Abstract Mammalian Derived Ruminococcaceae

ABSTRACT

MAMMALIAN DERIVED RUMINOCOCCACEAE POPULATION AS MOLECULAR TARGET OF FECAL CONTAMINATION

By Nilay J. Sheth

Domesticated animal’s fecal material, as agricultural runoff is a contributor to decreased water quality and subsequent, human exposure to waterborne illness. Fecal contamination of water sources is currently monitored by bacteria indicators such as Escherichia coli (E. coli), enterococci, or fecal coliforms. However, these bacteria typically comprise less than 1% of the intestinal bacterial population, and may not precisely predict the human health risk associated with fecal contamination. The mammalian intestinal tract is dominated by the bacterial phylum Firmicutes and a major population is the family Ruminococcaceae. The main objective of this research was to determine if animal derived Ruminococcaceae can be used as an alternative molecular indicator target for the detection of fecal pollution in surface water. To address this objective a 16S rRNA molecular target for mammalian specific Ruminococcaceae was created using the characterized diversity of rumen gastrointestinal/fecal microbial community. The Ruminococcaceae target was amplified from the fecal DNA samples of dairy cow, pig, goat, camel, bison, and water buffalo. The target population was not amplified from horse, chicken, goose, and human sewage influent DNA samples. In natural water spiked with fresh cow fecal material; the culturable indicator E. coli was detected for four days as compared to one day post inoculation, illustrating rapid signal loss. Following rain events, this population was detected in surface water samples from Spring Brook creek (n=10) and only one sample from Fox River, indicating this creek receives animal fecal pollution. The bacterial community analysis of Spring Brook and Fox River water samples demonstrated an increased diversity in the Spring Brook samples. This is most likely attributed to the surrounding agricultural area, and may suggest a significant input of agricultural related fecal pollution. Overall, this molecular target may prove useful in detecting mammalian fecal contaminations, specially that of agricultural.

ACKNOWLEDGMENTS

First of all, I would like to thank my advisor, Dr. Sabrina Mueller-Spitz, for encouraging me and allowing me to work in her lab. Thank you for all your guidance, support, teaching, and encouragement for past three years. I would not be where I am today if it were not for you. I would also like to thank my committee member, Dr.

Colleen McDermott, for her guidance and being there for me and answering my many questions. I also would like thank Dr. McDermott for helping me obtain dairy cow fecal samples which I was able to use in different aspects of my research. I would like to acknowledge my committee member, Dr. Toivo Kallas, for his guidance and support that played an important part in my development as a scientist. Dr. Kallas always welcomes questions and has played a vital role in completion of my thesis. I also want to thank numerous people for allowing me to come to their personal farm and letting me collect fecal samples from various different animals (who have requested to be kept anonymous). This has helped me fulfill my research and understandings to completion.

There are not enough words to thank my family whose support and guidance has helped me to get where I am. They gave remarkable amount of time and personal sacrifice so that I could work towards my goals in life.

TABLE OF CONTENTS

Pages

LIST OF TABLES...... v

LIST OF FIGURES...... vii

INTRODUCTION...... 1

Global Perspective on Water Pollution………………………… 1 Sources of Water Pollution…………………………………….. 2 Fecal Indicator Bacteria……………………………………….. 3 Regulations for Water Quality………………………………… 5 Molecular Methods for Fecal Source Tracking……………….. 6 Bacterial Diversity in Mammalian GI tract…………………… 7

PROJECT FRAME WORK…………………………………………… 11

PROJECT OBJECTIVES……………………………………………… 12

MATERIALS AND METHODS...... 13

Creation of Ruminococcaceae 16S rRNA Primers………… …. 13 Collection of Animals Fecal Samples………………………….. 17 Bacterial DNA Extraction and Concentration…………………. 18 Determination of Optimal Annealing Temperature……………. 18 Detection of Fecal Ruminococcaceae and Bacteroides Populations…………………………………………………….. 19 PCR Cloning of 16S rRNA Gene Sequences…………………. 20 Plasmid Extraction for Ruminococcaceae Animal Clones……. 20 DNA Concentrations and Confirmation of Extracted Plasmids… 21 DNA Sequencing and Analysis of Clones…………………… 21 Development of Real-time PCR Analysis for Ruminococcaceae……………………………………………. 23 Microcosm Experiment to Detect Signal Loss of E. coli And Ruminococcaceae………………………………………….. 25 Characterization of Fecal Pollution in Upper Fox Watershed….. 26 Bacterial Community Analysis of Surface Waters……………… 27

RESULTS………………………………………………………………. 29

Specificity of Ruminococcaceae Primer Set……………………. 29 Sensitivity of Ruminococcaceae Primer Set……………………. 34

iii

TABLE OF CONTENTS (Continued)

Pages

Gene Sequence Analysis of PCR Ruminococcaceae 16S rRNA Animal Clones…………………………………………………... 34 Microcosm Experiment to Detect Signal Loss of E. coli and Ruminococcaceae...... 39 Field Test of Ruminococcaceae as Fecal Pollution Target……… 41 Bacterial Community Analysis………………………………….. 46 Bacterial Populations Associated with Fecal Signature of Mammals……………………………………………………… 52

DISCUSSION………………………………………………………….. 56

APPENDIX…………………………………………………………….. 62

REFERENCES………………………………………………………… 79

LIST OF TABLES

Page

Table 1. Twelve out groups within the known Clostridiales ……….. 15

Table 2. Sequences and the annealing temperatures of the primers used in this study…………………………………………… 16

Table 3. Mammalian specific Ruminococcaceae 16S rRNA primer sequences ………………………………………………….... 31

Table 4. The predicated overall genera for the family of Ruminococcaceae sequences for 120RF, CF and 430R primer Sequences…………………………………………………... . 32

Table 5. Detection of Ruminococcaceae in various agricultural animals and human fecal material……………………………………. 33

Table 6. Diversity of Ruminococcaceae population targeted with 120RF/430R primer set……………………………………....… 36

Table 7. The predicated genera for Ruminococcaceae sequences at 95% similarity level for cow, goat and pig clone library origin using RDP classifier …………………………………… 36

Table 8. Ruminant animal sources matching cow, goat and pig fecal clones ……………………………………………….. 37

Table 9. Non-ruminant animal sources matching cow, goat and pig fecal clones …………………………………………... 38

Table 10. Detecting signal loss of fecal indicators in seven day microcosm … 40

Table 11. Spring Brook samples were analyzed for average E. coli count and real-time SYBR green PCR of Ruminococcaceae target………………………………………… 42

LIST OF TABLES (Continued)

Page

Table 12. Fox River samples were analyzed for average E. coli count and SYBR green real-time PCR of Ruminococcaceae target………………………………………………………… 44

Table 13. Average E. coli for the five different water samples used for pyrosequence analysis ………………………………………… 51

Table 14. Number of sequences of Gram positive species commonly identified within the GI tract of mammals from five different surface water samples. …………………………………………. 53

Table 15. Number of sequences of Gram negative species commonly identified within the GI tract of mammals from five different surface water samples. ………………………………………… 54

LIST OF FIGURES

Page

Figure 1. The major steps in the CAP3 assembly algorithm…………….. 14

Figure 2. Standard curve for SYBR Green real-time qPCR analyses of cowB1 plasmid clone……………………………… 24

Figure 3. Map of the lower Fox basin (Oshkosh, WI)…………………… 26

Figure 4. The correlation between rainfall totals and culturable E. coli levels for Spring Brook surface water samples………… 43

Figure 5. Relationship between Ruminococcaceae abundance and culturable E. coli levels in Spring Brook surface water samples……..………………………………………………….. 43

Figure 6. The correlation between rainfall totals and culturable E. coli levels for Fox River surface water samples……………. 44

Figure 7. The correlation between Spring Brook and the Fox River E. coli levels for surface water samples ………………………. 45

Figure 8. Relative abundance of dominant bacteria phylum from five different water sample sequences…………………………… 48

Figure 9. Rarefaction analysis to determine species richness and diversity………………………………………………………… 49

Figure 10. Principal coordinate plot generated utilizing MG-RAST analysis comparing five different sample dates………………… 50

Figure 11. Principal component analysis using UniFrac metric to analyze fecal population cluster sequence diversity ………………………. 55

vii

INTRODUCTION

Global Perspective on Water Pollution

Currently, 1.1 billion people lack access to safe water and over 2.6 billion people do not have proper sanitation. These conditions exist primarily in developing countries, leading to unsafe drinking water (90). Globally, the restricted access to safe drinking water results in 1.6 million deaths per year, with more than 99% of those deaths occurring in developing countries (109). These easily preventable diarrheal diseases contribute to 6.1% of all health-related deaths (3). The World Health Organization

(WHO) estimates that there are four billion cases of waterborne diarrhea each year associated with the lack of access to clean drinking water (15). The main risk is associated with fecal pollution contaminating drinking water and ingestion of pathogens.

The human pathogens of concern include viral, bacterial, and protozoan agents, which are spread via the fecal-oral route (3). To decrease the human health risk from poor water quality and limited sanitation, the WHO and the United Nations Children’s Fund have launched as a millennium development goal to halve the population without access to safe drinking water and basic sanitation by 2015 (110).

In the United States, fecal pollution is also one of the top three contaminants impacting the quality of recreational waters and drinking water sources (101). Most bodies of water that exist as drinking water supplies and for recreational activities in the

USA are prone to microbial contamination. In the 2002, the National Water Quality

Inventory Report of the U.S. Environmental Protection Agency, 32-47% of the streams,

rivers, lakes, ponds, bays and estuaries were determine not clean enough to support their designated uses because of presence of a fecal pollution (101). Accidental contact with contaminated water can result in diarrhea, vomiting, and other GI illnesses, which can be life threatening for populations such as the elderly, children, and those with impaired immune systems (101).

Sources of Water Pollution

Waterborne diseases are spread through water containing human or animal feces.

The fecal material can originate from point and non-point source discharges. Point sources of fecal pollution include stormwater, raw sewage from sewer overflows, and effluent from wastewater treatment plants (40). These sources can release human pathogens into surface waters, creating serious public health risk and environmental concerns. If the entry point of contamination is not obvious, the source is categorized as non-point sources, which includes wildlife (Deer) and domesticated animals (Dairy cow, pig, goat etc.) fecal material, as agricultural and urban runoff (40, 92).

Agricultural pollution can increase the loading rates of pathogens to soils and waters from agricultural related animal operations, such as grazing (39). The gastrointestinal (GI) tract microbiota of cattle can affect human health. The presence of pathogenic bacteria such as E. coli 0157:H7 in the bovine gastrointestinal tract has been linked to disease outbreaks due to consumption of contaminated beef, milk, and drinking water (5). Modern dairy farms generate large amounts of animal waste and can house more than 1,000 cows, each producing approximately 35 kg of manure a day (74).

Hence, dairy farms could produce approximately 35,000 kg of manure per day and more than 12,000,000 kg per year (69). In comparison, humans contribute 0.7% of the feces in the U.S. while dairy cows produce 10% and beef cattle produce about 44% of the total feces (30). This amount of animal manure is of concern when considering the potential for pollution associated with agricultural runoff. Wisconsin has approximately 78,000 farms and over 1.2 million dairy cows, so agricultural animals could be a major cause of decreased water quality (18). The EPA states that agriculture has a greater impact on stream and river contamination than any other non-point source. Negative impacts downstream can include the potential contamination of drinking water supplies (14).

Fecal Indicator Bacteria

Public health concerns associated with water-related illness in humans are partially due to bacterial pathogens (e.g. enteropathogenic Escherichia coli, Shigella spp,

Vibrio cholerae). However, these pathogens in surface water are present in low numbers and can be difficult to culture (21). Fecal contamination is commonly monitored by using culture based methods to detect and enumerate facultative-anaerobic bacteria associated with feces of warm blooded animals, such as Escherichia coli, enterococci, or total coliforms (32, 71). Total coliforms include bacteria that are found in human feces, but some can also be present in animal manure, soil, and submerged wood or from other places outside the human body (2). A subgroup of total coliforms, called fecal coliform bacteria, are different from the total coliform group because they grow at higher temperatures, and are found only in the fecal waste of warm-blooded animals (2). The

fecal coliform includes both pathogenic and nonpathogenic bacteria (111). An example of a fecal coliform bacterium is E. coli. The presence of fecal coliforms is indicative of fecal contamination and subsequently the potential presence of enteric pathogens (111).

Some of the anaerobic bacteria are essential component of the fecal flora. Anaerobic bacteria such as Clostridium perfringens, Bacteroides spp., and Bifidobacteria have been proposed as indicators of fecal contamination (2, 12). Clostridium perfringens also is associated with the feces of warm-blooded animals (12). Although C. perfringens has been considered a useful fecal indicator species for several years, its use largely has been limited (12). The main concern over its use as a fecal indicator is its long persistence in the environment, which is considered to be considerably longer than enteric pathogens

(12). Until bifidobacteria were suggested as fecal indicators, C. perfringens was the only obligately anaerobic, enteric bacterium used as a possible indicator of the sanitary quality of water (12). However, because bifidobacteria are highly oxygen sensitive, there has been a limited use for them as fecal indicator in surface waters. Although the anaerobic clostridia are much less prevalent than bifidobacteria in feces of human, the spore- forming capability of the clostridia makes them highly environmental resistant (12).

Regulations for Water Quality

In 1986, the United States Environmental Protection Agency (EPA) developed water quality criteria for recreational waters based on indicator bacteria. These criteria were based on public health studies conducted in the 1950s-1980s. EPA recommended the use of two indicators: enterococci for marine waters and E. coil or enterococci for freshwaters (100). Enterococci and E. coli are used as water quality indicators because they have been linked to human gastroenteritis through the use of epidemiological studies

(100, 103). The monitoring of both E. coli and enterococci has higher correlation with human pathogens than with either organism alone (100). However, humans and animals species contain both different numbers and different ratios of E. coli and enterococci in their feces (99). In light of this risk to human health, it is important to identify the source of fecal contamination to better facilitate resource management and remediation.

To summarize, there are about seven criteria for ideal indicator organisms (11): 1. It should be a member of the intestinal microflora of warm-blooded animals. 2. It should not multiply in the environment. 3. It should be present when pathogens are present, and absent in uncontaminated samples. 4. It should be present in greater numbers than the pathogen. 5. It should be at least equally resistant as the pathogen to environmental factors 6. It should be detectable by means of easy, rapid, and inexpensive methods. 7. The indicator organism should be non pathogenic

The species traditionally used as fecal indicators have multiple limitations. The fecal indicator organism should have similar survival and transport properties as pathogens known to cause disease (21). Second, the fecal indicator bacteria also should not reproduce outside the animal host. E. coli and enterococci are able to survive, grow, and establish populations in natural environments such as fresh water lakes and streams,

beach sand, soils and sediments (30). The survival rates and re-growth potential of some bacteria vary depending on water temperature, sunlight, nutrient status, and turbidity

(30). The fecal indicator organism should be correlated with the presence of pathogens; and it should have a survival similarity to the pathogens whose presence it indicates. E. coli and enterococci are not well correlated with pathogenic Salmonella spp.

