Risk Analysis and Rapid Detection of the Genus Thermoascus, Food Spoilage Fungi

Food Control 41 (2014) 7e12

Contents lists available at ScienceDirect

Food Control

journal homepage: www.elsevier.com/locate/foodcont

Risk analysis and rapid detection of the genus Thermoascus, food spoilage fungi

Kouichi Hosoya a, Motokazu Nakayama a, Daisuke Tomiyama a, Tetsuhiro Matsuzawa b, Yumi Imanishi c, Seiichi Ueda d, Takashi Yaguchi b,* a Global R&D-Safety Science, Kao Corporation, 2606 Akabane, Ichikai-machi, Haga-Gun, Tochigi 321-3497, Japan b Medical Mycology Research Center, Chiba University, 1-8-1 Inohana, Chuo-ku, Chiba 260-8673, Japan c Department of Public Health, School of Medicine, Kurume University, 67 Asahimachi, Kurume, Fukuoka 830-0011, Japan d Graduate School of Human Health Science, University of Nagasaki, Japan article info abstract

Article history: Recently the numbers of spoilage incidents in food industry by the species of Thermoascus are increasing, Received 30 August 2013 but the risk of food spoilage have never been evaluated. Received in revised form It became obvious that their heat-resistances were higher than those of other heat-resist fungi, 2 December 2013 Byssochlamys, Hamigera and Neosartorya by our analyses. On the other hand, Thermoascus aurantiacus Accepted 17 December 2013 and Byssochlamys verrucosa had the idh gene, but they showed no patulin production in Potato dextrose broth or Czapek-glucose medium. Therefore, Thermoascus must be discriminated from other fungi in the Keywords: food industry. We developed a rapid and highly-sensitive method of detecting Thermoascus in the genus D value ﬁ Food spoilage level by using PCR. This method is expected to be extremely bene cial for the surveillance of raw ma- Heat-resistance terials in the food production process. idh gene Ó 2013 Elsevier Ltd. All rights reserved. PCR

1. Introduction a variety of agricultural products, including maize stored in sub- Sahara Africa and olive and olive cake in Morocco (Roussos et al., The history of food spoilage due to heat-resistant fungi dates 2006; Wareing, 1997), and in food-related environments (Ueda & back to the 1930s when incidents in canned foods were first re- Udagawa, 1983; Yaguchi, Someya, & Udagawa, 1995). Recently the ported; numerous incidents of spoilage of processed fruit juices and numbers of spoilage accidents in various processed tea and fruit beverages have been reported to date (Dijksterhuis, 2007; Kikoku, juice products by this genus are increasing (in our experience). In Tagashira, & Nakano, 2008; Samson, Hoekstra, Lund, Filtenborg, & addition, the production of high amounts of amylase and cellulase Frisvad, 2004; Tournas, 1994). It is known that asexual fungi have by Thermoascus spp. can markedly alter food product properties, hyphae and conidia that are susceptible to heat, and these fungi are and the high thermostability of these enzymes makes them diffi- typically killed with heat treatment at 70 C for 10 min. However, cult to inactivate with heat treatment (Adams, 1992; Feldman, heat-resistant fungi in genera Byssochlamys, Neosartorya, Hamigera, Lovett, & Tsao, 1988). However, the risk for food industry caused and Thermoascus belonging to the order Eurotiales (Plectomycetes) by this genus has never been reported, except for the article by form ascospores that are highly heat-resistant, allowing these King, Michener, and Ito (1969) who reported that ascospores of species of the genera to survive heat sterilization and be present in T. aurantiacus were survived at 88C for 60 min. Therefore, we finished acidic beverages (Samson et al., 2004). analyzed the growth, heat resistance and formation temperature The species of Thermoascus can, as the genus name implies, range of ascospores of Thermoascus spp. in order to establish risk grow at high temperatures. The major species of this genus include analysis data for each Thermoascus spp. Furthermore, we demon- Thermoascus crustaceus, Thermoascus thermophilus, Thermoascus strate that the species of Thermoascus and the closely related aurantiacus, and Thermoascus aegyptiacus (Houbraken & Samson, Byssochlamys genera produce patulin. To evaluate the risk of patulin 2011; Ueda & Udagawa, 1983). These fungi have been detected in production, we analyzed the detection of homology of the gene that encodes isoepoxydon dehydrogenase (idh) an important enzyme in patulin production (Dombrink-Kurtzman & Engberg, 2006; * Corresponding author. Tel.: þ81 43 226 2790; fax: þ81 43 226 2486. Paterson, Kozakiewicz, Locke, Brayford, & Jones, 2003; Puel, E-mail address: [email protected] (T. Yaguchi). Galtier, & Oswald, 2010). A detailed risk analysis based on patulin

