Mycologia. 93(4), 2001, pp. 689-703. © 2001 by The Mycological Society of America, Lawrence, KS 66044-8897 Aspergillus bombycis, a new aflatoxigenic species and genetic variation in its sibling species, A. nomius* Stephen W. Peterson1 Key Words: aflatoxin, fungi, molecular systemat­ Microbial Properties Research Unit, National Center ics, ribosomal DNA sequence for Agricultural Utilization Research, Agricultural Research Service, V. S. Department ofAgriculture, 1815 N. University St., Peoria, Illinois 61604-3999 USA INTRODUCTION Yoko Ito Aflatoxin has been the subject of many studies be­ National Research Institute for Vegetables, Ornamental cause of its deadly toxicity to certain domesticated Plants and Tea, Ministry ofAgriculture forestry and animals, including turkeys, ducks and trout (Hessel­ Fisheries, Ano, Mie, 514-2392Japan tine et al 1966). In addition increased incidence of Bruce W. Horn human hepato-carcinoma is associated with ingestion National Peanut Research Laboratory, Agricultural of sublethal doses of aflatoxin (Scholl and Groopman Research Service, V. S. Department ofAgriculture, 1995). Because of aflatoxin's effects on animal and 1011 Forrester Dr. SE, Dawson, Georgia 31742 USA human health, it is essential to determine which spe­ Tetsuhisa Goto cies produce this toxin, as well as details about their National Food Research Institute, Ministry of life histories and distribution in nature. Currently, Agriculture forestry and Fisheries, Kannondai, there are three known aflatoxigenic species from As­ Tsukuba, 305-8642Japan pergillus section Flavi: A. flavus, A. parasiticus, and A. nomius (Cotty et al 1994). An additional species, A. ochraceoroseus (section Circumdati) has recently Abstract: A new aflatoxigenic species of Aspergillus, been reported as aflatoxigenic (Klich et al 1998). As­ A. bombycis, was discovered during isolation of fungi pergillus ochraceoroseus is phylogenetically part of the from insect frass collected in silkworm rearing houses Aspergillus versicolor group (Peterson unpubl). in Japan. The new species resembles A. flavus, but During a study of the incidence of aflatoxigenic produces Band G aflatoxins. It is distinguished from fungal isolates from soil (Peterson et al 2000), we A. flavus and A. nomius by differences in growth encountered some isolates that produced aflatoxins rates at 37 and 42 C, from A. nomius by roughness and whose phylogenetic position (based on ribosom­ of the stipe, and from both of these species by dif­ al DA sequences) among the other species of As­ ferences in the nucleotide sequences in the beta-tu­ pergillus section Flavi suggested that they might be bulin, calmodulin, norsolorinic acid reductase, ITS, undescribed species. One of these unusual fungal and lsu-rD A genes. Aspergillus bombycis is known types (Goto et al 1996) has recently been described from nine isolates, eight collected in silkworm-rear­ as the new aflatoxigenic species Aspergillus pseudota­ ing houses in Japan and one collected in a silk-worm marii Ito et al on the basis of its unique morphology rearing house in Indonesia. Phylogenetic analysis of and phylogenetic distinction from the other species the DNA sequences shows that A. bombycis is a phy­ of section Flavi (Ito et al 2001). logenetically distinct species which is most closely re­ We also isolated fungi from the dust and insect lated to A. nomius and which belongs in Aspergillus frass found in silk worm-rearing houses in eastern section Flavi. Analysis by partition homogeneity did Asia. Some of these isolates have been identified as not reveal evidence of genetic recombination in A. A. nomius, whereas others resemble A. nomius but bombycis, but in A. nomius the patterns of polyrnor­ differ genetically and morphologically from the typ­ phisms in different genes strongly suggest cryptic ge­ ical A. nomius strain. Species from the A. flavus clade netic recombination. can have overlapping character states, making iden­ tification of some isolates by phenotype problemati­ Accepted for publication February 3, 2001. cal (Klich and Pitt 1988, Horn et al 1996), and sev­ I Corresponding author, E-mail: [email protected] eral species are not identifiable phenotypically, such * Names are necessary to report factually on available data; however, as the cryptic species found (but not named) by Geis­ the USDA neither guarantees nor warrants the standard of the prod­ uct, and the use of the name by USDA implies no approval of the er et al (1998). In order to determine the phyloge­ product to the exclusion of others that may also be suitable. netic placement of these isolates and whether they 689 690 MYCOLOGIA might represent a new species, we have sequenced region) were amplified by means of polymerase chain re­ portions of the large subunit rD A and ITS regions action (PCR). Conditions and buffers were those of White as well as the genes for beta-tubulin, calmodulin