比較三種樣本前處理方式並以極致液相層析/串聯式質譜儀檢測血清中全氟碳化合物、類雌激素物質及真菌毒素

Abstract

全氟碳化合物 (PFASs) 與類雌激素物質，包括鄰苯二甲酸酯類 (Phthalates) 及雙酚A等，被廣泛應用於各式消費性產品中。真菌毒素為一群對人體具多重毒性之黴菌二級代謝物，於食物栽種或保存條件不當時，由滋生之真菌產生。這些物質由於具持久性、生物累積性、內分泌干擾性質及致癌性而受到關注。上述物質廣泛流佈於環境中，對人類造成重複且經常性的暴露，因此，藉由生物偵測了解這些物質在一般大眾體內累積量為一值得探討的議題。然而，迄今能於血清中同時檢測上述污染物之分析方法仍相當缺乏。另一方面，在液相層析串聯式質譜的定量分析中，基質效應一直以來都被視為一個棘手的問題，因此，一個適當的樣品前處理方法對於清除基質中的干擾物，進而提高分析效能是相當重要的。本研究開發並比較三種不同的樣品前處理技術，包括：固相萃取、支持性液液萃取，及液液萃取搭配固相萃取(產品名稱:Ostro,後續將以產品名代稱此法)，搭配高效能液相層析串聯式質譜儀，以多重離子監測模式獲取質荷比資訊，來檢測血清中十種全氟碳化合物、五種鄰苯二甲酸酯類代謝物、四種真菌毒素及雙酚A代謝物，並以同位素稀釋技術進行定量。黃麴毒素M1與黃麴毒素B1以正離子電灑游離法作為游離源，並利用ACE Excel 2 C18-PFP HPLC管柱於液相層析儀中搭配移動相：(A) 5mM醋酸氨水溶液 (pH = 6.54)、(B)甲醇，進行梯度流析；十種全氟碳化合物、五種鄰苯二甲酸酯類代謝物、雙酚Ａ葡萄糖苷酸、赭麴毒素Ａ及橘黴素以負離子模式之電灑游離法作為游離源，並利用Waters CORTECS UPLC C18管柱，於液相層析儀中搭配移動相： (A)乙腈：水 = 5:95 (v/v), 0.04%乙酸、(B)乙腈：水 = 90:10 (v/v), 0.04%乙酸，進行梯度流析。 Sirocco蛋白質沉澱盤用於固相萃取前移除血清中之蛋白質，相較於傳統的有機溶劑/離心方法較為省時。固相萃取法最佳化的過程中比較：(1)兩種調節溶劑；(2)五種強度之清洗溶劑；(3)兩種回溶溶劑，以獲取最終處理流程。支持性液液萃取的部分，在測試兩種稀釋緩衝溶劑與五種萃取溶劑後得到最佳流程。Ostro則測試兩種不同衝擊溶液(crash solvent)以得出最佳結果。固相萃取、支持性液液萃取及Ostro三種方法的偵測極限範圍分別為：0.09–4.67 ng/mL、0.19–2.74 ng/mL 以及0.12–2.58 ng/mL。三種方法的基質效應因子如下：固相萃取98–179%、支持性液液萃取67–122%、以及Ostro 81–180%。固相萃取、支持性液液萃取及Ostro三法之萃取效率範圍分別是：3.3–67%、36–95%及26–82%。定量偏差及相對標準偏差多數均低於20%。 本研究使用臺大醫院新竹分院提供之119位受試者剩餘血清檢體各350微升(IRB編號：104-007-F)，應用新開發之三種前處理技術批次分析。其中，8碳以上之全氟碳化合物具高檢出率(>90%)；而6碳以下之全氟碳化合物僅於少數樣本中偵測到低濃度。鄰苯二甲酸酯代謝物及雙酚A葡萄糖苷酸在部分樣本中被測得，其中鄰苯二甲酸單乙酯、鄰苯二甲酸單丁酯及雙酚A葡萄糖苷酸之檢出率落在10–20%，較本研究中其他塑化劑代謝物來的高。而真菌毒素在本研究中檢出率相當低；赭麴毒素A在三中方法中具有相對高的檢出率(4.1–8.1%)。 本研究評估新開發之三種前處理方法之基質效應因子、樣本前處理萃取效率、操作通量、方法靈敏度、方法穩定度及價格等因子。總體而言，Ostro能達到同時淨化基質並維持良好的萃取效率，且較其他兩方法來得快速、簡單、穩定及便宜，為最具成本效益的方法，適用於血液類樣品(全血、血漿、血清)之多重殘留分析。

