2003 Fairfax County Regional Science and Engineering Fair

The project description below was written by the student. The Northern Virginia Soil and Water Conservation District made no editorial changes.

Abstract

Fecal pollution is a leading cause of global enteric waterborne diseases, from which over two million children die annually (WHO, 2000). Fecal contamination is tested by cultivation of Enterococci or E. coli. Recent studies (Bernhard, Langendjik) suggest that using molecular methods rather than traditional culturing, which may dramatically reduce bacterial diversity (Suau) and misrepresent the source population, Bacteroides and Bifidobacteria could be used to not only indicate but track sources of fecal contamination. These bacterial groups are more prevalent in feces than the current indicators (Sghir). In this study, terminal restriction fragment length polymorphism (T-RFLP) was used to assess variations within the Bacteroides and Bifidobacterium communities in stool from species that were potential sources of contamination and in fecal-contaminated waters. Multiple samples were collected from different geographical regions to assess the spatial variation of populations. DNA extracted from the samples was amplified by PCR using published primers for the Bacteroides-Prevotella and Bifidobacterium groups. Restriction fragments from different enzymes were analyzed and compared using principal component analysis to develop criteria to distinguish between different sources of feces. Results of this study indicate that the Bacteroides populations vary between the stool of different source species and that Bifidobacterium populations vary geographically within a single source species. Both bacterial groups were also detected in waters contaminated with feces and linked through similar T-RFLP patterns to the known source of the contamination. These results suggest that molecular techniques using Bacteroides and Bifidobacteria may be useful in detecting and identifying sources of fecal contamination.

Introduction

Contamination of drinking and recreational water such as wells, rivers, or ponds by sewage or animal feces leads to dangerous illnesses (WHO, 1998). If further outbreaks are to be prevented, the cause of the contamination must be located, be it a sewage pipe, a cattle farm, or local wildlife. For legal and practical purposes this source must be identified rapidly and accurately. The origin of the pollution may indicate the pathogens that may be expected, the risk of infection, and the treatment required to control the spread of the disease (Ha).

The current indicators of fecal pollution, E. coli and Enterococci (EPA), are extracted from water samples and cultured on plates for quantification. However, Enterococci and E. coli do not constitute a significant proportion of the fecal flora compared to other bacterial groups such as the Bacteroides and Bifidobacteria (Sghir, Suau). E. coli and Enterococci still cannot be used to determine sources of contamination, in part because they are neither numerous nor diverse enough in feces to vary among different source species. Natural reservoirs of enterococci exist in the environment (AAM), creating the potential for false positives and skewed data.

Culturing the indicator organism to measure the amount of contamination introduces potential bias, for laboratory conditions favor some bacterial strains over others (Suau). Preserving the initial proportions of different strains is even more important when investigating the source of contamination, for current methods use the ratio of various strains present in the stool to differentiate among sources.

A simple, quick means to identify sources of fecal pollution has not yet been developed. To be a practical method for testing bodies of water around the world, the procedure must be rapid, easily repeatable, and relatively inexpensive. This experiment intends to evaluate the viability of Bifidobacteria and Bacteroides as potential tools for fecal source tracking, based on the possibility that Bacteroides and Bifidobacterium species isolated from fecal material and water contaminated with stool will be distinctive to the source species.

Methods and Materials

Sample collection and DNA extraction: Bovine fecal material was collected from two farms located in Virginia (5 samples each), 3 in Indiana (4-5 samples each), and ten in West Virginia (43 samples) prior to the start of the experiment. Five or six stool samples were taken from each of two Virginian farms and three Indiana farms. In Indiana four water samples were also taken from the Iroquois River and Sugar Creek from locations downstream of wastewater treatment plants, adjacent to cattle farms, affected by wildlife, or distant from any apparent source of contamination. The West Virginian samples were suspended in glycerol and frozen at -20°C, while the rest (hereafter referred to as stock samples) were frozen and stored at -20°C within six hours of collection. Stock and glycerol samples of horse, dog, and goose feces as well as of sewage were also collected following the same procedures.

Stock samples were extracted using the Qiagen QIamp® DNA Stool Mini Kit (Qiagen, Valencia, CA). About 2g of fecal material were homogenized in 10 ml of ASL buffer (Qiagen). DNA was extracted from 2 ml of the homogenate following the protocol for isolation of DNA stool for pathogen detection in the QIamp® DNA Stool Mini Kit Handbook included in the kit (Qiagen, 2002). The glycerol samples were thawed and added to 5 ml of ASL buffer. DNA was extracted from the entire fecal slurry following the same protocol as above except adding a proportionately more of each reagent. All DNA samples were stored at -20°C after extraction.

