Final Report: Novel Antibiotic-Resistant Bacteria Formed in the Environment as a Result of Fecal Pollution

EPA Grant Number: CR830396
Title: Novel Antibiotic-Resistant Bacteria Formed in the Environment as a Result of Fecal Pollution
Investigators: Field, Katharine G.
Institution: Oregon State University
EPA Project Officer: Packard, Benjamin H
Project Period: January 1, 2002 through December 31, 2003
Project Amount: $207,483
RFA: Futures Research in Natural Sciences (2001) RFA Text |  Recipients Lists
Research Category: Hazardous Waste/Remediation , Ecological Indicators/Assessment/Restoration , Land and Waste Management


This research project was proposed after we measured the widespread occurrence of fecal Bacteroidales spp. in surface waters. Furthermore, we found that tetQ, a gene for tetracycline resistance commonly carried by Bacteroidales bacteria in the genera Bacteroides and Prevotella, is distributed widely in estuarine and river waters. We showed that laboratory strains of Bacteroides bacteria are able to transfer antibiotic resistance from one strain to another in 15° seawater microcosms. This suggested that these bacteria could provide an important source of antibiotic resistance, possibly leading to the development of novel tetracycline resistant bacteria through an environmental route.

The spread of antibiotic-resistant strains of pathogenic bacteria is a major concern worldwide and has led to a major reduction in the effectiveness of antibiotics. It is clear that the discovery of novel antibiotics is a short-term solution at best; the only practical solution is to interrupt the cycle of evolution of resistance. As part of this cycle, the release of antibiotic-resistant fecal bacteria into aquatic environments is an issue that may have grave consequences. Antibiotic resistance genes from fecal bacteria may pass into environmental bacteria and serve as a reservoir of resistance; alternately, resistant fecal bacteria in the environment may be acquired by other hosts.

Because Bacteroidales bacteria are obligate anaerobes, their survival in the extraintestinal environment is thought to be limited. For fecal Bacteroidales bacteriato serve as a source of tetracycline resistance in the environment, these bacteria would have to persist in the environment (once released from the original host) at least transiently. In addition, tetracycline-resistant Bacteroidales bacteria would either have to conjugate with bacteria under environmental conditions (fecal Bacteroidales acting as donors of antibiotic resistance genes), or Bacteroidales cells or their tetracycline resistance genes would have to be transferred into other (animal or human) hosts, where they could spread by conjugation or other mechanisms (fecal Bacteroidales acting as vectors for antibiotic resistance genes). The original objectives of the research project were to: (1) determine the survival and growth of Bacteroides in estuarine waters and suspended and bottom sediments; (2) measure the rate of conjugation with environmental bacteria under natural conditions; and (3) survey the current level of biopollution by Bacteroides and tetQ, its tetracycline resistance gene.

The scope of the project was reduced when key personnel left the principal investigator’s laboratory to accept permanent positions elsewhere. In addition, the initial results we obtained from our studies of tetQ occurrence in wild animal feces suggested that Bacteroidales bacteria might be more important as vectors, rather than donors, of antibiotic resistance genes; environmental conjugation might not be as important as direct horizontal transfer of fecal bacteria among host species. In a separately-funded study, we presented strong evidence for direct horizontal transfer of fecal bacteria among hosts (Dick, et al., 2005).

Because fewer personnel were available, we concentrated on the first and third objectives. For the first objective, we measured persistence and growth of fecal Bacteroidales under environmental conditions with bromodeoxyuridine (BrdU) labeling. The third objective was addressed by extensive measurements of Bacteroidales biopollution in the environment, measurement of the extent to which tetQ genes are found in wild animal species, phylogenetic analyses of tetQ gene sequences in wild animal feces to test whether they were likely to have originated from human and domestic animals,and comparison of the frequency of occurrence of tetQ in wild animal (coyote) feces in areas of high fecal contamination (suburban Oregon) and low fecal contamination (wilderness Alaska) of surface waters. The specific objectives of this research project were to: (1) measure survival and growth of fecal Bacteroidales by labeling with BrdU, a thymidine analog, separating BrdU-containing DNAs immunochemically, and analyzing labeled (growing) and unlabeled (surviving but not growing) DNAs with Bacteroidales-specific PCR primers; (2) survey the occurrence, frequency, and dynamics of fecal Bacteroidales throughout a watershed; (3) measure the distribution and phylogeny of tetQ in wild animal feces in Oregon; and (4) compare occurrence and phylogeny of tetQ in wild coyotes in Oregon and Alaska. Instead of a postdoctoral researcher as requested in the original grant proposal, the personnel accomplishing these objectives were a graduate student and two undergraduates. Two postdoctoral researchers paid on other grants also contributed to the project.

