目的 构建包含9个属15个种15株菌株的食品中常见产毒霉菌基质辅助激光解吸电离飞行时间质谱(matrix-assisted laser desorption ionization-time of flight mass spectrometry,MALDI-TOF MS)鉴定数据库,为标准 GB 4789.16-2016《食品卫生微生物学检验常见产毒霉菌的鉴定》中产毒霉菌的鉴定提供补充手段,提高其时效性和准确性.同时探究不同固体培养基、前处理方法、培养时间对15株常见产毒霉菌MALDI-TOF MS鉴定结果的影响.方法 2022年3月-2023年8月,将15株产毒霉菌经沙氏液体培养基28℃培养24 h,利用得到的蛋白谱图建立其数据库.通过建立的数据库比较不同固体培养基、培养时间、前处理方法对其鉴定结果的影响.结果 以圆弧青霉为例,其在沙氏葡萄糖琼脂培养基(SDA)、查氏培养基(CA)、马铃薯葡萄糖琼脂培养基(PDA)、孟加拉红培养基(RBC)不同固体培养基的鉴定分值分别为鉴定错误、1.555±0.007、1.875±0.022和2.138±0.048,不同固体培养基对鉴定分值差异有统计学意义(F=88.436,P＜0.05);采用直接涂抹法、甲酸提取法、甲酸-乙腈提取法3种不同前处理方法的鉴定分值分别为1.813±0.008、2.159±0.094和2.138±0.048,不同前处理方法对鉴定分值差异有统计学意义(F=25.852,P＜0.05);在3～7天不同培养时间的鉴定分值分别为2.315±0.037、2.224±0.016、2.138±0.048、2.171±0.016和2.166±0.010,不同培养时间对鉴定分值差异有统计学意义(F=22.893,P＜0.05).对于其余14株常见产毒霉菌,不同固体培养基、培养时间、前处理方法对其MALDI-TOF MS鉴定分值差异均有统计学意义(均P＜0.05).结论 由于细交链孢菌与暗孢节菱孢不能选用PDA作为其鉴定用固体培养基,最终以常见产毒霉菌自建数据库为鉴定平台,采用马铃薯葡萄糖琼脂培养基于28℃培养4～5天,采用甲酸-乙腈法提取蛋白质进行质谱鉴定,以分值≥1.80的高置信度作为评判标准,可使15株中86.7％的产毒霉菌得到有效鉴定.
Establishment of common toxigenic moulds in food identification database by matrix-assisted laser desorption ionization time-of-flight mass spectrometry
Objective To establish common toxigenic moulds in food identification database including 9 genus,15 species,and 15 strains in sum by a matrix-assisted laser desorption ionization time-of-flight mass spectrometry(MALD1-TOF MS),aiming to provide supplementary means for the identification of toxigenic moulds in the standard GB 4789.16-2016"Microbiological identification of common toxigenic molds in food"and improves the efficiency and accuracy.Simultaneously explore the effects of different culture conditions(media,time)and pretreatment methods on the MALDI-TOF MS identification results of 15 common toxigenic moulds.Methods From March 2022 to August 2023,15 strains of toxigenic mouldswere cultured in sabouraud liquid media at 28 ℃ for 24 hours were assigned for database construction,and their protein spectrum were used to establish their database.Compare the effects of different solid culture media,culture time,and pretreatment methods on the identification results through the established database.Results Taking penicillium purpureum as an example,the identification scores of different solid culture media,including sabouraud dextrose agar(SDA),chalet agar media(CA),potato dextrose agar(PDA)and Bengal red medium(RBC)were identification errors,1.555±0.007,1.875±0.022,and 2.138±0.048,respectively.There was a statistically significant difference in identification scores between different solid culture media(F=88.436,P＜0.05).The identification scores of three different pretreatment methods,namely direct application method,formic acid extraction method,and formic acid acetonitrile extraction method were 1.813±0.008,2.159±0.094,and 2.138±0.048 respectively.The differences in identification scores between different pretreatment methods were statistically significant(F=25.852,P＜0.05).The identification scores for different cultivation times from 3 to 7 days were 2.315±0.037,2.224±0.016,2.138±0.048,2.171±0.016 and 2.166±0.010,respectively.There was a statistically significant difference in identification scores for different cultivation times(F=22.893,P＜0.05).For the remaining 14 common toxigenic moulds,there were statistically significant differences in MALD1-TOF MS identification scores among different solid culture media,culture time,and pretreatment methods(all P＜0.05).Conclusion Due to the inability to use PDA as the solid culture medium for the identification of Streptomyces and Fusarium oxysporum,a self built database of common toxigenic moulds was ultimately used as the identification platform.Potato glucose agar was cultured at 28 ℃ for 4-5 days,and protein was extracted using formic acid acetonitrile method for mass spectrometry identification.A high confidence score of ≥ 1.80 was used as the evaluation criterion,which effectively identified 86.7％of the toxigenic moulds in 15 strains.
Matrix-assisted laser desorption ionization time-of-flight mass spectrometryToxigenicmouldsEstablishment identification databaseInfluence factors
NETL
NSTL
万方数据
