Identification of unique bacteria among different animal species using denaturing gradient gel electrophoresis (DGGE) and single strand conformation polymorphism (SSCP) methods

Fecal contamination of water bodies is a widespread problem throughout the world. The United States Environmental Protection Agency (EPA) reported about 21,000 water bodies in the United States are contaminated with feces from both animal and human sources (EPA, 2001).;This study aimed to find unique fingerprints among different animal species to use for source identification and calculating the Total Maximum Daily Load (TMDL) for a body of water. Two methods were used (DGGE and CE-SSCP) to separate the bacterial species contained in the fecal samples collected from different animal species including cow, pig, chicken, chicken litter, deer, goose and horse.;Comparing the DGGE band patterns by location using Idrisi software revealed unique bacteria for the animal species from Arkansas and Tennessee. Comparing the DGGE band patterns using the Dice Similarity Coefficient showed very large difference between the bacterial species with regard to their band patterns (>90% similarity level). This revealed high % of difference between the animal species indicate that there is chance to established fingerprints out of this study using DGGE. In addition, similar species comparison showed some similarity but also showed uniqueness for each species. This finding might indicate the weakness of choosing the total fecal bacteria to establish the fingerprints instead of selecting one bacterium indicator instead.;This study used CE-SSCP method in addition to DGGE to evaluate the discriminatory power for differentiating between the bacterial species.;The CE-SSCP did not reveal the number of peaks expected and calculated from the DGGE band patterns (;In order to resolve the lower number of peaks resulted from the CE-SSCP, the Electropherograms generated for each animal species were regenerated in the "PeakFit" (V. 4.12) program to reveal the hidden peaks. All of the Electropherograms were overlapped and compared and regardless the lower number of peaks, the CE-SSCP revealed unique fingerprints for the horse and for chicken litter samples that DGGE did not. This finding indicates that both CE-SSCP and DGGE methods are using different baths to differentiate between the bacterial species and/or other factors associated with DGGE process might contributed to this including denaturing gradient concentration, PCR bias and co-migration.;However, the findings using Dice Similarity Coefficient indicated that the present study has potential for bacterial fingerprints to be established using DGGE. These bacterial fingerprints were suppose to help with TMDL calculations. Therefore, this study proposed two methods for allocating the bacterial loads among the different animal species which are Luminex beads and Microarrays. The principle for these two methods depends on quantifying the intensity of DNA from the bacterial fingerprints by measuring the hybridization intensity between the target DNA and its correspondent probe. The bead color in the Luminex beads method and the different slide spots (different probes) on the Microarray methods can tell from what source the bacterial fingerprint DNA came from. So, The Luminex beads and Microarrays methods are not only used to identify the source of fecal contamination but also allocate the bacterial loads among the different sources.;Results of BST analysis, combined with the output of simulation models, could give considerable scientific justification to TMDL allocation scenarios and implementation plans for water bodies contaminated by fecal bacteria.;On the other hand, this study relies on the bacterial content of the animal feces. Depending on the fact that two animals of the same species do not have the same genetic profile for their bacteria even if they are living under similar conditions and consuming the same diet. This means that each individual of one animal species has to be included in the source tracking study which is not possible. This makes the initiation of the watershed model hard and complicated. (Abstract shortened by UMI.)...