不同加熱處理及還原乳添加對牛乳蛋白質的影響

Abstract

本研究之目的為探討不同加熱處理及還原乳添加對牛乳蛋白質之影響，並且建立指紋圖譜，作為鮮乳造假攙偽之檢驗方法。實驗以生乳、高溫短時殺菌乳 (high-temperature short-time, HTST)、超高溫殺菌乳 (ultra-high temperature, UHT) 及還原乳為樣品，藉由二維膠體電泳 (two-dimensional gel electrophoresis, 2D) 分析牛乳中的蛋白質，再利用2D電腦影像軟體ImageMaster 2D Platinum 7.0分析比較，生乳、HTST乳、UHT乳與還原乳組間蛋白質點表現之差異。本研究進一步針對上述的蛋白質點以液相層析串聯質譜儀 (liquid chromatograph tandem mass spectrometry, LC/MS/MS) 進行蛋白質序列分析鑑定，並製作生乳、HTST乳、UHT乳混攙不同比例的還原乳樣品 (0%、2%、5%、10%、20%、30%、40%、50%)，根據預測迴歸模型的結果顯示，各組之決定係數分別為1.000、0.703及0.922，另外各組之偵測極限 (limit of detection, LOD) 分別為0.0029%、17.2598%及13.8385%，定量極限 (limit of quantification, LOQ) 分別為0.0089%、52.3024%及41.9348%。最後，以驗證試驗 (verification test) 測試預測迴歸模型的準確性，製作生乳、HTST乳、UHT乳混攙4%及8%的還原乳樣品，結果顯示，生乳混攙8%還原乳的樣品可被準確預測生乳內實際攙還原乳的含量，HTST乳樣及UHT乳樣的預測結果與定量極限相符合，需在高攙偽比例下才能準確預測。本研究中所建立之預測迴歸模型方可當作鑑定鮮乳造假攙偽之新穎方法。

The objective of this study was to investigate the effects of different heat treatments and addition of reconstituted milk on milk proteins. Furthermore, fingerprints of milk proteins were constructed to identify adulteration of milk powder in fresh milk. First, proteins from raw milk, high-temperature short-time milk (HTST milk), ultra-high temperature milk (UHT milk), and reconstituted milk were separated by two-dimensional gel electrophoresis (2D) and their expression levels were analyzed by using ImageMaster 2D Platinum 7.0. These protein spots were then identified by liquid chromatograph tandem mass spectrometry (LC/MS/MS). Moreover, raw milk, HTST milk, and UHT milk were respectively adulterated with different percentages of reconstituted milk (0%, 2%, 5%, 10%, 20%, 30%, 40%, and 50%). The results of estimated regression models demonstrated that the coefficients of determination in three groups were 1.000, 0.703, and 0.922, respectively. In addition, the limits of detection (LODs) were 0.0029%, 17.2598%, and 13.8385%. The limits of quantification (LOQs) were 0.0089%, 52.3024%, and 41.9348%. The effectiveness of estimated regression models was further confirmed by the verification test. Raw milk, HTST milk, and UHT milk were respectively adulterated with 4% and 8% reconstituted milk. The result of prediction from raw milk sample was in good agreement with real percentage of adulteration (8%). The data from HTST milk sample and UHT milk sample matched the results of LOQs; therefore, only if high level of reconstituted milk is respectively added in HTST milk and UHT milk, models will accurately predict the percentage of adulteration. The estimated regression models constructed in this study may be used as a novel method for identification of adulteration in fresh milk.