Analysis of the Fungal Population Involved in Katsuobushi Production
Total Page:16
File Type:pdf, Size:1020Kb
Advance Publication J. Gen. Appl. Microbiol. doi 10.2323/jgam.2019.09.003 ©2020 Applied Microbiology, Molecular and Cellular Biosciences Research Foundation 1 The Journal of General and Applied Microbiology 2 Short Communication 3 4 Running title: 5 Fungi in Katsuobushi 6 7 Analysis of the fungal population involved in Katsuobushi production 8 (Received August 5, 2019; Accepted September 24, 2019; J-STAGE Advance publication date: January 31, 2020) 9 Chihiro Kadooka,1,2† Eri Nakamura,1† Shingo Kubo,3 Kayu Okutsu,1 Yumiko 10 Yoshizaki,1,2 Kazunori Takamine,1,2 Hisanori Tamaki,1,2 Taiki Futagami1,2* 11 12 1 Education and Research Center for Fermentation Studies, Faculty of Agriculture, 13 Kagoshima University, Korimoto1-21-24, Kagoshima, 890-0065, Japan. 14 2 United Graduate School of Agricultural Sciences, Kagoshima University, Korimoto 15 1-21-24, Kagoshima, 890-0065, Japan. 16 3 Division of Instrumental Analysis, Research Support Center, Kagoshima University, 17 Korimoto 1-21-24, Kagoshima, 890-0065, Japan. 18 19 †These authors have contributed equally to this work. 20 21 *To whom correspondence should be addressed. 22 Education and Research Center for Fermentation Studies, Faculty of Agriculture, 1 23 Kagoshima University, 1-21-24, Korimoto, Kagoshima, 890-0065, Japan 24 Tel/Fax: 81-99-285-3536 25 Email: [email protected] 26 Summary 27 Naturally occurring fungi have been used in the traditional production of dried bonito, 28 Katsuobushi, in Japan. In this study, we analyzed the fungal population present during 29 Katsuobushi production. Amplicon sequence analysis of ITS1 indicated that Aspergillus 30 spp. are predominant throughout the production process. In addition, culture-dependent 31 analyzes identified three species Aspergillus chevalieri, Aspergillus montevidensis, and 32 Aspergillus sydowii, based on sequencing of benA, caM, and rpb2 genes. A. chevalieri 33 isolates were classified into teleomorphic and anamorphic strains based on 34 morphological analysis. A. chevarieri was the dominant species throughout the 35 production process, whereas A. montevidensis increased and A. sydowii decreased in 36 abundance during Katsuobushi production. Our study will enhance the understanding of 37 fungal species involved in traditional Katsuobushi production. 38 39 Key words: Aspergillus chevalieri; Aspergillus montevidensis; Aspergillus sydowii; 40 fungi; Katsuobushi 41 42 Katsuobushi is a dried bonito, which is essential for traditional Japanese food. It 43 is sliced as dried bonito flakes and used for “dashi,” which is a stock. For Katsuobushi 44 production, molding and sun-drying processes are repeated several times. Molding 45 enhances the drying of Katsuobushi. In addition, it plays a significant role in the 46 decomposition of proteins and lipids, as well as the flavor formation of Katsuobushi 47 (Aoki et al., 2012; Dimici and Wada, 1994; Doi et al., 1992; Doi et al., 1989; Doi and 2 48 Shuto, 1995; Kunimoto et al., 1996; Kaminishi et al., 1999). 49 The fungal starter Aspergillus glaucus (formerly Eurotium herbariorum) is 50 available from the Japan Katsuobushi Association (Tokyo, Japan) for molding (Miyake 51 et al., 2009); however, traditional Katsuobushi production is performed using naturally 52 occurring fungi in a room for molding. Prior studies reported that species from the A. 53 glaucus group such as Aspergillus ruber and Aspergillus pseudoglaucus (formerly 54 Aspergillus repens) were dominant in Katsuobushi based on the morphological 55 characteristics of the fungi (Nakazawa et al., 1934; Yoshikawa and Kosugi, 1937). 56 However, the recent classification method based on DNA sequence (Chen et al., 2017; 57 Samson et al., 2014) has not been applied for the classification of the fungal population 58 in Katsuobushi. In this study, we characterized the fugal population in Katsuobushi 59 production using DNA sequencing. 60 The Katsuobushi samples with different molding steps were prepared at a 61 Katsuobushi manufacturing factory (Makurazaki, Kagoshima, Japan) using the same 62 method used to make the commercial product. Briefly, the head and viscera of bonito, 63 Katsuwonus pelamis, were removed and fileted, then further separated into the back and 64 stomach sides. Thus, Katsuobushi can be distinguished into “obushi” and “mebushi,” 65 which are made from back side and stomach side of bonito, respectively. Then, the filets 66 were dipped in hot water at 90–95°C for 60–90 min. After cooling down, bones and a 67 part of the skin were removed. Then, the filets were repeatedly dried in a room with 68 burned firewood from 10 to 15 times. The surface layer was removed and preserved in 69 the room for molding for 2–3 weeks and then sun-dried. The molding and sun-drying 70 steps were repeated up to three times. Each sample was obtained after each sun-drying 71 step, yielding the first, second, and third steps of obushi or mebushi samples. 