以超高效液相層析搭配飛行時間串聯式質譜儀篩查食物中之有機污染物

Abstract

化學物質的應用隨著工業發展及因應日常所需日漸增加，人體可能藉由食物的攝取而暴露多種化學品，進而產生危害。標的性分析方法針對預先設定的物質進行定量，因此無從得知未知化學物質的暴露。另一方面，透過非標的性分析能夠在不事先設定待測物之下，篩檢食物中常見及未知之污染物以協助釐清食物中潛在之化學性危害。 本研究開發適合非標的性分析之前處理方法，並以超高效能液相層析搭配四極桿-飛行時間串聯式質譜儀(UHPLC-QTOF MS)對彰化縣大二林地區及連江縣南竿鎮市場之23項食材(7種蔬菜、7種海鮮、4種肉類、蛋、奶、米、麵粉、黃豆)各進行三次食物中化學污染物殘留之篩查。同一種食物樣本各由一家傳統市場及一家生鮮超市採買，採買時間分別為彰化2019/5/10、2019/8/2、2019/10/25及馬祖2019/12/25、2020/3/16、2020/4/27。食物樣本經均質後進行QuEChERS萃取，以PRiME HLB cartridge淨化後上機分析。樣本資料以全離子MS/MS模式採集訊號，後端比對資料庫以鑑定樣本中所殘留之化學物質種類。最後，針對篩查出之化學物質，以其於樣本中之篩查陽性率及訊號強度，並連結暴露情形、生物活性及每日容許攝取量(ADI)做為參考依據，建置優先化學品清單。 樣本淨化步驟之優化結果顯示，食物混樣以EMR-lipid或PRiME HLB淨化後，分子特徵數量相似。然而，PRiME HLB對色素有較好的去除能力，並且操作簡易、所需時間較短(包含萃取步驟，以PRiME HLB淨化一個樣本僅需25分鐘)，其較EMR-lipid節省兩倍時間，故選擇PRiME HLB。綜合三次採樣之分析結果，彰化及馬祖之多類食物樣本分別篩查出188及175個化學物質。蔬菜類主要關注未經批准使用之農藥篩查結果。Empenthrin為一種環境衛生用藥，於台灣並未批准用於農業用途，其篩查陽性率於彰化及馬祖之葉菜類樣本皆大於90%。海鮮類及肉類中鑑定出之化學物質以荷爾蒙類物質及人類治病藥物為大宗。Desogestrel為人類口服避孕藥成分，於61%之海鮮樣本中篩查出來；mesterolone為治療男性不孕症之藥物，僅於彰化之海鮮(31.4%)及肉類(54.5%)樣本中篩查出來。Allopurinol為治療痛風之藥物，與其代謝物oxipurinol於肉類及海鮮類樣本中篩查率皆大於50%。蒽醌類物質(emodin、phscion、chrysophanol)於所有黃豆樣本中檢出；牛奶樣本相對有較多塑化劑殘留。 檢出物質之優先化排序整合化學品之實驗數據(篩查陽性率、訊號強度)，並依據外部資料(生物活性、暴露資料、ADI)之可得性分為兩組；A組為具有至少一個外部資料之化學品，B組則為沒有外部資料之化學品。4-tert-octylphenol及塑化劑(dibutyl phthalate、bis(2-ethylhexyl) phthalate、di(2-ethylhexyl) adipate、dioctyl phthalate)於各類食物中(葉菜類、穀類/根莖類、海鮮類及禽畜類)之A組化學品中皆位在排序前十名；vidarabine為抗皰疹病毒之藥物，則於各類食物之B組化學品排名前三。Empenthrin (環境衛生用藥)及hexylresorcinol (食品抗氧化劑)於葉菜類及穀類/根莖類中A組化學品排名前五。在海鮮類及禽畜類中，A組化學品之abietic acid (黏著劑)及B組化學品中之mesterolone皆排名第一。 本研究利用經優化之前處理方法進行食物中化學污染物之非標的性篩查，釐清一般族群可能經常由食物暴露之化學性危害項目，並且進行食物中潛在關注化學物質清單之排序。此排序結果可以提供未來食物中關注化學物質定量及健康風險評估的優先化參考依據。

