Antibiotic Resistance in Aquaculture Production DISSERTATION

Home , Josamycin, Piromidic acid, Pseudoclavibacter, Rifamycin, Thiamphenicol

Antibiotic Resistance in Aquaculture Production

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Ying Huang M.S.

Graduate Program in Food Science and Technology

The Ohio State University

2014

Dissertation Committee:

Dr. Hua H. Wang, Advisor

Dr. John H. Litchfield

Dr. Gireesh Rajashekara

Dr. Zhongtang Yu

Copyrighted by

Ying Huang

2014

Abstract

The rapid emergence of antibiotic resistance (AR) has become a major public health concern. Recent findings provided solid evidences suggesting that multiple risk factors contributed to AR development, enrichment, dissemination and persistence, and a comprehensive understanding of AR ecology is essential for targeted mitigation.

Following the investigation of AR in aquaculture products from China, a study on AR in fish and aquaculture production-related samples from a U.S. fish farm with controlled practices and no history of antibiotic applications was further conducted to better understand the potential impact of aquaculture production practice on the prevalence of antibiotic resistant (ART) bacteria in the aquaculture ecosystem.

Phenotypic resistant populations against sulfamethoxazole with trimethoprim

(Sul/Tri), tetracycline (Tet), erythromycin (Erm) or cefotaxime (Ctx) were screened by conventional plating with the corresponding antibiotics, followed by population assessments with denaturing gradient gel electrophoresis (DGGE) and 16S rDNA next generation sequencing (NGS). Despite the absence of antibiotic application in the farm, our results showed that antibiotic resistant (ART) bacteria were abundant in all samples examined, including fish intestine, surface rinsing water, feed, pond water and mud samples. By NGS, a total of 569 genera were identified in Tetr and Ctxr bacteria from five types of samples. Certain correlations of ART bacteria between different types of ii

samples were observed. Various AR genes measured by quantitative PCR (qPCR) were significantly more abundant in fish intestine and feed than farm environmental samples.

Certain dominant ART bacteria subpopulations in feed and pond water were also identified in fish intestine.

About 79% of 4747 ART isolates from aquaculture samples showed resistance to more than one antibiotic. Some ART isolates showed MIC of Sul/Tri, Tet, Erm or Ctx no less than 512 μg/ml. Various AR genes were detected in ART isolates including sul1, sul2, tetS, tetL, tetM, ermB or ermC. Identified AR gene carriers belong to 18 genera. In addition, the AR traits in many isolates were quite stable, even in the absence of selective pressure.

A fosmid genomic library generated from a pooled DNA of 6 ART isolates was used to screen for new AR genes against tetracycline. Tetr clones were further subjected to subcloning using pBluescript vector, followed by DNA sequence analysis. A 4.7-kb fragment from a Tetr subclone T61 contained two divergently transcribed open reading frames (ORFs). The larger ORF encoded a 398-amino-acid protein with 88% identity to a major facilitator superfamily (MFS) transporter and 87% identity to class D tetracycline/H+ antiporter. The smaller ORF encoded a 199-amino-acid protein with 86% identity to TetR family transcriptional regulator. Deletion mutagenesis confirmed the involvement of the two ORFs in Tetr. The original host of the new tetD variant, designated tetD(Y), was identified as Providencia sp.

iii

Our results suggested that additional risk factor(s), rather than direct exposure to antibiotics, are responsible for AR in aquaculture production. Results would contribute to an improved understanding of AR ecology in the aquaculture production system.

Acknowledgments

I would like to thank Dr. Hua Wang, for offering me precious encouragement, guidance, and support throughout my entire Ph.D. study. Also thank her for providing valuable advice on many aspects of my life.

I would also like to thank my committee members: Dr. John Hyland Litchfield, Dr.

Gireesh Rajashekara and Dr. Zhongtang Yu, for their kind help and patience throughout my Ph.D. study.

I would also like to thank Dr. Laura Tiu, who is the aquaculture extension specialist and has helped me get access to the samples.

I am very grateful to all of my lab mates in these four years: Lu Zhang, Xinhui Li, Linlin

Xiao, Andrew Wassinger, Wenfei Wang, Yi Shao, Lei Ye, Qianying Yao, Yang Zhou,

Yu Li and He Yan for their help and encouragement during my research. We had a wonderful time together.

I want to thank all my friends and my family for their unconditional support.

Vita

2007...... B.S. Bioengineering, China Agricultural

University

2010...... M.S. Food Science, China Agricultural

University

2010 to present ...... Graduate Research Associate, Department

of Food Science and Technology, The Ohio

State University

Publications

Peer-reviewed articles:

Zhang L, Huang Y, Zhou Y, Buckley T & Wang HH. (2013) Antibiotic administration routes significantly influence the levels of antibiotic resistance in gut microbiota. Antimicrobial agents and chemotherapy. 57 (8): 3659-3666.

Ye L, Zhang L, Li X, Shi L, Huang Y & Wang HH. (2013) Antibiotic-Resistant Bacteria Associated with Retail Aquaculture Products from Guangzhou, China. Journal of Food Protection. 76 (2): 295-301.

Zhang L, Kinkelaar D, Huang Y, Li Y, Li X & Wang HH. (2011) Acquired Antibiotic Resistance: Are We Born With It? Applied and Environmental Microbiology. 77(20): 7134-7141. vi

Xie Y, An H, Hao Y, Qin Q, Huang Y, Luo Y & Zhang L. (2011) Characterization of an anti-Listeria bacteriocin produced by Lactobacillus plantarum LB-B1 isolated from koumiss, a traditionally fermented dairy product from China. Food Control. 22(7): 1027- 1031.

Hao Y, Zhao L, Zhang H, Zhai Z, Huang Y, &Zhang L. (2010) Identification of the bacterial biodiversity in koumiss by denaturing gradient gel electrophoresis and species- specific PCR. Journal of Dairy Science. 93(5): 1926-1933.

Huang Y, Luo Y, Zhai Z, Zhang H, Yang C, Tian H, Li Z, Feng J, Liu H & Hao Y. (2009) Characterization and application of an anti-Listeria bacteriocin produced by Pediococcus pentosaceus 05-10 isolated from Sichuan Pickle, a traditionally fermented vegetable product from China. Food Control. 20(11): 1030-1035.

Li R, Zhai Z, Yin S, Huang Y, Wang Q, Luo Y, & Hao Y. (2009) Characterization of a rolling-circle replication plasmid pLR1 from Lactobacillus plantarum LR1. Current Microbiology. 58 (2): 106-110.

Yin S, Hao Y, Zhai Z, Li R, Huang Y, Tian H, & Luo Y. (2008) Characterization of a cryptic plasmid pM4 from Lactobacillus plantarum M4. FEMS Microbiology Letters. 285(2): 183-187.

Hao Y, Huang X, Mei X, Li R, Zhai Z, Yin S, Huang Y, & Luo Y. (2008) Expression, purification and characterization of pectin methylesterase inhibitor from kiwi fruit in Escherichia coli. Protein Expression and Purification: 60(2), 221-224.

Presentations at conferences:

Huang Y, Zhang L, Ye L, Wang WF, Tiu L, Wang HH. 2014. Antibiotic resistant microbiome and its potential evolution in aquaculture production. 2014 ASM General Meeting. Boston, MA. 5/20/2014

Zhou Y, Zhang L, Huang Y, Wang HH. The impact of antibiotic administration routes and environmental exposure on antibiotic resistance ecology in poultry gut microbiota. 2014 ASM General Meeting. Boston, MA.

Huang Y, Zhang L, Ye L, Wang WF, Tiu L, Wang HH. 2014. Identification of a new tetD subgroup from domestic aquaculture products by functional metagenomics. 2014 OVIFT Annual Meeting. Columbus, OH. 3/17/2014

Huang Y, Zhang L, Ye L, Wang WF, Tiu L, Wang HH. 2013.. Antibiotic resistance in aquaculture products from domestic production. 2013 OARDC Annual research meeting. Columbus, OH.

vii

Huang Y, Zhang L, Ye L, Wang WF, Tiu L, Wang HH. 2013. Antibiotic resistance in aquaculture products from domestic production. 2013 Ohio Branch ASM General Meeting. Ashland, OH.

Huang Y, Zhang L, Ye L, Wang WF, Tiu L, Wang HH. .2012. Antibiotic resistance in aquaculture products from domestic production. 2012 IFT Annual Meeting. Las Vegas, NV.

Huang Y, Zhang L, Ye L, Wang WF, Tiu L, Wang HH. .Antibiotic resistance in aquaculture products from domestic production. 2012 PHPID Annual Meeting. Columbus, OH.

Zhang L, Kinkelaar D, Huang Y, Wang HH. 2011. Acquired antibiotic resistance: drug, food Exposure or are we born with it? 2011 ASM General Meeting. New Orleans, LA.

Fields of Study

Major Field: Food Science and Technology

viii

Table of contents

Abstract ...... ii

Acknowledgments...... v

Vita ...... vi

Table of contents ...... ix

List of Tables ...... xi

List of Figures ...... xiii

Chapter 1 : Literature Background and Rationale of the Study ...... 1

1.1 Antibiotics and their modes of action ...... 2 1.2 Antibiotic resistance ...... 5 1.3 AR disseminated in ecosystem ...... 9 1.4 Aquaculture development and production ...... 11 1.5 Antibiotics usage in aquaculture ...... 15 1.6 AR in aquaculture ...... 23 1.7 Rationale and objectives of this study ...... 32 Chapter 2 : Profiles of antibiotic resistant bacteria associated with domestic aquaculture

production ...... 34

2.1 Abstract ...... 34 2.2 Introduction ...... 35

2.3 Materials and methods ...... 38 2.4 Results ...... 41 2.5 Discussion ...... 52 Chapter 3 : Characterization of antibiotic resistant bacteria isolates from aquaculture

production ...... 61

3.1 Abstract ...... 61 3.2 Introduction ...... 62 3.3 Materials and methods ...... 64 3.4 Results ...... 68 3.5 Discussion ...... 74 Chapter 4 : Identification of a new tetD variant from domestic aquaculture products ..... 78

4.1 Abstract ...... 78 4.2 Introduction ...... 79 4.3 Materials and methods ...... 81 4.4 Results ...... 85 4.5 Discussion ...... 93 Summary and future direction ...... 96

Bibliography ...... 98

List of Tables

Table 1.1. Antibiotics commonly used in therapeutic treatment and their modes of action.

...... 3

Table 1.2. The production and growth of top 15 aquaculture country producers in 2012.

...... 14

Table 1.3. Antibiotics approved for use in aquaculture by FDA. Adapted from FDA ..... 17

Table 1.4. Reported antibiotic usage in top aquaculture producing countries...... 18

Table 1.5. Reported antibiotic usage in aquaculture in some North American and

European countries ...... 22

Table 1.6. Reported AR determinants in aquaculture system...... 25

Table 2.1. Primers used in this study ...... 40

Table 3.1. Primers used for conventional PCR...... 66

Table 3.2 The percentage of MDR isolates from aquaculture samples ...... 69

Table 3.3 The identity of representative AR gene carriers ...... 71

Table 3.4. Resistance persistence of ART isolates ...... 73

Table 3.5. The donors, recipients and the resulting transformants in transformation

experiments...... 74 xi

Table 4.1. ART bacterial isolates used in the study...... 82

Table 4.2. Putative ORF located in plasmid pT61...... 87

Table 4.3 PCR primers for identification of tetD(Y) gene carrying host...... 91

xii

List of Figures

Figure 1.1. Dissemination of antibiotics and antibiotic resistance within the ecosystem..

...... 10

Figure 2.1. ART bacterial populations in fish and related samples collected in October

2012...... 43

Figure 2.2 Distribution of bacterial phylum of Tetr (A) and Ctxr (B) populations on BHI

plates associated with different types of samples...... 47

Figure 2.3 Distribution of bacterial family of Tetr (A) and Ctxr (B) populations on BHI

plates associated with different types of samples...... 48

Figure 2.4 Distribution of bacterial genus of Tetr (A) and Ctxr (B) populations on BHI

plates associated with different types of samples...... 49

Figure 2.5 Quantitative PCR of 16S rDNA gene and AR gene pools in aquaculture

samples collected in October 2012...... 51

Figure 2.6 ART bacterial population from BHI plates (A) and MACG plates (B) collected

at the other four different time points ...... 57

Figure 2.7 DGGE assessment results of cultured ART bacteria from aquaculture samples.

...... 59

xiii

Figure 4.1 Determination of the size of the insertion fragment in plasmid pT61...... 85

Figure 4.2. Schematic diagram of the plasmid pT61 from the Tetr subclone T61...... 88

Figure 4.3. Schematic diagram of phylogenetic relationship of TetD(Y) ...... 89

Figure 4.4. Scheme of transmembrane structure of TetD (Y) by SMART...... 90

Figure 4.5. Complementary DNA sequence of the 80-bp tetD(Y)–tetR(Y) intergenic

region...... 91

Figure 4.6 Electrophoresis and southern hybridization of genomic DNA of six isolates.

...... 92

xiv

Chapter 1 : Literature Background and Rationale of the Study

Antibiotics have been widely used in human and veterinary medicine to protect human and animal health against pathogens. People once thought almost all the infectious diseases could be treated with antibiotics. However, the incidences of antibiotic resistance (AR), once sporadic until 1960s, have turned into a global concern today due to the rapid emergence of infectious diseases by antibiotic resistant (ART) pathogens, ranging from well-known methicillin-resistant Staphylococcus aureus (MRSA), vancomycin-resistant Enterococci (VRE), to carbapenem-resistant Enterobacteriaceae

(CRE) and fluoroquinolone-resistant Clostridium difficile. According to CDC 2013 report

(1), deaths related to C. difficile infection increased 400% between 2000 and 2007 in the

United States, in part due to the emergence of fluoroquinolone-resistant C. difficile at

2000. At least 2 million people acquire health care for ART bacterial infections and at least 23 thousands people die for the infections each year in the United States (1). In addition, infections causing by ART pathogens contribute to about $20 billion in excess direct healthcare costs and about $35 billion of additional costs to society each year in the

U.S. (1). This global antibiotic resistance (AR) issue has become a priority for many agencies and organizations across the world and significant investments have been made to combat the AR challenge. Current mitigation strategies were primarily focused on

limiting the applications of antibiotics in human medicine and food animal production.

However, the resistance issue is more severe and complicated than previously thought, and the existing AR surveillance and control strategy might have not been perfectly targeted (2–4).

Aquaculture has become a significant industry segment due to its high social and economy impact. The annual production of aquaculture worldwide has increased from less than 1 million tonnes in 1950 to 66.6 million tonnes in 2012 with an estimated value of US $137.7 billion (5). The safety of aquaculture products has become an issue of concern for consumers, with the primary focus on the presence of antibiotic residues and foodborne pathogens. AR has emerged as a food safety concern in aquaculture production in recent years, but the potential correlation between aquaculture production practice and AR within the aquaculture system has yet to be revealed. A comprehensive understanding of the risk factors during the aquaculture production is essential for achieving effective AR mitigation.

1.1 Antibiotics and their modes of action

Antibiotics are compounds that kill or inhibit bacterial growth (bactericidal vs bacteriostatic). They can be either synthetic or produced naturally by microbes. Since penicillin was discovered by Alexander Fleming in 1928, more antibiotics have been found or synthesized. So far more than 5000 antibiotics have been discovered and about

100 (belong to fifteen classes) of those have been widely used in therapeutic treatment (6).

Table 1.1 lists antibiotics commonly used in therapeutic treatment and their targets.

Table 1.1. Antibiotics commonly used in therapeutic treatment and their modes of action.

Adapted from (7)

Antibiotic class Examples Target

β-lactam Penicillin, Cephalosporin, Penem, Monobactam Peptidoglycan synthesis

Aminoglycoside Gentamicin, Streptomycin, Spectinomycin Translation

Glycopeptide Vancomycin, Teicoplanin Peptidoglycan synthesis

Tetracycline Oxytetracycline, Minocycline, Tigecycline Translation

Macrolide Erythromycin, Azithromycin Translation

Lincosamide Clindamycin Translation

Streptogramin Synercid Translation

Oxazolidinone Linezolid Translation

Phenicol Chloramphenicol Translation

Quinolone Ciprofoxacin DNA replication

Pyrimidine Trimethoprim C1 metabolism

Sulfonamide Sulfamethoxazole C1 metabolism

Rifamycin Rifampin Transcription

Lipopeptide Daptomycin Cell membrane

Cationic peptide Colistin Cell membrane

So far five main modes of action of the antibiotics have been recognized. First, some antibiotics can interfere with cell wall synthesis, such as β-Lactams and glycopeptides. Biosynthesis of cell wall peptidoglycan requires a crosslinking of peptidyl moieties on adjacent glycan strands. The D-alanine–D-alanine transpeptidase, which catalyzes this crosslinking, is the target of β-lactams. Glycopeptides, in contrast, do not inhibit an enzyme, but directly bind to D-alanine–D-alanine and prevent subsequent crosslinking by the transpeptidase (8). Second, some antibiotics exert their action through inhibiting protein synthesis. The bacterial ribosome comprises of two subunits, 30S and

50S. Because of the importance of rRNA in protein synthesis, the rRNA –rich surface on the 30S and 50S subunits are the targets of most ribosome inhibitors. For example, spectinomycin, tetracycline, pactamycin, hygromycin B, streptomycin, geneticin, and paromomycin target at 30S subunits while there are target sites on 50S subunits for

Thiostrepton, Avilamycin, Streptogramin, Chloramphenicol, Puromycin, Pleuromutilins and Macrolides (9).Therefore, the 30S subunit loses its main function of deciphering the genetic information encoded in the mRNA by such an inhibitor. And the inhibitor of 50S subunit hinders its principle function, including controlling GTP hydrolysis, the formation of peptide bonds, and translocating the peptide through the subunit channel

(10). Third, some antibiotics can interfere with nucleic acid synthesis. Due to its DNA cleavage activity, fluoroquinolones disrupt DNA synthesis during DNA replication.

Through binding to the β-subunit of the polymerase, rifampin can hinder bacterial RNA polymerase, which plays a role in RNA synthesis. Fourth, certain metabolic pathways are the targets of some antibiotics. For example, trimethoprim (a folic acid analogue)

combining with sulfamethoxazole (a sulfonamide) can block the pathway of folic acid synthesis which is important in DNA synthesis. Finally, certain antibiotics can disrupt bacterial membrane structure, such as polymyxins and daptomycin. Polymyxins can increase bacterial membrane permeability leading to the leakage of cell contents. The insertion of the lipid tail of daptomycin into the bacterial cell membrane results in membrane depolarization and eventual death of the cell.

1.2 Antibiotic resistance

Antibiotic resistance (AR) was recognized shortly after the discovery of antibiotics. Resistance to penicillin in some strains of Staphylococcus was found in 1947, just four years after the drug started being mass-produced. Likewise, very soon after their introduction in the late 1940s, resistance to streptomycin, chloramphenicol and tetracycline was noted (http://www.textbookofbacteriology.net/resantimicrobial.html).

Methicillin-resistant Staphylococcus aureus (MRSA) was first detected in Britain in 1961, and now is ‘quite common’ in hospitals all over the world. MRSA strains account for >60% of S. aureus clinical isolates in Japan, Singapore and Taiwan, >50% in Italy and Portugal, and 34% in the United States (11). It is important to recognize the main mechanisms of resistance and the pathways that lead to the development of resistance in bacteria.

1.2.1 Intrinsic AR

Bacteria may be inherently resistant to an antibiotic due to intrinsic (natural) mechanisms. Inherited structural or functional characteristics lead to tolerance of a

particular antibiotic within a particular species. For instance, natural insensitivity may be due to barriers for antibiotics to penetrate into the bacterial cell or the lack of an appropriate target site in the bacterial cell, etc. Gram-negative bacteria are naturally resistant to vancomycin because this drug cannot penetrate the gram-negative outer membrane. The intrinsic resistance of Leuconostoc to vancomycin results from its lacking of appropriate cell wall precursor target for vancomycin to bind and inhibit cell wall synthesis (8).

