2002 Progress Report: Molecular Detection of Anaerobic Bacteria as Indicator Species for Fecal Pollution in WaterEPA Grant Number: R827639
Title: Molecular Detection of Anaerobic Bacteria as Indicator Species for Fecal Pollution in Water
Investigators: Field, Katharine G.
Institution: Oregon State University
EPA Project Officer: Packard, Benjamin H
Project Period: November 1, 1999 through October 31, 2002
Project Period Covered by this Report: November 1, 2001 through October 31, 2002
Project Amount: $223,829
RFA: Ecological Indicators (1999) RFA Text | Recipients Lists
Research Category: Ecological Indicators/Assessment/Restoration , Ecosystems
The overall objectives of this research project are to: (1) develop additional Bacteroides-Prevotella indicators from species that are believed to make a significant contribution to fecal pollution in the waters of Tillamook Bay and its feeder rivers; (2) analyze clone libraries of Bacteroides-Prevotella 16S rDNAs amplified from cow, human, and other species; (3) identify the strains unique to different host species; (4) design oligonucleotide probes; and (5) quantify the amount of Bacteroides-Prevotella fecal bacteria in samples, and the proportional contribution from cattle, human, and other sources, using real-time quantitative polymerase chain reaction (PCR). An additional objective of this research project is to test a method of establishing the source of fecal pollution in water, based on amplification of marker genes from Bacteroidetes fecal anaerobic bacteria.
Our method of specifically detecting human and ruminant fecal pollution already is in use in several parts of the United States. Practical applications of this part of the research include indicators for additional species, and a method of estimating the proportional contribution of different sources.
Twelve candidate species-specific primers were designed and tested (summer, fall, and winter in 2001-2002). For each one, annealing temperature empirically was established by gradient PCR in a 96-well microtiter plate, and reactant concentrations experimentally were optimized. If the primer, paired with a general primer, gave strong amplification with the target DNAs, it was then screened against other non-target DNAs to search for nonspecific amplification.
A total of four 5' elk primers were designed, optimized, and tested to distinguish wild from domestic ruminant species. At least one of these primers showed promise, but most amplified targets in certain domestic ruminants as well. Three primers from cat, one from dog, and one specific to both dog and cat, were designed and tested to distinguish fecal pollution from domestic pets from that of humans and other animals. One showed promise, but most also amplified targets in human samples. Three primers were designed and tested from pig. One did not work, one showed weak amplification but good specificity, and one was both sensitive and specific. This primer distinguished pig fecal pollution from that of all other animals tested.
Many of the new primers amplified a signal from a species other than the target species. Although each of the clone libraries is quite large, it still does not represent the entire diversity of Bacteroidetes sequences found within each species. We concluded that it is not practical to construct clone libraries large enough to sample the entire range of sequence diversity found within each species (because of the large cost associated with this amount of sequencing, and the diminishing return of unique groups as more sequences are analyzed). Nevertheless, the observed patterns of diversity indicate that there is substantial endemism within the Bacteroides-Prevotella group; sequence variability within this group is very great. Thus, we were determined to apply a new methodology, involving subtractive hybridization in microplates, to identify unique primer sequences.
Subtractive hybridization utilizes DNA: DNA hybridization between "target" and "subtractor" sequences created by PCR and digested into manageably sized fragments. A molar excess of subtractor sequences are anchored to the bottom of microplate wells and denatured. Target sequences are denatured, added to the wells with subtractor sequences, and allowed to hybridize. Unique target sequences are left unhybridized. These are removed from the wells, cloned, sequenced, and used to design primer sequences.
Our initial goal was to optimize this system to design host-species-specific Bacteroidetes rDNA primers. We paired our Bacteroidetes-specific forward primer (Bac32F) with a 23S rDNA primer (23S422R), adapted from published reports, designed and optimized restriction-site-extended primers and specific linkers for target and subtractor, and tested various restriction enzymes.
We now have performed four different subtractive hybridization experiments. The first two utilized amplified Bacteroidetes rDNA from elk feces as target with either human, or human plus cow, fecal rDNAs as subtractors. Both of these experiments resulted in the recovery of two bands, one of which was very faint. The more intense bands were cloned and sequenced. Eight clones were initially analyzed. All clones represented SSU rDNAs from 940 to 1370 (E. coli numbering). We used these sequences to design two new primers, which used together, appear to preferentially amplify elk (wild ruminant) fecal pollution, allowing us to distinguish between fecal contamination from wild and domestic ruminants.