0956-7135/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodcont.2013.12.021 8 K. Hosoya et al. / Food Control 41 (2014) 7e12 production was made for the species of Thermoascus grown in beverage model. The ascospores were adjusted to 106 spores/ml in culture medium. standard glucose-tartrate solution, and heat-resistance testing of On the other hand, the detection and identification of fungi has each species was performed using the thermal death time (TDT) been based on morphological observation, and there has been great test tube method (Ueda, Kawara, Yaguchi, & Udagawa, 2010). To interest in a rapid and versatile detection method. Due to the lack of assess the survival, germination rate was determined by counting informative genetic data, a rapid detection method for the species the number of germinated ascospores in triplicate (Ueda & Kawara, of Thermoascus had not previously been established. We success- 2010). A survival curve was plotted from the number of surviving fully developed a PCR method to detect and identify Byssochlamys fungi, and the D-value (min) was calculated. and Neosartorya to the genus and species levels (Hosoya et al., 2012; Nakayama et al., 2010; Yaguchi et al., 2012). 2.4. Evaluation of patulin production in culture and by detection of In this study, we first evaluated the risk of food spoilage on the the isoepoxydon dehydrogenase (idh) gene species of Thermoascus and found the necessity to distinguish Thermoascus spp. from other heat-resistant fungi. Next, we Evaluation for the presence or absence of the idh gene in analyzed various genes used in the phylogenetic classification of members of the Thermoascus genus and related species was per- fungi, designed new specific primers targeting gene sequence formed using the following region-specific primers that encode patterns specifictoThermoascus, and established a PCR method to functional domains (Hosoya et al., 2012): idh2444 (50- ATGCA- rapidly detect and distinguish Thermoascus at the genus level. CATGGAAGGCGAGAC-30) and idh2887 (50-CAAVGTGAATTCCGC- CATCAACCAAC-30). PCR was performed with 1 ml of 5 ng/mlDNA, 2. Materials and methods 1 ml of each primer (10 pmol), 22 ml of DW, and 25 ml of Sap- phireAmp Fast PCR Master Mix (Takara Bio Inc.) subjected to 35 2.1. Phylogenetic analysis of Thermoascus and related species cycles of denaturation at 98 C for 5 s, annealing at 59 C for 5 s, and extension at 72 C for 10 s. Fungi were cultured on potato dextrose agar (PDA) medium The PCR products were electrophoresed on a 2% agarose gel (Eiken Chemical Co., Tokyo, Japan) in the dark at 25 or 37 Cfor7 for 45 min at 100 V, and the presence or absence of bands and days. Fungal DNA was extracted using Dr. GenTLEÔ High Re- band sizes were evaluated. The PCR products were purified using covery Kit (Takara Bio, Inc. Ohtsu, Japan) and adjusted to 5 ng/ml High Pure PCR Product Purification Kits (Roche, Mannheim, in TE buffer. The region coding the large subunit of the RNA Germany). The purified PCR products were labeled with idh2444 polymerase II gene (RPB1) was amplified and the PCR products or idh2887 primers using a Terminator Cycle Sequencing Ready were labeled using the Terminator Cycle Sequencing Ready Re- Reaction Kit (Applied Biosystems). The base sequence was action Kit (Applied Biosystems, Foster City, CA), following the determined using an ABI PRISM 3130 (Applied Biosystems). Ho- method of Samson et al. (2011). The DNA sequences were mology between the idh gene of members of the Thermoascus determined using an ABI PRISM 3130 Genetic Analyzer (Applied genus and Byssochlamys nivea (which produces patulin) was