Virulence of Flavobacterium psychrophilum, Antimicrobial Treatment and Intestinal Microbiota of Rainbow Trout Oncorhynchus mykiss
by
Maureen Jarau
A Thesis presented to The University of Guelph
In partial fulfilment of requirements for the degree of Doctor of Philosophy in Pathobiology
Guelph, Ontario, Canada
© Maureen Jarau, September, 2018
ABSTRACT
VIRULENCE OF FLAVOBACTERIUM PSYCHROPHILUM, ANTIMICROBIAL TREATMENT AND INTESTINAL MICROBIOTA OF RAINBOW TROUT ONCORHYNCHUS MYKISS
Maureen Jarau Advisors: University of Guelph, 2018 Dr. J. S. Lumsden Dr. J. I. MacInnes
Flavobacterium psychrophilum, the aetiologic agent of bacterial cold-water disease (BCWD), is responsible for significant economic losses in rainbow trout Oncorhynchus mykiss aquaculture.
This thesis provides insights into the a) virulence potential of F. psychrophilum isolates from
Ontario in rainbow trout, b) efficacy of oral treatment in controlling F. psychrophilum infection and c) impact of prophylactic treatment on the intestinal microbiota and subsequent susceptibility to BCWD. Following a preliminary infection trial, a broad range (0 to 63% mortality) of virulence was observed for 21 Ontario F. psychrophilum isolates. The virulent isolate FPG101 produced significant mortality (p < 0.05) at the infection dose of 108 cfu fish-1 compared to the control group. Plate counting rather than the rpoC qPCR assay was found to be more sensitive for detection of splenic F. psychrophilum. A minimum inhibitory concentration test showed little variation in the antimicrobial susceptibility of six virulent isolates of F. psychrophilum. One F. psychrophilum isolate with moderate susceptibility to erythromycin (FPG101) and another isolate with low susceptibility to erythromycin (FPG105) were used in an experimental infection trial. Relative to untreated controls, erythromycin treatment significantly (p < 0.05) reduced mortality of rainbow trout infected with FPG101. Erythromycin treatment was more effective
than florfenicol treatment at splenic bacterial clearance during infection period. Furthermore, using Illumina Miseq sequencing, five bacterial phyla (Proteobacteria, Tenericutes, Spirochaetes,
Fusobacteria and Firmicutes) plus “unclassified Bacteria” were identified as major constituents of the intestinal microbiota in rainbow trout throughout the experimental trial period. Before prophylactic oral treatment, Tenericutes and Proteobacteria were the most abundant phyla;
Mycoplasma was the most abundant genus. No significant change was observed in the relative abundance of intestinal bacterial phyla after 10-day oral treatment with erythromycin. In contrast, oral treatment with florfenicol significantly increased (p < 0.05) Proteobacteria;
Sphingomonas was the most abundant bacterial genus. Neither treatment increased the susceptibility of fish to subsequent F. psychrophilum infection.
ACKNOWLEDGEMENTS
I would like to express my infinite gratitude to my advisors, Dr. John Lumsden and Dr.
Janet MacInnes, for their guidance, moral support, patience and generosity that they have extended towards me from the very beginning of my studies. Without their contributions and supervision, this thesis would simply not have happened. I would also like to thank my advisory committee, Dr. Roselynn Stevenson, who also has played an important role in assisting my success as a PhD candidate.
I would like to thank the financial supporters of my program, the Ontario Veterinary
College (OVC) Scholarship, the research funding from the Natural Sciences and Engineering
Research Council (NSERC), and the Ontario Ministry of Agriculture, Food, and Rural Affairs
(OMAFRA) awarded to Dr. John Lumsden in support of my PhD research.
I would like to extend my personal gratitude to the members of Fish Pathology
Laboratory (Lumsden’s laboratory), past and present, for their support and friendship: Paul
Huber, Dr. Elena Contador, Dr. Lowia Al-Housinee, Dr. Ehab Misk, Jessica Grice, Ryan
Horricks, Dr. Juan-Ting Liu, Dr. Jaramar Balmori-Cedeño, Paige Vroom, Dr. Doran Kirkbright and Ryan Eagleson. Many thanks to Donna Kangas, Karla De Uslar, Cathy Bernardi and Marni
Stryuk, for their administrative supports.
I also would like to express my special gratitude to the staff of Hagen Aqualab of
University of Guelph, especially Matthew Cornish and Mike Davis, for their technical assistance and support during my experimental infection trial. Thank you also to Jeffery Gross from the
Advanced Analytic Centre, for his technical expertise and assistance with MiSeq Illumina sequencing.
iv Last but not least, I would like to express my appreciation to my family for their unconditional love and support on this journey. I would like to thank my best friend, Dr. Jacklyn
Ng for her continuous support, understanding and friendship. I would like to give my special thanks to my boyfriend, Dr. Mark Gosselink, who has been very supportive in every possible way, with all the love, encouragement, patience and understanding throughout this thesis. Thank you for your love, support and friendship through it all.
v DECLARATION OF WORK PERFORMED
All work in this thesis was performed by myself, with the following exceptions:
1. The 16S rRNA partial sequencing (Chapter 2) was done by the Animal Health
Laboratory, University of Guelph (Guelph, Ontario).
2. The antimicrobial residue analyses of fish and feed samples (Chapters 3 and 4) were
performed by Eurofins Analytical Laboratories Inc. (New Orleans, Louisiana).
3. MiSeq Illumina Next-generation sequencing (NGS) (Chapter 4) was done at the
Advanced Analysis Centre- Genomics Facility at the University of Guelph (Guelph,
Ontario).
vi TABLE OF CONTENTS
ABSTRACT ...... ii
ACKNOWLEDGEMENTS ...... iv
DECLARATION OF WORK PERFORMED ...... vi
TABLE OF CONTENTS ...... vii
LIST OF TABLES ...... x
LIST OF FIGURES ...... xi
LIST OF SYMBOLS AND ABBREVIATIONS ...... xv
CHAPTER 1: Literature Review ...... 1 1.1 General Introduction ...... 1 1.2 Flavobacterium psychrophilum ...... 2 1.2.1 Initial discovery of F. psychrophilum ...... 2 1.2.2 Taxonomy and classification ...... 3 1.2.3 Growth, phenotypic and biochemical characteristics ...... 4 1.2.4 Genomic characteristics ...... 6 1.2.5 Virulence factors ...... 8 1.3 Disease Associated with Flavobacterium psychrophilum ...... 12 1.3.1 Bacterial cold-water disease (BCWD) ...... 13 1.3.2 Rainbow trout fry syndrome (RTFS) ...... 14 1.4 Clinical Signs and Pathology of F. psychrophilum Infection ...... 14 1.5 Diagnosis of F. psychrophilum Infections ...... 17 1.5.1 Primary isolation and identification ...... 17 1.5.2 Serodiagnosis ...... 18 1.5.3 Molecular techniques ...... 19 1.6 Prevention, Control and Treatment ...... 20 1.6.1 Management practices ...... 20 1.6.2 Vaccine development ...... 21 1.6.3 Antimicrobial therapy ...... 22
vii 1.6.4 Disinfection ...... 24 1.7 Microbiota in Fish Intestines ...... 25 1.8 Rationale, Hypotheses and Objectives ...... 26
CHAPTER 2: Virulence of Flavobacterium psychrophilum Isolates in Rainbow Trout Oncorhynchus Mykiss (Walbaum) ...... 29 2.1 Abstract ...... 29 2.2 Introduction ...... 29 2.3 Materials and Methods ...... 32 2.3.1 Bacterial isolates and culture conditions ...... 32 2.3.2 Bacterial suspension preparation ...... 33 2.3.3 Fish and rearing conditions ...... 33 2.3.4 Experimental infection trials ...... 34 2.3.5 Histology ...... 35 2.3.6 Bacterial isolation and identification ...... 36 2.3.7 DNA extraction ...... 36 2.3.8 16S rRNA PCR ...... 37 2.3.9 Splenic bacterial load ...... 39 2.3.10 Statistical analysis ...... 41 2.4 Results ...... 41 2.4.1 Experimental infection ...... 41 2.4.2 Enumeration of F. psychrophilum in fish spleen ...... 48 2.4.3 Histopathology ...... 56 2.5 Discussion ...... 60
CHAPTER 3: Erythromycin and Florfenicol Treatment of Rainbow Trout Oncorhynchus mykiss (Walbaum) Experimentally Infected with Flavobacterium psychrophilum ...... 64 3.1 Abstract ...... 64 3.2 Introduction ...... 65 3.3 Materials and Methods ...... 67 3.3.1 In vitro susceptibility testing ...... 67 3.3.2 In vivo oral antibiotic treatment ...... 69 3.3.3 Splenic DNA extraction and enumeration of splenic bacterial load using qPCR ..... 75
viii 3.3.4 Statistical analysis ...... 76 3.4 Results ...... 76 3.4.1 In vitro susceptibility test ...... 76 3.4.2 In vivo oral antibiotic treatment ...... 79 3.4.3 Splenic bacterial load ...... 83 3.4.4 Antibiotic concentration analyses ...... 89 3.5 Discussion ...... 89
CHAPTER 4: Effect of Prophylactic Treatment with Erythromycin and Florfenicol on the Intestinal Microbiota of Rainbow Trout (Oncorhynchus mykiss) Juveniles ...... 94 4.1 Abstract ...... 94 4.2 Introduction ...... 95 4.3 Materials and Methods ...... 96 4.3.1 Fish and rearing conditions ...... 96 4.3.2 Preparation of medicated feed ...... 97 4.3.3 Preparation of bacterial inoculum ...... 98 4.3.4 Experimental design ...... 98 4.3.5 Sample collections ...... 102 4.3.6 Microbiota study ...... 103 4.3.7 Enumeration of splenic bacterial load ...... 106 4.3.8 Statistical analysis ...... 107 4.4 Results ...... 107 4.4.1 Characterization of intestinal microbiota ...... 107 4.4.2 Diversity of rainbow trout intestinal microbiota ...... 110 4.4.3 Flavobacterium psychrophilum infection ...... 130 4.4.4 Antimicrobial concentration analysis ...... 134 4.5 Discussion ...... 136
CHAPTER 5: General Discussion and Conclusions ...... 143
REFERENCES ...... 156
ix LIST OF TABLES
Table 2.1 Primers and probe used in this study...... 38
Table 2.2 Percent mortality of fish groups infected with F. psychrophilum strains in Trial 1. 45
Table 3.1 Details of fish groups used...... 72
Table 3.2 Minimum inhibition concentration values of six F. psychrophilum isolates determined using the broth microdilution method...... 78
Table 3.3 Number of fish detected positive with F. psychrophilum as determined using qPCR.84
Table 4.1 Fish groups and intestinal contents sampling...... 101
Table 4.2 Relative abundances (≥ 0.1%) of bacterial phyla found in rainbow trout intestinal microbiota in antibiotic-treated and untreated groups before and after challenge with F. psychrophilum...... 111
Table 4.3 Relative abundances of the 18 most abundant bacterial genera found in rainbow trout intestinal microbiota in antibiotic-treated and untreated groups before (D-11 and D0) and after (D12 and D24) challenge with F. psychrophilum...... 112
Table 4.4 Bacterial OTUs in the microbiota in both groups F (florfenicol-treated) and E (erythromycin-treated) before (-D11 p.i.) and after (D0 p.i.) prophylactic oral antibiotic treatment...... 120
Table 4.5 Antimicrobial concentration analysis of medicated feed samples...... 135
x LIST OF FIGURES
Figure 2.1 Gross lesions from an affected rainbow trout after experimental i.p. injection with F. psychrophilum during Trial 1. There was a pale kidney and liver, splenomegaly and hemorrhaging of the serosal surfaces, peritoneal cavity and muscles near the site of the injection...... 43
Figure 2.2 Cumulative survival of rainbow trout following i.p. injection with F. psychrophilum in Trial 2. Duplicate groups of rainbow trout were injected with FPG101 at low (106, square), medium (107, triangle) and high (108, cross) doses compared to one sham-injected control group. Mortality was first observed on Day 5 p.i. A significant difference (p < 0.001) was observed between the high (108) and the low (106) dose injected group. No mortalities occurred in the control fish group...... 47
Figure 2.3 Standard curve of rpoC qPCR assay used in this study. Each point represents the mean of threshold cycle (Ct) values of nine F. psychrophilum standards prepared in triplicate. The standards were between 2 x 100 and 2 x 108 rpoC copies reaction-1. Error bars denote standard deviation from triplicate samples...... 49
Figure 2.4 Number of fish detected positive for splenic cfus and rpoC qPCR throughout the infection Trial 2. At most time points, more fish were positive by culture than by qPCR assay. Flavobacterium psychrophilum could be detected by both methods on Day 3 p.i. although no clinical signs were exhibited in these fish...... 51
Figure 2.5 Splenic bacterial count of asymptomatic (×, cross) and moribund (O, circle) fish from Trial 2 (108 cfu fish-1 injection group) assayed by rpoC qPCR. A few asymptomatic fish had moderate bacterial loads, typically during early infection. There was a significant difference (p < 0.05) in splenic bacterial loads between moribund and asymptomatic fish over time. Darker symbols indicate different samples with the same bacterial load...... 53
Figure 2.6 Correlation of F. psychrophilum load determined by an rpoC qPCR assay and viable colony counting (cfu) of 96 spleen samples from infected fish group from Trial 2. Linear -1 regression log10 (cfu) = -1.99 + 1.84 (log10 rpoC copies reaction ) is shown by the straight line. There was a very strong positive correlation between the two counting methods (R2 = 0.915, p < 0.05)...... 55
Figure 2.7 Light microscopy of rainbow trout spleen sampled in Trial 1. A) Normal spleen from fish sampled on Day 0 with expected organization of red and white pulp; splenic capsule and ellipsoids are fully intact. B) Spleen of affected fish sampled on D12 p.i. with rarefaction of splenic parenchyma, extensive hemorrhage (blue arrow) and necrosis (black arrow). (H&E. Bar = 200 µm) ...... 57
Figure 2.8 Section of epidermis and peritoneum from close to the injection site of an infected rainbow trout. There was epidermal ulceration, attenuation of the dermis, marked inflammation
xi (black arrow) of the peritoneal adipose tissue and large numbers of bacteria (blue arrow) noted. (H&E. Bar = 200 µm) ...... 59
Figure 3.1 Experimental timeline...... 71
Figure 3.2 Mean cumulative survival of triplicate groups of fish infected with F. psychrophilum FPG101 and untreated (UT1) or treated orally with erythromycin (ERY1), florfenicol (FFN1). Survival was significantly higher (p < 0.05) in groups of fish treated with erythromycin compared to untreated groups. Vertical dashed line indicates initiation of FFN and ERY oral treatment (Day 8 p.i.)...... 80
Figure 3.3 Mean cumulative survival triplicate groups of fish infected with F. psychrophilum FPG105 and untreated (UT2) or treated orally with erythromycin (ERY2) or florfenicol (FFN2). No significant difference (p > 0.05) was observed in fish survival rate between groups. Vertical dashed line indicates initiation of FFN and ERY oral treatment (Day 8 p.i.)...... 82
-1 Figure 3.4 Splenic F. psychrophilum FPG101 load (mean ± SEM log10 rpoC copies reaction ) detected from positive fish. Increase in the bacterial loads in untreated (UT1) and florfenicol- treated (FFN1) group from Day 5 p.i. to Day 12 p.i. were observed; however, there were statistically insignificant. There was a significant increase in the bacterial load in the erythromycin-treated (ERY1) group from Day 5 to Day 9 p.i. and a significant reduction observed from Day 9 to Day 12 p.i. At Day 12 p.i., erythromycin-treated fish had significantly less F. psychrophilum than both untreated or florfenicol-treated groups of fish. Significance for all comparisons was p < 0.05. Error bars denote standard error mean from samples in triplicate...... 86
-1 Figure 3.5 Splenic F. psychrophilum FPG105 load (mean ± SEM log10 rpoC copies reaction ) detected positive from infected fish groups. There was a significant increase in bacterial load in untreated (UT2) and florfenicol-treated (FFN2) group and on Day 9 p.i. compared to Day 5 p.i. On Day 12 p.i., both of these groups had significant reductions in splenic bacterial count compared to Day 9 p.i. There was no significance change in splenic bacterial loads of fish treated with erythromycin (ERY2) at any time. Significance for all comparisons was p < 0.05. Error bars denote standard error mean from samples in triplicate...... 88
Figure 4.1 Experimental design...... 100
Figure 4.2 Rarefaction curves comparing the number of OTUs (clustered at 97% similarity) with the number of sequences found in the intestinal contents of fish groups sampled throughout the trial...... 109
Figure 4.3 Relative abundance of the top phyla (≥ 0.1%) found in samples of intestinal contents collected from groups before (-D11 p.i.) and after prophylactic oral antibiotic treatment (D0 p.i.). a) Before oral treatment, Tenericutes and Proteobacteria were the most abundant phyla in all groups. b) After oral treatment, Proteobacteria was the most abundant phylum in all groups