1.2.2 Acquired AR

Bacteria may develop AR by spontaneous mutation(s) such as point mutation(s), deletion(s), inversion(s), insertion(s), etc. within the bacterial genome, selected by environmental factors due to survival advantage, and passing to offspring through vertical proliferation. But the mutation frequency leading to resistance is generally low and may vary in different species and is affected by many factors. For example, there is a big difference in mutation frequency in developing resistance against β-lactams among

Enterobacter cloacae, Citrobacter freundii and E. coli due to the significantly increasing expression level of chromosomal AmpC β–lactamase. E. coli resistant mutants, overproducing the enzyme, occurred at the low frequency of about 10-9 (12, 13) whereas the resistant mutants of E. cloacae and C. freundii occurred at higher frequency of about

10-6~10-7 (14, 15). In addition, bacterial mutation rates can transiently increase due to the environmental conditions. Starvation and other situations that cause bacterial stress, including induction of the SOS response, all affect the mutation rate (16).

Meanwhile, bacteria may develop AR by acquiring resistance-encoding genes through horizontal transfer gene (HGT) mechanisms. HGT is the most important avenue in rapid dissemination of resistance among the bacterial populations within or across genus or species. HGT involves three main mechanisms: conjugation, transformation and transduction (17). In conjugation, a previously susceptible recipient bacterium acquires

DNA from a resistant donor cell by direct contact. Transformation refers to the uptake of short fragment of naked DNA and incorporating it into the genome of the transformable bacterial cell. Transduction involves transfer of DNA from one bacterium into another mediated by bacteriophage(s).

Besides the chromosome, the genes encoding AR can also be carried by mobile genetic elements such as plasmids, transposons or integrons. These mobile elements can serve as vectors, effectively disseminating resistance encoding genes to recipient cells via

HGT mechanisms. A plasmid is an extra-chromosomal element, often a circular DNA, capable of autonomous replication within a suitable host. Since the first resistance (R) plasmid was detected in the 1950s, plasmids carrying drug resistance genes have been detected from most bacterial groups, indicating a large pool of R plasmids among antibiotic-resistant bacteria (18). Transposons encode a site-specific transposase which is responsible for site specific insertions and excisions. An integron consists of a site- specific integrase and the corresponding DNA target sequence. Resistance gene cassettes are integrated into such target sites mediated by the integrase. Moreover, integrons are believed to play a critical role in the rapid dissemination of multi-drug resistance (MDR) among bacteria.

1.2.3 Mechanisms of resistance

Enzymatic inactivation of antibiotics is the most well-known mechanism of resistance. This resistance is specific for a single drug class or some members of the class.

It occurs when the bacteria produce one or more enzymes that chemically degrade (such as β-lactamase) or modify (such as aminoglycoside acetyltransferase) the antibiotic(s) making them inactive against the bacteria. β-lactamase is the hydrolytic enzyme that disrupts the amide bond of the characteristic β-lactam ring, before the antibiotic can get to the site of cell wall synthesis, leading to the antibiotics inactivation (19). Robicsek et al. reported that aminoglycoside acetyltransferase can reduce the activity of ciprofloxacin

(one type of fluroquinolone) by N-acetylation at the amino nitrogen on the piperazinyl substituent of the drug (20).

Extrusion by efflux pumps. Besides taking up the antibiotics like sensitive ones, certain resistant bacteria employ an effective transport system to pump the antibiotic molecules out of the cell. The active function of those pumps prevents the intercellular antibiotics reaching toxic concentration. Since the tetracycline efflux system was first reported in E. coli, many efflux pumps for other different antibiotics have been discovered in many organisms (21, 22).

Modification of the antibiotic target site is another mechanism which involves changing the structure of the target, resulting in reduced affinity for the antibiotics. For example, a rifampin-resistant mutant of Staphylococcus was due to the structure change of RNA polymerase resulting from mutation in rpoB, the gene encoding the subunit of

RNA polymerase (23). The mutants with altered ribosomes protein S12 became resistant 8

to streptomycin (24). Modifications in the structural conformation of penicillin-binding proteins (PBPs) were observed in certain types of penicillin resistant strains (25).

Furthermore, another mechanism, called “bypass,”occurs when a bacterium develops an alternative, additional system to bypass the existing, primary target. The case of MRSA is a good example. Methicillin resistance in MRSA strains is due to the acquisition of the mecA gene via HGT from an unidentified species. The mecA encodes

PBP2a, a newly discovered PBP unlike any of the PBPs normally produced by S. aureus

(<21% sequence identity). Since PBP2a has unusually low β-lactam affinity, it remains active to allow cell wall synthesis at normally lethal β-lactam concentrations (26).

1.3 AR disseminated in ecosystem

In the past several decades, it has been accepted that the clinical setting is a major avenue where AR develops. And restriction of the use of antibiotics during clinical treatment has been regarded as the primary AR mitigation strategy for a long time.

However, besides clinical applications, antibiotics are also widely used in agriculture, particularly in food animal production for growth promotion, prophylaxis and therapy. It has been reported that 29.9 million pounds of antibiotics were sold in the United States for meat and poultry production while only 7.7 million pounds were sold for human medicine in the same period (http://www.pewtrusts.org/en/multimedia/data- visualizations/2013/recordhigh-antibiotic-sales-for-meat-and-poultry-production).

Therefore, clinical usage of antibiotics is one important reason but not the only reason for

the development of AR, and the resistance issue is more complicated than previously thought.

Since 2005, results from several lines of studies provided solid evidence suggesting that multiple risk factors contributed to AR development, enrichment, dissemination and persistence. It has been proposed that the whole ecosystem is involved in AR dissemination (27, 28). As shown in Fig.1.1, each compartment within the ecosystem is not isolated but rather related to or affected by each other.

Figure 1.1. Dissemination of antibiotics and antibiotic resistance within the ecosystem.

Adapted from (28).

The prevalence and abundance of ART bacteria in retail and restaurant foods including many ready-to-eat items, with transmissible AR genes in a broad-spectrum of commensal and even beneficial bacteria, indicated that the food chain serves as a critically important avenue transmitting AR to the general public (2, 29–32). ART bacteria are further enriched in host gastrointestinal (GI) system, even without the direct exposure to antibiotics (3, 33, 34). The AR-rich feces from both human and animals likely served as the most impactful source for the environmental AR gene pool (35).

More importantly, the targeted AR gene pool was not observed in feces in the mouse model receiving antibiotic treatment, if the host did not receive oral inoculation of AR gene-containing bacteria before the treatment (36). These data clearly illustrated that oral exposure to ART bacteria is a critical risk factor for the rapid dissemination of AR in the global ecosystem, independent from antibiotic treatment. In addition, the food chain serves as an important avenue bridging the AR flow between the environment and hosts.

Therefore AR mitigation in the food chain is essential to effectively combat this significant food safety and public health challenge (2, 3).

1.4 Aquaculture development and production

Aquaculture is the rearing of aquatic organisms including finfish, shellfish and aquatic plants. As the rudiment of modern aquaculture, the ancient fish farming could be traced back to 2500 B.C. in Egypt and 1100 B.C. China, respectively (37). The rapid development of modern aquaculture in the global context, especially in Asian countries, 11

has occurred from 1950 (38). Its main purpose is to provide food fish for human consumption, as well as providing ornamental fishes and recreational fishes.

Compared to terrestrial farming, aquaculture has a richer diversity which resulting from the differences in aquatic environments (freshwater, brackish water, saltwater), species (567 species registered in 2012 FAO statistics), production systems (open, semi- closed, closed system) and practices (39). In the rearing process of aquaculture, the aquatic organisms are cultivated by individuals, groups or corporations and some interventions are applied to enhance the production, such as formulated feed, medical treatment, various stocking system, controlled breeding (38).

Aquaculture as an important industrial segment has grown rapidly in the past 60 years. Its rapid development was stimulated by large food fish demands of the fast- growing consumer population and gradually over-exploited capture fisheries worldwide.

In 2010, fish accounted for 16.7 % of animal protein consumed by the global population

(40). In contrast to world capture fisheries production, which has remained almost stagnant since the mid-1980s, aquaculture worldwide has expanded at an annual average rate of 8.6% between 1980 and 2012. The annual production of aquaculture worldwide

(exclude aquatic plants) has increased from less than 1 million tonnes in 1950 to 66.6 million tonnes in 2012 with an estimated value of US $137.7 billion. Globally, aquaculture accounted for 42.2% of total world food fish production (158 million tonnes) for human consumption in 2012, up from 9% in 1980 (40). In China, the world’s largest aquaculture producer, 80.2% of seafood consumed was derived from aquaculture in 2008, up from 23.6 percent in 1970 (41). 12

There exists an imbalance of the aquaculture development and production across different regions and countries. As shown in Table 1.2, the aquaculture production of the top 15 countries accounted for 92.7% of world aquaculture production in 2012. Out of 15 countries, 10 countries were from Asia which suggested that Asia-Pacific region dominates the world aquaculture. China had the largest share (61.7%) of total aquaculture production worldwide in 2012; while the United States ranked 15th with 2.9% of average annual growth rate between 1990 and 2012.

Table 1.2. The production and growth of top 15 aquaculture country producers in 2012.

Adapted from (40, 41).

Average annual growth rate Production Share in world total Country between 1990 and 2012 (tonnes) (percentage) (percentage)

China 41108306 61.7 20.3

India 4209415 6.3 15.3

Viet Nam 3085500 4.6 34.4

Indonesia 3067660 4.6 19.9

Bangladesh 1726066 2.6 24.5

Norway 1321119 2 24.2

Thailand 1233877 1.9 15.5

Chile 1071421 1.6 42.1

Egypt 1017738 1.5 32.3

Myanmar 885169 1.3 62.3

Philippines 790894 1.2 7.6

Brazil 707461 1.1 37.2

Japan 633047 1 -2.4

South Korea 484404 0.7 2.5

United states 420024 0.6 2.9

Top 15 subtotal 61762101 92.7 NA

Rest of world 4871152 7.3 NA

world 66633253 100 NA

1.5 Antibiotics usage in aquaculture

The rapid growth of aquaculture has been impeded by sporadic outbreaks of infectious diseases (42–50). Disease outbreaks inevitably resulted in some loss of aquaculture production. In 2010, the loss of aquaculture production in China caused by disease has reached to 295,000 tonnes (5). The occurrence of disease outbreaks was largely related to the suboptimal hygienic and stressful conditions present in aquaculture environment of high stocking density (51). To prevent and treat bacterial infection and then reduce the mortality loss, antibiotics have been used in aquaculture as therapeutic and/or prophylactic agents. The first research record of use of antimicrobials in aquaculture was done by Gutsell (52). In this study, the mortality of Trout resulting from the furunculosis declined after the administration by sulfa drugs including sulfamerazine, sulfathiazole, furacin, sulfanilamide and sulfadiazine. Currently, antibiotics are commonly delivered in aquaculture in the ways of bath immersion and/or oral feeding

(http://www.fda.gov/AnimalVeterinary/DevelopmentApprovalProcess/Aquaculture/ucm1

32954.htm).

Generally speaking, aquacultural use of antibiotics in industrialized countries has been subjected to more stringent restriction than that in developing countries which have larger aquaculture production. Four types of antibiotics are approved by U.S. Food and

Drug Administration (FDA) for use in aquaculture (Table 1.3). However, there are a large number of aquaculture facilities worldwide and some aquaculture farmers inappropriately use of antibiotics for economic reasons. In some countries, regulatory agencies are not successful in tracking and limiting antibiotics usage. Therefore information regarding the

specific types and the amounts of antibiotics used in the aquaculture industry worldwide has not been completely publicly available. Some authors have reported or summarized antibiotic usage in aquaculture in several countries (38, 51, 53–56). But the data were inevitably biased and incomplete depending on the availability of tracking systems and research investigations in different countries. Here, Tables 1.4 and 1.5 summarize the information reported on antibiotic use in aquaculture. Of the data from 20 aquaculture- producing countries, oxytetracycline, oxolinic acid and sulfa/trimethoprim were the most commonly used antibiotics, used in 17 countries and 13 countries, respectively. Japan,

Vietnam, Thailand and China used a larger number of antibiotics, reaching to 22, 21, 14 and 14, respectively.

It has been recognized that the inappropriate use of antibiotics in the aquaculture leads to some problems including bacterial resistance, antibiotic residues in tissues of food products and aquaculture environment, and the cost of studying and preventing unintended effects. The occurrences of antibiotic residues in the aquaculture environment as well as aquaculture products have been often reported. In a study conducted in Viet

Nam, trimethoprim, sulfamethoxazole, norfloxacin and oxolinic acid were detected in water and mud in shrimp ponds and surrounding canals with the concentrations of 1.04 ppm to 820.49 ppm (57). A total of 132 samples from Baiyangdian Lake in China showed that sulfonamides were the dominant antibiotics in the water whereas quinolones were prominent in sediments, aquatic animals and plants (58). According to FDA, chloramphenicol, nitrofurans, and fluoroquinolones were the most commonly detected antibiotic residues in imported aquaculture products during the period between 2004 and

2007, especially from Vietnam, Bangladesh, PRC, Indonesia, India, Chile, and Taiwan

(http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformati on/Seafood/ucm150954.htm). Antibiotic residues in the food products as well as the environment, at lower concentration but over longer time periods usually, would lead to the development of the ART bacteria and AR genes in the environment, food products and humans exposed to such environment and food. Moreover, when the food products with antibiotic residues are eaten, the normal intestinal microflora might be altered.

Table 1.3. Antibiotics approved for use in aquaculture by FDA. Adapted from FDA website

(http://www.fda.gov/AnimalVeterinary/DevelopmentApprovalProcess/Aquaculture/ucm132954.h tm)

Approved antibiotics Administration route Species/Class

Oxytetracycline Immersion Finfish fry and fingerlings hydrochloride

Freshwater-reared finfish including warm Florfenicol Oral via feed water finfish, salmonids and catfish

Oxytetracycline Salmonids, catfish, Oncorhynchus mykiss Oral via feed dihydrate and lobster

Sulfadimethoxine/ Oral via feed Catfish, Salmon, Trout ormetoprim

South Bangla- Antibiotics China India Japan Philippines Indonesia Thailand Chile Norway Vietnam USA Korea desh

Sulfonamides category

sulfamerazine Y Y

Sulphamonometh- Y oxine

Sulfamethoxazole Y

Sulfadimidine Y

Sulfadimethoxine Y Y Y Y

8 Ormetoprim Y Y Y

trimethoprim/ Y Y Y Y Y sulfadiazine Tetracycline category

Chlortetracycline Y Y Y Y

Oxytetracycline Y Y Y Y Y Y Y Y Y Y Y

Doxycycline Y Y

Continued

Table 1.4. Reported antibiotic usage in top aquaculture producing countries. Adapted from (38, 51, 53–56). Y: this antibiotic

was used as reported.

Table 1.4. continued

Penicillin category

Ampicillin Y Y Y Y

Amoxycillin Y Y Y Y Y

Benzyl penicillin Y Y Y

Quinolone category

Ciprofloxacin Y Y

Levofloxacin Y

Enrofloxacin Y Y Y Y

Norfloxacin Y Y Y

Oxolinic Acid Y Y Y Y Y Y Y Y

Perfloxacin Y

Flumequine Y Y Y Y

Piromidic acid Y

Macrolide category

Erythromycin Y Y Y Y Y Y Y

Spiramycin Y

Kitasamycin Y

Continued

Table 1.4. continued

Josamycin Y

Nitrofurans category

Furazolidone Y Y Y Y

Aminoglycoside category

Neomycin Y

Kanamycin Y

Gentamicin Y Y Y Y Y

20 Aparamycin Y

Other categories

Chloramphenic- Y Y Y Y Y Y Y Y ol

Florfenicol Y Y Y Y

Thiamphenicol Y Y Y

Tiamulin Y

Nalidixic acid Y Y Y

Miloxacin Y

Continued

Table 1.4. continued

Lincomycin Y

Novobiocin Y

Colistin Y Y

Cephalexin Y

Reported # of 14 4 22 10 4 3 4 14 9 7 21 5

antibiotics used

by country

Table 1.5. Reported antibiotic usage in aquaculture in some North American and European countries. Adapted from (51, 53,

55)

Canada United Spain Italy Iceland Greece Finland Denmark

Kingdom

Oxytetracycline Y Y Y Y Y Y

Florfenicol Y

Sulfa/trimethoprim derivatives Y Y Y Y Y Y Y

Oxolinic acid Y Y Y Y Y

Sarafloxacin Y Y

Amoxycillin trihydrate Y Y Y

Co-trimazine Y

Ampicillin Y

Chlotetracycline Y

Flumequin Y Y

Florfenicol Y Y

1.6 AR in aquaculture

1.6.1 AR in the aquaculture system

Certain ART pathogens were reported to be associated with the aquaculture system, such as Aeromonas (59–70), Vibrio (70–76) and Salmonella (70, 77–79). In addition, large numbers of ART bacteria associated with a wide spectrum of commensal bacteria have been found in various aquaculture products and the farm production systems (80–90). Overall, research on AR patterns found that ART bacteria including multi-drug resistant bacteria were common in aquaculture. Bacteria resistant to oxytetracycline, sulfa/trimethoprim and oxolinic acid have been repeatedly found in aquaculture farms (51).

Regarding the widespread of AR in aquaculture, the presence of AR determinants such as AR genes, insertion sequence, integrons, genomic islands, and transposons, also gained attention. Those AR determinants potentially could disseminate and circulate within the ecosystem via HGT mechanism. Bacteria in aquaculture systems have been reported as a source of AR determinants. Table 1.6 lists some reported AR determinants in aquaculture systems. As shown in the Table 1.6, numerous AR genes were detected in various aquaculture organisms and aquaculture environments. Tetracycline resistant genes were commonly detected and the diversity was very broad. Different resistance gene cassettes were found to be carried by class1 integron. Moreover, some of these AR genes detected in aquaculture systems were found to be carried by mobile genetic elements including plasmids (64, 83, 88, 90–92) and integrons (59, 66, 67, 71, 93). The plasmids carrying AR genes were transferred to E. coli by conjugation or transformation 23

from various genera, such as Aeromonas spp.(64, 67, 90), Vibrio spp.(76, 90) and

Edwardsiella spp.(92).

AR was more prevalent in aquaculture products than did in wild-caught seafood products. For example, more ceftriaxone- and tetracycline-resistant bacteria isolates were detected in farm-raised imported shrimps than wild-caught South Carolina shrimps (94).

Besides the aquaculture products, AR was also more prevalent in the aquaculture environment (water and sediment) compared to some other aquatic environment, such as the unpolluted stream (95) and distant control site (96). Therefore, it has been proposed that the use of antibiotics usage in aquaculture would promote AR issue in aquaculture systems, involving of selection for ART bacteria, change of AR pattern and alteration of the biodiversity (51). Various studies have investigated the impact of antibiotic treatment in aquaculture on the AR ecology and suggested a causal relationship resulting from a selective pressure (61, 81, 97–99). However, a strong causal relationship was not always present for increased AR. For example, remarkable proportions of resistant bacteria with similar resistance pattern were recovered in both florfenicol-treated and untreated scallop larval cultures from a commercial hatchery (84). And water from farms with recent oxytetracycline treatment had significantly higher tetr detection frequencies than did water from farms without recent oxytetracycline treatment whereas sediment samples showed the opposite results (100). Therefore, it is now recognized that antibiotic usage is an important but not only factor led to AR, and multiple risk factors contribute to the AR ecology in aquaculture.

Detected AR determinants Sample Bacteria genera Reference

sul1, sul2, dfrA1, intI1 Sediments (total DNA) - (101)

tetA, tetC, tetH, tetM, tetE, tetG, tetW Sediments (total DNA) - (102)

tetM, tetL Sediments (total DNA - (103)

tetM, tetO, tetT, tetW, sul1, sul2 Water and sediment Acinetobacter lwoffii, Bacillus cereus, B. (82)

megaterium, B. subtilis

tetB, tetD, blaTEM, aadA, class 1 integron carrying Commercial fish and E. coli (93)

gene cassette array dfrA12-aadA2 or aadB- seafood

25 aadA2, class 2 integron carrying gene cassette

array dfrA1-sat-aadA1

blaCTX-M-14, blaCTX-M-79, blaSHV-1, blaSHV-11, blaSHV-25, Farmed fish E. coli (104)

blaSHV-26, blaSHV-27, blaLEN-17, blaLEN-26, qnrB,

qnrS, qnrD, aac(6’)-Ib-cr

tetL, tetM, tetK, blaZ, msrC Seawater and sediment Enterococci spp. (80)

Continued

Table 1.6. Reported AR determinants in aquaculture system.

