Browsing by Author "Hänel, Kristina"

Now showing 1 - 1 of 1

Open Access
Design and application of a DNA microarray for the identification of intestinal pathogens during gastroenteritis and monitoring of the resident intestinal microbiota
(2012) Hänel, Kristina; Schmid, Rolf D. (Prof. Dr.)
Bacterial gastroenteritis is a commonly occurring disorder with three to five billion cases of acute diarrhea annually worldwide, particularly in developing countries where it is a leading cause of childhood morbidity and mortality. Diarrheal diseases can quickly reach epidemic dimensions, especially in the case of food-borne pathogens. Fast identification of the etiologic agent remains a key task in clinical diagnosis. Currently, it is mainly based on phenotypic classification using stool culture and a set of morphological, physiological, and serological tests. However, cultivation of the intestinal pathogen can often be a challenging and time-consuming task. Microarrays can fulfil the need for a rapid diagnostic tool that generates reliable information with a minimum of laboratory effort and do not rely on cultivation of the organisms. Here, a diagnostic oligonucleotide microarray was developed to detect the most common bacterial species associated with gastroenteritis. As a new concept, the array aimed at also providing information about the residential flora and potential probiotics. This combination allows not only pathogen identification but also a monitoring of the patients gut flora recovery during therapy. The developed array comprised probes for 23 species and 11 genera following a multiple-probe concept. This covers also 13 common bacterial gastroenteritis-related pathogens. The DNA array performance was initially verified with pure cultures from clinical isolates and DSM strains and confirmed by sequencing. The final array allowed unambiguous identification of the target species by combining the multiple-probe concept with a robust cut-off. The limit of detection was determined to be 103 genome equivalents. The clinical applicability was examined by processing stool samples from 58 patients with gastroenteritic symptoms and 6 healthy individuals using the diagnostic microarray. For 30 samples where the etiologic pathogen had been detected by a combination of culture and specific real-time PCR assays, 67% of these samples were correctly identified. It was assumed that a higher recovery rate would require a lower detection limit of the array. Furthermore, highly individualized communities of the detected resident bacteria were found in the analysed samples. A difference of this microbiota between healthy and infected subjects was not observed by multivariate analysis of the array data. This was attributed to the inhomogeneous sample group in terms of patient age and clinical diagnose. Additionally, the array was applied to investigate the influence of a viral, intestinal infection on the composition of the infantile microbiota and the establishment of human and porcine intestinal community in inoculated piglets. In both trials, also bacterial pathogens could be identified. Multivariate analysis of the microarray data revealed a significant difference of the resident flora between healthy and infected children and between the pig and human intestinal flora. Partial least squares analysis and one-way analysis of variance allowed identification of resident species or genera, which significantly differed between these groups. According to literature survey, this was the first application of principle component and partial least squares analysis for statistical evaluation of identification microarray data. The results indicated the applicability of the present microarray for studying the resident microbiota under a certain question. Nevertheless, it was concluded that more species should be included in the array to allow in-depth analysis of faecal microbiota. To address the problem of reliable quantification of intestinal bacteria and pathogens with a concentration range over several orders of magnitude using microarrays, a FRET-based system was developed. A black hole quencher was used to gradually quench the Cyanine 3 reporter signal allowing detection of a target species in the linear scanner range independent from its initial concentration. The proof of concept based on E. coli-specific probes showed that an effective quenching of the fluorophore label in the target was possible upon hybridization and dependent on the amount of immobilized quencher. It was concluded that two spots would be sufficient to detect 104 up to 2*106 genome equivalents in the linear range of the scanner. This setup partly overcomes the technical limitation of linear fluorescence signal acquisition for different amounts of bacteria. In summary, in this study solutions for two problems of molecular pathogen detection in the clinic were developed: (1) the time for pathogen detection and (2) the assay coverage. The analytical sensitivity for clinical application was not yet sufficient due to limitations of the pre-analytical procedures. A general concept to expand the linear range of a fluorescence intensity based detection system was achieved by a FRET-based system with immobilized quencher on the array surface.