Monoclonal antibodies (MAbs) against lasalocid and semduramicin were prepared using keyhole limpet hemocyanin conjugates for the immunization of mice. With these MAbs, we developed quantitative enzyme-linked immunosorbent assay (ELISA) methods for lasalocid and semduramicin. The ELISAs were quantitative in the ranges of 0.1-50 ng/mL for lasalocid and 0.05-12.5 ng/mL for semduramicin, and showed 50% inhibition concentrations of 1.2 ng/mL for lasalocid and 0.5 ng/mL for semduramicin. The coefficient of variations (CV%) of lasalocid were 0.3-4.4% for intra-assay and 0.5-5.1% for inter-assay and those of semduramicin were 0.1-4.6% for intra-assay and 0.3-5.2% for inter-assay. The detection limits for lasalocid and semduramicin were 10 ng/g and 5 ng/g in chicken liver and muscle, respectively. Based on the immunochromatographic method, rapid test kits for lasalocid and semduramicin were also developed. With these kits, the detection limits of lasalocid were 50 ng/mL for standard solution and 125 ng/g for chicken muscle, and those of semduramicin were 10 ng/mL for standard solution and 100 ng/g for chicken muscle.

View full abstract

An enrichment procedure and a polymerase chain reaction (PCR) method for the detection of injured Escherichia coli O157 in foods were examined. Freeze-injured E. coli O157 inoculated in boiled spaghetti could be detected in 6-h culture within 12 h by the PCR method. Cells injured by heating in boiled spaghetti and cells injured by chlorine treatment in raw lettuce and carrot did not grow sufficiently to be detected in 6-h culture but were detected in 18-h culture using selective agar media. The injured cells could be also detected in 18-h culture within 24 h by the PCR method. Enrichment at 42°C in trypticase soy broth (TSB) was more effective than that at 42°C in modified EC broth with novobiocin. These results indicated that the usage of enrichment in TSB for 18 h at 42°C in combination with the PCR method is suitable for screening for E. coli O157 in boiled or chlorinated foods, even if the O157 cells are injured.

View full abstract

Extracts of sausage, sauce, cookie, cereal and pasta sauce spiked with milk standard protein at a level of 5-20 ng/mL as sample solutions were analyzed in replicate in 10 laboratories. Coefficients of variation (CVs) of the three ELISA methods using a Milk Protein Casein ELISA Kit (Casein kit), a Milk Protein β-Lactoglobulin ELISA Kit (β-Lactoglobulin kit) and a FASTKIT^TM Milk ELISA Kit (Milk ELISA kit) were mostly below 10%. Mean recoveries of the milk standard protein from the food extracts were over 40% in the three ELISA methods with a few excertions. The recoveries of milk standard protein from the sauce extract in Casein kit were improved by adjusting the extract to neutrality before the Casein kit assay. The recoveries of milk standard protein from cookie, cereal and pasta sauce were improved by the increasing the amount of antibody coated in the Milk ELISA kit.
The detection limits of all the ELISA methods were 1 ng/mL in sample solutions. These results suggested that the notified ELISA methods are reliable and reproducible for the inspection of milk protein levels in extracts of sausage, sauce, cookie, cereal and pasta sauce.

View full abstract

Extracts of sausage, sauce, pasta sauce, fish paste and cereal spiked with wheat standard protein at a level of 5-20 ng/mL as sample solutions were analyzed in replicate in 10 laboratories. Coefficients of variation (CVs) of both ELISA methods using a Wheat Protein ELISA Kit (Gliadin kit) and a FASTKIT^TM Wheat ELISA Kit (Wheat ELISA kit) were mostly below 10%. Mean recoveries of the wheat standard protein from the food extracts were over 40% in the two ELISA methods except those from cereal extract determined using the Wheat ELISA kit. Repeatability relative standard deviations of wheat standard protein in the five food extracts were in the ranges of 16-26.9% and 3.7-36.2% for the Gliadin kit and the Wheat ELISA kit, respectively. Reproducibility relative standard deviations of wheat standard protein in the five food extracts were 21.6-38.5%, 29.7-53.8% for the Gliadin kit and the Wheat ELISA kit, respectively. The recoveries of wheat standard protein from the cereal extract were improved by the increasing the amount of antibody coated on the plate in the Wheat ELISA kit. The detection limits of both ELISA methods were 1 ng/mL in sample solutions. These results suggested that the notified ELISA methods are reliable and reproducible for the inspection of wheat protein levels in extracts of sausage, sauce, pasta sauce, fish paste and cereal.

View full abstract

A sensitive and selective method using liquid chromatography-electrospray mass spectrometry (LC-ESI-MS) for the determination of carbadox metabolites, quinoxaline-2-carboxylic acid (QCA) and desoxycarbadox (Desoxy-CDX), in swine muscle and liver has been developed. The LC separation was performed on a Cadenza CD-C18 column (10 cm×2 mm i.d.) with a gradient system of 0.01% acetic acid-acetonitrile as the mobile phase at a flow rate of 0.2 mL/min. Negative ionization produced the [M-H]^- molecular ion of QCA. On the other hand, the positive mode produced the [M+H]⁺ ion of Desoxy-CDX. The calibration graphs for QCA and Desoxy-CDX were rectilinear from 0.01 to 0.5 ng with selected ion monitoring (SIM). The drugs were extracted with 0.3% metaphosphoric acid-methanol (7 : 3), and the extracts were cleaned up on an Oasis HLB cartridge (60 mg) and by liquid-liquid extraction. The recoveries of QCA and Desoxy-CDX from swine muscle and liver fortified at 2.5 and 5 ng/g were 70.2-86.3%, and the detection limits were 1 ng/g for both drugs.

