比較傳統液相層析與極致液相層析搭配質譜質譜儀分析鮮乳中之黃麴毒素M1

Abstract

動物飼料遭受黃麴菌的汙染，可能會導致飼料中含有黃麴毒素。當牛隻食入遭受黃麴毒素B1污染的飼料後，會在乳汁中產生肝毒性及致癌性極高的黃麴毒素M1 ( AFM1 )。AFM1在乳品的處理過程不易受到破壞，因此世界各國目前仍嚴格把關乳品中AFM1的含量。歐盟自1999年起，將AFM1在鮮乳中容許量由 0.5 μg/L 降至 0.05 μg/L，對國內現有的檢測方法形成挑戰。 我國藥物食品檢驗局現階段公告鮮乳中AFM1的檢測方法，是利用免疫親和性管柱 ( immunoaffinity column, IAC ) 進行樣品的前處理，再搭配液相層析儀並以螢光偵測器檢測。本研究欲改善樣本前處理以縮短樣本前處理時間，並提升樣品偵測靈敏度而做了以下探討 (1)使用RAM pre-column與Shodex column進行樣本的前處理以縮短實驗時間 (2)分別比較BEH HILIC極致液相層析管柱、HSS T3極致液相層析管柱、BEH C18 極致液相層析管柱與傳統的C18液相層析管柱等四種管柱搭配極致液相層析/串聯式質譜儀的分析效能。並根據偵測靈敏度、層析時間、偵測極限、基質效應，從極致液相層析管柱中選擇一支最佳者。 本研究樣本前處理結果顯示：管柱材質為C4的限制接觸材質管柱 ( RAM pre-column ) 無法將AFM1滯留於管柱內，導致AFM1隨鮮乳中蛋白質等大分子物質一併排除至廢液，因此C4的RAM pre-column不適合做為鮮乳中AFM1的前處理管柱；以聚合材質層析管柱 ( Shodex column )為樣本前處理與分析管柱，雖可以有效的分離大分子蛋白質與AFM1，縮短前處理與分析時間，但AFM1波峰過寬 ( ＞3 分鐘 )影響解析度。因此本研究仍然使用免疫親和性管柱作為樣本前處理的方法。 本實驗使用免疫親和性管柱進行鮮乳樣本前處理，使用HSS T3極致液相層析管柱搭配極致液相層析/串聯式質譜儀偵測AFM1的優點為：(1)基質干擾較其它極致液相層析管柱小 ( 63.4% )；(2)較傳統C18液相層析管柱可獲得較低的儀器偵測極限：0.011 ng/mL ( 0.11 pg )；(3)更低的粉狀乳檢出極限 2.01 ng/kg ( 優於我國藥物食品檢驗局於93年公佈粉狀乳的檢出極限 20 ng/kg )。雖然HSS T3 極致液相層析管柱已為所測試三支極致液相層析管柱中基質干擾較低者，但其訊號抑制 ( ion suppression ) 仍高達63.4%。因此本實驗配置了AFM1基質添加檢量線，發現其線性結果良好，R2可達0.997，且使用基質添加檢量線即可有效減低分析時樣本基質干擾的問題，可以提供更準確的樣本定量結果。另外，黃麴毒素G1 ( AFG1 )會對AFM1的定量子離子造成干擾，不適合作為本研究的內標。 本研究結果顯示使用極致液相層析管柱優於傳統液相層析管柱的主要特點為層析速度較快 ( AFM1於分析管柱滯留時間由 4.6分鐘縮短為2.6分鐘 )，較低的儀器偵測極限 ( 儀器偵測極限由4.7 pg降低至 0.11 pg ) 與提升訊號雜訊比 ( 訊號雜訊比提高16−58倍 )。

Feedstuff are subject to the pollution of Aspergillus, which is able to produce aflatoxin B1 ( AFB1 ). After injestion by lacting cow, AFB1 can be transfered into aflatoxin M1 ( AFM1 ) and is secreted in milk, which is carcinogenic and toxic to liver. AFM1 has been classified as a group 1 carcinogen by the International Agency for Research on Cancer ( IARC ) since 2002. Unfortunately, AFM1 is relatively stable either in preparation and storage of milk products. The European Union has reduced the AFM1 maximum allowance from 0. 5 μg/L to 0.05 μg/L since 1999, and this has become a challenge for most of the existing analytical methods. The official analytical method of Bureau of Food and Drug Analysis, Taiwan, processes milk samples with immunoaffinity column ( IAC ) and detects AFM1 using high-performance liquid chromatography coupled with a fluorescence detecter ( HPLC/FLD ). The aims of this study were to modify pre-treatment method in order to shorten analytical time and improve detection limits. There were two strategies in our study. The first one was to simplify the milk sample preparation using restricted access materials ( RAM ) pre-column and shodex column. The second one was to increase the throughput and sensitivity with ultra performance liquid chromatography ( UPLC ). Three UPLC columns were compared with a traditional HPLC C18 column. RAM pre-column did not retain the AFM1 in milk, and the proteins in milk were co-eluted with AFM1. Thus, RAM pre-column was not suitable for the sample preparation. Shordex column can separate the proteins in milk with AFM1 but the peak width of AFM1 is too broad ( >3 min ). Consequently, we still used IAC columns for sample preparation. AFM1 was analyzed using an Acquity UPLC system and tandem mass spectrometry with positive electrospray ionization. The HSS T3 UPLC column gave the best performance on selectivity and sensitivity. The on-column detection limit of AFM1 was 0.011 ng/mL ( 0.11 pg ) of AFM1 and the method detection limit of AFM1 in milk powder was 2.01 ng/kg. The ion suppression was 63.4% for milk； therefore, matrix-matched standard curves were set up to eliminate the matrix effect. A good linearity was obtained with the square of correlation coefficient larger than 0.997. Aflatoxin G1 is not a good internal standard , which influenced the quantification of AFM1. Use of UPLC lowed the detection limit ( 0.11 pg ), shortened the run time ( 2.6 min ), and the signal-to-noise ratios increased 16−58 times than the HPLC column because of their narrow peaks.