Biosystems). Using the sequences determined here and se- calculated using DNASIS Pro (Hitachi), and patulin production quences from other species of the genus Thermoascus and related potential was evaluated. species obtained through an ARSA search of the DNA Data Bank Patulin production was evaluated in the culture media of Ther- of Japan (DDBJ) (http://arsa.ddbj.nig.ac.jp/html/) and our previ- moascus spp. cultures. The fungi tested were T. thermophilus IFM ously determined sequences, a sequence alignment and neighbor 60075, T. crustaceus IFM 60077, T. aurantiacus IFM 57325, joining trees were prepared using ClustalX software (http:// Byssochlamys verrucosa IFM 48423, and B. nivea NBRC 57325. Fungi clustalx.ddbj.nig.ac.jp/top-.html). was inoculated on 100 ml of Czapek-glucose medium (Wako Pure Chemical Industries, Ltd., Osaka, Japan) or potato dextrose broth 2.2. Growth and ascospore formation of Thermoascus spp. (PDB) (Difco) and cultured in the dark at 35 C for 7 days without agitation. Then, the culture broth containing each fungus was To evaluate the potential for food spoilage due to individual adjusted to pH 3.6 with aqueous acetic acid, extracted with ethyl Thermoascus spp. over a range of temperatures, the growth and acetate, and washed with alkali. Patulin was quantified by high- ascospore formation were examined for the following test fungi: performance liquid chromatography (Nanospace SI-2, Shiseido, T. aegyptiacus IFM 61569, T. aurantiacus IFM 57325, T. crustaceus IFM Tokyo, Japan) fitted with a Mightysil RP-18GP column (Kanto 57326, and T. thermophilus IFM 59664. Each species was inoculated Kagaku and with detection at 290 nm). The detection limit for on oatmeal agar medium (Difco, BD, Sparks, MD) or PDA medium patulin was 0.05 ppm. (Eiken Chemical Co., Tokyo, Japan) in triplicate and cultured at a temperature range of 20e60 C for 14 days. Growth temperature 2.5. Design of Thermoascus genus-specific primer set range was assessed by visual inspection, and ascospore formation temperature range was assessed by microscopic examination. For Thermoascus spp. listed in Table 1, the homology of the RPB1 gene was calculated by multiple alignment analysis (DNASIS Pro, 2.3. Heat resistance of ascospores on Thermoascus Hitachi Software, Tokyo, Japan), conserved sequences in Thermoascus were identified and used to design genus-specificprimers.In T. aegyptiacus IFM 61569, T. aurantiacus IFM 57325 and designing genus-specific primer set, conservation at the 30 end of T. crustaceus IFM 57326 were cultured at 35 C and T. thermophilus extension in each Thermoascus spp., selection of base sequences with IFM 59664 was cultured at 30 C on oatmeal agar medium (Difco) in little similarity to fungi of other genera, Tm of 60 2 C, about 20 the dark for 30 days. The ascocarps were collected in centrifuge bases, and about 50% GC content were conditions as primer target tubes containing 0.1 M phosphate buffer (pH 7.0). Then, the hypha regions. Based on these criteria, a set of Thermoascus genus-specific were removed with sterilized gauze. After centrifugation (700 g, primers was also designed: RPB_F2 (50-ATCTGCCGGCGT- 10 min), the supernatant was discarded, and the pellets were GATGTGTTCCTG -30) and RPB_R2 (50-GTTGTGCAGAAGCCAGTTGACC- washed 3 times with 0.1 M phosphate. 30). A search for homology of the designed base sequences was per- A standard glucose-tartrate solution (glucose, 16 g; DL-tartaric formed using the Basic Local Alignment Search Tool search (http:// acid, 0.5 g; DW, 100 ml; adjusted to pH 3.6) was used as an acidic www/ncbi.nlm.nih.gov). K. Hosoya et al. / Food Control 41 (2014) 7e12 9