xii except for group Cb (untreated & infected). Tenericutes was the most predominant phyla in group Cb. Groups Ca & Cb = untreated, F = florfenicol-treated, E = erythromycin-treated. 114
Figure 4.4 Relative abundance of the top 10 bacterial genera found in samples of intestinal contents collected from groups before (-D11 p.i.) and after prophylactic oral antibiotic treatment (D0 p.i.). a) Before oral treatment, Mycoplasma was the most abundant genus in all groups. Fish group Ca (untreated) had the most diverse community compared to groups Cb (untreated); F (florfenicol-treated); and E (erythromycin-treated). b) After oral treatment, group Ca had the most diverse community followed by group Cb; group E and F. Sphingomonas mostly occupying the intestines of group F while Mycoplasma and Methylobacterium were the most abundant genera in group E...... 117
Figure 4.5 Numbers of OTUs in rainbow trout intestinal microbiota before (-D11 p.i.) and after (D0 p.i.) prophylactic oral antibiotic treatment. a) Group F (florfenicol-treated): total OTUs richness was 194 and total number of OTUs shared by groups before (F_n11) and after (F_0) treatment was 23. b) Group E (erythromycin-treated): total OTUs richness is 346 and total number of OTUs shared by groups before (E_n11) and after (E_0) treatment was 16...... 119
Figure 4.6 Non-metric dimensional scaling (NMDS) plots illustrating the intestinal microbiota beta-diversity of a) F (florfenicol-treated & infected) and b) E (erythromycin-treated & infected). Each circle represents an individual intestinal contents sample based on subsample of 2,962 OTUs...... 123
Figure 4.7 Relative abundance of the top phyla (≥ 0.1%) found in samples of intestinal contents collected from groups after F. psychrophilum infection (D12 & D24 p.i.). a) On D12 p.i., Proteobacteria was predominant in groups F (florfenicol-treated & infected), E (erythromycin- treated & infected) and Ca (control). b) On D24 p.i., an increase of Tenericutes was observed in all groups. Proteobacteria remained the most abundant phyla in groups F, E and Ca...... 125
Figure 4.8 Relative abundance of the top 10 bacterial genera found in rainbow trout samples of intestinal contents collected from groups after F. psychrophilum infection (D12 & D24 p.i.). a) On D12 p.i.: group Ca (control) and E (erythromycin-treated & infected) had more diverse bacterial community compared to group F (florfenicol-treated & infected) and group Cb (untreated & infected). b) On D24 p.i.: groups Ca and E had the most diverse community followed by group F and group Cb. Acinetobacter and Pseudomonas were the most abundant phyla in group F while Mycoplasma was the most abundant phyla occupying the intestines of groups Ca, Cb and E...... 129
Figure 4.9 Cumulative survival curves of groups following i.p. infection with F. psychrophilum after prophylactic oral treatment with ERY and FFN. Triplicate groups of 2 oral treated (F: florfenicol-treated; E: erythromycin-treated) and one untreated (Cb) rainbow trout were injected with FPG101 at 108 cfu fish-1 compared to triplicate mock-infected control group (Ca). Mortality was first observed on D6 p.i. There was no significant difference (p > 0.05) between survival curves of all fish groups. Group E had the highest percent survival followed by group F and group Cb. No mortalities occurred in the control fish group Ca...... 131
xiii -1 Figure 4.10 Splenic F. psychrophilum loads (mean ± SEM log10 rpoC copies reaction ) of infected fish randomly sampled at 4 time points (D3, D6, D9 & D12 p.i.) as determined using qPCR. Error bars denote the standard error of the mean. On D3 p.i., bacterial loads ranging from -1 1.61-1.80 log10 rpoC copies reaction were detected in all groups. There was a significant increase (p < 0.05) in bacterial loads in F (florfenicol-treated & infected) group on D6 to D12 p.i. and D9 p.i. to D12 p.i. There was no significance change (p > 0.05) in splenic bacterial loads of fish groups E (erythromycin-treated & infected) and Cb (untreated & infected) at any time. ** refers to statistically significant as p < 0.05...... 133
xiv LIST OF SYMBOLS AND ABBREVIATIONS
% Percentage ˚C Degree Celsius α Alpha ß Beta γ Gamma µ Micron µg Microgram(s) µg L-1 Microgram per litre µL Microlitre(s) µg µL-1 Microgram per microlitre µM Micromolar A Adenine or adenosine A260 Absorbance at 260 nm A280 Absorbance at 280 nm ANOVA Analysis of variance ATCC American Type Culture Collection BCWD Bacterial cold-water disease bp Base pairs C Cytosine or cystidine CA Cytophaga agar CAMHB Cation-adjusted Muller-Hinton broth CB Cytophaga broth cfu Colony-forming unit(s) d Day(s) dNTPs Deoxyribonucleotide triphospates DNA Deoxyribonucleic acid EDTA Ethylenediaminetetraacetic acid g Gram(s) G Guanine h Hour(s) H & E Hematoxylin and eosin HPLC High performance liquid chromatography i.p. Intraperitoneal i.m. Intramascular kg Kilogram(s) L Litre(s) LC-MS/MS Liquid chromatography tandem-mass spectrophotometry L min-1 Litre per minute M Molar Mg Megagram mg Milligram(s) MgCl Magnesium chloride mg L-1 Milligram per liter MIC Minimum inhibitory concentration(s)
xv Min Minute(s) mL Milliliter mM Millimolar Mo Month(s) MW Molecular weight ng Nanogram NMDS Non-metric dimentional scaling OD Optical density ON Ontario OTUs Operational taxonomic units OVC Ontario Veterinary College p Probability PBS Phosphate buffered saline PCR Polymerase chain reaction pH Puissance hydrogen (Hydrogen ion concentration) p.i. Post-infection ppm Parts per million QC Quebec qPCR Quantitative polymerase chain reaction ROS Reactive oxygen species RNA Ribonucleic acid rRNA Ribosomal ribonuclease acid rpm Round per minute rpoC ß´ DNA-dependent RNA polymerase RTFS Rainbow trout fry syndrome s Second(s) SD Standard deviation SEM Standard error mean TBE Tris-base EDTA buffer v/v Volume per volume wk Week(s) w/v Weight per volume
xvi CHAPTER 1: Literature Review
1.1 General Introduction
Oncorhynchus mykiss (rainbow trout) is the most common trout species farmed in
Ontario and Ontario is the largest rainbow trout producer in Canada. According to Aquastats
2016, Ontario produced approximately 5,060 tonnes of rainbow trout with a production value of
$26.8 M (Moccia and Bevan, 2017). Although rainbow trout are hardy and relatively easy to culture, infectious disease remains an economic concern.
Bacterial cold-water disease (BCWD) is one of the most common diseases affecting rainbow trout and is the most significant cause of mortality and decreased growth of fish production in farmed rainbow trout in Ontario (Lumsden et al., 2004). The disease, caused by
Flavobacterium psychrophilum, occurs at any time of the year, but usually in low water temperatures (4˚C to 15 °C) (Holt, 1987).
Control of the disease in fish farms involves good management practices such as minimizing fish movement and handling (Groff and LaPatra, 2000; LaFrentz and Cain, 2004;
Cipriano and Holt, 2005). Since no commercial vaccines are currently available, antimicrobial therapy has been widely used to control and to prevent the BCWD outbreaks. Of the four antimicrobials currently available for use in Ontario (florfenicol, oxytetracycline, ormetoprim- sulfadimethoxine, trimethoprim-sulfamethoxazole), florfenicol is the agent of first choice for controlling BCWD in the fish farms (J. S. Lumsden, personal communication). However, there are significant concerns regarding the development of antimicrobial resistance, particularly given the lack of alternative antimicrobials.
1 Fish intestinal microbiota plays an important role in the health and the development of the immune system in fish and factors such as age, environment, diet, stress and infectious disease have been associated to the alteration of its composition (Llewellyn et al., 2014). During
BCWD outbreaks, continuous antimicrobial therapy as well as various stress factors likely impact the microbiota of the affected fish. However, to date, there have been no reports on the intestinal microbiota of rainbow trout with BCWD and studies on the impact of F. psychrophilum infection on intestinal microbiota of rainbow trout are needed.
1.2 Flavobacterium psychrophilum
Flavobacterium psychrophilum is a ubiquitous bacterium in the freshwater aquatic environment (Groff and LaPatra, 2000). Like two other species of the genus Flavobacterium, F. branchiophilum and F. columnare, F. psychrophilum can be a primary pathogen of freshwater- reared and wild fish populations (Bernardet and Nakagawa, 2006).
1.2.1 Initial discovery of F. psychrophilum
The first description of F. psychrophilum was reported by Davis (1946) following recurring epizootic events at a hatchery in West Virginia, USA, affecting rainbow trout fingerlings in 1941 and 1945. Although he was unable to isolate the causative agent, Davis
(1946) observed masses of bacterial rods in scrapings from the caudal peduncle lesions of affected fish.
Later, in 1948, F. psychrophilum was isolated for the first time from the kidney and external lesions of a diseased coho salmon (Oncorhynchus kisutch) in Washington state, USA
(Borg, 1948). The pathological findings in diseased fish and the morphological characteristics of
2 the suspected causative bacteria were similar to that described by Davis (1946). Borg (1948) further described the pathogen as a Gram-negative bacterium with gliding motility and an absence of fruiting bodies or microcysts. He also observed that the bacterium was unable to grow at temperatures higher than 25 ˚C.
Following the initial discovery of F. psychrophilum, numerous occurrences of disease associated with this bacterium have been reported worldwide. Moreover, Bernardet and Bowman
(2006) suggested that this freshwater fish pathogen very likely occurred much earlier in many areas. They suspected that it was unnoticed due to the fact that F. psychrophilum did not appear to be responsible for serious outbreaks before intensive aquaculture was put into practice and because the appropriate media and temperature for its isolation were not commonly used before its first isolation in 1948.
1.2.2 Taxonomy and classification
There has been considerable debate about the taxonomy and classification of F. psychrophilum since its first isolation by Borg (Bader and Shotts 1998, 2001; Bernardet, 2001).
This bacterium was initially placed in the order Myxobacterales and named Cytophaga psychrophila based on its phenotypic characteristics and biochemical analysis (Borg, 1948,
1960). The name Cytophaga psychrophila remained in use until Bernardet and Grimont (1989) performed DNA homology studies. From this work, they proposed that the bacterium be reassigned to the order Cytophagales as a member of the genus Flexibacter and renamed the bacterium Flexibacter psychrophilus.
In the following years, various genomic and proteomic studies were conducted on
Flexibacter psychrophilus. In 1996, Bernardet et al. suggested that Flexibacter psychrophilus
3 should be renamed Flavobacterium psychrophilum based on their studies on the guanine plus cytosine (G+C) composition of genomic DNA, DNA-rRNA hybridization, and fatty acid and protein profiles of the bacterium (Bernardet et al., 1996). This change in nomenclature remains widely accepted worldwide. Flavobacterium psychrophilum was subsequently confirmed to belong to the family Flavobacteriaceae in the Cytophaga-Flavobacterium-Bacteriodes group based on its genomic and proteomic characteristics (Bernardet and Bowman 2006; Bernardet,
2011).
1.2.3 Growth, phenotypic and biochemical characteristics
Due to its slow growth and fastidious nature, culture of F. psychrophilum from clinical samples can be easily suppressed by many fast-growing and less fastidious bacteria and fungi
(Wiklund et al., 2000). Selective culture media with specific low nutrient levels such as cytophaga agar (CA; Anacker and Ordal, 1959), Shieh’s medium (Shieh, 1980); Hsu-Shotts medium (Bullock et al., 1986), tryptone-yeast extract salts medium (TYES; Holt, 1987), and modified CA medium (Daskalov et al., 1999) have been developed to support the growth of F. psychrophilum. The addition of antimicrobial agents to these media suppresses the growth of less fastidious bacteria and fungi found in clinical samples thus improving their selectivity for F. psychrophilum (Schmidt et al., 2000; Nematollahi et al., 2003b).
Flavobacterium psychrophilum is a strictly aerobic psychrophilic bacterium that grows well at temperatures between 4 ˚C and 25 ˚C (Holt, 1987; Bernardet and Kerouault, 1989;
Ostland et al., 1997; Madetoja et al., 2001). Optimum growth in vitro occurs from 15 ˚C to 20 ˚C
(Holt et al., 1989; Uddin and Wakabayashi, 1997) with the shortest generation time (~2 h) at 15
4 ˚C (Holt et al., 1993). The ability to grow at 25 ˚C varies with different strains (Lorenzen et al.,
1997; Madetoja et al., 2001; Cepeda et al., 2004).
Agar cultures of F. psychrophilum exhibit raised glossy, non-diffusible yellow-, flexirubin-type pigmented colonies with a diameter of approximately 1 to 5 mm. These colonies do not adhere to the surface of the medium (Bernardet and Bowman, 2006) and usually have a thin, spreading margin (Pacha, 1968). Pure cultures of F. psychrophilum strains can display both spreading and non-spreading phenotypes on the same plate (Pacha, 1968; Bernardet and
Bowman, 2006). Two distinct colony phenotypes: S (smooth, hydrophobic, adhesive (auto- agglutinating)) and R (rough, hydrophilic (non-agglutinating)) have been described (Högfors-
Rönnholm and Wiklund, 2010), and both types can be isolated from the same BCWD outbreak
(Sundell et al., 2013).
Flavobacterium psychrophilum cells are Gram-negative, weakly refractile, slender rods with rounded ends (Holt, 1987; Rangdale and Way, 1995; Madetoja et al., 2001). The size of the bacterium is approximately 0.3 to 0.75 µm in diameter and between 2 to 7 µm in length (Pacha,
1968; Bernardet and Grimont, 1989). Flavobacterium psychrophilum exhibits poor gliding motility (Pacha, 1968; Bernardet and Grimont, 1989) which does not appear to involve pili or flagella (Hunnicutt and McBride, 2000; Duchaud et al., 2007). The extent of gliding motility varies substantially among strains (Álvarez et al., 2006).
Flavobacterium psychrophilum produces flexirubin-pigment; is weakly oxidase and catalase positive; and has no ability to hydrolyze esculin, starch, chitin or xanthine (Bernardet and Kerouault, 1989; Lorenzen and Olesen, 1997; Madetoja et al., 2001). It does not reduce nitrite to nitrate or produce indole or lysine (Pacha, 1968; Nematollahi et al., 2003b). All strains are able to degrade chondroitin sulfate (Otis, 1984; Bernardet and Grimont, 1989; Holt et al.,
5 1993; Bertolini et al., 1994) while carbohydrate degrading ability is strain dependent (Bernardet and Kerouault, 1989; Lorenzen et al., 1995; Lorenzen and Olesen, 1997; Austin and Austin,
2007; Hesami et al., 2008). As well, F. psychrophilum is highly proteolytic and can hydrolyze a wide range of substrates including gelatin, casein, albumin, fibrinogen, collagen and hemoglobin
(Bernardet and Kerouault, 1989; Bertolini et al., 1994; Dalsgaard and Madsen, 2000). The ability to degrade elastin and to produce hydrogen sulfide (H2S) varies among strains (Holt et al., 1993;
Bertolini et al., 1994; Dalsgaard and Madsen, 2000; Madetoja et al., 2001; Nematollahi et al.,
2003b).
1.2.4 Genomic characteristics
Duchaud et al. (2007) reported the first complete genome sequence of F. psychrophilum,
JIP02/86 (ATCC 49511). The JIP02/86 genome is circular and has a G+C content of 32.5%. It is
2.86 Mb in size, has six 16S rRNA operons and has 2,432 putative protein-coding genes including those predicted to be involved in the colonization, invasion and destruction of host tissues. Since this initial report, additional isolates have been sequenced (Wiens et al., 2014; Wu et al., 2015; Castillo et al., 2016; Neiger et al., 2016; Shimizu et al., 2016; Rochat et al., 2017).
As of June 2018, there are 67 genome assemblies available at the National Center for
Biotechnology Information (NCBI) website; however, of the 67 genomes deposited, only nine are complete. The average length of the nine complete F. psychrophilum genomes is 2.76 Mb, with an average protein count and G+C content of 2,473 and 32.5% respectively.