Campylobacter spp. Cryptosporidium, Giaridia spp. or human entroviruses (30).

Connections between high levels of enterococci and E. coli and disease outbreaks have been supported by research findings (19). However, there is little or no predictive value in determining the specific sources of pollution or correlating human pathogens with these bacterial indicators (91).

Molecular Methods for Fecal Source Tracking

The inability to differentiate between human fecal and animal fecal contamination using culturable indicators requires alternative microbial source tracking targets to identify contamination in surface water. Currently, monitoring of surface water for fecal indicator bacteria is done using culture based technology that can require more than a day for laboratory analysis. These culture methods are time-consuming and have not detected approximately 99% of the bacteria within the gut (76).

A basic concept of Microbial Source Tracking (MST) is the assumption that some attribute in or associated with feces identifies a particular fecal type or host source; and that this identifying trait can be used to detect the population in water (30). Molecular methods for source tracking as become the focus to pinpoint the source of contamination

in surface water, which permit the direct measurement of the organisms without incubation and have the possibility to decrease the detection period to less than an hour.

These technologies provide a method which increases the number and types of microbiological indicator that can be used to detect fecal contamination in water.

Bacterial Diversity in Mammalian GI tract

The need to differentiate between human and nonhuman source of fecal pollution has stimulated the search for species-specific indicators. Most relevant to microbial source tracking is the analysis of gastrointestinal microbes, which indicates that over 500 different species of bacteria may be found in mammalian intestines (114). These microbial populations are in order of 1011 g-1 of luminal contents (114). However, these bacteria are not readily cultured in the laboratory, which has previously limited their use as fecal indicators. But the incorporation of molecular methods for MST allows for their detection (30).

A number of studies have outlined specific microbial population within GI tract of different mammals (16, 26, 32). These studies indicate that microbial populations are dominated by strict anaerobes such as Clostridium spp., Bacteroides spp., and

Bifidobacterium spp. while facultative anaerobes, such as the Enterobacteriaceae (e.g. E. coli), are typically reported to occur in numbers at least 100-fold lower than the strict anaerobes (26, 32). The bacterial order of Bacteroidales are found in large numbers in mammalian intestines and have been shown to correlate with the presence of fecal pathogens such as Escherichia coli O157:H7, Salmonella and Campylobacter (105).

Fecal anaerobes, Bacteroides spp. have been suggested as alternative indicators to the fecal coliform group, because they are present at 1,000-fold higher densities in the GI tract and are less prone to survive in the environment (10, 13). Several studies have suggested that some species in the genus Bacteroides have host-specific distributions (9,

53). Bernhard and Field developed a method to detect Bacteroides of human and ruminant specific 16S rRNA gene sequences (9). Detection of these bacteria allows for a highly sensitive indicator of fecal pollution because there is little possibility for growth or persistence for long time in the environment (54).

The use of the molecular techniques has been utilized to amplify and detect the

16S rRNA genes of Bacteroides spp. from water and feces (9, 23, 53, 55). Numerous studies have focused on detecting Bacteroidales found in human, bovine, swine and equine sources (9, 31). Swine-associated molecular methods for targeting Prevotella species have been proposed to discriminate swine fecal contamination from other possible fecal contributions (i.e., bovine and human) (23, 78). Similarly, bovine- associated Bacteroides 16S rRNA PCR-based methods have been proposed to discriminate bovine fecal pollution from other possible pollutants (9, 93).

Characterizations of Bacteroidales in other host sources of fecal pollution have included cats, dogs, geese, horses, and seagulls (23, 31). Jeter et al. (46) identified low occurrence of Bacteroidales from gulls at each beach tested from Lake Michigan beaches. The progression of molecular target detection analysis for Bacteroidales has provided a reliable and fast means by which to identify the sources of fecal contamination. Culture and culture-independent methods have revealed that Bacteroides species account for only

20 to 52% of the human fecal flora (27, 43). The 16S rRNA gene clone libraries indicate slightly less Bacteroides in nonhuman hosts (10, 24, 59). In general, animal and human intestinal microflora has identified Firmicutes in higher abundance then Bacteroidetes.

However, the dominant of the bacterial phyla Firmicutes as fecal indicator bacteria remains underutilized.

Ley et al. (60) demonstrated that humans and 59 other mammalian species have

17 bacteria phyla in common. The majority of sequences belong to the Firmicutes

(65.7%), Bacteroidetes (16.3%) and to the Proteobacteria (8.8%) (60). Of the phyla detected, only the Firmicutes were detected in all samples. In comparative assessment of human and farm animal fecal microbiota using real-time qPCR, the most commonly represented bacterial groups in human stool were Clostridium leptum and the Clostridium coccoides groups of the Firmicutes (32). However, the presence of C. leptum group significantly differed between human and horse, cow, goat and sheep microbiota (32).

Firmicutes have been detected at high levels in both human feces (46-58%) and the distal gut (47-74%) (36, 115). Eckburg et al. (28) demonstrated that Firmicutes corresponded to over 70% of the phylotypes of the human intestinal microbial flora.

McLellan et al. (70) showed that members of Firmicutes, Bacteroidetes, and

Actinobacteria accounted for 37.5% of the sewage dataset. However, it accounted for

97% of the tags in the human fecal samples. The surface water showed taxonomic profile that differed from both sewage and human fecal samples. Betaproteobacteria and

Alphaproteobacteria (51% and 19%) dominated the surface water microbial population.

However, Firmicutes were less than 0.5% of the surface water tags (70).

The phylum Firmicutes has populations that are present at higher abundance in other animal’s GI tract in comparison to humans. The bacterial diversity of bovine rumen fecal samples commonly detected Ruminococcaceae as the dominant family (16,

82, 94). The family of Ruminococcaceae is more abundant (37%) in bovine GI tract flora compared to human (5%) (28, 82).

The family of Ruminococcaceae is also illustrated as a dominant bacterial population of other ruminant animals (59, 88). In studies which looked at fecal bacterial community of Holstein cattle, pig, and dromedary camel determined that the majority of the sequences belonged to the phyla Firmicutes (81.3%, 80%, and 48%) and the family of

Ruminococcaceae were 41.7%, 48%, and 35% of those sequences (59, 80, 88). Various studies have now characterized Ruminococcaceae as the dominant family within GI tract of ruminant animals (16, 51, 59, 88). However, the use of Ruminococcaceae as fecal indicator bacteria remains underutilized and the high abundance within GI tract of ruminant animals makes it a potential target for a molecular detection of fecal pollution.

PROJECT FRAME WORK

Bacterial diversity studies based on 16S rRNA gene sequences of bovine rumen and fecal samples have characterized Ruminococcaceae, phylum Firmicutes as the dominant family (16, 82, 94). Members of the taxa from Ruminococcaceae are also common within the gastrointestinal tracts and feces of other mammals (e.g. cow, goat, and pig) (16, 25, 60). We propose that mammalian derived Ruminococcaceae could be utilized as an alternative molecular target for detection of agricultural related fecal pollution. The hypothesis tested in this research is: Agriculturally derived fecal pollution from multiple sources can be detected in recreational surface waters by specifically targeting members of mammalian fecal Ruminococcaceae.

PROJECT OBJECTIVES

The main objective of this research is to determine if mammalian derived Ruminococcaceae can be used as an alternative indicator for the detection of fecal pollution from ruminant in surface water. Specifically, this study addressed the following:

1. Utilize diversity of rumen GI/fecal microbial community data to create Ruminococcaceae (family) specific PCR 16S rRNA gene primers.

2. Test specificity of Ruminococcaceae primer set(s) against various animal sources.

3. Characterize diversity of Ruminococcaceae amplified from cow, pig, and goat fecal sample.

4. Compare specificity and sensitivity of Ruminococcaceae target with previously generated molecular fecal targets for detecting fecal pollution.

5. Validate the use of Ruminococcaceae as a target by bacterial community composition using pyrosequencing (n = 5); Characterized any populations that could be associated with fecal pollution.

6. Compare culture based (E. coli) against molecular Ruminococcaceae and cow-specific Bacteroides for detecting fecal pollution.

MATERIALS AND METHODS

Creation of Ruminococcaceae 16S rRNA Gene Primers

Bovine gastrointestinal specific (n=692), bacterial Ruminococcaceae 16S rRNA gene sequences from two separate studies (16, 82) were used as the template to create

Ruminococcaceae specific primers. The sequences were aligned using BioEdit sequence alignment editor (41) to determine a unique region specific to Ruminococcaceae within the 16S rRNA gene. The target sequences were further aligned with 76 Clostridiales named sequences which were selected from the Ribosomal Database Project (RDP) browser (56, 106). The total 768 sequences were analyzed using Contig Assembly

Program (CAP3) (44). The assembly algorithm has three major phases to generate consensus sequences, as summarized in Figure 1. The CAP3 algorithm reduced the 768 sequences to 105 sequences based on 95% sequence similarity (family level cut-off based on 16S rRNA gene). These 105 sequences were aligned with 12 Clostridiales named sequences as out-groups for comparison to find unique sequence regions within the family Ruminococcaceae (Table 1).

Three Ruminococcaceae specific primers sequences were created: two forward primers (120RF, 5’ CGCGTGAGCAACCTGCC 3’ and CF, 5’

ACTGAGAGGTTGAACGGCCACAT 3’) and one reverse primer (430R, 5’

ACGTCCTTCGTCCCTCATAA 3’) (Table 2). The primer combination of 120RF and

430R yielded a PCR product size of 310 bp. CF and 430R yielded a PCR product size of

200 bp. The three 16S rRNA Ruminococcaceae specific primer sequences were cross-

referenced with ProbeMatch tool to find sequences in the RDP database of 16S rRNAs based on 95% sequence similarity (106), which allowed predicting coverage of

Ruminococcaceae sequences.

Figure 1. The major steps in the CAP3 assembly algorithm (44).

Table 1. Twelve out groups within the known Clostridiales (106).

Clostridiales named sequences Family Accession number

Roseburia intestinalis Lachnospiraceae AJ312386

Clostridium scindens* Lachnospiraceae AB020730

Anaerostipes caccae Lachnospiraceae AJ270487

Blautia producta Lachnospiraceae L76595

Coprococcus catus Lachnospiraceae AB038359

Butyrivibrio fibrisolvens Lachnospiraceae AF349662

Dialister pneumosintes Veillonellaceae ATCC3304

Veillonella parvula Veillonellaceae X84005

Clostridium perfringens* Clostridiaceae 1 AB075767

Clostridium bifermentans Peptostreptococcaceae AB075769

Anaerococcus prevotii Clostridiales_Incertae Sedis XI AF542232

Finegoldia magna Clostridiales_Incertae Sedis XI AB109770

*These two sequences were part of the original 768 sequence alignment.

Table 2. Sequences and the annealing temperatures of the primers used in this study.

Target Annealing Amplicon Reference Primer Sequence (5'–3') (16S rRNA) temp. (° C) size (bp)

HF183F ATCATGAGTTCACATGTCCG Bacteroides- 59 500 9 Bac708R CAATCGGAGTTCTTCGTG Prevotella

CF128F CCAACYTTCCCGWTACTC CF - Bacteroides 58 600 9 Bac708R CAATCGGAGTTCTTCGTG

120RF CGCGTGAGCAACCTGCC Ruminococcaceae 60 310 This study 430R ACGTCCTTCGTCCCTCATAA

CF ACTGAGAGGTTGAACGGCCACAT Ruminococcaceae 60 200 This study 430R ACGTCCTTCGTCCCTCATAA

Collection of Animals Fecal Samples

Fecal samples were collected from dairy cows, pig, goats, horses, chickens, geese, camels, water buffalos, bisons, and human sewage influent. The fecal samples for dairy cows were collected from five different Wisconsin dairy farms (Collected by Dr. Robert

Harrison Jr., Nutrition Professionals Inc.). Each dairy cow fecal sample was collected in sterile 50 mL centrifuge tube and stored on ice until processed. The goat and pig fecal sample was collected from a local breeder (Unverzagt, A., Oshkosh, WI). The fresh samples were collected and stored on ice until processed. The horse and chicken fecal samples were collected from two separate farms and processed in the same manner as previously stated. The camel, water buffalo, and bison fecal samples were collected from a farm in Oshkosh, WI. The fecal samples were collected during winter time (8 February

2011). These samples were frozen in snow. The goose fecal sample was collected fresh from a local beach (Menominee Park, Oshkosh, WI) where constant goose populations have been continually monitored. Sewage influent sample was collected in sterile 50 mL centrifuge tube from City of Oshkosh Waste Water Treatment Plant (Oshkosh, WI) and processed in the same manner as other animal fecal samples. A half gram of fecal material from these animals was weighed and stored at -80o C for DNA analysis. A total of six fecal samples per animal were stored.

Bacterial DNA Extraction and Concentration

Bacterial DNA was extracted from the collected waste samples using MO-BIO

Power Soil® DNA Isolation Kit according to manufactures instructions (MO-BIO

Laboratories, Inc., Carlsbad, CA.). The manufactures protocol is summarized briefly below. First, the DNA isolation kit used detergent to lyse the cells and release the DNA.

The samples were washed and centrifuged various times to remove proteins, humic acid and other impurities. Next, the DNA adhered to a silica membrane in a spin column was washed with ethanol for purification. Next, an elution buffer removed the DNA from the spin column into a sterile microfuge tube for storage at -20o C. The extracted DNA was quantified using Eppendorf BioPhotometer (λ = 260 and 260/280) (Eppendorf Inc.

Westbury, N.Y.). This provides specific DNA concentrations in ng/µL and ensures that the extracted DNA contained presence of good quality DNA.

Determination of Optimal Annealing Temperature

The optimal annealing temperature for Ruminococcaceae specific primer was determined by comparing dairy cow, pig, goat, horse, chicken, goose and sewage fecal

DNA samples at five specific annealing temperatures. The temperature gradient PCR assay included denaturing step at 95o C for 5 minutes, followed by 35 cycles of 95o C for

30 seconds, specific annealing temp (54o C, 54.5o C, 55.6o C, 57.2o C, 59.1o C, 60.7o C,

61.6o C and 62.0o C) for 30 seconds, and 72o C for 1 minute, followed by a final extension step at 72o C for 5 minute. The PCR reaction mixture (20 µL) comprised 10 µL of 2x master mix (New England BioLabs, Ipswich, MA), 0.5 µL (10 µM) of both forward and

reverse primers, and 1 µL DNA extract. All PCR products were visualized using 1.5% agarose gels and stained with Ethdium Bromide (TEKnova, Hollister, CA). All PCR products were compared against 100 bp DNA ladder (New England BioLabs, Ipswich,

MA).

Detection of Fecal Ruminococcaceae and Bacteroides Populations

The fecal DNA samples were analyzed by conventional PCR to confirm the presence or absence of the Ruminococcaceae target population and other traditional fecal indicators molecular targets (Table 2). Specificity and sensitivity of Ruminococcaceae specific primers were tested against cow, pig, and goat fecal DNA samples at 100, 50, 25,

10, 5, and 1 ng/µL of target DNA. This was repeated with the bovine-specific

Bacteroides (CF128F/708R) primers (Table 2) (9). The PCR conditions for Bacteroides primer were detailed by Bower et al. (13).

