f u n g a l b i o l o g y x x x ( 2 0 1 2 ) 1 e8 journal homepage: www.elsevier.com/locate/funbio Patulin and secondary metabolite production by marine-derived Penicillium strains Marieke VANSTEELANDTa, Isabelle KERZAONa,1, Elodie BLANCHETa, Olivia FOSSI TANKOUAa, Thibaut ROBIOU DU PONTa, Yolaine JOUBERTa, Fabrice MONTEAUb, Bruno LE BIZECb, Jens C. FRISVADc, Yves François POUCHUSa,*, Olivier GROVELa a LUNAM universit
e, Universit
e de Nantes, MMS, Nantes, France ONIRIS, L’UNAM, Atlanp^ole la Chantrerie, 44307 Nantes Cedex 3, France c Center for Microbial Biotechnology, Department of Systems Biology, Technical University of Denmark, DK-2800 Kgs. Lyngby, Denmark b article info abstract Article history: Genus Penicillium represents an important fungal group regarding to its mycotoxin produc- Received 13 October 2011 tion. Secondary metabolomes of eight marine-derived strains belonging to subgenera Received in revised form Furcatum and Penicillium were investigated using dereplication by liquid chromatography 20 June 2012 (LC)eDiode Array Detector (DAD)emass spectrometry (MS)/MS. Each strain was grown on Accepted 22 June 2012 six different culture media to enhance the number of observable metabolites. Corresponding Editor: Thirty-two secondary metabolites were detected in crude extracts with twenty first obser- Stephen W. Peterson vations for studied species. Patulin, a major mycotoxin, was classically detected in extracts Keywords: secondary metabolome is still to be done. These detections constituted the first descrip- LCeDADeMS/MS dereplication tions of patulin in marine strains of Penicillium, highlighting the risk for shellfish and their Marine-derived Penicillium consumers due to the presence of these fungi in shellfish farming areas. Neuroactivity Patulin induced acute neurotoxicity on Diptera larvae, indicating the interest of this bioas- Patulin say as an additional tool for detection of this major mycotoxin in crude extracts. of Penicillium expansum, and was also isolated from Penicillium antarcticum cultures, whose Secondary metabolites ª 2012 The British Mycological Society. Published by Elsevier Ltd. All rights reserved. Subgenus Furcatum Introduction Filamentous fungi are present in cultured shellfish and their immediate environment (seawater and sediments) and a clear predominance of genus Penicillium was observed on the French western coast (Sallenave-Namont et al. 2000). Penicillia are known to produce a wide range of mycotoxins such as patulin (e.g. Penicillium expansum), ochratoxin A (e.g. Penicillium verrucosum), penitrem A (e.g. Penicillium crustosum), or roquefortine C (e.g. Penicillium roqueforti) (Frisvad et al. 2004; Paterson et al. 2004). Their presence in shellfish farming areas could then constitute a risk for shellfish and their consumers. They may particularly be a possible origin for unexplained toxicities and mortalities of marine cultured shellfish (cockles, blue mussels, and oysters) observed along the French coast in recent years (Marcaillou-LeBaut & Amzil 1995). Ability for fungal toxic metabolites to contaminate shellfish has been demonstrated in vitro with gliotoxin produced by a marine-derived strain of Aspergillus * Corresponding author. Tel.: þ33 253484191; fax: þ33 251125679. E-mail addresses: marieke.vansteelandt@univ-nantes.fr, jcf@bio.dtu.dk, yves-francois.pouchus@univ-nantes.fr 1 Present address: University of Lyon 1, UMR 5557 CNRS-UCBL, Ecologie Microbienne, CESN, ISPB, F69373 Lyon cedex, France. 1878-6146/$ e see front matter ª 2012 The British Mycological Society. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.funbio.2012.06.005 Please cite this article in press as: Vansteelandt M, et al., Patulin and secondary metabolite production by marine-derived Penicillium strains, Fungal Biology (2012), http://dx.doi.org/10.1016/j.funbio.2012.06.005 2 M. Vansteelandt et al. fumigatus (Grovel et al. 2003), and in situ with peptaibols, neurotoxic linear peptides produced by Trichoderma sp. found in mus
(Poirier et al. 2007). sels and sediments in Ars-en-Re Numerous works concerning mycotoxin production by terrestrial Penicillium spp. have been published. On the contrary there are few data on the presence of such toxic compounds in marine environment and marine food. In fact, published works only concern exogenous Penicillium mycotoxins brought to fish, for instance in vegetable ingredients used in farming feed (El-Sayed et al. 2009), but none described their direct production in situ. Genus Penicillium is divided into four subgenera: Penicillium, Furcatum, Biverticillium, and Aspergilloides. Among them, subgenus Penicillium has been much more studied than the three others. Subgenus Furcatum, which has been often observed in the marine environment (Amagata et al. 1998; Bugni et al. 2003; Tsuda et al. 2004; Smetanina et al. 2007), represents then a promising source of investigation for metabolite production. For that purpose, dereplication methods are useful, allowing a rapid identification of compounds in crude extracts without the need of a previous purification (Nielsen & Smedsgaard 2003; Lang et al. 2008). HPLC separations coupled with spectrometric analysis (UV, MS, MS/MS, high resolution mass spectrometry (HRMS)) and comparison of acquired data with published databases allow identification of metabolites with a