A wide range of techniques have been used to characterize F. psychrophilum strains including: PCR-restriction fragment length polymorphism (PCR-RFLP) (Izumi et al., 2003;
Soule et al., 2005; Izumi et al., 2007), random amplification of polymorphic DNA (RAPD)
6 (Crump et al., 2001; Chakroun et al., 1997), ribotyping (Cipriano et al., 1996), pulsed-field gel electrophoresis (PFGE) (Arai et al., 2007), multiplex PCR-based serotyping, plasmid profiling
(Madsen and Dalsgaard, 2000; Lorenzen et al., 1997; Chakroun et al., 1998), and multilocus sequence typing MLST (Nicolas et al., 2008; Strepparava et al., 2013; Avendaño-Herrera et al.,
2014; Nilsen et al., 2014). Early genetic studies indicated that F. psychrophilum isolates obtained from a variety of host species and different continents were highly homogenous (Bernardet and
Kerouault, 1989; Nicolas et al., 2008; Valdebenito and Avendaño-Herrera, 2009). Later studies reported genetic heterogeneity among F. psychrophilum isolates. For example, strong associations between F. psychrophilum ribotypes (Chakroun et al., 1998) as well as F. psychrophilum genotypes (Izumi et al., 2003) and host of origin have been reported. Chen et al.
(2008) reported that the population structure of F. psychrophilum isolated from spawning coho salmon (Oncorhynchus kisutch) was more genetically diverse than that of farmed rainbow trout while Madetoja et al. (2001) reported that 37 isolates F. psychrophilum isolated from Finland comprised 13 different ribotypes. Hesami et al. (2008) reported four distinct restriction pattern types and nine nucleotide sequence types among 75 F. psychrophilum isolates collected from mortality events in Ontario salmonids. Recently, based on phylogenetic analysis, Loch et al.
(2013) reported that clinical F. psychrophilum isolates recovered from Michigan, USA consisted of two distinct and well-supported F. psychrophilum genotypes: i) cluster XIIIa of isolates that were recovered predominantly from feral-spawning Chinook salmon (Oncorhynchus tshawytscha) and, ii) cluster XIIIb of isolates recovered predominantly from hatchery-reared
Atlantic salmon (Salmo salar).
7 1.2.5 Virulence factors
1.2.5.1 Adhesion and colonization
Bacterial adhesion mechanisms are important for colonization and disease. Little is known about adhesins of F. psychrophilum. This bacterium is reported to be able to adhere to fish skin mucus (Kondo et al., 2002; Högfors-Rönnholm et al., 2015) as well as other mucosal tissues such as fins, gills, skin, and eyes (Kondo et al., 2002; Nematollahi et al., 2003b;
Papadopoulou et al., 2017). Early scanning electron microscopy studies by Kondo et al. (2002) showed that F. psychrophilum cells were able to attach to lower jaw and caudal peduncle after destruction of the epidermis tissue but the mechanism of attachment was not elucidated. A highly virulent strain (Strain Dubois) of F. psychrophilum was reported to attach better to rainbow trout gill tissue compared to a less virulent strain (Strain 99/10A) and the degree of adhesion may have been influenced by environmental factors such as water quality and water temperature
(Nematollahi et al., 2003a) but again, the mechanisms of attachment were not elucidated in these studies. A recent study by Högfors-Rönnholm et al. (2015) suggested that smooth cells
(hydrophobic, adhesive, auto-agglutinating) of F. psychrophilum are more likely to adhere much sooner to abiotic surfaces than rough cells (hydrophilic, non-agglutinating) in vitro; however, very little is known about the adhesion mechanisms of either cell phenotype.
Based on analysis of the genome, Duchaud et al. (2007) suggested 27 genes in F. psychrophilum JIP02/86 (ATCC 49511) may be involved in bacterial adhesion. Fifteen of these genes are predicted to encode proteins with leucine-rich repeats similar to BspA and LrrA-- proteins involved in the attachment to human oral tissue by the periodontal pathogenic bacteria
Bacteroides forsythus and Treponema denticola (Sharma et al., 1998; Ikegami et al., 2004).
Other adhesin homologs (i.e., FP1830, FP0016 and FP0595) contain fibronectin type III domains
8 thought to be involved in cell surface binding while FP2413 contains five fibronectin type III domains and a CUB domain (found in extracellular proteins of sperm adhesin and in eukaryotic peptidase families) (Duchaud et al., 2007).
Gliding motility in F. psychrophilum plays an important role in colonization and movement in the host (LaFrentz et al., 2009; LaFrentz et al., 2011; Pérez-Pascual et al., 2017).
The GldN protein, one of several proteins involved in gliding motility of F. psychrophilum, was predicted from the genome sequence by Duchaud et al. (2007) and later demonstrated to be immunogenic in fish (LaFrentz et al., 2009, 2011). Further, it has been suggested that F. psychrophilum uses its proteolytic activity and gliding motility to gain access into fish fin tissue
(Martínez et al., 2004). Gene knock-out studies of two gliding genes (gldD and gldG) in F. psychrophilum strain THCO2-90 showed that disruption of either gene impairs the secretion of extracellular enzymes and adhesins thus reducing the ability of the bacterium to colonize and infect the host (Pérez-Pascual et al., 2017).
Bacteria are generally found in complex surface-associated microbial communities known as biofilms (Huq et al., 2008). A biofilm consists of surface-attached bacteria enclosed in a self-produced extracellular polymeric substance (EPS) matrix (Costerton et al., 1999). Biofilms have been associated with virulence in several pathogenic bacteria. Sundell and Wiklund (2011) showed that F. psychrophilum is able to form biofilms and suggested that mature biofilms (>107
CFU mL-1) are less susceptible to oxytetracycline and flumequine if exposed to subinhibitory concentrations (three times minimum inhibitory concentrations (MICs) and an environment concentration) of each antimicrobial. Álvarez et al. (2006) suggested that gliding motility and biofilm formation are antagonistic properties as a tlpB mutant impaired in virulence and gliding motility exhibited enhanced ability to form biofilms. Other studies have reported positive
9 association between virulence of F. psychrophilum and biofilm formation (Nematollahi et al.,
2003a; Castillo et al., 2015; Högfors-Rönnholm et al., 2015). Most recently, Levipan and
Avendaño-Herrera (2017) investigated the expression of virulence- and biofilm-related genes
(VBRGs) in planktonic and biofilm phases of F. psychrophilum isolates. Based on transcriptomes analysis (RNA-Seq), these authors identified differential expression of putative
VBRGs involved in several functions (i.e., polysaccharide biosynthesis, lateral gene transfer, membrane transport, sensory mechanisms and adhesion). Over 60% of VBRGs involved in these processes were significantly upregulated in the biofilm state (Levipan and Avendaño-Herrera,
2017). Further, Levipan et al. (2018) performed cytotoxicity assays and an experimental infection trial with rainbow trout fry. Their findings were in agreement with the transcriptomic findings of Levipan et al. (2018) and supported that notion that biofilm formation could contribute to the virulence of F. psychrophilum.
1.2.5.2 Immune response inhibitors
Several studies have suggested that host phagocytes play a major role in the outcome of
F. psychrophilum infection in rainbow trout. For example, F. psychrophilum induces the production of reactive oxygen species (ROS) in head kidney phagocytes; although enzymes that protect the bacterium from ROS such as superoxide dismutase (SOD) and catalase activity are also produced. Survival in head kidney phagocytes is considerably poorer due to higher ROS activities compared to those in the spleen (Sanchez-Moreno et al., 1989; Kawai et al., 2000;
Lammens et al., 2000; Nematollahi et al., 2003b). Viable F. psychrophilum have been noted inside splenic phagocytes (Rangdale et al., 1999; Decostere et al., 2000). Splenic macrophages have less antimicrobial action against F. psychrophilum than head kidney macrophages, and survival in this intracellular location protects F. psychrophilum against humoral defense
10 mechanisms such as complement and lysozyme activity (Nematollahi et al., 2005). Many genes encoding enzymes that could counteract deleterious effects of ROS and aid in the resistance of the bacterium to ROS killing have been identified in the F. psychrophilum JIP02/08 genome
(Duchaud et al., 2007).
1.2.5.3 Toxins
Flavobacterium psychrophilum is actively proteolytic and extracellular proteases have been reported to be important virulence factors, which can lead to rapid and massive tissue destruction (Bertolini et al. 1994). Two F. psychrophilum metalloproteases, Fpp1 and Fpp2, have been purified and characterized (Secades et al., 2001, 2003). Both enzymes can degrade tissue proteins in vitro and are greatly induced in the presence of calcium levels similar to those present in fish tissue (Secades et al., 2001, 2003). They were able to degrade gelatin, laminin, fibronectin, fibrinogen, collagen type IV, actin and myosin--all basic elements of the fish muscular system. Recently, a gene knockout study using the wild-type F. psychrophilum
THC02/90 strain showed that the fpp1 and fpp2 genes had no effect on bacterial virulence in a rainbow trout challenge model (Pérez-Pascual et al., 2011). Extracellular collagenase secreted by
F. psychrophilum is thought to aid in the colonization of connective tissues of fish (Ostland et al., 2000). The ability of F. psychrophilum to degrade elastin has also been reported (Madsen and
Dalsgaard, 2000); however, elastinase production varies among F. psychrophilum isolates (Soule et al., 2005; Nilsen et al., 2011).
Bacterial hemolysins have been demonstrated in F. psychrophilum infecting rainbow trout fry and Lorenzen (1994) further suggested that the distinctive anemia developed by infected rainbow trout fry might be due to the ability of the bacterium to partially lyse and agglutinate rainbow trout erythrocytes. Later studies showed that F. psychrophilum strains partially lyse calf
11 and rainbow trout red blood cells and the organism appears to grow best on media supplemented with calf or fish blood cells (Lorenzen et al., 1997).
1.2.5.4 Iron acquisition
The role of iron acquisition mechanisms in promoting F. psychrophilum infections within fish hosts has been reported (Álvarez et al., 2006, 2008). Mutants deficient in growth in iron- limited conditions (to mimic the condition in the host) have been generated by Álvarez et al.
(2006) using a Tn4351-mutagenesis system. For example, F. psychrophilum mutant (FP1033) led to the identification of a TonB system involving ExbB, ExbD1, ExbD2 and TonB proteins
(Álvarez et al., 2008). Further, an experimental infection of rainbow trout fry showed that an exbD2 mutant was attenuated approximately 450-fold compared to the wild strain consistent with the notion that the TonB system plays a significant role in virulence of F. psychrophilum
(Álvarez et al., 2008).
1.3 Disease Associated with Flavobacterium psychrophilum
Since the first description of F. psychrophilum, disease outbreaks associated with this pathogen have been reported worldwide. Flavobacterium psychrophilum infections cause significant economic and production losses particularly in salmonid aquaculture (Antaya, 2008).
The two most significant diseases caused by F. psychrophilum infection are bacterial cold-water disease (BCWD) and rainbow trout fry syndrome (RTFS). These diseases typically occur at water temperatures between 4 ˚C to 15 ˚C and are most prevalent and serious below 10 ˚C (Holt,
1987).
12 1.3.1 Bacterial cold-water disease (BCWD)
Bacterial cold-water disease (BCWD), also known as ‘peduncle disease’ or ‘tail-rot’, was first described by Borg in 1948. The disease outbreaks most often occur in farmed rainbow trout and coho salmon (LaFrentz and Cain, 2004; Antaya, 2008; Chen et al., 2008). Although BCWD affects all life stages, young fish especially fry and fingerlings are most susceptible to F. psychrophilum infection (Holt, 1987; Brown et al., 1997). The disease is most acute in fry and fingerlings whereas in larger fish, it tends to be more chronic. A wide range of mortality rates from BCWD has been reported. Mortality may be as low as 1%, but it can reach up to 75% in a severe epizootic (Post, 1987). The highest mortality rate reported to date was 90% in a population of rainbow trout in Norway (Nilsen et al., 2011).
Although BCWD primarily affects salmonid fish, other aquatic vertebrates are also negatively impacted. For example, F. psychrophilum has been isolated from Japanese eel
Anguilla japonica (Izumi et al., 2003), ayu Plecoglossus altivelis (Iida and Mizokami, 1996; Lee and Heo, 1998), and pale chub Zaco platypus (Wakabayashi et al., 1994; Iida and Mizokami,
1996). European eel Anguilla anguilla, common carp Cyrpinus carpio, crucian carp Carassius carassius, and tench Tinca tinca are also affected (Lehmann et al., 1991). The occurrence of F. psychrophilum infections in wild fish including perch Perca fluviatilis and roach Rutilis rutilis has also reported (Madetoja et al., 2002). That said, infection in non-salmonid fish species appears to be less severe than infections in salmonids (Nematollahi et al., 2003b).
Flavobacterium psychrophilum has also been isolated from the brain tissue of a newt (Brown et al., 1997).
13 1.3.2 Rainbow trout fry syndrome (RTFS)
In Europe, F. psychrophilum has been most commonly associated with a bacteremic/septicemic disease condition in rainbow trout fry and fingerlings known as rainbow trout fry syndrome (RTFS) (Lorenzen et al., 1991; Santos et al., 1992). The condition was named
RTFS because it was primarily observed in rainbow trout fry ranging from ~ 0.2 to 2.0 g
(Lorenzen et al., 1991). The disease was first described in 1985 and involved systemic disease outbreaks in rainbow trout hatcheries throughout the United Kingdom (Rangdale, 1995).
Mortality rates from RTFS have been reported ranging from 10 to 30% (Bruno, 1992) to as high as 70% (Lorenzen et al., 1991; Santos et al., 1992). In fish fry, the most serious losses occur approximately six to seven weeks following the first feeding (Lorenzen et al., 1991; Nematollahi et al., 2003b). The term ‘fry mortality syndrome’ (FMS) has also been used to describe the same condition (Lorenzen et al., 1991; Blazer et al., 1996).
1.4 Clinical Signs and Pathology of F. psychrophilum Infection
The clinical signs and pathology of fish affected by BCWD are varied and depend on the size and age of the infected host (Cipriano and Holt, 2005). In eggs and yolk-sac fry of coho salmon, the disease causes softening of the egg, premature hatching as well as rupture of yolk with mortalities up to 50% of the affected group (Holt, 1987). In fry, fingerlings, and larger fish, affected fish usually become inappetent and lethargic and have visible erosions on their fin tips
(Holt, 1987). Other clinical signs often present in fingerlings include distended abdomen with variable ascites, enlarged spleen, and bilateral exophthalmia (Lorenzen et al., 1991; Bruno,
1992). Some fish exhibit severe anemia and the presence of pale gills, liver and kidney (Holt,
1987; Lorenzen et al., 1991; Iida and Mizokami, 1996; Miwa and Nakayasu, 2005). Branchial
14 lesions associated with F. psychrophilum infections are reported to occur but are less severe than those caused by F. columnare, the cause of columnaris disease, or F. branchiophilum, the etiologic agent of bacterial gill disease (BGD) (Ostland et al., 1999). Diseased fry and fingerlings usually exhibit skin ulceration on the peduncle area, anterior to the dorsal fin, at the anus, or on the lower jaw (Holt, 1987). Ulcerations inside the mouth are also present in some affected fish
(Wood and Yasutake, 1956).
Skeletal muscle is also very commonly affected in fish with BCWD (Holt, 1987;
Lumsden et al., 1996). Foci of myonecrosis can develop into bullae or cavernous ulcers which subsequently rupture once the dermis is breached (Lumsden et al., 1996; Ostland et al., 1997). In some advanced cases of BCWD, severe necrosis of the caudal region may occur, which can progress further thereby exposing the caudal vertebra. Once the disease progresses to the brain, fish, especially fry, can exhibit disrupted swimming behavior (Kent et al., 1989). Ocular disorders such as opthalmitis and blindness have also been reported in chronic forms of BCWD
(Ostland et al., 1997). During this chronic stage of BCWD, affected fish may appear to swim erratically or in spiral pattern and may have dark pigmentation of caudal regions with or without spinal column deformities (Kent et al., 1989; Blazer et al., 1996). Fish that recover from acute
BCWD infection, especially young salmon and trout, often display abnormal swimming behavior due to the compression of nervous tissue by cranial lesions and spinal deformities (Kent et al.,
1989; Ostland et al., 1997). However, abnormal swimming behavior has also been reported in fish with no recent history of BCWD (Blazer et al., 1996) and with no detectable lesions (J. S.
Lumsden, personal communication). For example, fish with whirling disease caused by
Myxobolus cerebralis (Blazer et al., 1996) may have signs very similar to those seen with
BCWD. Correct diagnosis of BCWD can be obtained based on the case history (i.e., BCWD
15 outbreak history at the farm, farm management, fish behavior, fish size and clinical signs) and isolation of F. psychrophilum from affected fish tissues (Starliper, 2011).