Fecal-contaminated water: To evaluate the similarity of the biota in contaminated water to that of the original fecal material, about 1g of fecal material from a single cow was added to 1 liter of pond water. This water was used because of its similarity with respect to nutrient and chemical composition to other environmental water supplies that might require testing for fecal pollution. The original pond water and fecal-contaminated water were each filtered onto a 0.2 [M Sterivex filter. DNA was extracted from the filters with the PUREGENE DNA Isolation Kit. DNA was extracted from the original fecal sample with the Qiagen QIAMP DNA Stool Kit following procedures mentioned above.

Bacterial survival rate in environment: To assess the survival rate of fecal bacteria in the environment, four grams of bovine fecal material was added to two liters of water from the aforementioned USGS pond water. After the stool was dissolved, 800 ml of the fecal-contaminated water was added to each of two 1-L polypropylene bottles. To simulate environmental conditions, the sealed bottles were submerged in a 19°C water bath and mixed continually by stir plates. In order to account for the affects of ultraviolet radiation, the water bath was placed outdoors where the bottles could receive sunlight. Fifty-ml samples were taken from each bottle at the onset of the experiment and after 1, 3, 5, and 12 days. The water samples were filtered onto Sterivex filter units and stored at -20°C until DNA extraction by the PUREGENE DNA Isolation Kit.

DNA amplification: Amplification of Bacteroides DNA that had been extracted from fecal and water samples was performed with the 16S rRNA primer set specific for the Bacteroides-Prevotella group, Bac32-f and Bac-708r used in similar studies (Bernhard). The PCR reaction components consisted of a final concentration of 1X PCR buffer pre-mixed in lab, 1.2 mg/ml bovine serum albumin (BSA), 0.005% DMSO, 0.2 mM deoxynucleoside triphosphates (dNTPs), 0.2 [M of forward and reverse primers, 1[l Taq mixed in laboratory, 1 [l DNA at appropriate concentration, and sterile tissue culture water to bring reaction volume up to 50 [l. DNA amplification was carried out in a Perkin Elmer 2400 Gene Amp PCR system (Perkin Elmer-Cetus, Norwalk, CN) with the following conditions: 95° for 7 min; 34 cycles of: 94°C for 3 sec, 92°C for 30 sec, 40°C for 1 min, 65°C for 8 min; a final elongation of 16 min at 65°C; and a final hold at 4°C.

Amplification of Bifidobacterium DNA from the bovine fecal and water samples was performed with the genus-specific 16S rRNA primers Lm26-f and Lm3-r. The Bifidobacteria PCR reaction, after having been optimized using modified Taguchi methods (Cobb), consisted of the same reagent concentrations as that of the Bacteroides reaction and was performed in the same machine. The thermocycling program was modified from previous experiments that had used the same primer set (Satokari, Ventura): 94°C for 5 minutes; 35 cycles of 30 s at 94°C, 60 s at 57°C, 2 minutes at 72°C; a final elongation for 6 minutes at 72°C; and a final hold at 4°C.

PCR tubes were stored at 4°C for the first two weeks and then moved to long term storage at -20°C. The presence of amplified DNA was confirmed by analyzing 5 ml of PCR product by 1% agarose gel electrophoresis and ethidium bromide staining.

Restriction digestion and DNA precipitation: Restriction enzymes were selected on the basis of the diversity of recognition sites among different strains or species of the same genus and the quantity of restriction fragments produced. The web-based research tool located at the Ribosomal Database Project web site (Marsh) as well as lab experimentation provided this data. Amplified Bacteroides DNA was cut with single digests of HaeIII, AciI, and MspI restriction enzymes. Bifidobacteria DNA was cut with single digests of HaeIII, AciI, and HhaI. HaeIII and HhaI restriction digests were composed of 2 ml enzyme buffer, 0.2 ml BSA, 0.5 ml enzyme, 9.3 ml sterile water and 8 ml PCR product. AciI digests omitted the BSA, the amount of water increased to maintain the same total digest volume. Digests were incubated at 37°C for 8 hours or overnight, then stored at -20°C until precipitation.

After digestion each sample received 2 ml 3M sodium acetate and 45 ml 100% ethanol. The DNA was precipitated for at least 2 hours before tubes were centrifuged at 4°C for 15 minutes at high speeds. The supernatant was immediately poured off and discarded. 500 ml of 70% ethanol was added to wash pellet, the tubes were vortexed, and samples were centrifuged for another 5 minutes at 4°C at high speeds. Supernatant was poured off a second time, and tubes were left inverted on a kim wipe to air dry for at least 30 minutes. Samples were then heated to 70-90°C on a heat block until the rest of the fluid had evaporated, about 2-7 minutes. 8 ml sterile water was added. Samples were maintained at 4°C for 2-3 days while DNA hydrated and then stored at -20°C.