high degree of certainty. The aim of the present work was to conduct a risk assessment associated to marine environmental toxigenic fungi exposure both for shellfish and their consumers by an identification of known mycotoxins produced by marine-derived Penicillium spp.. For that purpose, marine-derived strains of genus Penicillium, especially of the subgenus Furcatum, isolated from shellfish farming areas, were selected in the fungal collection of the laboratory. They were chosen for their novelty or their taxonomical interest among toxigenic Penicillium strains after a screening of bioactivity (internal unpublished data). A peculiar attention was held to production of major mycotoxins such as patulin which could constitute a high risk for shellfish consumption and whose presence is regulated in food. Materials and methods Fungal strains All studied fungal strains were previously isolated from marine natural samples of shellfish, sediments or seawater collected on the West Atlantic coast of France and maintained in the laboratory fungal collection (MMS e Marine Fungal Collection, Nantes University). They were identified using a phenotypic approach (morphology, comparison to ex-type cultures, physiology and secondary metabolite profiles) and by sequencing the internal transcribed spacers (ITS-1 and -2) and beta-tubulin regions. Selected Furcatum species were Penicillium antarcticum, a recently described marine species (McRae et al. 1999) for which very few chemical studies have already been published (Shiono et al. 2008) and a new marine-derived species, still undescribed, temporarily called Penicillium SP MMS351 (GenBank accession number JN676192 for ITS sequence and JN794530 for beta-tubulin sequence), belonging to the section Divaricatum series Canescentia. As a reference, Penicillium canescens, a well-known metabolite producer (Brian et al. 1953; Keromnes & Touvenot 1985; DiMenna et al. 1986; Kozlovskii et al. 1997; Nicoletti et al. 2007) belonging to the same section/series was also studied. In order to compare subgenera Furcatum and Penicillium, marine-derived strains belonging to toxigenic species of this last subgenus, Penicillium chrysogenum and Penicillium expansum, were studied in the same operating conditions. Table 1 gives taxonomic positions and origins of the eight strains selected. Culture and extraction OSMAC approach (‘One Strain Many Compounds’) (Bode et al. 2002), consisting in cultures of a strain on different media in order to enhance the number of observable metabolites, was used. Each strain was grown on six culture media, chosen for their different compositions in mineral, carbon, and nitrogen sources, all prepared with 15 g L 1 of agar in sterilized natural seawater (salinity of 32.8 g L 1) and containing: 5 g of Czapek extract, 5 g of yeast extract, and 30 g of saccharose for Czapek Yeast extract Agar medium (CYA); 40 g of glucose and 10 g of casein enzymatic digest for Dextrose Casein Agar medium (DCA); 2.4 g of MgSO4, 2.4 g of NH4NO3, 1.21 g of Tris buffer, and 5 g of glucose for Kohlmeyer-Medium Solid (KMS); 5 mg of CuSO4, 10 mg of ZnSO4, 1 g of peptone, 20 g of malt extract, and 20 g of glucose for Malt Extract Agar medium (MEA); 5 mg of CuSO4, 10 mg of ZnSO4, 4 g of potato extract, and 20 g of glucose for Potato Dextrose Agar medium (PDA); 5 mg of CuSO4, 10 mg of ZnSO4, 0.5 g of MgSO4, 20 g of yeast Table 1 e Investigated Penicillium strains (classification according to Frisvad et al. (2004) for the strains of subgenus Penicillium). Taxonomic position Genus Subgenus Penicillium Furcatum Penicillium Section Series Strain Species Authority Reference Origin Place Hocking & McRae MMS14 MMS15 MMS351 MMS747 MMS194 MMS460 Shellfish Shellfish Seawater Sediments Seawater Sediments Le Croisic Le Croisic
e La Pre La Couplasse La Baule Le Croisic MMS5 MMS42 Shellfish Sediments Le Croisic Le Croisic Furcatum Citrina antarcticum Divaricatum Canescentia SP MMS351 Chrysogena Penicillium Chrysogena Expansa canescens Sopp chrysogenum expansum Thom Link Please cite this article in press as: Vansteelandt M, et al., Patulin and secondary metabolite production by marine-derived Penicillium strains, Fungal Biology (2012), http://dx.doi.org/10.1016/j.funbio.2012.06.005 Patulin and secondary metabolite production by marine-derived Penicillium strains extract, and 150 g of saccharose for Yeast Extract Sucrose medium (YES). Cultures were realized in Erlenmeyers containing 50 mL of solid medium and incubated at 27  C for 11e13 d depending on strains. After incubation culture substrates and mycelia were ground together and extracted twice with 100 mL of ethylacetate, pure or mixed with methylene-chloride (1:1 v/v) to enhance extraction yield. Ground mixtures were ultrasonicated for 30 min and then shaked for 1 h. After dehydration on Na2SO4 and filtration under vacuum on regenerated cellulose € ttingen, Germany), organic membrane (0.45 mm, Sartorius, Go phases were evaporated to dryness leading to crude extracts. LCeUV/DAD (liquid chromatographyeUV/Diode Array Detector) and LCeESIeITeMS/MS (liquid chromatographye electrospray ionizationeion trapemass spectrometry/MS) analyses Separations were performed on a RP-C18 Yperspher BDS column (2  125 mm, 3 mm particle size) fitted with a BC18 Yperspher BDS pre-column (2  10 mm, 3 mm, Interchim, Montluçon, France). Crude