In RTFS, affected fish usually display clinical signs similar to those seen in fish affected by BCWD (e.g., lethargy, anemia, dark skin pigmentation, loss of appetite, abdominal distension due to ascites and exophthalmia (Lorenzen et al., 1991; Bruno, 1992). Fry mortality due to RTFS typically involves acute systemic F. psychrophilum infection with bacteremia and septicemia
(Lorenzen, 1994; Henriksen et al., 2013). Fin erosion, ulcers on the body and head, mouth and eye lesions may also occur in both fry and fingerlings (Bernardet and Kerouault, 1989; Lehmann et al., 1991; Dalsgaard, 1993; Martínez et al., 2004). Necropsy findings typically reveal splenomegaly (the most consistent internal lesion), pale gills, liver and kidney and sometimes a hemorrhagic protruding anus (Lorenzen et al., 1991; Bruno, 1992).
Similar histopathological changes have been reported in acute F. psychrophilum infections of fish with BCWD or RTFS (Ekman, 2003). In RTFS, histopathology reveals numerous filamentous bacterial rods dispersed throughout the parenchyma of the spleen
(Rangdale et al., 1999; Ekman et al., 2003; Ekman and Norrgren, 2003). The presence of filamentous bacteria in the spleen is considered pathognomonic for RTFS (Rangdale et al.,
1999). Changes in the fish spleen from RTFS outbreaks usually consist of edema, congestion, hemorrhages and most importantly necrosis, particularly of the ellipsoids and capsule (Lorenzen,
1994). These changes often result in complete destruction of the spleen and spread of the inflammatory response to the surrounding peritoneum (Rangdale et al., 1999).
In fish with BCWD necrosis in kidney (including hematopoietic tissue), heart, as well as liver is often seen (Wood and Yasutake, 1956; Lorenzen et al., 1991; Bruno, 1992). Since 2013
(though not earlier) rainbow trout juveniles with BCWD in Ontario outbreaks have exhibited
16 necrotizing lesions in the heart with the presence of long filamentous Gram-negative bacteria (J.
S. Lumsden, personal communication). This change in pathogenesis is unexplained but does illustrate the shifting patterns of lesions and disease that can occur. In the chronic form of
BCWD, histologic studies reveal a wide range of pathological conditions including: periostitis, osteitis, meningitis, osteochondritis, myeloencephalitis as well as periosteal proliferation of vertebrae at the junction of the vertebral column and cranium (Kent et al., 1989; Meyers, 1991;
Ostland et al., 1997; Madsen and Dalsgaard, 1999; Madsen et al., 2001). In some cases, histopathology of eye lesions reveals necrotic scleritis, which causes blindness in the affected fish (Ostland et al., 1997).
1.5 Diagnosis of F. psychrophilum Infections
Clinical signs alone are not sufficient for a definitive diagnosis as there are several bacteria that can produce similar signs of systemic disease. Therefore, several laboratory-based diagnostic methods have been developed to identify and confirm the etiologic agent of the disease.
1.5.1 Primary isolation and identification
Flavobacterium psychrophilum infection is most commonly confirmed by direct isolation of the bacterium from infected fish. Flavobacterium psychrophilum can be recovered from numerous external and internal sites especially the spleen of fish showing characteristic signs of
BCWD or RTFS (Schmidtke and Carson, 1995; Miwa and Nakayasu, 2005). Traditionally, identification of the bacterium was routinely done by evaluating its morphological, biochemical, and physiological characteristics (Bernardet and Kerouault, 1989; Schmidtke and Carson, 1995;
17 Bernardet et al., 1996; Cipriano et al., 1996; Hesami et al., 2008). API ZYM, a commercial biochemical identification kit, has also been used for rapid biochemical testing and identification of the bacterium (Lorenzen et al., 2007; Hesami et al., 2008). However, there are limitations in these methods for identification of the bacterium, especially when using clinical samples obtained from disease outbreaks. Direct culture is time consuming and can be insensitive
(Madetoja et al., 2001; Madetoja and Wiklund, 2002) and viable F. psychrophilum may be difficult to recover from clinical samples (Michel et al., 1999; Nematollahi et al., 2003b; Álvarez and Guijarro, 2007). Therefore, other methods are often employed in addition to bacterial culture.
1.5.2 Serodiagnosis
A number of serodiagnostic methods have been used to detect F. psychrophilum in water, fish, and fish sex products, as well as to diagnose or to confirm BCWD/RTFS. For example, antisera raised against F. psychrophilum have been applied in an immunofluorescence antibody test (IFAT) (Amita et al., 2000; Madetoja et al., 2002; Aoki et al., 2005; Madetoja et al., 2008;
Lindstrom et al., 2009) and in immunohistochemistry (IHC) (Lorenzen and Karas, 1992;
Evensen and Lorenzen, 1996; Madetoja et al., 2000). Immunohistochemistry is particularly useful as routine light microscopy stains often do not differentially stain F. psychrophilum in tissues. Enzyme-linked immunosorbent assays (ELISAs) have been developed using antibodies to F. psychrophilum cell surface components for detection of the pathogen in fish (Lorenzen and
Karas, 1992; Rangdale and Way, 1995; Crump et al., 2003; Lindstrom et al., 2009; Long et al.,
2012). Serological typing methods including slide agglutination and ELISAs developed for routine identification of F. psychrophilum and have also been used for epidemiological studies
18 (Mata et al., 2002). Colony blotting and immunostaining have also been used to quantify viable
F. psychrophilum from ovarian fluid and kidney samples of chum salmon Oncorhynchus keta
(Misaka et al., 2008). However, these methods are not always useful because of the existence of several different serotypes of the bacterium (Faruk, 2002).
1.5.3 Molecular techniques
Molecular methods targeting the genetic material have been augmenting or replacing bacteriological culture and serological and histological techniques (Wagener, 1997; McKeever and Rege, 1999). They have been increasingly employed to detect F. psychrophilum and to confirm the bacterial identification initially made by other methods such as standard phenotypic characterization. A number of procedures using polymerase chain reaction (PCR) amplification
(Bader and Shotts, 1998) and the sequencing of the 16S rRNA gene have been described (Suzuki and Arai, 2008). A comparison of the sensitivities and specificities of PCR assays targeting 16S rDNA, the DNA gyrase subunit genes (gyrA, gyrB), and the peptidyl-prolyl cis-trans isomerase C gene (ppiC) have been reported by Suzuki and Arai (2008). These authors suggested that PCR assays targeting gyrB and the ppiC are best for the detection of F. psychrophilum. A multiplex
PCR assay has been developed to detect several major fish pathogens (i.e., Aeromonas hydrophila, Yersinia ruckeri and Aeromonas salmonicida subsp. salmonicida) simultaneously with F. psychrophilum (Del Cerro et al., 2002). Nested PCR assays have also been developed to detect F. psychrophilum from samples such as water and fish tissue (Amita et al., 2000; Wiklund et al., 2000; Baliarda et al., 2002; Madetoja and Wiklund, 2002; Izumi et al., 2005) and also from formalin-fixed wax-embedded tissue sections (Crumlish et al., 2007). The use of a TaqMan PCR assay (targeting a 971-bp fragment of the 16S rRNA) as a diagnostic tool for F. psychrophilum
19 infection has also been described (Del Cerro et al., 2002). As well, fluorescent in situ hybridization (FISH) has been used to detect this bacterium in infected tissues (Vatsos et al.,
2002; Strepparava et al., 2012). Also, real-time or quantitative PCR assays have been developed to detect and quantify F. psychrophilum from environmental and fish samples (Marancik and
Wiens, 2013; Orieux et al., 2013; Strepparava et al., 2014).
1.6 Prevention, Control and Treatment
1.6.1 Management practices
Various efforts have been employed by fish farmers; research institutions/companies or goverments to minimize the impact of disease outbreaks in aquaculture facilities. Farm management plays an important role in preventing the introduction of infectious agents and minimizing their impact. Proper husbandry and sanitation are thought to be the most important routine practices (Nematollahi et al., 2003b ; Bebak et al., 2007). For example, rearing the most susceptible fish groups in pathogen-free water facilities separate from others is highly recommended (Nematollahi et al., 2003b). Introduction of wild or novel fish should be avoided as they may be the carriers of the pathogen and transmit disease (Post, 1987; Groff and LaPatra,
2000). Periodic fish health screening and pathogen inspection for early detection of the F. psychrophilum as well as confinement of the affected fish for further treatment have also been suggested (Starliper, 2011). Reducing stocking density as well as minimizing stress and physical handling are also recommended to prevent BCWD/RTFS outbreaks (Groff and LaPatra, 2000;
LaFrentz and Cain, 2004; Taylor, 2004; Cipriano and Holt, 2005). Moreover, maintenance of optimal water quality can also reduce the impact of F. psychrophilum infections (Groff and
LaPatra, 2000; Taylor, 2004).
20 Control measures need to be taken promptly once the disease has been diagnosed.
Removal of dead and moribund fish from the tanks or ponds is highly recommended to minimize the possibility of uncontrollable disease outbreaks in the farm facility (Madetoja et al., 2000).
Elevation of water temperatures is also recommended (Holt et al., 1989; Groff and LaPatra,
2000; Chen et al., 2008); however, this is not practical in most systems. In addition, several researchers have suggested using genetic selection for F. psychrophilum resistance (Henryon et al., 2005; Hadidi et al., 2008; Silverstein et al., 2008). It has been proposed that rainbow trout resistance to BCWD is moderately heritable, thus allowing for genetic improvements (Silverstein et al., 2008).
1.6.2 Vaccine development
Vaccination is a common practice in the aquaculture industry. However, to date, there are no commercial vaccines available for protection of fish from F. psychrophilum infection. Several studies have focused on immunogenic proteins of F. psychrophilum as vaccine candidates including heat shock proteins (Hsp) 60 and 70 (Plant et al., 2009), elongation factor-Tu (EF-TU),
SufB Fe-S assembly protein (SufB), and ATP synthaseβ (Plant et al., 2011). However, it has been reported that Hsp60 and 70 are not protective antigens when used either singly or in combination (Plant et al., 2009). As well, studies by Plant et al. (2011) reported that a single intraperitoneal (i.p.) injection of both recombinant EF-TU (rEF-TU) and rSufB does not provide protection (Plant et al., 2011). These authors also reported that rATP synthaseβ is not a suitable protein candidate when administered as a precipitated protein by the same route and suggested the possibility of combining these three immunogenic proteins (rEF-TU, rSufB and rATP
21 synthaseβ) with other immunogenic antigens for F. psychrophilum vaccine development. A detailed discussion of this topic is beyond the scope of the present literature review.
1.6.3 Antimicrobial therapy
Since vaccination is currently not an option to prevent F. psychrophilum infections, antimicrobial therapy is the most common control measure used in salmonid farms worldwide.
The key to the success of antimicrobial treatment is early detection and accurate diagnosis of the etiological agent of the disease (Starliper, 2011). Administration of antimicrobial agents in the feed requires early detection of the disease since one of the first signs of the disease associated with F. psychrophilum infection is loss of appetite (Starliper, 2011). Oxytetracycline and florfenicol have been widely applied to control BCWD outbreaks worldwide (Branson, 1989;
Groff and LaPatra, 2000; LaFrentz and Cain, 2004; Lumsden et al., 2004). Amoxicillin and oxolinic acid are commonly used in Europe (Branson, 1998; Bruun et al., 2000).
Some regions (e.g., Europe, North America and Japan) have implemented strict regulations on the use of antimicrobial agents and only a few are licensed for use in aquaculture
(Smith et al., 2009). In Canada, only four antimicrobial agents: ormetoprim-sulfadimethoxine, trimethoprim-sulfamethoxazole, florfenicol and oxytetracycline, are available for use in food fish. Of the four, only two drugs, i.e., oxytetracycline and florfenicol are suitable for use in food fish to control F. psychrophilum. Florfenicol is the most effective drug clinically and is the drug of choice for farmers treating smaller fish with BCWD (J. S. Lumsden, personal communication). Erythromycin has also been used in BCWD treatment; however, it is only available on an emergency release application basis.
22 The extensive use of antimicrobial agents in the control of disease outbreak in fish farms can result in the development of antimicrobial-resistant strains of F. psychrophilum. The resistance of F. psychrophilum to certain antimicrobial agents can either be inherent or acquired.
This pathogen has been found to be inherently resistant to a variety of antibiotics including phosphomycin, penicillin (Bustos et al., 1995), gentamicin, neomycin, polymyxin B (Rangdale et al., 1997), trimethoprim/sulfadiazine, ormetoprim-sulfadimethoxine, and oxolinic acid (Rangdale et al., 1997; Bustos et al., 1995; Dalsgaard and Madsen, 2000; Bruun et al., 2003; Hesami et al.,
2010). Reduced susceptibility of F. psychrophilum over time has been reported for oxytetracycline and amoxicillin (Rangdale et al., 1997; Bruun et al., 2000; Dalsgaard and
Madsen, 2000), suggesting that this bacterium can acquire resistance against at least some antibiotics. In studies by Hesami et al. (2010), most isolates of F. psychrophilum collected in
Ontario had higher minimum inhibitory concentrations (MICs) (> 2 µg mL-1) for oxytetracycline than for florfenicol. Henríquez-Núñez et al. (2012) suggested that the high MICs for oxytetracycline, florfenicol and oxolinic acid among Chilean isolates of F. psychrophilum in their study were likely associated with the high levels of antimicrobials applied at the farms to control disease outbreaks. On the other hand, Hesami et al. (2010) reported that many isolates of
F. psychrophilum were susceptible (i.e., had low MIC values) to erythromycin and that it might be useful for the treatment of BCWD outbreaks. However, the authors suggested that an experimental treatment trial using in vivo infection was needed to confirm their in vitro studies.
At high bacterial densities (greater than 107 cfu mL-1), F. psychrophilum is able to produce extensive antibiotic resistant biofilms (Sundell and Wiklund, 2011). When exposed to sub-inhibitory concentration of antimicrobial agents, these biofilm cells can lead to rapid development of antibiotic resistance which may promote recurrent infection events (Sundell and
23 Wiklund, 2011). In view of the increase in antimicrobial resistance demonstrated by F. psychrophilum, Nematollahi et al. (2003b) emphasized the importance of wider use of antimicrobial susceptibility testing.
1.6.4 Disinfection
Methods involving multiple chemicals and/or non-antibiotic treatment have been suggested to control BCWD. Prophylactic bath treatment of fish with anti-bacteriological chemicals such as quaternary ammonium compounds, sodium chloride and chloramine-T are sometimes used to treat mild outbreaks of the disease in fish that have primarily external lesions
(Groff and LaPatra, 2000; LaFrentz and Cain, 2004). Since F. psychrophilum may be present in ovarian fluid and on egg-surfaces, disinfection of eggs after fertilization is vital to minimize the infectious stress in hatcheries (Holt, 1987; Rangdale et al., 1996; Madetoja et al., 2002).
Povidone-iodine is most often used to disinfect fertilized and eyed eggs (Wedemeyer, 2001).
New alternatives to chemical disinfectants include the use of F. psychrophilum-specific phage therapy (Stenholm et al., 2008; Kim et al., 2010) as well as the use of probiotic bacteria
(Ström-Bestor and Wiklund, 2011); both have been found to have the potential to reduce the impact of F. psychrophilum infections (Castillo et al., 2012; Castillo et al., 2014; LaPatra et al.,
2014).
Since F. psychrophilum infects fish externally, internally, and intracellularly, researchers have conducted research on different treatment delivery methods for effective control of the disease. Treatment alternatives pairing chemicals and antimicrobial therapy at the same time have been described. For example, oral antibiotic treatments in combination of prophylactic chemical baths such as hydrogen peroxide or quaternary ammonium or potassium permanganate
(Schachte, 1983; Post, 1987; Gultepe and Tanrikul, 2006). The greatest impact of bath treatments
24 is likely to reduce the horizontal spread of the agent between fish and to reduce bacterial numbers in the water column.
1.7 Microbiota in Fish Intestines
In the intestine, like other mucosal surfaces (i.e. gills and skin), microbiota exist from the earliest stages of fish development (Romero and Navarrete, 2006). Microbiota are complex communities of microorganisms that reside on sites exposed to the environment. Similar to other vertebrates, the epithelial surfaces of fish are colonized by large numbers of microbes that can form commensal or mutual relationship with their host (Spor et al., 2011). For example, the intestinal microbiota of fish plays many important roles and can assist in host digestion, promote angiogenesis, help in the development of the mucosal immune system, and serve as a protective barrier against infections (Rawls et al., 2004; Ringø et al., 2006; Semova et al., 2012).
Accordingly, maintaining a normal and protective microbiota (excluding potential invaders) is fundamental to good fish health (Gómez and Balcázar, 2008). In fish, the composition of the microbiota can be influenced by diet (Martin-Antonio et al., 2007; He et al., 2010; Liu et al.,
2012; Green et al., 2013; Kühlwein et al., 2013), stress (Boutin et al., 2013), the environment
(Skrodenyte-Arbaciauskiene et al., 2006; Martin-Antonio et al., 2007); and disease (Gómez and
Balcázar 2008; Boutin et al., 2013).