Table 1 Strains used in this study.

Species Strain No. GenBank accession No. DNA template No RPB_F2/R2 detection

Thermoascus aegyptiacus IFM 61569T ¼ NHL 2914T AB849502 4 þ T. aurantiacus IFM 57325 ¼ NBRC 6766 þ T. aurantiacus IFM 60076 3 þ T. aurantiacus IFM 60078 ¼ CBS 256.34 þ T. aurantiacus IFM 60080 ¼ CBS 100054 þ T. aurantiacus NBRC 31693 þ T. aurantiacus CBS 396.78 JN121671 None examined T. crustaceus IFM 57326 ¼ NBRC 9129 þ T. crustaceus IFM 60077T ¼ CBS 181.67T JN121591 2 þ T. crustaceus IFM 60232 ¼ CBS 374.62 þ T. thermophilus IFM 59664 ¼ SUN47 þ T. thermophilus IFM 59665 ¼ SUN49 þ T. thermophilus IFM 60075NT ¼ CBS 528.71NT JN121697 1 þ Byssochlamys verrucosa IFM 48423T ¼ IAM 13423T JN680311 5 þ B. fulva IFM 51213T ¼ CBS 132.33T AB849503 13 B. nivea IFM 51243T ¼ CBS 100.11T JN121511 12 B. spectabilis IFM 52963T ¼ CBS 101075T JN121554 11 Hamigera avellanea IFM 52957isoT ¼ CBS295.48isoT JN121632 15 H. striata IFM 52958NT ¼ CBS 377.48NT JN121665 14 Rasamsonia emersonii IFM 52961T ¼ CBS 393.64T JN121670 10 Talaromyces ﬂavus IFM 52962NT ¼ CBS 310.38NT JN121639 6 Ta. trachyspermus IFM 52964T ¼ CBS 373.48T JN121664 7 Ta. wortmanii IFM 53866T ¼ CBS 91.48T JN121669 8 Ta. luteus (Out group) IFM 53168T ¼ CBS 348.51T JN121656 9 A. fumigatus DAOM 215394 JN985124