extracts were dissolved in methanol (10 mg mL 1) and filtered through a 0.45 mm regenerated cellulose filter (Minisart, Sartorius, Hannover, Germany). Injection volume was 5 mL and the separation was performed at a flow rate of 0.3 mL min 1 with an acidified watereacetonitrile gradient (0.005 % of trifluoroacetic acid in water) adapted from Nielsen & Smedsgaard (2003). After 5 min of equilibration, a linear gradient started at 85 % acidified water and 15 % acetonitrile, changing to 100 % acetonitrile in 40 min, and maintaining the isocratic stage for 5 min. Equilibrium delay before next injection has been fixed at 8 min. In order to normalize retention time and to calculate retention index (RI), 1 mL of a mixture of seven alkylphenones (in MeOH at the concentration of 0.35 M each) was injected according to the protocol published by Frisvad & Thrane (1987). LCeDAD/MS/MS analyses were performed on a HPLC Spectraphysics Spectra System P2000 equipped with an automatic injector AS 100XR (Thermo Separation Products, San Jose, CA, USA) coupled to a UVevisible DAD (Agilent 1200 series) and a LCQ (Finnigan Thermo Separation Products, San Jose, CA, USA) IT mass spectrometer. The mass spectrometer was operated in ESI positive mode with a source voltage of 4.5 kV, sheath gas (N2) flow rate of 75 AU, auxiliary gas (N2) flow rate of 25 AU, capillary temperature of 200  C, and capillary voltage of 19 V. For LCeMS/MS analysis, data-dependant scan mode was used, and led to the selection of the most abundant ion in each scan which was subjected to MS/MS analysis. MS/MS spectra acquisitions were carried out with collision energy of 35 % and an isolation width of the precursor ion of 2 u. MS data were collected and analysed using LCQ Xcalibur 1.3 software (Thermo Fisher Scientific). For each chromatographic peak, the corresponding RI was calculated according to Frisvad & Thrane (1987). UV- and mass-spectra data were recorded. All data collected were compared with those of the fungal metabolite database of Nielsen & Smedsgaard (2003) or from specific bibliography. 3 HRMS (high resolution mass spectrometry) analyses HRMS analyses were performed on a linear IT coupled to an orbitrap mass spectrometer (LTQ/Orbitrap Thermo Fisher Scientific, Bremen, Germany) allowing the determination of molecule accurate mass and molecular formula. Crude extract was analysed by LCeHRMS/MS in the same operating conditions as previously, in LCeMS/MS. Methanolic solution of pure compound (10 mg mL 1) was analysed by direct infusion. Isolation and identification of patulin Isolation of patulin was performed by bioguided fractionation, using its neurotoxic properties. MEA extract of MMS14 (100 mg) was first separated by vacuum LC on a silica gel column (60  A, 35e70 mm, SDS, Val de Reuil, France) with a CH2Cl2/MeOH non-linear gradient as mobile phase (from 100:0 to 0:100 (v/v)). Twelve fractions were collected and their neuroactivity was evaluated on Diptera larvae. The most active fraction (eluted by CH2Cl2/MeOH 97:3 (v/v)) was then fractionated by normal-phase solid-phase extraction (SPE) (silica 60  A, 35e70 mm, SDS, Val de Reuil, France), with a non-linear gradient of hexane/acetone (100:0 to 0:100 (v/v)) and pure MeOH, leading to 14 subfractions. Most active fraction (eluted by hexane/acetone 70:30 (v/v)) was purified by HPLC on a Modulo-cart QS Nucleosil NS-25QS column (4.6  250 mm, 5 mm particle size Interchrom (Interchim, Montluçon, France)) eluted by CH2Cl2/EtOAc 95:5 (v/v), at 0.5 mL min 1 flow rate, leading to a pure active compound which was identified by HRMS and NMR analysis. NMR (nuclear magnetic resonance) analyses 1 H and 13C NMR spectra were recorded in CDCl3 on a Bruker 500 MHz equipped with a TXI cryoprobe. Chemicals Organic solvents used in extractions and purification were purchased from Carlo Erba SDS (Val de Reuil, France). LCeMS analyses were performed using HPLC-grade methanol and acetonitrile (Baker, Deventer, The Netherlands). Water was purified to HPLC-grade quality with a Millipore-QRG ultrapure water system (Millipore, Milford, CT, USA). Trifluoroacetic acid was obtained from SigmaeAldrich (Saint-Quentin Fallavier, France). Alkylphenones were purchased from SigmaeAldrich. Neurotoxicity assessment on Diptera larvae All assays were performed according to Ruiz et al. (2010), with larvae of the fly Calliphora vomitoria L. purchased from fishing stores. Crude extracts were dissolved at the concentration of 100 mg mL 1 in an aqueous solution containing 15 % of dimethyl sulfoxide (DMSO). Injections were performed at the rate of 0.1 mL mg 1 of larvae body weight using a 10 mL microsyringe with a thin hypodermic needle (Hamilton, Bonaduz, Switzerland). Solutions were injected in the last abdominal segment of larvae in order to determine the neuroactivity Please cite this article in press as: Vansteelandt M, et al., Patulin and secondary metabolite production by marine-derived Penicillium strains, Fungal Biology (2012), http://dx.doi.org/10.1016/j.funbio.2012.06.005 4 M. Vansteelandt et al. certain identification. In the same way, patulin was identified on the basis of RI and UV-spectrum. It is a well-known terrestrial mycotoxin produced by different species of fungi, especially Penicillium spp.. A publication reported its production by a marine-derived Aspergillus varians (Smetanina et al. 2005) but it has never been observed yet in marine-derived Penicillium isolates. As this compound is a representative metabolite of Penicillium expansum, obtained data (RI and UV) can be considered as sufficient for its identification in this species. On the other hand, identification based only on these criteria is more hazardous for a first observation as proposed in Penicillium antarcticum strains. Additional experiments were needed for a complete identification for such a major mycotoxin (see Confirmation of patulin production in Penicillium antarcticum). Several compounds were not observed in species in which they have already been described in literature. It is the case of griseofulvin, penicillic acid, and citrinin which were not detected in extracts of P. expansum, Penicillium chrysogenum, and Penicillium canescens respectively (Leistner & Pitt 1977; Keromnes & Touvenot 1985; Paterson & Kemmelmeier 1989). As these compounds were visualized in other strains, their non-observation was not due to a failing of detection but probably to a lack of production by these marine-derived strains in experimental conditions used. A second hypothesis could be a misidentification of species studied in previous published works, as it has been noticed by Frisvad et al. (2004). On the contrary, it was the first observations of verruculotoxin and maculosin in P. expansum and/or P. chrysogenum indicating metabolic differences between terrestrial and marinederived fungal strains. There have been very limited investigations on the metabolome of P. antarcticum with only two secondary metabolites index (NI) which is calculated as follow: NI ¼ 500/T where T (expressed in seconds) is the time of appearance of symptoms (total or partial immobilization) after injection. If no symptoms occurred after 10 min, the test was considered as negative. To determine its median effective dose (ED50) on this bioassay, solutions of pure patulin solubilized in an aqueous solution containing 15 % of DMSO at various concentrations were prepared and injected in the larvae as previously reported, in the range of 0.1e8 mg mg 1 of larvae. Results and discussion Dereplication of crude extracts Analysis of the 48 crude extracts obtained led to the identification of 32 compounds or group of compounds. All compounds were identified using their RI and spectral data. For a majority of them, identification was performed using four criteria (RI, UV, MS, and MS/MS). For others, a lack of one or more criteria, especially comparative MS/MS data in the literature, led to putative identifications which must be confirmed by other physicochemical means. Detected metabolites are listed in Table 2 with their number of appearances among the six crude extracts prepared for each strain. There were notable similarities between strains of same species and differences between species and subgenera. Only five compounds were found to be produced by different strains of the two subgenera: maculosin, aurantioclavine, chrysogine, patulin, and terrestric acid. Identification of this last compound was only based on its observation in low resolution MS without MS/MS data which is not sufficient for Table 2 e Number of observations of identified metabolites among the six crude extracts prepared for each strain. Identified compounds De or epidechlorogriseofulvin Griseofulvin Penicillic acid Meleagrin Roquefortine D Roquefortine C Cladosporin 5'-hydroxyasperentin Nortryptoquivaline Citrinin Communesin A Communesin B Chaetoglobosins (MW=528) Chaetoglobosins (MW=530) Antarone A Violaceic acid Festuclavine Agroclavine Expansolide A and/or B Penitremone A Penitremone C Penitremone B Verruculotoxin Glandicoline A Glandicoline B Terrestric acid Patulin Orsellinic acid Penicillium Oxalin P. antarcticum P. antarcticum P. SP MMS351 P. SP MMS351 P. canescens P. canescens P. chrysogenum P. expansum Maculosin Species Furcatum Aurantioclavine Subgenus Retention index UV spectrum MS MS/MS Strain MMS14 MMS15 MMS351 MMS747 MMS194 MMS460 MMS5 MMS42 Chrysogine Identification criteria observed Possible identification X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X - X X - X X - X X - X X - X X - X X - X X - X X - 1 3 1 - 1 5 6 2 1 - 2 1 - 1 4 5 5 - 4 5 6 6 0 4 3 4 1 0 0 - 6 3 0 2 5 3 6 6 - 1 2 - 5 6 - 0 0 4 5 6 5 4 4 5 - 2 - 5 5 - 5 5 - 1 5 6 - 3 5 - 3 5 - 3 1 3 - 3 - 4 4 3 5 2 - 2 2 4 2 3 4 1 - - Compound name: indole alkaloid. X: criteria used for identification. -: Criteria not observed or metabolite not detected. 0: Metabolite not detected although cited in literature. Grey frame: 1st observation in the species. Please cite this article in press as: Vansteelandt M, et al., Patulin and secondary metabolite production by marine-derived Penicillium strains, Fungal Biology (2012), http://dx.doi.org/10.1016/j.funbio.2012.06.005 Patulin and secondary metabolite production by marine-derived Penicillium strains described, the polyketides antarones A and B, produced by an endophytic strain isolated from the stem of Cedrus deodara (Shiono et al. 2008). In extracts of the marine-derived strains of P. antarcticum, a chromatographic peak (RI between 957 and 998) could correspond to antarone A with a UV e lmax at 284 nm (as observed by Shiono et al. 2008) and concordant results obtained in HRMS with observation of [M þ H]þ at m/z