Summary/Accomplishments (Outputs/Outcomes):

Survival and growth of fecal Bacteroidales

We used a cultivation independent DNA labeling technique with BrdU to measure the persistence and proliferation of these bacteria. Uptake of BrdU and its subsequent incorporation into DNA can be used to identify actively growing bacteria. Separation of BrdU-labeled DNA using immunocapture, followed by PCR of ribosomal genes in the labeled fraction, allows species level identification of growing bacteria in mixed communities. Following immunocapture, successful PCR amplification of unlabeled (supernatant) DNA demonstrates persistence of targeted cells or DNAs, whereas PCR amplification of labeled (immunocaptured) DNA demonstrates their growth.

Because strains containing the fecal- and human-specific Bacteroidales markers have not been cultivated, we first tested the ability of Bacteroides vulgatus, the closest cultivated phylogenetic relative, to take up and incorporate BrdU into newly synthesized DNA. We successfully amplified B. vulgatus genes from immunocaptured DNA, indicating that B. vulgatus took up BrdU during growth, and incorporated it into newly synthesized DNA.

However, we obtained inconsistent results using 30 cycles of PCR to detect BrdU-labeled DNA. Frequently, amplification occurred in the immunocaptured fraction of the unlabeled control (Figure 1); if the antibody technique isolated BrdU-labeled DNA only, the immunocaptured unlabeled control should not have amplified.

Figure 1. Agarose Gel Electrophoresis Following 30 Cycles of 27F and 338R Primed PCR on Immunocaptured Fractions of B. vulgatus BrdU-labeled and Unlabeled DNA, Showing Robust Amplification of the Unlabeled Control (lane 9). Lanes: 1, 100 bp ladder as size standard; 2, PCR positive control; 3, PCR negative control; 4, No DNA bead supernatant; 5, No DNA immunocaptured fraction; 6, BrdU-labeled bead supernatant; 7, BrdU-labeled immunocaptured DNA; 8, unlabeled bead supernatant; 9, unlabeled immunocaptured DNA.

To test the ability of the antibody assay to separate BrdU-labeled from unlabeled DNA, we used a more sensitive technique, length-heterogeneity (LH)-PCR. Targets were B. vulgatus and Fulvimarina pelagi. These bacteria have a naturally occurring 37 base pair difference in length between 16S rRNA amplicons, allowing fluorescently-labeled amplicons to be separated and quantified on a DNA automated sequencer. First we showed that by using no more than 15 cycles of amplification, PCR products were recovered in the same proportions as their proportions in the initial template (Table 1).

Table 1. Ratios of Relative Fluorescence Units Resulting From LH-PCR Products, Following 15 and 20 Cycles of PCR, Where Known Template Ratios of Unlabeled F. pelagi and BrdU-Labeled B. vulgatus DNA Were Amplified Using the 27F and 338R Primer Pair

F. pelagi:B. vulgatus ratio in template

F. pelagi:B. vulgatus
ratio in PCR products
(15 cycles)

F. pelagi:B. vulgatus
ratio in PCR products
(20 cycles)










Second, we tested whether the BrdU immunocapture technique separated BrdU-labeled from unlabeled DNAs. Immunocapture followed by 15-cycle LH-PCR on bead supernatants, as well as immunocaptured fractions, revealed that immunomagnetic separation enriched for BrdU-labeled DNA (Figure 2). In repeated experiments, unlabeled F. pelagi DNA was never identified in the immunocaptured fraction when detected with 15 cycles of PCR. These results demonstrated that immunocapture was specific for BrdU-labeled DNA; however, at high cycle number unlabeled background DNA was occasionally amplified as a result of blocker leakage or PCR sensitivity. We eliminated amplification of background DNA using fluorescent detection of PCR amplicons following 15 cycles.

Figure 2. LH-PCR Electropherogram After 15 Cycles of PCR Depicts Unlabeled F. pelagi and BrdU-labeled B. vulgatus Proportions Before and After Immunocapture, When Amplified Using 27F and 338R.