Table 1.6. continued

class 1 integrons carrying 12 different gene Diseased fish Aeromonas spp. (59)

cassettes dfrA28-orfV, dfrA12-orfF, orf-aadA5,

dfrB3-aadA1, dfrA1-aadA1, dfrA14-recombined

aadA6, catB3-aadA2, dfrA12-orfF-aadA2, dfrB4-

catB3-aadA1, dfrB1-aadA1-catB2, dfrA12-orfF

with integrated IS630-aadA2, arr-3–aacA4–

blaOXA-10–aadA1

class 1 integrase, aadA, sul1, qac1, tetA, tetC Farm-raised fish and Aeromonas spp. (60)

sediments

class 1 integrons, ICEVflTha1, ICEVchTha1, Farmed marine Vibrio spp. (71)

ICEVpaTha1, ICEVchTha2, ICEVflTha2, shrimps

ICEVchTha3, ICEVvuTha1, mutation in the

quinolone resistance determining region (QRDR)

of gyrA

Continued

Table 1.6. continued

tetA, tetB, tetC, tetD, tetE, tetG, tetM, Water and sediments (total DNA) - (100)

tetO, tetQ, tetS

tetM/O/S, tetM, tetS, ermB Diseased olive flounder Streptococcus spp (105)

sul1, sul2, sul3 Water Acinetobacter spp., Aeromonas spp., (91)

Arthrobacter spp., Bacillus spp.,

Brachybacterium spp., Cellulosimicrobium

sp., E.coli, Enterobacter sp.,

Pseudoalteromonas spp., Wautersiellai sp.,

Shigella sp.,

tetA, tetB Farm-raised catfish intestine Citrobacter spp. (106)

tetA, tetD, tetE, tetM Fish, prawn, water Aeromonas spp., Edwardsiella sp., (90)

Flavobacterium sp, Vibrio spp.,

tetM, tetS Fish and seawater Lactococcus garvieae, Photobacterium, (107)

Vibrio spp.,

tetE Farmed fish, water and sediment Aeromonas spp. (64)

Continued

Table 1.6. continued

tetA, tetD, tetE, tetM Fish, prawn, water Aeromonas spp., Edwardsiella sp., (90)

Flavobacterium sp., Vibrio spp.

tetA, tetD, tetB, tetG Diseased fish Edwardsiella tarda (92)

tetA, tetB, tetD, tetE and tetH, class 1 integron Tilapia, trout and koi Aeromonas spp. (66)

carrying resistance gene cassettes ant(3’’)Ia, aquaculture systems

aac(6’)Ia, dhfr1, oxa2a and/or pse1

tetA, tetB, tet34, tetH, tetE, tet35, tetL, Tn5706- Water, salmon A.hydrophila, A. radioresistens, (83)

associated IS element IS1597 fingerlings, food Brevundimonas vesicularis E. coli, E.

pellets sakazakii, Moraxella sp., Morganella

morganii, Pseudomonas spp., Serratia

liquefaciens, Stenotrophomonas

maltophilia,

Continued

Table 1.6. continued

tetA, tetD, tetE, class 1 integron carrying Fish, water, sediment Aeromonas spp. (67)

resistance gene cassettes dhfr2a-ant(3’’)1a,

dhfr1-ant(3’’)1a, dhfr1-ant(3’’)1a-catB2, dhfr1-

ant(3’’)1a-orf-catB2

sul1, sul2, intI1, tetE, tetM, tetL, tetS, blaTEM, Multiple aquaculture Acinetobacter spp., Actinobacterium (88)

blaCMY, ermB, ermC products including spp., Aeromonas spp., Arthrobacter

tilapia, crucian carp, sp., Brevibacterium sp., C. freundii, E.

catfish, and shrimp coli, Klebsiella spp., Kocuria

rhizophila, Morganella sp., Myroides

sp., Paracoccus sp., Plesiomonas sp.,

Proteus mirabilis, Pseudomonas sp.,

Psychrobacter sp., Shewanella sp.,

1.6.2 Integrated aquaculture and its impact on AR

Integrated fish farming has become popular practice throughout Southeast Asia, and manure enrichment is also practiced in the U.S. In the integrated system, different organisms are co-cultivated so the waste of one cultured species can be utilized as food for another cultured species directly or indirectly. The common practice is to combine livestock production with fish production. Animals often chicken and pigs are kept in cages above or adjacent to fish ponds, and animal manure is excreted into the ponds and directly consumed by fish. In addition, the released nutrients can support the growth of photosynthetic organisms which are also the food of fish (108, 109). However, previous studies have indicated that using animal manure in integrated fish farms, which usually use antibiotics containing feed as growth promoters or to prevent or treat diseases, could be associated with increased AR in bacteria from fish intestines (110) and water-sediment

(111, 112) than traditional fish farming. Moreover, the input of manure may also have an impact on species composition. It was reported that E. faecium and E. faecalis were most prevalent isolates in the integrated chicken-fish farms and exerted the highest prevalence of resistance, whereas E. casseliflavus and E. mundtii were the predominant species in traditional fish farms (113). Indeed, the choice of indicator organisms in such studies may influence the results, and the multiple factors in different aquaculture systems would make things more complicated. For instance, a significant increase in the resistance to antibiotics was observed for the indicator microorganisms Acinetobacter spp. but not

Enterococcus spp. in water-sediment samples from an integrated broiler chicken-fish farm (111). In addition, Aeromonas spp. have been successfully employed as indicator

organisms in multiple studies showing the AR in various aquatic environments (114, 115) but this was not always the case (110).

1.6.3 The impact of AR in aquaculture on human health risk

Depending on the effectiveness of food processing on microbial inactivation, aquaculture products may also represent an avenue for disseminating AR to the general public through seafood consumption. Certain AR determinants have been found in both aquaculture and clinical isolates. For instance, Tetr genes (including tetB, tetC, tetD, tetG and tetY), which were detected in isolates from fish farms in Japan, showed high identity

(92% ~100%) with the corresponding genes reported in clinical isolates (116). One gene cassette containing blaCMY-2, sugE and blc detected in Aeromonas salmonicida subsp. salmonicida isolates from Atlantic Canadian salmon farms was identical to a transponson-like element which was widely distributed among clinical and food-borne

Salmonella and other Enterobacteriaceae throughout Asia and the United States (117).

Dissemination of tetA involving Tn1721 and Tn1721-like elements was reported to occur between different Aeromonas species and E. coli and between the human and aquaculture environments in distinct geographical locations (including Norway, Scotland, England, and Germany) (118). HGT mechanism likely plays an important role in the dissemination of AR determinants in microbiota, since such genetic flow has at least been illustrated in multiple experimental settings (64, 67, 76, 90, 92, 119, 120).

Aquaculture creatures such as fish are unique in that they are constantly exposed to microorganisms associated with feed and the production environment (water, soil, etc.)

through the oral route affecting their gut microbiota, while their fecal bacteria further impact the microbiome in the water system that the fish live in. Based on data from human and other animals, it is anticipated that the fish GI tract may selectively enrich certain groups of bacteria in the gut, but it is unclear whether the circulated flow of microorganisms may lead to a homogenized AR populations in the aquaculture production environment, or to what extent other host and environmental factors may contribute to a diversified microbial populations in the ecosystem.

1.7 Rationale and objectives of this study

As mentioned, AR is a complicated issue demanding a comprehensive understanding of AR ecology for targeted mitigation, and the prevalence of ART bacteria in food products is a major food safety and public health concern. Our recent study found a large number of ART bacteria in seafood products from South China. A question of significance to mitigation yet to be addressed is the potential impact of aquaculture production practice on the AR prevalence in aquaculture products and the environment.

Therefore this study targeted at a fish farm in Ohio, with controlled production practices (no antibiotic application). The objectives of the study were: 1) to examine the

AR magnitude in fish (intestine and surface rinsing water), feed and farm environmental samples (pond water and mud); 2) to investigate the comprehensive profiles of ART populations associated with different types of samples from the aquaculture farm; 3) to determine the pheno- and geno- characteristics of the ART isolates from aquaculture

production without antibiotic application; and 4) to identify undescribed AR genes in those ART isolates. The results would contribute to an improved understanding of AR ecology in the aquaculture production system.

Chapter 2 : Profiles of antibiotic resistant bacteria associated with domestic

aquaculture production

2.1 Abstract

Antibiotic resistance (AR) is a complicated issue demanding a comprehensive understanding of AR ecology for targeted mitigation. To assess potential AR risk factors in aquaculture production, this study examined AR in aquaculture products and environmental samples in a 1-year period from a fish farm in Ohio where antibiotics were not used in the production. Phenotypic resistant populations against sulfamethoxazole with trimethoprim (Sul/Tri), tetracycline (Tet), erythromycin (Erm) or cefotaxime (Ctx) were screened by conventional plating with the corresponding antibiotics, followed by population assessments with denaturing gradient gel electrophoresis (DGGE) and 16S rDNA next generation sequencing (NGS). Despite the absence of antibiotic application in the farm, our results showed that antibiotic resistant (ART) bacteria were prevalent in all samples examined, including fish intestine, surface rinsing water, feed, pond water and mud samples. By NGS, a total of 569 genera were identified in Tetr and Ctxr bacteria from five types of samples whereas 167 genera were common in Tetr and Ctxr subpopulations in all samples. Plesimonas spp. were among the dominant Tetr 34

subpopulations in pond water, fish surface rinsing water and fish intestine samples, and the genus of Enterococcus spp. was one of the dominant Ctxr genera in fish feed and intestinal samples. Various AR gene pools measured by quantitative PCR (qPCR) were significantly more abundant in fish intestine and feed than farm environment samples.

Our results suggested that AR risk factor(s) other than direct exposure to antibiotics contribute to AR in aquaculture production. Concentration of AR from feed/water or

ART bacteria amplification in the host should be taken into account for data interpretation and when considering targeted AR mitigation.

2.2 Introduction

The rapid emergence of antibiotic resistance (AR) is a major public health threat, resulting in an estimated 23,000 deaths in the United States each year. Current mitigation strategies were primarily focused on limiting the use of antibiotic application in human clinics and food animal production. However, recent data suggested that AR is a more complicated issue than previously thought (2–4). Multiple risk factors, including food, environmental exposure and selective enrichment by the host gastrointestinal (GI) tract , contributed to AR development, amplification, dissemination and persistence (33, 34), demanding a more thorough consideration of mitigation strategies. For instance, the prevalence of AR in food commensal bacteria was recognized as a significant avenue of

AR exposure (2, 29–32). On the other hand, using a mouse model, it was found that short-term antibiotic exposure did not lead to detectable AR genes in the fecal microbiome, if the corresponding AR gene carriers were not present in the mouse GI tract 35

before antibiotic administration (36). The above studies illustrated a complex picture of

AR ecology independent from antibiotic application.

Aquaculture farming is an important food industrial segment worldwide and has expanded at an average rate of 8.6% annually between 1980 and 2012, driven by increased demand for seafood. In 2012, farmed fish represented 42.2% of total world fish production, resulting in an estimated value of $137.7 billion (40). Seafood creatures are also susceptible to infectious diseases, and antibiotics have found applications in aquaculture production. The presence of antibiotic residues has been considered an important food safety risk factor besides foodborne pathogens, particularly in imported seafood products. According to FDA, chloramphenicol, nitrofuran, and fluoroquinolone were the most commonly detected antibiotic residues in imported aquaculture products during the period between 2004 and 2007, especially the aquaculture products from

Vietnam, Bangladesh, China, Indonesia, India, Chile, and Taiwan

(http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformati on/Seafood/ucm150954.htm). Meanwhile, application of antibiotics during aquaculture production has been considered a risk factor contributing to AR in certain pathogens associated with fish production, such as Aeromonas (60–70, 121), Vibrio (70–76) and

Salmonella (70, 77–79). In addition, large AR gene pools associated with a wide spectrum of commensal bacteria have been found in various aquaculture products and the farm production systems (80, 82–84, 87–89, 122).

As mentioned, AR is a complicated issue with multiple risk factors. AR in aquaculture including fish production is even more complex, since many factors such as 36

the aquaculture creatures themselves, the feces released, the production practices (feed, antibiotics application, manure enrichment, etc.), and environmental condition (water and mud) all impact the microbial parameters of the relatively closed production environment and subsequently the overall health and the composition of the gut and fecal microbiota of the animals. While ART bacteria and AR genes have been reported in aquaculture environments by many authors, and several studies examined the total microbial profiles of aquaculture products and environmental samples (123, 124), there is lack of comprehensive profiling of ART bacteria associated with aquaculture production and proper interpretation of the impact and connection of various risk factors. Such information is essential for effective management controls.

To better understand the potential impact of aquaculture production practice on the prevalence of ART bacteria in the aquaculture ecosystem, following the investigation of AR in aquaculture products from China (88), we studied AR in fish samples raised in a

U.S. fish farm with controlled practices and no history of antibiotic applications. Besides fish, feed and farm environmental samples (pond mud, pond water) were also assessed for resistant microbiome parameters associated with the aquaculture production system.

The results would contribute to an improved understanding of AR ecology in the aquaculture production system.

2.3 Materials and methods

2.3.1 Sampling and bacterial enumeration

Bluegill (Lepomis macrochirus) samples were collected from a farm in Ohio with no history of antibiotic application. The fish were reared for 18 months starting from

April 2011. Three fish were harvested from each of the four ponds included in the study, and a total of sixty fish samples were sampled in five different time points between July

2011 and October 2012. Feed samples, pond mud (top 5 cm) and pond water from each pond were also collected. All samples were transported in sterile containers on ice and processed in the laboratory within 24 hours.

Each fish was briefly rinsed with 50 ml sterile 0.1% peptone water, then the whole fish was placed in a sampling bag with equal volume by weight of 0.1% peptone water, and hand massaged for three minutes. Fifteen ml of resulting liquid was collected and those from fish in the same pond were pooled (rinsing water). Five g of intestine from each fish was collected and pooled in similar manner. Ten ml pond water, 10 g pond mud and 10 g feed for each pond were also collected. All samples were enumerated on

Brain Heart Infusion (BHI) agar and MacConkey (supplemented with 1% glucose) agar containing 100 μg/ml cycloheximide (Sigma-Aldrich, St. Louis, MO) and one of the four following antibiotics, 152 μg/ml of sulfamethoxazole (Sigma-Aldrich) with 8 μg/ml of trimethoprim (Sigma-Aldrich), 16 μg/ml of tetracycline (Sigma-Aldrich), 100 μg/ml of erythromycin (Fisher Scientific, Waltham, MA), or 2 μg/ml of cefotaxime (Sigma-

Aldrich). The plates were then incubated at 30 ˚C for 48 h.

2.3.2 DNA preparation

Pond water (1 L) was concentrated by filtering through a 0.22 µm membrane, which was then cut into pieces and the bacteria were suspended in 10 ml of 0.1% peptone water using a vortex mixer. All samples were subjected to total DNA extraction following the procedures described by Yu and Morrison (125). DNA were extracted in duplicates and used as templates for qPCR.

ART bacteria DNA were extracted from total resistant bacteria isolates on plates with the corresponding antibiotics according to the procedures described previously (29).

Extracted DNA of samples from four ponds were pooled and subjected to DGGE and

NGS assessment.

2.3.3 DGGE analysis on ART microbiota from aquaculture samples

DNA samples were diluted to 50 ng/μl and used as templates to amplify the partial 16S rRNA gene with the primers (16S-357F-GC and 16S-518R) (126) by PCR.

The PCR products were subjected to DGGE using the linear denaturing gradient of 30% to 60%. Specific DGGE bands were excised, purified, amplified and sequenced as previously described (33).

2.3.4 ART microbiota analysis by NGS

The 16S rRNA amplicon targeting V3 and V4 region was prepared with sequencing adapters and Illumina index according to the protocol from Illumina (San

Diego, CA). Purified amplicons were sequenced on Illumina MiSeq system at the Iowa

State University DNA Facility. Data were further processed using the 16S Metagenomics

app (v1.0) at Illumina BaseSpace website

(https://developer.basespace.illumina.com/docs/content/documentation/getting- started/overview).

2.3.5 Assessment of AR gene pools.

Copy numbers of 16S rDNA and AR genes were quantified by TaqMan qPCR on a CFX96 system (Bio-Rad). 16S rDNA served as quality control. The probes (Biosearch

Technology Inc., Novato, CA) and primers used in the study were listed in Table 1.

Primer Sequence (5’-3’) Reference

16S realF TCCTACGGGAGGCAGCAGT (127)

16S realR GGACTACCAGGCTATCTAATCCTGTT (127)

16S probe CGTATTACCGCGGCTGCTGGCAC (127)

sul1 realF CACCTTCGACCCGAAG (128)

sul1 realR TTGAAGGTTCGACACCACG (128)

sul1 probe TCGACGAGATTGTGCGGTTCTTCG (128)

sul2 realF GATATTCGCGGTTTTCCAGA (33)

sul2 realR CAAAGAACGCCGCAATGT (33)

sul2 probe ATCATCTGCCAAACTCGTCGTTATGC (33)

tetS realF GTATGTTCATCTTTCTAAG (30)

tetS realR GCAATAACATCTTTTCAAC (30)

Continued

Table 2.1. Primers used in this study 40

Table 2.1. continued tetS probe CCATGTGTCCAGGAGTATCTAC (30) tetL realF CGTCTCATTACCTGATATTGC (32) tetL realR AGGAGTAACCTTTTGATGCC (32) tetL probe AACCACCTGCGAGTACAAACTGG (32) tetM realF GAACATCGTAGACACTCAATTG (33) tetM realR CAAACAGGTTCACCGG (33) tetM probe CGGTGTATTCAAGAATATCGTAGTG (33) ermB realF GAAAGCCRTGCGTCTGACATC (33) ermB realR CGAGACTTGAGTGTGCAAGAGC (33) ermB probe ACCTTGGATATTCACCGAACACTAG (33) blaTEM realF CACTATTCTCAGAATGACTTGGT (129) blaTEM realR TGCATAATTCTCTTACTGTCATG (129) blaTEM probe CCAGTCACAGAAAAGCATCTTACGG (129)

16S-357F-GC CGCCCGCCGCGCGCGGCGGGCGGGGCGGG (126)

GGCACGGGGGGCCTACGGGAGGCAGCAG

16S- 518R ATTACCGCGGCTGCTGG (126)

2.4 Results

2.4.1 The prevalence and abundance of ART bacteria in aquaculture samples

Bacteria resistant to Sul/Tri, Tet, Erm and Ctx were found in all five types of samples including fish surface rinsing water, fish intestine, mud, pond water and fish feed samples during the entire investigation period (July 2011 to October 2012). As illustrated 41

in Fig. 2.1A, Sul/Trir and Tetr bacteria were the most abundant ART microbiota in fish rinsing water and intestine on BHI plates (102-103 CFU/g, corresponds to 0.2% to 3.7% in total population); while the abundance of Ermr (below detection limit in rinsing water and less than 0.01% in total population) and Ctxr (10-102 CFU/g, corresponds to 0.01% to

0.4%) bacteria ranged from low to moderate. In contrast, levels of ART bacteria were found to be similar among different resistance phenotypes in gram-negative bacteria within the same sample (Fig.2.1B). For instance, 10 to 102 CFU/g Sul/Trir, Tetr, Ermr and

Ctxr bacteria were found in in rinsing water and intestine, which counted 0.05% to 1.2% in total gram-negative population, except that Ctxr bacteria were below detection limit in intestine. Pond mud contained 103 to 105 CFU/g Sul/Trir, Tetr and Ctxr bacteria (0.01% to

1.3%) and 102 CFU/g Ermr bacteria (0.004%) on BHI plates (Fig.2.1A). In feed samples,

102 to 104 CFU/g of Sul/Trir, Tetr, Ermr and Ctxr bacteria had a high portion in total bacteria number on BHI plates, accounting for 95.0%, 63.5%, 0.2% and 7.2% (Fig.2.1A).