theoretical mass ¼ 431.24335, 431.24191 (C25H35O6, Dppm ¼ 3.34). In MS/MS analysis, main fragments were observed at m/z 422.3, m/z 413.2 (loss of 18 u, [M þ H H2O]þ), m/z 395.4 (loss of 36 u, [M þ H (H2O)2]þ), m/z 387.2 (loss of 44 u, [M þ H CO2]þ), m/z 373.1, m/z 318.7, m/z 298.3, m/z 267.1, m/z 167.1. Antarone B was not observed. Six other secondary metabolites were identified from the cultures of the two strains. Among them, cladosporin (¼asperentin) was observed in all P. antarcticum culture extracts, and could then be considered as a consistent metabolite for this species. It was originally discovered in Cladosporium cladosporioides (Scott et al. 1971) and then observed in various Ascomycota genera, Microascus (Fujimoto et al. 1999), Oidiodendron (John et al. 1999) and Eurotium (Slack et al. 2009). In the genus Penicillium, it has been mentioned in Penicillium brevicompactum (subgenus Penicillium) isolated as airborne fungi in biowaste (Fischer et al. 2000) and in Penicillium soppii (subgenus Furcatum) from different terrestrial origins (Christensen et al. 1999). Present observation is the first one in a marine-derived fungus. Two other polyketides derivatives, 50 -hydroxyasperentin (a derivative of cladosporin) and violaceic acid, and two alkaloids, chrysogine and aurantioclavine, were also observed but as occasional metabolites. The four strains of the two other species of subgenus Furcatum series Canescentia (P. canescens and P. SP MMS351) all produced a group of four polyketides: griseofulvin, de- or epidechlorogriseofulvin, penicillic acid, and orsellinic acid. This last metabolite was hidden in MS analysis by penicillic acid but has a characteristic UV-spectrum allowing its identification. Although in this study these four compounds were only observed in this fungal group, they are widely spread in the fungal world. Since its first description from Penicillium griseofulvum by Oxford et al. (1939), griseofulvin has often been observed in Penicillium sp. of the subgenera Penicillium (Frisvad et al. 2004) and Furcatum (Petit et al. 2004). Nevertheless, besides this genus, it has been observed in only one other genus, Nigrospora sp. (in Nigrospora oryzae and its teleomorph, Khuskia oryzae) (Furuya et al. 1967; Hemming et al. 1969). Penicillic acid and its biosynthetic precursor orsellinic acid are often associated in Penicillium spp. for instance in Penicillium fennelliae (Van Eijk 1969). Orsellinic acid is not specific of the genus Penicillium, it can also be found as a major constituent of depsides in lichens (Narui et al. 1998). The main difference between the two studied species of ser. Canescentia was their indole alkaloid production. Both strains produced these derivatives of tryptophan but belonging to different chemical families. The new species P. SP MMS351 produced clavines (agroclavine and festuclavine) and nortryptoquivaline whereas P. canescens strains produced a group of three metabolites which could correspond to penitremones derivatives. Agroclavine has already been noticed alone in a strain of Penicillium sizovae, belonging to the 5 Table 3 e 1H and 13C NMR of the purified toxic metabolite. d (ppm) in methanol-d4 C 13 C NMR 1 2 3 4 5a 171.6 111.8 148.7 90.7 61.1 5b 61.1 6 7 110.5 154.3 1 H NMR e 6.06 (s) e 5.98 (s) O 4.67 (dd, 2J5a5b 17 Hz, 3 J5a6 2.5 Hz) 4.4 (dd, 2J5a5b 17 Hz, 5 3 J5b6 3 Hz) 6,04 (m) e OH 4 3 2 1 7 O O 6 subgenus Furcatum (Kozlovskii et al. 1985). Associated with its derivative festuclavine it has been observed in different species of the subgenus Penicillium: P. roqueforti, P. carneum (Nielsen et al. 2006a, 2006b), and P. commune (Vinokurova et al. 2003). Nortryptoquivaline, which was detected in 11 out of the 12 extracts of P. SP MMS351, has only been € chi et al. observed before in the genus Aspergillus (Bu 1977; Fischer et al. 2000). It could then be considered as a chemical marker of this new species in the genus Penicillium. Penitremones are complex polycyclic molecules close to penitrem toxins, already described in an unidentified Penicillium (Naik et al. 1995). In this chemical family, only penitrem A had already been observed in P. canescens (DiMenna et al. 1986). More generally, indole-alkaloids seemed to be widespread in the investigated strains, with 15 out the 32 compounds identified belonging to this chemical family. In P. antarcticum extracts, only one observation of aurantioclavine was done. This lack of tryptophan derivatives could then be considered as a specificity of this species. However, all compounds detected in the extracts of P. antarcticum have not been identified and the complete investigation of the metabolome of this species is still to be done. Aside from previous cited metabolites, various other indole-alkaloids already known in P. expansum and P. chrysogenum were observed: communesins, chaetoglobosins, oxalin, meleagrin, and roquefortines (Frisvad et al. 2004; Kerzaon et al. 2009). This indicates that, Fig 1 e Acute toxicity of patulin on Diptera larvae. Please cite this article in press as: Vansteelandt M, et al., Patulin and secondary metabolite production by marine-derived Penicillium strains, Fungal Biology (2012), http://dx.doi.org/10.1016/j.funbio.2012.06.005 6 M. Vansteelandt et al. Table 4 e Neuroactivity of P. antarcticum and P. expansum extracts and presence of patulin detected