Since the fish microbiome has been described as the first line of defense against opportunistic pathogens (Boutin et al., 2012), there has been an increased interest in the study of microbiomics. To date, most studies involving microbiota in fishes have focused on the effects of diet on the intestinal microbiota (Ingerslev et al., 2014a; Desai et al., 2012; Sealey et al., 2015;
Ringø et al., 2016). Although rainbow trout is a popular fish species in the aquaculture industry,
25 only a few studies of the microbiota of this species have been reported (Huber et al., 2004; Kim et al., 2007; Mansfield et al., 2010; Desai et al., 2012). Traditional culture-dependent, as well as culture-independent techniques (e.g., molecular-based method targeting the bacterial 16S rRNA gene) have been applied to study the composition of the microbiota of fish intestines. Using both culture and culture-independent techniques, Huber et al. (2004) showed that the most abundant bacteria are Betaproteobacteria and Gammaproteobacteria. These authors also demonstrated that the presence of unculturable species that could be revealed by fluorescence in situ hybridization
(FISH). In a study by Kim et al. (2007) using culture and molecular techniques (16S/denaturing gradient gel electrophoresis-DGGE and Sanger sequencing) the intestinal microbiota of rainbow trout was found to be mostly dominated by Proteobacteria and Fusobacteria. Using
HSP60/Sanger clone libraries. Mansfield et al. (2010) determined that a fish meal diet associated with a high diversity microbiota was dominated by both Gammaproteobacteria and Firmicutes, while fish on a plant-based diet had low diversity microbiota dominated by Firmicutes. However, a study conducted by Desai et al. (2012) using 16S/454 pyrosequencing and DGGE showed that
Proteobacteria and Firmicutes are more abundant in the microbiota of fish on a plant-based diet compared with that of fish on a fish meal diet. Since there is limited knowledge of rainbow trout intestinal microbiota, more studies are needed to further evaluate the indigenous rainbow trout intestinal microbiota, which is likely crucial in fish fitness and survival.
1.8 Rationale, Hypotheses and Objectives
Given the current limitations for the control and treatment of F. psychrophilum outbreaks in rainbow trout farms in Ontario, it is crucial to characterize the virulence potential of F. psychrophilum isolates and also to investigate the efficacy of antimicrobial treatment in vivo.
26 Since oral antimicrobial administration is a common practice in treating and/or preventing
BCWD outbreak, its impact on intestinal microbiota also needs to be studied.
The first hypothesis of this research is that there will be differences in the virulence of F. psychrophilum isolates obtained from BCWD mortality events. To test this hypothesis, experimental trials using juvenile rainbow trout injected i.p. with 108 cfu fish-1 of 21 representative F. psychrophilum clinical isolates of Ontario were conducted. Virulence were determined based on the degree of mortality. Selected virulent F. psychrophilum isolates from infection trials were maintained and stored in glycerol stock at -80 ˚C for future use.
The second hypothesis is that at some point during infection, the F. psychrophilum load in the spleen will be correlated with fish morbidity. This hypothesis was tested by conducting periodic sampling of the fish throughout an infection trial. Spleen samples were collected and F. psychrophilum load evaluated by culture (cfus mg-1) and by quantitative PCR (qPCR) using a method of Strepparava et al. (2014).
The third hypothesis is that oral treatment with ERY will significantly reduce mortality rate of rainbow trout infected with F. psychrophilum. First, the MIC profiles of virulent isolates were evaluated using the broth microdilution method (CLSI, 2014). Two virulent isolates of F. psychrophilum were selected for in vivo treatment studies based on their MIC profiles. Oral treatments were carried out during an infection trial using the selected F. psychrophilum isolates.
Florfenicol and erythromycin were administered in feed at a rate of 10 mg kg-1 of fish body weight day-1 and 75 mg kg-1 fish body weight day-1, respectively, for 10 consecutive days.
Kaplan-Meier survival analysis (Log-Rank) was used to compare survival rates and mortality percentage between groups will be analyzed using ANOVA.
27 The fourth hypothesis is that the prophylactic supplementation of erythromycin or florfenicol in the diet will alter the composition of intestinal microbiota in rainbow trout juveniles. And finally, the fifth hypothesis of this work is that fish treated prophylactically with erythromycin or florfenicol will have increased susceptibility to F. psychrophilum infection compared to untreated fish. For these studies, an experimental trial was performed starting with a prophylactic oral treatment period (10 days) followed by an infection period (24 days). Intestinal contents and spleen samples were collected throughout the trial for study of the intestinal microbiota and to evaluate splenic bacterial load at different time points. For these studies, the
V3 to V4 region (Klindworth et al., 2013) of the 16S rRNA gene was sequenced using Illumina
Miseq technology. Sequence analysis was done using the mothur software package (Schloss et al., 2009). Splenic bacterial load was enumerated using qPCR assay (Strepparava et al., 2014).
Kaplan-Meier survival analysis (Log-Rank) was used to compare survival rates and mortality percentage between groups will be analyzed using ANOVA.
28 CHAPTER 2: Virulence of Flavobacterium psychrophilum Isolates in Rainbow Trout Oncorhynchus Mykiss (Walbaum)
2.1 Abstract
Flavobacterium psychrophilum, the causative agent of bacterial cold-water disease
(BCWD) in freshwater-reared salmonids, is also a common commensal organism of healthy fish.
The virulence potential of F. psychrophilum isolates obtained from BCWD cases in Ontario between 1994 and 2009 was evaluated. In preliminary infection trials of rainbow trout juveniles, significant differences (0 to 63% mortality) in the virulence of the 22 isolates tested was noted following intraperitoneal infection with 108 cfu fish-1. A highly virulent strain, FPG101, was selected for further study. When fish were injected intraperitoneally with a 106, 107, or 108 cfu fish-1 of F. psychrophilum FPG101, the 108 cfu fish-1 dose produced significantly greater mortality (p < 0.05). The bacterial load in spleen samples collected from fish every three days after infection was determined using rpoC qPCR amplification and by plate counting. Bacterial culture and rpoC qPCR highly correlated (R2 = 0.92); however, culture was more sensitive than the qPCR assay for the detection of F. psychrophilum in spleen tissue. Ninety-seven percent of
-1 the asymptomatic and the morbid fish had splenic bacterial loads of < 2.8 log10 gene copies and
-1 8 -1 > 3.0 log10 gene copies reaction , respectively following infection with 10 cfu fish .
2.2 Introduction
Flavobacterium psychrophilum, a Gram-negative bacterium, is an important fish pathogen in freshwater salmonid aquaculture. It is the aetiologic agent of bacterial cold-water
29 disease (BCWD), a disease that is associated with significant economic losses particularly in rainbow trout Oncorhynchus mykiss and coho salmon Oncorhynchus kisutch (Walbaum) (Holt,
1987; Cipriano and Holt, 2005; Lumsden et al., 2006). BCWD is typically reported at water temperatures ranging between 4 ˚C and 15 ˚C and is most prevalent and severe at 10 ˚C and below (Holt, 1987). Also known as ‘peduncle disease’ or ‘tail-rot’, BCWD affects all life stages of affected fish, with fish fry and fingerlings being the most susceptible to F. psychrophilum infections (Brown et al., 1997). In rainbow trout, mortality has been reported to be as high as
100% (Madsen and Dalsgaard, 2000; Nilsen et al., 2011).
Flavobacterium psychrophilum was first isolated from a kidney of infected coho salmon in the USA (Borg, 1948, ATCC 49418) and subsequently from rainbow trout in numerous location worldwide (Bernardet et al., 1988; Lorenzen et al., 1991; Santos et al., 1992; Toranzo and Barja, 1993; Bustos et al.,1995; Lumsden et al., 1996). It has also been isolated from
Atlantic salmon Salmo salar (Linnaeus) in Australia (Schmidtke and Carson, 1995) and from ayu
Plecoglossus altivelis (Temminck & Schlegel) and pale chub Zacco platypus (Temminck &
Schlegel) in Japan (Iida and Mizokami, 1996; Amita et al., 2000). Early reports suggested that strains of this bacterium are phenotypically homogenous (Madsen and Dalsgaard, 2000;
Madetoja et al., 2001; Valdebenito and Avendaño-Herrera, 2009) and some genetic studies of F. psychrophilum isolates recovered from multiple host fish species and from different continents showed low genetic diversity (Bernardet and Kerouault, 1989; Nicolas et al., 2008).
Subsequently, phenotypic and genotypic heterogeneity of F. psychrophilum isolates have been reported. For example, studies of 75 isolates of F. psychrophilum from Ontario, Canada showed that the isolates could be grouped into two major lineages based on their phenotypic and genotypic properties (Hesami et al., 2008). In addition, Högfors-Rönnholm and Wiklund (2010)
30 found that F. psychrophilum has two colony phenotypes, smooth and rough. They noted that the smooth type auto-agglutinated in broth media and was able to switch into a non-agglutinating rough phenotype. The rough type did not switch, but rather maintained the same non- agglutinating phenotype in all growth conditions. There was no relationship between the two phenotypic variants and virulence in rainbow trout (Högfors-Rönnholm and Wiklund 2010).
In addition to phenotypic and genotypic heterogeneity, studies on the pathogenicity of F. psychrophilum isolates in vivo have shown some differences in virulence. For example, Madsen and Dalsgaard (2000) reported F. psychrophilum strains produced cumulative mortality ranging from 12 to 100% in experimentally in rainbow trout infected via the intraperitoneal (i.p.) route.
As well, Madetoja et al. (2002) reported 5 to 100% cumulative mortality of experimentally infected rainbow trout juveniles injected subcutaneously with F. psychrophilum strains.
During infection, F. psychrophilum is able to survive inside macrophages, especially those in the spleen. Experimental infection studies of rainbow trout of different ages have shown viable F. psychrophilum cells in spleen phagocytes from young (10 wk old) fish while in older fish (5 or 15 mo old), no F. psychrophilum was detected (Decostere et al., 2001). It has also been demonstrated that rainbow trout head kidney macrophages produce more reactive oxygen species
(ROS) and are able to kill F. psychrophilum better than splenic macrophages (Nematollahi et al.,
2005). Thus, in rainbow trout, the spleen may be a relatively ‘safe site’ for F. psychrophilum to reside (Nematollahi et al., 2005). Further, it has been suggested that innate immunity against F. psychrophilum infection is positively correlated to the fish spleen size in rainbow trout (Hadidi et al., 2008).
Our laboratory, Fish Pathology Laboratory of University of Guelph, has collected over
100 F. psychrophilum isolates from BCWD outbreaks in Ontario. Although many of these
31 isolates have been phenotypically and genotypically characterized (Hesami et al., 2008), the virulence of these isolates has not been investigated. Thus, the objectives of the present study were to investigate the virulence potential of 21 Ontario F. psychrophilum isolates in juvenile rainbow trout and further, to compare splenic bacterial loads as enumerated by plate counting and quantitative polymerase chain reaction (qPCR) with mortality.
2.3 Materials and Methods
2.3.1 Bacterial isolates and culture conditions
A total of 22 F. psychrophilum isolates were used in Trial 1. Twenty-one of the isolates were from clinical cases of F. psychophilum infection submitted from several different rainbow trout farms in Ontario, Canada from 1994 to 2009. A reference strain, F. psychrophilum ATCC
49418, was included in both experimental trials (see below). Most of the 21 isolates were previously identified and characterized in vitro (Hesami et al., 2008). All isolates were verified to be F. psychrophilum using the 16S rRNA PCR amplification test of Orieux et al. (2011) as described below and partial 16S rRNA sequencing (Animal Health Laboratory, University of
Guelph, ON, Canada). Low passage (< 4) pure cultures of each isolate were maintained in modified cytophaga broth (CB; Anacker and Ordal, 1959) pH 7.5 with 15% glycerol and stored at -80 ˚C. The CB consisted of 0.05% (w/v) tryptone (Fisher Scientific, ON, Canada), 0.05%
(w/v) yeast extract (Difco Laboratories, MI, USA), 0.02% (w/v) beef extract (Difco), 0.02%
(w/v) sodium acetate (Fisher Scientific), 0.005% (w/v) potassium chloride (Fisher Scientific),
(0.005%) (w/v) calcium chloride (Fisher) and 0.005% (w/v) magnesium chloride (Fisher
Scientific). Cytophaga agar (CA: Anacker and Ordal, 1959) was prepared by the addition of
1.5% (w/v) agar (Fisher Scientific) and 0.2% (w/v) gelatin (Difco) to CB.
32 2.3.2 Bacterial suspension preparation
Frozen stocks of F. psychrophilum were sub-cultured on CA and following incubation at
15 ˚C for 72 h, cells were harvested and suspended in CB. The cells suspensions were adjusted using a Novaspec Plus spectrophotometer (GE Healthcare Life Sciences, ON, Canada) to an optical density of 0.6 at 600 nm (~ 2.24 x 109 cfu mL-1) then diluted to ~ 1 x 109 cfu mL-1. To confirm the cell numbers, suspensions were enumerated by plating 0.1 mL of ten-fold serial dilutions on CA plates in triplicate. After incubation at 15 ˚C for 4 d, visible colonies were counted on plates containing 25-250 colonies and the number of cfu mL-1 calculated as described previously (Maturin and Peeler, 2001). In Trial 2, three different concentrations (107, 108 and 109 cfu mL-1) of the virulent F. psychrophilum isolate FPG101 were prepared and enumerated as described above.
2.3.3 Fish and rearing conditions
Hatchery-raised rainbow trout from the same genetic stock weighing approximately 30 to
40 g were obtained from Lyndon Fish Hatcheries Inc. and housed in the Hagen Aqualab facility,
University of Guelph, ON, Canada. Fish were kept in circular fiberglass tanks (125 L) with continuous aeration and single pass well water at a rate of 2 L min-1 and at a temperature of 10 ±
1 ˚C. The photoperiod was 12 h dark:light and fish were fed once per day with commercial trout pellets (Profishent, Martin Mills Inc., ON, Canada) at a rate of 1.5% biomass day-1 for the duration of the experimental trials. Fish were observed to evaluate fish behavior as well as fish morbidity and mortality. The fish were acclimated for at least one week before the start of the experiments. One day prior to infection, five fish were euthanized by exposure to benzocaine
(ethyl-4-aminobenzoate; 50 mg L-1) and tested for the presence of F. psychrophilum in spleen
33 and kidney by bacterial culture followed by confirmation using a 16S rRNA PCR assay (Orieux et al., 2011; as described below). The University of Guelph Animal Care Committee approved all experimentation.
2.3.4 Experimental infection trials
2.3.4.1 Trial 1
The virulence potential of 21 isolates F. psychrophilum obtained from BCWD outbreaks was determined in rainbow trout juveniles (30 to 40 g body weight). Prior to injection, fish were anesthetized with benzocaine (50 mg L-1). Fish were randomly divided into groups and placed in duplicate tanks (n fish = 15 tank-1). The various groups were injected i.p. with 1 x 108 cfu of F. psychrophilum suspended in 0.1 mL of CB. Control fish (single group of 15 fish) were sham- infected with 0.1 mL of sterile CB injected i.p. Following injection, fish were observed three times per day for signs of morbidity and mortality including loss of appetite, reduced fright response, lethargy, abnormal swimming behavior and visible signs (e.g., abdominal enlargement, open lesions near the injection site/peduncle area, moderate to severe exophthalmia, moderate to severe tail-rot). Asymptomatic fish were considered healthy when they showed no clinical signs of disease and/or infection. Moribund fish were removed, counted as mortalities and tested for the presence of F. psychrophilum by culture and a species-specific 16S rRNA PCR assay
(Orieux et al., 2011) as described below. On Day 9 post-infection (p.i.), live control and infected fish (n = 2 tank-1), were tested for the presence of F. psychrophilum by culture and 16S rRNA
PCR. In the first series of injections (Trial 1A), 12 F. psychrophilum isolates were tested; in the second series, the remaining nine isolates were tested (Trial 1B). As a measure of repeatability,
34 one virulent isolate (FPG101) identified in Trial 1A was included in Trial 1B. Trials 1A and 1B were both carried out for 15 d.
2.3.4.2 Trial 2
In Trial 2, the effect of dose was evaluated in experimental infections over a 30-d period.
Ten-fold dilution series (107, 108 and 109 cfu mL-1) of the virulent F. psychrophilum strain
FPG101 were prepared as described above. Rainbow trout (30 g average body weight) were randomly divided into three groups of duplicate tanks (n = 50 fish tank-1). Each of the three experimental challenge groups were injected intraperitoneally with 108, 107 and 106 cfu fish-1, respectively. A single tank of control fish (n = 50) was sham-infected with 0.1 mL of sterile CB by i.p. injection. Fish (n = 3) were sampled from each tank every 3 d after the injection. For sampling, the abdominal body wall of individual fish was removed aseptically. The spleen was then collected and divided into two parts and processed for determination of bacterial load by plate counting (as described above) and by qPCR as described below. Spleen samples for bacterial counts were saved in sterile sampling bags (Fisher Scientific) and stored at -80 ˚C.
