Bacillus sp.於市售豆漿中之生長預測模型

豆漿為亞洲地區傳統飲料之一。根據國家膳食資料庫2017年的營養調查顯示，19-65歲消費者每日豆漿的平均攝食量為50 g。豆漿含豐富之醣類及蛋白質，因此若保存條件不當，容易造成微生物的生長。但是透過各種培養條件來控制細菌的生長，並且可藉由病原體建模軟體（IPMP 2013）利用數學模式描述微生物的生長行為，以預測食物的保存期限。本研究目的對從市售豆漿分離出之Bacillus sp.進行生長條件測試。將分離之Bacillus sp.培養於各種溫度條件下，並將生長速率用於預測生長模型。總生菌數檢驗結果顯示，採集自早餐店之58件豆漿樣品中，6件包裝豆漿的生菌數皆小於10 cfu/mL；而52件散裝豆漿樣品中，則有15件的總生菌數是大於104 cfu/mL。進一步根據菌落顏色與型態為篩選原則，分離出最頻繁出現在散裝豆漿中的菌株，並以菌株之16S rDNA片段進行序列分析，結果可能為Bacillus sp.。將純化出之菌株進行不同溫度下（7、15、25、30及35 °C）下之細胞生長試驗，觀察菌數變化，再分別利用IPMP軟體中的Huang、Baranyi、修飾之Gompertz、與Buchanan三段線性模式，建立該菌株於豆漿中生長之預測模型，根據決定係數（r2）與均方根誤差（RMSE）之結果判定，Huang模式對Bacillus sp.在35 °C與30 °C豆漿中的生長情形有較好的擬合程度；而Buchanan三段線性模式對Bacillus sp.於25 °C與15 °C豆漿中的生長描述有較佳的預測能力，然而Bacillus sp.於7 °C豆漿中的生長情況則並不明顯。最後透過生長預測模式，可以得到Bacillus sp.在35、30、25及15 °C豆漿中的遲滯生長期為0.7、1.2、3.5、14.2小時，最大比生長速率為每小時1.3、1.0、0.6、0.09 log cfu/mL，而最大生長菌數皆約8 log cfu/mL。綜合以上，生長預測模式可預測Bacillus在豆漿中的生長情形。

關鍵字

豆漿；芽胞桿菌屬；生長預測模型；遲滯生長期；比生長速率

並列摘要

Soy milk is one of the popular beverages in Asia. According to national nutrition survey in Taiwan (NAHSIT) by National Food Consumption Database in 2017, the average daily intake of soymilk for consumers aged 19-65 is 50 g / person. Soy milk is rich in protein and carbohydrate. Thus, it is susceptible for bacterial growth. However, the growth of bacteria can be manipulated by various cultural conditions. Microbial growth behavior can be defined by mathematical model using Integrated Pathogen Modeling Program 2013 (IPMP 2013) in order to predict the shelf life of food products. In this study, we aim to investigate the growth of Bacillus sp. isolated from soy milk sold in restaurants. The isolated Bacillus sp. was exposed to various temperature condition and the growth rate was used to predict the growth model. Total plate counts of bacteria from soy milk made by restaurants around National Taiwan University were performed (n=58). The results indicated that six paked products had less bacterial colonies (< 10 cfu / mL) detected. However, 15 bulk products had high total plate counts (> 104 cfu / mL). With colony color and morphology, Bacillus sp. were isolated and the 16S rDNA was sequenced. Bacterial growth was investigated at various temperature (7、15、25、30 and 35 °C) and the results were fitted into Huang, Baranyi, modified Gompertz and Buchanan three-phase linear growth model using IPMP software to establish predictive growth model. The results show that Huang growth model and Buchanan three-phase linear growth model can fit for Bacillus sp. at higher (35、30 °C) and lower (25、15 °C) temperature successfully according to little small akaike information criterion (AIC) and root mean square error (RMSE). Besides that, the growth of Bacillus sp. in soy milk at 7 °C was not obvious. Finally, the lag time of cell growth in soymilk at different temperatures (35, 30, 25 and 15 °C) were 0.7, 1.2, 3.5, 14.2 hr, and the maximum specific growth rate were 1.3, 1.0, 0.6, 0.09 log cfu/mL/hour. Besides, the maximum growth population was about 8 log cfu / mL. The predictive models were applied to estimate the proliferation of Bacillus sp. in the products. In conclusion, it might be useful tool for the management of the shelf-life.