The PCR conditions for the Ruminococcaceae specific primer PCR assay included a denaturing step at 95o C for 5 minutes, followed by 35 cycles of 95o C for 30 seconds, 60o C for 30 seconds, and 72o C for 1 minute, followed by a final extension step at 72o C for 5 minute. PCR conditions for Bacteroides assays included denaturing at 94o

C for 4 minute followed by 35 cycles of 94o C for 30 seconds, 58o C for 30 seconds, and

72o C for 30 seconds, followed by final extension step of 72o C for 6 minutes (13). All

PCR reaction mixtures were set-up and examined as previously mentioned.

PCR Cloning of 16S rRNA Gene Sequences

Cow, pig, and goat fecal material were extracted for DNA as previously mentioned. The extracted template DNA was used for three separate PCR reactions (25

µL) for each fecal DNA using Ruminococcaceae specific primer. PCR conditions were set-up as previously mentioned. PCR reaction mixture (75 µL) of each cow, pig, and goat fecal DNA was cleaned up using kit manual methods (MO-BIO PCR Clean-up Kit).

Cleaned PCR products (1.5 µL, 1/10 diluted) were ligated into vector pCR®2.1 following manufacturer instructions (Invitrogen Life Technologies, Carlsbad, CA.). The

TOPO chemically 10 competent E. coli were transformed with the vector pCR®2.1

TOPO. Cells were plated on LB agar with ampicillin (50 μg/mL) and x-gal (Invitrogen

Life Technologies, Carlsbad, CA.). For each animal fecal sample, a total of 48 successful clones were picked from each cloning reaction. Overnight cultures of the transformed cells were inoculated into 200 µL LB broth containing ampicillin (50 μg/mL) and incubated at 37o C for 24 hours. The cultures were stored in -80o C until needed for plasmid extraction.

Plasmid Extraction for Ruminococcaceae Animal Clones

A total of 24 clones from each animals fecal were inoculated in 1.8 mL LB broth containing ampicillin (50 μg/mL) and incubated at 37o C for 24 hours in 96 well plate.

The pellet was re-suspended in 100 µL solution 1 (25 mM Tris pH 8; 10mM EDTA pH 8 and 15% sucrose), mixed thoroughly and incubated at room temperature for 5 minutes. A

200 µL of solution 2 (0.2N NaOH; 1% SDS), was added to the reaction, mixed

thoroughly and incubated on ice for 5 minutes. A 150 µL cold 3M KOAC (pH 5), was added to the reaction, mixed well and incubated on ice for 5 minutes. Following the incubation, samples were spun for 2 minutes at 13,000 x g and the supernatant was transferred to a new tube. A 900 µL of 95% EtOH was added to the supernatant and incubated at room temperature for 3 minutes, spun for 2 minutes at 13,000 x g and supernatant was discarded. The pellet was washed in 70% EtOH (100 µL) and mixed by very gentle method. The plasmid DNA was spun for 1 minute at 13,000 x g, the supernatant was discarded and air-dried for 1 hour or until ALL trace of ethanol is evaporated. The pellet was re-suspended in 30 µL 10 mM TE buffer.

DNA Concentrations and Confirmation of Extracted Plasmids

All extracted plasmids were quantified using Eppendorf BioPhotometer

(Eppendorf Inc. Westbury, N.Y.). All plasmids (2 µL) were visualized using 0.7% agarose gels. All plasmid products were compared against 1 kb DNA ladder (New

England BioLabs, Ipswich, MA). All plasmid clones were screened by PCR prior to

DNA sequencing to confirm the presence of fragment of interest. PCR assay was ran as previously detailed.

DNA Sequencing and Analysis of Clones

Each plasmid clone was sequenced using standard M13 primers (F-5'

GTAAAACGACGGCCAGT 3'; R - 5' CACACAGGAAACAGCTATGACCAT 3') by

University of Chicago Cancer Research Center DNA Sequencing Facility. Each

individual sequence chromatogram was analyzed for well-resolved peaks and no ambiguities using BioEdit sequence alignment editor program (41). Segments of vector origin for the plasmids that were sequenced using M13 primers were identified using

NCBI (National Center for Biotechnology Information) VecScreen system (1). BioEdit sequence alignment editor program was used to trim the vector from the sequences (41).

The segment specific to 16S rRNA Ruminococcaceae gene (270 - 350 bp) was selected for sequence analysis. Total of 68 high quality clones were obtained (Cow clones n=23;

Goat n=24; Pig n=21). CAP 3 sequence assembly program was used to identify contiguous sequences (contigs) based on 97% criteria between cow, pig and goat clone sequences (44). The RDP classifier was used to determine the identity of the 68 sequences (106).

Clone sequences (57 of the 68 total) matching the family of Ruminococcaceae sequences at 95% criterion (Family level cut-off based on 16S rRNA gene) within RDP classifier was used to demonstrate how well these primers work against the other putative fecal pollution target sources. Using NCBI blastn algorithm database 57 clones (Cow clones 18; Goat 20; Pig 19) were individually uploaded and matched with the top ten hits within the database, which allowed for a comparison of overlap within the three clone libraries (1). The database was utilized to obtain the target source (e.g. animal type) of the 570 sequence hits.

Development of Real-time PCR Analysis for Ruminococcaceae

In the real-time PCR assay, each reaction (20 µL) contained 10 µL of SYBR

Green 2x master mix (KAPA SYBR FAST, Kapa Biosystem, Woburn, MA), 0.4 µL (10

µM) of each forward and reverse primer, 0.4 µL of ROX High, and 5 µL of DNA template. Samples without any target DNA was used as a negative control for each assay. All real-time PCR assays were performed using StepOnePlus instrument (Applied

Biosystems, Bedford, MA). The qPCR conditions for the Ruminococcaceae specific primer PCR assay included incubation for 10 minutes at 95o C, followed by 40 cycles of denaturation at 95o C for 15 seconds, annealing at 60o C for 1 minute, and extension at

95o C for 15 seconds. A melting curve analysis was performed after amplification to distinguish the target product from any nonspecific product. All PCR products were visualized using 1.5% agarose gels and Ethdium Bromide stain (TEKnova, Hollister,

CA). All PCR products were compared against 100 bp DNA ladder (New England

BioLabs, Ipswich, MA).

A standard curve was generated using cowB1 plasmid clone. The target copy numbers were calculated based on mass of one plasmid molecule (M = [4250bp][1.096 x

10-21 g/bp]). The second step was to calculate the mass of plasmid containing the copy numbers of interest. The final concentration of plasmid DNA (g/µL) was calculated based on 5 µL of plasmid DNA used into each PCR reactions. The top four of the six

(1,000,000; 100,000; 10,000; 1,000; 100; 10; 1) plasmid copy numbers were used to generate a calibration curve (Y = -4.058x + 41.431; R² = 0.9945) (Figure 2). The low range plasmid copy numbers (1000; 100; 1) did not amplify a signal and were not used.

All the real-time assays were performed using the SYBR Green assay. All samples were run in triplicates. The results of Ruminococcaceae specific marker quantifications were expressed in 16S rRNA gene target copy number per gram of feces or 100 ml of water sample.

Cycle threshold (Ct) threshold Cycle 10 y = -4.058x + 41.431 R² = 0.9945 5

0 0 1 2 3 4 5 6 7

Plasmid clone copy number [log10]

Figure 2. Standard curve for SYBR Green real-time qPCR analyses of cowB1 plasmid clone. The equation and the square regression coefficient (R2) of the qPCR are given. Values are of four plasmid copy numbers ± standard errors of the mean (SEM) (error bars).

Microcosm Experiment to Detect Signal Loss of E. coli and Ruminococcaceae

Two identical microcosm experiments were set up to determine the signal loss of alternative fecal indicator in surface water. The detection of Ruminococcaceae signal was compared with E. coli culture based method. The fecal material slurry for the experiment was created by mixing 1g of fresh (collected same day) cow fecal to 100 mL of sterilized water. The fecal slurry (1 mL) was mixed in 1L of fresh Fox River (FR) water (Oshkosh, WI) in an Erlenmeyer flask. The experiment was run for seven consecutive days at 22o C while slowly stirring. A 10 mL of water sample was filtered using 0.45-µM gridded Millipore cellulose filters. The filter was cultured for E. coli using modified m-TEC media (BD Difco, Franklin Lakes, NJ). E. coli levels (CFU/100 mL) were monitored for seven days for both microcosm experiments (duplicates).

For PCR and qPCR studies, a 50 mL water sample from both microcosm experiments was removed each day for seven days using serological pipette for molecular analysis. Water sample were concentrated at 8,000 x g at 4o C for 20 minutes and the pellets were stored for DNA analysis. The Fox River sample without the addition of fecal slurry was also concentrated and stored for DNA analysis. Ruminococcaceae primer (RF120/430R) was compared against the bovine-specific Bacteroides

(CF128F/708R) primers for all eight DNA samples. The same PCR and qPCR assays and conditions were used as described previously.

Characterization of Fecal Pollution in Upper Fox Watershed

Water samples were collected from two distinct locations within June and August

2010 (Figure 3). The water samples were collected from ten rain and non-rain event days (Rain event – 24/48 hours post rain event). The samples were collected from the

Fox River (Oshkosh, WI) and a creek located in a rural area (Spring Brook - Oshkosh,

WI). A total of two 500 mL of water samples was collected into sterile bottle from each location on the same day and were transported to the lab on ice. The Fox River and

Spring Brook samples were filtered in duplicate using sterile 0.45-µm gridded Millipore cellulose filters. The filtered samples were cultured for E. coli using modified m-TEC media (BD DIFCO, Franklin Lakes, NJ). For the two sample locations, 200 mL of water was filtered using 0.22-µM polycarbonate Millipore DNA filters (Billerica, MA).

Duplicate filters were frozen in 1.5-mL tubes at -80o C for future DNA analysis.

Figure 3. Map of the lower Fox basin (Oshkosh, WI) (108). Spring Brook (SB) and Fox River (FR) locations where the water samples were collected.

Bacterial Community Analysis of Surface Waters

Five different water samples were analyzed utilizing bacterial 16S-based (28F –

519R) tag-encoded FLX amplicon pyrosequencing (bTEFAP) approach that is able to perform diversity analyses of bacterial population in water samples (Research and

Testing Laboratory, Lubbock, Texas). Three of the samples were collected from the Fox

River (Oshkosh, WI), on three separate dates (9 June, 15 July, and 21 August). The other two samples (15 July and 21 August) were collected from Spring Brook (Figure 3).

A total of 90,394 bacterial sequences were generated from the five samples (FR69

– 28,146, FR715 – 24,687, FR821R – 10,327, SB715R – 15,847, and SB821 – 11, 387).

After quality control, the remaining 78,971 sequences (FR69 – 23,806, FR715 – 21,965,

FR821R – 8,560, SB715R – 14,319, and SB821 – 10, 321) were analyzed by MG-RAST by comparing it to RDP using a maximum e-value of 1e-0, a minimum identity of 3%, and a minimum alignment length of 50 (72). The server was used to compare sequences generated by bTEFAP to those in a RDPs 16S rRNA gene database (72). The server also allowed for binning of sequences and phylogenetic comparisons. Overall bacterial diversity of all 16S rRNA reads was analyzed with Principal Coordinate Analysis (PCoA) and rarefaction curve was used to assess total diversity (72). PCoA was used to detect similarity across samples with different sample dates and rain and non-rain factor.

Rarefaction curve analysis was used to detect total diversity of samples.

Dominant bacterial population of mammalian GI and those that are used as indicators of fecal pollution were selected from the bTEFAP generated data sets. The bacterial sequences selected for analysis were identified from previous research which

has identified those specific bacterial populations within animal or human intestinal microbial flora (17, 69, 85, 113, 28). The 50 sequences were aligned in ClustalX2 (56).

The selected sequences were used to demonstrate population abundance or loss from rural area to urbanized area. These bacterial sequences were used to generate relative abundance table within the five sample dates and location (Fox River and Spring Brook).

Cluster environmental analysis of the 50 bacterial sequences was performed using

UniFrac server to demonstrate species diversity within the five sample dates and location

(67). The server was also used to perform principal component analysis to demonstrate which environments are most closely related based upon fecal pollution signatures (67).

RESULTS

Specificity of Ruminococcaceae Primer Set

Two forward and one reverse 16S rRNA gene specific Ruminococcaceae primers sequences were created. The 120RF and CF primer sequences had predicted coverage of

64% and 76% to the 56,054 Ruminococcaceae sequences from RDP, while 430R only had 2% specificity to Ruminococcaceae sequences (Table 3). The 430R primer sequence is highly specific to Ruminococcaceae 16S rRNA gene sequences with only

0.015% of Clostridiales matching the 16S rRNA sequences. The most common genera matching both 120RF and CF primer sequences were Faecalibacterium (31% and 29%),

Oscillibacter (12% and 8%), Ruminococcus (7% and 10%), and Clostridium IV (6% and

7%) (Table 4). A large portion (33%) of unclassified Ruminococcaceae was detected with Ruminococcaceae primers. Other notable genera targets were Acetivibrio,

Sporobacter, Butyricicoccus and Clostridium III. Only six genera were predicted to be targeted with 430R primer sequence (Oscillibacter 1%, Flavonifractor 2%,

Faecalibacterium 0.4%, Sporobacter 0.8%, Clostridium IV 0.3% and

Saccharofermentans 0.07%) (Table 4). This suggests that these specific genera will have higher specificity for six Ruminococcaceae populations.

The 16S rRNA Ruminococcaceae specific primers (120RF/430R and CF/430R) were tested against fecal DNA samples from dairy cows, pig, goat, water buffalo, bison, horse, chicken, goose and sewage influent. Using conventional PCR, both

Ruminococcaceae primer sets yield positive results against fecal DNA samples from

dairy cow, pig, goat and camel. The primer sets failed to amplify the target population from horse, chicken, goose, and sewage influent DNA samples (Table 5). The

Bacteroides primer was tested against the same fecal DNA samples as the

Ruminococcaceae specific primers, which amplified the targets in fecal DNA samples from dairy cows, pig, goat, and camel. By conventional PCR assay, both primer sets tested negative against fecal DNA samples from water buffalo, bison, horse, chicken, goose, and sewage influent (Table 5). The Ruminococcaceae real-time PCR assay was able to amplify target population from cow, pig goat, camel, water buffalo and bison fecal DNA (Table 5). The target copy numbers for goat (2.24 x 107 gram/fecal material) and the camel (1.74 x 107 gram/fecal material) fecal samples had the highest amount of the Ruminococcaceae population. While the bison (2.76 x 106 gram/fecal material) and dairy cow (2.96 x 106 gram/fecal material) fecal sample had the lowest detection per gram of fecal material.

Table 3. Mammalian specific Ruminococcaceae 16S rRNA primer sequences.