The use of chemicals is increasing in industry, agriculture, and consumer products, which may enter the food chain and result in adverse health effects. Targeted analysis only quantifies pre-selecting chemicals; thus, there are no clues on the exposure of unknown chemicals. In contrast, non-targeted analysis can identify common and unknown contaminants without information prior to chemicals and helps to clarify potential chemical hazards in foods. This study developed an assay for screening organic contaminants in food using a universal sample preparation and analyzing sample residues with ultra-high performance liquid chromatography/quadrupole time-of-flight mass spectrometer (UHPLC-QTOF MS). Twenty-three items of food (including seven vegetables, seven seafood, four meats, egg, milk, rice, flour, and soybean) were collected three times during May to October in 2019 at Changhua and December 2019 to April 2020 at Matsu from traditional markets and supermarkets, respectively. The homogenized samples were extracted with QuEChERS and cleaned up with PRiME HLB, then were analyzed at both positive and negative electrospray ionization. Data were acquired at All Ions MS/MS mode, and mass spectra were matched libraries for chemical identification. The identified chemicals were prioritized using chemical detection frequency, peak abundance, exposure information, bioactivity, and acceptable daily intake; priority chemical lists were established. The results of optimization on cleanup during sample preparation showed that composite food samples after the cleanup with EMR-lipid and PRiME HLB obtained comparable numbers of molecular features. However, PRiME HLB offered better cleanup performance on removing pigments as well as higher throughput that only required a half of time as EMR-lipid (the extraction and the cleanup using PRiME HLB of a sample only cost 25 minutes). A total of 188 and 175 compounds were identified in three batches of multiple food samples in Changhua and Matsu, respectively. Residues of non-approved pesticides were the major concern in the vegetable samples. Empenthrin, an environmental agent that is not allowable to be used for agricultural purposes in Taiwan, was observed in more than 90% of the vegetable samples from either Changhua or Matsu. A number of hormone-related compounds and human medicines were identified in seafood and poultry/livestock samples. Desogestrel, a contraceptive of humans, was detected in 61% of the seafood samples; mesterolone, a medicine for treating male infertility, was only detected in the seafood (31.4%) and the poultry/livestock (54.5%) samples from Changhua. The anti-gout medicine allopurinol and its metabolite oxipurinol were found in more than 50% of the seafood and the poultry/livestock samples. Fugal anthraquinones, such as emodin, physcion, and chrysophanol, were found in all the soybean samples. More plasticizer residues were observed in the milk samples. The empirical data (i.e., detection frequency, peak abundance) and external information (i.e., exposure data, bioactivity, and acceptable daily intake (ADI)) of the identified chemicals were taken into consideration for prioritization. The identified chemicals were classified into two groups based on the availability of external information; group A were chemicals with at least one available items of external information; group B were chemicals without any available external information. 4-tert-octylphenol and plasticizers (dibutyl phthalate, bis(2-ethylhexyl) phthalate, di(2-ethylhexyl) adipate, dioctyl phthalate) were positioned within top 10 of group A chemicals in all the food categories (leafy vegetables, grains/root crops, seafood, poultry/livestock). Vidarabine, an anti-virus drug, ranked at the top 3 of group B chemicals in all the food categories. Empenthrin (an environmental agent) and hexylresorcinol (an antioxidant of food) were both located at the top 5 in group A chemicals in leafy vegetables and grains/root crops. Abietic acid (an adhesive) and mesterolone were both ranked at the 1st places in group A and group B chemicals, respectively, in the seafood and poultry/livestock. The optimized non-targeted screening method was applied on identifying contaminants in 23 types of food to clarify the common chemical hazards that humans are often exposed to from food ingestion. This study provided lists of prioritized chemicals for further quantification and risk assessment.