仙人掌桿菌群之分子檢驗方法及非溶血性腸毒素表現之研究

Abstract

仙人掌桿菌 (Bacillus cereus) 是重要的食品病原菌，在台灣地區由1991至2006年之食品中毒案件病因物質中排名第三位，僅次於腸炎弧菌及金黃色葡萄球菌。其致病因子至少包括五種腹瀉型腸毒素以及一種嘔吐型毒素。本研究建立多重組對引子聚合酶鏈反應 (multiplex polymerase chain reaction, multiplex PCR) 方法，同時以12組引子偵測所有的腸毒型腸毒素基因及嘔吐型菌株特異序列，並將方法應用於162株仙人掌桿菌群分離菌株。結果顯示此些菌株可分為10種毒素型別，而且每株菌株至少攜帶一種毒素基因。雖然仙人掌桿菌與蕈狀桿菌之類緣關係相當接近，但超過70%之蕈狀桿菌 (Bacillus mycoides) 菌株沒有攜帶任何已知的毒素基因，且其毒素型別以及毒素基因分布與仙人掌桿菌菌株之間顯著不同，因此推測許多蕈狀桿菌菌株可能較不易導致食品中毒，也顯示檢驗菌株是否攜帶毒素基因與檢驗仙人掌桿菌群菌株同樣重要。傳統之仙人掌桿菌群菌株之分離及鑑定耗時費力，PCR方法雖可快速檢驗但無法精確定量。實務上，有必要進行食品中仙人掌桿菌群之定量檢驗。因此，本研究使用非溶血性腸毒素 (nonhemolytic enterotoxin, Nhe) 之基因nhe作為目標基因，建立即時聚合酶鏈反應 (real-time PCR) 檢驗方法，直接由食品及糞便中定量仙人掌桿菌群菌株，以達快速檢驗之目的。共使用60株仙人掌桿菌群菌株以及28株其他菌株進行即時聚合酶鏈反應方法之評估，並應用於不同之食品基質及糞便檢體。結果顯示此方法具有足夠的線性關係 (r2>0.993)、定量範圍 (米飯及雞肉檢體由102至107 CFU/g、牛奶檢體由103至107 CFU/mL、糞便檢體由104至107 CFU/g) 以及相對定量準確度 (85.5至101.1%)。另為解決定量低污染菌量之問題，結合最確數 (most probable number, MPN) 方法與real-time PCR建立最確數即時聚合酶鏈反應 (MPN real-time PCR) 檢驗方法，將提高靈敏性至100 CFU/mL。Real-time PCR以及MPN real-time PCR方法應用於市售食品檢體，結果顯示皆適合食品中仙人掌桿菌群菌株之檢驗及定量。仙人掌桿菌之Nhe腸毒素表現量在不同菌株之間相差甚多，而Nhe可能是導致食品中毒最主要的致病因子。為探究影響不同仙人掌桿菌菌株表現Nhe之機制，本研究比較Nhe表現量相異菌株組別之生長情形，並以三種軟體評估仙人掌桿菌mRNA表現研究可用之五種候選內部對照基因，再將評估結果用於建立定量nhe mRNA之反轉錄即時聚合酶鏈反應 (real-time reverse transcription PCR) 方法。結果顯示Nhe表現量相異菌株組別之生長曲線並無顯著差異。各候選之內部對照基因在菌株各生長階段皆有穩定的表現，比較Nhe表現量相異菌株組別之nhe mRNA表現量，發現高Nhe表現量之組別顯著高於低Nhe表現量之組別。因此，影響Nhe表現之差異應非導因於菌株生長速率之不同，推測應為nhe基因表現轉錄之層次所調控。

Bacillus cereus foodborne diseases are a major concern worldwide. In Taiwan for the period 1991 to 2006, outbreaks due to B. cereus were exceeded only by Vibrio parahaemolyticus and Staphylococcus aureus. Five different enterotoxins and one emetic toxin of B. cereus have been characterized. To amplify all of the enterotoxin and emetic-specific sequences of the species in the B. cereus group, a multiplex polymerase chain reaction (multiplex PCR) with 12 primer pairs was established. The assay was successfully applied to analyze the toxigenic potential of 162 isolated B. cereus group strains. Results showed that there were 10 toxigenic patterns for all the test strains. All of the B. cereus strains carried at least one toxin gene. More than 70% of B. mycoides strains carried no known toxin genes. The toxin profiles and toxin genes of B. mycoides strains were significantly different from B. cereus strains although the two species were closely related. The results suggested that many B. mycoides strains might be less prone to cause food poisoning. It also indicated the importance of detecting the toxin genes together with the detection of the species in the B. cereus group. Conventional bacteriological methods for the detection and identification of species of the B. cereus group require individual biochemical confirmation and are laborious and time-consuming. PCR is a choice of rapid detection but can not quantify the contamination level. In practice, it is necessary to quantify the contamination level of B. cereus group cells. We selected nhe coding for Nhe as the target and developed a real-time PCR assay to quantify enterotoxigenic strains of the B. cereus group. The real-time PCR assay was evaluated using 60 B. cereus group strains and 28 others. The assay was also used to construct calibration curves for different food matrices and feces. The assay has an excellent quantification capacity, as proved by its linearity (R2>0.993), wide dynamic quantification range (102-107 CFU/g for cooked rice and chicken, 103-107 CFU/mL for milk, and 104-107 CFU/g for feces), and adequate relative accuracy (85.5-101.1%). For the low level contaminations, a most probable number (MPN) real-time PCR assay was developed that could detect as low as 100 CFU/mL. Both assays were tested with real food samples and shown to be considerably appropriate for B. cereus group detection and quantification. The cytotoxicity titers of Nhe components varied considerably and the level of Nhe seems to explain most of the cytotoxic activity of B. cereus isolates. In order to examine the regulatory mechanism of different Nhe-producing strains, the expression level of their nhe mRNA was determined by real-time reverse transcription PCR. Five candidates of internal controls were evaluated by three softwares. Meanwhile, the growth curves of different Nhe-producing strains were also compared. All studied candidate of internal control genes reached high expression stability. The growth curves showed no significant difference but the nhe mRNA expression level of high Nhe-producing strains was significantly higher than low Nhe-producing strains. The results indicated that the different Nhe expression levels between B. cereus strains may not be due to their growth difference but may be controlled at the transcribed level of nhe gene expression.