[Show abstract][Hide abstract] ABSTRACT: 仙人掌桿菌是主要的食品病原菌之一，與炭疽桿菌、蕈狀桿菌及蘇力菌同被歸類在仙人掌桿菌群，彼此在表型特徵及染色體序列上具有高度相似性。溶血性腸毒素HBL是仙人掌桿菌的主要毒性因子，亦為檢測仙人掌桿菌腸瀉型腸毒素商業化套組常見的檢測目標。在檢測仙人掌桿菌群菌株後發現，雖然僅有43%具有HBL活性，但所有的菌株皆對中國倉鼠卵巢細胞株(CHO cell)產生細胞毒性，因此不應僅限於檢測HBL活性，亦應同時檢測仙人掌桿菌群菌株。本研究利用16S rRNA基因設計對仙人掌桿菌群具特異性的PCR引子組B16S1與B16S2，另設計出一組用來檢測HBL基因的引子組Hm1及Hm2，利用Hm1及Hm2引子組進行PCR的檢測結果與BCET-RPLA免疫分析套組的檢測結果相符。將B16S1、B16S2、Hm1及Hm2兩組引子組合併成用來同時檢測仙人掌桿菌群與HBL腸毒素的多套式PCR，經過預培養步驟，此一PCR系統的檢測靈敏度可達n x 100 CFU/每克食品樣品。 蘇力菌廣泛使用於生物殺蟲劑，在傳統分類上只有以產生殺蟲毒蛋白的特徵來與仙人掌桿菌做區分，Giffel與Yamada等學者曾提出可利用16S rRNA基因及gyrB基因上的差異將兩菌區分，但部分學者認為兩種菌無法以分生方法區分而建議將兩種菌歸為同一種(species)，本研究分別以16S rRNA基因及gyrB基因為目標，以PCR方法進行仙人掌桿菌與蘇力菌區分性之評估，結果發現82株仙人掌桿菌中，有6株被判定為蘇力菌，而73株蘇力菌中，高達67株被誤判成仙人掌桿菌，利用選殖及對引子黏合部位進行序列分析後發現，仙人掌桿菌與蘇力菌在16S rRNA基因及gyrB基因上並無絕對的相異點，無法用來做為區分的依據，就目前而言傳統的結晶毒蛋白及cry基因的檢測仍是區分兩者較可靠的方法。 以抗生素敏感性分析、腸毒素圖譜、隨機擴增多形性分析(RAPD)與脈衝式膠體電腸電泳(PFGE)分析82株仙人掌桿菌與73株蘇力菌，抗生素敏感性分析可得到27種圖譜，腸毒素圖譜8型，以NotⅠ剪切的PFGE圖譜有121型，而以3F及6F引子分別可得到112及120種RAPD圖譜，不論是PFGE分析或RAPD分析皆無法將仙人掌桿菌與蘇力菌區分成二群不同的類型，也沒有出現可用來做為區分兩者的RAPD marker或PFGE條帶，由PFGE圖譜及親緣性樹狀圖可知在仙人掌桿菌與蘇力菌同菌種及相互間，並無法獲得相似的PFGE圖譜，其原因可能是由於仙人掌桿菌與蘇力菌染色體內的基因具有高度的重排性(rearrangement)，導致染色體序列呈現異質性(heterogeneity)的關係。 Bacillus cereus is one of the major foodborne pathogens. It is very closely related to B. anthracis, B. mycoides and B. thuringiensis. These species share many phenotypic properties and high levels of chromosomal sequence similarity. Hemolysin BL (HBL) is a major virulence factor for Bacillus cereus group strains. It is also a target enterotoxin for the most commonly used B. cereus detection kit, i.e., the BCET-RPLA kit. A survey of the hemolysin BL (HBL) activities and the cytotoxicities to the Chinese hamster ovary (CHO) cells for the B. cereus group strains, however, showed that although only part of the B. cereus group strains are HBL active, all strains show the cytotoxicity to the CHO cells. Thus, methods that allow the detection of not only the HBL, but also all of the B. cereus group strains are important. In this study, by comparison of the gene sequences of the 16S rRNA for B. cereus group and other bacteria strains, we designed primers B16S1 and B16S2 specific to all the B. cereus group strains. In addition, since HBL is a major enterotoxin, we also designed HBL gene specific PCR primers, i.e., Hm1 and Hm2, which generated the same results as those of the hemolysis and BCET-RPLA assays. Primers B16S1/B16S2 and Hm1/Hm2 could be combined into a multiplex PCR system for the simultaneous detection of B. cereus group cells and the possible presence of their HBL enterotoxins. Also, all these PCR systems allowed the detection of N x 100 CFU B. cereus cells per g of food sample if an 8 h enrichment step was performed prior to the PCR. B. thuringiensis is a widely used biopesticide. It can be distinguished from B. cereus by its ability to produce an insecticidal crystal inclusion inside the cell during sporulation. According to the differences in variable region V1 of 16S rRNA or gyrB gene between Bacillus cereus and B. thuringiensis, Giffel and Yamada et al. have suggested the use of DNA method to differentiate B. cereus from B. thuringiensis strains. However, other researchers have suggested that these two Bacillus species are indistinguished by molecular biological methods and should belong to the same species. In this study, we re-evaluated such DNA methods and tried to explain the results with DNA sequencing data. The results of PCR showed that six of eighty-two B. cereus strains were identified as B. thuringiensis and sixty-seven of seventy-three B. thuringiensis strains were identified as B. cereus. Sequence analysis of the primer annealing site showed that there is no clear-cut difference in the 16S rRNA gene or gyrB gene between B. cereus and B. thuringiensis strains. Therefore, we believe that of the methods to discriminate B. cereus and B. thuringiensis, parasporal crystal observation and cry gene detection may be the reliable methods. The comparison of the antibiograms, enterotoxin profiles, randomly amplified polymorphic DNA (RAPD), and pulsed-field gel electrophoresis (PFGE) patterns were also used to discriminate the strains of B. cereus and B. thuringiensis. 82 B. cereus and 73 B. thuringiensis strains were subtyped by antibiotic susceptibility testing, enterotoxin profile, RAPD and PFGE and results showed that for all those strains tested, 27 antibiogram and 8 enterotoxin profiles were found. Digestion with NotI generated 121 PFGE patterns. RAPD with primer 3F and 6F yielded 112 and 120 RAPD patterns, respectively. Thus, it is difficult to classify B. cereus ad B. thuringiensis strains by their RAPD or PFGE patterns. No specific RAPD or PFGE marker band could be found for the discrimination of these two Bacillus species. Due to the high frequency of genome rearrangements for these B. cereus ad B. thuringiensis strain, these strains display a higher level of heterogeneity within the chromosomal gene organization. This may also cause the diversified RAPD and PFGE patterns for B. cereus and B. thuringiensis strains.