Length Target Predicted* Primer Sequence (5'–3') of Population Coverage Primer

120RF CGCGTGAGCAACCTGCC 17 Ruminococcaceae 64%

CF ACTGAGAGGTTGAACGGCCACAT 23 Ruminococcaceae 76%

430R ACGTCCTTCGTCCCTCATAA 20 Ruminococcaceae 2%

* The sequences were specifically compared to known Clostridiales as out groups. 120RF and CF primers were paired with 430R. The predicted coverage was determined using the Ribosomal Database Project (RDP) probe match classifier. The target sequences were compared to RDPs online data base of 56,054 Ruminococcaceae sequences (Date analyzed: 08/25/2012).

Table 4. The predicated overall genera for the family of Ruminococcaceae sequences for 120RF, CF and 430R primer sequences.

Percent of Predicted Hits for Each Primer Sequence Total Sequence Genera 120RF (%) CF (%) 430R (%) Comparison Faecalibacterium 83.0 93.4 0.045 13407 Acetivibrio 27.2 56.5 0 1723 Subdoligranulum 80.9 95.7 0 141 Anaerofilum 8.5 43.4 0 129 Acetanaerobacterium 29.2 41.7 0 24 Ethanoligenens 43.1 7.8 0 116 Anaerotruncus 39.5 92.8 0 539 Ruminococcus 49.6 74.8 0 5419 Papillibacter 59.3 51.9 0 27 Sporobacter 65.3 70.7 8.2 147 Fastidiosipila 3.8 1.9 0 53 Butyricicoccus 75.5 90.1 0 968 Saccharofermentans 52.6 75.4 0.096 1043 Flavonifractor 83.6 85.4 2.7 854 Pseudoflavonifractor 75.3 2.2 0 493 Oscillibacter 79.0 64.5 0.274 5478 Hydrogenoanaerobacterium 76.2 81.0 0 21 Clostridium III 30.9 70.6 0 1720 Clostridium IV 58.1 87.9 0.14 3606 Unclassified Ruminococcaceae 59.4 69.2 7.0 20146

* The target sequences were compared to RDPs online data base of 56,054 Ruminococcaceae sequences. The primer 120RF sequence had a total of 36,095 hits, CF had 42,642 hits and 430R primer sequence had 1,417 hits.

Table 5. Detection of Ruminococcaceae in various agricultural animals and human fecal material. The Ruminococcaceae primers were compared against the established bovine-specific Bacteroides (CF128F/708R) primers. (neg: negative; +: positive; NT: Not Tested).

Ruminococcaceae Bacteroides Cow Ruminococcaceae

PCR PCR Real-time PCR

Target per Gram DNA 120RF/430R CF/430R CF128/708R 120RF/430R CF/430R of Fecal Material Sample (120RF/430R) Dairy + + + + + 2.96 x 106 Cow Pig + + + + + 7.10 x 106 Goat + + + + + 2.24 x 107 Camel + + + + + 1.74 x 107 Water neg neg neg + + 4.38 x 106 Buffalo Bison neg neg neg + + 2.76 x 106 Chicken neg neg neg neg neg neg Goose neg neg neg neg neg neg Horse neg neg neg neg neg neg Sewage neg neg neg NT NT NT Influent*

* Oshkosh sewage Influent tested positive against human-specific Bacteroides (HF183/708R) primers (Appendix: Figure 3)

Sensitivity of Ruminococcaceae Primer Set

Out of the two Ruminococcaceae primer sets, the 120RF/430R demonstrated a slight higher sensitivity than CF/430R in the animal fecal DNA (ng/µL). The

120RF/430R amplified a signal from cow, pig, and goat fecal DNA at 1 ng/ µL and the

CF/430R amplified a signal at 25 (Cow), 1 (pig), and 5 ng/µL (goat) (Appendix: Table

1). Bovine-specific Bacteroides (CF128F/708R) primer had similar sensitivity. The

Ruminococcaceae primers had higher sensitivity in detection of pig fecal DNA (1 ng/µL) than Bacteroides primer (10 ng/µL). The dairy cow and goat fecal DNA was detected at the same concentration (1 ng/µL) with both primer sets (Appendix: Table 1). The

Ruminococcaceae 120RF/430R primer set had increased sensitivity in detection of animal fecal DNA in comparison to CF/430R and also had similar sensitivity in comparison to established bovine specific Bacteroides primer.

Gene Sequence Analysis of PCR Ruminococcaceae 16S rRNA Animal Clones

A total of 68 16S rRNA gene clones were obtained from cow, goat and pig fecal

DNA samples (Table 6). Out of the 23 cow fecal clone sequences, 78% of the sequences belonged to Ruminococcaceae, 9% to unclassified Bacteria, 4% to unclassified

Firmicutes, and 9% were unclassified Clostridiales. Out of the 24 goat fecal clone sequences, 83.3% of the sequences were Ruminococcaceae, 8.3% were unclassified

Bacteria and 8.3% were unclassified Firmicutes. Within the 21 pig fecal clone sequences, 90% of the sequences were assigned to Ruminococcaceae, 5% to unclassified

Bacteria, and 5% of the clone sequence to unclassified Firmicute.

Of the 57 high quality Ruminococcaceae sequences classified, based on RDPs

95% sequence similarity, a total of five genera were detected: Clostridium IV,

Faecalibacterium, Oscillibacter, Papillibacter and Sporobacter. Only Oscillibacter and

Sporobacter were identified within all three clone libraries. Oscillibacter was the most predominant species identified within each individual clone library (Cow: 67%, Goat:

90% and Pig: 68%) (Table 7). The genus Faecalibacterium was only identified within the goat clone library. Papillibacter was only identified within the pig clone library. The pig clone library also had the highest diversity, with four out of the five species detected

(Table 7).

The clone libraries were used to demonstrate how well these primers work against the other putative fecal pollution target sources. Using the NCBI blastn algorithm database, the Ruminococcaceae specific clones (18 cow, 20 goat and 19 pig clones) were matched to the top ten hits (Total = 570 hits) within the database. A total of 56 potential fecal sources were identified. Of the 56 sources identified, 24 (43%) were specific to ruminant mammals (Table 8). However, the 24 ruminant sources identified matched

60% of the possible 570 top hits. The ruminant specific sources include cattle, sheep, goat, gazelle, yak, urial, reindeer, buffalo and camel. Half of the ruminant animals identified can be considered agricultural related animals and are potential sources for fecal pollution in rural area. The other non-ruminant species included babirusa, elephant,

Francois langur, pig, European rabbit, Hoatzin (Opisthocomus hoazin), flying squirrel, rat, colobus, human, horse, and gorilla (Table 9).

Table 6. Diversity of Ruminococcaceae population targeted with 120RF/430R primer set. Ruminococcaceae specific clone contiguous sequences based on 97% criteria for cow, pig, and goat clones.

Clone

Library Clones (n) Unique Clones OTUs* Ruminococcaceae Origin* Cow 23 15 18 18

Goat 24 20 22 20

Pig 21 12 15 19

* The number of operational taxonomic units (OTUs) was established based on 97% cut-off level. The clone sequences matching the family of Ruminococcaceae sequences at 95% criterion was identified using RDP classifier. (RDP analyses: 7/6/2012). See Appendix for clone library origin assignment break down based on RDP classifier (Table 1, 2, and 3).

Table 7. The predicated genera for Ruminococcaceae sequences at 95% similarity level for cow, goat and pig clone library origin using RDP classifier.

Goat Fecal Predicted Genera Cow Fecal Clones Pig Fecal Clones Clones

Clostridium IV 3 ND 1 Faecalibacterium ND 1 ND Oscillibacter 12 18 13 Papillibacter ND ND 3 Sporobacter 3 1 2

Table 8. Ruminant animal sources matching cow, goat and pig fecal clones. They were identified using NCBI blastn algorithm database to detect animal fecal pollution sources.

Hits (n) Hits (n) Hits (n) Matching Cow Matching Goat Matching Pig Source Reference Fecal Clones Fecal Clones Fecal Clones 19 32 - Cow feces 45 3 2 - Cattle feces 80 1 4 11 Cow 29 1 - 10 Cow 51 13 30 17 Bovine 16 28 13 25 Rumen fluid 34 - 13 14 Bovine rumen contents 65 1 2 - Holstein heifer 77 - 2 - Cattle 49 - 1 - Holstein heifer 97 1 - - Holstein calves 107 - 1 2 Sheep 33 9 1 - Argali Sheep 60 6 5 - Bighorn Sheep 60 - 1 - Sheep rumen 95 6 2 - Goat 22 10 25 - Speke's gazelle feces 60 - 2 - Okapia johnstoni 60 2 1 1 Qinghai Yak 8 - 5 - Ovis vignei (Urial) 60 - 2 - Reindeer 96 2 6 7 Svalbard reindeer 87 - 1 - Buffalo 84 1 - 3 Dromedary camel 88

Table 9. Non-ruminant animal sources matching cow, goat and pig fecal clones. They were identified using NCBI blastn algorithm database to detect animal fecal pollution sources.

Hits (n) Hits (n) Hits (n) Matching Cow Matching Goat Matching Pig Source Reference Fecal Clones Fecal Clones Fecal Clones 7 1 7 Pig 89 1 - 13 Sus scrofa (Pig) 102 1 2 11 Pig 86 6 2 2 Swine feces 112 - 1 1 Pig 47 - - 8 Pig 50 - - 3 Pig manure 63 - - 3 Pig 59 - 1 - Pig 45 - - 2 Hog lagoon 4 32 2 - European rabbit feces 60 2 - - Hoatzin 37 6 - - Hoatzin 38 2 - - Flying squirrel 68 1 15 18 Rat feces 83 - - 4 Rat feces 15 - 1 - Angola colobus feces 60 - 2 - Equus asinus 60 1 2 - Horse feces 60 - 1 - Grevy's Zebra feces 60 - 2 - Western lowland gorilla feces 60 Eastern Black and White - 3 3 60 Colobus Feces - - 3 Human feces 62 - - 2 Forearm, Human 52 - - 2 Human 64 Human - Ulcerative colitis 2 - - 104 patient - - 2 Human Colon 42 5 3 - Babirusa feces 60 - 2 - African elephant feces 60 6 7 - Asiatic elephant feces 60 3 - - Hamadryas Baboon feces 60 2 2 16 Francois langur 60

Microcosm Experiment to Detect Signal Loss of E. coli and Ruminococcaceae

After spiking river water with bovine fecal material, the E. coli concentrations were detectable (n=2) up to four days after inoculation. The highest level of E. coli (830

CFU/100 mL) was detected one day post inoculation (Table 10). The E. coli counts decreased by almost 75% by day two. The E. coli concentrations continued to decline and could not be detected by culture by day five. The high concentrations of E. coli post inoculation and 24 hours post inoculation indicates that a fecal pollution will more likely be detected shortly after a contamination event.

Inhibition played a significant role in the molecular detection of

Ruminococcaceae and bovine specific Bacteroides target populations within the microcosm samples. Neither of the molecular targets were detected in the original surface water sample. Post DNA clean-up, the post inoculation and day one samples were positive for both molecular targets (Table 10). The extracted DNA from all eight samples from the second identical microcosm experiment did not result in any detection of Ruminococcaceae with conventional or real-time PCR. Removal of PCR inhibitors was not effective in increasing the sensitivity. The additional DNA clean up step may have potentially removed the presence of the target signal from the samples.

Table 10. Detecting signal loss of fecal indicators in seven day microcosm. (Day 1: Post 24 hours inoculation; Day 2: Post 48 hour’s inoculation; Day 3: Post 72 hours; etc.) (neg: negative; +: positive)

E. coli (CFU/100 mL PCR Sample water) (n=2), (SD) Ruminococcaceae Bacteroides Fox River Water 20, (0.71) neg neg

Post Inoculation 560, (99) + + Day 1 830, (99) + + Day 2 215, (21) neg neg Day 3 205, (7) neg neg Day 4 70, (0) neg neg Day 5 0 neg neg

Day 6 0 neg neg

 See appendix (Figure 6) for PCR inhibition and agarose gel visualization of Ruminococcaceae specific primer set (120RF/430) and bovine-specific Bacteroides (CF128F/708R) primers.

Field Test of Ruminococcaceae as Fecal Pollution Target

Water samples from Spring B rook and the Fox River were collected within June and August 2010. All analyzed samples for Spring Brook and the Fox River were positive for E. coli by the culture method (Table 11 and 12). The highest E. coli counts detected from Spring Brook samples were found on 15 July and on 21 August, which corresponded to the largest rainfall event (Table 11). The highest E. coli counts detected from the Fox River samples were found on 12 July and 15 July, which also corresponds to the largest rainfall measured (Table 12). The amount of rainfall plotted against the E. coli values for all 20 Spring Brook samples and all 12 Fox River samples were weakly correlated (R2 values 0.5447 and 0.3835) (Figure 4 and 6).

All samples analyzed from Spring Brook samples were positive for

Ruminococcaceae (Table 11). However, only one sample (15 July) was positive in among the Fox River samples (Table 12). Five of the ten samples (28 July; 4, 21, 29

August; 12 September) had higher sensitivity (104 copies/100 mL) in detection of the

Ruminococcaceae specific population. There was no correlation between (R² = 0.078)

Ruminococcaceae target copy numbers and the average E. coli levels (Figure 5). The only Fox River sample (15 July) positive for molecular detection of Ruminococcaceae population had a detection value of 1.78 x 103 target copy number per 100 mL water sample. The high rainfall 48 hours prior to 15 July could have increased the abundance of this agricultural specific population into the Fox River. A minimal correlation (R² =

0.222) was detected for E. coli levels between Spring Brook and The Fox River samples

(Figure 7).

Table 11. Spring Brook samples were analyzed for average E. coli count and real-time SYBR green PCR of Ruminococcaceae target. Ruminococcaceae 16S rRNA gene copy number per 100 mL of water sample was determined for each sample. The rainfall data were collected using weather underground history data search (Oshkosh, WI).

Average Target Rainfall Amount E. coli Average copy Sample (CFU/100mL), Real-Time PCR number Date 24hrs prior (cm) 48hrs prior (cm) (n=2) (SD) Ruminococcaceae (n=3) SB76 1.42 0.03 3230 (948) + 1.54 x 103 SB715 2.62 7.06 13050 (4172) + 1.56 x 103 SB 719 1.78 0 730 (523) + 5.47 x 103 SB728 0.46 0 420 (170) + 2.13 x 104 SB84 NR NR 790 (156) + 1.15 x 104 SB819 NR NR 925 (389) + 7.51 x 103 SB821 2.82 0 10500 (1839) + 1.98 x 104 SB829 NR NR 280 (170) + 1.57 x 104 SB912 1.96 0 2685 (445) + 1.13 x 104 SB924 0.20 0.99 9850 (1909) + 3.61 x 103

 See Appendix (Figure 9 and 10) for agarose gel visualization of Ruminococcaceae specific primer set (120RF/430) and Real time SYBR green PCR Melt-Curve analysis for ten Spring Brook water DNA samples.

18000

16000

14000

12000

10000

8000

count (CFU/100mL) 6000 y = 1293.2x + 1745

4000 R² = 0.5447 E. coli E. 2000

0 0 1 2 3 4 5 6 7 8 9 10 Amount of Rain (cm)

Figure 4. The correlation between rainfall totals and culturable E. coli levels for Spring Brook surface water samples.

