Amplicon pyrosequencing and ion torrent sequencing of wild duck eubacterial microbiome from fecal samples reveals numerous species linked to human and animal diseases

Introduction

Throughout the history of medicine there has been an awareness of animal to human transmission of disease, and the etiological pathogens have been collectively described as zoonoses¹. Water fowl and wild birds have been identified as reservoirs for the virus Influenza A^2,3; a highly mutable and infectious pathogen that infects avian and mammalian species⁴. Ducks are observed in a multitude of fresh water sources including ponds, water fountains and pools where they can defecate; bacteria have been shown to be distributed through aerosols from ornamental fountains^5,6 and reclaimed water dispensed through an irrigation system⁷. Humans may also have direct contact with ducks and their excrement through the recreational sport of duck hunting⁸. Ducks can also shed pathogens near chicken farms or other animals—such as pigs—that have access to outside areas. An avian influenza A virus (H7N7) epidemic in the Netherlands in 2003 thought to be initiated from a migratory water fowl resulted in the culling of 30 million poultry in an area of the country where free-range poultry farming was common⁹. Due to the migratory nature and unrestrained behavior of the wild duck, Aythya Americana, our study set out to investigate the bacterial microbiome of a wild duck and to identify its bacterial flora relative to the same bacterial species that have been reported to cause disease in farm animals and humans.

Methods

Amplicon pyrosequencing (bTEFAP) was originally described by Dowd et al.¹⁰ and has been used in describing a wide range of environmental and health related microbiomes including the intestinal populations of a variety of animals and their environments including cattle^11–15. Fecal samples obtained from wild ducks, Aythya americana, that were killed during duck hunting season (December 2012) by licensed hunters, were aseptically swabbed onto a Whatman FTA cards (GE Healthcare Life Sciences) using sterile swabs and gloves being careful to avoid environmental contamination. The flaps of the FTA cards were placed over the FTA paper and placed into a sterile pouch, and the FTA cards were stored at room temperature prior to DNA amplification. 2 mm punches were washed with FTA reagent and TE (10 mM Tris-HCL, 1 mM EDTA, pH 8.0) according to the manufacturer’s protocol, and the dried punches were used as template DNA for thermal cycling. DNA was also isolated from 50 mL of pond water as a negative comparison and sampled from a source of water visited by numerous avian species but not at the source of the fecal sampling but within the migratory range of Aythya americana. The pond water DNA was isolated using water RNA/DNA purification kit (0.45 µm) [Norgen Biotek Corp, Thorold, ON, Canada]. For thermal cycling and DNA amplification we used the 16S universal Eubacterial primers 27f 5´-AGAGTTTGATCCTGGCTCAG-3´ and 1492r primer 5´-ACGGCTACCTTGTTACGACTT-3´ (Integrated DNA Technologies). A single-step 30 cycle PCR using EconoTaq PLUS 2X Master Mix (Lucigen, Middleton, WI) were used under the following conditions: 94°C for 2 minutes, followed by 30 cycles of 95°C for 120 seconds; 42°C for 30 seconds and 72°C for 4 minutes; after which a final elongation step at 72°C for 20 minutes was performed. Following PCR, DNA products were resolved in a 1% agarose, 1X TAE gel stained with ethidium bromide and 1.5 Kb products were excised from the gel purified using a cyclo-prep spin column (Amresco, Solon, OH). All the DNA products were purified using Agencourt Ampure beads (Agencourt Bioscience Corporation, MA, USA). Samples were sequenced using Roche 454 FLX titanium instruments and reagents following manufacturer’s guidelines. The Q25 sequence data derived from the sequencing process was processed using a proprietary analysis pipeline (www.mrdnalab.com, MR DNA, Shallowater, TX). Sequences were depleted of barcodes and primers. Next, short sequences < 200bp, sequences with ambiguous base calls, and sequences with homopolymer runs exceeding 6bp were removed. Sequences were then denoised and chimeras removed. Operational taxonomic units (OTUs) were defined after removal of singleton sequences, clustering at 3% divergence--97% similarity^10,15. OTUs were then taxonomically classified using BLASTn against a curated GreenGenes database^16,17 and compiled into each taxonomic level into both “counts” and “percentage” files.

For ion torrent sequencing, the 16S rRNA gene V4 variable region PCR primers 515/806⁶⁰ were used in a single-step 30 cycle PCR using the HotStarTaq Plus Master Mix Kit (Qiagen, USA) under the following conditions: 94°C for 3 minutes, followed by 28 cycles of 94°C for 30 seconds, 53°C for 40 seconds and 72°C for 1 minute, after which a final elongation step at 72°C for 5 minutes was performed. Sequencing was performed at MR DNA (www.mrdnalab.com, Shallowater, TX, USA) on an Ion Torrent PGM following the manufacturer’s guidelines.

