C, MichelA, Aust HJ, Draeger S, Schulz B, Antus S. 1997. Dihydroisocoumarinsfrom fungi: isolation, structure elucidation, circular dichroism and bio-logical activity. Phytochemistry 45:313–320. http://dx.doi.org/10.1016/S0031-9422(96)00854-0.

61. Bestmann HJ, Kern F, Schafer D, Witschel MC. 1992. Pheromones. 86.3,4-Dihydroisocoumarins, a new class of ant trail pheromones. AngewChem Int Ed Engl 31:795–796. http://dx.doi.org/10.1002/anie.199207951.

62. Herzner G, Schlecht A, Dollhofer V, Parzefall C, Harrar K, Kreuzer A,Pilsl L, Ruther J. 2013. Larvae of the parasitoid wasp Ampulex compressasanitize their host, the American cockroach, with a blend of antimicrobi-als. Proc Natl Acad Sci U S A 110:1369 –1374. http://dx.doi.org/10.1073/pnas.1213384110.

63. Kern F, Klein RW, Janssen E, Bestmann HJ, Attygalle AB, Schafer D,Maschwitz U. 1997. Mellein, a trail pheromone component of the ant

Mellein Synthase Gene from Parastagonospora nodorum

January 2015 Volume 81 Number 1 aem.asm.org 185Applied and Environmental Microbiology

on Decem

ber 18, 2014 by Australian N

ational Univ.

http://aem.asm

.org/D

ownloaded from

Lasius fuliginosus. J Chem Ecol 23:779 –792. http://dx.doi.org/10.1023/B:JOEC.0000006410.35938.49.

64. Dai J, Carte BK, Sidebottom PJ, Sek Yew AL, Ng S, Huang Y, Butler MS.2001. Circumdatin G, a new alkaloid from the fungus Aspergillus ochra-ceus. J Nat Prod 64:125–126. http://dx.doi.org/10.1021/np000381u.

65. Djoukeng JD, Polli S, Larignon P, Abou-Mansour E. 2009. Identifica-tion of phytotoxins from Botryosphaeria obtusa, a pathogen of black deadarm disease of grapevine. Eur J Plant Pathol 124:303–308. http://dx.doi.org/10.1007/s10658-008-9419-6.

66. Ramirez-Suero M, Benard-Gellon M, Chong J, Laloue H, StempienE, Abou-Mansour E, Fontaine F, Larignon P, Mazet-Kieffer F, Farine

S, Bertsch C. 2014. Extracellular compounds produced by fungi asso-ciated with Botryosphaeria dieback induce differential defence gene ex-pression patterns and necrosis in Vitis vinifera cv. Chardonnay cells. Pro-toplasma 251:1417–1426. http://dx.doi.org/10.1007/s00709-014-0643-y.

67. Keller B, Winzeler H, Winzeler M, Fried PM. 1994. Differential sensi-tivity of wheat embryos against extracts containing toxins of Septorianodorum—first steps towards in-vitro selection. J Phytopathol 141:233–240. http://dx.doi.org/10.1111/j.1439-0434.1994.tb01466.x.

68. Essad S, Bousquet JF. 1981. Action de l’ochracine, phytotoxine de Sep-toria nodorum Berk., sur le cycle mitotique de Triticum aestivum L.Agronomie 1:684 – 694.

Chooi et al.

186 aem.asm.org January 2015 Volume 81 Number 1Applied and Environmental Microbiology

on Decem

ber 18, 2014 by Australian N

ational Univ.

http://aem.asm

.org/D

ownloaded from