25000

20000

15000

y = -0.4102x + 11671 10000 R² = 0.0777

5000

0 0 2000 4000 6000 8000 10000 12000 14000

Average 16S rRNA16SAverage gene copytarget number E. coli Average (CFU/100 mL)

Figure 5. Relationship between Ruminococcaceae abundance and culturable E. coli levels in Spring Brook surface water samples.

Table 12. Fox River samples were analyzed for average E. coli count and SYBR green real-time PCR of Ruminococcaceae target. (neg: negative; +: positive).

Average Rainfall Amount E. coli Target Sample Average Real-Time PCR copy Date (CFU/100mL), Ruminococcaceae number 24hrs prior (cm) 48hrs prior (cm) (n=2) (SD) (n=3) FR712 2.67 1.75 745 (785) neg ND FR715 2.62 7.06 690 (438) + 1.78 x 103 FR728 0.46 0 195 (7) neg ND FR84 NR NR 100 (57) neg ND FR821 2.82 0 185 (120) neg ND FR829 NR NR 34 (23) neg ND

 See Appendix (Figure 11 and 12) for agarose gel visualization of Ruminococcaceae specific primer set (120RF/430) and Real time SYBR green PCR Melt-Curve analysis for six Fox River water DNA samples.

1400 y = 69.49x + 123.54 1200

R² = 0.3835

1000

800

600 count (CFU/100mL)

400 E. coli E. 200

0 0 2 4 6 8 10 Amount of Rain (cm)

Figure 6. The correlation between rainfall totals and culturable E. coli levels for Fox River surface water samples.

16000

14000

12000

10000

8000 y = 7.3353x + 1884.6 R² = 0.2221 6000

count for Spring Brook (CFU/100mL) 4000

E. coli E. 2000

0 0 200 400 600 800 1000 1200 E. coli count for Fox River (CFU/100 mL)

Figure 7. The correlation between Spring Brook and the Fox River E. coli levels for surface water samples.

Bacterial Community Analysis

Bacterial diversity was determined by tag-encoded FLX 16S rRNA amplicon pyrosequencing (bTEFAP) which allowed for comparison between small stream to large open water. Out of the 62,589 bacterial 16S rRNA gene sequences, the dominant bacteria phyla were Cyanobacteria, Proteobacteria, Actinobacteria, Firmicutes, and

Bacteroidetes. Cyanobacteria were significantly higher in the Fox River samples than the

Spring Brook samples (28% and 0.41% of the bacterial sequences) (Figure 8). While the

Proteobacteria, Firmicutes, and Bacteroidetes population were slightly higher in the two

Spring Brook samples (Figure 8). The Proteobacteria were 50% and 41% of the total bacteria sequences in the Spring Brook samples (SB715 and SB821) and only 25%, 19% and 19% of the Fox River samples (FR69, FR715 and FR821). Similarly, the Firmicutes were 20% and 30% of the total bacteria sequences in the Spring Brook samples (SB715 and SB821) and only 18%, 7% and 6% (FR69, FR715 and FR821) of the Fox River samples (Figure 8).

A higher relative abundance of the soil specific bacterial population was detected in the Spring Brook water samples in comparison to the Fox River samples. Eight

(Acidobacteria, Actinobacteria, Bacilli, Chloroflexi, Clostridia, Deltaproteobacteria,

Flavobacteria, and Planctomycetacia) of the most frequently encountered bacterial lineages in soil were analyzed for diversity within the samples. These lineages encountered for 44% (SB715) and 51% (SB821) of the total bacterial sequences in the

Spring Brook samples. Meanwhile, it was only 38% (FR69), 23% (FR715) and 18%

(FR821) of the total bacterial sequences in the Fox River samples. Bacillus species was

predominantly higher in Spring Brook (13 and 17% of the total bacterial population) in comparison to the Fox River (0.7, 0.9, and 1%) samples (Append Table 8).

Rarefaction analysis was conducted to estimate species richness and diversity within five different samples. The steep slope for Spring Brook indicates higher species richness than in the Fox River samples. Additional sequencing would be required to fully describe the bacterial diversity of the Spring Brook water samples. Fox River bacterial community has lower species richness and additional sequencing would not generate increased diversity (Figure 9).

A small number of Ruminococcaceae sequences were identified within all samples (n=54). The Spring Brook samples had higher relative abundance of

Ruminococcaceae sequences than Fox River samples (Table 13). The agricultural surroundings by the Spring Brook potentially play a significant role in the increased bacterial diversity. Principal coordinate analysis plot (PCA) allows for samples that exhibit similar taxonomic abundance to be grouped together. However, there was limited similarity across the Fox River and Spring Brook surface water bacterial community, which also illustrates that the Fox River bacterial community was very different across the sampling dates (Figure 10).

70%

60%

50%

FR69NR 40% FR715R 30% FR821R

SB715R PercentBacteriaof Phylum 20% SB821R

10%

0% Cyanobacteria Proteobacteria Actinobacteria Firmicutes Bacteroidetes

Bacteria Phylum

Figure 8. Relative abundance of dominant bacteria phylum from five different water sample sequences. Three Fox River and two from Spring Brook (FR – Fox River; SB – Spring Brook; R – Indicates rain; NR – No Rain).

 See appendix (Table 5 and 6) for Taxonomic distribution.

Figure 9. Rarefaction analysis to determine species richness and diversity. Three Fox River and two Spring Brook sample sequences were analyzed (FR – Fox River; SB – Spring Brook; R – Indicates rain) (FR69 – 23,806, FR715 – 21,965, FR821R – 8,560, SB715R – 14,319, and SB821 – 10, 321 sequences). Pyrosequence data was compared to RDP using a maximum e-value of 1e-0, a minimum identity of 3 %, and a minimum alignment length of 50 by MG-RAST (Analysis date: 6/13/2012).

 See appendix (Table 5 and 6) for Taxonomic distribution.

Figure 10. Principal coordinate plot generated utilizing MG-RAST analysis comparing five different sample dates. Three Fox River and two from Spring Brook (FR – Fox River; SB – Spring Brook; R – Rain event). The data was compared to RDP using a maximum e-value of 1e-0, a minimum identity of 3 %, and a minimum alignment length of 50.

Table 13. Average E. coli for the five different water samples used for pyrosequence analysis. Three Fox River and two from Spring Brook (FR – Fox River; SB – Spring Brook; R – Indicates rain) (ND – Not determined) (SD; Standard deviation). (ND – Not detect).

Relative E. coli Average Number of 16S Number of Abundance of (CFU/100mL) n=2, rRNA Bacterial Ruminococcaceae Ruminococcaceae (SD) Sequences Sequences (%) FR69NR ND 19,342 6 0.031 FR715R 690 (438.4) 17,096 6 0.035 FR821R 185 (120.2) 6,222 ND ND SB715R 13050 (4171.9) 11,325 21 0.19 SB821R 10500 (1838.5) 8,604 21 0.24

Bacterial Populations Associated with Fecal Signature of Mammals

The large scale bacterial communities to minor population within the GI tract of mammals can be associated with fecal indicator signature of surface water pollution. A total of 36 bacterial species that can be associated with fecal material of mammals were found from the bTEFAP generated data. The bacterial sequences associated with fecal signature were used to generate relative abundance table to show relative representation of the species within the five sample dates and location (Fox River and Spring Brook).

Out of the 36 species selected for analysis, 67% of the species were identified within the

SB821 sample. However, only 8% of those species were in the FR821 sample. A large majority of the species were only identified from the Spring Brook samples in comparison to the Fox River samples (Table 14 & 15). The Ruminococcaceae associated species (Acetivibrio cellulolyticus, Ethanoligenens harbinense,

Faecalibacterium prausnitzii, Ruminococcus albus, Ruminococcus bromii, Ruminococcus flavefaciens, and Ruminococcus gnavus) were predominantly detected within the Spring

Brook samples. Only Acetivibrio cellulolyticus was detected within the Fox River sample (FR715) (Table 14).

Certain species that are predominantly present within gastrointestinal tract of animals were only found within the SB samples demonstrating increase input of animal related fecal pollution (Clostridium lituseburense, Cellulosilyticum ruminicola,

Selenomonas ruminantium, Ruminococcus sp., and Faecalibacterium prausnitzii) (Table

14). While Alistipes shahi which is predominantly present within the human intestinal microbial flora was only targeted from the FR samples (Table 15).

UniFrac server was used to perform principal component analysis to demonstrate which environments are most closely related based upon fecal pollution signature (67).

None of the environments grouped together (Figure 11).

Table 14. Number of sequences of Gram positive species commonly identified within the GI tract of mammals from five different surface water samples. (ND – Not detected).

Species FR69NR FR715R FR821R SB715R SB821R Acetivibrio cellulolyticus ND 6 ND ND 1 Ethanoligenens harbinense ND ND ND 4 ND Faecalibacterium prausnitzii ND ND ND 14 4 Ruminococcus albus ND ND ND ND 4 Ruminococcus bromii ND ND ND 3 ND Ruminococcus flavefaciens ND ND ND ND 9 Ruminococcus gnavus ND ND ND ND 3 Clostridium difficile ND 2 1 56 5 Clostridium novyi ND ND ND ND 58 Clostridium lituseburense ND ND ND 57 130 Clostridium perfringens 58 ND 1 ND 2 Roseburia cecicola ND ND ND ND 2 Butyrivibrio fibrisolvens ND ND ND ND 4 Cellulosilyticum ruminicola ND ND ND ND 14 Selenomonas ruminantium ND ND ND ND 2 Enterococcus faecalis ND ND ND ND 13 Streptococcus sobrinus ND ND ND ND 1 Streptococcus suis ND ND ND 1 ND Bifidobacterium animalis ND 3 ND 2 3 Bifidobacterium bifidum 22 ND ND ND ND

Table 15. Number of sequences of Gram negative species commonly identified within the GI tract of mammals from five different surface water samples. (ND – Not detected).

Species FR69NR FR715R FR821R SB715R SB821R Bacteroides acidifaciens ND ND ND 14 ND Bacteroides nordii ND ND ND 14 1 Bacteroides rodentium 1 ND ND ND ND Bacteroides sp. ND ND ND 7 11 Bacteroides thetaiotaomicron ND 4 ND ND 2 Bacteroides vulgates ND ND ND ND 12 Prevotella bivia ND 2 ND ND ND Prevotella brevis ND ND ND ND 1 Prevotella multisaccharivorax ND ND ND ND 1 Prevotella oulorum ND ND ND 8 ND Prevotella salivae ND ND ND 5 ND Alistipes finegoldii ND ND ND 2 ND Alistipes shahii 29 6 1 ND ND Escherichia coli ND 1 ND 1 ND Klebsiella pneumoniae ND ND ND 2 8 Salmonella enteric 6 ND ND ND 2

Figure 11. Principal component analysis using UniFrac metric to analyze fecal population cluster sequence diversity. Three Fox River and two Spring Brook were calculated from a neighbor-joining tree.

DISCUSSION

The molecular detection of microbes within fecal and gastrointestinal tract has provided a reliable and fast means by which to identify the sources of fecal contamination. Various studies have targeted 16S rRNA gene of different fecal bacterial population to identify fecal pollution and the sources (12, 30, 78, 93). The molecular targets improve the detection for fecal pollution in surface water. In this study the

Ruminococcaceae 16S rRNA gene primers were detected from dairy cow, goat, pig, buffalo, bison, and camel fecal DNA. In contrast, Oshkosh sewage influent sample was negative for Ruminococcaceae specific signal. This indicates that the target population is higher in abundance within animal fecal material than human. Various bacterial populations can co-exist within both human and animal gastrointestinal tract but can reside at significant different percentage from host to host (28, 73, 98, 113). Certain bacterial populations can be present at significant higher levels in gastrointestinal tract of ruminant animals (e.g cattle, pig, sheep) in comparison to human and can be selected for identification of fecal signature specific to agricultural related fecal pollution (59, 60, 80,

95). This provides a reason for the difference in the detection limits.

The specificity and sensitivity of Ruminococcaceae primers compared against established Bacteroides cow primers were very similar when tested against fecal DNA samples from cow, pig, goat and camel. The ability of both Ruminococcaceae and

Bacteroides primers to detect signal from various animals’ fecal material indicates possible cross-reactivity and low diversity of Ruminococcaceae population within

various animals, specifically agricultural and ruminant animals. Shanks et al. (93) showed a broad range of specificity and cross-reactivity within animals that share similar digestive physiologies. Mieszkin et al. (73) examined pig fecal contamination in the environment found pig-specific Bacteroidales primer (Pig-Bac2) to amplify DNA extracted from bovine, sheep, human, and horse feces. Layton et al. (57) demonstrated higher similarity between the Bacteroides 16S rRNA gene sequences obtained from swine feces and the Bacteroides 16S rRNA gene sequences obtained from human feces than to bovine feces. In contrast, the human-associated and bovine-associated

Bacteroides 16S rRNA gene sequences did not have similarity (57). In this study, we also experienced cross-reactivity when using bovine specific Ruminococcaceae 16S rRNA primers and detected a signal from dairy cow, pig, goat, camel, water buffalo, and bison fecal material. The 16S rRNA gene copy numbers per gram of feces were between

2.76 x 106 and 2.24 x 107 receptively, indicating that Ruminococcaceae population has a relatively similar abundance within the fecal material of cow, pig, goat, camel, water buffalo, and bison.

The gene sequence analysis of PCR Ruminococcaceae 16S rRNA gene clones of cow, pig and goat fecal DNA helped confirm the overlap of the primer set within

Ruminococcaceae populations. Out of the 68 clone sequences, 84% of the sequences were identified as Ruminococcaceae taxa. The majority (75%) of the sequences corresponded to Oscillibacter as the predicted genera. The closest related sequences of the Ruminococcaceae 16S rRNA gene clones were used to demonstrate how well these primers work against the other putative fecal pollution target sources. The 43% of the

animal sources were ruminant mammals. The high concentration of Ruminococcaceae population within the GI tract of agricultural related animals and the low abundance within the GI tract of humans make it a good target for agricultural related fecal pollution

(16, 25, 60). The water samples from Spring Brook (local creek surrounded by farm fields) were all positive for Ruminococcaceae; which also had an elevated presence of E. coli levels (CFU/100mL). The detection of Ruminococcaceae population in the Spring

Brook maybe related to the surrounding farm fields and the higher input of agricultural animal fecal material into the creek. Similar results were obtained by examining streams near farms that contain pastures with and without cattle (58). Lee et al. (58) detected ruminant-specific Bacteroidales marker (CF128F) in all three of the local streams analyzed. Overall, they determined that a general fecal marker was identified frequently in their streams; however the source of contamination varied (58). The Spring Brook creek has constant level of pollution flowing into the stream, which may obtain more fecal material and subsequently the microbial populations for longer period. Also, the detection of Ruminococcaceae population from Fox River was limited because of the decreased fecal pollution flowing into the Fox River.