T; ex type, NT; neotype.

2.6. Evaluation of Thermoascus genus-specific primer set electrophoresed on 2% agarose gel for 45 min at 100 V, and the presence or absence of bands and band size was evaluated. To evaluate the specificity of designed primer set, DNA from We evaluated the effect of contamination of other fungal DNA in Thermoascus spp. cultures and other related fungi frequently iso- Thermoascus detection method. T. crustaceus IFM 60232 DNA (1 ng) lated from food processing environments (Table 2) was subjected to and Penicillium citrinum NBRC 6352 DNA (1e1000 ng), was ampli- PCR with primers set. The DNA was extracted using a Gen TLEÔ fied using a GenomiPhi V2 DNA Amplification Kit (GE Healthcare, High Recovery Kit (Takara Bio, Inc.). Following extraction, DNA used Wauwatosa, WI), were added to PCR mixtures using the primer pair in these assays was stored at 20 C and used in assays over a RPB_F2/R2. period of 6 months. PCR was performed with a mixture consisting To evaluate the detection of Thermoascus in beverages, of 1 ml of 5 ng/mlDNA,1ml of each primer (10 pmol), 22 mlofDW, T. crustaceus IFM 60232 ascospores (10 cfu) were added to 500 ml of and 25 ml of SapphireAmp Fast PCR Master Mix (Takara Bio, Inc.) an acidic isotonic beverage (pH 3.4). This was filtered using a and amplification was carried out by heating at 97 C for 10 min MicrofilVfiltration device (MERCK Millipore, Billeric, MX) and followed by 30 cycles of 97 C for 1 min, 62 C for 1 min, and 72 C cultured on PDA medium (Difco) at 30 C for 22 h. Then DNA was for 1 min. extracted using PrepMan Ultra Sample Preparation Reagent PCR products were electrophoresed on 2% agarose gels for (Applied Biosystems), and PCR was performed with conditions 45 min at 100 V, and the presence or absence of bands and band described above. sizes were evaluated. To determine detection sensitivity for the designed primers, genomic DNA from T. crustaceus IFM 60232 was 3. Results serially diluted (10 nge10 pg), and the first PCR and nested PCR were performed. For nested PCR, the first PCR products were pu- 3.1. Phylogenetic analysis of the species of Thermoascus and related rified using High Pure PCR Product Purification Kits (Roche, Man- species nheim, Germany), and PCR was performed again with the same conditions mentioned above. These PCR products were also The phylogenetic tree of Thermoascus and related species based on RPB1 gene sequences was shown in Fig. 1. All species of Ther- moascus were clustered together. B. verrucosa placed in the Ther- Table 2 moascus-clade, thus showing high relatedness to the species of Growth and ascospore formation temperature of members of genus Thermoascus Thermoascus. The phylogenetic trees made based on other gene and Byssochlamys verrucosa. sequences, such as the b-tubulin gene, showed similar grouping Species Strain Temperature (C) (data not shown). Thermoascus aegypticus IFM 61569 Growth 20e55 Ascospore forming 25e40 3.2. Growth and ascospore formation for the species of T. aurantiacus IFM 57325 Growth 35e57 Thermoascus Ascospore forming 35e40 T. crustaceus IFM 57326 Growth 25e50 Ascospore forming 25e35 Table 2 showed the temperature range at which growth and T. thermophilus IFM 59664 Growth 20e53 ascospore formation was observed for each species on oatmeal Ascospore forming 20e45 medium. T. aegyptiacus, T. crustaceus, T. thermophilus, and Byssochlamys verrucosa IFM 48423 Growth 20e53 B. verrucosa grew between 25 and 50 C. The temperature range for Ascospore forming None T. aurantiacus was higher at 33e57 C. Meanwhile, the highest 10 K. Hosoya et al. / Food Control 41 (2014) 7e12