Collection of T-RFLP data: The DNA fragments migrated up a filament in the GeneScan machine according to their size, the smaller fragments migrating more quickly than the larger ones. The 6-carboxyfluorescein phosphoramidite (FAM) label on the 5' end of each DNA strand was recorded as it passed by, allowing the computer program to compute the quantity of each size of fragment. The software then graphed this terminal restriction length polymorphism (T-RFLP) data for each digest, the peaks representing restriction fragments of high instance.
Data Analysis: T-RFLP data was analyzed both visually and with the aid of principle component analysis (PCA), which similar studies of bacterial community dynamics over time (Dollhopf) found useful in analyzing large quantities of complex data. The quantity of each size of restriction fragment was compared across the different samples. PCA was performed with the software PCOrd. For logistical reasons only three samples from each location were used in the PCA of the Bifidobacteria data, although all samples were visually examined and the best representatives were chosen.

Discussion and Conclusions

The Bacteroides data not only suggest that Bacteroides communities vary among stool of different source species, but that the communities are similar among the same source species. While no single "marker" restriction fragment was identified that could incontrovertibly indicate the cause of a contamination event, the combination of T-RFLP data from the three different restriction enzymes proved sufficient to separate fecal samples with respect to their source species in many cases.

The Bifidobacteria study focused on the geographical diversity among the different bovine fecal samples. While Bifidobacteria T-RFLP profiles of bovine feces from different regions were more similar to one another than Bacteroides data from different species, geographical variations in Bifidobacterium communities were found within the bovine stool. However, these variations could not be detected significantly until extensive PCA analysis was performed.

T-RFLP profiles from fecal-contaminated water resembled the profiles of the stool added (see Appendix), testifying to the accuracy of the molecular methodology. Bifidobacteria survived at high levels in fecal-contaminated water for over two weeks at 19°C, suggesting that sources could be determined long after the contamination event took place. This study had little difficulty amplifying Bacteroides and Bifidobacteria from mammal feces, although other studies using different Bifidobacterium primer sets (Wang, Bernhard, Langendijk) have reported varied degrees of success. Amplifying DNA from avian stool and from pond or river water was feasible, albeit more difficult due to the inhibitors present in environmental water and the nature of birds' digestive tracts and feces. Any source tracking technique relying on the variability fecal bacterial communities among source species may not be as effective with geese, whose fecal flora vary drastically depending on freshness of sample and season.

If further refined, molecular techniques using Bacteroides and Bifidobacteria as indicator species could potentially be used to determine sources of fecal pollution. Significant fragments identified in the T-RFLP profiles could potentially be sequenced and used as DNA probes for detection of fecal contamination and identification of sources. However, extensive statistical analysis is often required to draw conclusions from the complex T-RFLP profiles, and for the sake of accuracy a large sample set must be studied. Because a PCR machine, proper enzymes and reagents, and a GeneScan machine are necessary, this procedure is neither as efficient nor as inexpensive as might be desired for testing global water supplies. In light of the current dearth of accurate methods for fecal source tracking (AAM), however, this procedure proves promising.

Bibliography

American Academy of Microbiology (AAM), 2001, "Reevaluation of Microbial Water Quality: Powerful New Tools for Detection and Risk Assessment," Washington D.C.: AAM.

Bernhard, Anne E. and Katharine G. Field, 2000, "Identification of Nonpoint Sources of Fecal Pollution in Coastal Waters by Using Host-Specific 16S Ribosomal DNA Genetic Markers from Fecal Anaerobes," Applied and Environmental Microbiology, April 2000, pp. 1587-1594.

Bernhard, Anne E. and Katharine G. Field, 2000, "A PCR Assay To Discriminate Human and Ruminant Feces on the Basis of Host Differences in Bacteroides-Prevotella Genes Encoding 16S rRNA," Applied and Environmental Microbiology, October 2000, pp. 4571-4574.

Clarke, R. T. J. and T. Bauchop, 1977, Microbial Ecology of the Gut, New York, NY: Academic Press. pp. 47-59.

Cobb, Benjamin D. and John M. Clarkson, 1994, "A simple procedure for optimizing the polymerase chain reaction (PCR) using modified Taguchi methods," Nucleic Acids Research, Vol. 22, No. 18, pp. 3801-3805.

Dollhopf, S. L., Hashham, S. A., Tiedje, J. M., 2001, "Interpreting 16S rDNA T-RFLP Data: Application of Self-Organizing Maps and Principle Component Analysis to Describe Community Dynamics and Convergence," Microbial Ecology, Vol. 42, pp. 495-505.

Drasar, B. S., Barrow, P. A., 1985, "Intestinal Microbiology," Aspects of Microbiology, Washington, D.C.: American Society for Microbiology. pp. 25-27

Franks, Alison H., Hermie J. M. Harmsen, Gerwin C. Raangs, Gijsbert J. Jansen, Frits Schut, and Gjalt W. Welling, 1998, "Variations of Bacterial Populations in Human Feces Measured by In Situ Hybridization with Group-Specific 16S rRNA-Targeted Oligonucleotide Probes," Applied and Environmental Microbiology, September 1998, pp. 3336-3345.