The prevalence of pathogenic bacteria was estimated by the means and 95% confidence intervals.

Results

Due to the aquatic nature of the animal, we initially expected that the biodiversity of bacterial species in the duck feces would reflect numerous bacterial species present in the pond water, and since we observed multiple species of aquatic birds in the pond we expected to find eubacteria in common. Figure 1 is a modified heat map showing differences and similarities among the classes of eubacteria sequenced and identified. The figure demonstrates clear differences at the taxonomical level of Class with few common classes of bacteria namely Actinobacteria, Clostridia and Gammaproteobactera.

Figure 1. Comparison of Classes of Eubacteria present in the Duck to the Classes of Eubacteria present in pond water using a modified heat map.

Darker colors represent a higher representation of the bacterial class.

However, similarities at the level of Genus and species included only Agrobacterium tumefaciencs and a species of Porphyromonas and a species of Ruminococcaceae (Figure 2). This analysis indicated distinct differences between the eubacteria present in the duck fecal sample and the pond water sample, and it also indicated that our sampling of the duck feces was devoid of any obvious pond water eubacterial constituents.

Figure 2. Bacterial species present in both duck feces and pond water.

The taxonomical classification of OTU at the level of genus and species was compiled in relation to percentages of the Eubacterial microbiome (Table 1). In Table 2, we referenced reported cases of diseases related to the bacteria sequenced from the duck’s fecal sample reflecting the eubacterial microbiome’s potential to cause disease in humans and other mammals. The largest representative bacterial species—relative to percentage—was Fusobacterium mortiferum at 31.6%. Fusobacterium mortiferum reports related to human disease are sparse, but Fusobacterium have been associated with rare but serious cases of bacteremia^18,19, and a 6 year study of “other gram-negative anaerobic bacilli” (OGNAB) isolated from anaerobic infections at the Wadsworth Clinical Anaerobic Clinical Anaerobic Bacteriology Research Laboratory in Los Angeles, CA reported that most strains of Fusobacteria—outside of Fusobacterium nucleatum—were resistant to erythromycin²⁰. The pathogen, Fusobacterium nucleatum, on the other hand, is well-known for its association with disease and its ability to adhere to Gram-positive and Gram-negative bacteria in dental biofilms such as plaque²¹.

Streptobacillus moniliformis was also identified as a major constituent of the duck fecal eubacterial microbiome at 30.1%. Several well-studied and documented cases of disease are attributed to S. moniliformis including rat bite fever or Haverhill disease²², osteomyelitis²³, epidural abscesses²⁴, fever and polyarthralgia²⁵, bacteremia²⁶ and contaminated drinking water related disease²⁷.

Other organisms and their respective illnesses included Lactobacillus intermedius (11.02%) in a renal transplant infection²⁸, Actinomyces suimastitidis (4.47%) in pig mastitis²⁹ and Campylobacter canadensis (3.69%) in drinking water related disease²⁷. Enterococcus cecorum was another identified pathogen at 3.59% of the sequenced Eubacterial microbiome, and E. cecorum has been reported to cause disease in chicks^30,31 and humans including aortic valve endocardititis³², empyema thoracis³³, septicemia in a malnourished adult³⁴ and recurrent bacteremic peritonitis in a patient with liver cirrhosis³⁵. Actinomyces odontolyticus (0.70%) has recently been reported to cause bacteremia in immunosuppressed patients³⁶, and members of the genus Actinomyces have been known to cause actinomycosis for some time. A. odonolyticus was reported by Michell, Hintz and Haselby in 1997 to be the cause of a malar mass in soft tissue in a human³⁷. A species of the genus Leptotrichia (0.36%) was also identified, a genus that has been associated with bacteremia in multiple myeloma patients receiving chemotherapy³⁸. Another Actinomyces present in the wild duck eubacterial microbiome was Actinomyces turicensis at 0.3%, a bacterium associated with a spectrum of diseases including genital infections, urinary tract infections, skin infections, post-operative wound infection, abscesses, appendicitis, ear and nose and throat infection and bacteremia³⁹. In addition, Actinomyces europaeus (0.14%) was reported in human abscesses⁴⁰, Actinomyces neuii (0.03%) was reported to cause endophthalmitis⁴¹ and periprosthetic infection⁴², Actinomyces vaccimaxillae (0.01%) was isolated from a cow jaw lesion⁴³ and Actinomyces hongkongensis (0.004%) was reported to cause high-mortality bacteremia in humans⁴⁴.