by UV. Species Strain Culture medium CYA DCA KMS MEA Patulin e UV NI* Patulin e UV NI Patulin e UV NI PDA YES Patulin e UV NI Patulin e UV NI Patulin e UV NI P. antarcticum MMS14 MMS15 e e e e e e e e e e e e xx xx 15 16 xx xx 19 8 e e e e P. expansum MMS42 e 4 xx 29 e 4 xx 5 xx 18 xx 8 NI*: Neuroactivity Index xx: Patulin detected. e: Patulin not detected. independently of their origin and their growth conditions, there are similarities between terrestrial and marine-derived strains. This implies that the presence of known mycotoxinproducing species in shellfish farming areas could constitute a real risk for shellfish and their consumers. Fifty-three percent of compounds detected were found as constant for a strain as they were produced on at least five out of the six media used. This indicates that these metabolites are produced by the considered strain, whatever environmental conditions are. On the contrary, 15 % could be considered as sporadic (identified in less than three extracts for a same strain). In this case, their production is dependent on the environment, and the risk, if they are toxic, is lowered because ad hoc productive conditions must be realized. Confirmation of patulin production in Penicillium antarcticum In order to confirm the presence of patulin in P. antarcticum, MEA-MMS14 extract was fractionated leading to a pure compound which was analysed by HRMS. In the positive ion mode, the base peak was observed at m/z 177.01588 corresponding to a sodium adduct [M þ Na]þ with the formula C7H6O4Na (theoretical mass 177.01638, Dppm ¼ 2.82). In the negative ion mode it appeared at m/z 153.01906 [M H] (C7H5O4, theoretical mass 153.01878, Dppm ¼ 1.83). These formulae may correspond to patulin. Observed compound also appeared in positive mode as a methanolesodium adduct [M þ Na þ (CH3OH)2]þ at m/z 241.06825, and in negative mode as a methanol adduct ([M H þ (CH3OH)2] ) at m/z 217.07124. 1 H and 13C NMR analyses (Table 3) allowed a final identification of the compound as patulin, confirming the presence of this mycotoxin in studied extract. Pure patulin was also analysed by LCeMS/MS in positive ionization mode. Even at a high concentration (1 mg mL 1), only a low abundant chromatographic peak could be observed, revealing a poor ionization efficiency in the MS source conditions used. It explains as well its non-detection by LCeMS/MS in studied crude fungal extracts. For that kind of metabolites, other chromatographic and MS conditions have to be used as observed by Nielsen et al. (2006a, 2006b). Diptera larvae toxicity: a complementary bioassay for patulin detection Patulin is known to present several biological activities and to induce toxicity in vitro and in vivo. It is considered as mutagenic, genotoxic, immunotoxic, neurotoxic to rodents, and teratogenic to chicken (Andersen et al. 2004). This mycotoxin also induces cytotoxicity on different cell lines (Iwamoto et al. 1999; Heussner et al. 2006). This justifies its regulation by FAO/JEFCA (2004). In this work, patulin was purified by bioguided fractionation of MEA-MMS14 extract, using neuroactivity on Diptera larvae as bioassay. To determine its ED50 in this model, a series of decreasing doses of commercial patulin were injected to the larvae (Fig 1). All Penicillium antarcticum and Penicillium expansum extracts were tested with the same bioassay, as these species produced patulin (Table 4). It clearly appeared that when patulin was detected by UV, neuroactivity was observed. However, other mycotoxins and bioactive metabolites can also induce a neuroactivity, such as communesins and particularly communesin B (Kerzaon et al. 2009) which was observed in all P. expansum extracts. This bioassay seems to be less sensitive than acute toxicity in mice with ED50 ¼ 3.9 mg mg 1 whereas median lethal dose (LD50) in mice ¼ 10 mg kg 1 subcutaneous, 48 mg kg 1 per os and 7.5 mg kg 1 intraperitoneal (McKinley & Carlton 1980). Nevertheless, it presents an interest as a simple complementary tool to LCeUV analysis for patulin detection. Conclusion Dereplication of crude fungal extracts is a powerful tool for rapid detection thanks to direct identification of known metabolites. It was particularly effective in this study where a total of 245 observations concerning 32 compounds were completed in the 48 extracts prepared, with a first observation for 20 of them in the studied species. Nevertheless, dereplication procedures have some limits such as difficulty to identify important metabolites such as patulin, even with published data, because of their poor observability in operating conditions. Nevertheless, concomitant use of a bioassay, indicating the presence of a toxic compound, appeared to be an efficient Please cite this article in press as: Vansteelandt M, et al., Patulin and secondary metabolite production by marine-derived Penicillium strains, Fungal Biology (2012), http://dx.doi.org/10.1016/j.funbio.2012.06.005 Patulin and secondary metabolite production by marine-derived Penicillium strains complement for detection of toxins, such as acute toxicity in Diptera larvae for patulin detection. Patulin, citrinin and roquefortine C, which have been produced by studied strains, are regulated mycotoxins, highlighting the risk of the presence of such toxigenic fungi in marine environment for shellfish and their