However, the presence of ART gram-negative bacteria in feed was minimal on MACG plates (10 CFU/g Tetr and Ctxr, 10% in gram-negative population, remaining resistance phenotypes were below detection limit) (Fig.2.1B). Low to moderate levels of ART bacteria were found on both types of medium plates in pond water.

8.0 (A) 7.0

6.0 BHI 5.0 Sul/Tri 4.0 Tet 3.0

(Log10 CFU/g) (Log10 Erm 2.0 Ctx ART bacterial populations bacterial ART 1.0 ND0.0 W I M PW F

8.0 (B)

7.0

6.0 MACG 5.0 Sul/Tri 4.0 Tet (Log10 CFU/g) (Log10 3.0 Erm

ART bacterial populations bacterial ART 2.0 Ctx 1.0

ND0.0 W I M PW F

Figure 2.1. ART bacterial populations in fish and related samples collected in October

2012. Bacterial population from BHI plates (A) and MACG plates (B). W: surface rinsing water; I: fish intestine; M: mud; PW: pond water; F: feed; ND: detection limit; CFU: colony forming unit. The error bars represent standard deviations among 4 fish ponds.

2.4.2 16S NGS analysis on culturable ART bacteria in fish and related samples

Fig. 2.2A and 2.2B show that four bacteria phyla, including Proteobacteria,

Firmicutes, Actinobacteria and Bacteroidetes, predominated (99%) in ART microbiota from all types of samples, though their distribution varied with respect to resistance phenotype and type of samples. As shown in Fig.2.2A, the most abundant Tetr bacteria from surface rinsing water (98.0%), intestine (97.5%) and pond water (64.8%) samples were Proteobacteria, while Tetr Firmicutes were found to dominate in pond mud (89.7%) and feed (74.6%) samples. In Ctxr subpopulations (Fig. 2.2B), Proteobacteria were dominant in surface rinsing water (71.6%) and pond water (81.6%), and the most prevalent populations in intestine (87.0%) and feed (51.3%) samples were Actinobacteria while Firmicutes were dominant in mud (61.8%) samples. Therefore at the bacterial phylum level, the Tetr subpopulations recovered on BHI plates of fish intestine and fish surface rinsing water were mostly correlated to those of pond water, and the Ctxr subpopulations in fish surface rinsing water were mostly correlated to those of pond water, and the Ctxr subpopulations of fish intestine samples were mostly correlated to those of fish feed and mud. Moreover, there existed a dramatic difference between dominant Tetr and Ctxr bacteria subpopulations in intestine.

Fig. 2.3 listes the composition of ART bacteria from five types of samples at the family level. Enterobacteriaceae counted majority of the Tetr bacteria from fish samples

(87.6% in intestine, 68.0% in rinsing water) while more Planococcaceae and

Moraxellaceae were found in the Tetr subpopulations from mud (66.9%, 1.1%), pond water (31.1%, 42.4%) and feed (56.8%, 9.2%). Similar as the abundance of Tetr

subpopulations in intestine, Ctxr subpopulations from intestine were also composed of one highly dominant bacterial family, which was Cellulomonadaceae (84.4%), and other less abundant populations such as Enterococcaceae (11.5%). Pseudomonadaceae, and

Sphingobacteriaceae were more prevalent in Ctxr subpopulations from rinsing water.

Some Ctxr bacteria families were more prevalent in both pond water and mud, including

Bacillaceae, Moraxellaceae, Caulobacteraceae and Flavobacteriaceae. The dominant Ctxr subpopulations from feed samples composed of Brevibacteriaceae (48.8%),

Enterococcaceae (21.2%), Planococcaceae (15.7%) and Carnobacteriaceae (5.6%).

Therefore at the family level, among the Tetr subpopulations on BHI plates,

Enterobacteriaceae were common in fish surface rinsing water, intestinal and pond water samples, Planococcaceae of pond water and those of fish feed and pond mud were common. In addition, among the Ctxr populations Pseudomonadaceae subpopulations were common in pond water and fish raising water, and Enterococcaceae were common in fish feed and fish intestine samples.

Among 569 identified genera in ART bacteria, 14 genera had over 100,000 reads including Plesiomonas spp., Sporosarcina spp., Pseudomonas spp., Acinetobacter spp.,

Brevibacterium spp., Oerskovia spp., Bacillus spp., Enterococcus spp., Arthrobacter spp.,

Paenisporosarcina spp., Lysinibacillus spp., Citrobacter spp., Virgibacillus spp. and

Shewanella spp., As shown in Fig. 2.4, Tetr subpopulations of Plesimonas spp. (73.7% in intestine, 44.9% in rinsing water), Pseudomonas spp. (1.3% in intestine, 19.3% in rinsing water), were commonly found in fish samples but the spectrum of Ctxr bacteria in intestine was quite different whereas Oerskoia spp. (81.3%) and Enterococcus spp.

(10.9%) were predominant. In spite of different distributions of Tetr and Ctxr bacteria from environmental samples, some dominant genera were found among both Tetr and

Ctxr subpopulations, such as Pseudomonas spp., Sporosarcina spp. and Acinetobacter spp.

It is worth noting that the large portion of dominant ART bacteria (Tetr and Ctxr) from feed samples were gram-positive bacteria. For example, the dominant Tetr subpopulations from feed included Lysinibacillus spp., Paenisporosarcina spp., Macrococcus spp.,

Acinetobacter spp., Brevibacterium spp. and Corynebacterium spp.; while

Brevibacterium spp., Enterococcus spp., Rummeliibacillus spp. and Carnobacterium spp. were more prevalent in Ctxr subpopulations from feed. At the genus level, Plesimonas spp. was among the dominant Tetr subpopulations in pond water, fish surface rinsing water and fish intestine samples, and Enterococcus spp. was one of the dominant Ctxr genera in fish feed and intestinal samples.

100% (A)

80% Others 60% Bacteroidetes Actinobacteria 40% Firmicutes Proteobacteria

Relativeabundance 20%

0% W I M PW F

100% (B)

80% Others 60% Bacteroidetes Actinobacteria 40% Firmicutes Proteobacteria Relativeabundance 20%

0% W I M PW F

Figure 2.2 Distribution of bacterial phylum of Tetr (A) and Ctxr (B) populations on BHI plates associated with different types of samples. W: surface rinsing water; I: intestine; M: mud; PW: pond water; F: fish feed.

100% (A) Others Enterococcaceae 80%

Streptococcaceae Microbacteriaceae Bacillaceae 60% Corynebacteriaceae Brevibacteriaceae 40% Moraxellaceae Staphylococcaceae

Relative abudance Relative Planococcaceae 20% Pseudomonadaceae Comamonadaceae 0% Shewanellaceae W I M PW F Enterobacteriaceae

(B) 100% Others Microbacteriaceae 80% Sphingobacteriaceae

Xanthomonadaceae

Pseudomonadaceae 60% Flavobacteriaceae Caulobacteraceae Moraxellaceae 40% Micrococcaceae

Carnobacteriaceae Relative abundance Relative Brevibacteriaceae 20% Bacillaceae Planococcaceae Enterococcaceae 0% W I M PW F Cellulomonadaceae

Figure 2.3 Distribution of bacterial family of Tetr (A) and Ctxr (B) populations on BHI plates associated with different types of samples. W: surface rinsing water; I: intestine; M: mud; PW: pond water; F: fish feed.

(A) 100% Others Corynebacterium 80% Brevibacterium Leucobacter Shewanella 60% Macrococcus Paenisporosarcina 40% Lysinibacillus Acinetobacter Virgibacillus

Relative abundance Relative 20% Sporosarcina Citrobacter 0% Pseudomonas W I M PW F Plesiomonas

(B) 100% Others Carnobacterium Rummeliibacillus

80% Brevibacterium Flavobacterium Brevundimonas 60% Acinetobacter Arthrobacter Bacillus 40% Sporosarcina Oerskovia Enterococcus

Relative abundance Relative 20% Microbacterium Stenotrophomonas Pedobacter 0% Pseudomonas W I M PW F

Figure 2.4 Distribution of bacterial genus of Tetr (A) and Ctxr (B) populations on BHI plates associated with different types of samples. W: surface rinsing water; I: intestine; M: mud; PW: pond water; F: fish feed.

2.4.3 AR gene pools in fish and environmental samples from fish farm

As illustrated in Fig 2.5, feed contained all examined AR genes and the largest

AR gene pools (106 to 109 copies/g). Another significant source of AR genes was fish intestine which also contained all tested AR genes (up to 104 copies/g). The least abundant AR gene pools were detected in environmental samples especially in pond water containing AR gene pools from below detection limit to 10 copies/ml. Among 7 representative AR genes, tetM was the most abundant (from 10 copies/mL in pond water to 109 copies/g in feed), followed by tetS.

10.0 (A)

8.0

Copies/g)

10 6.0 I

4.0 M PW 2.0

F AR gene pools (Log pools gene AR ND0.0 sul1 sul2 tetL tetM tetS ermB blaTEM

1.0E+00 (B)

1.0E-02 W 1.0E-04 I M 1.0E-06 PW F 1.0E-08

1.0E-10 Ratio of AR gene to 16S rDNA 16S to gene AR ofRatio sul1 sul2 tetL tetM tetS ermB

Figure 2.5 Quantitative PCR of 16S rDNA gene and AR gene pools in aquaculture samples collected in October 2012. ND: detection limit. A: Gene pool of surface rinsing water

(W), fish intestine (I), mud (M) and pond water (PW) and feed (F) were assessed by qPCR; B: the calculated ratio of AR genes to 16S rDNA gene. Average values and standard deviation of quadruplicate trial were calculated based on the results from four ponds.

2.5 Discussion

The association between antibiotic application and presence of AR in seafood is well documented. For example, various studies suggested a causal relationship resulting from selective pressure (61, 81, 97–99). Meanwhile, abundant ART bacteria and AR genes were detected in aquaculture products from developing countries, where restriction of antibiotic usage is less stringent (65, 88). Our study examined fish samples from an antibiotic-free farm, yet ART bacteria and AR genes were still abundant in fish, feed and related environmental samples. Our results are in agreement with previous studies, which discovered significant ART bacteria in human (33, 130), swine (131, 132) and rodents

(133) with no history of direct antibiotic exposure. These data clearly illustrated the contribution of additional risk factors besides antibiotic applications to the AR issue in the global ecosystem.

Here we presented the population count results of samples collected in October

2012, when fish were harvested. Additional fish, farm environment and feed samples in

July 2011, September 2011, November 2011 and March 2012 were also collected and analyzed (Fig. 2.6). In agreement with the result of October 2012 (Fig. 2.1), ART bacteria were detected in all samples collected in different seasons of the year. However, intestine samples from different seasons did present variety in enumeration results of total and

ART bacteria, which was higher in the summer season (July 2011) but lower in the winter season (Nov. 2011). For instance, 106 CFU/g Tetr bacteria in summer were detected while only 10 CFU/g was found in winter. Tetr populations from other four types of samples also exhibited a similar trend, but the differences in the enumeration numbers

were smaller. Compared with other time points, the enumeration results of samples from

October 2012 were moderate, and the microbial profiling data from this moderate season would be a good representation of bacteria persistent in aquaculture samples in general.

Even in an aquaculture farm with controlled practices, production as well as microbial parameters would still be affected by many uncontrollable factors, as in any outdoor, semi-natural systems. Therefore data interpretation for potential AR risk factors in such systems is still challenging. The general survey of microbial enumeration of total and ART bacteria as well as qPCR results showed that fish feed is an important source of

ART bacteria and AR gene pools, with as high as 109 copies/g tetM, and all AR genes examined found in this study. Farm environmental samples, on the other hand, presented moderate levels of AR, though no history of antibiotic application was reported.

Nevertheless, it is worth noting that fish filter through a large volume of water daily, which also represents a significant avenue for AR intake. But these data were still insufficient to illustrate the AR causative agents in the aquaculture environment.

Microbiota profiling of the ART populations were further conducted in this study.

The composition of ART bacteria in all samples were examined by NGS assessment. A total of 383, 421, 329, 364 and 354 Tetr genera from rinsing water, intestine, mud, pond water and feed samples, respectively, were identified. Correspondingly, 339, 278, 407,

408 and 323 Ctxr genera were detected. Among them, 283, 246, 292, 327 and 278 genera were found in both Tetr and Ctxr subpopulations from rinsing water, intestine, mud, pond water and feed samples, respectively. Moreover, 167 genera having Tetr and Ctxr subpopulations were detected in all five types of samples, such as Sporosarcina spp., 53

Pseudomonas spp., Acinetobacter spp., Brevibacterium spp., Oerskovia spp., Bacillus spp., Enterococcus spp., Arthrobacter spp., Lysinibacillus spp. and Virgibacillus spp.

We have also performed DGGE analysis of related samples (Fig. 2.7). Identified

ART bacteria genera by DGGE, which were in agreement with the identified dominant populations by NGS, included Tetr bacteria of Plesiomonas spp. (in intestine, rinsing water), Pseudomonas spp. (in rinsing water), Enterobacteriaceae (in rinsing water),

Sporosarcina spp. (in pond water, mud), Bacillus spp. (in mud), Macrococcus spp. (in feed), Acinetobacter spp. (in feed); Ctxr bacteria of Enterococcus spp. (in intestine, feed),

Oerskovia spp. (in intestine), Brevibacterium spp. (in feed), Pseudomonas spp. (in rinsing water, pond water), Pedobacter spp. (in rinsing water), Flavobacterium spp. (in pond water), Bacillus spp. (in pond water), Brevundimonas spp. (in pond water), Sporosarcina spp. (in mud), Arthrobacter spp. (in mud); Sulr bacteria of Enterococcus spp. (in intestine, rinsing water), Providencia spp. (in intestine), Enterobacter spp. (in intestine),

Pseudomonas spp. (in rinsing water, pond water), Brevundimonas spp. (in pond water),

Sporosarcina spp. (in mud), Carnobacterium spp. (in mud), Bacilus spp. (in mud, feed),

Weissella spp. (in feed), Macrococcus spp. (in feed); Ermr bacteria of Enterobacteriae (in intestine), Pseudomonas spp. (in pond water, mud), Acinetobacter spp. (in pond water, mud), Citrobacter spp. (in mud), Enterobacter spp. (in mud), Enterococcus spp. (in feed),

Staphylococcus spp. (in feed), Corynebacterium spp. (in feed). Some genera were present among multiple samples. For example, Pseudomonas sp. were found in rinsing water, pond water and mud samples, and showed the resistance to all four types of antibiotics.

As illustrated in Fig 2.3 and 2.4, at the family and genus levels, at least part of the dominant ART bacteria in fish intestine were correlated to the same types of organisms in fish feed and/or pond water samples. Common ART subpopulations were also found in both fish surface rinsing water and pond water samples. Thus fish feed could be an important source of ART bacteria and AR genes for aquaculture ecosystem. Currently, an important ingredient of commercial fish feed is animal byproducts, which are ART bacteria- and AR genes- rich, serving as the protein source. Fish meal is another important ingredient of fish feed. As demonstrated in this study, the skin and intestine of fish were found to contain large number of ART bacteria and AR genes. Therefore it is not difficult to explain the high density of ART bacteria and AR genes in the feed. In addition, pond water also impacts AR in both fish intestine and fish surface.

Meanwhile, the composition of ART bacteria in the fish intestine was still rather distinct from that in farm environment and feed samples. For instance,

Cellulomonadaceae and Oerskoia spp. were the most dominant Ctxr subpopulations in fish intestine but not in any other samples. Likewise many dominant ART bacteria in fish feed at family and genus levels did not overlap with those in fish intestine. These data indicate that fish as hosts selectively enriched certain ART bacterial populations in the gut microbiota. While results from microbial profiling studies indicated potential AR risk factors in aquaculture production, further genetic characterization of ART isolates will provide additional evidence regarding the potential correlation and dissemination of AR among segments in the aquaculture production system.

Besides the potential risk factors assessed in this study, other factors may also contribute to AR in the aquaculture ecosystem. As mentioned previously, antibiotic resistance was developed in infant GI tract within days after birth (33). So is it possible that some of those ART bacteria may have already colonized in fish intestine when they were fingerlings? Also, since pond water has an impact on ART bacteria associated with fish intestine and fish surface, manure (often rich in ART bacteria) treatment as a popular practice in aquaculture may also be a potential AR risk factor. Further studies for a comprehensive understanding of risk factors and targeted mitigation are essential to improve seafood safety.

Figure 2.6 ART bacterial population from BHI plates (A) and MACG plates (B) collected

at the other four different time points. W: surface rinsing water; I: fish intestine; M: mud;

PW: pond water; F: feed; ND: detection limit; CFU: colony forming unit. A1 and B1: results of samples collected at July 2011; A2 and B2: results of samples collected at

September 2011; A3 and B3: results of samples collected at November 2011; A4 and B4: results of samples collected at March 2012; The error bars represent standard deviations among 4 fish ponds.

Figure 2.7 DGGE assessment results of cultured ART bacteria from aquaculture samples. W: rinsing water; pW: pond water; M: mud; I: intestine; F: fish feed; M1: 1 kb ladder; M2: 100 bp ladder; (A) Sul/Trir bacteria. 1-3, 6-11: Pseudomonas spp.; 4-5, 17-18: Enterococcus spp.; 12:

Brevundimonas sp.; 13, 15: Sporosarcina spp.; 14: Carnobacterium sp.; 16, 24: Bacilus spp.; 19, 59

21-22: Providencia spp.; 20: Enterobacter sp.; 23: Weissella sp.; 25: Macrococcus sp.; (B) Tetr bacteria. 1-7: Pseudomonas spp.; 8-9, 11, 25-27: Plesiomonas spp.; 10: Enterobacteriaceae; 12:

Elizabethkingia sp.; 13: Lactococcus sp.; 14-15, 19-21, 23: Sporosarcina spp.; 16-18, 22, 24, 30:

Bacillus spp.; 28, 31: Macrococcus spp.; 29: Weissella sp.; 32-33: Acinetobacter spp.; (C) Ermr bacteria. 1-4, 6-13, 17: Pseudomonas spp.; 5, 14: Acinetobacter spp.; 15: Citrobacter sp.; 16:

Enterobacter sp.; 18-22: Enterobacteriae; 23-24: Enterococcus spp.; 25-26: Staphylococcus spp.;

27-28: Corynebacterium spp.; (D) Ctxr bacteria. 1-2, 6-10: Pseudomonas spp.; 3-4: Pedobacter sp.; 5: Sphingobacterium sp.; 11-12: Flavobacterium spp.; 13, 27: Bacillus spp.; 14:

Brevundimonas sp.; 15-17: Sporosarcina spp.; 18: Arthrobacter sp.; 19-20, 23-24, 26:

Enterococcus spp.; 21-22: Oerskovia spp.; 25: Lactobacillus sp.; 28: Brevibacterium sp

Chapter 3 : Characterization of antibiotic resistant bacteria isolates from

aquaculture production

3.1 Abstract

To understand the correlation among environmental, foodborne and host AR ecology, a total of 4747 Sul/Trir, Tetr, Ermr and Ctxr isolates, originated from fish intestine, fish surface rinsing water, feed, mud and pond water from an aquaculture farm with no antibiotic application history were examined in this study. About 79% of 4747

ART isolates showed resistance to more than one antibiotic. MIC of some ART isolates no less than 512 μg/ml of sulfamethoxazole (Sul) with 8 μg/ml of trimethoprim (Tri), 512

μg/ml of tetracycline (Tet), 512 μg/ml of erythromycin (Erm), or 512 μg/ml cefotaxime

(Ctx), respectively. ART isolates were found to carry various AR genes including sul1, sul2, tetS, tetL, tetM, ermB or ermC. Identified AR gene carriers belong to 18 genera. In addition, the AR traits in many isolates were quite stable, even in the absence of selective pressure. The results suggested AR risk factor(s) independent from direct antibiotic exposure in aquaculture production.