Fig. 2. PCR amplification to detect the idh gene using DNA of genera Thermoascus and Byssochlamys with primer set idh2444 and idh2887. Numbers on the lanes correspond to the following markers and genomic DNA templates: M, 100 bp ladder; 1, T. thermophilus IFM 60075; 2, T. thermophilus IFM 59665; 3, T. crustaceus IFM 60077; 4, T. Fig. 1. Neighbor joining tree based on RPB1 gene sequences for Thermoascus and crustaceus IFM 57326; 5, T. aurantiacus IFM 57325; 6, T. aurantiacus IFM 60076; 7, related species. Bootstrap samplings based on 1000 samplings supporting the internal Byssochlamys nivea NBRC 57325; 8, T. aegyptiacus IFM 61569 and 9, B. nivea IFM 48423. branches with a probability of 50% or higher are shown. growth temperature observed for B. verrucosa was 53 C, which was 3.4. Evaluation for presence or absence of the idh gene and patulin a higher temperature range than was observed for other Byssochl- production in culture media amys spp. T. crustaceus showed ascospore formation at 25e35 C, while T. aegyptiacus and T. thermophilus showed ascospore forma- The PCR results using the idh gene-specific primers (idh2444 tion at temperatures that were 5e10 C higher. B. verrucosa did not and idh2887) are shown in Fig. 2.ForB. nivea, which produces form ascospores under these culture conditions. patulin, a 400-bp PCR product was detected. Signals identical in size to those produced for B. nivea were also detected for 3.3. Heat resistance of ascospores formed by the species of T. aurantiacus and B. verrucosa and showed 99.9% base sequence Thermoascus homology and complete matching at the amino acid level. How- ever, no PCR products were detected for T. aegyptiacus, T. crustaceus The results of heat resistance of ascospores on Thermoascus and T. thermophilus. species are shown at Table 3. The ascospores of T. aegyptiacus IFM The results for patulin production in the culture media for 61569 and T. thermophilus IFM 59664, IFM 60075 did not germinate Thermoascus and related species were shown in Table 4. Patulin after heat treatment under 90 C. The heating at 90 C provided the production was 55.6 ppm in PDB and 42.8 ppm in Czapek-glucose best germination of these ascopores, therefore only D-values of medium for B. nivea, which has the idh gene. On the other hand, these strains at 90 C were determined, but the z-values were not despite high homology with the B. nivea idh gene, T. aurantiacus calculated. Those of T. aurantiacus IFM 57325 and NBRC 31693 did and B. verrucosa showed no patulin production in PDB or Czapek- not germinate after heating at 90 C. Inactivation of ascospore glucose medium. Patulin production was also not detected for germination by heating were observed in the range 80e85 C, other fungal species for which the absence of the idh gene was therefore D-values at 80, 83 and 85 C were determined and the z- suggested. values were subsequently calculated. The D-value for T. crustaceus IFM 57326 could not be accurately determined, but survival was confirmed even after heating at 90 C for 90 min. Table 4 Patulin production of genus Thermoascus and related species.

Table 3 Species Strain Medium Patulin (ppm) Heat resistance of strains of genus Thermoascus based on D and z-values. Thermoascus thermophilus CBS 528.71 CzLa N.D. b PDB N.D. Species Strain D-value (min) z-value ( C) T. crustaceus IFM 60077 CzL N.D. Heating temperature ( C) PDB N.D. T. aurantiacus IFM 57325 CzL N.D. 80 83 85 90 PDB N.D. Thermoascus aegyptiacus IFM 61569 eee56.2 e Byssochlamys verrucosa IFM 48423 CzL N.D. T. aurantiacus IFM 57325 57.1 13.2 10.8 e 5.2 PDB N.D. T. aurantiacus NBRC 31693 288 50.1 16.2 e 4.0 Byssochlamys nivea NBRC 31351 CzL 42.8 T. thermophilus IFM 59664 eee21.3 e PDB 55.6 T. thermophilus IFM 60075 eee20.5 e a CzL ¼ Czapek-glucose liquid. e: not evaluated. b PDB ¼ Potato dextrose broth. K. Hosoya et al. / Food Control 41 (2014) 7e12 11

