Toxigenic Aspergillus and Penicillium Isolates from Weevil-Damaged Chestnuts

Aspergillus and Penicillium were among the most common genera of fungi isolated on malt-salt agar from weevil-damaged Chinese chestnut kernels (16.8 and 40.7% occurrence, respectively). Chloroform extracts of 21 of 50 Aspergillus isolates and 18 of 50 representative Penicillium isolates, grown for 4 weeks at 21.1 C on artificial medium, were toxic to day-old cockerels. Tweleve of the toxic Aspergillus isolates were identified as A. wentii, eight as A. flavus, and one as A. flavus var. columnaris. Nine of the toxic Penicillium isolates were identified as P. terrestre, three as P. steckii, two each as P. citrinum and P. funiculosum, and one each as P. herquei (Series) and P. roqueforti (Series). Acute diarrhea was associated with the toxicity of A. wentii and muscular tremors with the toxicity of P. terrestre, one isolate of P. steckii, and one of P. funiculosum.     

Toxigenic species of Penicillium, Fusarium, and Aspergillus from weevil-damaged pecans.

Of a total of 2392 fungi isolated from weevil-damaged pecans, 46.4% were Alternaria and Epicoccum, 23.9% were Penicillium, 12.4% were Pestalotia and Monochaeta, 6.5% were Cladosporium, 6.4% were Fusarium, and less than 2% each were Phoma, Aspergillus, Rhizopus, Trichothecium, and miscellaneous. Chloroform extracts of 34 of 105 representative Penicillium isolates, 3 of 28 Fusarium isolates, and 3 of 23 Aspergillus isolates were toxic to day-old cockerels during three bioassays. Eight of the toxic extracts from Penicillium spp. were tremorgenic. One tremorgenic isolate was identified as P. paxilli, four were identified as P. lanoso-coeruleum, and three as P. cyclopium. Nine of the non-tremorgenic isolates were identified as P. citrinum, five as P. aurantio-virens, three as P. oxalicum, and two as P. meleagrinum. Others were identified as one each of P. brevi-compactum (Series), P. nigricans (Series), P. roqueforti, P. rugulosum (Series), P. terrestre, and P. stoloniferium. One was unidentified. Toxigenic Aspergillus isolates were all A. flavus. Two of the toxic Fusarium isolates were F. moniliforme, and one was unidentified.     

Toxigenic Fungi in Food

Forty-five fungal isolates from moldy supermarket foods were tested for toxicity to brine shrimp, and twenty-two of these isolates were subsequently tested for toxicity to chicken embryos. Highly toxigenic fungi were Cladosporium sphaerospermum from a bakery product, Fusarium oxysporum from carrots, F. solani from cabbage, Aspergillus niger and Penicillium corylophilum from bread, P. cyclopium and P. herguei from corn meal, P. lanosum from onions,P. steckii from chocolate syrup, Penicillium sp. from jelly, and Rhizopus nigricans isolates from sweet potato, applesauce, and strawberries. Approximately one-third of the fungal cultures were moderately to highly toxigenic to brine shrimp and chicken embryos, while several additional cultures were slightly toxigenic.     

Toxicity to Chicks of Aspergillus and Penicillium Species Isolated from Moldy Pecans

Isolates of Aspergillus chevalieri, A. flavus, A. ochraceus, A. repens, and Penicillium funiculosum and complexes of P. citrinum-P. implicatum isolated from moldy pecan meats were toxic to chicks.     

Toxin production by Fusarium species from sugar beets and natural occurrence of zearalenone in beets and beet fibers.

