Prevalence of cpe-positive Clostridium perfringens in surface-attached soil of commercially available potatoes and its significance as a potential source of food poisoning.

OBJECTIVE: We aimed to examine the surface-attached soil of commercially available potatoes in Japan to determine the association between foodborne infection and the circulation of Clostridium perfringens through vegetables, soil, and environments. METHODS: C. perfringens spores were isolated from 30 surface-attached soil samples of potatoes obtained from six regions in Japan. We performed multiplex polymerase chain reaction (PCR) and sequencing to detect the presence of six toxin and plasmid-related genes in the isolates. RESULTS: Sulfite-reducing clostridial spores were detected in 28 (93%) of 30 potato samples, and toxin gene PCR was performed using 613 isolates. The C. perfringens α toxin gene (cpa) was detected in 288 isolates (288/613; 47%) from 25 potato samples (83%), and these isolates were presumed to be the strains of C. perfringens. The toxin types of C. perfringens were classified into type A, in which 73% of isolates had only cpa, followed by type F in 20%, type C in 6%, and type E in 0.003% (1 isolate). The enterotoxin gene (cpe) related to food poisoning was detected in 64 isolates from 9 potato samples (3%). Of these, 59 isolates had cpa and cpe, whereas five had cpa, C. perfringens ß toxin gene, and cpe. All tested cpe-positive isolates had plasmid-type cpe. CONCLUSIONS: The isolation of culturable cpe-positive C. perfringens from the surface-attached soil of commercially available potatoes indicates that potatoes are a potential source of foodborne transmission of C. perfringens.

[In vivo and In vitro Mitigation Effects of Lactic Acid Bacteria Derived from Fresh Vegetables on Aflatoxins].

Aflatoxins (AFs) are known to be oncogenic mycotoxins. This study investigated the mitigation effects of lactic acid bacteria (LAB) isolated from four types of vegetable, cucumber, Chinese cabbage, Japanese radish and eggplant, which are used to make Japanese traditional fermented pickles, on AFs. Using aflatoxin M1 (AFM1) binding assay for screening, four representative strains were selected (one from each vegetable) from total 94 LAB strains, based on the highest binding ratio. The ranges of the binding ratio of these representative strains to aflatoxin B1 (AFB1), aflatoxin B2, aflatoxin G1, aflatoxin G2 and AFM1 were 57.5%-87.9% for the LAB strain derived from cucumber, 18.9%-43.9% for the LAB strain derived from Chinese cabbage, 26.4%-41.7% for the LAB strain derived from Japanese radish, and 15.0%-42.6% for the LAB strain derived from eggplant. The strains isolated from cucumber, Chinese cabbage, Japanese radish and eggplant were identified as Lactococcus lactis subsp. lactis, Weissella cibaria, Leuconostoc mesenteroides and Leu. mesenteroides, respectively. An in vitro binding assay of the four strains under acidic conditions showed that the number of living bacteria decreased, while the binding ratio increased in some strains, suggesting that the LAB maintained their capacity to bind aflatoxins even in an environment that imitated the stomach. An in vivo experiment using L. lactis subsp. lactis derived from cucumber revealed that the bacteria significantly inhibited the absorption of AFB1 into blood. These results showed that the LAB used for Japanese vegetable pickles was an effective binding agent of AFs and suggested that they might play a role in mitigating AF absorption.

Biology of fowl adenovirus type 1 infection of heterologous cells.

The JM1/1 strain of fowl adenovirus (FAV) serotype 1 isolated from gizzard erosion was used to investigate the biology of FAV in homologous (susceptible) and heterologous cells. The FAV JM1/1 strain is capable of efficient multiplication in primary chicken kidney (CK) cells, but not in Crandell-Rees feline kidney (CRFK) cells or Vero cells. FAV adsorption in heterologous cells was slightly higher than in CK cells. An early gene encoding a DNA-binding protein and a late gene encoding the hexon protein were expressed in CK cells. Only the early gene was expressed in Vero cells. Neither of these genes was expressed in CRFK cells. These results suggest that the virus was unable to multiply effectively due to suppression of viral gene expression in the heterologous cells used in this study.

Preparation of Monoclonal Antibodies Specifically Reacting with the Trichothecene Mycotoxins Nivalenol and 15-Acetylnivalenol via the Introduction of a Linker Molecule into Its C-15 Position.