Fig. 3. PCR amplifications using the primer set, RBP_F2/R2, to detect strains in the genus Thermoascus and B. verrucosa. Numbers indicate the genomic DNA templates (1e15) shown in Table 1;16,Neosartorya fischeri IFM 57324; 17, N. spinosa IFM 47025; 18, Asperguillus fumigatus IFM 47042; 19, A. niger IFM 55890; 20, A. flavus IFM 48054; 21, Eupenicillium brefeldianum IFM 42321; 22, Penicillium griseofulvum IFM 49451; 23, Alterraria alternata IFM 56020; 24, Aureobasidium pullulans IFM 41411, 25, Chaetomium globosum IFM 40869; 26, Fusarium oxysporum IFM 50002; 27, Tricoderma viride IFM 51045; 28, Cladosporium cladosporioides IFM 46166; M, 100 bp ladder.

3.5. Evaluation of Thermoascus genus-specific primer set Byssochlamys fulva (D85 ¼ 12.3 min (Hosoya et al., 2012)) and Neosartorya fischeri (D85 ¼ 10e35 min (Dijksterhuis, 2007)), which PCR was performed using the RPB_F2 and RPB_R2 primer set, and has highly heat-resistant ascospores among the main heat- the presence or absence of PCR amplification products and the sizes resistance fungi. Specially, that of T. agyptiacus was the highest were evaluated by electrophoresis. Of the 28 strains tested in this (D90 ¼ 56.2 min). Additionally, all species of Thermoascus examined study, gene amplification products with a size of about 400 bp (Fig. 3), were able to grow more than at 50 C. For Thermoascus under which is consistent with the primer design, were detected for all standard sterilization conditions of vegetative cells to 80 C for Thermoascus species and strains listed inTable 1. Similar findings were 10 min, a value of more than 5D indicates that commercial detected for B. verrucosa, which is phylogenetically very closely pasteurization does not achieve complete sterilization. The food related to Thermoascus genus. However, no gene amplification prod- industry recognizes the species of Thermoascus as being difficult to ucts were detected for other species of Byssochlamys genus, Hamigera control and, thus, accurate detection of these species to the genus genus, or other fungi that are closely associated with spoilage in food level in raw materials and in the manufacturing environment is products (Fig. 3). In addition, gene amplification products with a size important in controlling and preventing contamination with of about 400 bp were produced for PCR conducted on DNA extracted Thermoascus spp. Therefore, the development of a rapid method for from T. aurantiacus that was added to an acidic beverage. detecting Thermoascus spp. is essential. But, Thermoascus has a Furthermore, because multiple fungal species are frequently similar type of sclerotioid cleistothecium some species of Eupeni- detected in food processing environments, we evaluated whether cillium and the order Thermoascaceae is composed of two closely this PCR assay could detect DNA from the Thermoascus genus in related clades (clades Thermoascus and Byssochlamys)(Houbraken samples that included DNA from other fungi. DNA (1 ng) from & Samson, 2011), making it difficult to detect and discriminate T. crustaceus IFM 60232 and amplified DNA (1e1000 ng) from members of genus Thermoascus using traditional methods based on P. citrinum NBRC 6352 were added to the PCR mixture and ampli- morphology. fied. Even with 1000 times the amount of P. citrinum DNA added, Samson, Houbraken, Varga, and Frisvad (2009) noted that this PCR detected Thermoascus genus DNA. B. verrucosa is misidentified in the genus Byssochlamys, and this The detection sensitivity for PCR assay using RPB_F2 and observation was subsequently confirmed by the clustering of this RPB_R2 was between 100 pg of template DNA for the first PCR. With species with Thermoascus on a phylogenetic tree based on the nested PCR, PCR products were detected even with 10 pg of tem- sequence of the RPB1 gene (Houbraken & Samson, 2011) and the b- plate DNA. Based on these results, detection sensitivity for RPB_F2 tubulin gene (Nakayama et al., 2010; Samson et al., 2009). and RPB_R2 on first PCR was 100 pg, and detection sensitivity Furthermore, B. verrucosa can grow even at 53 C, a temperature at improved 10-fold with nested PCR (Fig. 4). which other Byssochlamys spp. cannot grow. However B. verrucosa has never been reclassified to Thermoascus. Therefore we developed 4. Discussion a new detection method for detecting Thermoascus spp. and B. verrucosa. As PCR methods are rapid and useful for detecting We first evaluated the risk of food spoilage on each species of fungi, we developed a PCR method using the RPB1 gene and genus- Thermoascus. Their heat-resistances were higher than those of specific primer set (RPB_F2/R2) to detect Thermoascus spp. and B. verrucosa. Using this primer set, it was possible to detect Ther- moascus spp. and B. verrucosa and discriminate from other fungi. We found that B. nivea and Byssochlamys lagunculariae, which are related to the genus Thermoascus, have the idh gene and produce patulin (Hosoya et al., 2012) and that Byssochlamys and some species of Thermoascus have the same anamorph (Paecilomyces) (Houbraken & Samson, 2011). As T. aurantiacus was toxic to chicken embryos and weanling rats (Davis, Wagener, Morgan-Jones, & Diener, 1975), we hypothesized that Thermoascus spp. carries the risk of patulin production. We elucidated that T. aurantiacus and B. verrucosa had the idh gene and that each homology of this idh gene with that of B. nivea was 99.9%. However, T. aegyptiacus, T. crustaceus and T. thermophilus did not have the idh gene, and formed a monophyletic clade separated from T. aurantiacus and B. verrucosa in the phylogenic tree (Fig. 1). In this study, we did not Fig. 4. Detection limit of the PCR with the RBP_F2/R2 primer set: using T. crustaceus detect patulin in PDB and Czapek-glucose liquid medium despite IFM 60232 DNA. Lane numbers 1e4 refer to the first PCR only, and 5e8 refer to the nested PCR. The amount of DNA used in the PCR mixtures are as follows: 10 ng (1, 6), abundant fungal growth. The reason is due to a defect of the patulin 1 ng (2, 7), 100 pg (3, 8) and 10 pg (4, 9), 1 pg (5, 10). M: 100 bp-ladder. biosynthetic gene other than the idh gene, such as 6-MSA synthase 12 K. Hosoya et al. / Food Control 41 (2014) 7e12