View full abstract

Spectrophotometrically determined substances in Yucca extract, listed as “Yucca foam extract” in the “List of Existing Food Additives in Japan”, were investigated by GC/MS, TLC and GC. A TLC method using an anisaldehyde color developing reagent similar to that employed in spectrophotometry was established for selective detection of sapogenins in Yucca extract. Several steroidal sapogenins were found by GC/MS in the fractions corresponding to spots on the TLC plate, and these were assumed to have contributed to the color development in spectrophotometry. Sarsasapogenin and smilagenin were the dominant sapogenins. An on-column injection GC method to determine these sapogenins in Yucca extract was also developed. The sum of these two sapogenins in Yucca extract was 0.9%. The total amount of sapogenin estimated by GC was approximately 2%, which was similar to that measured by spectrophotometry.

View full abstract

During 1990 to 2003, the toxicity of the liver in 4,515 specimens of the puffer fish Takifugu rubripes (torafugu) cultured in netcages or on land were investigated by means of mouse bioassay and liquid chromatography-mass spectrometry (LC/MS). Other tissues (skin, muscles, gonads, etc.) were also investigated in some of them. All the livers and other parts examined were found to be non-toxic. The peak corresponding to tetrodotoxin (TTX) was not detected in the samples by LC/MS analysis for TTX (<0.1 MU/g). These results show that puffer fish fed on a non-toxic diet in netcages do not become intoxicated.

View full abstract

A method for the determination of 12 synthetic food dyes (Amaranth, Erythrosine, Allura Red AC, New Coccine, Phloxine, Rose Bengal, Acid Red, Tartrazine, Sunset Yellow FCF, Fast Green FCF, Brilliant Blue FCF, Indigo Carmine) in food was developed using capillary electrophoresis (CE) with photodiode array detection.
The dyes were extracted with water and 0.5% ammonia-ethanol (1 : 1) mixture, and cleaned up using solid-phase extraction (Sep-Pak Plus tC18). The dyes were eluted with methanol from the cartridge. The dyes were separated by CE on a bubble cell fused-silica capillary (72 cm to the detector, 75 μm i.d.) using 20% acetonitrile in a mixture of 10 mmol/L potassium phosphate, monobasic and 5 mmol/L sodium carbonate (pH 10.0) as the running buffer. Identifications of the dyes were performed on the basis of the migration time and the absorbance spectrum of each peak. The coefficients of variation of the migration times and the peak areas were 0.28-0.62% and 1.84-4.30%, respectively (n=5). The identification limits using the absorbance spectra of the dyes were 10 μg/mL for Brilliant Blue FCF and Fast Green FCF, and 5 μg/mL for the other 10 dyes. The recoveries of the 12 dyes from pickles, soft drinks and candies at the level of 10 μg/g were 70.0-101.5%. The method was applied to the analysis of dyes in foods. The dyes detected by CE were in agreement with those detected by paper chromatography.

View full abstract

Evaluation of commercial ELISA kits for the screening of deoxynivalenol (DON) was carried out. Three kinds of commercial kits supplied by different companies were used. Three lots of naturally contaminated wheat and DON-free wheat (<0.05 μg/g) were used as samples. The values obtained from ELISA were compared with those of the HPLC-UV method. In the results obtained from 14 institutions using ELISA, the CV values were less than 17.6% for all the commercial kits. The DON-free sample was not detected as positive with any commercial kits. Also there was no negative finding among the naturally contaminated samples used in this experiment. Coefficients of determination in ELISA and HPLC analysis were 0.979-0.999. These results indicate that ELISA using any of the three commercial kits is efficient for the screening of DON in wheat.

View full abstract

Residues of 14 kinds of β-lactam and macrolide antibiotics in bovine muscles were extracted with acetonitrile and the extract was partitioned with n-hexane to remove fat. High-performance liquid chromatography was carried out on a TSK-gel ODS-80TM column using gradient elution with acetonitrile-0.05% trifluoroacetic acid and the drugs were quantitated by diode array detection. The recoveries of the drugs from bovine muscles spiked at 0.1 ppm were over 63% and each quantitation limit was 0.04 ppm.

View full abstract

A collaborative study involving 8 laboratories was conducted to evaluate a method for the simultaneous determination of pesticide residues in 6 types of fruits and vegetables (spinach, tomato, apple, radish, cabbage and carrot). The method of analysis was the same as reported by Kakimoto et al. in 2003. One handred and thirty-nine pesticides were spiked by each of 8 laboratories at levels of 0.1 μg/g (pesticides analyzed by GC/MS) or 0.5 μg/g (pesticides analyzed by HPLC) into the 6 kinds of samples. Statistical analysis showed that 111 pesticides could be analyzed with practical precision by this method. For screening purposes, the method could analyze 118 pesticides. The median values of the limits of detection were 0.001-0.041 μg/g. The calibration curves were linear in the range of 0.5-5 μg/mL for most pesticides with median correlation coefficients of 0.983-1.000.

View full abstract