In the microcosm study, the Ruminococcaceae population was only detected for two days post cow fecal inoculation and the E. coli levels decreased by 75% by day two and continued to decline until day five. The shorter detection limits for the

Ruminococcaceae population suggest a possible source of a recent fecal pollution. These fecal anaerobes make a good target for tracing recent pollution.

The presence of inhibitory factors played a significant role in detection

Ruminococcaceae population within the microcosm experiments. The inhibitory compounds that are found in mixture of environmental samples can lessen the sensitivity of both conventional and quantitative PCR (6, 66). The same inhibition issues were noticed with both conventional and quantitative PCR when targeting Ruminococcaceae population within the two identical microcosm experiments inoculated with cow fecal slurry. The target population was only detected once the extracted DNA was further cleaned.

Bacterial community analysis of Fox River and Spring Brook water samples data illustrated higher species diversity and very different bacterial populations in the Spring

Brook samples than Fox River. The higher diversity in the Spring Brook water sample can be result of the surrounding agricultural areas and its effect on the surrounding surface water. The data also illustrated higher soil bacterial population in the two Spring

Brook water samples in comparison to the three Fox River samples. Eight of the most frequently encountered bacterial lineages in soil were detected at higher levels in the

Spring Brook samples than the Fox River samples. This suggests that there is a significant input of soil and the impact of the surrounding agricultural area into the

Spring Brook creek.

The relative abundance of fecal indicator bacteria in the bTEFAP data suggested that species that are predominantly present within gastrointestinal tract of animals were only found within the Spring Brook samples. The identified species (Clostridium lituseburense, Cellulosilyticum ruminicola, Selenomonas ruminantium, Ruminococcus

sp., and Faecalibacterium prausnitzii) sequences from Spring Brook samples are predominantly associated with fecal signature of animals (17, 69, 85, 113). Clostridium lituseburense has been isolated from dairy wastewaters by use of 16S rRNA (69).

Studies that have looked at bacterial diversity in yaks have isolated Cellulosilyticum ruminicola and Ruminococcus flavefaciens (17, 113). The species Selenomonas ruminantium has been isolated from the rumen of cattle (85). While Alistipes shahi which is predominantly present within the human intestinal microbial flora was only detected in the Fox River samples (28). The relative abundance of bacterial populations associated with fecal signature of mammals within the Spring Brook and Fox River water sample further demonstrate the increased diversity within the Spring Brook. This indicates that greater amount of fecal pollution is detectable in Spring Brook.

The reduction in detection of target fecal population within Fox River samples could be explained by the strict anaerobic nature of the Firmicutes and Bacteroidetes.

The bacterial cells will not remain survive being exposed to the environment after leaving the intestines of their specific hosts and through the transport of current and processing through the stream (20). Meanwhile, if the target population is found at a high concentration, it is likely recent and near the source (20). The detection of

Ruminococcaceae population within the two identical microcosm experiments inoculated with cow fecal slurry was only detected via PCR post inoculation. Balleste et al. (7) showed higher die-off rates for Bacteroides spp. than fecal coliform and enterococci.

Intestinal anaerobic bacterial populations may die off rapidly in environmental water

(75). This suggests that the anaerobic species, such as Ruminococcaceae populations, are more useful as a marker of recent fecal pollution in aquatic environments.

Overall, the results obtained in the present study suggest that the newly designed

Ruminococcaceae 16S rRNA gene primers should be utilized as an alternative molecular target for detection of agricultural related fecal pollution. The primers successfully in amplified this population from surface water samples, fecal material of dairy cow, pig, goat, camel, bison, and water buffalo. The bacterial community analysis of Spring Brook and Fox River water samples utilizing bTEFAP approach helped demonstrate an increased diversity and presence of other populations associated with fecal pollution in the Spring Brook samples. This is most likely attributed to the surrounding farm fields.

Further research and sampling downstream would be needed to support the loss of the target population before it approaches the Fox River. An extensive watershed assessment of the Spring Brook stream could help support the loss of Ruminococcaceae population.

APPENDIX

Figure 1. Ruminococcaceae specific primers (120RF/430R and CF/430R) were tested against fecal DNA samples from dairy cows (Co), pig (Pi), goat (Go), horse (Ho), chicken (Ch), and goose (Gs) for possible cross reactivity between mammalian gut bacterial flora.

Figure 2. The two Ruminococcaceae specific primers and bovine-specific Bacteroides primers were tested against fecal DNA samples from bison (1: 125.63ng and 2: 50.25ng), camel (3: 123ng and 4: 49ng), and Water buffalo (5: 2648ng and 6: 1059ng). Camel fecal DNA (1/2 dilution – Well 4) was the only sample amplified with each of the primer sequences.

Figure 3. Four specific primers as listed below were tested against four concentrations (1: 100 ng/µl, 2: 50 ng/µl, 3: 25 ng/µl 4: 10 ng/µl) of Oshkosh, WI influent sewage sample.

Figure 4. The lowest concentration where the signal was detected using Ruminococcaceae primer (120RF/430R) was compared against the established bovine- specific Bacteroides (CF128F/708R) primers. With Ruminococcaceae primer (120RF/430R), cow, pig, and goat fecal DNA samples were analyzed at six decreasing concentrations (Wells 1-6: 100 50, 25, 10, 5, and 1 (ng/µl)). With Bacteroides (CF128F/708R) primers, cow, pig, and goat fecal DNA samples were analyzed at eight decreasing concentrations (Wells 1-8: 300, 200, 100, 50, 25, 10, 5, and 1 (ng/µl)) (M - 100bp DNA ladder; NT – no template control).

Figure 5. PCR inhibition was assessed for the crude DNA extracts from microcosm experiment. Cow fecal DNA was combined with DNA extracts to characterize the level of PCR inhibition in the samples across 3 dilutions and a temperature gradient of 54 to 62 0C. (A) Ruminococcaceae specific (120RF/430R) primers and (B) Cow Bacteroides (128F/708R) primer were used to for the PCR analysis.

Figure 6. PCR analyses via Ruminococcaceae primer (120RF/430R) and cow Bacteroides primer (CF128F/708R) after crude DNA extract clean up to remove PCR inhibition. Three DNA extracts (FR, T0, and T1) were analyzed at 3 dilutions and 57 0C annealing temperature.

Figure 7. Standard curve analysis of B1 cow plasmid clone using Ruminococcaceae specific primer 120RF/430R in triplicate. (A; 1,000,000 target copy, B; 100,000 target copy, C; 10,000 target copy, D; 1000 target copy, and NTC; No Template Control). All were run in triplicate (A1 – D1, A2 – D2, and A3 – D3).

Figure 8. Real time SYBR green PCR Melt-Curve analysis of Standard Curve analysis of cow B1 plasmid clone at seven different concentrations. NTC: No Template Control. All samples were analyzed in triplicate.

Figure 9. The Ruminococcaceae specific primer (120RF/430R) was tested against 10 Springbrook water DNA sample via real-time SYBR green PCR. (A) (1: SB76R, 2: SB715R, 3: SB821R, and 4: SB924R). (B) (1: SB719R, 2: SB728R, 3: SB84NR, 4: SB819NR, 5: SB829NR, 6: SB912R). (M: 100bp DNA ladder; NT: no template control; +: Positive control/10,000 copies of plasmid) All samples were analyzed in triplicate.

Figure 10. Real time SYBR green PCR Melt-Curve analysis for 10 Springbrook water DNA samples, No template control and Positive control. All samples were analyzed in triplicate.

Figure 11. The Ruminococcaceae specific primer (120RF/430R) was tested against six Fox River water DNA sample via real-time SYBR green PCR. (A) (1: FR712R, 2: FR715R, 3: FR728R, 4: FR84NR, 5: FR821R, and 6: FR829NR). (M: 100bp DNA ladder; NT: no template control; +: Positive control/Cow Fecal DNA) All samples were analyzed in triplicate.

Figure 12. Real time SYBR green PCR Melt-Curve analysis for the Fox River surface water DNA samples. No template control and positive control. All samples were analyzed in triplicate.

Table 1. The lowest concentration where the signal was detected using Ruminococcaceae primer (120RF/430R) was compared against the established bovine-specific Bacteroides (CF128F/708R) primers. Cow, pig, and goat fecal DNA samples were analyzed at 100, 50, 25, 10, 5, and 1 (ng/µl) concentration. 120RF/430R has better specificity over the Ruminococcaceae primer CF/430R. The bovine specific Bacteroides primer had similar sensitivity in detection to 120RF/430R Ruminococcaceae primer.

Animal Fecal DNA (ng/µL)

Primer set Dairy Cow Pig Goat Ruminococcaceae 1 1 1 (120RF/430R) Ruminococcaceae (CF/430R) 25 1 5

Bacteroides (CF128/708R) 1 10 1

Table 2. Cow fecal clone sequences library origin assignment break down based on RDP classifier. Out of the 23 cow fecal clone sequences, 18 sequences belonged to Ruminococcaceae, two clone sequences to unclassified Bacteria (CowE1 and CowB2), one to unclassified Firmicutes (CowE2), and two were unclassified Clostridiales (CowF2 and CowC1). (RDP analyses: 7/6/2012)

Phylum Level Class Level Order Level Clones Breakdown Breakdown Breakdown Family Level Breakdown Species Level Breakdown

CowA3 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 47%

CowB1 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 43%

CowD3 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 42%

CowF3 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 36%

CowD1 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 31%

CowG2 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 31%

CowC2 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 99% Oscillibacter 33%

CowC3 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 98% Sporobacter 31%

CowD2 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 98% Sporobacter 31%

CowG1 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 98% Clostridium IV 28%

CowG3 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 98% Clostridium IV 28%

CowH2 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 98% Clostridium IV 28%

CowE3 Firmicutes 100% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Oscillibacter 28%

CowF1 Firmicutes 100% Clostridia 99% Clostridiales 99% Ruminococcaceae 97% Oscillibacter 24%

CowA2 Firmicutes 99% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Oscillibacter 57%

CowB3 Firmicutes 99% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Oscillibacter 41%

CowA1 Firmicutes 99% Clostridia 98% Clostridiales 98% Ruminococcaceae 96% Sporobacter 37%

CowF2 Firmicutes 98% Clostridia 97% Clostridiales 97% Ruminococcaceae 94% Oscillibacter 26%

CowH3 Firmicutes 97% Clostridia 97% Clostridiales 97% Ruminococcaceae 97% Oscillibacter 36%

CowC1 Firmicutes 97% Clostridia 96% Clostridiales 96% Ruminococcaceae 94% Sporobacter 35%

CowE2 Firmicutes 96% Clostridia 94% Clostridiales 94% Ruminococcaceae 94% Sporobacter 37%

CowE1 Firmicutes 90% Clostridia 90% Clostridiales 90% Ruminococcaceae 89% Oscillibacter 48%

CowB2 Firmicutes 92% Clostridia 91% Clostridiales 91% Ruminococcaceae 90% Oscillibacter 43%

Table 3. Goat fecal clone sequences library origin assignment break down based on RDP classifier. Out of the 24 goat fecal clone sequences, 20 sequences were Ruminococcaceae, two are unclassified Bacteria (GoatB10 and GoatF11) and two to unclassified Firmicutes (GoatB9 and GoatD11). (RDP analyses: 7/6/2012)

Phylum Level Class Level Order Level Species Level Clones Breakdown Breakdown Breakdown Family Level Breakdown Breakdown

GoatA9 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 62%

GoatG11 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 61%

GoatH9 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 59%

GoatC11 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 49%

GoatF10 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 44%

GoatD10 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 37%

GoatF9 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 33%

GoatA11 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 27%

GoatE11 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 27%

GoatH11 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 98% Oscillibacter 25%

GoatH10 Firmicutes 100% Clostridia 98% Clostridiales 98% Ruminococcaceae 98% Sporobacter 33%

GoatA10 Firmicutes 99% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Oscillibacter 53%

GoatE9 Firmicutes 99% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Oscillibacter 51%

GoatE10 Firmicutes 99% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Oscillibacter 47%

GoatC9 Firmicutes 99% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Oscillibacter 44%

GoatG9 Firmicutes 99% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Oscillibacter 41%

GoatB11 Firmicutes 99% Clostridia 98% Clostridiales 98% Ruminococcaceae 98% Oscillibacter 43%

GoatG10 Firmicutes 99% Clostridia 98% Clostridiales 98% Ruminococcaceae 98% Oscillibacter 40%

GoatD9 Firmicutes 99% Clostridia 98% Clostridiales 98% Ruminococcaceae 97% Oscillibacter 37%

GoatC10 Firmicutes 99% Clostridia 98% Clostridiales 98% Ruminococcaceae 95% Faecalibacterium 44%

GoatB9 Firmicutes 97% Clostridia 94% Clostridiales 94% Ruminococcaceae 93% Oscillibacter 38%

GoatD11 Firmicutes 96% Clostridia 94% Clostridiales 94% Ruminococcaceae 92% Oscillibacter 53%

GoatB10 Firmicutes 90% Clostridia 90% Clostridiales 90% Ruminococcaceae 89% Oscillibacter 45%

GoatF11 Firmicutes 86% Clostridia 84% Clostridiales 84% Ruminococcaceae 84% Oscillibacter 38%

Table 4. Pig fecal clone sequences library origin assignment break down based on RDP classifier. Of the 21 pig fecal clone sequences, 19 sequences were assigned to Ruminococcaceae, one to unclassified Bacteria (PigD7), and one clone sequence to unclassified Firmicute (PigG7). (RDP analyses: 7/6/2012)

Phylum Level Class Level Order Level Clones Breakdown Breakdown Breakdown Family Level Breakdown Species Level Breakdown

PigB7 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Papillibacter 60%

PigE5 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 48% pigD5 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 44%

PigA6 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 44%

PigG6 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Clostridium IV 44%

PigC7 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 44%

PigA7 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 43%

PigC5 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 42%

PigF7 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 33% pigB6 Firmicutes 100% Clostridia 100% Clostridiales 100% Ruminococcaceae 100% Oscillibacter 24%

PigD6 Firmicutes 100% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Oscillibacter 52%

PigF6 Firmicutes 100% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Oscillibacter 52%

PigC6 Firmicutes 100% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Oscillibacter 37%

PigA5 Firmicutes 100% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Oscillibacter 35%

PigH6 Firmicutes 100% Clostridia 98% Clostridiales 98% Ruminococcaceae 98% Papillibacter 34% pigE6 Firmicutes 99% Clostridia 99% Clostridiales 99% Ruminococcaceae 99% Papillibacter 40%

PigE7 Firmicutes 99% Clostridia 98% Clostridiales 98% Ruminococcaceae 98% Oscillibacter 41%

PigF5 Firmicutes 99% Clostridia 98% Clostridiales 98% Ruminococcaceae 97% Sporobacter 55%

PigB5 Firmicutes 98% Clostridia 97% Clostridiales 97% Ruminococcaceae 96% Sporobacter 31%

PigG7 Firmicutes 95% Clostridia 88% Clostridiales 88% Ruminococcaceae 83% Pseudoflavonifractor 14%

PigD7 Firmicutes 94% Clostridia 94% Clostridiales 94% Ruminococcaceae 93% Oscillibacter 30%

Table 5. Domain level taxonomic distribution of five water sample sequences based on data compared to RDP using a maximum e-value of 1e-0, a minimum identity of 3 %, and a minimum alignment length of 50 by MG-RAST (Metagenomics analysis server: Analysis date: 6/13/2012)

Bacteria Eukaryota Unassigned Unclassified Total

FR69NR 19,342 2,673 1,790 1 23,806 FR715R 17,096 3,427 1,442 0 21,965 FR821R 6,222 1,234 1,100 4 8560 SB715R 11,325 162 2,818 14 14,319 SB821R 8,604 289 1,423 5 10,321 Total 62,589 7,785 8,573 24 78,971

Table 6. Phylum level break down of 10 major Bacteria phyla sequences and unclassified sequences based on data compared to RDP using a maximum e-value of 1e-0, a minimum identity of 3 %, and a minimum alignment length of 50 by MG-RAST (Metagenomics analysis server: Analysis date: 6/13/2012).