Spleen samples for genomic DNA extraction were preserved in 1.0 mL RNAlater. Following incubation for 24 h at room temperature, these samples were stored at -80 ˚C until used.
2.3.5 Histology
In Trial 1, five fish were euthanized with benzocaine on Day 0 to serve as normal tissue controls. Following injection, fish with characteristic BCWD lesions were euthanized and tissues harvested. Skin, muscle, gill, kidney, spleen, liver, and heart samples were fixed with 10% buffered formalin and routinely processed by the Animal Health Laboratory (AHL), University
35 of Guelph. Tissue sections were stained with hematoxylin & eosin (H&E) and examined by light microscopy.
2.3.6 Bacterial isolation and identification
To isolate F. psychrophilum from infected fish in Trial 1, the abdominal body wall of the fish was removed aseptically and samples from spleen and kidney were streaked with a sterile loop onto CA plates. After incubation at 15 ˚C for 7 d, bacterial colonies were characterized.
Colonial and cell morphology, Gram stain, oxidase, catalase, flexirubin, and gelatinase tests were evaluated as described previously (Lumsden et al., 1996). Presumptive F. psychrophilum isolates were tested using the 16S rRNA PCR assay of Orieux et al. (2011) as described below. Low passage F. psychrophilum isolates recovered from the different fish groups were then placed in
15% glycerol and stored at -80 ˚C.
2.3.7 DNA extraction
Total cellular DNA from bacterial cultures and spleen samples was extracted using a
DNeasy Blood and Tissue® extraction kit (Qiagen, ON, Canada) according to the manufacturer’s instructions for purification of total DNA from animal tissues (Spin-Column Protocol). Prior to
DNA extraction and purification from bacterial culture, approximately five to eight F. psychrophilum colonies grown for 96 h at 15 ˚C on CA were scraped up and resuspended in
Buffer ATL following the manufacturer’s pretreatment protocol for Gram-negative bacteria. For
DNA extraction from spleen, tissues were thawed on ice prior to the extraction. Ten milligrams of tissue was minced and DNA was extracted and purified following the manufacturer’s
36 protocol. The final concentration and A260/280 ratio of the DNA samples were determined twice using a Nanodrop™ 2000 spectrophotometer (Thermo Fisher Scientific, ON, Canada).
2.3.8 16S rRNA PCR
DNA from the spleens of control and infected fish were amplified using F. psychrophilum-specific 16S rRNA primers Fp_16S1_fw and Fp_16Sint1_rev (Orieux et al.,
2011) (Table 2.1). The PCR reactions consisted of 25 µL of TopTaq Master Mix 2x (Qiagen) which contained 1.25 U TopTaq DNA polymerase, 2x PCR buffer and 200 µM of each dNTP; 1
µL of each primer (30 µM), 7 µL of template DNA, and sufficient sterile distilled water for a final volume of 50 µL. The PCR assay was done in a pre-programmed Eppendorf Mastercycler personal thermocycler (Eppendorf, ON, Canada). The PCR amplification steps were as follows: initial denaturation of the DNA template at 96 ˚C for 10 min, followed by 40 cycles of DNA amplification in which each cycle consisted of denaturation step at 96 ˚C for 30 s, primer annealing at 56 ˚C for 30 s and primer extension at 72 ˚C for 30 s with a final amplification step at 72 ˚C for five min. After amplification, 10 µL of the PCR products were resolved by electrophoresis in a 2% agarose gel containing 10% SYBR® Safe DNA gel stain (Life
Technologies Inc., ON, Canada) and visualized using a ChemiDoc™ XRS+ imager (Biorad, ON,
Canada). Reaction mixtures containing pure culture F. psychrophilum ATCC 49418 DNA and no template were used as positive and negative controls, respectively.
37 Table 2.1 Primers and probe used in this study.
Target Product Reference Primer or probe Description of sequence gene size (bp)
Primer for PCR
Fp_16S1_fw 5 ′-GAGTTGGCATCAACACAC -3 ′ 16S Orieux et al., 146 Fp_16Sint1_rev 5′-TCCGTGTCTCAGTACCAG-3′ rRNA 2011
Primer for qPCR
F. psychro_P1F 5' -GAAGATGGAGAAGGTAATTTAGTTGATATT-3'
F. psychro_P1R 5'-CAAATAACATCTCCTTTTTCTACAACTTGA-3' Strepparava et rpoC 164 al., 2014 Probe for qPCR
F. psychrophilum_probe 5' -FAM- AAACGGGTATTCTTCTTGCTACA -3'BHQ1 -MGB
38 2.3.9 Splenic bacterial load
2.3.9.1 rpoC qPCR assay
Extracted DNA samples from spleen tissue of fish that were infected with 108 cfu fish-1
(Trial 2) and those from the control group were used to enumerate splenic bacterial loads during infection period. The rpoC qPCR assay (which targets a single copy ß´ DNA-dependent RNA polymerase gene) used in this study was adapted from the method of Strepparava et al. (2014).
The primers and the probe used are shown in Table 2.1. Briefly, the qPCR mixture consisted of 1
µL of 10x primers-probe mix containing 30 µM of each primer and 6 µM of F. psychrophilum probe, 5 µL of Lightcycler® 480 Probes Master; 2x stock concentration (Roche Diagnostics, QC,
Canada) comprised of FastStart Taq DNA polymerase, reaction buffer, dNTP mix and MgCl2;
2.5 µL template DNA and PCR-grade water to final reaction volume of 10 µL. qPCR amplification was done in a LightCycler® 480 System (Roche Diagnostics) under the following conditions: initial DNA denaturation and activation of Taq polymerase at 95 ˚C for 10 min, 40 cycles of denaturation at 95 ˚C for 15 s, primer annealing at 60 ˚C for 1 min and primer extension at 72 ˚C for 1 s followed by a final cooling step at 40 ˚C for 10 s. All samples were tested in triplicate. A known positive biological control from experimental Trial 1 of F. psychrophilum FPG101 infection and a no DNA template control and a dilution series of F. psychrophilum DNA were included as standards (as described below).
Standards for the rpoC qPCR assay were prepared as described by Strepparava et al.
(2014) using genomic DNA of reference strain F. psychrophilum ATCC 49418. The PCR products were purified using a QIAquick PCR Purification kit (Qiagen) following the manufacturer’s instruction. The concentrations of the purified DNAs were determined using a
Nanodrop™ 2000 spectrophotometer. To estimate the number of gene copies in 1 µL of purified
39 product, the total amount of the amplified DNA was divided by 1.797 x 10-7 pg (i.e., weight of the 164 bp of rpoC gene fragment). Serial ten-fold dilutions from 1 x 108 to 1 x 100 copies µL-1 of amplified DNA were used to calculate the limit of detection (LOD) of the qPCR assay and as quantitative standards for further qPCR tests. A standard curve was obtained by plotting
-1 threshold cycle (Ct) values against log10 (rpoC copies reaction ). The regression coefficient and the R2 value were calculated from the standard curve (Strepparava et al., 2014).
To determine if the splenic tissue inhibited bacterial quantification, purified splenic DNA and spleen homogenate of healthy rainbow trout were used as described previously (Marancik and Wiens, 2013). A pure culture of F. psychrophilum ATCC 49418 (incubated at 15 ˚C for 96 h on CA plates) was harvested in sterile PBS and the optical density adjusted to 0.6 at OD600. Ten- fold dilutions of the ATCC 49418 suspension (100-10-8) were prepared in PBS and added in a 1:1 volume ratio to a spleen homogenate prepared from 10 mg of tissue diluted 1:10 (w/v) in PBS.
DNA was extracted from each sample; each diluted sample was run in triplicate. The experiment was repeated the following week so that the reproducibility of the results could be assessed. The bacterial cell numbers of both the bacterial suspension and spiked spleen were determined as described above. The quantification limit was determined by plotting cycle of quantification (Cq)
-1 against bacterial concentration (log10 [cfus reaction ]). The cfu value of the last linear point on the standard curve was determined to be the assay limit. Standard curve was obtained by plotting
-1 threshold cycle (Ct) values against log10 (rpoC copies reaction ). The regression coefficient and the R2 value were calculated from the standard curve.
2.3.9.2 Splenic colony-forming unit
Frozen spleen samples stored at -80 ˚C were thawed on ice and homogenized 1:10 (w/v) in PBS. A ten-fold serial dilution of the homogenized spleen was made and 0.1 mL of each
40 dilution was spread on a CA plate as described above. After incubation at 15 ˚C for 7 d, visible colonies were counted. Each spleen sample was evaluated twice; plating was performed in triplicate.
2.3.10 Statistical analysis
Statistical analyses were performed using SAS® 9.2 (SAS Institute Inc. Cary, NC, USA) and GraphPad Prism 7 (GraphPad Software Inc., Ja Lolla, CA, USA). A one-way analysis of variance (ANOVA) was used to analyze percent mortality between experimental fish groups.
Comparison of the survival rates over time and between groups in each trial was calculated using
Kaplan-Meier survival analysis (Log-Rank). If the Log-Rank survival curves were greater than what would be expected by chance, a pairwise multiple comparison procedure was performed using the Holm-Sidak method. Pearson’s correlation was used to determine the relationship between splenic cfus and gene copies extracted from fish spleen. Unless otherwise stated, p <
0.05 was considered statistically significant.
2.4 Results
2.4.1 Experimental infection
2.4.1.1 Trial 1
On Day 5 post-infection (p.i.), some infected fish exhibited early clinical signs of F. psychrophilum infection including lethargy, loss of appetite and darkening of skin color before mortality started to occur. Most infected fish developed lesions at the site of the injection and ascites then developed erosions and ulceration on the lateral body wall. Necrosis of the fins and
41 tail (tail-rot) were also observed in some affected rainbow trout following injection, as was bilateral exophthalmia. Consistent internal gross lesions included pallor, splenomegaly and hemorrhage of the fat, serosal surfaces and muscle (Figure 2.1). Yellow-pigmented bacterial
(YPB) colonies were isolated on CA from the spleens of all infected fish sampled on Day 9 p.i.
YPB isolates exhibiting phenotypic characteristics of F. psychrophilum were confirmed as F. psychrophilum by 16S rRNA PCR amplification (Orieux et al., 2011). There was no detectable
F. psychrophilum in the fish spleens sampled from on Day 0 or from sham-infected fish on Day
9 p.i.
42
Figure 2.1 Gross lesions from an affected rainbow trout after experimental i.p. injection with F. psychrophilum during Trial 1. There was a pale kidney and liver, splenomegaly and hemorrhaging of the serosal surfaces, peritoneal cavity and muscles near the site of the injection.
43 Mortality in different infected groups varied between 0 and 63% and deaths began to occur between Day 6 and Day 9 p.i. (Table 2.2). Of the 22 F. psychrophilum isolates tested, six isolates (FPG10, FPG101, FPG100, FPG108, FPG102 and FPG105) produced high mortality (50 to 63%). FPG10 and FPG101 were the most virulent isolates producing mortality of 63% and
62%, respectively. Three isolates (FPG111, FPG103 and FPG110) had moderate virulence potential (23 to 43% mortality) while the remaining 12 isolates (FPG6, FPG25, FPG35, FPG39,
FPG48, FPG59, FPG65, FPG84, FPG87, FPG92, FPG104 and FPG107) produced low or no mortality (0 to 17%). The reference strain F. psychrophilum ATCC 49418, reported to be
‘avirulent’ (Cain et al., 2002), produced low mortality (15%) in this study. Neither morbidity nor mortality occurred in the control fish group throughout the 15-day trial.
44 Table 2.2 Percent mortality of fish groups infected with F. psychrophilum strains in Trial 1.
Year of Cumulative Isolate/strain Tissue of isolation/diagnosis isolation mortality (%) FPG10 1994a Kidney, necrotic myositis 63 FPG101 2008a Kidney, systemic disease 62 FPG100 2008a Kidney, systemic disease 57 FPG108 2008a Not recorded 53 FPG102 2007a Kidney, systemic disease 50 FPG105 2008a Not recorded 50 FPG110 2008a Not recorded 43 FPG103 2009a Kidney, systemic disease 33 FPG111 2009a Not recorded 23 FPG107 2008a Not recorded 17 FPG87 2005a Ulcerative dermatitis 17 ATCC 49418 1947b Kidney 15 FPG104 2008a Not recorded 13 FPG25 1993a Ulcerative dermatitis, necrotic 3 myositis FPG35 1992a Necrotizing stomatitis 3 FPG48 1997a Ulcerative dermatitis 0 FPG59 1997a Not recorded 0 FPG84 2005a Ulcerative dermatitis 0 FPG92 2005a Ulcerative dermatitis 0 FPG6 1996a Ulcerative dermatitis 0 FPG65 1995a Necrotizing stomatitis 0 FPG39 1996a Ulcerative dermatitis 0 a originated from Ontario, Canada; isolated from rainbow trout b originated from Washington, USA; isolated from coho salmon
45 2.4.1.2 Trial 2
In the second trial, infected fish began to show clinical signs of F. psychrophilum infection (as described in Trial 1) on Day 4 p.i. and mortality started to occur on Day 5 p.i.
(Figure 2.2). The 106, 107 and 108 cfu fish-1 infection doses of FPG101 produced 16.5%, 22.4%,
50.5% mortality, respectively. Pairwise comparison using the Log-Rank Mantel-Cox test indicated that the replicate challenge and control groups were not significantly different (p >
0.05) and so, data from the replicates were pooled. There was a significant difference between cumulative survival of all fish groups except between the control and the low (106) dose injected groups (p > 0.05). The groups of fish injected with 108 cfu fish-1 experienced significantly higher mortality (p < 0.001) than those injected with 106 cfu fish-1 and the sham-injected control group.
Neither morbidity nor mortality occurred in control group throughout the 30-d trial.
46
Figure 2.2 Cumulative survival of rainbow trout following i.p. injection with F. psychrophilum in Trial 2. Duplicate groups of rainbow trout were injected with FPG101 at low (106, square), medium (107, triangle) and high (108, cross) doses compared to one sham-injected control group. Mortality was first observed on Day 5 p.i. A significant difference (p < 0.001) was observed between the high (108) and the low (106) dose injected group. No mortalities occurred in the control fish group.
47 2.4.2 Enumeration of F. psychrophilum in fish spleen
In the rpoC qPCR assay, there was a high positive correlation (R2 = 0.995) between (Ct) values and quantification of the standards (2 x 100 to 2 x 108 rpoC copies reaction-1) (Figure 2.3).
The limit of detection for this assay was determined to be 20 rpoC copies reaction-1. For the plate count method, the presence of a single F. psychrophilum colony was considered positive.
48 40
y = -3.15x + 38.20 35 R² = 0.995
30
25
20
15 Threshold cycle (Ct) cycle Threshold
10
5
0 0 1 2 3 4 5 6 7 8 9 -1 Log10 (rpoC copies reaction )
Figure 2.3 Standard curve of rpoC qPCR assay used in this study. Each point represents the mean of threshold cycle (Ct) values of nine F. psychrophilum standards prepared in triplicate. The standards were between 2 x 100 and 2 x 108 rpoC copies reaction-1. Error bars denote standard deviation from triplicate samples.
49 Spleen samples collected at various time points from fish injected with 108 cfu fish-1
(Trial 2) were used to directly compare the detection of F. psychrophilum by colony counting and by the rpoC qPCR assay (Figure 2.4). Flavobacterium psychrophilum could be detected in spleen samples as early as Day 3 p.i. by both culture and rpoC qPCR. Based on bacterial culture, the number of F. psychrophilum positive fish increased after Day 3 to Day 12 p.i. and then decreased. Relative to bacterial culture, F. psychrophilum was detected by rpoC qPCR amplification in fewer fish and at fewer times. In total, F. psychrophilum was detected in 33% of the samples tested by culture, but in only 15% of the samples tested by rpoC qPCR. No F. psychrophilum was detected in spleen samples collected on Day 0 by either method.
50
Figure 2.4 Number of fish detected positive for splenic cfus and rpoC qPCR throughout the infection Trial 2. At most time points, more fish were positive by culture than by qPCR assay. Flavobacterium psychrophilum could be detected by both methods on Day 3 p.i. although no clinical signs were exhibited in these fish.