Fifty-five Fusarium isolates belonging to nine species were collected from fungus-invaded tissue of stored sugar beets and identified as F. acuminatum (11 isolates), F. avenaceum (1 isolate), F. culmorum (1 isolate), F. equiseti (23 isolates), F. graminearum (4 isolates), F. oxysporum (1 isolate), F. solani (4 isolates), F. sporotrichioides (7 isolates), and F. subglutinans (2 isolates). All isolates were cultured on autoclaved rice grains and assayed for toxicity by feeding weanling female rats the ground-rice cultures of the isolates in a 50% mixture with a regular diet for 5 days. Fifty-eight percent of the isolates were acutely toxic to rats, 26% caused hematuria, 18% caused hemorrhages, and 29% caused uterine enlargement. In most cases, toxicity could not be accounted for by the known toxins found. The following mycotoxins were found in extracts of the rice cultures: zearalenone (22 to 6,282 micrograms/g), chlamydosporol (HM-8) (68 to 4,708 micrograms/g), moniliformin (45 to 400 micrograms/g), deoxynivalenol (10 to 34 micrograms/g), 15-acetyldeoxynivalenol (5 to 10 micrograms/g), diacetoxyscirpenol (22 to 63 micrograms/g), monoacetoxyscirpenol (21 to 26 micrograms/g), scirpenetriol (24 micrograms/g), T-2 toxin (4 to 425 micrograms/g), HT-2 toxin (2 to 284 micrograms/g), neosolaniol (2 to 250 micrograms/g), and T-2 tetraol (4 to 12 micrograms/g). F. equiseti was the predominant species found on visibly molded beets in the field. Six of 25 moldy sugar beet root samples collected in the field contained zearalenone in concentrations ranging between 12 and 391 ng/g, whereas 10 samples from commercial stockpiles were negative for zearalenone.(ABSTRACT TRUNCATED AT 250 WORDS)     

Toxin production by Fusarium species from sugar beets and natural occurrence of zearalenone in beets and beet fibers.

Fifty-five Fusarium isolates belonging to nine species were collected from fungus-invaded tissue of stored sugar beets and identified as F. acuminatum (11 isolates), F. avenaceum (1 isolate), F. culmorum (1 isolate), F. equiseti (23 isolates), F. graminearum (4 isolates), F. oxysporum (1 isolate), F. solani (4 isolates), F. sporotrichioides (7 isolates), and F. subglutinans (2 isolates). All isolates were cultured on autoclaved rice grains and assayed for toxicity by feeding weanling female rats the ground-rice cultures of the isolates in a 50% mixture with a regular diet for 5 days. Fifty-eight percent of the isolates were acutely toxic to rats, 26% caused hematuria, 18% caused hemorrhages, and 29% caused uterine enlargement. In most cases, toxicity could not be accounted for by the known toxins found. The following mycotoxins were found in extracts of the rice cultures: zearalenone (22 to 6,282 micrograms/g), chlamydosporol (HM-8) (68 to 4,708 micrograms/g), moniliformin (45 to 400 micrograms/g), deoxynivalenol (10 to 34 micrograms/g), 15-acetyldeoxynivalenol (5 to 10 micrograms/g), diacetoxyscirpenol (22 to 63 micrograms/g), monoacetoxyscirpenol (21 to 26 micrograms/g), scirpenetriol (24 micrograms/g), T-2 toxin (4 to 425 micrograms/g), HT-2 toxin (2 to 284 micrograms/g), neosolaniol (2 to 250 micrograms/g), and T-2 tetraol (4 to 12 micrograms/g). F. equiseti was the predominant species found on visibly molded beets in the field. Six of 25 moldy sugar beet root samples collected in the field contained zearalenone in concentrations ranging between 12 and 391 ng/g, whereas 10 samples from commercial stockpiles were negative for zearalenone.(ABSTRACT TRUNCATED AT 250 WORDS)     

PCR-RFLP of ITS rDNA for the rapid identification of Penicillium subgenus Biverticillium species

RFLP of ITS rDNA is proposed as a useful tool for molecular identification of the most common species of biverticillate penicillia. 60 isolates were analysed representing 13 species and 21 unique sequences were produced. The combination of five restriction enzymes was successful in separating 12 species. However, the variety Penicillium purpurogenum var. rubrisclerotium remained indistinguishable from Penicillium funiculosum. P. funiculosum appeared as the most confused species, being mis-identified with Penicillium miniolutum and Penicillium pinophilum, which were originally part of the species, and with P. purpurogenum perhaps because of the common production of red pigment. Penicillium variabile was difficult to investigate as introns were found on half of the isolates. Penicillium piceum, Penicillium rugulosum, Penicillium loliense, Penicillium erythromellis and P. purpurogenum were homogeneous from molecular and morphological positions and corresponded to a well circumscribed taxon. Furthermore, intraspecific variability was evidenced within P. pinophilum and P. funiculosum. The ex-type isolate of P. funiculosum produced a unique pattern. The method is sensitive, rapid and inexpensive and can be used for isolate identification of the biverticillate species. It is recommended particularly when many isolates have to be authentificated prior to analysis for phylogenetic assessment or population genetics.     