0.24% of the eubacterial population was composed of Plesiomonas shigelloides a well-documented pathogen associated with Travelers’ diarrhea, dysentery and gastroenteritis^45–48. Arcanobacterium pyogenes was also present (0.18%), a pathogen reported to cause soft tissue infections in humans⁴⁹. Atopobium vaginae (0.12%) was reported to cause bacterial vaginosis in a human⁵⁰ and Varibaculum cambriense (0.01%) was reported to cause complications with intrauterine devices and vaginal infections in Hong Kong⁵¹, Parvimonas micra (0.08%) was associated with odontogenic infection⁵² and human bacteremia was reported with Atopobium rima⁵³, Fusobacterium nucleatum⁵⁴, Corynebacterium freneyi⁵⁵ and Streptococcus suis⁵⁶. Finally, Veillonella dispar (0.02%) was reported in a case of septic arthritis⁵⁷ and Porphyromonas gingivalis (0.02%) is a well-studied pathogen reported decades earlier and associated with periodontitis⁵⁸.

Six additional fecal samples from Ayantha americana were processed and sequenced using the ion torrent platform. The number of OTUs increased from 85 to 163, and the rarefraction curve that included OTUs from only Ayantha americana indicated that additional OTUs can be expected due to the lack of a plateau on the graph (data not shown). However, the ion torrent sequencing data in Table 3 supported the 454 pyrosequencing data in that the Anaerotruncus genus was common along with Bacteroides, Blautia, Campylobacter, Cetobacterium, Clostridium, Coprococcus, Erysipelotrichaceae, Faecalibacterium, Fusobacterium, Peptostreptococcacease Pseudobutyrivibrio, Roseburia, Ruminococcaceae, Streptobacillus and Veillonella.

Forty six of the eubacterial OTUs have been reported as pathogens or opportunistic pathogens and on average they represented nearly 72% of the bacteria present, M=71.78, SD=5.96, 95% CI [57.19, 86.37]. Anaerotruncus spp., Campylobacter canadensis, Fusobacterium mortiferum, Helcococcus kunzii, Streeptobacillus moniliformis, Streptococcus suis, Arcanobacterium pyogenes, Clostridium ramosum, Corynebacterium xerosis, Enterococcus faecium, Flavobacterium spp. and Veillonella magna were recognized and evaluated as the most pathogenic representative bacteria. The twelve most pathogenic on average represented nearly 31% of the bacteria present, M=30.95, SD=9.70, 95% CI [7.19, 54.71]. The mean is not normally distributed for a small sample of seven; consequently the confidence intervals were estimated using a t-distribution with six degrees of freedom.

Discussion

Numerous pathogenic eubacterial species have been identified in the fecal samples obtained from the wild duck, Aythya Americana, using amplicon pyrosequencing and ion torrent sequencing, two widely accepted methods for analyzing the bacterial composition of microbial ecosystems. We were surprised to find that most of the species of eubacteria sequenced the duck feces were not present in a pond water sample from a water source that was known to be visited by numerous water fowl. Perhaps, the analyses of small samples from a pond or lake are not adequate when investigating the presence of avian contamination.

The summary in Table 2 indicates that many of the bacteria that are listed are clinically important causing severe diseases such as bacteremia and septicemia. The potential to cause disease can be appreciated when one considers that wild-duck feces can contaminate food, drinking water and open wounds. In addition, bird feces can easily contaminate pond fountains and ornamental fountains--where aerosols are produced—and the aerosols can carry the bacteria in a similar way to what has been reported for Legionella pneumophila⁴⁷. It is possible that many of the bacterial entities when disseminated to humans and other animals could also cause subclinical respiratory illnesses that are not reported due to patient resolution and the empiric treatment of respiratory illnesses with antibiotics without diagnostic testing.

It is only prudent to recommend that immunocompromised humans, the elderly and animals should limit their exposure to environments where ducks are abundant and may have polluted the water source—this includes ponds with fountains, outdoor pools and ornamental fountains that are not properly maintained. That realization also supports the practice of adequately chlorinating or sanitizing artificial pools and fountains to prevent opportunistic infections through aerosols or breaks in the skin.

Duck hunters should also be aware of the risk of bacterial contamination in addition to the risk posed by the influenza virus. Additionally, reclaimed water poses a threat to the elderly and other immunocompromised humans who might be exposed to aerosols that are produced when the reclaimed water is used as a source of irrigation such as in golf courses and gardens, a common practice that might warrant further inquiry.

When determining the cause of disease it is difficult—if not impossible—to identify the source of infection, and whether it has indeed originated from an animal that is migratory or aquatic in nature. Many of the bacterial species that were cited to cause infections among humans were also found in the excrement of a migratory and aquatic bird that is unrestrained and defecates in water supplies and defecates around farm animals that are raised as ‘free-range” and exposed to the environment.

Our analysis of the wild duck was focused on Ayantha americana, a wild duck common to Florida and the Southeastern region of the United States. The statistical analysis of seven fecal samples indicated that the wild duck eubacterial microbiome contained numerous bacteria that were pathogenic or opportunistic pathogens, and the wild duck should be recognized as a vector for bacterial contamination and disease to humans and farm animals.