Nivalenol (NIV) is a trichothecene mycotoxin that is more toxic than deoxynivalenol. It accumulates in grains due to infection with Fusarium species, which are the causative agents of scab or Fusarium head blight. An immunoassay, which is a rapid and easy analytical method, is necessary for monitoring NIV in grains. However, a specific antibody against NIV has not been prepared previously. To establish an immunoassay, we prepared NIV, introduced a linker, and generated antibodies against it. NIV was prepared from a culture of Fusarium kyushuense obtained from pressed barley through chromatographic procedures with synthetic adsorbents and silica gel. NIV was reacted with glutaric anhydride, and the reaction was stopped before mono-hemiglutaryl-NIV was changed to di-hemiglutaryl-NIV. 15-O-Hemiglutaryl-NIV was isolated via preparative HPLC and bound to keyhole limpet hemocyanin (KLH) using the active ester method. Two different monoclonal antibodies were prepared by immunizing mice with the NIV-KLH conjugate. The 50% inhibitory concentration values were 36 and 37 ng/mL. These antibodies also showed high reactivity in a direct competitive enzyme-linked immunosorbent assay and specifically reacted with NIV and 15-acetyl-NIV but not with deoxynivalenol and 4-acetyl-NIV.

Identification of Legionella londiniensis isolated from hot spring water samples in Shizuoka, Japan, and cytotoxicity of isolates.

As part of an epidemiological study on legionellosis, we attempted to isolate Legionella spp. from hot spring water and were able to isolate L. londiniensis HYKF-90505 (=JCM 16338), confirming that L. londiniensis inhabits hot spring water in Japan. To investigate the disease potential of L. londiniensis, we examined its ability to grow intracellularly within Acanthamoeba sp. JAC/E1 strain. The isolated HYKF-90505 was able to grow within Acanthamoeba sp. JAC/E1 strain, and we confirmed also that the HYKF-90505 strain showed cytotoxicity for cultured cells such as J774.1 (JCRB0018). However, in a culture of human U937 cells, the bacterial count was not increased by the intracellular growth of the HYKF-90505 strain. Cells infected for 24 h and stained using the Giménez method showed no intracellular growth of the HYKF-90505 strain. Thus, the isolate appears to be weakly pathogenic to humans.

Growth in Acanthamoeba sp. and antibiotic susceptibility of Legionella micdadei isolated from hot spring water samples.

As part of an epidemiological study on legionellosis, we attempted to isolate Legionella spp. from hot spring water samples, and were able to isolate Legionella micdadei from 3 (5.5%) of 55 samples. All of these isolates were able to grow within Acanthamoeba sp., suggesting that the isolates will be pathogens. We also confirmed that the K-2 strain from hot spring water grew in guinea pig monocytes. Sensitivity tests using 10 drugs showed that the isolates were most sensitive to imipenem, with the MIC90 of 0.032 microg/ml, were least sensitive to minocycline, with the MIC90 of 4 microg/ml, and were not sensitive to low amounts of other drugs.

[Studies on Vibrio vulnificus infection: molecular epidemiology of environment-derived strains and clinical isolates].

To clarify the route and source of Vibrio vulnificus infection, we conducted molecular epidemiological investigation by DNA analysis of 355 environmental isolates (seawater-derived strain: 86, sea mud-derived strain:36, and oyster-derived strain: 233) and 65 human clinical isolates, for a total of 420 isolates, using pulse field gel electrophoresis (PFGE), with the following results. 1. When DNA was cleaved with 2 enzymes, Not I and Sfi I, and subjected to PFGE, Not I DNA interpretation was 76.9%, and Sfi I cleavage was 97.9%, showing that Sfi I was superior in cleaving DNA of this bacteria. 2. Sfi I-interpreted strains were subjected to PFGE and migration patterns were analyzed by UPGMA, but close classification was not possible because similarity was low, this infectious disease clearly originated from multiple rather than a single-clone. In this cluster, we concluded that this infectious disease was acquired through contact between the environment and human beings and viceversa. We identified an assortment of clinical isolates and environment-derived strains among more than 89% of strain groups tested, none of which could be expected to have the same origin. We conclued DNA analysis on these two types of restriction enzymes using PFGE, but were unable to classify test results in detail due to the proliferation of migration patterns and low degree of similarity.

[Drug sensitivity test of Campylobacter jejuni strains isolated from human diarrheal stools, chicken meat, and chicken feces, and gene mutation of quinolone-resistant strains].