Mouse Toxicity of Fungi of Tobacco

A bioassay for fungal toxins based on the intraperitoneal injection of test materials into mice was used to screen 976 cultures isolated from tobacco and grown in a high-protein baby cereal and also to determine whether samples of tobacco damaged by fungi are more toxic than samples of apparently sound tobacco. Of 236 fungal isolates from noncured tobacco, 79% were lethal when homogenized cultures of these isolates were tested. Forty-nine per cent of 740 fungi isolated from cured tobacco were lethal. Of the genera from which 30 or more isolates were tested, Epicoccum, Alternaria, and Penicillium had the highest percentage of toxic isolates from non-cured tobacco, whereas Epicoccum, Aspergillus, and Alternaria had the highest percentage from cured tobacco. Samples of tobacco naturally infected with brown spot, caused by Alternaria tenuis, did not have a significantly different LD(50) value after 48 hr than comparable disease-free samples. However, animals which died from doses near the LD(50) dose of tobacco infected with Alternaria generally died in 24 to 48 hr with signs associated with a depressant rather than a stimulant, such as nicotine, which caused death in 15 to 30 min. These signs were duplicated by injecting homogenized pure cultures of Alternaria. These studies, although inconclusive with regard to the effects of fungal contaminants on the quality or usability of tobacco, have developed evidence that suggests the advisability of a study on smoke or smoke condensates from moldy and nonmoldy tobacco.     

Toxins of molds from decaying tomato fruit.

Among 27 mold isolates from decaying tomatoes, culture filtrates or ethyl acetate extracts of 8 isolates grown in yeast extract-sucrose medium were markedly toxic (mortality, greater than 50%) to brine shrimp larvae. The toxicity of six of these isolates could be attributed to the presence of citrinin, tenuazonic acid, or T-2 toxin. Ethyl acetate extracts of five Alternaria isolates and one Fusarium isolate were mutagenic for Salmonella typhimurium strains. In ripe tomatoes inoculated with toxin-producing isolates and incubated at 25 degrees C, one Alternaria alternata isolate produced tenuazonic acid in seven of seven tomatoes at levels of up to 106 micrograms/g and alternariol methyl ether in one of the seven tomatoes at 0.8 microgram/g. Another A. alternata isolate produced tenuazonic acid or alternariol methyl ether at much lower levels in only three of seven tomatoes. Patulin and citrinin were produced by a Penicillium expansum isolate at levels of up to 8.4 and 0.76 microgram/g, respectively. In tomatoes incubated at 15 degrees C, a Fusarium sulphureum isolate produced T-2 toxin, HT-2 toxin, and neosolaniol at levels of up to 37.5, 37.8 and 5.6 micrograms/g, respectively. If these mycotoxins are thermostable, they may occur at detectable levels in tomato products whenever partially moldy tomatoes are used as raw material.     

Toxins of molds from decaying tomato fruit.

Among 27 mold isolates from decaying tomatoes, culture filtrates or ethyl acetate extracts of 8 isolates grown in yeast extract-sucrose medium were markedly toxic (mortality, greater than 50%) to brine shrimp larvae. The toxicity of six of these isolates could be attributed to the presence of citrinin, tenuazonic acid, or T-2 toxin. Ethyl acetate extracts of five Alternaria isolates and one Fusarium isolate were mutagenic for Salmonella typhimurium strains. In ripe tomatoes inoculated with toxin-producing isolates and incubated at 25 degrees C, one Alternaria alternata isolate produced tenuazonic acid in seven of seven tomatoes at levels of up to 106 micrograms/g and alternariol methyl ether in one of the seven tomatoes at 0.8 microgram/g. Another A. alternata isolate produced tenuazonic acid or alternariol methyl ether at much lower levels in only three of seven tomatoes. Patulin and citrinin were produced by a Penicillium expansum isolate at levels of up to 8.4 and 0.76 microgram/g, respectively. In tomatoes incubated at 15 degrees C, a Fusarium sulphureum isolate produced T-2 toxin, HT-2 toxin, and neosolaniol at levels of up to 37.5, 37.8 and 5.6 micrograms/g, respectively. If these mycotoxins are thermostable, they may occur at detectable levels in tomato products whenever partially moldy tomatoes are used as raw material.     

Toxicity to Spodoptera exigua and Trichoplusia ni of individual P1 protoxins and sporulated cultures of Bacillus thuringiensis subsp. kurstaki HD-1 and NRD-12.