Finally, we measured persistence and survival of Bacteroidales in sewage influent with the BrdU immunocapture technique. We detected BrdU-labeled Bacteroidales DNA at 4, 8, 12, and 24 hours, indicating growth of fecal Bacteroidales cells occurred for at least 24 hours. The Bacteroidales molecular signal also persisted in the unlabeled DNA fraction over the entire 24-hour time course. However, fluorescent fragment analysis of PCR products obtained following 15 cycles did not detect Bacteroidales human-specific markers in either supernatants or immunocaptured fractions, even when using ten times the concentration of labeled sewage DNA in the immunocapture. Either the human-specific markers did not persist or grow, or fluorescent detection was not sensitive enough using 15 cycles of PCR.

The results of this study were presented at the American Society for Microbiology General Meeting (2004) and recently published in Applied and Environmental Microbiology. They may have significant impacts for microbial ecology studies using PCR of immunoseparated, BrdU-labeled DNA to detect active populations. Because of product inhibition, an unequal proportion of amplicon concentration to initial template concentration can result from high cycle number PCR. Unlabeled background DNA amplicons may reach a detectable concentration and appear as active members of a population or community. Inaccurate conclusions may be drawn about metabolic or biogeochemical processes within a population or community. The sensitivity of PCR is so great that to overcome the problem of amplifying unlabeled background DNA, a quantitative assay should be incorporated into this protocol.

In addition, we demonstrated that Bacteroidales fecal bacteria grew up to 24 hours in sewage, when incubated aerobically at the in situ temperature of sewage, and that the genetic markers for these bacteria persisted at least 24 hours under the same conditions. This is in agreement with our ability to detect these bacteria in surface waters (see below), and suggests that they are indeed available as environmental donors or vectors for antibiotic resistance genes.

To follow these studies, we are using RNA and DNA hybridization techniques and Q-PCR to quantitatively survey the persistence and growth of host-specific Bacteroidales bacteria in mesocosms simulating a variety of environmental parameters. Data from the mesocosm studies is still being analyzed and publication is in preparation.

Occurrence and Dynamics of Fecal Bacteroidales Throughout a Watershed

Sampling occurred over a 2-year period in the Tillamook basin in Oregon, at 30 sites along 5 river tributaries and in the Tillamook Bay. We performed Bacteroidales PCR assays with general, ruminant, and human Bacteroidales primers; compared occurrence of Bacteroidales to fecal indicator counts, general measurements of water quality, and climatic forces; and identified sites of intense exposure to specific sources of contamination.

Results of Quality Control/Quality Assurance Assays. Initial analysis of water samples in this study yielded both positive and negative results for each genetic marker, indicating that the 100 mL sample volume established in previous studies allowed for the isolation of detectable and nondetectable quantities of target DNA without coextracting other substances that inhibit the PCR assay. In 1,258 no-template and extraction blank PCR control reactions 1,255 (99.8%) were negative. We used Chi square tests to analyze 1 year of triplicate samples to test whether replication significantly increased the frequency of detection of Bacteroidales and found no significant increase in frequencies of detection when replicate samples were analyzed.

Logistic regression supported tighter linkage between Escherichia coli count and ruminant Bacteroidales than human Bacteroidales. The probability of detecting ruminant Bacteroidales was relatively high across all water bodies in the basin. The probability of detecting ruminant Bacteroidales fell to a minimum from April to July and then increased in the fall. For the least affected rivers, logistic models were best fit with a quadratic term, indicating a tendency toward lower detection probabilities during extremes of precipitation. This suggests a source loading potential consistent with rainfall runoff models, in which pollutants build up in the landscape between rain events and are washed off during subsequent precipitation. In waters where point sources were indicatedthe quadratic behavior was lost for both ruminant and human Bacteroidales, and trends were positively correlated with rainfall over the entire rainfall distribution.

Sites were identified in the watershed that had significantly higher or lower occurrence of human and ruminant Bacteroidales (Figure 3). Ruminant Bacteroidales was higher in all but one sampling site. This figure also illustrates a trend toward increasing frequencies of both human and ruminant Bacteroidales in a downstream direction in the watershed. In the part of the bay closest to the river outlets (Tillamook Bay stations 1 and 2), frequencies of occurrence of both ruminant and human Bacteroidales levels were very high; in most parts of the watershed, however, one or the other type predominated, indicating very specific sources of Bacteroidales antibiotic resistance at most sites.

Figure 3. Geometric Mean E. coli Count (Lines), Ruminant Bacteroidales (Gray Bars) and Human Bacteroidales (Black Bars) Frequencies Plotted by Water Body and Sampling Station (Sampling Stations Numbered From Upstream to Downstream). Error bars represent 95% confidence intervals. Asterisks indicate significantly greater or lower marker frequency than river-wide frequencies (0.75 and 0.35 for ruminant and human Bacteroidales, respectively) at p <0.05 in Chi-square tests.