Ha, Haejin, and Michael K. Stenstrom, 2003, "Methods to Identify Human and Animal Faecal Pollution in Water," Los Angeles, CA: UCLA.

Kimura, Katsunori, Anne L. McCartney, Michelle A. McConnell, and Gerald W. Tannock, 1997, "Analysis of Fecal Populations of Bifidobacteria and Lactobacilli and Investigation of the Immunological Responses of Their Human Hosts to the Predominant Strains," Applied and Environmental Microbiology, September 1997, pp. 3394-3398.

Langendijk, Petra S., Frits Schut, Gijsbert J. Jansen, Gerwin C. Raangs, Ger R. Kamphuis, Michael H. F. Wilkinson, and Gjalt W. Welling, 1995, "Quantitative Fluorescence In Situ Hybridization of Bifidobacterium spp. with Genus-Specific 16S rRNA-Targeted Probes and Its Application in Fecal Samples," Applied and Environmental Microbiology, August 1995, pp. 3069-3075.

Marsh, Terrence L., Paul Saxman, James Cole, and James Tiedje, 2000, "Terminal Restriction Fragment Length Polymorphism Analysis Program, a Web-Based Research Tool for Microbial Community Analysis," Applied and Environmental Microbiology, August 2000, pp. 3616-3620.

Matsuki, Takahiro, Koichi Watanabe, Ryuichiro Tanaka, and Hiroshi Oyaizu, 1998, "Rapid Identification of Human Intestinal Bifidobacteria by 16S rRNA-Targeted Species- and Group-Specific Primers," FEMS Microbiology Letters, Vol. 617, pp. 113-121.

Satokari, Reeta M., Elaine E. Vaughan, Antoon D. L. Akkermans, Maria Saarela, and Willem M. de Vos, 2001, "Bifidobacterial Diversity in Human Feces Detected by Genus-Specific PCR and Denaturing Gradient Gel Electrophoresis," Applied and Environmental Microbiology, February 2001, pp.504-512.
Sghir, Abdelghani, Genevieve Gramet, Antonia Suau, Violaine Rochet, Philippe Pochart, and Joel Doré, 2000, "Quantification of Bacterial Groups within Human Fecal Flora by Oligonucleotide Probe Hybridization," Applied and Environmental Microbiology, May 2000, pp. 2263-2266.

Suau, Antonia, Régis Bonnet, Maléne Sutren, Jean-Jacques Godon, Glenn R. Gibson, Matthew D. Collins, and Joel Doré, 1999, "Direct Analysis of Genes Encoding 16S rRNA from Complex Communities Reveals Many Novel Molecular Species within the Human Gut," Applied and Environmental Microbiology, November 1999, pp. 4799-4807.

United States Environmental Protection Agency (EPA), 2002, "Implementation Guidance for Ambient Water Quality Criteria for Bacteria," Washington, D.C.: EPA.

Ventura, Marco, Marina Elli, Roberto Reniero, Ralf Zink, 2001, "Molecular Microbial Analysis of Bifidobacterium Isolates from Different Environments by the Species-Specific Amplified Ribosomal DNA Restriction Analysis (ARDRA)," FEMS Microbiology Ecology, Vol. 36 pp. 113-121.

Ventura, Marco, Roberto Reniero, and Ralf Zink, 2001, "Specific Identification and Targeted Characterization of Bifidobacterium lactis from Different Environmental Isolates by a Combined Multiplex-PCR Approach," Applied and Environmental Microbiology, June 2001, pp. 2760-2765.

Wilson, Kenneth H., and Rhonda B. Blitchington, 1996, "Human Colonic Biota Studied by Ribosomal DNA Sequence Analysis," Applied and Environmental Microbiology, July 1996, pp. 2273-2278.

Wang, Rong-Fu, Wei-Wen Cao, and Carl E. Cerniglia, 1996, "PCR Detection and Quantitiation of Predominant Anaerobic Bacteria in Human and Animal Fecal Samples," Applied and Environmental Microbiology, April 1996, pp. 1242-1247.

World Health Organization (WHO), "Who 1998 Guidelines for Safe Recreational-water Environments: Coastal and Freshwaters," Geneva: WHO.

World Health Organization (WHO), 2000, "The Gateway to Children's Environmental Health." Online: www.who.int/peh/ceh/index.htm

Acknowledgments

I would like to thank Lisa Fogarty, for taking the time from her own work schedule to teach me the techniques and methodology. Dr. Mary Voytek also has my gratitude, for her invaluable assistance with my papers and allowing me to work in her laboratory, and Julie Kirshtein, for loading the Genescans and taking my digests out of the incubator.