The toxicities to neonate Spodoptera exigua and Trichoplusia ni of lyophilized powders obtained from sporulated liquid cultures (referred to as sporulated cultures) and Escherichia coli-expressed P1 [cryIA(a) cryIA(b) cryIA(c)] protoxins from three-gene strains of NRD-12 and HD-1 of Bacillus thuringiensis subsp. kurstaki were determined by using diet incorporation bioassays. Although sporulated cultures from both strains were more toxic to T. ni than S. exigua, there were no differences in toxicity between NRD-12 and HD-1. Toxicities of the three individual P1 protoxins against S. exigua varied by at least fivefold, with the cryIA(b) protein being the most toxic. These same protoxins varied in toxicity against T. ni by at least 16-fold, with the cryIA(c) protein being the most toxic. However, when tested against either S. exigua or T. ni, there were no differences in toxicity between an NRD-12 P1 protoxin and the corresponding HD-1 P1 protoxin. Comparing the toxicities of individual protoxins with that of sporulated cultures demonstrates that no individual protoxin was as toxic to S. exigua as the sporulated cultures. However, this same comparison against T. ni shows that both the cryIA(b) and cryIA(c) proteins are at least as toxic as the sporulated cultures. Results from this study suggest that NRD-12 is not more toxic to S. exigua than HD-1, that different protein types have variable host activity, and that other B. thuringiensis components are not required for T. ni toxicity but that other components such as spores might be required for S. exigua toxicity.     

Toxicity to Spodoptera exigua and Trichoplusia ni of individual P1 protoxins and sporulated cultures of Bacillus thuringiensis subsp. kurstaki HD-1 and NRD-12

The toxicities to neonate Spodoptera exigua and Trichoplusia ni of lyophilized powders obtained from sporulated liquid cultures (referred to as sporulated cultures) and Escherichia coli-expressed P1 [cryIA(a) cryIA(b) cryIA(c)] protoxins from three-gene strains of NRD-12 and HD-1 of Bacillus thuringiensis subsp. kurstaki were determined by using diet incorporation bioassays. Although sporulated cultures from both strains were more toxic to T. ni than S. exigua, there were no differences in toxicity between NRD-12 and HD-1. Toxicities of the three individual P1 protoxins against S. exigua varied by at least fivefold, with the cryIA(b) protein being the most toxic. These same protoxins varied in toxicity against T. ni by at least 16-fold, with the cryIA(c) protein being the most toxic. However, when tested against either S. exigua or T. ni, there were no differences in toxicity between an NRD-12 P1 protoxin and the corresponding HD-1 P1 protoxin. Comparing the toxicities of individual protoxins with that of sporulated cultures demonstrates that no individual protoxin was as toxic to S. exigua as the sporulated cultures. However, this same comparison against T. ni shows that both the cryIA(b) and cryIA(c) proteins are at least as toxic as the sporulated cultures. Results from this study suggest that NRD-12 is not more toxic to S. exigua than HD-1, that different protein types have variable host activity, and that other B. thuringiensis components are not required for T. ni toxicity but that other components such as spores might be required for S. exigua toxicity.     

Biological control of apple blue mold with Pseudomonas fluorescens

Pseudomonas fluorescens isolate 1100-6 was evaluated as a potential biological control agent for apple blue mold caused by Penicillium expansum or Penicillium solitum. Both the wild-type isolate 1100-6 and a genetically modified derivative labeled with the gene encoding the green fluorescent protein (GFP) were compared. The P. fluorescens isolates with or without GFP equally reduced the growth of Penicillium spp. and produced large zones of inhibition in dual culture plate assays. Cell-free metabolites produced by the bacterial antagonists reduced the colony area of Penicillium isolates by 17.3% to 78.5%. The effect of iron chelate on the antagonistic potential of P. fluorescens was also studied. The use of iron chelate did not have a major effect on the antagonistic activity of P. fluorescens. With or without GFP, P. fluorescens significantly reduced the severity and incidence of apple decay by 2 P. expansum isolates after 11 d at 20 degrees C and by P. expansum and P. solitum after 25 d at 5 degrees C when the biocontrol agents were applied in wounds 24 or 48 h before challenging with Penicillium spp. Populations of P. fluorescens labeled with the GFP were determined 1, 9, 14, and 20 d after inoculation at 5 degrees C. The log CFU/mL per wound increased from 6.95 at the time of inoculation to 9.12 CFU/mL (P < 0.05) 25 d after inoculation at 5 degrees C. The GFP strain did not appear to penetrate deeply into wounds based on digital photographs taken with an inverted fluorescence microscope. These results indicate that P. fluorescens isolate 1100-6 could be an important new biological control for apple blue mold.     