In basic studies on campylobacteriosis, we tested 53 strains from human diarrhea stools and 102 strains from chicken meat and feces obtained between 2002 and 2006 for drug sensitivity to different drugs and gene mutation in quinolone-resistant strains. 1) Of 15 drugs tested, all were resistant to one or more of the following 10 drugs: CEX, 99.4%: ABPC, 59.4%; NA, 40.6%; NFLX, 40.0%; TC and CPFX, 39.4%; PIPC, 38.1%; MINO, 30.3%; KM, 3.2%; and SM, 2.6%. 2) Of 155 drug-resistant strains, 28 (18.1%) were resistant to single drugs and 127 (81.9%) were resistant to multiple drugs. The most frequent pattern of multipledrug resistance was ABPC/PIPC/CEX, followed by ABPC/PIPC/CEX/TC/MINO/NA/NFLX/CPFX. 3) Mutation of GyrA (Thr86 --> Ile) was detected in 43 (97.7%) of 44 quinolone-resistant strains. We found that resistance to beta-lactams, quinolones, and tetracycline antibiotics was high, and most resistant strains were resistant to multiple drugs. We also found that most quinolone-resistant strains had GyrA mutation.

In vitro Anti-Influenza Virus Activity of Agaricus brasiliensis KA21.

　Agaricus is known to have immunostimulatory and anti-tumor effects. However, the antiviral effects of Agaricus have not yet been examined. In the present study, the antiviral effects of an extract of Agaricus brasiliensis KA21 (AE) on the H1N1 influenza virus (PR8 strain) were investigated.　The anti-influenza virus effects of AE were examined by using the plaque formation inhibition test. AE inhibited the plaque formation of PR8 in a dose-dependent manner: 98 and 50% (IC50) inhibition at 2.5 and 0.99 mg/mL, respectively. To elucidate the mechanisms of AE, the direct actions and adsorption and invasion inhibition of AE were examined, and were found to have no inhibitory effect on PR8 infection. Thus, in vitro antiviral effects may somehow inhibit PR8 after the viral invasion of cells. These results demonstrated that it is expected that AE can effectively prevent the spread of the influenza virus.

[Drug sensitivity of vero toxin-producing Escherichia coli O157:H7 (VTEC O157:H7) isolated from human diarrhea and healthy cows].

As a part of studies on the source of infection of Vero toxin-producing Escherichia coli (VTEC), O157:H7 strains isolated from human infectious enteritis between 1986 and 1995 and O157:H7 strains isolated from feces of milk cows between 2001 and 2003 were subjected to drug sensitivity test with drugs widely used as therapeutic drugs for various infectious diseases in humans and animals, and the following results were obtained. 1) Drug sensitivity tests with 20 drugs were performed in 52 strains derived human from diarrhea and 100 strains derived from milk cows, and resistance was noted in 115 strains (75.7%): 36 of the 52 human diarrhea-derived strains (69.2%) and 79 of the 100 milk cow-derived strains (79.0%). 2) The human diarrhea-derived strains and milk cow-derived strains were compared with regard to MIC90 of each drug. The antibacterial activity of the drugs was generally higher against the human diarrhea-derived strains than against the milk cow-derived strains. 3) In the 115 strains exhibiting resistance, the most frequent pattern of drug resistance was single drug resistance noted in 80 strains (68.4%), and multidrug resistance was noted in 35 strains (30.4%) consisting of 17 strains with resistance to 3 drugs, 14 strains with resistance to 2 drugs, and 2 strains each with resistance to 4 drugs and 5 drugs. More strains were multidrug-resistant in the milk cow-derived strains.

[Drug susceptibility and analysis using pulsed-field gel electrophoresis of Streptococcus pyogenes strains isolated from the patients with streptococcal toxic shock syndrome (STSS) in Japan].

Previously, we have performed T typing of Streptococcus pyogenes strains isolated from patients with streptococcal toxic shock syndrome (STSS) in Japan, and streptococcal pyrogenic exotoxin (SPE) typing for epidemiological examination. In this study, we conducted a drug sensitivity test using these strains, and investigated the results of gene analysis by pulse-field gel electrophoresis (PFGE) of S. pyogenes strains derived from patients with STSS, the patient's family, and patients other than those with STSS. To clarify the relationship between the host and bacterial factors, we investigated the association between clinical symptoms and T typing of the isolated strains/production of streptococcal pyrogenic exotoxin. There were no strains resistant to beta-lactams, and only 1 strain was resistant to multiple agents other than beta-lactams. The PFGE pattern of T1 type strains was classified into 2 ; the pattern was consistent between the strains derived from patients with STSS and those derived from the patient's family. The PFGE pattern of T3 type strains was classified into 5 (IV) ; Pattern I, which was most frequently observed, was detected in both the strains derived from patients with STSS/non-STSS. However, Patterns II and III were detected only in the strains derived from patients with non-STSS. Patterns IV and V were detected only in the strains derived from patients with STSS. When examining the association between clinical symptoms and bacterial factors, disseminated intravascular coagulation (DIC) was associated with T1-SPE B-producing strains, and pharyngitis was associated with T3-SPE A-producing strains. In the future, the relationship between the host and bacterial factors should be further investigated.