3.2 Introduction

The rapid emergence of antibiotic resistance (AR) has become a major public health concern. Since 2005, results from several lines of studies provided solid evidences suggesting that multiple risk factors contributed to AR development, enrichment, dissemination and persistence. The resistance issue is more severe and complicated than previously thought, and the existing AR surveillance and control strategy might have not been perfectly targeted (2–4). The prevalence and abundance of antibiotic resistant (ART) bacteria in retail and restaurant foods including many ready-to-eat items, with transmissible AR genes in a broad-spectrum of commensal and even beneficial bacteria, indicated that the food chain serves as a critically important avenue transmitting AR to the general public (2, 29–32). ART bacteria are further enriched in host gastrointestinal system, even without the direct exposure to antibiotics (3, 33, 34). The AR-rich feces from both human and animals likely served as the most impactful source for the environmental AR gene pool (35). More importantly, the targeted AR gene pool was not observed in feces in the mouse model receiving antibiotic treatment, if the host did not receive oral inoculation of AR gene-containing bacteria before the treatment (36). These data clearly illustrated that oral exposure to ART bacteria is a critical risk factor for the rapid dissemination of AR in the global ecosystem, independent from antibiotic treatment, and the food chain serves as an important avenue bridging the AR flow between the environment and hosts. Therefore AR mitigation in the food chain is essential to effectively combat this significant food safety and public health challenge (2, 3).

Seafood creatures are also susceptible to infectious diseases, and antibiotics have found applications in aquaculture production, which is considered a risk factor contributing to AR in aquaculture ecosystem. Certain ART pathogens were reported to be associated with aquaculture system , such as Aeromonas (60–70, 121), Vibrio (70–76) and Salmonella (70, 77–79). In addition, significant AR gene pools were found in a wide spectrum of commensal bacteria associated with the aquaculture system (80–88, 90, 134).

Depending on the effectiveness of food processing on microbial inactivation, aquaculture products may also represent an avenue disseminating AR to the general public through seafood consumption. Certain AR determinants have been found in both aquaculture and clinical isolates. For instance, Tetr genes (including tetB, tetC, tetD, tetG and tetY), which were detected in isolates from fish farms in Japan, showed high identity

(92% ~100%) with the corresponding genes reported in the clinical isolates (116). One gene cassette containing blaCMY-2, sugE and blc detected in Aeromonas salmonicida subsp. salmonicida isolates from Atlantic Canadian salmon farm was identical to a transponson-like element which was widely distributed among clinical and food-borne

Salmonella and other Enterobacteriaceae throughout Asia and the United States (117).

Rhodes et al. (118) reported the dissemination of tetA involving Tn1721 and Tn1721-like elements between different Aeromonas species and E. coli and between the human and aquaculture environments in distinct geographical locations (including Norway, Scotland,

England, and Germany). HGT mechanism likely plays an important role in the dissemination of AR determinants in microbiota, since such genetic flow has at least been illustrated in multiple experimental settings (119, 120).

As mentioned, AR is a complicated issue with multiple risk factors. Increasing evidences have illustrated that ART bacteria are prevalent in food, hosts and environmental samples without direct exposure to antibiotics. Characterization of ART isolates from such samples has already led to the discovery of various mechanisms involved in AR evolution (135, 136), enrichment (33) and persistence (32, 137, 138), contributed to an improved understanding of AR ecology.

The objective of this study is to characterize the ART bacteria isolates from aquaculture production without antibiotic application. Their phenotypic- and genetic- resistance profiles, resistance persistence and functionality in other hosts were investigated. The results can provide further information on potential factors contribute to

AR ecology, and thus benefit the development of strategies for targeted controls.

3.3 Materials and methods

3.3.1 Strain cultivation

A total of 4747 Sul/Trir, Tetr, Ermr and Ctxr isolates, originated from fish intestine

(1045 isolates), fish surface rinsing water (850 isolates), feed (150 isolates), mud (1273 isolates) and pond water (1429 isolates) from an aquaculture farm with no antibiotic application history were examined in this study. All isolates were cultured at Brain Heart

Infusion (BHI) media containing the corresponding antibiotics, including 152 μg/ml of sulfamethoxazole (Sigma-Aldrich, St. Louis, MO) with 8 μg/ml of trimethoprim (Sigma-

Aldrich), 16 μg/ml of tetracycline (Sigma-Aldrich), 100 μg/ml of erythromycin (Fisher

Scientific, Waltham, MA), or 2 μg/ml of cefotaxime (Sigma-Aldrich). Frozen stocks of all strains were stored in the media containing the corresponding antibiotics, supplemented with 20% glycerol and kept at -80°C.

3.3.2 Determination of the phenotypic resistant profiles of the ART isolates

Recovered Sul/Trir, Tetr, Ermr and Ctxr isolates were spotted onto all 4 types of

BHI agar plates containing the corresponding antibiotics as described above.

One third of AR gene carriers were collected and subjected to the minimum inhibition concentration (MIC) test in BHI broth containing each of four corresponding antibiotics (up to 512 μg/ml of Sul with 8 μg/ml of Tri, 512 μg/ml of Tet, 512 μg/ml of

Erm, and 512 μg/ml of Ctx) as previously described (2).

3.3.3 Detection of AR genes and identification of AR gene carriers

One fourth of ART isolates were subject to conventional PCR screening for representative AR genes following the published procedures (29). Table 3.1 listed the

PCR primers used in the study. Approximately 10% of the positive PCR products were confirmed by DNA sequence assessment, and 50% of the AR gene carriers were further identified by partial 16S rRNA gene sequence analysis, as described previously (2).

Primer Sequence (5’-3’) Reference sul1 F CGGCGTGGGCTACCTGAACG (139) sul1 R GCCGATCGCGTGAAGTTCCG (139) sul2 F GCAGGCGCGTAAGCTGA (33) sul2 R GGCTCGTGTGTGCGGATG (33) tetS F GAACGCCAGAGAGGTATT (29) tetS R TACCTCCATTTGGACCTCAC (29) tetL F TTGGATCGATAGTAGCC (29) tetL R GTAACCAGCCAACTAATGAC (29) tetM F CGAACAAGAGGAAAGCATAAG (29) tetM R CAATACAATAGGAGCAAGC (29) ermB F TGGTATTCCAAATGCGTAATG (140) ermB R CTGTGGTATGGCGGGTAAGT (140) ermC F GCTAATATTGTTTAAATCGTCAAT (141) ermC R TCAAAACATAATATAGATAAA (141) blaTEM F CATTTCCGTGTCGCCCTTATTC (142) blaTEM R CGTTCATCCATAGTTGCCTGAC (142) blaSHV F AGCCGCTTGAGCAAATTAAAC (142) blaSHV R ATCCCGCAGATAAATCACCAC (142) blaOXA F GGCACCAGATTCAACTTTCAAG (142) blaOXA R GACCCCAAGTTTCCTGTAAGTG (142)

Continued

Table 3.1. Primers used for conventional PCR.

Table 3.1. continued

blaCMY-2 F GACAGCCTCTTTCTCCACA (143) blaCMY-2 R TGGAACGAAGGCTACGTA (143) blaCTX MU1 ATGTGCAGYACCAGTAARGT (144) blaCTX MU2 TGGGTRAARTARGTSACCAGA (144) intl1 F ACGAGCGCAAGGTTTCGGT (145) intl1 R GAAAGGTCTGGTCATACATG (145)

16S F AGAGTTTGATCCTGGCTCAG (146)

16S R TACCTTGTTACGACTT (146)

3.3.5 Persistence of AR within resistant isolates

AR stability was determined according to published procedures (32, 88) with slight modifications. Overnight culture of each ART isolate was inoculated (1:100) into fresh BHI medium without antibiotics and incubated at 30°C for 12 h. After 30 days of consecutive inoculation (twice daily), the cultures were serially diluted and plated on BHI agar plates. Up to 100 colonies were randomly picked from each sample, and spotted on

BHI agar plates with and without the corresponding antibiotic. The ratio of resistant to total colonies was used to describe the resistance persistence in ART isolates under no antibiotic selective pressure.

3.3.6 Transformation.

Chemical transformation was conducted using plasmids from representative MDR isolates (9 gram-negative isolates) and E. coli DH5α as a recipient by the calcium chloride transformation method (147). For 5 gram-positive isolates, electroporation was employed and Lactococcus lactis 0230 was used as recipient following the procedures described previously (148).

3.4 Results

3.4.1 Phenotypic resistant profiles of the ART isolates

Of the recovered Sul/Trir, Tetr, Ermr and Ctxr isolates associated with fish intestine, surface rinsing water, feed, mud and pond water samples, 3772 of 4747 (about

79%) total isolates showed resistance to more than one antibiotics and multiple drug resistant (MDR) bacteria were common in all samples. As shown in Table 3.2, among

ART bacteria from fish intestine, 838 of 1045 (80%) were found resistant to at least two drugs, including 58.5% resistant to two, 12.1% to three, and 9.7% to four antibiotics.

Despite smaller number of ART bacteria were recovered from fish feed, which was 150 in total, 96 (64%) were resistant to additional antibiotics, with 31.3%, 22.0% and 10.7% resistant to two, three, and four of the antibiotics tested in this study, respectively.

Moreover, fish intestine and mud samples had higher portions of MDR isolates against

2AR (58.5%) and 3AR (50.65%), respectively.

Table 3.2 The percentage of MDR isolates from aquaculture samples

1ARa 2AR 3AR 4AR

18.7% 35.6% 32.2% 13.4% Rinsing water (159/850) (303/850) (274/850) (114/850) 19.8% 58.5% 12.1% 9.7% Fish intestine (207/1045) (611/1045) (126/1045) (101/1045) 33.8% 13.7% 34.6% 18.0% Mud (430/1273) (174/1273) (440/1273) (229/1273) 8.7% 15.2% 50.6% 25.5% Pond water (125/1429) (217/1429) (723/1429) (364/1429) 36.0% 31.3% 22.0% 10.7% Fish feed (54/150) (47/150) (33/150) (16/150) a1AR: isolates were resistant to one type antibiotic Sulr Tetr Ermr Ctxr

High MIC value (no less than 512 μg/ml) was detected against each antibiotic,

including 38 out of 40 (95%) Sul/Trir isolates, 12 out of 32 (37.5%) Terr isolates, 9 of 9

(100%) Ermr isolates, and 8 out of 15 (53.3%) Ctxr isolates. Especially, some isolates not

only exerted phenotypic resistance against multiple antibiotics but also owned extremely

high MIC value. For instance, one Bacillus isolate from feed sample was found to have

high MIC (more than 512 μg/ml) against all four antibiotics.

3.4.2 Prevalence of the AR genes and identification of the isolates

As shown in Table 3.3, feed, fish intestine and surface rinsing water samples were

found to contain high level of various AR genes while lower level of AR genes were

detected in mud and pond water samples. Within 13 AR determinants examined, tetM

had highest detection rate (8.1%) in all isolates tested which was consistent with AR gene

pools assessment result in Phase I of the study. High percentages and a high variety of

AR genes were found in fish feed samples, including sul1 (10.2%), sul2 (2.0%), tetL(13.3%), tetM (9.2%), tetS (6.1%) and ermB (1.0%). Among all isolates tested, tetM was the most abundant AR gene (8.1%).

As illustrated in Table 3.3, identified AR gene carriers belonged to 18 genera.

Aeromonas spp., Enterococcus spp., Enterobacter spp. and Plesiomonas spp. found in fish intestine samples were common microorganisms isolated from fish (149, 150). Some of isolates examined were found to carry multiple resistance encoding genes, such as

Plesiomonas spp., Enterococcus spp., Lactococcus spp., Carnobacterium spp.,

Psychrobacter spp. Despite of lower populations of ART bacteria, a high variety of these

AR gene carriers were found in fish feed samples. It is worth noting that some genera were found in different types of samples. For example, Plesiomonas spp. were detected in rinsing water, fish intestine and mud samples while Aeromonas spp. were present in fish intestine and mud samples. This result suggested the circulation of microorganisms within the ecosystem.

Detected AR genes Sample Gene carrier (x no. identified isolates, AR gene) (no. positive/ no. detected isolates)

sul1 (8/209), sul2 (1/209), tetM (30/209), Rinsing water Plesiomonas spp. [(x3, sul1),(x1, sul2), (x3, tetM), (x2, sul1+tetM)] sul1+tetM (5/209)

sul1 (7/357), sul2 (15/357), tetL (1/357), Aeromonas sp. (x1, sul1); Enterobacter spp. (x7, sul2);

Fish intestine tetM (9/357), sul1+ tetM (1/357), tetL+ Enterococcus spp. [(x1, tetL), (x1, tetL+ tetS)];

tetS(1/357), Plesiomonas spp. [(x3, sul1), (x6, sul2), (x1, sul1+ tetM)];

Aeromonas sp.(x1, sul1); Comamonas sp. (x1, int1);

Mud sul1 (3/269), sul2 (1/269), int1(1/269) Plesiomonas sp. (x1, sul1)

Exiguobacterium spp. (x2, sul1); Staphylococcus sp. (x1, ermC); sul1 (5/244), tetS (1/244), tetL+ tetS (3/244), Pond water Lactococcus spp. [(x1, tetS), (x2, tetL+tetS)]; ermC(1/244)

Continued

Table 3.3 The identity of representative AR gene carriers

Table 3.3 continued

Aerococcus spp. (x2, tetM); Bacillus spp.[(x1,tetS),(x1, ermB)];

Carnobacterium sp. (x1, tetL+tetS); Kocuria spp. (x4, sul1);

sul1 (9/98), sul2 (1/98), sul1+ sul2 (1/98), Corynebacterium spp. [(x1,sul1), (x2, tetL)]; Kurthia sp. (x1, tetM);

tetL (4/98), tetM (2/98), tetS (3/98), tetL+ Enterococcus spp. [(x1, sul1), (x2, tetS), (x5, tetL+ tetM), (x1, tetL+ Fish feed tetM(6/98), tetL+ tetS(2/98), tetS), (x1, tetL+ tetM+ tetS)]; Pseudoclavibacter sp. (x1, sul1);

tetL+tetM+tetS(1/98), ermB(1/98) Vagococcus sp. (x1, tetL); Lactobacillus sp. (x1, tetL);

72 Psychrobacter spp. [(x1, sul2), (x1, sul1+sul2)];

Staphylococcus sp. (x1, sul1);

3.4.3 Stability and functionality of the antibiotic resistance

As illustrated in Table 3.4, 90% to 100% of the progenies from 9 ART isolates retained their original AR traits. It is interesting that one Enterococcus strain kept its

Sur/Trir but almost lost its Tetr and Ctxr traits.

Table 3.4. Resistance persistence of ART isolates

Resistance Retention Rate (%) Isolate Sul/Trir Tetr Ermr Ctxr

AfS1 Enterococcus sp. 100 100 - -

BpwE5 Staphylococcus sp. 100 - 93 100

BfE1 Bacillus sp. 100 100 98 100

AiT14 Plesiomonas sp. 100 100 - -

BfT3 Enterococcus sp. 100 100 - 100

BfT16 Enterococcus sp. - 100 - 100

AfT7 Staphylococcus sp. - 99 - -

AfT4 Enterococcus sp. 100 100 - -

BiX15 Enterococcus sp. 100 1 - 2

BiX21 Enterococcus sp. 100 100 - 100

As shown in Table 3.5, plasmids from two gram-negative isolates (Aeromonas sp. bearing sul1 from fish intestine and Psychrobacter sp. bearing sul2 from fish feed) and plasmid from one gram-positive isolate (Vagococcus sp. bearing tetL from fish feed) 73

were successfully transferred to the corresponding recipients with acquired resistance.

The transformants had comparable MIC of the corresponding antibiotic with the donors, indicating the resistance genes were transmissible and functional in other bacteria if acquired via HGT events.

Table 3.5. The donors, recipients and the resulting transformants in transformation experiments

Donor MIC Transformant Donor sauce Donor Recipients (μg/ml) MIC (μg/ml)

Vagococcus sp. Fish feed Tetr: 256 L. lactics Tetr: 512 (tetL)

Psychrobacter sp. Fish feed Sulr: 512 E. coli Sulr: >512 (sul 2)

Fish Aeromonas sp. Sulr: >512 E. coli Sulr: >512 intestine (sul1)

3.5 Discussion

It is becoming recognized that the AR ecology, of ART bacteria and AR genes emergence, persistence and dissemination in the microbial ecosystem, is much more complicated than previously thought. As part of the 2-year study (Phase I), we have found that even without the application of antibiotics, ART bacteria were still detected in 74

domestic aquaculture products (134). In this study, we found majority of them were resistant to more than one antibiotic, and some of them had MIC values, higher than 512

μg/ml of Sul, 512 μg/ml of Tet, 512 μg/ml of Erm and 512 μg/ml of Ctx. Therefore, the aquaculture system is a rich reservoir of AR, even without direct exposure to antibiotics, with contributing risk factors yet to be carefully assessed.

Using next generation sequencing, the profiles of Tetr and Ctxr bacterial populations associated with fish intestine, surface rinsing water, fish feed, pond water and mud were determined in the Phase I of the study (134). This study screened for the presence of representative AR genes and identified carriers of these AR genes belong to

18 genera, in agreement with the findings from the AR microbiome profiling study

(Phase I). However, some dominant ART bacteria, for instance, Lysbacillus sp. in Tetr populations from fish feed, Sporossarcina sp. in Tetr populations from mud and

Acinetobacter sp. in Tetr populations from pond water, determined in the Phase I of our study, were not detected here. This could be due to the limited types of AR gene assessed.

Furthermore, as illustrated in Table 3.3, the dominant types of AR genes and carriers in fish intestine were not the same as those in feed samples. The results indicated that host factors have a role in the selective enrichment of ART bacteria.

Regarding 13 AR determinants examined, 30 (30.6%) of 98 ART isolates from fish feed samples were found to carry various AR genes and the positive detection rate was much higher than the ART isolates from other samples, being 9.5% (34 of 357) in fish intestine, 21% (44 of 209) in surface rinsing water, 4% (10 of 244) in pond water and

1.9% (5 of 269) in mud sample. This result was consistent with finding from that higher 75

variety and level of AR gene pools were detected in fish feed samples (Phase I profile study). As illustrated in the Phase I profile and this characterization studies, the skin and intestine of fish were found to contain large number of ART bacteria and AR genes.

Since fish feed contains animal and fish byproducts as important protein source, without proper processing, these byproducts may become the source of secondary contamination in fish feed which could further be a potential risk factors spreading the ART bacteria in the aquaculture systems (fish products and environmental samples).

It is worth noting that during the assessment of resistance persistence, one

Enterococcus strain from fish intestine sample lost Tetr and Ctxr traits but retained

Sur/Trir in most progenies after 30 days of consecutive inoculation without antibiotics which suggested that Tetr and Ctxr were likely encoded by plasmids. As mentioned, no antibiotics were applied in the farm examined in this study. The fact that the resistant traits were maintained in the particular isolate from the aquaculture system without the corresponding antibiotics suggested additional survival advantage mechanisms of the strain in the host, which may be worthy of further investigation.

In this study, the ART flora targeted were commensal bacteria. Although several types of pathogens due to their direct public health relevance have been the focus of research for the past decades, emerging evidence since 2005 strongly demonstrated the key roles of commensal bacteria in AR ecology from resistance development, amplification, dissemination to persistence due to their dominant populations and diversified genetic background in both natural, food and host environments (2–4, 33). In this case, both the resistance gene pools and resistance mechanisms likely are much more 76

prevalent in commensals than in pathogens. Moreover, the frequency of horizontal gene transfer (HGT) events is related to the size of the AR gene pool as well as with the genetic features and compatibility of the donor and recipient (151). In this study, the sul1 gene of Aeromonas sp. from fish intestine, the sul2 gene of Psychrobacter sp. and the tetL gene of Vagococcus sp. from fish feed were functional after being transferred to the corresponding recipients, suggesting they can serve as a source of AR genes if involved in HGT events.

Finally, representative AR genes were only found in a small percent of the ART isolates. Further functional genomic can be conducted to discover unknown AR determinants in these samples, for a better understanding of AR gene evolution within the ecosystem.