and et al (1990) except the primers used were ITS1 (White et norsolorinic acid reductase. In addition, morpholog­ al 1990) and D2R (Peterson et al 2000) and the thermal ical and physiological comparisons of these new iso­ profile (96 C, 30 s; 51 C, 45 s; 72 C, 120 s) was repeated 30 lates were made with other species in Aspergillus sec­ times, followed by 7 min at 72 C. The amplified fragment tion Flavi. was purified using GeneClean according to manufacturer's instructions, eluted in 1/10th strength TE and stored at -20 C. Part of the beta-tubulin gene (referred to as BT) MATERIALS AD METHODS was amplified using primers BenO (5'-ATGCGTGA­ GATTGTATGTI) or BenOb (5'-ATGCGTGAGATIGTATG) Isolation data, permanent accession numbers, and prove­ that are identical to the first exon of the gene from A. fla­ nance for the fungi used in this study are listed in TABLE 1. vus (GenBank M38265) as well as several bases in the first These fungi are permanently preserved in the Agricultural intron, and Ben2 (5'-ATCTGGAAACCCTGGAGGC) which Research Service Culture Collection (NRRL) , Peoria, Illi­ nois. is complementary to part of the gene sequence in exon 6. The thermal profile for beta-tubulin amplification was 96 C Isolation offungi.-Fungi were isolated by dispersing 1 g of for 2 min followed by 30 cycles (96 C, 20 s; 51 C, 45 s; 72 substrate (insect frass, dust, etc.) in 100 mL sterile 0.01 % C, 2 min) then 72 C for 5 min. Part of the norsolorinic acid Tween 80 and mixing 0.03-1.0 mL of the substrate dilution reductase gene (referred to as NOR) was amplified using with 15 mL of isolation medium, then pouring the molten the primers and conditions developed by Geisen (1996). A (ca 60 C) agar into a 90-mm petri plate. The isolation me­ portion of the calmodulin gene (referred to as CAL) was dium contained (per liter): 45 g malt extract (Difco); 30 g amplified using the primers and conditions described by NaCI; 30 mg chloramphenicol; 30 mg rose bengal; and 1 Feibelman et al (1998). mg DDVP (Diclorvos) (King et al 1979). Petri plates were incubated in darkness at 27 C for 1-6 d and checked daily DNA sequencing and analysis.-D A sequencing reactions for colony development. Individual colonies from isolation were carried out using Taq polymerase, ABI fluorescent dye plates were subcultured on Czapek agar (Cz) (Raper and labeled dideoxy nucleotides, and DA template and oligo­ Fennell 1965) or other suitable agar media. nucleotide primers. For the ID templates, primers ITS1, Growth and examination of cultures.-For temperature tol­ ITS2, ITS3, ITS4 (White et al 1990) and D1, D1R, D2, D2R erance and morphological examination studies, the isolates (Peterson et al 2000) were used for sequencing. For CAL were inoculated at three points on 90-mm petri plates (Pitt and OR, the primers used for amplification were also used 1979) containing 30 ml of Blakeslee's malt agar (MA). Fun­ for sequencing. For BT, sequences were determined using gal material was viewed with a Zeiss microscope equipped BenO, BenOb, Ben2, br (5'CCAGAAAGCGGCACC) and bf with phase and differential interference contrast. For SEM, (5'-GAGCCCGGTACCATGGA) . nincorporated dye was blocks of agar and fungal material (ca 5 mm X 5 mm) were removed from sequencing reactions by spun-eolumn chro­ cut from a petri plate culture, fixed overnight in 1% os­ matography (Maniatis et al 1982) over Pharmacia ultra-fine mium tetroxide, dehydrated in increasingly concentrated sephadex G-50. The column eluate was dried in a rotary ethanol baths, critical point dried, and sputter coated with vacuum drier and dissolved in 1.2 J.LL ultrapure formamide, gold-paladium (Peterson 1992). For D TA extraction, 100 with or without 20 mg/mL high molecular weight blue dex­ mL of malt extract (ME) broth (Raper and Fennell 1965) tran. The nucleotide sequences were determined by elec­ was autoclaved for 15 min in a 500-mL cotton stoppered trophoresis on an Applied Biosystems 377 DA sequencer. flask. An agar slant culture was flooded with 2-3 mL of 0.1 % The sequence of each fragment was determined from the sterile Triton X-100, spores were dislodged with a wire loop, base sequences of both DA strands. and the spore suspension was pipetted into the malt broth. Sequences were aligned using ClustalW (Thompson et al Flasks were incubated at 25 C on a rotary shaker (200 rpm) 1994) and alignments were checked visually using an ASCII for 1-2 d until 1-2 g biomass had accumulated. editor. D TA analysis was performed using PAUP* ver. DNA extraction and purification.-Biomass was separated 4.004a (Swofford 1998) for parsimony, partition homoge­ from the medium by vacuum filtration over cheesecloth or neity test and bootstrap analysis. Trees were redrawn from Whatman number 54 filter disks, and mycelium was frac­ PA p* tree files using TREEVIEW (Page 1996). tured by vortex mixing with glass beads using the method of Peterson et al (2000).