[Basic studies on Vibrio vulnificus infection: isolation of V. vulnificus from sea water, sea mud, and oysters].

To clarify the environmental distribution of Vibrio vulnificus, sea water, sea mud, and oysters were examined at 13 sites, i.e. 4 sites in the Tokyo Bay (eastern Japan) and 9 sites (5 sites for oysters) in Tokushima Prefecture (western Japan). 1. V. vulnificus was isolated from 80 (54.8%) of the 146 samples of sea water examined. It was isolated from 19 (41.3%) of the 46 samples from western Japan and 61 (61.0%) of the 100 samples from eastern Japan. 2. It was isolated from 40 (40.8%) of the 98 samples of sea mud obtained in eastern Japan. 3. It was isolated from 655 (30.3%) of the 2,165 samples of oysters. They were 30 (9.7%) of 309 samples from western Japan and 625 (33.7%) of 1,856 samples from eastern Japan. 4. The density of V. vulnificus was 0.3-1.1 x 10(6) MPN/L in seawater, 0.3-1.1 x 10(5) MPN/100 g in sea mud, and 0.3-1.1 x 10(7) MPN/100 g in oysters. 5. Seasonally, V. vulnificus was isolated from 44 (6.2%) of the 713 samples in spring, 450 (72.6%) of the 620 samples in summer, 264 (51.8%) of the 510 samples in fall, and 17 (3.0%) of the 56 samples in winter. Thus, the isolation rates of V. vulnificus from sea water and oysters tended to be higher in eastern Japan than in western Japan and to be highest in summer, then, in fall.

[A basic study of Vibrio vulnificus infection: serotyping and drug sensitivity test of environment-derived strains and human clinical isolates].

In attempt to elucidate the route and source of Vibrio vulnificus infection. serotyping and drug sensitivity tests of environment-derived strains and human clinical isolates were performed. 1) Serotyping of isolates from the two types of source were determined. Of environment-derived strains, 72.5% were classified into 18 types, and O7 was the most frequent type, accounting for 73.1%, and the second frequent type was O4, accounting for 6.1%. Of human clinical isolates, 87.1% were classified into eight types, and O4 was the most frequent, accounting for 73.5%, and O7 was the secondly most frequent, accounting for 12.9%. 2) Serotypes were investigated by regions. In eastern Japan, 69.2% were classified into 18 types, and O7 and O4 accounting for 44.6% and 5.7%, respectively. In western Japan, 64.8% were classified into eight types, and O7 was the most frequent, accounting for 20.4%, and secondly frequent type was O4, accounting for 11.1%. 3) Regarding the relationship between biotypes and serotypes, environment-derived biotype-I strains were widely distributed in the serotypes, but most biotype-I human clinical isolates were distributed in serotypes O1-O7, showing a difference between the two types of sources. However, many biotype-II strains from the two types of sources included in the serotype O7 group. 4) Drug sensitivity was compared based on MIC90 between strains from the two types of sources. Environment-derived strains were sensitive to ABPC, PIPC, CPZ, CTX, LMOX, MEPM, GM, EM, TC, DOXY, MINO, CP, NA and CPFX, but some strains were resistant to CER, CET, CTX, CMZ, KM and LCM. Human clinical isolates were sensitive to EM, TC, DOXY, MINO, CP, NA and CPFX, but some strains were resistant to ABPC, PIPC, CER, CET, CPZ, CTX, CMZ, LMOX, MEPM, KM, GM, AMK and LCM.

[Isolation and serotyping of Vero toxin-producing Escherichia coli (VTEC) from pig].