FR69NR FR715R FR821R SB715R SB821R Total

Actinobacteria 4,404 2,864 509 1,914 1,479 11,170 Bacteroidetes 1,479 670 79 1,205 941 4,374 Chloroflexi 8 8 6 55 41 118 Cyanobacteria 4,799 8,929 4,025 164 90 18,007 Firmicutes 3,411 1,209 362 2,167 2,472 9,621 Planctomycetes 19 26 24 8 8 85 Proteobacteria 4,738 3,061 1,097 5,499 3,449 17,844 Spirochaetes 27 13 0 5 12 57 Tenericutes 19 0 15 24 22 80 Verrucomicrobia 320 116 23 51 49 559 Unclassified (derived from 95 176 63 164 32 530 Bacteria) Total 19,319 17,072 6,203 11,256 8,595 62,445

Table 7. Phylum level break down of five major Eukaryote phyla sequences and unclassified sequences based on data compared to RDP using a maximum e-value of 1e- 0, a minimum identity of 3 %, and a minimum alignment length of 50 by MG-RAST (Metagenomics analysis server: Analysis date: 6/13/2012).

FR69NR FR715R FR821R SB715R SB821R Total

Ascomycota 4 4 6 11 2 27 Bacillariophyta 1,010 1,959 353 2 194 3,518 Chlorophyta 177 916 617 38 49 1,797 Euglenida 8 8 23 0 0 39 Streptophyta 85 275 126 75 23 584 Unclassified (derived from 1,387 260 109 30 21 1,807 Eukaryota) Total 2,671 3,422 1,234 156 289 7,772

Table 8. Eight of the most frequently encountered bacterial lineages in soil were analyzed for percent of total bacterial sequences within each of the five water samples. The data was compared to RDP using a maximum e-value of 1e-0, a minimum identity of 3 %, and a minimum alignment length of 50 by MG-RAST (Metagenomics analysis server: Analysis date: 11/13/2012).

Class FR69NR FR715R FR821R SB715R SB821R Acidobacteria 0.03% 0.03% 0.06% 0.20% 0.10% Actinobacteria 21% 15% 10% 17% 18% Bacilli 0.70% 0.90% 1% 13% 17% Chloroflexi 0.05% 0.05% 0.10% 0.40% 0.40% Clostridia 14% 5% 4% 8% 11% Deltaproteobacteria 1% 0.80% 0.9% 1% 1% Flavobacteria 1% 1% 0.30% 4% 4% Planctomycetacia 0.09% 0.10% 0.30% 0.10% 0.3%

REFERENCES

1. Altschul, S.F., Madden,T.L., Schaffer,A.A., Zhang,J., Zhang,Z., Miller,W. and Lipman,D.J. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25: 3389–3402.

2. APHA. (1992) Standard methods for the examination of water and wastewater. 18th ed. American Public Health Association, Washington, DC.

3. Ashbolt N.J. (2004) Microbial contamination of drinking water and disease outcomes in developing regions. Toxicology 198: 229-38.

4. Arfken, A.M. and Song, B. (2012) Identification and Comparison of Airborne Bacterial Communities form a Hog Farm and Spray Field Using Culture and Non-Culture Based. Unpublished manuscript.

5. Armstrong G.L., Hollingsworth J, and Morris J.G. (1996) Emerging foodborne pathogens: Escherichia coli 0157:H7 as a model of entry of a new pathogen into the food supply of the developed world. Epidemiol. Rev. 18: 29-51.

6. Audemard, C., Reece, K.S., Burreson, E.M., (2004) Real-time PCR for detection and quantification of the protistan parasite Perkinsus marinus in environmental waters. Appl. Environ. Microbiol. 70: 6611-6618.

7. Ballesté, E., and Blanch, A. R. (2010) Persistence of Bacteroides species populations in a river as measured by molecular and culture techniques. Appl. Environ. Microbiol. 76: 7608-16.

8. Bao, L., Zhou, J., Li, C., Quan, Z., Sun, Q. and Lu, H. (2008) Diversity of Qinghai Yak Ruminal Bacteria and Fungi in China. Unpublished manuscript.

9. Bernhard, A. E., and K. 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. Appl. Environ. Microbiol. 66: 4571-457 10. Bernhard, A. E., and K. 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. Appl. Environ. Microbiol. 66: 1587– 1594.

11. Bitton, G. (2005) Wastewater Microbiology, 3rd. Edition, Wiley-Liss, Hoboken, NJ., 746 pp.

12. Bitton, G. (2005) Microbial Indicators of Fecal Contamination: Application to Microbial Source Tracking. Department of Environmental Engineering Sciences. 3-50.

13. Bower, P. A., C. O. Scopel, E. T. Jensen, M. M. Depas, and S. L. McLellan. (2005) Detection of genetic markers of fecal indicator bacteria in Lake Michigan and determination of their relationship to Escherichia coli densities using standard microbiological methods. Appl. Environ. Microbiol. 71: 8305–8313.

14. Brown, A. (1994) Ohio’s Hydrologic Cycle. AEX 461. Ohio State University Extension

15. Brooks, S.P.J., Abnous, K., Kwan, J., Matias, F., Green-Johnson, J., Selinger, L.B., Thomas, M. and Kalmokoff, M.L. (2009) Diets enriched in oat bran or wheat bran temporally and differentially alter the composition of the fecal community of the rat. Unpublished manuscript.

16. Brulc J.M., Antonopoulos D.A., Berg Miller M.E., Wilson M.K., Yannarell A.C., et al. (2009) Gene-centric metagenomics of the fiber-adherent bovine rumen microbiome reveals forage specific glycoside hydrolases. Proc Natl Acad Sci U S A. 106:1948–1953

17. Cai, S. C., Li, J. B., Hu F. Z., Zhang K. G., Luo Y. M., Janto B., Boissy R., Ehrlich G., and Dong X. Z. (2010) Cellulosilyticum ruminicola, a newly described rumen bacterium that possesses redundant fibrolytic-protein-encoding genes and degrades lignocellulose with multiple carbohydrate-borne fibrolytic enzymes. Appl. Environ. Microbiol. 76:3818–3824.

18. Center for Integrated Agricultural Systems. (2008) The Wisconsin Beef Cattle Industry: A snapshot. UW-Madison College of Agricultural and Life Sciences.

19. Chao, K.K., Chao C.C., and Chao W.L. (2003) Suitability of the traditional microbial indicators and their enumerating methods in the assessment of fecal pollution of subtropical freshwater environments. J Microbiol Immunol Infect 36, 288–293.

20. Coakley, Tricia L. (2011) Locating hot spots of fecal pollution in an urban watershed of central Kentucky using Bacteroides 16S rRNA markers. Theses and Dissertations--Earth and Environmental Sciences. Paper 4.

21. Colford, J.M. Jr., Wade T.J., Schiff K.C., Wright C.C., Griffith J.F., Sandhu S.K., Burns S., Sobsey M., Lovelace G., and Weisberg S.B. (2007) Water quality indicators and the risk of illness at beaches with nonpoint sources of fecal contamination. Epidemiology 18: 27-35.

22. Cunha, I.S., Barreto C.C., Costa O.Y., Bomfim M.A., Castro A.P., Kruger R.H., Quirino B.F. (2011) Bacteria and Archaea community structure in the rumen microbiome of goats (Capra hircus) from the semiarid region of Brazil. Anaerobe 17:118–124.

23. Dick, L. K., Bernhard A. E., Brodeur T. J., Santo Domingo J. W., Simpson J. M., Walters S. P., and Field K. G. (2005) Host distributions of uncultivated fecal Bacteroides bacteria reveal genetic markers for fecal source identification. Appl. Environ. Microbiol. 71:3184–3191.

24. Daly, K., Stewart C. S., Flint H. J., and Shirazi-Beechey S. P. (2001) Bacterial diversity within the equine large intestine as revealed by molecular analysis of cloned 16S rRNA genes. FEMS Microbiol. Ecol. 38:141–151.

25. Dehority, B. A., and Grubb J. A. (1977) Characterization of the predominant bacteria occurring in the rumen of goats (Capra hircus). Appl. Environ. Microbiol. 33:1030–1036.

26. Dowd, S. E., T. R. Callaway, Wolcott R. D., Sun Y., McKeehan T., Hagevoort R. G., Edrington, T. S. (2008) Evaluation of the bacterial diversity in the feces of cattle using 16S rDNA bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP) BMC Microbiol 8: 125

27. Duerden, B. I. (1980) The isolation and identification of Bacteroides spp. from normal human fecal flora. J. Med. Microbiol. 13: 69–78.

28. Eckburg, P.B., Bik, E.M., Bernstein, C.N., Purdom, E., Dethlefsen, L., Sargent, M., Gill, S.R., Nelson, K.E., Relman, D.A., (2005) Diversity of the human intestinal Microbial flora. Science 308, 1635-1638.

29. Fernando, S.C., Purvis H.T 2nd, Najar F.Z., Sukharnikov L.O., Krehbiel C.R. et al. (2010) Rumen microbial population dynamics during adaptation to a high-grain diet. Appl Environ Microbiol 76: 7482–7490.

30. Field, K. G., and S. Mansour. (2007) Fecal source tracking, the indicator paradigm, and managing water quality. Water research. 41: 3517-3538

31. Fogarty, L. R., and M. A. Voytek. (2005) Comparison of bacteroides-prevotella 16S rRNA genetic markers for fecal samples from different animal species. Appl. Environ. Microbiol. 71: 5999–6007.

32. Furet, J.P., Firmesse O, Gourmelon M, Bridonneau C, Tap J, Mondot S, Doré J, Corthier G. (2009) Comparative assessment of human and farm animal fecal microbiota using real-time quantitative PCR. FEMS Microbiol Ecol. 68: 351-362.

33. Gagen, E.J., Mosoni P., Denman S.E., Al Jassim R., McSweeney C.S. & Forano E. (2012) Methanogen colonisation does not significantly alter acetogen diversity in lambs isolated 17 hours after birth and raised aseptically. Microbial. Ecol. 64(3):628-40.

34. Garcia, S.L, Jangid K, Whitman W.B., Das K.C. (2011) Transition of microbial communities during the adaption to anaerobic digestion of carrot waste. Bioresour Technol. 15:7249-56.

35. Gleick H. Peter. (2002) Dirty Water: estimated Death from Water-Related Disease 2000-2020. Pacific Institute for Studies in Development, Environment, and Security.

36. Gill, S. R., et al. (2006) Metagenomic analysis of the human distal gut microbiome. Science, 312: 1355–1359.

37. Godoy-Vitorino F, Ley R.E, Gao Z, Pei Z, Ortiz-Zuazaga H, Pericchi L.R et al. (2008) The crop bacterial community of the hoatzin, a neotropical folivorous flying bird. Appl Environ Microbiol 74: 5905–5912.

38. Godoy-Vitorino F, Goldfarb K.C., Brodie E.L., Garcia-Amado M.A., Michelangeli F., Domínguez-Bello M.G. (2010) Developmental microbial ecology of the crop of the folivorous hoatzin. ISME Journal. 4:611–620.

39. Guan, T. Y., and R. A. Holley. (2003) Pathogen survival in swine manure environments and transmission of human enteric illness—A review. J. Environ. Qual. 32:383–392.

40. Haack, S. K., Fogarty L. R., and Wright C. (2003) Escherichia coli and enterococci at beaches in the Grand Traverse Bay, Lake Michigan: sources, characteristics, and environmental pathways. Environ. Sci. Technol. 37:3275-3282.

41. Hall, Thomas. (2011). BioEdit. Ibis Biosciences. A division of Abbott Laboratories. Carlsbad, CA

42. Harrell L, Wang Y, Antonopoulos D, Young V, Lichtenstein L, Huang Y, Hanauer S, Chang E. (2012) Standard colonic lavage alters the natural state of mucosal- associated microbiota in the human colon. PLoS One. 7:e32545.

43. Hold, G. L., Pryde S. E., Russell V. J., Furrie E., and Flint H. J. (2002) Assessment of microbial diversity in human colonic samples by 16S rDNA sequence analysis. FEMS Microbiol. Ecol. 39: 33–39.

44. Huang, X. and Madan, A. (1999) CAP3: a DNA sequence assembly program. Genome Res., 9, 868–877.

45. Jeong J.Y, Park H.D, Lee K.H, Weon H.Y, and Ka J.O. (2011) Microbial community analysis and identification of alternative host-specific fecal indicators in fecal and river water samples using pyrosequencing. J Microbiol. 49(4): 585-94.

46. Jeter, S. N., McDermott C. M., Bower P. A., Kinzelman J. L., Bootsma M. J., Goetz G. W., and McLellan S. L. (2009) Bacteroidales diversity in ring-billed gulls (Laurus delawarensis) residing at Lake Michigan beaches. Appl. Environ. Microbiol. 75: 1525–1533.

47. Kalmokoff M, Waddington L.M, Thomas M, Liang K.L, Ma C, Topp E, Dandurand U.D, Letellier A, Matias F, Brooks S.P. (2011) Continuous feeding of antimicrobial growth promoters to commercial swine during the growing/finishing phase does not modify faecal community erythromycin resistance or community structure. J Appl Microbiol, 110: 1414-1425.

48. Kauffman, J.B. and Krueger, W.C. (1984) Livestock Impacts on Riparian Ecosystems and Streamside Management Implications: A review. J Range Manage. 37: 430-438.

49. Kim M, Morrison M, Yu Z. (2011) Phylogenetic diversity of bacterial communities in bovine rumen as affected by diets and microenvironments. Folia Microbiol. 56: 53– 458

50. Kobayashi Y, Itoh A, Miyawaki K, Koike S, Iwabuchi O, Iimura Y, Kobashi Y, Kawashima T, Wakamatsu J, Hattori A, Murakami H, Morimatsu F, Nakaebisu T, Hishinuma T. (2011) Effect of liquid whey feeding on fecal microbiota of mature and growing pigs. Anim Sci J. 82(4): 607-15.

51. Kong Y, Teather R, Forster R. (2010) Composition, spatial distribution, and diversity of the bacterial communities in the rumen of cows fed different forages. FEMS Microbiol Ecol 74: 612–622.