consumers. Acknowledgement Authors want to acknowledge the French Ministry of Higher Education and Research, the region Pays de la Loire and Atlantic Bone Screen for granting Ph.D. students, and the region Pays de la Loire for the research program ChimiMar. references Amagata T, Minoura K, Numata A, 1998. Cytotoxic metabolites produced by a fungal strain from a Sargassum alga. Journal of Antibiotics 51: 432e434. Andersen B, Smedsgaard J, Frisvad JC, 2004. Penicillium expansum: consistent production of patulin, chaetoglobosins, and other secondary metabolites in culture and their natural occurrence in fruit products. Journal of Agricultural and Food Chemistry 52: 2421e2428. € fs R, Zeeck A, 2002. Big effects from small Bode HB, Bethe B, Ho changes: possible ways to explore nature’s chemical diversity. ChemBioChem 3: 619e627. Brian PW, Hemming HG, Moffatt JS, Unwin CH, 1953. Canescin an antibiotic produced by Penicillium canescens. Transactions of the British Mycological Society 36: 243e247. € chi G, Luk KC, Kobbe B, Townsend JM, 1977. Four new mycoBu toxins of Aspergillus clavatus related to tryptoquivaline. Journal of Organic Chemistry 42: 244e246. Bugni TS, Bernan VS, Greenstein M, Janso JE, Maiese WM, Mayne CL, Ireland CM, 2003. Brocaenols AeC: novel polyketides from a marine-derived Penicillium brocae. Journal of Organic Chemistry 68: 2014e2017. Christensen M, Frisvad JC, Tuthill D, 1999. Taxonomy of the Penicillium miczynskii group based on morphology and secondary metabolites. Mycological Research 103: 527e541. DiMenna ME, Lauren DR, Wyatt PA, 1986. Effect of culture conditions on tremorgen production by some Penicillium species. Applied and Environmental Microbiology 51: 821e824. El-Sayed YS, Khalil RH, Saad TT, 2009. Acute toxicity of ochratoxin-A in marine water-reared sea bass (Dicentrarchus labrax L.). Chemosphere 75: 878e882. FAO, 2004. Worldwide Regulation for Mycotoxins in Food and Feed in 2003. FAO Food and Nutrition Paper No. 81. Rome, 180 pp. € ller T, Schwalbe R, Ostrowski R, Dott W, 2000. SpeFischer G, Mu cies-specific profiles of mycotoxins produced in cultures and associated with conidia of airborne fungi derived from biowaste. International Journal of Hygiene and Environmental Health 203: 105e116. Frisvad JC, Thrane U, 1987. Standardized high-performance liquid chromatography of 182 mycotoxins and other fungal metabolites based on alkylphenone retention indexes and UVeVIS spectra (diode array detection). Journal of Chromatography A 404: 195e214. Frisvad JC, Smedsgaard J, Larsen TO, Samson RA, 2004. Mycotoxins, drugs and other extrolites produced by species in Penicillium subgenus Penicillium. Studies in Mycology 49: 201e242. 7 Fujimoto H, Fujimaki T, Okuyama E, Yamazaki M, 1999. Immunomodulatory constituents from an ascomycete, Microascus tardifaciens. Chemical and Pharmaceutical Bulletin 47: 1426e1432. Furuya K, Enokita R, Shirasaka M, 1967. Antibiotics from fungi. II. New griseofulvin producer, Nigrospora oryzae. Annual Report of the Sankyo Research Laboratory 19: 91e95. Grovel O, Pouchus YF, Verbist JF, 2003. Accumulation of gliotoxin, a cytotoxic mycotoxin from Aspergillus fumigatus, in blue mussel (Mytilus edulis). Toxicon 42: 297e300. Hemming HG, Lehan M, Giles D, 1969. Griseofulvin. Patent DE1813572, 8 pp. Heussner AH, Dietrich DR, O’Brien E, 2006. In vitro investigation of individual and combined cytotoxic effects of ochratoxin A and other selected mycotoxins on renal cells. Toxicology in Vitro 20: 332e341. Iwamoto C, Minoura K, Oka T, Ohta T, Hagishita S, Numata A, 1999. Absolute stereostructures of novel cytotoxic metabolites, penostatins AeE from a Penicillium species separated from an Enteromorpha alga. Tetrahedron 55: 14353e14368. € rke U, Aust HJ, Draeger S, Schulz B, 1999. John M, Krohn K, Flo Biologically active secondary metabolites from fungi. 12.1 Oidiolactones AeF, labdane diterpene derivatives isolated from Oidiodendron truncata. Journal of Natural Products 62: 1218e1221. Keromnes J, Touvenot D, 1985. Role of penicillic acid in the phytotoxicity of Penicillium cyclopium and Penicillium canescens to the germination of corn seeds. Applied and Environmental Microbiology 49: 660e663. Kerzaon I, Pouchus YF, Monteau F, Le Bizec B, Nourrisson MR, Biard JF, Grovel O, 2009. Structural investigation and elucidation of new communesins from a marine-derived Penicillium expansum Link by liquid chromatography/electrospray ionization mass spectrometry. Rapid Communications in Mass Spectrometry 23: 3928e3938. Kozlovskii AG, Vepritskaya IG, Gayazova NB, Il’chenko VY, 1985. Effect of exogenous tryptophan on biosynthesis of ergot alkaloids in Penicillium sizovae. Mikrobiologyia 54: 883e888. Kozlovskii AG, Vinokurova NG, Zhelifonova VP, Ozerskaya SM, 1997. Alkaloid formation by penicillia of the series Fellutana and Canescentia. Microbiology (Moscow) (Translation of Mikrobiologiya) 66: 429e433. Lang G, Mayhudin N, Mitova MI, Sun L, Van der Sar S, Blunt JW, Cole ALJ, Ellis G, Laatsch H, Munro MHG, 2008. Evolving trends in the dereplication of natural product extracts: new methodology for rapid, small-scale investigation of natural product extracts. Journal of Natural Products 71: 1595e1599. Leistner L, Pitt JI, 1977. Miscellaneous Penicillium toxins. In: Rodricks JV, Hesseltine CW, Mehlman MA (eds), Mycotoxins in Human and Animal