72 To investigate the population of fungi in Katsuobushi, we performed sequence 3 73 analysis of the ITS1 (Internal Transcribed Spacer 1) region amplified from DNA 74 isolated from Katsuobushi. Approximately 5 mm of the surface layer of obushi or 75 mebushi samples were scraped with a chisel and treated with bead beating at 4.5 m/s for 76 30 s using a FastPrep 120 Cell Disrupter System (Thermo Savant; Carlsbad, CA) and 77 glass beads BZ-4 (0.35–0.5 mm; AS ONE Corporation, Osaka, Japan), then subjected to 78 DNA extraction. DNA was extracted from 70 mg of each sample using the ZR fecal 79 DNA MiniPrep kit (Zymo Research, Irvine, CA) according to the manufacturer’s 80 protocol. Sequencing analysis of ITS1 amplified using the ITS1F_KYO1 and 81 ITS2_KYO2 primer set (Toju et al., 2012) was performed using MiSeq (Illumina, San 82 Diego, CA) at the Bioengineering Lab. Co., Ltd. (Kanagawa, Japan). The data were 83 deposited to the DNA Data Bank of Japan (DDBJ) under the accession numbers 84 DRR186742-DRR186744. The sequence reads were subjected to quality filtering using 85 FASTX-Toolkit (http://hannonlab.cshl.edu/fastx_toolkit) and low quality reads with 86 quality values of less than 20 or with lengths of 40 bases were removed after trimming 87 the primer sequences. Then, the pair-end reads were generated and subjected to chimera 88 check using UCHIME (Edgar et al., 2011) with a 97% representative operational 89 taxonomic units (OTUs) set of the UNITE database (Kõljalg et al, 2013). After the 90 quality filtering and chimera check of sequencing reads, the OTUs were predicted using 91 a Quantitative Insight Into Microbial Ecology (QIIME) (Caporaso et al., 2010) without 92 using any external set of reference sequences. The QIIME analysis was performed with 93 default settings. The result indicated that approximately 99.9% of ITS1 amplicons were 94 classified into genus Aspergillus at the first, second, and third molding steps (Table S1). 95 This result was inconsistent with the previous report that Penicillium spp. were also 96 present with Aspergillus spp. during Katsuobushi production (Yoshikawa and Kosugi, 97 1937). Conceivable explanations for this difference could be: (i) taxonomic 4 98 identification difficulties by morphology (e.g., misidentification in the previous study) 99 and (ii) the fungal populations might depend on Katsuobushi samples (e.g., 100 factory-dependent fungal population). 101 A culture-dependent approach was used to perform species-level identification. 102 Approximately 5 mm of the surface layer of obushi and mebushi samples were scraped 103 with a chisel and dissolved in sterilized water. Then, the suspension solutions were 104 spread on Potato Dextrose Agar (PDA) (BD Difco, Franklin Lakes, NJ); single colonies 105 were transferred to new media and classified based on colony morphology with a focus 106 on size, shape, and color (spore color and bottom color of petri dish). Approximately 107 100 colonies (from 82 to 156 colony for each sample, 2110 total colonies) were 108 morphologically classified from each obushi and mebushi sample, and four different 109 types of colonies, namely strains M1, M2, M3, and M4, were obtained (Fig. 1A). 110 Strains M1, M2, M3, and M4 showed better growth in M40Y medium (40% [wt/vol] 111 sucrose, 2% [wt/vol] malt extract, 0.5% [wt/vol] yeast extract, 2% [wt/vol] agar) 112 (DSMZ-Medium 187), indicating that they are osmophilic fungi. There is no significant 113 difference in the relative abundance of strains M1, M2, M3, or M4 between obushi and 114 mebushi samples. However, the relative colony numbers of strain M1 and M2 are stable 115 at the first, second, and third steps of molding, whereas those of strains M3 and M4 tend 116 to increase and decrease over the molding step, respectively (Fig. 1B). 117 For the identification of strains M1, M2, M3, and M4, the ITS1–5.8S rDNA–ITS2 118 region was amplified by colony PCR analysis using a primer set of ITS1 and ITS4 119 (White et al., 1990), and subsequent sequencing analysis indicated that all strains were 120 classified into genus Aspergillus (data not shown). This was consistent with the result 121 that nearly all sequence reads of the ITS1 amplicon sequence analysis were classified 122 into genus Aspergillus (Table S1). The benA, caM, and rpb2 genes, which encode 5 123 β-tubulin, calmodulin, and the second largest subunit of RNA polymerase II, 124 respectively, were then sequenced to further identify strains M1, M2, M3, and M4. The 125 benA, caM, and rpb2 genes have been used for species-level identification of 126 Aspergillus (Samson et al., 2014; Chen et al., 2017). Colony PCR analysis was 127 performed using the following primer sets: Bt2a and Bt2b for benA (Glass and 128 Donaldson, 1995); CMD5 and CMD6 for caM (Hong et al., 2005); 5F and 7CR for rpb2 129 (Liu et al., 1999). The results of sequence analysis were deposited to the DDBJ under 130 the accession numbers LC494250-LC494265.