Chapter 4 : Identification of a new tetD variant from domestic aquaculture products

4.1 Abstract

To better understand the antibiotic resistance (AR) ecology, functional genomic analyses were conducted to identify undescribed AR genes associated with samples from a fish farm in Ohio where antibiotics were not used in the production. A fosmid genomic library generated from a pooled DNA of 6 ART isolates was used to screen for AR genes against tetracycline. Tetr clones were further subjected to subcloning using pBluescript II

KS (+) vector, followed by DNA sequence analysis. A 4.7-kb fragment from a Tetr subclone T61 contained two divergently transcribed open reading frames (ORFs). The larger ORF encoded a 398-amino-acid protein with 88% identity to major facilitator superfamily (MFS) transporter and 87% identity to class D tetracycline/H+ antiporter.

The smaller ORF encoded a 199-amino-acid protein with 86% identity to TetR family transcriptional regulator. Deletion mutagenesis confirmed the involvement of the two

ORFs in Tetr. The original host of the new tetD variant, designated tetD(Y), was identified as Providencia sp. The data illustrated the prevalence of new, functional and potentially transmissible AR determinants in fish gastrointestinal (GI) microbiota without direct exposure to antibiotics. 78

4.2 Introduction

The rapid emergence of AR has become a major public health concern. Emerging data suggested that the food chain may serve as an important avenue of disseminating

ART bacteria to humans, bridging the AR microbiota between the environment and the hosts (2, 29–31). Particularly, the prevalence of ART or multidrug-resistant (MDR) bacteria and resistance determinants associated with aquaculture products and environment have received increasing attention and have been investigated by many research groups (62, 67, 81–83, 87, 88, 91, 106, 122, 152). For instance, Schmidt et al.

(67) determined AR determinants in a collection of 313 Aeromonans isolated from

Danish rainbow trout farms and found tetA, tetD, tetE and class 1 integron carrying different resistance gene cassettes such as dhfr2a-ant(3’’)1a, dhfr1-ant(3’’)1a, dhfr1- ant(3’’)1a-catB2 or dhfr1-ant(3’’)1a-orf-catB2. AR genes including tetM, tetO, tetT, tetW, sul1 and sul2 were detected in ART commensal bacteria from aquaculture environment in Tianjin, northern China (82). In another study conducted by Ye et al. (88), high levels of ART bacteria were found in all ten aquaculture products and multiple AR determinants were detected, such as sul1, sul2, intI1, tetE, tetM, tetL, tetS, blaTEM, blaCMY, ermB and ermC. Despite the potential contribution of antibiotic usage in aquaculture production, we found in Phase I of a 2-year study that ART bacteria were also abundant in fish products from a farm in Ohio without antibiotic application (134). Since no antibiotics were applied in this particular aquaculture farm production, characterization of these ART bacteria can potentially lead to better understanding on AR risk factors other than direct antibiotic exposure, as well as potential insights on AR gene evolution. In

agreement with previous studies (2, 29, 88), only approximately 10% of the ART isolates from feed, aquaculture products and environmental samples were found containing representative AR genes examined in Phase I of the study. Therefore it becomes important to reveal the genetic background of the rest of the ART bacteria. Functional genomic approach has been reported as a useful tool to identify diverse AR genes in recent years (135, 153–155). Sommer et al. (135) functionally characterized AR reservoir in human microbiome from both culture-independent sampling and cultured aerobic isolates. The results showed that AR genes identified in cultured aerobic isolates were more closely related to known genes harbored by pathogens while most AR genes identified using culture-independent approach were distant from previously identified resistance determinants. While the culture-independent approach recovers potential AR determinants including those from non-culturable subpopulations, the lack of backup cultures made it difficult for further molecular characterization and interpretation of AR ecology. In this study, the functional genomic method was adapted to examine AR determinants in culture-recovered isolates. Strain identification and additional molecular characterizations were further conducted for potential insights on AR evolution, dissemination and persistence in the corresponding aquaculture ecosystem.

4.3 Materials and methods

4.3.1 Bacterial strains and culture condition

Bacterial isolates exhibiting resistance against two to four antibiotics sulfamethoxazole with trimethoprim (Sul/Tri), tetracycline (Tet), erythromycin (Erm) or cefotaxime (Ctx) from multiple aquaculture samples (including fish intestine and surface rinsing water) were used in the study. The isolates lacked representative AR determinants

r r (including tetL, tetM and tetS for Tet , sul1 and sul2 for Sul , blaCMY-2, blaCTX, blaTEM,

r r blaSHV and blaOXA for Ctx and ermB and ermC for Erm ) examined by PCR in our 2-year study (Phase I). Those isolates were cultivated in BHI media containing corresponding antibiotics at 30°C. Six isolates exerting higher MIC value against Tet (no less than 512

µg/ml) were selected for functional genomics analysis (Table 4.1).

Table 4.1. ART bacterial isolates used in the study.

MIC against Tet AR phenotypes Isolate Source (µg/ml)

Surface Rinsing Sulr, Tetr, Ctxr, Ermr 512 No.1 Water

Sulr, Tetr, Ctxr, Ermr >512 No.2 Fish Intestine

Sulr, Tetr, Ctxr, Ermr 512 Providencia sp. No.3 Fish Intestine

Sulr, Tetr, Ctxr, Ermr 512 No.4 Fish Intestine

Sulr, Tetr, Ctxr, Ermr 512 No.5 Fish Intestine

Sulr, Tetr 512 Aeromonas sp. No.6 Fish Intestine

4.3.2 Genomic library construction and subcloning

Genomic DNA from isolates was extracted according to the method described by

Sambrook and Russell (156). A fosmid genomic library was generated using a

CopyControl pCC1FOS fosmid library production kit according to the protocol from

Epicentre (Madison, WI). Briefly, extracted DNA samples were pooled, sheared, and then end-repaired. DNA fragments of 25-40 kb were purified and ligated into pCC1FOS followed by packaging into phage and introducing into the Escherichia coli EPI300. The transformation mixture was plated onto Luria-Bertani (LB) medium containing chloramphenicol (Chl, 12.5 µg/ml) (Sigma-Aldrich, St. Louis, MO) or Tet (16 µg/ml) 82

(Sigma-Aldrich), and incubated at 37°C for 24 h. All Chlr colonies were washed from the

Chl containing plates and stocked in pools. Five hundred Tetr colonies were randomly selected and spotted onto Chl containing LB plates to confirm its resistance to Chl.

Tetr clones with MIC of no less than 64 μg/ml were randomly selected and fosmid

DNA from each clone was isolated by alkaline lysis method (156) and their restriction patterns were determined using seven common restriction enzymes (BamHI, EcoRI,

HindIII, KpnI, SacI, XbaI and XhoI) (New England Biolabs, Ipswich, MA) independently.

The clones showing unique restriction pattern with DNA fragment between 3 and 10 kb were selected. The digested fosmid DNA was ligated into pBlueScript II KS (+) and transformed into E.coli DH5α (Invitrogen, Grand Island, NY). The transformation mixtures were plated on LB plates containing Tet and X-Gal (20 mg/ml) (Life

Technologies, Carlsbad, CA), and then incubated at 37°C for 24 h.

4.3.3 DNA sequencing and bioinformatics analysis

Plasmid DNA from Tetr subclones were extracted using QIAprep Spin Miniprep

Kit (Qiagen Inc., Valencia, CA). A 4.7-kb fragment inserted in the recombinant plasmid pT61 was sequenced from both ends using T7 and T3 primers followed by primer walking. Sequences were assembled and restriction enzyme sites were analyzed using

SeqMan (DNAstar, Madison, WI). Putative open reading frames (ORFs) were identified and annotated using ORF Finder program at NCBI website. The predicted resistance related genes were translated using EditSeq (DNAstar, Madison, WI), transmembrane domains were analyzed by the Simple Modular Architecture Research Tool

(http://smart.embl.de/) and helix-turn-helix motif was predicted using online tool 83

(http://npsa-pbil.ibcp.fr/cgi-bin/npsa_automat.pl?page=/NPSA/npsa_hth.html).

Phylogenetic analysis was conducted with the neighbor-joining method using Ｍegalign

(DNAstar, Madison, WI).

4.3.4 Determination of the locus of Tetr determinants

To test the functionality of each pair of TetR and MFS transporter, the plasmid pT61 was digested by SpeI and two fragments were produced. The vector-containing larger fragment was purified, self-ligated, while the smaller fragment was purified, ligated into a new pBluescript II KS (+) vector. The two ligated DNA products were independently transformed into E.coli DH5α and plated onto LB plates containing X-Gal and Tet, and then incubated at 37°C for 24 h.

4.3.5 Identification of tetD(Y) gene carrying host

To identify the bacterial source of this new tetD(Y) gene, specific primers were designed based on the tetD(Y) gene sequence. The six isolates were assessed for the presence of the tetD(Y) gene by PCR following the procedure as described previously

(29). The positive bands were confirmed by DNA sequencing at the Ohio State

University Plant Genomic facility.

4.3.6 Southern hybridization

Genomic DNA from isolates were extracted and southern blotting was performed as described by Sambrook and Russell (156). The strain Lactococcus lactis 2301 was used as a negative control. The hybridization probes were synthesized by PCR using

DNA from Tetr subclone as the template and primer T1 from this study. The probe was

labeled and detected with the DIG DNA Labeling and Detection Kit (Roche Diagnostics,

IN, USA).

4.4 Results

4.4.1 Construction of genomic library and identification of the Tetr subclones

One genomic library from six Tetr isolates was constructed in E. coli EPI300 and screened for clones that conferred resistance against both Chl and Tet. Fosmids were extracted and the restriction digestion results showed digested fosmids by HindIII were suitable for subcloning. One Tetr subclone T61 was selected. Moreover, HindIII digestion confirmed one insertion DNA fragment of about 4.7kb was ligated to pBlueScript II KS

(+) vector (Fig. 4.1).

Figure 4.1 Determination of the size of the insertion fragment in plasmid pT61. M: 1 kb ladder; Lane1: pT61 digested by HindIII. The lower band showed pBluescript vector.

4.4.2 DNA sequencing and bioinformatics analysis results

Predicted ORFs within the 4.7-kb DNA insertion fragment were listed in Table

4.2. Among them, there were two ORFs encoding for major facilitator superfamily (MFS) transporter and two ORFs encoding for TetR family transcriptional regulator which could potentially related to the Tet resistance of the subclone T61. No significant similarities were found between these two MFS or two TetR family transcriptional regulators. Based on their locus, these four ORFs could be divided into two pairs of MFS transporter and

TetR (ORF 2 and 4, ORF 7 and 9). Besides having high identity with MFS transporter,

ORF 4 also showed identity with TetD. In addition, ORF 2 and 4 divergently transcribed which agreed with the feature of TetR regulation system. Therefore, ORF 2 and 4 much more likely contributed to the tetracycline resistance compared with ORF 7 and 9.

Table 4.2. Putative ORF located in plasmid pT61.

ORF ORF ORF Accession no. of closest Predicted function of closest match No. start stop match (% identity)

1 92 241 Fumarate reductase subunit D WP004258620.1 (100)

2 904 305 TetR family transcriptional WP004258624.1(86)

regulator

3 876 992 Hypothetical protein YP001338865.1 (71)

4 985 2181 MFS transporter a, WP004258627.1(88)a

Class D tetracycline/H+ antiporter WP004908113.1(87)

5 1266 1523 tRNA modification GTPase mnmE CDM37125.1 (41)

6 1334 1492 Aminopeptidase 2 EPE07584.1 (39)

7 2286 2867 TetR family transcriptional WP004908111.1(90)

regulator

8 3089 2976 Amidase WP007801731.1 (65)

9 4197 3013 MFS transporter WP004258633.1(85)

10 4668 4249 Hypothetical protein WP004908109.1 (69) a: The annotation of the sequence was changed into Tet resistant protein after the author has completed the defense.

Sequence analysis revealed a SpeI digestion site between the two pairs of putative

TetR and MFS transporter within the 4.7-kb insertion fragment (Fig. 4.2). Thus T61 was digested by SpeI and the subclones containing each of the two DNA fragments were examined for Tetr phenotype, respectively. Results showed that the new recombinant

plasmid containing ORF2 (305-904 bp) and ORF4 (985-2181bp) conferred the resistance against Tet.

SpeI (25) HindIII (61)

TetR family transcriptional regulator 1 pBluescript II KS (+) backbone

MFS transporter, TET resistance protein pT61

7687bp SpeI (2182)

TetR family transcriptional regulator HindIII (4787)

MFS transporter Figure 4.2. Schematic diagram of the plasmid pT61 from the Tetr subclone T61.

The ORF4 predicted a 398-animo-acid protein which was most closely related to the class D tetracycline/H+ antiporter from Providencia sp. (GenBank accession no.

WP004908113.1 and EKT59059.1; 87.11% identity). But it only exhibited 68.75% identity to class D Tetr protein from other genera (Fig. 4.3). Less than 80% amino acid identity is the standard for a new tetracycline resistant determinant according to the current nomenclature system (157). Therefore, the Tetr determinant found in this study belonged to class D Tet resistance gene, and represented a new variant of tetD, designated tetD(Y).

8 YP 007349484.1 Escherichia coli EUM99568.1 Enterobacter sp. 2 Cluster 1: EZR10850.1 Klebsiella pneumoniae 13 9 ADF47467.1 Pseudomonas aeruginosa 68.75% identity 89 YP 002995632.1 Aeromonas hydrophila 11 AFV98769.1 Candidatus Snodgrassella YP 001102244.1 Yersinia P33733.1 Salmonella enterica Tet D(Y) Cluster 2: 100 WP 004908113.1 Providencia rettgeri 100 EKT59059.1 Providencia rettgeri Dmel1 87.11% identity

Figure 4.3. Schematic diagram of phylogenetic relationship of TetD(Y)

Consistent with the feature of the group1 efflux pumps (158), the TetD (Y) protein was predicted to have 12 transmembrane regions (located at amino acids 7 to 29,

39 to 61, 73 to 95, 99 to 118, 130 to 152, 157 to 179, 211 to 233, 248 to 267, 274 to 293,

297 to 319, 332 to 354, and 364 to 386) (Fig. 4.4).

The ORF2 which divergently transcribed from tetD(Y) predicted a 199-animo- acid protein TetR(Y). It has 86% identity to the TetR family transcriptional regulator from Providencia rettgeri (NCBI assession no. WP004258624.1). The presence of helix- turn-helix motif (amino acids 25 to 46) supported its putative function of transcriptional regulator (159, 160).

In the 80-nucleotide intergenic region between tetD (Y) and tetR (Y), there existed two putative operator sequences, which were separated by 24-nucleotides, for

TetR (Y) binding (TCTATTGACACTCTA and TCTATCACTGATAGA) (Fig. 4.5). The second operator was highly similar to the recognized TetR binding sites (160). The first half of the first operator matched the second operator except for 2 nucleotides, although the complete sequence didn’t closely match the second operator. Also, the first operator sequence was not perfect palindromic and differed by 3 bases. Similar variations were also reported by other researchers (161). The putative promoters were also identified.

Figure 4.4. Scheme of transmembrane structure of TetD (Y) by SMART.

Figure 4.5. Complementary DNA sequence of the 80-bp tetD(Y)–tetR(Y) intergenic region. The sequences with border around referred to the putative TetR(Y) binding operators.

The sequences under the arrows indicated the putative promoters.

4.4.3 Identification of tetD(Y) gene carrying host

To identify the bacterial source of this new tetD(Y) gene, two pairs of primers were designed (Table 4.3). The original host of tetD(Y) was identified as 3 isolates of

Providencia spp. by PCR.

Table 4.3 PCR primers for identification of tetD(Y) gene carrying host.

Primer Sequence Product Size

T1 F GCG TTT GGC GTG GGT TTA AT 626 bp T1 R GAC CCC TGT GGC ATT GGT TA

T2 F CAG CAT TAA CGA TCA CCG CC 984 bp T2 R ACC CCT TGG AGC TTT CCT TG

4.4.4 Southern hybridization

Since no plasmid was isolated from these isolates using different methods

(including traditional methods and commercial kit), genomic DNA was used in southern blotting. Only the chromosomal DNA from three Providencia isolates showed the positive band which agreed with the results of PCR detection. Also, the results proved that the tetD(Y) gene was located in chromosome (Fig. 4.6).

Figure 4.6 Electrophoresis and southern hybridization of genomic DNA of six isolates.

Lane M: 1 kb Marker; Lane 1: Lactococcus lactis 2301; Lane 2: Isolate No.1; Lane 3: Isolate

No.6; Lane 4: Isolate No.2; Lane 5: Isolate No.3; Lane 6: Isolate No.4; Lane 7: Isolate No.5.

4.4.5 Nucleotide sequence accession numbers

Sequence data from this work were deposited in GenBank with the following accession numbers: KP137701 and KP137702.

4.5 Discussion

In this study, we found a new tetD varient gene in Providencia sp. associated with the fish intestine samples from a domestic aquaculture farm without antibiotic applications. This gene could also function in E. coli with MIC of 256 μg/ml. Results suggested the presence of new, functional and potentially transmissible AR determinants in fish GI tract. Since antibiotics were not used in the aquaculture production in the specific Ohio farm, further identification of the mechanism and potential risk factors contributing to the AR gene evolution and persistence in the aquaculture environment will help understand the AR ecology.

Providencia genus is gram-negative urea-producing bacilli which naturally presents in water, soil and sewage (162). Providencia spp. were reported in various aquaculture products and the environment (83, 84, 87). Current interests in fish gut microbiota focus on two aspects, mainly probiotics and pathogens. Only limited information about Providencia was available. Among Tetr Providencia spp., several tetr genes were reported, including tetB, tetE, tetG, tetM, tet39

(http://faculty.washington.edu/marilynr/tetweb2.pdf), but tetD was rarely reported in

Providencia spp., with only two records found in NCBI website. This is the first thorough study of tetD in Providencia spp.

There are several reasons that motivated us to investigate the ART bacteria and

AR genes in fish intestine. First, the microbiota in fish intestine greatly affects the host health, such as through impacting nutrition utilization and immunity development. Also, it is now recognized that the host GI tract plays an important role in the enrichment of 93

ART bacteria as well as AR genes in the ecosystem, independent of selective pressure by antibiotics (33). During the seafood processing, those ART bacteria can contaminate other sterile fish parts, spread to the processing equipment and transmit to the workers.

These ART bacteria may enter into the human GI tract through consumption of uncooked or incompletely cooked aquaculture products. In addition, the release of ART bacteria- and AR gene-rich feces of fish further impacts the aquaculture environment and therefore the ecosystem. Moreover, some fish remedies are used as ingredients of fish feed, thus can facilitate the dissemination and circulation of ART bacteria and AR genes within the aquaculture system. Hence, an improved knowledge of AR associated with fish intestine microbiota is essential for the design of targeted AR mitigation in the aquaculture ecosystem and seafood safety.

Our results illustrated the presence of AR determinants in the absence of the selective pressure of manmade antibiotics. Based on the DNA sequence information, we hypothesize that two possible routes might be involved in tetD(Y) evolution. First, this gene could evolve from a tetD which could be previously acquired through previous horizontal gene transfer (HGT) event. So far, tetD was reported in gram-negative bacteria including Providencia, Alteromoas, Candidatus, Citrobacter, Edwardsiella, Enterobacter,

Escherichia, Klebsiella, Morganella, Plesiomonas, Salmonella, Shigella, Yersinia,

Aeromonas, Brevundimonas, Burkholderia, Chryseobacterium, Halomonas, Pasteurella,

Photobacterium, Pseudomonas, Rhizobium, Shewanella, Vibrio (NCBI website record and http://faculty.washington.edu/marilynr/tetweb2.pdf). Among them, 13 genera belong to the Enterobacteriaceae family including Providencia. The similar genetic background

within this family would facilitate the HGT event. The tetD(Y) gene was found to be located in the chromosome. Despite lower frequency, the gene encoded by chromosome could also involve HGT event, which has been illustrated previously (120). Second, this gene was mutated from MFS transporter within this genus of Providencia. This type of transporter was highly prevalent in Providencia and the TetD(Y) protein showed high identity (88%) with the reported MFS transporter in Providencia rettgeri

(WP004258627.1). Therefore, the tetD(Y) gene could be derived from a mutation in the

MFS transporter-encoding gene. While Tet is absent in this production system, the chromosomal location of the gene explained persistence of the trait. However, the evolutional advantage of the mutation to ART bacterium is worth further investigation.