Antagonistic activity of Penicillium oxalicum Corrie and Thom,Penicillium decumbens Thom and Trichoderma harzianum Rifai isolates against fungi,bacteria and insects in vitro

The antibiotic activity of 70 isolates belonging to the genera Aspergillus, Penicillium, Fusarium, Alternaria and Trichoderma was tested as preliminary screening. The highest activity was obtained with three Penicillium oxalicum isolates, one Penicillium decumbens isolate and the Trichoderma harzianum isolate. After that, we chose these five isolates in order to carry out other studies with bacteria, fungi and insects. Extracts from these isolates were obtained. The extracts were tested for antibiotic activity with positive results, which implies that metabolite production is involved in this antagonistic effect. The highest activity was shown by T. harzianum and P. oxalicum extracts, but there was high variability among P. oxalicum isolates.     

Distribution, Frequency, and Diversity of Bacillus thuringiensis in an Animal Feed Mill

Bacillus thuringiensis was isolated from 36 of 50 residue samples obtained from an animal feed mill (a stored-product environment). Of 710 selected colonies having Bacillus cereus-B. thuringiensis morphology isolated from the samples, 477 were classified as B. thuringiensis because of production of parasporal delta-endotoxin crystals. There was a diverse population of B. thuringiensis, as revealed by differentiation of the isolates into 36 subgroups by using (i) their spectra of toxicity to the lepidopterans Heliothis virescens, Pieris brassicae, and Spodoptera littoralis and the dipteran Aedes aegypti and (ii) their parasporal crystal morphology. A total of 55% of the isolates were not toxic to any of these insects at the concentrations used in the bioassays; 40% of all isolates were toxic to one or more of the Lepidoptera; and 20, 1, and 1% of the isolates were toxic to only P. brassicae, H. virescens, and S. littoralis, respectively. The most frequent toxicity was toxicity to P. brassicae (36% of all isolates); 18% of the isolates were toxic to A. aegypti (5% exclusively), 10% were toxic to H. virescens, and 4% were toxic to S. littoralis. Toxicity to P. brassicae was more often linked with toxicity to H. virescens than with toxicity to S. littoralis. The frequency of toxicity was significantly greater in isolates that produced bipyramidal crystals than in isolates that produced irregular pointed, irregular spherical, rectangular, or spherical crystals.     

Investigation of Penicillium chrysogenum Isolates for their Suitability as Starter Cultures

One hundred twenty three isolates ofP. chrysogenum were biologically tested in brine shrimp test and screened withStaphylococcus aureus for the detection of antibacterial activity. Furthermore, they were chemically examined (thin layer chromatographic method, TLC) for the synthesis of 8 mycotoxins (citrinin, cyclopiazonic acid, mycophenolic acid, patulin, penicillic acid, PR-toxin, ochratoxin A, and roquefortine). The results indicated that 85% of the tested isolates produce roquefortine and one isolate produces cyclopiazonic acid. Considering the results of the chemical assay for mycotoxins as well as the results of the brine shrimp test and the detection of antibacterial activity, 119 (97%) of the tested isolates are not suitable to be used as starter cultures for mould-fermented meats. The extracts of only 4 isolates were subjected to further biological tests in mice and the results indicated that only one isolate was non-toxinogenic.     

Iodine sensitivity of bacteria isolated from iodinated water systems

Fourteen bacterial isolates, predominantly Pseudomonas sp., from two water systems disinfected by iodinated anion-exchange resins were studied and compared with an isolate of Pseudomonas aeruginosa from a povidone-iodine solution and four other isolates. Pseudomonas cepacia and P. aeruginosa grown in brain heart infusion were 3 to 5 logs less sensitive to 1 mg/L I2 (pH 7.2, 1 min) when compared with cultures grown in phosphate buffer. Another P. cepacia isolate was the least sensitive culture when grown in brain heart infusion (1 log decrease) but was more sensitive after cultivation in phosphate buffer (5 logs). Isolates from an iodinated potable water system, including P. cepacia, Staphyloccus warneri, and a Bacillus sp., were all less sensitive to iodine than a "resistant" P. aeruginosa and three other isolates when grown in brain heart infusion. A clinical isolate of P. aeruginosa exhibited intermediate sensitivity. The sensitivity of bacteria to iodine is thus highly variable, depending on the organism as well as the growth conditions.     