(Puel et al., 2010) or deactivation of IDH (Puel, Tadrist, Delaforge, Feldman, K. A., Lovett, J. S., & Tsao, T. (1988). Isolation of the cellulase enzymes from Oswald, & Lebrihi, 2007). From this finding, we determined that the thermophilic fungus Thermoascus aurantiacus and regulation of enzyme production. Enzyme Microbial Technology, 10, 262e272. Thermoascus spp. has a very low risk of patulin production. This is Hosoya, K., Motokazu, N., Matsuzawa, T., Imanishi, Y., Hitomi, J., & Yaguchi, T. (2012). the first report about the risk evaluation of patulin produced by Risk analysis and development of a rapid method for identifying four species of e Thermoascus spp. This genus is used to produce heat-resistant en- Byssochlamys. Food Control, 26,169 173. Houbraken, J., & Samson, R. A. (2011). Phylogeny of Penicillium and the segregation zymes, making our study very important for not only food safety of Trichocomaceae into three families. Studies in Mycology, 70,1e51. but also for enzymatic industry. Kikoku, Y., Tagashira, N., & Nakano, H. (2008). Heat resistance of fungi isolated from Using our detection method by PCR, even dead fungi can be frozen blueberries. Journal of Food Protection, 71, 2030e2035. King,D.A.,Jr.,Michener,H.D.,&Ito,K.A.(1969).ControlofByssochlamys and detected if the DNA is extracted. Therefore, the causative organisms related heat-resistant fungi in grape products. Applied Microbiology, 18, of food spoilage can be determined, even in cases in which 166e173. morphologically identifiable fungi are not isolated from the food Nakayama, M., Hosoya, K., Matsuzawa, T., Hiro, Y., Sako, A., Tokuda, H., et al. (2010). A rapid method for identifying Byssochlamys and Hamigera. Journal of Food product. Furthermore, this new detection method is unaffected by Protection, 73, 1486e1492. acidic beverages. As this method can be performed rapidly with Paterson, R. R. M., Kozakiewicz, Z., Locke, T., Brayford, D., & Jones, S. C. B. (2003). high sensitivity and requires no special expertise, it is expected to Novel use of the isoepoxydon dehydrogenase gene probe of the patulin be extremely beneficial for the surveillance of raw materials in the metabolic pathway and chromatography to test penicillia isolated from apple production systems for the potential to contaminate apple juice with patulin. food production process. Food Microbiology, 20,359e364. Puel, O., Galtier, P., & Oswald, I. P. (2010). Biosynthesis and toxicological effects of patulin. Toxins, 2,613e631. 5. Conclusion Puel, O., Tadrist, S., Delaforge, M., Oswald, I. P., & Lebrihi, A. (2007). The inability of Byssochlamys fulva to produce patulin is related to absence of 6-methylsalicylic fi acid synthase and isoepoxydon dehydrogenase genes. International Journal of We rst evaluated the risk of food spoilage on the species of Food Microbiology, 115,131e139. Thermoascus. Their heat-resistances were higher than those of Roussos, S., Zaouia, N., Salih, G., Tantaoui-Elaraki, A., Lamrani, K., Cheheb, M., et al. other heat-resist fungi. On the other hand, T. aurantiacus and (2006). Characterization of filamentous fungi isolated from Moroccan olive and olive cake: toxinogenic potential of Aspergillus strains. Molecular Nutrition & B. verrucosa has the idh gene, but they showed no patulin produc- Food Research, 50,500e506. tion in PDB or Czapek-glucose medium. Therefore, Thermoascus Samson, R. A., Hoekstra, E. S., Lund, F., Filtenborg, O., & Frisvad, J. C. (2004). Methods must be discriminated from other fungi in the food industry. We for the detection, isolation and characterization of food-borne fungi. In R. A. Samson, E. S. Hoekstra, & J. C. Frisvad (Eds.), Introduction to food and developed a PCR method of detecting Thermoascus spp. in the airborne fungi (7th ed.). Utrecht: Centraalbureau voor Schimmelcultures. genus level. As the PCR method can be performed rapidly with high Samson, R. A., Houbraken, J., Varga, J., & Frisvad, J. C. (2009). Polyphasic taxonomy of sensitivity, it is effective to detect heat-resistant fungi in the food the heat resistant ascomycete genus Byssochlamys and its Paecilomyces ana- morphs. Persoonia, 22,14e27. production process. Samson, R. A., Yilmaz, N., Houbraken, J., Spierenburg, H., Seifert, K. A., Peterson, S. W., et al. (2011). Phylogeny and nomenclature of the genus Talar- omyces and taxa accommodated in Penicillium subgenus Biverticillium. Studies Acknowledgement in Mycology, 70,159e183. Tournas, V. (1994). Heat-resistant fungi of importance to the food and beverage e This work was supported in part by the National BioResource industry. Critical Reviews in Microbiology, 20, 243 263. Ueda, S., & Kawara, N. (2010). Development of a new D-value testing method for Project e Pathogenic Microbes in Japan (http://www.nbrp.jp/). heat-resistant molds using microscopical counting of ascospore germination. Nippon Kingakukai Kaiho, 51,27e34. Ueda, S., Kawara, N., Yaguchi, T., & Udagawa, S. (2010). Identification and heat- References resistance characteristics of Byssochlamys lagunculariae, a mold isolated from citrus raw material of spoiled fruit jelly products. Nippon Kingakukai Kaiho, 51, e Adams, P. R. (1992). Growth and amylase production of Thermoascus aurantiacus 48 58. Miehe. Biotechnology and Applied Biochemistry, 15,311e313. Ueda, S., & Udagawa, S. (1983). Thermoascus aegyptiacus, a new thermophilic e Davis, N. D., Wagener, R. E., Morgan-Jones, G., & Diener, U. L. (1975). Toxigenic ascomycete. Transactions of the Mycological Society of Japan, 24,132 142. thermophilic and thermotolerant fungi. Applied Microbiology, 29, 455e457. Wareing, P. W. (1997). Incidence and detection of thermotolerant and thermophilic Dijksterhuis, J. (2007). Heat-resistant ascospores. In J. J. Dijksterhuis, & R. A. Samson fungi from maize with particular reference to Thermoascus species. International e (Eds.), Food mycology. A multifaceted approach to fungi and food (p. 17). Boca Journal of Food Microbiology, 35,137 145. Raton, FL: CRC Press. Yaguchi, T., Imanishi, Y., Matsuzawa, T., Hosoya, K., Hitomi, J., & Nakayama, M. Dombrink-Kurtzman, M. A., & Engberg, A. E. (2006). Byssochlamys nivea with (2012). Method for identifying heat-resistant fungi of the genus Neosartorya. e patulin-producing capability has an isoepoxydon dehydrogenase gene (idh) Journal of Food Protection, 75, 1806 1813. with sequence homology to Penicillium expansum and P. griseofulvum. Myco- Yaguchi, T., Someya, A., & Udagawa, S. (1995). Two interesting cleistothecial asco- e logical Research, 110, 1111e1118. mycetes from soils. Mycoscience, 36, 151 154.