The second two utilized amplified Bacteroidetes rDNA from dog feces as target with either human, or human plus cat, fecal rDNAs as subtractors. These experiments resulted in recovery and cloning of bands similar to the elk bands recovered in previous experiments. The sequences currently are being analyzed and used to design dog-specific primers.
Both sets of subtractive hybridization experiments resulted in the recovery of fragments from an area of the Bacteroidetes 16S rRNA gene that we previously had not targeted in our clone libraries. To maximize specificity to the Bacteroidetes group when constructing clone libraries, we used group-specific primers that only amplified the first half of the gene. Thus, to analyze the new fragments, we must rely on the few Bacteroidetes sequences that are available in public databases (we use ARB). This suggests that we need to construct new clone libraries that will provide more sequence data from this new area of the gene. Unlike the previous libraries, these probably could be quite small. However, we have concluded that subtractive hybridization will work best when paired with clone library analysis.
For that reason, we have begun to analyze a new Bacteroidetes rDNA clone library from seagulls. This will be used in tandem with our next subtractive hybridization experiment, which will target gulls. The primer pair we used in previous subtractive hybridizations does not amplify a band from gull fecal DNA, so we are trying different primer pairs. Also, since preliminary evidence suggests that Bacteroidetes in birds are different from those found in mammals, we might be able to design gull primers from the initial clone library.
We have designed three real-time quantitative-PCR (Q-PCR) nuclease (TaqMan) assays for Bacteroidetes fecal pollution: one that targets all Bacteroidetes (to measure total fecal pollution) and two that target human-specific Bacteroidetes. Our objectives were to develop a rapid method of water quality assessment, and to develop a method of quantifying the contribution of different sources to fecal pollution in water. The assays extensively have been tested with both cloned sequences and feces. The general Bacteroidetes assay also has been validated with cultured cells from B. ovatus and B. fragilis. However, none of the sequences used to design specific PCR primers are from cultured Bacteroides species. Therefore, we have established stable fecal bacterial cultures of each of the markers growing in consortia. We have developed fluorescence in situ hybridization (FISH) assays to quantify the number of marker cells in each consortium. This will allow us to validate the Q-PCR assays for the specific markers against cells.
We have begun experiments to calculate the proportional contribution of human fecal contamination to total fecal contamination in water samples. To do this, we will first establish, by Q-PCR, the ratio between a human-specific marker and the general Bacteroidetes marker in human fecal DNAs from sewage. Then, we can measure concentrations of the human marker and the general Bacteroidetes marker in a contaminated water sample with Q-PCR, and use the previously established ratio to solve the proportional contribution of human fecal contamination to total fecal contamination.
We currently are taking part in a national comparative study of fecal source detection methods, managed by the Southern California Coastal Water Research Project (SCCWRP). For the study, SCCWRP prepared sample sets consisting of various combinations of human, cow, dog, and gull feces diluted into several matrices, and distributed them along with reference samples to study participants. We have analyzed our samples with our available general and specific primers, and have been able to identify human, dog, and cow fecal contamination in the samples. Results of this study will be presented at a meeting in February 2003.
This year, we established and tested several new experimental protocols in the laboratory (subtractive hybridization, Q-PCR, FISH) and are entering into a period of rapid experimental progress as a result. We requested a no-cost extension to carry out further primer development and testing by means of subtractive hybridization, and to finish testing our method to calculate the proportional contribution of human fecal contamination to total fecal contamination in water samples.
Journal Articles on this Report : 2 Displayed | Download in RIS Format
|Other project views:||All 38 publications||10 publications in selected types||All 8 journal articles|
||Bernhard AE, Goyard T, Simonich MT, Field KG. Application of a rapid method for identifying fecal pollution sources in a multi-use estuary. Water Research 2003;37(4):909-913.||
||Bernhard AE, Colbert D, McManus J, Field KG. Microbial community dynamics based on 16S rRNA gene profiles in a Pacific Northwest estuary and its tributaries. FEMS Microbiology Ecology 2005;52(1):115-128.||