Finally, in order to control AR in aquaculture products and the ecosystem, other risk factors should be considered besides the application of the antibiotics. For instance, the potential contribution of production methods, management practices and environment conditions to AR in the aquaculture system needs to be reassessed.

Summary and future direction

AR is a complicated issue with multiple risk factors. AR in aquaculture including fish production is even more complex, since many factors such as the aquaculture creatures themselves, the feces released, the production practices (feed, antibiotics application, manure enrichment, etc.), and environmental conditions (water and mud) all directly impact the ART bacterial community in the aquaculture ecosystem. To assess potential AR risk factors in aquaculture production, this study examined AR in aquaculture products and environmental samples for one-year from a fish farm in Ohio where antibiotics were not used in the production.

Despite the absence of antibiotic application in the farm, our results showed that

ART bacteria were prevalent in all samples examined, including fish intestine, surface rinsing water, feed, pond water and mud samples. By NGS, a total of 569 genera were identified in Tetr and Ctxr bacteria from five types of samples. AR genes measured by quantitative PCR (qPCR) were significantly more abundant in fish intestine and feed than farm environmental samples. Characterization of the ART isolates from the aquaculture samples found that 79% of 4747 ART isolates examined were resistant to more than one antibiotic. Some ART isolates showed the MIC of Sul/Tri, Tet, Erm or Ctx higher than

512 μg/ml. Various AR genes, including sul1, sul2, tetS, tetL, tetM, ermB or ermC, were 96

detected in samples, and majority of the AR genes tested were stably retained in the isolates at the absence of the corresponding antibiotics. A new tetD variant, designated tetD(Y) was identified from a fosmid genomic library generated from a pooled DNA of 6

ART isolates. The original host of the new tetD(Y) was identified as Providencia sp.

Our results contributed to an improved understanding regarding the nature of

ART commensal bacteria in domestic aquaculture production. Also, the results suggested that besides direct exposure to antibiotics, additional AR risk factor(s) also significantly contribute to AR in aquaculture production. Fish feed might be a potential risk factor and the host GI tract likely plays important role in the selective enrichment of ART bacteria and AR genes. Therefore effective AR mitigation should base on a comprehensive understanding of all critical risk factors, and to address each of them in the production.

The data further indicated that once the resistant bacteria derived, many will become persistent in the environment due to various molecular mechanisms, even after the selective pressure was lifted. Therefore AR mitigation should focus on prevention instead of responding. Results from this study are consistent with findings from our laboratory in the past ten years crossing the retail foods, human, food and non-food animals, and thus contribute to the broad understanding of the AR ecology.

Future studies will be conducted using controlled aquaculture units to assess the details of potential AR risk factors. In addition, more molecular characterization of ART isolates will be conducted to reveal novel AR genes and molecular mechanisms involved in AR persistence. The results of these studies will contribute to an improved understanding of AR ecology in the aquaculture production system. 97

Bibliography

1. Centres for Disease Control and Prevention (US). 2013. Antibiotic resistance threats in the United States, 2013.

2. Wang HH, Manuzon M, Lehman M, Wan K, Luo H, Wittum TE, Yousef A, Bakaletz LO. 2006. Food commensal microbes as a potentially important avenue in transmitting antibiotic resistance genes. FEMS Microbiol. Lett. 254:226–231.

3. Wang HW. 2009. Commensal bacteria, microbial ecosystems, and horizontal gene transmission: adjusting our focus for strategic breakthroughs against antibiotic resistance., p. 267–281. In Jaykus, LA, Wang, HH, Schlesinger, LS (eds.), . ASM Press.

4. Wang HH, Schaffner DW. 2011. Antibiotic resistance: how much do we know and where do we go from here? Appl. Environ. Microbiol. 77:7093–7095.

5. FAO. 2012. The State of World Fisheries and Aquaculture 2012. Rome.

6. Khardori N. 2006. Antibiotics—past, present, and future. Med. Clin. North Am. 90:1049–1076.

7. Morar M, Wright GD. 2010. The genomic enzymology of antibiotic resistance. Annu. Rev. Genet. 44:25–51.

8. Kuzin AP, Sun T, Jorczak-Baillass J, Healy VL, Walsh CT, Knox JR. 2000. Enzymes of vancomycin resistance: the structure of d-alanine–d-lactate ligase of naturally resistant Leuconostoc mesenteroides. Structure 8:463–470.

9. Bockstael K, Van Aerschot A. 2009. Antimicrobial resistance in bacteria. Cent. Eur. J. Med. 4:141–155.

10. Poehlsgaard J, Douthwaite S. 2005. The bacterial ribosome as a target for antibiotics. Nat. Rev. Microbiol. 3:870–881. 98

11. Diekema DJ, Pfaller MA, Schmitz FJ, Smayevsky J, Bell J, Jones RN, Beach M. 2001. Survey of infections due to Staphylococcus species: frequency of occurrence and antimicrobial susceptibility of isolates collected in the United States, Canada, Latin America, Europe, and the Western Pacific region for the SENTRY Antimicrobial Surveillanc. Clin. Infect. Dis. 32:S114–S132.

12. Edlund T, Grundström T, Normark S. 1979. Isolation and characterization of DNA repetitions carrying the chromosomal β-lactamase gene of Escherichia coli K-12. MGG Mol. Gen. Genet. 173:115–125.

13. Jaurin B, Grundström T, Edlund T, Normark S. 1981. The E. coli β-lactamase attenuator mediates growth rate-dependent regulation. Nature 290:221–225.

14. Lindberg F, Lindquist S, Normark S. 1987. Inactivation of the ampD gene causes semiconstitutive overproduction of the inducible Citrobacter freundii beta- lactamase. J. Bacteriol. 169:1923–1928.

15. Lindberg F, Lindquist S, Normark S. 1988. Genetic basis of induction and overproduction of chromosomal class I β--lactamase in nonfastidious Gram- negative Bacilli. Clin. Infect. Dis. 10:782–785.

16. Martínez JL, Baquero F, Andersson DI. 2007. Predicting antibiotic resistance. Nat. Rev. Microbiol. 5:958–965.

17. Andersson DI, Hughes D. 2010. Antibiotic resistance and its cost: is it possible to reverse resistance? Nat. Rev. Microbiol. 8:260–271.

18. Sørum H, L’Abée-Lund TM. 2002. Antibiotic resistance in food-related bacteria—a result of interfering with the global web of bacterial genetics. Int. J. Food Microbiol. 78:43–56.

19. De Pascale G, Wright GD. 2010. Antibiotic resistance by enzyme inactivation: from mechanisms to solutions. Chembiochem 11:1325–1334.

20. Robicsek A, Strahilevitz J, Jacoby GA, Macielag M, Abbanat D, Park CH, Bush K, Hooper DC. 2006. Fluoroquinolone-modifying enzyme: a new adaptation of a common aminoglycoside acetyltransferase. Nat. Med. 12:83–88.

21. McMurry L, Petrucci RE, Levy SB. 1980. Active efflux of tetracycline encoded by four genetically different tetracycline resistance determinants in Escherichia coli. Proc. Natl. Acad. Sci. 77:3974–3977.

22. Poole K. 2005. Efflux-mediated antimicrobial resistance. J. Antimicrob. Chemother. 56:20–51.

23. Didier J-P, Villet R, Huggler E, Lew DP, Hooper DC, Kelley WL, Vaudaux P. 2011. Impact of ciprofloxacin exposure on Staphylococcus aureus genomic alterations linked with emergence of rifampin resistance. Antimicrob. Agents Chemother. 55:1946–1952.

24. Finken M, Kirschner P, Meier A, Wrede A, Böttger EC. 1993. Molecular basis of streptomycin resistance in Mycobacterium tuberculosis: alterations of the ribosomal protein S12 gene and point mutations within a functional 16S ribosomal RNA pseudoknot. Mol. Microbiol. 9:1239–1246.

25. Lambert PA. 2005. Bacterial resistance to antibiotics: modified target sites. Adv. Drug Deliv. Rev. 57:1471–1485.

26. Lim D, Strynadka NCJ. 2002. Structural basis for the β lactam resistance of PBP2a from methicillin-resistant Staphylococcus aureus. Nat. Struct. Mol. Biol. 9:870–876.

27. Doyle M, Busta F, Cords B, Davidson P. 2006. Antimicrobial resistance: implications for the food system. Compr. Rev. Food Sci. Food Saf.

28. Davies J, Davies D. 2010. Origins and evolution of antibiotic resistance. Microbiol. Mol. Biol. Rev. 74:417–433.

29. Li X, Wang H. 2010. Tetracycline resistance associated with commensal bacteria from representative ready-to-consume deli and restaurant foods. J. Food Prot. 73:1841–1848.

30. Manuzon MY, Hanna SE, Luo H, Yu Z, Harper WJ, Wang HH. 2007. Quantitative assessment of the tetracycline resistance gene pool in cheese samples by real-time TaqMan PCR. Appl. Environ. Microbiol. 73:1676–1677.

31. Duran GM, Marshall DL. 2005. Ready-to-eat shrimp as an international vehicle of antibiotic-resistant bacteria. J. Food Prot. 68:2395–2401.

32. Li X, Alvarez V, Harper WJ, Wang HH. 2011. Persistent, toxin-antitoxin system-independent, tetracycline resistance-encoding plasmid from a dairy Enterococcus faecium isolate. Appl. Environ. Microbiol. 77:7096–7103.

33. Zhang L, Kinkelaar D, Huang Y, Li Y, Li X, Wang HH. 2011. Acquired antibiotic resistance: are we born with it? Appl. Environ. Microbiol. 77:7134–7141.

34. Li X, Sun K, Zhang L, Li YL, Wang. HH. 2010. The involvement of animal host in the enrichment of antibiotic resistance. IFT Annual Meeting. Chicago, IL.

100

35. Koike S, Krapac IG, Oliver HD, Yannarell AC, Chee-Sanford JC, Aminov RI, Mackie RI. 2007. Monitoring and source tracking of tetracycline resistance genes in lagoons and groundwater adjacent to swine production facilities over a 3-year period. Appl. Environ. Microbiol. 73:4813–4823.

36. Zhang L, Huang Y, Zhou Y, Buckley T, Wang HH. 2013. Antibiotic administration routes significantly influence the levels of antibiotic resistance in gut microbiota. Antimicrob. Agents Chemother. 57:3659–3566.

37. Landau M. 1992. Introduction to aquaculture. John Wiley & Sons, New York.

38. Sapkota A, Sapkota AR, Kucharski M, Burke J, McKenzie S, Walker P, Lawrence R. 2008. Aquaculture practices and potential human health risks: current knowledge and future priorities. Environ. Int. 34:1215–1226.

39. Tidwell JH. 2012. Aquaculture Production Systems. Wiley-Blackwell.

40. FAO. 2014. The State of World Fisheries and Aquaculture 2014. Rome.

41. FAO. 2010. The State of World Fisheries and Aquaculture 2010. Rome.

42. Rodger H, Colquhoun D. 2008. Clinical vibriosis in farmed Atlantic cod (Gadus morhua) in Ireland. Vet. Rec. 162:94–95.

43. Ibrahem M, Hatem M. 2009. Mass mortality caused by Pseudomonas fluorescens and Cryptocaryon irritans in gilthead sea bream (Sparus aurata L.) in Egyptian mariculture., p. 275–289. In Proceedings of the 2nd Global Fisheries and Aquaculture Research Conference. Massive Conferences and Trade Fairs.

44. Altinok I, Balta F, Capkin E, Kayis S. 2007. Disease of rainbow trout caused by Pseudomonas luteola. Aquaculture 273:393–397.

45. Yiagnisis M, Vatsos I, Kyriakou C, Alexis M. 2007. First report of Vibrio anguillarum isolation from diseased big scale sand smelt, Atherina boyeri Risso 1810, in Limnos, Greece. Bull. Eur. Assoc. Fish Pathol. 27:61–69.

46. Yashima T, Odajima T. 1992. An outbreak of Vibriosis in wild ayu (Plecoglossus altivelis) and characteristics of the iolates. J. Japan Vet. Med. Assoc. 45:203–206.

47. Oraić D, Zrnčić S, Šoštarić B. 2002. Occurrence of enteric redmouth disease in rainbow trout (Oncorhynchus mykiss) on farms in Croatia. Acta Vet. Hung. 50:283–291.

101

48. Rodríguez L, Castillo A, Gallardo C, Nieto T. 1999. Outbreaks of Yersinia ruckeri in rainbow trout in north west of Spain. Bull. Eur. Assoc. Fish Pathol. 19:130–132.

49. Eissa A, Moustafa M, Abdelaziz M, Ezzeldeen N. 2008. Yersinia ruckeri infection in cultured Nile tilapia, Oreochromis niloticus, at a semi-intensive fish farm in lower Egypt. African J. Aquat. Sci. 33:283–286.

50. Soltani M, Nikbakht G, Ebrahimzadeh Moussavi H, Ahmadzadeh N. 2008. Epizootic outbreaks of Lactoccoccosis caused by Lactococcus garvieae in farmed rainbow trout (Onchorhynchus mykiss) in Iran. Bull. Eur. Assoc. Fish Pathol. 28:95–106.

51. Cabello FC, Godfrey HP, Tomova A, Ivanova L, Dölz H, Millanao A, Buschmann AH. 2013. Antimicrobial use in aquaculture re-examined: its relevance to antimicrobial resistance and to animal and human health. Environ. Microbiol. 15:1917–1942.

52. Gutsell J. 1946. Sulfa Drugs and the Treatment of Furunculosis in Trout. Science (80-. ). 104:85–86.

53. Alderman D, Hastings T. 1998. Sulfa drugs and the treatment of Furunculosis in Trout. Int. J. Food Sci. Technol. 33:139–155.

54. Rico A, Phu TM, Satapornvanit K, Min J, Shahabuddin a. M, Henriksson PJG, Murray FJ, Little DC, Dalsgaard A, Van den Brink PJ. 2013. Use of veterinary medicines, feed additives and probiotics in four major internationally traded aquaculture species farmed in Asia. Aquaculture 412-413:231–243.

55. Costello BMJ, Grant A, Davies IM, Cecchini S, Papoutsoglou S, Quigley D, Saroglia M. 2001. The control of chemicals used in aquaculture in Europe. J. Appl. Ichthyol. 17:173–180.

56. Thuy HTT, Nga LP, Loan TTC. 2011. Antibiotic contaminants in coastal wetlands from Vietnamese shrimp farming. Environ. Sci. Pollut. Res. Int. 18:835– 841.

57. Le TX, Munekage Y. 2004. Residues of selected antibiotics in water and mud from shrimp ponds in mangrove areas in Viet Nam. Mar. Pollut. Bull. 49:922–929.

58. Li W, Shi Y, Gao L, Liu J, Cai Y. 2012. Occurrence of antibiotics in water, sediments, aquatic plants, and animals from Baiyangdian Lake in North China. Chemosphere 89:1307–1315.

102

59. Kadlec K, von Czapiewski E, Kaspar H, Wallmann J, Michael GB, Steinacker U, Schwarz S. 2011. Molecular basis of sulfonamide and trimethoprim resistance in fish-pathogenic Aeromonas isolates. Appl. Environ. Microbiol. 77:7147–7150.

60. Ndi OL, Barton MD. 2011. Incidence of class 1 integron and other antibiotic resistance determinants in Aeromonas spp. from rainbow trout farms in Australia. J. Fish Dis. 34:589–599.

61. Naviner M, Gordon L, Giraud E, Denis M, Mangion C, Le Bris H, Ganière J- P. 2011. Antimicrobial resistance of Aeromonas spp. isolated from the growth pond to the commercial product in a rainbow trout farm following a flumequine treatment. Aquaculture 315:236–241.

62. Penders J, Stobberingh EE. 2008. Antibiotic resistance of motile aeromonads in indoor catfish and eel farms in the southern part of The Netherlands. Int. J. Antimicrob. Agents 31:261–265.

63. Akinbowale OL, Peng H, Grant P, Barton MD. 2007. Antibiotic and heavy metal resistance in motile aeromonads and pseudomonads from rainbow trout (Oncorhynchus mykiss) farms in Australia. Int. J. Antimicrob. Agents 30:177–82.

64. Agersø Y, Bruun MS, Dalsgaard I, Larsen JL. 2007. The tetracycline resistance gene tet(E) is frequently occurring and present on large horizontally transferable plasmids in Aeromonas spp. from fish farms. Aquaculture 266:47–52.

65. Hatha M, Vivekanandhan a. a., Julie Joice G. 2005. Antibiotic resistance pattern of motile aeromonads from farm raised fresh water fish. Int. J. Food Microbiol. 98:131–134.

66. Jacobs L, Chenia HY. 2007. Characterization of integrons and tetracycline resistance determinants in Aeromonas spp. isolated from South African aquaculture systems. Int. J. Food Microbiol. 114:295–306.

67. Schmidt AS, Bruun MS, Dalsgaard I, Larsen JL. 2001. Incidence, distribution, and spread of tetracycline resistance determinants and integron-associated antibiotic resistance genes among motile aeromonads from a fish farming environment. Appl. Environ. Microbiol. 67:5675–5682.

68. Nawaz M, Sung K, Khan SA, Khan AA, Steele R. 2006. Biochemical and molecular characterization of tetracycline-resistant Aeromonas veronii isolates from catfish. Appl. Environ. Microbiol. 72:6461–6466.

103

69. Radu S, Ahmad N, Ling FH, Reezal A. 2003. Prevalence and resistance to antibiotics for Aeromonas species from retail fish in Malaysia. Int. J. Food Microbiol. 81:261–266.

70. Woodring J, Srijan A, Puripunyakom P, Oransathid W, Wongstitwilairoong B, Mason C. 2012. Prevalence and antimicrobial susceptibilities of Vibrio, Salmonella, and Aeromonas isolates from various uncooked seafoods in Thailand. J. Food Prot. 75:41–47.

71. Kitiyodom S, Khemtong S, Wongtavatchai J, Chuanchuen R. 2010. Characterization of antibiotic resistance in Vibrio spp. isolated from farmed marine shrimps (Penaeus monodon). FEMS Microbiol. Ecol. 72:219–227.

72. Devi R, Surendran PK, Chakraborty K. 2009. Antibiotic resistance and plasmid profiling of Vibrio parahaemolyticus isolated from shrimp farms along the southwest coast of India. World J. Microbiol. Biotechnol. 25:2005–2012.

73. Nakayama T, Ito E, Nomura N, Nomura N, Matsumura M. 2006. Comparison of Vibrio harveyi strains isolated from shrimp farms and from culture collection in terms of toxicity and antibiotic resistance. FEMS Microbiol. Lett. 258:194–199.

74. Rebouças RH, de Sousa OV, Lima AS, Vasconcelos FR, de Carvalho PB, Vieira RHS dos F. 2011. Antimicrobial resistance profile of Vibrio species isolated from marine shrimp farming environments (Litopenaeus vannamei) at Ceará, Brazil. Environ. Res. 111:21–24.

75. Oh E-G, Son K-T, Yu H, Lee T-S, Lee H-J, Shin S, Kwon J-Y, Park K, Kim J. 2011. Antimicrobial resistance of Vibrio parahaemolyticus and Vibrio alginolyticus strains isolated from farmed fish in Korea from 2005 through 2007. J. Food Prot. 74:380–386.

76. Li J, Yie J, Foo R, Ling JM, Xu H, Woo N. 1999. Antibiotic resistance and plasmid profiles of Vibrio isolates from cultured Silver Sea Bream, Sparus sarba. Mar. Pollut. Bull. 39.

77. Budiati T, Rusul G, Wan-Abdullah WN, Arip YM, Ahmad R, Thong KL. 2013. Prevalence, antibiotic resistance and plasmid profiling of Salmonella in catfish (Clarias gariepinus) and tilapia (Tilapia mossambica) obtained from wet markets and ponds in Malaysia. Aquaculture 372-375:127–132.

78. Zhao S, McDermott PF, Friedman S, Qaiyumi S, Abbott J, Kiessling C, Ayers S, Singh R, Hubert S, Sofos J, White DG. 2006. Characterization of antimicrobial-resistant Salmonella isolated from imported foods. J. Food Prot. 69:500–507.