Results of this study were presented at the American Society for Microbiology General Meeting (2006) and recently published in Applied and Environmental Microbiology.

Distribution and Phylogeny of tetQ in Wild Animal Feces

Antibiotic resistant bacteria often are assumed to arise under selective pressure from exposure to low levels of antibiotics. Our purpose was to investigate the extent to which antibiotic resistance can spread independent of exposure to antibiotics. To that end, we investigated the occurrence of tetracycline resistance in wild animals. These animals are likely to have been exposed to antibiotic-resistant fecal bacteria in surface waters but not to have been exposed to antibiotics.

We obtained fresh samples of wild animal feces from near Corvallis, Oregon, and from the Oregon Coast Range, from Chintimini Wildlife Rehabilitation Center, and from donations from hunters. None of the animals had been given antibiotics; the rehabilitation center samples were collected from animals immediately after their arrival to the center. We extracted fecal DNAs and detected the presence of tetQ by PCR.

The following animal species contained fecal tetQ sequences (frequency of occurence of tetQ in individual samples in parentheses): elk (100%), red fox (100%), coyote (100%), black tail deer (50%), opossum (100%), pheasant (100%), squirrel (100%), and raccoon (50%). Seal, mallard, pigeon, bobcat, and skunk did not contain tetQ sequences in their feces.

We obtained 10 tetQ sequences from these samples. Sequence data revealed two different tetQ sequence types. The first of the two sequence types had greater than 97 percent sequence identity to a Bacteroides conjugative transposon associated tetQ (CTnDOT, from Bacteroides fragilis from a human source). Four wild animal fecal sequences were homologous to the CTnDOT-associated tetQ. These came from two different animal species: coyote and black tail deer. The second of the sequence types had greater than 97 percent sequence identity to a Prevotella ruminicola plasmid (pRRI4) associated tetQ (from a ruminant source). Six sequences were homologous to the pRRI4 associated tetQ. These came from three different species: coyote, opossum, and red fox.

High sequence identity among wild animal fecal sequences and known tetQ gene sequences suggested that the genes were acquired recently in the gut microbiota of the wild animals. Because fecal samples were collected from animals immediately after arrival at the rehabilitation facility, the animals are likely to have acquired the antibiotic resistance genes in their local environment. Because research suggests that even in environments likely to be contaminated by animal manure or sewage, tetracycline is not detectable in groundwater, and animals sampled had not been fed antibiotics at the rehabilitation facility, the animals appear to bear antibiotic resistance genes in the absence of exposure to antibiotics. The most likely path of exposure is through fecal contamination of surface waters. These animals were all from populous rural areas with many farms, commercial establishments, and housing developments as well as septic tanks and sewage treatment plants. All of these are potential outlets for fecal contamination that could seep into the waterways. Based on the sequence of the resistance genes found in the wild animals it is likely that human and ruminant fecal Bacteroidales were the source. The CTnDOT-like tetQ genes very likely are because of human fecal contamination and the pRRI4-like tetQ genes likely are caused by ruminant (cow) fecal contamination.

This work, part of an undergraduate Microbiology research project, was presented at the American Society for Microbiology General Meeting in 2003 and is being prepared for publication.

Occurrence and Phylogeny of tetQ in Wild Coyotes in Oregon and Alaska

To test our hypothesis that animals are acquiring tetQ from fecally contaminated surface waters, we compared the occurrence of tetQ in wild animal feces from a contaminated (Willamette Valley, Oregon) and uncontaminated (Alaskan Range) area. We hypothesize that local wild animals are acquiring tetQ from contaminated surface waters. If so, then tetQ should be present in local surface waters. In addition, if wild animals are acquiring tetQ from contaminated surface waters, then wild animals living in areas free from contaminated surface waters should not carry the gene. We focused on feces from coyote (Canis latrans), an animal species that occurs in both the Willamette Valley and the Alaska Range areas. We obtained fecal and water samples from these two areas. Of the Oregon coyote fecal samples, 77.8 percent were positive for tetQ. Of the Alaskan coyote fecal samples, 46 percent were positive for tetQ.

This part of the project was accomplished by an undergraduate as part of a microbiology honors project. We are in the process of finishing the analyses of Alaskan and Oregon water samples, completing statistical analyses, obtaining tetQ sequences from the samples, and preparing the results for publication.