52. Kong H. H., Oh J., Deming C. et al., (2012) “Temporal shifts in the skin microbiome associated with disease flares and treatment in children with atopic dermatitis,” Genome Research, vol. 22, no. 5, pp. 850–859.

53. Kreader, C. A. (1995) Design and Evaluation of Bacteroides DNA probes for the Specific Detection of Human Fecal Pollution. Appl. Environ. Microbiol. 61 (4), 1171- 1179.

54. Kreader, C. A. (1998) Persistence of PCR-detectable Bacteroides distasonis from human feces in river water. Appl Environ Microbiol 64: 4103-4105.

55. Lamendella, R., Santo Domingo J. W., Oerther D. B., Vogel J. R., and Stoeckel D. M. (2007) Assessment of fecal pollution sources in a small northernplains watershed using PCR and phylogenetic analyses of Bacteroidetes 16S rRNA gene. FEMS Microbiol. Ecol. 59: 651–660

56. Larkin M.A, Blackshields G, Brown N.P, Chenna R, McGettigan P.A, et al. (2007) Clustal W and Clustal X version 2.0. Bioinformatics 23: 2947–2948

57. Layton, A., McKay L., Williams D., Garrett V., Gentry R., and Sayler G. (2006) Development of Bacteroides 16S rRNA gene TaqMan-based real-time PCR assays for estimation of total, human, and bovine fecal pollution in water. Appl. Environ. Microbiol. 72: 4214-4224.

58. Lee, J.J., and M. Molina. (2009) An Assessment of Fecal Contamination in Rural Streams Impacted by Litter Applications. Proceedings of the Georgia Water Resources Conference.

59. Leser T. D., Amenuvor J. Z., Jensen T. K., Lindecrona R. H., Boye M., and Moller K. (2001) Culture-independent analysis of gut bacteria: the pig gastrointestinal tract microbiota revisited. Appl. Environ. Microbiol. 68: 673–690.

60. Ley R, et al. (2008) Evolution of mammals and their gut microbes. Science, 320:1647–1651.

61. Ley R.E, Peterson D.A, Gordon J.I. (2006) Ecological and evolutionary forces shaping microbial diversity in the human intestine. Cell 124: 837–848.

62. Ley R.E, Turnbaugh P.J, Klein S, Gordon J.I. (2006) Microbial ecology: Human gut microbes associated with obesity. Nature 444: 1022–1023.

63. Liu F.H, Wang S.B, Zhang J.S, Zhang J, Yan X, Zhou H.K, Zhao G.P, Zhou Z.H (2009) The structure of the bacterial and archaeal community in a biogas digester as revealed by denaturing gradient gel electrophoresis and 16S rDNA sequencing analysis. J. Appl. Microbiol. 106: 952-966.

64. Li E, Hamm C.M, Gulati A.S, Sartor R.B, Chen H, et al. (2011) Inflammatory bowel diseases phenotype, C. difficile and NOD2 genotype are associated with shifts in human ileum associated microbial composition PLoS ONE. Accepted for publication.

65. Li M, Zhou M, Adamowicz E, Basarab J.A., Guan L.L. (2011) Characterization of bovine ruminal epithelial bacterial communities using 16S rRNA sequencing, PCR- DGGE, and qRT-PCR analysis. Vet Microbiol. 155(1): 72-80.

66. Loge, F.N., Thompson, D.E., Call, D.R., (2002) PCR detection of specific pathogens in water: A risk-based analysis. Environ. Sci. Technol. 36: 2754-2759.

67. Lozupone C and Knight R (2005) UniFrac: a new phylogenetic method for comparing microbial communities. Appl Environ Microb 71: 8228–8235.

68. Lu, H.P., Wang, Y.B., Huang, S.W., Lin, C.Y., Wu, M., Hsieh, C.H. and Yu, H.T. (2012) Metagenomic analysis reveals a functional signature for biomass degradation by cecal microbiota in the leaf-eating flying squirrel (Petaurista alborufus lena). BMC Genomics 13: 466

69. McGarvey, J. A., Miller W. G., Sanchez S., and Stanker L. (2004) Identification of bacterial populations in dairy wastewaters by use of 16S rRNA gene sequences and other genetic markers. Appl. Environ. Microbiol. 70: 4267–4275.

70. McLellan, S.L., Huse, S.M., Mueller-Spitz, S.R., Andreishcheva, E.N., and Sogin, M.L. (2010) Diversity and population structure of sewage derived microorganisms in wastewater treatment plant influent. Environ Microbiol 12: 378–392.

71. McQuaig, S. M., Scott, T. M., Harwood, V. J., Farrah, S. R., and Lukasik, J. O. (2006) Detection of human derived fecal pollution in environmental waters using a PCR- based human polyomavirus assay. Appl. Environ. Microbiol. 72: 7567-7574.

72. Meyer, F., Paarmann, D., D’Souza, M., Olson, R., Glass, E.M., Kubal, M., Paczian, T., Rodriguez, A., Stevens, R., Wilke, A. et al. (2008) The metagenomics RAST server—a public resource for the automatic phylogenetic and functional analysis of metagenomes. BMC Bioinformatics. 9, 386

73. Mieszkin, S., Furet, J. P., Corthier, G., and Gourmelon, M. (2009) Estimation of pig fecal contamination in a river catchment by real-time PCR using two pig-specific Bacteroidales 16S rRNA genetic markers. Appl. Environ. Microbiol. 75: 3045–3054.

74. Miner, R.J., Humenik, F.J., and Overcash, M.R. (2000) Managing livestock wastes to preserve environmental quality, 1st ed. Iowa State University Press, Ames.

75. Nebra, Y., Jofre, J., and Blanch, A. R. (2002) The effect of reducing agents on the recovery of injured Bifidobacterium cells. J. Microbiol. Methods 49: 247-254.

76. Nocker, A., Burr, M., and Camper, A.K. (2007) Genotypic microbial community profiling: a critical technical review. Microb Ecol. 54: 276-289.

77. Ohene-Adjei, S., Teather, R.M., Penner, G., and Forster, R.J. (2008) Rumen eubacterial 16S rRNA gene analysis implicates the Bacteroidetes as a major particulate phase associated phylum. Unpublished manuscript.

78. Okabe, S., Okayama, N., Savichtcheva O., and Ito, T. (2007) Quantification of host-specific Bacteroides-Prevotella 16S rRNA genetic markers for assessment of fecal pollution in freshwater. Appl. Microbiol. Biotechnol. 74: 890-901.

79. Orpin, C. G., Mathiesen, S. D., Greenwood, Y., and Blix, A. S. (1985) Seasonal changes in the ruminal microflora of the high-arctic Svalbard reindeer (Rangifer tarandus platyrhynchus). Appl. Environ. Microbiol. 50: 144–151.

80. Ozutsumi Y, Hayashi H, Sakamoto M, Itabashi H, Benno Y. (2005) Culture independent analysis of fecal microbiota in cattle. Biosci Biotechnol Biochem. 69: 1793- 1797.

81. Page, R. D. (1996) TreeView: an application to display phylogenetics trees on personal computers. Comput. Appl. Biosci. 12: 357-358.

82. Patton, T.G., Scupham, A.J., Bearson, S.M., and Carlson, S.A. (2008) Characterization of fecal microbiota from a Salmonella endemic cattle herd as determined by oligonucleotide fingerprinting of rDNA genes. Vet Microbiol 136: 285– 292.

83. Patil, D.P., Satoor, S.N., Dhotre, D.P., Pawar, V., Patole, M.S., Shouche, Y.S. and Hardikar, A.A. (2010) Fecal microbiota study. Unpublished manuscript.

84. Paul, S.S., Deb, S.M., and Singh, D. (2012) Diversity of bacteria in rumen of buffaloes. Unpublished manuscript.

85. Petri, R.M, Forster, R.J., Yang, W., McKinnon, J.J., McAllister, T.A. (2012) Characterization of rumen bacterial diversity and fermentation parameters in concentrate fed cattle with and without forage. J Appl Microbiol. 112(6):1152-62

86. Pieper, R., Bindelle, J., Rossnagel, B., Van Kessel, A. G., and Leterme, P. (2009) Effect of carbohydrate composition in barley and oat cultivars on microbial ecophysiology and the proliferation of Salmonella enterica in an in vitro model of the porcine gastrointestinal tract. Appl. Environ. Microbiol. 75: 7006–7016.

87. Pope, P.B., Mackenzie, A.K., Gregor, I., Smith, W., Sundset, M.A., McHardy, A.C., Morrison, M., Eijsink, V.G.H. (2012) Metagenomics of the svalbard reindeer rumen microbiome reveals abundance of polysaccharide utilization loci. PLoS One. 7: e38571.

88. Samsudin, A.A., Evans, P.N., Wright, A.D., and Al Jassim R. (2011) Molecular diversity of the foregut bacteria community in the dromedary camel (Camelus dromedaries). Appl. Environ. Microbiol. 11: 3024-35.

89. Schmidt, B., Mulder, I.E., Musk, C.C., Aminov, R.I., Lewis, M., Stokes, C.R., Bailey, M., Prosser, J.I., Gill, B.P., Pluske, J.R. (2011) Establishment of normal gut microbiota is compromised under excessive hygiene conditions. PLoS One. 6, e28284.

90. Schwarzenbach, R.P., Egli, T., Hofstetter, T.B., von Gunten, U., and Wehrli, B. (2010) Global water pollution and human health. Annual Reviews of Environment and Resources 35: 109–136

91. Scott, T. M., Rose, J. B., Jenkins, T. M., Farrah, S. R., and Lukasik, J. (2002) Microbial source tracking: current methodology and future directions. Appl. Environ. Microbiol. 68: 5796-5803.

92. Selvakumar, A., and Borst, M. (2006) Variation of microorganism concentrations in urban stormwater runoff with land use and seasons. J. Water Health. 4: 109–124.

93. Shanks, O.C., White, K., Kelty, C.A., Hayes, S., Sivaganesan, M., et al. (2010) Performance Assessment PCR-Based Assays targeting Bacteroidales Genetic Markers of Bovine Fecal Pollution. Appl. Environ. Microbiol. 76: 1359-1366.

94. Shanks, O.C., Kelty C.A., Archibeque, S., Jenkins, M., Newton, R.J., McLellan, S.L., Huse, S.M., and Sogin, M.L. (2011) Community Structures of Fecal Bacteria in Cattle from Different Animal Feeding Operations. Appl. Environ. Microbiol. 77 (9): 2992-3001.

95. Shinkai, T., Ueki, T., and Kobayashi, Y. (2010) Detection and identification of rumen bacteria constituting a fibrolytic consortium dominated by Fibrobacter succinogenes. Anim. Sci. J. 81: 72–79.

96. Sundset, M., Praesteng, K., Cann, I., Mathiesen, S., and M. R. I. (2007) Novel rumen bacterial diversity in two geographically separated sub-species of reindeer. Microb. Ecol. 54: 424–438.

97. Tajima, K., Nonaka, I., Higuchi, K., Takusari, N., Kurihara, M., Takenaka, A., Mitsumori, M., Kajikawa, H., Aminov, R.I. (2007) Influence of high temperature and humidity on rumen bacterial diversity in Holstein heifers. Anaerobe 13: 57-64.

98. Topo TA cloning Kit for Sequencing Hand Book. Version V. Invitrogen Life Sciences. Carlsbad, CA. April 7, 2004. P 1-31.

99. U.S. Environmental Protection Agency. (1976) Quality Criteria Water Washignton, D.C. U.S. Environmental Protection Agency.

100. U.S. Environmental Protection Agency. (1986) Ambient Water Quality Criteria for Bacteria-1986. EPA 440/5-84-002. Office of Water, Regulations and Standards Criteria and Standards Division, Washington D.C.

101. U.S. Environmental Protection Agency. (2002) National water quality inventory: 2000 report. Publication EPA-841-R-02-001. U.S. Environmental Protection Agency, Washington, DC.

102. Van Kessel, A.G. and Hill, J.E. (2010) Qualitative and Quantitative Determination of Microbial Succession in the Gastro-Intestinal Tract of the Preweaned Pig. Unpublished manuscript.

103. Wade, T.J., Calderon, R.L., Brenner, K.P., Sams, E., Beach, M., Haugland, R., Wymer, L., and Dufour, A.P. (2008) High sensitivity of children to swimming- associated gastrointestinal illness: Result using a rapid assay of recreational water quality. Epidemiology. 19: 375-383.

104. Walker, A.W., Sanders, J.D., Churcher, C. (2011) High-throughput clone library analysis of the mucosa-associated microbiota reveals dysbiosis and differences between inflamed and non-inflamed regions of the intestine in inflammatory bowel disease. BMC Microbiol. 10: 11–17.

105. Walters, S.P., Gannon, V.P.J., and Field, K.G. (2007) Detection of Bacteroidales fecal indicators and the zoonotic pathogens E. coli O157:H7, Salmonella, and Campylobacter in river water. Environ Sci Technol 41: 1856-1862.

106. Wang, Q, G., Garrity, M., Tiedje, J. M., and Cole, J. R. (2007) Naïve Bayesian Classifier for Rapid Assignment of rRNA Sequences into the New Bacterial Taxonomy. Appl Environ Microbiol. 73(16): 5261-7

107. Wang, J.Q., Yu, P., Bu, D.P., Liu, K.L., Li, D. and Zhao, S.G. (2008) Effect of Bacillus subtilis Natto on Rumen Microbial Diversity of Weaning Calves.

108. Wisconsin lower Fox basin map. Wisconsin Dept. of Natural Resources. http://dnr.wi.gov/water/basin/lowerfox/

109. World Health Organ. (WHO). (2009) Global Health Risks: Mortality and Burden of Disease Attributable to Selected Major Risks. Geneva, Switz: WHO

110. World Health Organ. (WHO)/UN Child. Fund (UNICEF). (2006) Meeting the MDG Drinking-water and Sanitation Target: The Urban and Rural Challenge of the Decade. Geneva, Switz./New York: WHO/UNICEF.

111. World Health Organ. (WHO). (1996) Guidelines for Drinking water Quality, volume 2: Health criteria and other supporting information. Geneva, Switz: WHO

112. Wu, Z., Wang, H., Zhang, A., Yang, X., and Guan, Z. (2012) Analysis of fecal bacterial community structures in different species of swine growing in the same conditions. Unpublished manuscript.

113. Yang, L.Y., Chen, J., Cheng, X.L., Xi, D.M., Yang, S.L., Deng, W.D., Mao, H.M. (2010) Phylogenetic analysis of 16S rRNA gene sequences reveals rumen bacterial diversity in Yaks (Bos grunniens). Mol Biol Rep. 37 (1): 553-62.

114. Zoetendal, E. G., Collier, C. T., Koke, S., Mackie, R. I., and Gaskins, H. R. (2004) Molecular ecological analysis of the gastrointestinal microbiota: a Review. J. Nutr. 134: 465-472.

115. Zoetendal, E.G., Vaughan, E.E., and de Vos, W.M. (2006) A microbial world within us. Mol Microbiol 59: 1639–1650.