104

79. Hatha AAM, Lakshmanaperumalsamy P. 1995. Antibiotic resistance of Salmonella strains isolated from fish and crustaceans. Lett. Appl. Microbiol. 21:47–49.

80. Di Cesare A, Vignaroli C, Luna GM, Pasquaroli S, Biavasco F. 2012. Antibiotic-resistant enterococci in seawater and sediments from a coastal fish farm. Microb. Drug Resist. 18:502–509.

81. DePaola A, Peeler JT, Rodrick GE. 1995. Effect of oxytetracycline-medicated feed on antibiotic resistance of gram-negative bacteria in catfish ponds. Appl. Environ. Microbiol. 61:2335–2340.

82. Gao P, Mao D, Luo Y, Wang L, Xu B, Xu L. 2012. Occurrence of sulfonamide and tetracycline-resistant bacteria and resistance genes in aquaculture environment. Water Res. 46:2355–2364.

83. Miranda CD, Kehrenberg C, Ulep C, Schwarz S, Roberts MC. 2003. Diversity of tetracycline resistance genes in bacteria from Chilean Salmon farms diversity of tetracycline resistance genes in bacteria from Chilean Salmon farms. Appl. Environ. Microbiol. 47:883–888.

84. Miranda CD, Rojas R, Garrido M, Geisse J, González G. 2013. Role of shellfish hatchery as a reservoir of antimicrobial resistant bacteria. Mar. Pollut. Bull. 74:334–343.

85. Sahul Hameed A, Rahaman K, Alagan A, Yoganandhan K. 2003. Antibiotic resistance in bacteria isolated from hatchery-reared larvae and post-larvae of Macrobrachium rosenbergii. Aquaculture 217:39–48.

86. Sarter S, Kha Nguyen HN, Hung LT, Lazard J, Montet D. 2007. Antibiotic resistance in Gram-negative bacteria isolated from farmed catfish. Food Control 18:1391–1396.

87. Shah SQA, Cabello FC, L’abée-Lund TM, Tomova A, Godfrey HP, Buschmann AH, Sørum H. 2014. Antimicrobial resistance and antimicrobial resistance genes in marine bacteria from salmon aquaculture and non-aquaculture sites. Environ. Microbiol. 16:1310–1320.

88. Ye L, Lu Z, Li X, Shi L, Huang Y, Wang HH. 2013. Antibiotic-resistant bacteria associated with retail aquaculture products from Guangzhou, China. J. Food Prot. 76:295–301.

105

89. Zhang Y Bin, Li Y, Sun XL. 2011. Antibiotic resistance of bacteria isolated from shrimp hatcheries and cultural ponds on Donghai Island, China. Mar. Pollut. Bull. 62:2299–2307.

90. Akinbowale OL, Peng H, Barton MD. 2007. Diversity of tetracycline resistance genes in bacteria from aquaculture sources in Australia. J. Appl. Microbiol. 103:2016–2025.

91. Phuong Hoa PT, Nonaka L, Hung Viet P, Suzuki S. 2008. Detection of the sul1, sul2, and sul3 genes in sulfonamide-resistant bacteria from wastewater and shrimp ponds of north Vietnam. Sci. Total Environ. 405:377–384.

92. Jin Jun L, Bum Jeong J, Huh M-D, Chung J-K, Choi D, Lee C-H, Jeong H Do. 2004. Detection of tetracycline-resistance determinants by multiplex polymerase chain reaction in Edwardsiella tarda isolated from fish farms in Korea. Aquaculture 240:89–100.

93. Ryu S-H, Park S-G, Choi S-M, Hwang Y-O, Ham H-J, Kim S-U, Lee Y-K, Kim M-S, Park G-Y, Kim K-S, Chae Y-Z. 2012. Antimicrobial resistance and resistance genes in Escherichia coli strains isolated from commercial fish and seafood. Int. J. Food Microbiol. 152:14–18.

94. Boinapally K, Jiang X. 2007. Comparing antibiotic resistance in commensal and pathogenic bacteria isolated from wild-caught South Carolina shrimps vs. farm- raised imported shrimps. Can. J. Microbiol. 53:919–924.

95. Spanggaard B, Jorgensen F, Gram L, Huss HH. 1993. Antibiotic resistance in bacteria isolated from thlree freshwater fish farms and an unpolluted stream in Denmark. Aquaculture 115:195–207.

96. Buschmann AH, Tomova A, López A, Maldonado MA, Henríquez LA, Ivanova L, Moy F, Godfrey HP, Cabello FC. 2012. Salmon aquaculture and antimicrobial resistance in the marine environment. PLoS One 7:e42724.

97. Guardabassi L, Dalsgaard A, Raffatellu M, Olsen JE. 2000. Increase in the prevalence of oxolinic acid resistant Acinetobacter spp. observed in a stream receiving the effluent from a freshwater trout farm following the treatment with oxolinic acid-medicated feed. Aquaculture 188:205–218.

98. Tendencia EA, de la Pena LD. 2001. Antibiotic resistance of bacteria from shrimp ponds. Aquaculture 195:193–204.

99. Giraud E, Douet D-G, Le Bris H, Bouju-Albert A, Donnay-Moreno C, Thorin C, Pouliquen H. 2006. Survey of antibiotic resistance in an integrated marine

106

aquaculture system under oxolinic acid treatment. FEMS Microbiol. Ecol. 55:439– 448.

100. Seyfried EE, Newton RJ, Rubert IV KF, Pedersen JA, McMahon KD. 2010. Occurrence of tetracycline resistance genes in aquaculture facilities with varying use of oxytetracycline. Microb. Ecol. 59:799–807.

101. Muziasari WI, Managaki S, Pärnänen K, Karkman A, Lyra C, Tamminen M, Suzuki S, Virta M. 2014. Sulphonamide and trimethoprim resistance genes persist in sediments at Baltic Sea aquaculture farms but are not detected in the surrounding environment. PLoS One 9:e92702.

102. Tamminen M, Karkman A, Lõhmus A, Muziasari WI, Takasu H, Wada S, Suzuki S, Virta M. 2011. Tetracycline resistance genes persist at aquaculture farms in the absence of selection pressure. Environ. Sci. Technol. 45:386–391.

103. Di Cesare A, Luna GM, Vignaroli C, Pasquaroli S, Tota S, Paroncini P, Biavasco F. 2013. Aquaculture can promote the presence and spread of antibiotic- resistant Enterococci in marine sediments. PLoS One 8:e62838.

104. Jiang HX, Tang D, Liu YH, Zhang XH, Zeng ZL, Xu L, Hawkey PM. 2012. Prevalence and characteristics of β-lactamase and plasmid-mediated quinolone resistance genes in Escherichia coli isolated from farmed fish in China. J. Antimicrob. Chemother. 67:2350–2353.

105. Park YK, Nho SW, Shin GW, Park SB, Jang HB, Cha IS, Ha MA, Kim YR, Dalvi RS, Kang BJ, Jung TS. 2009. Antibiotic susceptibility and resistance of Streptococcus iniae and Streptococcus parauberis isolated from olive flounder (Paralichthys olivaceus). Vet. Microbiol. 136:76–81.

106. Nawaz M, Khan AA, Khan S, Sung K, Steele R. 2008. Isolation and characterization of tetracycline-resistant Citrobacter spp. from catfish. Food Microbiol. 25:85–91.

107. Kim S-R, Nonaka L, Suzuki S. 2004. Occurrence of tetracycline resistance genes tet(M) and tet(S) in bacteria from marine aquaculture sites. FEMS Microbiol. Lett. 237:147–156.

108. Little DC, Edwards P. 1999. Alternative strategies for livestock-fish integration with emphasis on Asia. Ambio 28:118–124.

109. Mathias JA, Charles AT, Baotong H. 1997. Integrated Fish Farming. CRC Press, New York.

107

110. Petersen A, Dalsgaard A. 2003. Antimicrobial resistance of intestinal Aeromonas spp. and Enterococcus spp. in fish cultured in integrated broiler-fish farms in Thailand. Aquaculture 219:71–82.

111. Petersen a., Andersen JS, Kaewmak T, Somsiri T, Dalsgaard a. 2002. Impact of integrated fish farming on antimicrobial resistance in a pond environment. Appl. Environ. Microbiol. 68:6036–6042.

112. Dang STT, Petersen A, Van Truong D, Chu HTT, Dalsgaard A. 2011. Impact of medicated feed on the development of antimicrobial resistance in bacteria at integrated pig-fish farms in Vietnam. Appl. Environ. Microbiol. 77:4494–4498.

113. Petersen A, Dalsgaard A. 2003. Species composition and antimicrobial resistance genes of Enterococcus spp., isolated from integrated and traditional fish farms in Thailand. Environ. Microbiol. 5:395–402.

114. Schmidt AS, Bruun MS, Dalsgaard I, Larsen JL. 2000. Occurrence of antimicrobial resistance in fish-pathogenic and environmental bacteria associated with four Danish Rainbow Trout farms. Appl. Environ. Microbiol. 66:4908–4915.

115. Goni-Urriza M, Capdepuy M, Arpin C, Raymond N, Caumette P, Quentin C. 2000. Impact of an urban effluent on antibiotic resistance of riverine Enterobacteriaceae and Aeromonas spp. Appl. Environ. Microbiol. 66:125–132.

116. Furushita M, Shiba T, Maeda T, Yahata M, Kaneoka A, Takahashi Y, Torii K, Hasegawa T, Ohta M. 2003. Similarity of tetracycline resistance genes isolated from fish farm bacteria to those from clinical isolates. Appl. Environ. Microbiol. 69:5336–5342.

117. McIntosh D, Cunningham M, Ji B, Fekete FA, Parry EM, Clark SE, Zalinger ZB, Gilg IC, Danner GR, Johnson KA, Beattie M, Ritchie R. 2008. Transferable, multiple antibiotic and mercury resistance in Atlantic Canadian isolates of Aeromonas salmonicida subsp. salmonicida is associated with carriage of an IncA/C plasmid similar to the Salmonella enterica plasmid pSN254. J. Antimicrob. Chemother. 61:1221–1228.

118. Rhodes G, Huys G, Swings J, Mcgann P, Hiney M, Smith P, Pickup RW. 2000. Distribution of oxytetracycline resistance plasmids between Aeromonads in hospital and aquaculture environments : Implication of Tn 1721 in dissemination of the tetracycline resistance determinant Tet A. Appl. Environ. Microbiol. 66:3883–3890.

108

119. Guglielmetti E, Korhonen JM, Heikkinen J, Morelli L, von Wright A. 2009. Transfer of plasmid-mediated resistance to tetracycline in pathogenic bacteria from fish and aquaculture environments. FEMS Microbiol. Lett. 293:28–34.

120. Neela FA, Nonaka L, Rahman MH, Suzuki S. 2009. Transfer of the chromosomally encoded tetracycline resistance gene tet(M) from marine bacteria to Escherichia coli and Enterococcus faecalis. World J. Microbiol. Biotechnol. 25:1095–1101.

121. Kadlec K, von Czapiewski E, Kaspar H, Wallmann J, Michael GB, Steinacker U, Schwarz S. 2011. Molecular basis of sulfonamide and trimethoprim resistance in fish-pathogenic Aeromonas isolates. Appl. Environ. Microbiol. 77:7147–7150.

122. Le TX, Munekage Y, Kato S. 2005. Antibiotic resistance in bacteria from shrimp farming in mangrove areas. Sci. Total Environ. 349:95–105.

123. Xing M, Hou Z, Yuan J, Liu Y, Qu Y, Liu B. 2013. Taxonomic and functional metagenomic profiling of gastrointestinal tract microbiome of the farmed adult turbot (Scophthalmus maximus). FEMS Microbiol. Ecol. 86:432–443.

124. Tang Y, Tao P, Tan J, Mu H, Peng L, Yang D, Tong S, Chen L. 2014. Identification of bacterial community composition in freshwater aquaculture system farming of Litopenaeus vannamei reveals distinct temperature-driven patterns. Int. J. Mol. Sci. 15:13663–13680.

125. Yu Z, Morrison M. 2004. Improved extraction of PCR-quality community DNA from digesta and fecal samples. Biotechniques 36:808–812.

126. Muyzer G, de Waal EC, Uitterlinden AG. 1993. Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl. Environ. Microbiol. 59:695–700.

127. Nadkarni MA, Martin FE, Jacques NA, Hunter N. 2002. Determination of bacterial load by real-time PCR using a broad-range (universal) probe and primers set. Microbiology 148:257–266.

128. Zhang L. 2011. Establishment and development of antibiotic resistant bacteria in host gastrointestinal tract—food, drug, or are We Born with It?. Ohio State University.

129. Lachmayr KL, Kerkhof LJ, Dirienzo a G, Cavanaugh CM, Ford TE. 2009. Quantifying nonspecific TEM beta-lactamase (blaTEM) genes in a wastewater stream. Appl. Environ. Microbiol. 75:203–211.

109

130. Bartoloni A, Bartalesi F, Mantella A, Dell’Amico E, Roselli M, Strohmeyer M, Barahona HG, Barrón VP, Paradisi F, Rossolini GM. 2004. High prevalence of acquired antimicrobial resistance unrelated to heavy antimicrobial consumption. J. Infect. Dis. 189:1291–1294.

131. Stanton TB, Humphrey SB, Stoffregen WC. 2011. Chlortetracycline-resistant intestinal bacteria in organically raised and feral Swine. Appl. Environ. Microbiol. 77:7167–7170.

132. Kazimierczak K a, Scott KP, Kelly D, Aminov RI. 2009. Tetracycline resistome of the organic pig gut. Appl. Environ. Microbiol. 75:1717–1722.

133. Gilliver M, Bennett M, Begon M, Hazel S, Hart C. 1999. Enterobacteria: antibiotic resistance found in wild rodents. Nature 401:233–234.

134. Huang Y, Zhang L, Tiu L, Wang HH. 2014. Profiles of antibiotic resistant bacteria associated with aquaculture production. Manuscr. Prep.

135. Sommer MOA, Dantas G, Church GM. 2009. Functional characterization of the antibiotic resistance reservoir in the human microflora. Science 325:1128–1131.

136. Bhullar K, Waglechner N, Pawlowski A, Koteva K, Banks ED, Johnston MD, Barton HA, Wright GD. 2012. Antibiotic resistance is prevalent in an isolated cave microbiome. PLoS One 7:e34953.

137. Schwarz F V, Perreten V, Teuber M. 2001. Sequence of the 50-kb conjugative multiresistance plasmid pRE25 from Enterococcus faecalis RE25. Plasmid 46:170–187.

138. Rosvoll TCS, Pedersen T, Sletvold H, Johnsen PJ, Sollid JE, Simonsen GS, Jensen LB, Nielsen KM, Sundsfjord A. 2010. PCR-based plasmid typing in Enterococcus faecium strains reveals widely distributed pRE25-, pRUM-, pIP501- and pHTbeta-related replicons associated with glycopeptide resistance and stabilizing toxin-antitoxin systems. FEMS Immunol. Med. Microbiol. 58:254–268.

139. Kerrn MB. 2002. Susceptibility of Danish Escherichia coli strains isolated from urinary tract infections and bacteraemia, and distribution of sul genes conferring sulphonamide resistance. J. Antimicrob. Chemother. 50:513–516.

140. Malhotra-Kumar S, Lammens C, Piessens J, Goossens H. 2005. Multiplex PCR for simultaneous detection of macrolide and tetracycline resistance determinants in streptococci. Antimicrob. Agents Chemother. 49:4798–4800.

110

141. Chung WO, Werckenthin C, Schwarz S, Roberts MC. 1999. Host range of the ermF rRNA methylase gene in bacteria of human and animal origin. J. Antimicrob. Chemother. 43:5–14.

142. Dallenne C, Da Costa A, Decré D, Favier C, Arlet G. 2010. Development of a set of multiplex PCR assays for the detection of genes encoding important beta- lactamases in Enterobacteriaceae. J. Antimicrob. Chemother. 65:490–495.

143. Zhao S, White DG, McDermott PF, Friedman S, English L, Ayers S, Meng J, Maurer JJ, Holland R, Walker RD. 2001. Identification and expression of cephamycinase bla(CMY) genes in Escherichia coli and Salmonella isolates from food animals and ground meat. Antimicrob. Agents Chemother. 45:3647–3650.

144. Mugnaioli C, Luzzaro F, De Luca F, Brigante G, Amicosante G, Rossolini GM. 2005. Dissemination of CTX-M-type extended-spectrum beta-lactamase genes to unusual hosts. J. Clin. Microbiol. 43:4183–4185.

145. Li X, Shi L, Yang W, Li L, Yamasaki S. 2006. New array of aacA4-catB3- dfrA1 gene cassettes and a noncoding cassette from a class-1-integron-positive clinical strain of Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 50:2278–2279.

146. Weisburg WG, Barns SM, Pelletier DA, Lane DJ. 1991. 16S ribosomal DNA amplification for phylogenetic study. J. Bacteriol. 173:697–703.

147. Dagert M, Ehrlich SD. 1979. Prolonged incubation in calcium chloride improves the competence of Escherichia coli cells. Gene 6:23–28.

148. Mcintyre DA, Harlander SK. 1989. Improved electroporation efficiency of intact Lactococcus lactis subsp . lactis cells grown in defined mediat 55:2621–2626.

149. Cahill MM. 1990. Bacterial flora of fishes: A review. Microb. Ecol. 19:21–41.

150. Austin B. 2002. The bacterial microflora of fish. ScientificWorldJournal. 2:558– 572.

151. Norman A, Hansen LH, Sørensen SJ. 2009. Conjugative plasmids: vessels of the communal gene pool. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 364:2275–2289.

152. Alcaide E, Blasco M, Esteve C. 2005. Occurrence of drug-resistant bacteria in two European eel farms occurrence of drug-resistant bacteria in two European eel farms. Appl. Environ. Microbiol. 71:3348–3350.

111

153. Martiny AC, Martiny JBH, Weihe C, Field A, Ellis JC. 2011. Functional metagenomics reveals previously unrecognized diversity of antibiotic resistance genes in gulls. Front. Microbiol. 2:238.

154. Zhou W, Wang Y, Lin J. 2012. Functional cloning and characterization of antibiotic resistance genes from the chicken gut microbiome. Appl. Environ. Microbiol. 78:3028–3032.

155. Allen HK, Moe LA, Rodbumrer J, Gaarder A, Handelsman J. 2009. Functional metagenomics reveals diverse beta-lactamases in a remote Alaskan soil. ISME J. 3:243–251.

156. Sambrook J, Russell DW. 2001. Molecular cloning : a laboratory manual3rd ed. Cold Spring Harbor, N.Y. : Cold Spring Harbor Laboratory.

157. Levy SB, McMurry LM, Barbosa TM, Burdett V, Courvalin P, Hillen W, Roberts MC, Rood JI, Taylor DE. 1999. Nomenclature for new tetracycline resistance determinants nomenclature for new tetracycline resistance determinants. Antimicrob. Agents Chemother. 43:1523–1525.

158. Roberts M. 1996. Tetracycline resistance determinants: mechanisms of action, regulation of expression, genetic mobility, and distribution. FEMS Microbiol. Rev. 19:1–24.

159. Chopra I, Roberts M. 2001. Tetracycline Antibiotics : Mode of Action , Applications , Molecular Biology , and Epidemiology of Bacterial Resistance. Microbiol. Mol. Biol. Rev. 65:232–220.

160. Ramos JL, Martínez-Bueno M, Molina-Henares AJ, Terán W, Watanabe K, Zhang X, Gallegos MT, Brennan R, Tobes R. 2005. The TetR family of transcriptional repressors. Microbiol. Mol. Biol. Rev. 69:326–356.

161. Brown MG, Mitchell EH, Balkwill DL. 2008. Tet 42, a novel tetracycline resistance determinant isolated from deep terrestrial subsurface bacteria. Antimicrob. Agents Chemother. 52:4518–4521.

162. Clifford RJ, Hang J, Riley MC, Onmus-Leone F, Kuschner RA, Lesho EP, Waterman PE. 2012. Complete genome sequence of Providencia stuartii clinical isolate MRSN 2154. J. Bacteriol. 194:3736–3737.

112