51 To evaluate the bacterial load in spleen over the infection period, all fish (asymptomatic and moribund) injected with the highest dose (108 cfu fish-1) were tested. The splenic bacterial load of these fish (n = 96) sampled over time was estimated using the rpoC qPCR assay (Figure
2.5). Most (97%) of the moribund fish had high splenic bacterial loads ranging between 3.1 and
-1 5.3 log10 rpoC copies reaction while 97% of the asymptomatic fish had a lower bacterial DNA
-1 load (1 to 2.5 log10 rpoC copies reaction ). Similar trends were observed in splenic bacterial
DNA load in asymptomatic and moribund fish (determined either by rpoC qPCR assay, Figure
2.5 or by plate counts, data not shown) with a rapid increase (from Days 3 to 9 p.i.) followed by a gradual decline in the splenic bacterial load throughout the 30-d trial period.
52 6 Asymptomatic Moribund 5 ) -1
4
3
2 (rpoC gene copies reaction (rpoCcopies gene 10
Log 1
0 0 5 10 15 20 25 30 Time post-infection (day)
Figure 2.5 Splenic bacterial count of asymptomatic (×, cross) and moribund (O, circle) fish from Trial 2 (108 cfu fish-1 injection group) assayed by rpoC qPCR. A few asymptomatic fish had moderate bacterial loads, typically during early infection. There was a significant difference (p < 0.05) in splenic bacterial loads between moribund and asymptomatic fish over time. Darker symbols indicate different samples with the same bacterial load.
53 There was a significant correlation (p < 0.05) between splenic counts enumerated by rpoC qPCR and culture (Figure 2.6). The Pearson correlation coefficient (R2) between the two
-1 methods was 0.915. Regression analysis yielded the following equation: log10 (cfu mg ) = -
-1 1.99+1.84 (log10 rpoC copies reaction ) of splenic bacterial count.
54
Figure 2.6 Correlation of F. psychrophilum load determined by an rpoC qPCR assay and viable colony counting (cfu) of 96 spleen samples from infected fish group from Trial 2. Linear -1 regression log10 (cfu) = -1.99 + 1.84 (log10 rpoC copies reaction ) is shown by the straight line. There was a very strong positive correlation between the two counting methods (R2 = 0.915, p < 0.05).
55 2.4.3 Histopathology
In Trial 1, fish sampled on Day 0 and from the control group did not exhibit any significant histopathological lesions. In infected fish, histological lesions were observed predominantly in the spleen, peritoneum and in the abdominal skin and muscle. In contrast to uninfected fish (Figure 2.7A), most of the spleens of the affected fish lost definition of both red pulp (RP) and white pulp (WP) and were necrotic (Figure 2.7B), particularly the ellipsoids.
56
Figure 2.7 Light microscopy of rainbow trout spleen sampled in Trial 1. A) Normal spleen from fish sampled on Day 0 with expected organization of red and white pulp; splenic capsule and ellipsoids are fully intact. B) Spleen of affected fish sampled on D12 p.i. with rarefaction of splenic parenchyma, extensive hemorrhage (blue arrow) and necrosis (black arrow). (H&E. Bar = 200 µm)
57 Histological lesions in the skin and muscles seen in affected fish were mostly localized at the site of injection and were dominated by necrosis. There were numerous filamentous bacteria present in affected tissues (Figure 2.8).
58
Figure 2.8 Section of epidermis and peritoneum from close to the injection site of an infected rainbow trout. There was epidermal ulceration, attenuation of the dermis, marked inflammation (black arrow) of the peritoneal adipose tissue and large numbers of bacteria (blue arrow) noted. (H&E. Bar = 200 µm)
59 2.5 Discussion
The virulence potential of 21 Ontario F. psychrophilum isolates was evaluated over a 15- day infection period. A broad range in pathogenicity was observed (0 - 63% mortality) despite the fact that all of the isolates originated from BCWD mortality events. Our findings are consistent with previous experimental trial studies where a broad range of virulence of clinical F. psychrophilum isolates has been reported. For example, Rangdale (1995) reported that six different strains of F. psychrophilum at 108 cfus mL-1 infectious dose yielded mortalities ranging from 17 to 74% following i.p. injection of rainbow trout fry. In another study, cumulative mortality in rainbow trout following i.p. injection with 105 cfu fish-1 ranged from 12 to 100%
(Madsen and Dalsgaard, 2000). More recently, a broad range in virulence potential (0 to 94% > mortality) was demonstrated following intramuscular injection with 109 cfu mL-1 infection dose of Chilean-derived field F. psychrophilum isolates in Atlantic salmon fry (Fredriksen et al.,
2016). A similar trend was also seen in the present study; however, there were a number of F. psychrophilum isolates in the present study for which there was no information regarding anatomic origin. In these studies, the isolates that produced high mortality (≥ 50%) were isolated mostly from fish diagnosed with systemic infection whereas avirulent isolates, which produced little or no mortality (0 - 3%), were mostly recovered from fish with superficial lesions. The well-studied F. psychrophilum strain ATCC 49418, which was recovered from a kidney of coho salmon and which has been characterized as ‘avirulent’ by Cain et al. (2002) also produced low mortality (17%) in rainbow trout in the current study. Co-infection of F. psychrophilum and other pathogens have been previously reported. Evensen and Lorenzen (1997) reported the presence of F. psychrophilum and infectious pancreatic necrotic virus (IPNV) from rainbow trout
60 obtained from RTFS outbreak. Due to that, the authors unable to determine the main causative agent for fish mortality during the outbreak.
Splenomegaly was the most notable lesion found in the present study in the fish that died following i.p. injection of F. psychrophilum. The spleen (rather than the head kidney) has been demonstrated to be the target organ of F. psychrophilum (Rangdale et al., 1999). Due to their lower production of ROS, spleen macrophages are less active against F. psychrophilum than head kidney macrophages, and thus, spleen is thought to provide a ‘safe site’ where the bacterium can reside (Nematollahi et al., 2005). Since the studies performed to date have injected isolates systemically, a likely source of variability would be the ability to survive in splenic macrophages. In these cells, bacteria are protected against complement and lysozyme activity and other humoral defense mechanisms of the host and the ability to survive in phagocytes is a critical virulence trait (Decostere et al., 2001). For F. psychrophilum to be able to survive within phagocytes, they must be able to escape from the bactericidal mechanisms such as ROS and superoxide dismutase (SOD). Virulence in F. psychrophilum has been reported to be associated with the high macrophage cytotoxicity and resistance to the bactericidal activities of splenic macrophages (Nematollahi et al., 2005). In the present study, isolates from systemic tissues were were more likely to be virulent (50% mortality) than those associated with superficial sites (0-
17% mortality) as shown in Table 2.2. Since F. psychrophilum is ubiquitous in the freshwater environment, it is possible that some of the isolates that were recovered from the surface lesions of fish with BCWD were not the causative agent of the disease outbreak.
Innate immunity in rainbow trout to F. psychrophilum has been associated with the spleen size (Hadidi et al., 2008). Fish with larger spleen size were significantly more resistant to
F. psychrophilum; however, the immune mechanism for this is not known (Hadidi et al., 2008).
61 Age-related resistance to F. psychrophilum based on differences in bacterial survival in spleen macrophages in younger (10 wk) vs. older (≥ 5 mo) rainbow trout has also been reported
(Decostere et al., 2001). In studies by Semple et al. (2018), a sharp reduction in macrophage respiratory burst occurred in experimentally infected rainbow trout seven days after infection and corresponded with the development of severe clinical signs.
In the current study, the splenic bacterial load in fish infected with 108 cfu fish-1 was determined by culture and an rpoC qPCR assay. Despite the ability of the rpoC qPCR assay to detect both viable and dead target organisms, the colony-counting method had greater sensitivity for detection of F. psychrophilum in fish spleen homogenates, presumably because it was possible to use more tissue (≥ 50 mg). The plate count method, although sensitive, is very time- consuming and labor intensive while the qPCR assay does offer the advantages of speed and ease of performance. To date, relatively few qPCR assays have been developed to detect and quantify
F. psychrophilum in fish tissues and field samples such as water and soil (Orieux et al., 2011;
Marancik and Wiens, 2013; Strepparava et al., 2014). The rpoC qPCR assay is based on a single copy gene, ß´ DNA-dependent RNA polymerase (rpoC), which would correspond to one F. psychrophilum cell (Strepparava et al., 2014); and in the current study the assay was not as sensitive as desired. A qPCR assay that targets the six copies of the rRNA gene in F. psychrophilum such as developed by Orieux et al. (2011) should be more sensitive.
In the current study, 97% of the asymptomatic and the morbid fish had mean splenic
-1 -1 bacterial loads of < 2.8 log10 rpoC copies reaction and > 3.0 log10 rpoC copies reaction , respectively, following injection with 108 cfu fish-1. Three percent of asymptomatic fish had
-1 higher bacterial loads (3.3 to 3.9 log10 rpoC copies reaction ) compared to the majority of the asymptomatic fish. These higher splenic bacterial loads were detected at the early stages (Day 4
62 to Day 6 p.i.) of infection. Strepparava et al. (2014) reported that asymptomatic fish had between
0 to 1.5 x 104 cfus spleen-1 whereas spleens from morbid fish had bacterial densities in the range of 7 x 104 to 7.7 x 108 cfus spleen-1. Direct comparison of bacterial numbers in the current study and that of Strepparava et al. (2014) is not possible because the weight of the spleens of both asymptomatic and morbid fish prior to DNA extraction were not reported in the previous study.
However, there does seem to be a relationship between splenic bacterial load and morbidity.
Asymptomatic fish have been presumed to act as carriers of virulent F. psychrophilum isolates, potentially spreading the disease via a horizontal and vertical transmission (Madetoja et al.,
2000).
In conclusion, the 21 F. psychrophilum clinical isolates from Ontario that were evaluated in the present study had a large range of virulence potential (0 to 63% mortality). Isolates originally obtained from systemic tissues more likely to produce higher mortality than those from superficial sites when rainbow trout juveniles were i.p. injected with 108 cfu fish-1. Based on the rpoC qPCR test, splenic bacterial loads were significantly higher in moribund fish than in those that were asymptomatic. Culture proved to be more sensitive than the rpoC qPCR assay to detect F. psychrophilum in spleen samples as it was possible to use more tissue. In the future, comparison of the genomes of virulent and avirulent isolates of F. psychrophilum should help us to better identify potentially virulent strains and to design better diagnostic targets.
63 CHAPTER 3: Erythromycin and Florfenicol Treatment of Rainbow Trout Oncorhynchus mykiss (Walbaum) Experimentally Infected with Flavobacterium psychrophilum
3.1 Abstract
Flavobacterium psychrophilum is responsible for significant economic losses in rainbow trout (Oncorhynchus mykiss) aquaculture. Antimicrobial treatment remains the primary means of control; however, there are limited choices available for use. The objectives of the study were therefore to determine the minimum inhibitory concentrations (MICs) of erythromycin (ERY) and florfenicol (FFN) in selected F. psychrophilum isolates and to evaluate their clinical treatment efficacy in experimentally infected rainbow trout juveniles. All isolates tested had moderate susceptibility to FFN and ERY except one isolate, which had low susceptibility to
ERY. Two isolates (one with moderate and one with low susceptibility to ERY) were used in an experimental infection trial. Rainbow trout juveniles were injected intraperitoneally with 108 cfu fish-1 and after mortality had begun, fish were given ERY- and FFN-medicated feed at a rate of
75 mg kg-1 day-1 and 10 mg kg-1 day-1 fish body weight, respectively, for 10 consecutive days.
The splenic F. psychrophilum load was determined using an rpoC quantitative polymerase chain reaction (qPCR) during the 30 d trial. Relative to antibiotic-free controls, ERY treatment significantly (p < 0.05) reduced mortality of rainbow trout infected with F. psychrophilum
FPG101, even when treatment was initiated after clinical signs developed.
64 3.2 Introduction
Bacterial cold-water disease (BCWD) is one of the most concerning bacterial diseases in rainbow trout farming worldwide including Canada. This disease has been responsible for significant economic and production losses due to growth reduction and mortality (Lumsden et al., 2004; Valdebenito and Avendaño-Herrera, 2009). BCWD is caused by the Gram-negative filamentous bacterium Flavobacterium psychrophilum. It affects rainbow trout at all stages of growth, with young fish typically the most susceptible to infection (Nematollahi et al., 2005;
Long et al., 2014). Despite the significant impact of BCWD in aquaculture, there is no commercial licensed vaccine that is widely or available in Canada to protect cultured fish from
F. psychrophilum infection. Currently, antimicrobial treatment is the most common approach used to treat infection and control the horizontal transmission of F. psychrophilum in the farm. In
Canada, florfenicol (FFN) and oxytetracycline (OTC) are presently the only two approved antimicrobials for food fish that have some efficacy against BCWD. Although FFN has been the first antimicrobial of choice for treatment, rainbow trout farmers in Ontario have been alternating between it and OTC to treat infected fish (J.S. Lumsden personal communication). Even with these antimicrobials, BCWD remains a significant management challenge.
The acute form of BCWD is associated with relatively high mortality (Holt, 1987), particularly in small fish, so rapid and accurate diagnosis is crucial for controlling the disease on the farm. Oral treatment is the most feasible route of antimicrobial delivery in aquaculture; however, one of the earliest signs of BCWD is inappetence (Dalsgaard, 1993; Starliper, 2011).
Frequent monitoring of feed response and rapid administration of oral medication is critical to ensure the greatest treatment success. In addition, antimicrobial susceptibility testing should be routinely performed to ensure the best outcome when treating a BCWD outbreak. A number of
65 antimicrobial susceptibility test methods such as disk diffusion (Kum et al., 2008; Durmaz et al.,
2012; Sundell et al., 2013), agar dilution (Bruun et al., 2000; Schmidt et al., 2000; Michel et al.,
2003) and broth microdilution (Schmidt et al., 2000; Del Cerro et al., 2010; Hesami et al., 2010;
Gieseker et al., 2012; Sundell and Wiklund, 2015) have been used to determine the antimicrobial susceptibility of various F. psychrophilum strains.
With the current limitations on BCWD control measures, excessive reliance on repeated antimicrobial treatment, particularly when there are few approved for use in aquaculture, can promote the emergence of antimicrobial-resistant strains of fish pathogens. Most F. psychrophilum isolates from Spain have high (2.4-9.7 µg mL-1) minimum inhibitory concentrations (MICs) to OTC while all isolates have low (≤ 0.1 - 2.44 µg mL-1) MICs to FFN, which is a more recently introduced antimicrobial agent (Del Cerro et al., 2010). On the other hand, most of F. psychrophilum isolates isolated from rainbow trout in Finland also demonstrated reduced sensitivity to oxolinic despite the fact that the antimicrobial has been banned in the country since 2001 (Sundell et al., 2013). Ontario isolates of F. psychrophilum are inherently resistant (MIC values are high; 1-4 µg mL-1) to ormetoprim-sulfadimethoxine and trimetroprim-sulfanethoxazole, but highly susceptible to erythromycin (ERY) (Hesami et al.,
2010). As well, the MICs of FFN in most isolates are low or moderate (Hesami et al.,
2010). Erythromycin is only available in Canada for treatment of fish as an emergency release antimicrobial and no studies on its efficacy to treat BCWD have been reported.
Several F. psychrophilum isolated from Ontario rainbow trout were found to be virulent by an experimental infection trial (Chapter 2; Table 2.2); however, their susceptibility to in vivo antimicrobial treatments had not been tested. Therefore, the objectives of this study were to: 1) determine the antimicrobial resistance of selected virulent F. psychrophilum isolates to FFN and
66 ERY using the broth microdilution test and 2) determine the ability of ERY- and FFN-medicated diets to control experimentally-induced F. psychrophilum infection in juvenile rainbow trout once clinical signs of disease were apparent.
3.3 Materials and Methods
3.3.1 In vitro susceptibility testing
3.3.1.1 Bacterial isolates and culture conditions
Six virulent F. psychrophilum isolates (FPG10, FPG100, FPG101, FPG102, FPG105 and
FPG108) that produced at least 50% mortality in rainbow trout juveniles experimentally infected by an intraperitoneal (i.p.) injection were used in this study (Chapter 2; Jarau et al., 2018). Two control strains, Escherichia coli (ATCC 25922) and Aeromonas salmonicida subsp. salmonicida
(ATCC 33658), were included in the MIC test to validate the antimicrobial susceptibility test results. The bacterial maintenance and preparation of F. psychrophilum isolates was as described previously (Chapter 2; Section 2.3.2). Briefly, the F. psychrophilum isolates were maintained in
Cytophaga broth (CB; Anacker and Ordal, 1959) containing 15% glycerol and stored at -80 ˚C until use. The E. coli and A. salmonicida control strains were maintained in tryptic soy broth
(TSB) (Difco Laboratories, MI, USA) containing 15% glycerol and stored at -80 ˚C. Prior to use, frozen stocks of these control strains were sub-cultured onto tryptic soy agar (TSA) (Difco
Laboratories). Plate cultures of E. coli and A. salmonicida were incubated at 37 ˚C for 24 h and at 25 ˚C for 48 h, respectively. Frozen stocks of F. psychrophilum isolates were streaked on
Cytophaga agar (CA; Anacker and Ordal, 1959) plates followed by incubation at 15 ˚C for 96 h prior to use.