Perfluoroalkyl substances (PFASs) and feminizing compounds including phthalates and bisphenol A (BPA) are widely used in many consumer products. Mycotoxins are a diverse group of secondary fungal metabolites, which possess multiple toxicities to humans. They are concerned because of the persistence, bioaccumulation, interfering with hormone functions, and carcinogenicity. All these compounds are ubiquitous in food and the environment. The general population is exposed to these contaminants continuously; hence, it is crucial to investigate the body burdens of the above compounds in humans by biomonitoring. However, limited methods are available to determine these contaminants together in serum. On the other hand, matrix effect is critical in quantitative liquid chromatography tandem mass spectrometry (LC-MS/MS) analysis. An appropriate sample preparation that can remove unwanted matrix components is essential to solve this problem. Therefore, this study developed and compared three sample preparation methods, which are solid-phase extraction (SPE), supported liquid-liquid extraction (SLE), and liquid-liquid extraction coupled with solid-phase extraction (LLE+SPE, i.e., the product name: Ostro) to simultaneously determine ten PFASs, five phthalate metabolites, BPA glucuronide (BPA-G) and four mycotoxins in serum using ultra-performance liquid chromatography/tandem mass spectrometry (UPLC-MS/MS) with multiple reaction monitoring (MRM) and quantified by isotope dilution techniques. Aflatoxin M1 and aflatoxin B1 were ionized by positive electrospray ionization mode (ESI+) and were separated on liquid chromatography by an Advanced Chromatography Technologies (ACE) Excel 2 C18-PFP HPLC column with the mobile phases composed of (A) 5-mM ammonium acetate(aq) (pH = 6.54) and (B) methanol; ten PFASs, five phthalate metabolites, BPA-G, ochratoxin A and citrinin were ionized by negative electrospray ionization mode (ESI-) and were separated on a Waters CORTECS UPLC C18 column with mobile phases composed of (A) 0.04% acetic acid in acetonitrile/water = 5:95 (v/v) (pH = 3.45) and (B) 0.04% acetic acid in acetonitrile/water = 90:10 (v/v) (pH = 3.91). A Sirocco protein precipitation plate was used to remove the protein in serum before the extracting step of SPE, in which was relatively quick comparing with traditional protein precipitation approach. The optimization of SPE included the tests of two condition solvent, five strengths of wash solvent, and two reconstitution solvents. The optimization of SLE involved the tests of two different buffers and five extraction solvents. Two crash solvent were tested for optimizing Ostro steps. The LODs of SPE, SLE and Ostro ranged from 0.09–4.67 ng/mL, 0.19–2.74 ng/mL and 0.12–2.58 ng/mL, respectively. The matrix effect factors of three assays were as follows: 98–179% of SPE, 67–122% of SLE, and 81–180% of Ostro. The extraction efficiencies of SPE, SLE and Ostro ranged from 3.3–67%, 36–95% and 26–82%, respectively. Most of the quantitative bias and relative standard deviations were below 20%. This study applied the three assays to analyze serum samples from 119 human subjects, each with 350 μL of serum. Most long-chain PFASs containing eight or more carbons were highly detected in the samples (>90%); short-chain PFASs containing six or fewer carbons were detected in few samples at low concentrations. Phthalate metabolites and BPA-G were detected in a few samples; monoethyl phthalate (MEP), monobutyl phthalate (MBP) and BPA-G were detected in about 10% to 20% of samples, which were higher than other phthalate metabolites in this study. Four mycotoxins were barely found in samples except for ochratoxin A, which was the most frequently detected among the four with the detection rates from 4.1% to 8.1% in the three assays. Matrix effect, extraction efficiency, throughput, detection sensitivity, robustness, and the cost were considered for evaluating the three methods. In overall, Ostro both provided clean extracts and good extraction recoveries on most of the analytes. It was faster, less complex, more robust, and cheaper than SPE and SLE. Thus, Ostro was the most cost-effective one among the three, which would be suitable for multiple-residue analysis of serum samples.