The overall import of this research is that tetracycline resistance from fecal bacteria is found in the environment and appears to be spreading into wild animals, supporting an environmental pathway for the spread of antibiotic resistance independent of exposure to antibiotics. This research has demonstrated that fecal Bacteroidales, and the tetracycline resistance genes they carry, are widespread in surface waters of an Oregon coastal watershed, including in rivers and in an estuary. Occurrence of the fecal bacteria in water is seasonal and related to rainfall. In upstream stations and less fecally polluted waters, the pattern of contamination follows a rainfall-runoff distribution pattern. In downstream stations and more fecally polluted waters, where levels of contamination are high, contamination is correlated positively with rainfall and can be directly related to sources such as sewage treatment plants and confined animal feedlot operations. Laboratory and mesocosm experiments suggest that Bacteroidales fecal bacteria survive in the environment for far longer than one might expect of strictly anaerobic bacteria. BrdU uptake data demonstrate growth of Bacteroidales cells during aerobic incubation of sewage influent, suggesting that Bacteroidales fecal bacteria may be able to persist and grow in low oxygen refugia within streams, lakes, estuaries, and bays. tetQ, a gene for tetracycline resistance found in fecal Bacteroidales, also was widespread in the environment and was recovered from feces of eight species of wild animals that had not been exposed to tetracycline. Sequence evidence supports a recent acquisition of tetQ in the animals by horizontal transfer from human and ruminant fecal bacteria. A comparison of the occurrence of tetQ in the feces of wild coyotes from Oregon and Alaska showed that the gene was significantly more common in the Oregon animals than in the Alaska animals.

The usual concern from fecal contamination in water is potential exposure to pathogens. Until recently, the focus of concern for water-related illness has been contamination by human effluent, and most epidemiological studies have been conducted in locations where human sewage was the predominant contamination source. Nonhuman host-animal feces can spread many pathogens and also may serve as a source for emergent zoonotic disease. This research demonstrates that the spread of antibiotic resistance genes by fecal bacteria from both humans and domestic animal sources must also be a human and animal health concern. Because fecal Bacteroidales can be detected in the environment, and in most cases their host species of origin can be assigned, it might be possible to assign or estimate risk of antibiotic resistance from fecal contamination based on knowledge of its source.


Dick, et al. Applied and Environmental Microbiology 2005;71:3184-3191.

Journal Articles on this Report : 3 Displayed | Download in RIS Format

Other project views: All 8 publications 3 publications in selected types All 3 journal articles
Type Citation Project Document Sources
Journal Article 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. CR830396 (Final)
R827639 (2001)
R827639 (2002)
R827639 (Final)
  • Abstract from PubMed
  • Full-text: ScienceDirect - Full Text HMTL
  • Abstract: ScienceDirect - Abstract
  • Other: ScienceDirect - Full Text PDF
  • Journal Article Shanks OC, Nietch C, Simonich M, Younger M, Reynolds D, Field KG. Basin-wide analysis of the dynamics of fecal contamination and fecal source identification in Tillamook Bay, Oregon. Applied and Environmental Microbiology 2006;72(8):5537-5546. CR830396 (Final)
  • Abstract from PubMed
  • Full-text: AEM Full Text
  • Other: AEM PDF
  • Journal Article Walters SP, Field KG. Persistence and growth of fecal Bacteroidales assessed by bromodeoxyuridine immunocapture. Applied and Environmental Microbiology 2006;72(7):4532-4539. CR830396 (Final)
  • Abstract from PubMed
  • Full-text: AEM Full Text
  • Other: AEM PDF
  • Supplemental Keywords:

    risk assessment, human health, antibiotic resistance, fecal contamination, water quality, environmental management, biochemistry, estuarine research, exposure assessment, ecology and ecosystems,, Scientific Discipline, Health, PHYSICAL ASPECTS, ENVIRONMENTAL MANAGEMENT, Water, Ecosystem Protection/Environmental Exposure & Risk, estuarine research, Health Risk Assessment, Risk Assessments, Environmental Monitoring, Physical Processes, Ecological Risk Assessment, Ecology and Ecosystems, Risk Assessment, exposure, biopollution, public health, human exposure, estuarine waters, Bacteroids, water quality, antibiotic resistant bacteria, fecal polllution, exposure assessment, human health risk

    Progress and Final Reports:

    Original Abstract
  • 2002