Identification of Bacteroides fragilis from clinical samples

Abstract

The members of Bacteroides fragilis group are Gram-negative anaerobic, non-spore forming rods that are present in the normal flora of human intestine. B. fragilis group consists of ten species: B. caccae, B. distasonis, B. eggerthi, B. fragilis, B. thetaiotaomicron, B. vulgatus, B. ovatus, B. stercoris, B. uniformis and B. merdae. The dominating species in the colon microflora are B. thetaiotaomicron, B. vulgatus and B. distasonis, while B. fragilis constitutes only 0.5% of the cultivable colon microflora. B. fragilis species account for roughly one-fourth of all anaerobic bacteria isolated from clinical specimen. Therefore, B. fragilis is regarded as the most important anaerobic pathogen causing intra-abdominal abscesses and bloodstream infections.

Recent investigations have revealed that some strains of B. fragilis may also be associated with diarrheal disease in animals and humans due to the production of an enterotoxin. The enterotoxin of enterotoxigenic B. fragilis (ETBF) could induce fluid changes in ligated intestinal segments and cytotoxic response in HT29/C1 cells. It has been suggested that ETBF may be endemic in communities. Thus, determination of the potential role of ETBF in human diarrheal diseases is of considerable importance and interest. In addition, the majority of B. fragilis isolates are resistant to ß-lactam antimicrobial agents as well as other antibiotics. Therefore, an early and rapid identification of B. fragilis is an important factor for the treatment of diseases caused by B. fragilis species.

The routine methods to identify B. fragilis are mainly based on the use of biochemical assays and gas liquid chromatography. Compared with other Gram-negative bacteria, B. fragilis species grow slower and with some difficulty. The limitations of the above-mentioned assays are that they are time consuming and cumbersome. For the detection of ETBF, some assays, i.e. HT29/C1 cell line assay, animal model and PCR have been established. However, the sensitivities and the need of pure cultures make them less suitable for rapid diagnosis.

In this study, two monoclonal antibodies (mAb) against the lipopolysaccharide (LPS) of B. fragilis have been produced. The mAb 4H8 (IgG3), binds to an immunodominant epitope in the LPS of a majority of B. fragilis isolates, and mAb C3 (IgG2b), most likely binds to a common epitope present in the inner core region of B. fragilis LPS. Using these two monoclonal antibodies alone or in combination with PCR assay, three different assays for identification of B. fragilis from clinical samples have been set up.

Co-agglutination is regarded as a simple and rapid method to identify bacteria, In the first study, mAb C3 was adsorbed to Staphylococcus aureus Cowan I bacteria. The reagent was used for identification of B. fragilis. Almost 98% of the B. fragilis strains were positive in co-agglutination. Among the 283 non-B. fragilis strains only two showed false positive reaction. The results showed that mAb C3 has a high specificity and can be used in coagglutination for rapid identification of the B. fragilis species.

A modified Immuno Polymerase Chain Reaction (mIPCR) using mAb 4H8 for identification of B. fragilis from bacterial colony as well as from clinical samples (pus), with the detection limit of 104 CFU/mL bacterial suspension has been developed. A number, of bacterial strains were examined, including B. fragilis, Bacteroides spp. other than B. fragilis and other genera. All the B. fragilis reactive with the mAb 4h8 were positive. None of the other strains showed any positive reaction. The mIPCR assay was also successful in order to detect B. fragilis from clinical pus samples. These results indicated that mIPCR assay with mAb 4H8 has a high specificity and high sensitivity, and mIPCR may be used to detect B. fragilis in pus samples or other type of clinical sample.

For the detection of ETBF, a pair of well-characterized specific monoclonal antibodies (mAb C3 and mAb 4H8) was used. The antibodies were coated on magnetic beads and used for subsequent capture of bacteria by immunomagnetic separation (IMS). In this study, an IMS-PCR assay, based on the use of B. fragilis mAbs, followed by PCR amplification of the enterotoxin gene for a rapid, specific and sensitive detection of ETBF directly from clinical samples has been developed. The detection limit was found to be ~50 cfu/g of feces. The minimum time to retrieve the final result by the IMS-PCR assay is 36h.

To investigate the association between ETBF and diarrheal disease, the IMS-PCR assay was used for detection of ETBF in fecal samples from diarrhea] and healthy Swedish adults. A total of 922 fecal samples were analyzed, including 728 samples from patients with diarrhea and 194 samples from controls. ETBF was detected in 195 of 728 (26.8%) of the patients with diarrhea and in 24 (12.4%) of the healthy controls. The difference between the two groups was significant (p<0.01). ETBF was also found as the only potential diarrheal pathogen in 91 out of 728 (12.5%) patients. The data from this study showed a high carriage of ETBF in Swedish adults that might be associated with diarrheal disease.

Further analyzes of the distribution of ETBF among the B. fragilis strains isolated from the patients with diarrhea and from healthy controls were performed. From the selected patients harboring ETBF, a total of 278 B. fragilis colonies were isolated. A majority (64%) of them were identified as ETBF. Only 32% of the B. fragilis strains isolated from healthy controls were enterotoxin positive. In about 90% of the patients it is shown that ETBF strains dominate among B. fragilis isolates. The corresponding figure for the controls was 12%. There is a significant difference in individual percentages of ETBF among isolated B. fragilis strains between patients and controls. It indicates that the higher counts of ETBF in the patients with diarrhea as compared with asymptomatic carriers may result in higher concentration of enterotoxin production.