Health. Pathotox Publishers, Park Forest South, USA, pp. 639e653. Marcaillou-LeBaut C, Amzil Z, 1995. Unexplained toxicity in molluscs found in the course of phytoplankton monitoring. Toxicon 33: 1127. McKinley ER, Carlton WW, 1980. Patulin mycotoxicosis in Swiss ICR mice. Food and Cosmetics Toxicology 18: 181e187. McRae CF, Hocking AD, Seppelt RD, 1999. Penicillium species from terrestrial habitats in the Windmill Islands, East Antarctica, including a new species, Penicillium antarcticum. Polar Biology 21: 97e111. Naik JT, Mantle PG, Sheppard RN, Waight ES, 1995. Penitremones AeC, Penicillium metabolites containing an oxidized penitrem carbon skeleton giving insight into structure-tremorgenic relationships. Journal of the Chemical Society, Perkin Transactions 1: Organic and Bio-organic Chemistry 9: 1121e1125. Narui T, Sawada K, Takatsuki S, Okuyama T, Culberson CF, Culberson WL, Shibata S, 1998. NMR assignments of depsides and tridepsides of the lichen family Umbilicariaceae. Phytochemistry 48: 815e822. Please cite this article in press as: Vansteelandt M, et al., Patulin and secondary metabolite production by marine-derived Penicillium strains, Fungal Biology (2012), http://dx.doi.org/10.1016/j.funbio.2012.06.005 8 Nicoletti R, Lopez-Gresa MP, Manzo E, Carella A, Ciavatta ML, 2007. Production and fungitoxic activity of Sch 642305, a secondary metabolite of Penicillium canescens. Mycopathologia 163: 295e301. Nielsen KF, Smedsgaard J, 2003. Fungal metabolite screening: database of 474 mycotoxins and fungal metabolites for dereplication by standardised liquid chromatographyeUVemass spectrometry methodology. Journal of Chromatography A 1002: 111e136. Nielsen KF, Sumarah MW, Frisvad JC, Miller JD, 2006a. Production of metabolites from the Penicillium roqueforti complex. Journal of Agricultural and Food Chemistry 54: 3756e3763. Nielsen KF, Sumarah MW, Frisvad JC, Miller JD, 2006b. Production of metabolites from the Penicillium roqueforti complex. Journal of Agricultural and Food Chemistry 54: 5216 (erratum). Oxford AE, Raistrick H, Simonart P, 1939. Studies in the biochemistry of microorganisms. LX. Griseofulvin, C17H17O6Cl, a metabolic product of Penicillium griseofulvum Dierckx. Biochemical Journal 33: 240e248. Paterson RRM, Kemmelmeier C, 1989. Gradient high-performance liquid chromatography using alkylphenone retention indices of insecticidal extracts of Penicillium strains. Journal of Chromatography A 483: 153e168. ^ ncio A, Lima N, 2004. Solutions to Penicillium Paterson RRM, Vena taxonomy crucial to mycotoxin research and health. Research in Microbiology 155: 507e513. Petit K, Mondeguer F, Roquebert MF, Biard JF, Pouchus YF, 2004. Detection of griseofulvin in a marine strain of Penicillium waksmanii by ion trap mass spectrometry. Journal of Microbiological Methods 58: 59e65. Poirier L, Amiard JC, Mondeguer F, Quiniou F, Ruiz N, Pouchus YF, Montagu M, 2007. Determination of peptaibol trace amounts in marine sediments by liquid chromatography/electrospray ionizationeion trapemass spectrometry. Journal of Chromatography A 1160: 106e113. Ruiz N, Petit K, Vansteelandt M, Kerzaon I, Baudet J, Robiou-DuPont T, Amzil Z, Grovel O, Pouchus YF, 2010. Enhancement of M. Vansteelandt et al. domoic acid neurotoxicity on Diptera larvae bioassay by marine fungal metabolites. Toxicon 55: 805e810. Sallenave-Namont C, Pouchus YF, Robiou Du Pont T, Lassus P, Verbist JF, 2000. Toxigenic saprophytic fungi in marine shellfish farming areas. Mycopathologia 149: 21e25. Scott P, Van Walbeek W, MacLean W, 1971. Cladosporin, a new antifungal metabolite from Cladosporium cladosporioides. Journal of Antibiotics 24: 747e755. Shiono Y, Seino Y, Koseki T, Murayama T, Kimura KI, 2008. Antarones A and B, two polyketides from an endophytic Penicil€ r Naturforschung, B: Chemical lium antarcticum. Zeitschrift fu Sciences 63: 909e914. Slack GJ, Puniani E, Frisvad JC, Samson RA, Miller JD, 2009. Secondary metabolites from Eurotium species, Aspergillus calidoustus and A. insuetus common in Canadian homes with a review of their chemistry and biological activities. Mycological Research 113: 480e490. Smetanina OF, Kalinovskii AI, Khudyakov YV, Moiseenko OP, Pivkin MV, Menzorova NI, Sibirtsev YT, Kuznetsova TA, 2005. Metabolites of the marine fungus Aspergillus varians KMM 4630. Chemistry of Natural Compounds 41: 243e244. Smetanina OF, Kalinovsky AI, Khudyakova YV, Pivkin MV, Dmitrenok PS, Fedorov SN, Ji H, Kwak JY, Kuznetsova TA, 2007. Indole alkaloids produced by a marine fungus isolate of Penicillium janthinellum Biourge. Journal of Natural Products 70: 906e909. Tsuda M, Kasai Y, Komatsu K, Sone T, Tanaka M, 2004. Citrinadin A, a novel pentacyclic alkaloid from marine-derived fungus Penicillium citrinum. Organic Letters 6: 3087e3089. Van Eijk GW, 1969. Isolation and identification of orsellinic acid and penicillic acid produced by Penicillium fennelliae Stolk. Antonie van Leeuwenhoek 35: 497e504. Vinokurova NG, Ozerskaya SM, Baskunov BP, Arinbasarov MU, 2003. The Penicillium commune Thom and Penicillium clavigerum Demelius fungi producing fumigaclavines A and B. Microbiology 72: 149e151. Please cite this article in press as: Vansteelandt M, et al., Patulin and secondary metabolite production by marine-derived Penicillium strains, Fungal Biology (2012), http://dx.doi.org/10.1016/j.funbio.2012.06.005