As a part of basic studies to elucidate the source of infection of Verotoxin-producing Escherichia coli (VTEC) infectious disease, fresh feces were collected from pigs raised in Kanto District (A and B Prefectures) and Kyushu District (C and D Prefectures) between April and October in 2000, and isolation, serotyping, toxin typing, and drug sensitivity test of VTEC were performed. 1) Of 411 fecal samples tested, VTEC was isolated from 44 samples (10.7%), consisting of 12 of 112 samples (10.7%) from A Prefecture, nine of 100 samples (9.0%) from B Prefecture, 18 of 99 samples (18.2%) from C Prefecture, and five of 100 samples (5.0%) from D Prefecture. 2) Forty-five isolates were serotyped. Four isolates (8.9%) were typed as type 3, but the remaining 41 isolates (91.1%) could not be typed. The four typed isolates consisted of two O112ac:H- isolates and one each of O126:H- and O157:H7. 3) Toxin was typed in 45 isolates. Twenty-seven (60.0%) and 17 isolates (37.8%) produced VT 2 and VT1, respectively, and one isolate (2.2%) produced both VT1 and VT2. 4) Drug sensitivity tests of 45 isolates were performed. All 45 isolates (100%) were multidrug-resistant that were resistant to multiple drugs. Nineteen, nine, four, four, seven, one, and one isolates were resistant to five, six, two, three, four eight, and nine drugs, respectively. The above findings confirmed contamination in all districts, although the VTEC isolation rate varied among the sampling districts. Serotyping clarified the presence of O157:H7 and O112ac:H- that are detected in human VTEC infectious disease. The drug sensitivity tests clarified the presence of many multidrug-resistant strains.

[Isolation and serotypes of Vero toxin-producing Escherichia coli (VTEC) from pigeons and crows].

To clarify the source and route of infection with Vero toxin-producing Escherichia coli (VTEC) in humans, we sampled gastrointestinal contents and isolated VTEC from wild birds captured to exterminate harmful birds between August 1997 and January 1998. Pigeons were caught in Sagamihara-shi and crows were caught in Sagamihara-shi, Kawasaki-shi, Yokohama-shi, and the Tokyo metropolitan area. The following results were obtained. 1) VTEC was isolated from 32 of 521 birds (6.1%) examined. Among pigeons, VTEC was isolated from 25 of 262 birds (9.5%) captured in Sagamihara-shi. Among crows, VTEC was isolated from 7 of 184 birds (3.8%) captured in Sagamihara-shi, but not isolated from any bird of 11.4, and 60 birds captured in Yokohama-shi, Kawasaki-shi, and the Tokyo metropolitan area, respectively. 2) Toxin was typed in 33 isolates. There were four VT1-producing isolates (6.5%), 27 VT2-producing isolates (88.7%), and two VT1, VT2-producing isolates (4.8%). 3) The serotypes of the isolates were: O78: H-, 10; O152: H-, 7; O153: H19.2; O164: H-, 1; O128: H-, 1; O164/143: H-, and O1: HUT, 1. The serotype was unknown in 10 isolates. Among 10 isolates for which the serotype could not be determined, auto-aggregation was observed in one isolate. 4) EaeA was investigated in the 33 isolates, and 31 isolates (93.9%) possessed eaeA. The above findings showed that strains with same toxin types and serotypes of human diarrhea-derived VTEC were isolated from pigeons and crows, and the isolates frequently possessed eaeA, which is considered to have an important association with its pathology, suggesting that birds are involved in VTEC infection in humans as a source of infection.

Powder infant formula milk contaminated with Enterobacter sakazakii.

To clarify the route and source of Enterobacter sakazakii infection in a basic study, we analyzed powder infant formula milk (PIF), which may be an important source of infantile infection, regarding contamination with Enterobacteriaceae including this type of bacteria, and conducted drug sensitivity tests with various antimicrobial agents. Enterobacteriaceae was isolated 36 (24.2%) of 149 PIF samples. These comprised of 12 (19.7%) of 61 domestically produced samples and 24 (27.3%) of 88 imported samples. E. sakazakii was isolated in 9 (6.6%) of the 149 PIF samples. These comprised 4 (6.6%) of 61 domestically produced samples and 5 (5.7%) of 88 imported samples. In 8 of the 9 samples in which E. sakazakii was isolated, the bacterial levels were estimated to be 0.36 MPN/100 g. However, one imported sample showed a bacterial level of 0.91 MPN/100 g. In the drug sensitivity tests of E. sakazakii isolated from PIF, we compared the MIC(90) values. E. sakazakii was highly sensitive to 9 agents: cefotaxime, ceftriaxone, cefoperazone, ceftazidime, cefpirome, cefozopran, gentamicin, meropenem, and ciprofloxacin, and moderately sensitive to 5 agents: piperacillin, erythromycin, minocycline, chloramphenicol, and rifampicin. However, it was resistant to 2 agents, ampicillin and lincomycin.