Protein-Lipopolysaccharide Complexes Produced by Virulent and Non-Virulent Isolates of Verticillium albo-atrum and V. dahliae are Non-Specific Toxins to Tomato and Potato

Toxic protein-lipopolysaccharide complexes (PLPC) were isolated from culture filtrates of Verticillium albo-atrum (isolate WCS 800) and V. dahliae (WCS 070) isolates, both virulent to tomato and potato cultivars, and from an isolate of,V. albo-atrum (V22W) which was non-virulent to these hosts. The virulent isolates each produced one major PLPC in culture and the non-virulent isolate produced two (fractions 1 and 2, characterized by their elution pattern after gel filtration). Virulence in these three isolates was not related to quantity of PLPC produced in culture. However, PLPC from the virulent isolates were toxic to tomato in a leaf bioassay at 4μg ml^-1 (WCS 800) and 20μg ml^-1 (WCS 070) but the two PL, PC from the non-virulent isolate required concentrations of 100 and 1000 μg ml^-1 for toxicity. Production of a modified, less toxic PLPC in V22W may partly account for its non-virulence. Gel filtration of PLPC from the three isolates on a calibrated Sephacryl S-400 column, eluted in phosphate buffer plus NaCl, indicated a compound of heterogeneous molecular mass with an average of 126,000 daltons for the PLPC of WCS 800, WCS 070 and, fraction 1 of V22W, and 25,000 daltons for fraction 2 of V22W. Attempts to extract a low molecular weight toxin from the PLPC by extended dialysis were unsuccessful. The susceptibility of 12 tomato and 19 potato cultivars to the Verticillium isolates was compared with their sensitivity to PLPC. Susceptibility was not correlated with toxin sensitivity and the PLPC were concluded to be non-specific toxins in these hosts.     

Organization of the gene cluster for biosynthesis of penicillin in Penicillium nalgiovense and antibiotic production in cured dry sausages

Several fungal isolates obtained from two cured meat products from Spain were identified as Penicillium nalgiovense by their morphological features and by DNA fingerprinting. All P. nalgiovense isolates showed antibiotic activity in agar diffusion assays, and their penicillin production in liquid complex medium ranged from 6 to 38 microgram. ml-1. We constructed a restriction map of the penicillin gene cluster of P. nalgiovense and found that the organization of the penicillin biosynthetic genes (pcbAB, pcbC, and penDE) is the same as in Penicillium chrysogenum and Aspergillus nidulans. The pcbAB gene is located in an orientation opposite that of the pcbC and penDE genes in all three species. Significant amounts of penicillin were found in situ in the casing and the outer layer of salami meat during early stages of the curing process, coinciding with fungal colonization, but no penicillin was detected in the cured salami. The antibiotic produced in situ was sensitive to penicillinase.     

Biological, Immunological, and Genetic Analysis of Bacillus thuringiensis Isolated from Granary in Korea

To isolate a naturally occurring novel Bacillus thuringiensis strain, we investigated the distribution, toxicity, morphology, H serotype, and gene type of B. thuringiensis from residue samples of granary in Korea. A total of 163 B. thuringiensis isolates out of 411 samples producing spore and crystal were obtained. In toxicity tests, 80% of all isolates were toxic to lepidoptera, and 12% were not toxic to any of tested insects. And dipteran-active and lepidopteran/dipteran-active isolates were rare (2% and 6%, respectively). 152 B. thuringiensis isolates produced typical rhomboidal crystals, and the remainder produced parasporal inclusions with various morphologies. Serological test showed that B. thuringiensis isolates in granary represented 12 H serotypes, indicating varied distribution of B. thuringiensis. Of these, the serotype 3ab predominated, followed by the serotype 7 and 4ac. B. thuringiensis isolates of the serotype 3ab, 4ac, 5ab, 7, 8ab, 9, and 23 were toxic to lepidoptera, and the serotype 8bd, 12, 18, and 20ac were nontoxic, while 14 isolates were untypable by 33 B. thuringiensis H antisera. The frequency of toxicity against lepidoptera and diptera was primarily highly toxic. PCR analysis using cryI gene type-specific primers showed that cryIA(b) genes are frequently found and cryIE gene exists in only one isolate. Analysis of B. thuringiensis crystals and plasmid DNAs indicated a diversity of crystal and gene types. Received: 15 January 1998 / Accepted: 18 February 1998