67 3.3.1.2 Preparation of bacterial suspensions
Bacterial suspensions for MICs testing were prepared following the guidelines of the
Clinical & Laboratory Standards Institute (CLSI) (2014). After incubation of the E. coli (ATCC
25922) and A. salmonicida subsp. salmonicida (ATCC 33658) control strains for 24 h or 48 h, respectively, single bacterial colonies (1-3) of each isolate were picked using a sterile loop and used to inoculate 5 mL of sterile deionized water. The bacterial suspensions were mixed well and the turbidity was adjusted to a 0.5 McFarland standard (Thermo Fisher Scientific, ON, Canada).
Similarly, single colonies (3-5) from plate cultures of F. psychrophilum incubated at 15 ˚C were suspended in 5 mL of sterile deionized water and after vigorous mixing, the suspensions were adjusted to a 0.5 McFarland standard. Then, aliquots E. coli, A. salmonicida, and F. psychrophilum suspensions were transferred into sterile glass tubes containing 11 mL of cation- adjusted Muller-Hinton broth (CAMHB) (Thermo Fisher Scientific) to produce a standardized inoculum of all isolates of approximately 5 x 105 cfu mL-1.
3.3.1.3 Minimum inhibitory concentration (MIC) test
Minimum inhibitory concentration testing was done using the broth microdilution method according to CLSI (2014) in a sterile 96-well, round-bottom microtitre plates (Corning, NY,
USA). Each plate contained: a) a positive control (bacteria without addition of antimicrobial), b) a negative control (medium only), and c) serial 2-fold dilutions of 10 concentrations of each of the antimicrobials (in triplicate) in the CAMHB medium. Two-fold serial dilutions of CAMHB were done in the plates and wells containing antimicrobials had final concentrations from 0.03 to
16.0 µg mL-1 for FFN (Sigma Aldrich, ON, Canada) or 0.25 to 128 µg mL-1 for ERY (Sigma
Aldrich). A total of 100 µL of adjusted bacterial inoculum containing 5 x 105 cfu mL-1 was transferred using a micropipette to the 100 µL of antibiotic-containing and positive control wells.
68 The inoculum + antibiotic suspensions were mixed by carefully pipetting the suspension in and out several times before covering the plates with an adhesive seal. The 96-well plates inoculated with F. psychrophilum, A. salmonicida or E. coli were incubated at 15 ˚C for 96 h, 25 ˚C for 48 h or 37 ˚C for 24 h, respectively. The number of cfus in the inoculum + antibiotic suspensions was determined by plate counting (Maturin and Peeler, 2001). After incubation, the MIC results were viewed using a mirror reader (Trek Diagnostic Systems, OH, USA) and the MIC value was recorded as the lowest concentration of antimicrobial that inhibited visible growth. The MIC tests of the control strains were repeated three times over several days in triplicate to determine the reproducibility of this method. For both antimicrobials, the lowest four concentrations were designated as low, the middle three as medium and the highest three as high (Hesami et al.,
2010).
3.3.2 In vivo oral antibiotic treatment
3.3.2.1 Preparation of bacterial inoculum
Two F. psychrophilum isolates (FPG101 and FPG105) were used for experimental infections based on their MIC values. The bacterial inoculum for the infection trial was prepared as described previously (Chapter 2; Section 2.3.2). Briefly, after 72 h incubation at 15 ˚C, F. psychrophilum cells were harvested from CA plates using sterile CB and adjusted to OD600 of 0.6
(~2.24 x 109 cfu mL-1) using a spectrophotometer (GE Healthcare Life Sciences, ON, Canada).
The cell suspensions were then diluted to ~ 1 x 109 cfu mL-1 and the number of cfus in the suspensions was determined by plate counting (Maturin and Peeler, 2001).
3.3.2.2 Fish and rearing conditions
Rainbow trout juveniles weighing approximately 30-40 g were obtained from Lyndon
Fish Hatcheries Inc. (New Dundee, ON, Canada) and housed in the Hagen Aqualab, University
69 of Guelph, ON, Canada. Fish (n = 50 tank-1) were kept in circular fiberglass tanks (125 L) with continuous aeration and flow-through ground water at a rate of 2 L min-1 at 10 ± 1˚C. The photoperiod was 12 h dark:light exposure and fish were fed once daily with a commercial trout feed pellet (Profishent, Martin Mills Inc., ON, Canada) at a rate of 1.5% of fish biomass day-1 for the duration of the trial. Daily observation was conducted to evaluate fish behavior as well as fish morbidity and mortality. Prior to infection, the fish were acclimatized for at least one week in the infection room. Once acclimatized, five fish individually taken from random tanks (n fish
= 1 tank-1) were euthanized using benzocaine (ethyl-4-aminobenzoate; 50 mg L-1) for health screening using gross examination, bacterial culture and F. psychrophilum detection using a 16S ribosomal RNA nucleic acid (16S rRNA) polymerase chain reaction (PCR) assay (Orieux et al.,
2011). These fish were confirmed to be free of F. psychrophilum based on these tests. All animal procedures and protocols were approved by the University of Guelph Animal Care Committee and were in accordance with the guidelines of the Canadian Council of Animal Care.
3.3.2.3 Experimental infection trial
Figure 3.1 shows the experimental design of this study and details of fish groups used in this study are shown in Table 3.1. Two virulent F. psychrophilum isolates (FPG101 and
FPG105) with the lowest MIC values to ERY and FFN were selected for this infection trial. Fish were anaesthetized with benzocaine (50 mg L-1) and infected fish groups were injected i.p. with
0.1 mL (~ 1 x 108 cfu fish-1) of F. psychrophilum inoculum. A control fish group was mock- infected with 0.1 mL of sterile CB. Following infection, fish were observed three times per day for signs of morbidity and mortality. Dead and moribund fish with characteristics of BCWD infection were removed from the tank and recorded as mortalities. Fish spleens were removed, preserved in RNAlater and stored at -80 ˚C until use.
70
*
Figure 3.1 Experimental timeline.
* 10-d oral antibiotic treatment started two days after the first mortality in any groups
71 Table 3.1 Details of fish groups used.
n of tank Fish group F. psychrophilum isolate Experimental details replicates* UT1 Infected, untreated 3 FFN1 FPG101 Infected, FFN-treated 3 ERY1 Infected, ERY-treated 3 UT2 Infected, untreated 3 FFN2 FPG105 Infected, FFN-treated 3 ERY2 Infected, ERY-treated 3 FFN3 Uninfected, FFN-treated 1 ERY3 Uninfected, ERY-treated 1 C Mock-infected, untreated 1
* 50 fish per tank replicate
72 3.3.2.4 Preparation of medicated feed
The medicated diet for treatment was prepared at least two days prior to oral administration. Aquaflor® (Merck, NJ, USA) and Gallimycin®-50 (Vetoquinol N.A Inc., QC,
Canada) commercial powdered antibiotic premixes were used to prepare the FFN- and ERY- medicated diet, respectively. FFN was delivered at a rate of 10 mg kg-1 fish body weight day-1 and ERY was given at 75 mg kg-1 fish body weight day-1. Both medicated feeds were prepared by surface coating the antibiotics on the commercial fish feed pellet (Profishent 3.0 mm) following the manufacturer’s instruction. For FFN-medicated diet preparation, Aquaflor® was combined with 5% vegetable oil prior to mixing with the feed. For ERY-medicated diet preparation, Gallimycin®-50 was dissolved in absolute ethanol prior to mixing with 5% vegetable oil. Then, the feed pellets and the antibiotic suspensions were mixed thoroughly to ensure even distribution to all pellets. The coated feed was then air-dried by spreading the pellets on a flat tray overnight. Approximately 25 g aliquots of the dried medicated feed were dispensed in small bags and stored at -20 ˚C until used. Control feed was surface coated with oil without antibiotic. Samples of control and medicated feed were kept at -20 ˚C until submission to an independent laboratory (Eurofins Analytical Laboratories Inc., LA, USA) for antibiotic concentration analysis. The concentrations of FFN and ERY in the feed samples were analyzed by high performance liquid chromatography (HPLC). The limit of detection of FFN and ERY were 0.1 g Mg-1 and 1 mg kg-1, respectively.
3.3.2.5 Oral treatment for infection trial
Oral antibiotic treatment was administered two days after the first mortality in any infected group. Medicated feed was given at a rate of 1.5% of fish body weight per day for 10 consecutive days. Untreated (UT1 and UT2) and control (C) fish group were fed with control
73 feed at the same feed ration as above. Fish response to feeding was recorded (data not shown) and uneaten feed was removed 15 min after being offered.
3.3.2.6 Oral treatment for antibiotic trace analysis
To determine the concentration of antibiotics in fish tissues, two separate uninfected fish groups (FFN3 and ERY3) were treated with FFN and ERY, respectively. Medicated feed was given at a rate of 1.5% of fish body weight per day for 10 consecutive days. The following day after oral treatments were ended, 12 fish from each medicated group were sampled and euthanized using benzocaine solution. Sampled fish were immediately stored in sterile bags and kept frozen at -20 ˚C until submission to an independent laboratory (Eurofins Analytical
Laboratories Inc., LA, USA) for antibiotics concentration analysis. The concentrations of FFN and ERY in muscle were measured by Eurofins Analytical Laboratories Inc. using liquid chromatography tandem-mass spectrophotometry (LC-MS/MS) was performed to analyze the florfenicol and erythromycin residue in fish muscles. The limit of detection of FFN and ERY were 0.3 µg kg-1 and 50 µg kg-1, respectively.
3.3.2.7 Fish sampling
Fish sampling was performed on Days 5, 9, 12,15, 18, 23 and 30 post-infection (p.i.) to determine splenic F. psychrophilum load during the infection period. The spleens of sampled fish were cultured to test for the presence of F. psychrophilum. The identity of putative F. psychrophilum isolates obtained was then confirmed using a 16S rRNA PCR assay (Orieux et al.,
2011). As well, F. psychrophilum was quantified using an rpoC qPCR assay (see below). Three fish were euthanized from each tank at each sampling time points except for Day 5 p.i. when two fish were taken from each tank. Morbid and freshly dead fish observed during sampling period were also included in the sampling and evaluated by splenic cultre and rpoC qPCR assay.
74 3.3.3 Splenic DNA extraction and enumeration of splenic bacterial load using qPCR
Prior to DNA extraction, frozen spleen samples were thawed on ice. Once thawed, 10 mg of each spleen was minced and then total cellular DNA from spleen samples were extracted using a DNeasy Blood and Tissue® extraction kit (Qiagen, ON, Canada) according to the manufacturer’s instructions for purification of total DNA from animal tissues (Spin-Column
Protocol). The final concentration and purity of the DNA samples were determined twice using a
Nanodrop™ 2000 spectrophotometer (Thermo Fisher Scientific). The rpoC qPCR assay of
Strepparava et al. (2014) was used to estimate the bacterial splenic load as described previously
(Chapter 2; Section 2.3.9.1). Briefly, the qPCR mixture consisted of 1 µL of 10x primers-probe mix containing 30 µM of each primer and 6 µM of F. psychrophilum probe, 5 µL of
Lightcycler® 480 Probes Master; 2x concentration (Roche Diagnostics, QC, Canada) comprised of FastStart Taq DNA polymerase, reaction buffer, dNTP mix and MgCl2; 2.5 µL template DNA and PCR-grade water to final reaction volume of 10 µL. The qPCR amplification was performed in a LightCycler® 480 System (Roche Diagnostics) under the following conditions: initial DNA denaturation and activation of Taq polymerase at 95 ˚C for 10 min, 40 cycles of denaturation at
95 ˚C for 15 s, primer annealing at 60 ˚C for 1 min and primer extension at 72 ˚C for 1 s followed by a final cooling step at 40 ˚C for 10 s. All samples were tested in triplicate. A known positive biological control from a clinical case of F. psychrophilum infection, a no DNA template control, and a dilution series of purified F. psychrophilum DNA were included in each test as standards.
75 3.3.4 Statistical analysis
Statistical analyses were performed using GraphPad Prism 7.0 (GraphPad Software, Inc.,
San Diego, CA, USA). A one-way analysis of variance (ANOVA) was used to analyze percent mortality between groups. Survival rates of each fish group over time were calculated using
Kaplan-Meier survival analysis (Log-Rank). If the Log-Rank survival curves were greater than what would be expected by chance, a pairwise multiple comparison procedure was performed using the Holm-Sidak method. The number of spleens that were F. psychrophilum positive was analyzed using a Chi-square test. Splenic bacterial load of different fish groups was compared using one-way analysis of variance (ANOVA) and Tukey's multiple-comparison test. A p value of < 0.05 was considered to be significant.
3.4 Results
3.4.1 In vitro susceptibility test
The MIC values of ERY and FFN in each of the six F. psychrophilum isolates tested were recorded after 96 h incubation (Table 3.2). Five F. psychrophilum isolates (FPG10,
FPG100, FPG101, FPG102 and FPG108) were inhibited by medium concentrations of both FFN and ERY; i.e., 1-2 µg mL-1 and 4 µg mL-1, respectively. However, one F. psychrophilum isolate,
FPG105, had medium (2 µg mL-1) and low (1 µg mL-1) MIC values for FFN and ERY, respectively. Two Flavobacterium psychrophilum isolates, FPG101 and FPG105, were selected for the in vivo study with FPG101 being a representive isolate of strains with medium susceptibility to both antimicrobials and FPG105 due to its lower MIC value to ERY. Both strains were previously demonstrated to be virulent in experimental infections of rainbow trout
(Chapter 2; Table 2.2). The E. coli and A. salmonicida controls had expected sensitivity; the MIC
76 of FFN in E. coli ATCC 26922 was 8 µg mL-1 (acceptable range: 2-8 µg mL-1) and 8 µg mL-1
(acceptable range: not available) for ERY. The MIC of FFN in A. salmonicida ATCC 33568 was
2 µg mL-1 (acceptable range: 0.5-2 µg mL-1) and the MIC of ERY was 4 µg mL-1 (acceptable range: 4-32 µg mL-1). No bacterial growth was observed in the negative control wells.
77 Table 3.2 Minimum inhibition concentration values of six F. psychrophilum isolates determined using the broth microdilution method.
-1 Year of MIC value (µg mL ) No. Isolate/strain Tissue/disease isolation FFN Interpretation ERY Interpretation 1 FPG10 1994a Kidney, necrotic myositis 2 M 4 M 2 FPG100 2008b Kidney, systemic disease 2 M 4 M 3 FPG101 2008b Kidney, systemic disease 2 M 4 M 4 FPG102 2007b Kidney, systemic disease 2 M 4 M 5 FPG105 2008b Not recorded 1 M 2 L 6 FPG108 2008b Not recorded 1 M 4 M a: originated from a private hatchery b: originated from a different private hatchery L: Low M: Medium
MIC interpretation FFN: Low (0.0375-0.25 µg mL-1); Medium (0.5-2 µg mL-1); High (4-16 µg mL-1) ERY: Low (0.25-2 µg mL-1); Medium (4-16 µg mL-1); High (32-128 µg mL-1)
78 3.4.2 In vivo oral antibiotic treatment
Upon infection day, FPG101- and FPG105-infected fish groups received final concentrations of bacterial inoculum at 1.83 x 108 cfu mL-1 and 1.56 x 108 cfu mL-1, respectively, as determined by triplicate plate counts of serially diluted bacterial inoculum. Fish infected with
F. psychrophilum exhibited clinical signs of BCWD (e.g., lethargy, loss of appetite and darkening of skin color) before mortality started to occur. Oral treatment of FPG101- and
FPG105-infected groups receiving ERY and FFN were initiated on Day 8 p.i. (two days after the first mortality began and after inappetence was noted in all infected groups) and ended on Day
17 p.i. During the treatment period, infected fish from both treatment groups had a reduced feeding response and did not consume all of the medicated feed (data not shown). Of the three groups of fish infected with FPG101, fish treated with ERY had the highest cumulative survival
(95%) followed by fish given FFN with a cumulative survival of 83% (Figure 3.2). The untreated fish group had the lowest survival (75%) and there was significantly greater (p < 0.05) survival in fish groups that were treated with ERY compared with untreated controls.
79
Figure 3.2 Mean cumulative survival of triplicate groups of fish infected with F. psychrophilum FPG101 and untreated (UT1) or treated orally with erythromycin (ERY1), florfenicol (FFN1). Survival was significantly higher (p < 0.05) in groups of fish treated with erythromycin compared to untreated groups. Vertical dashed line indicates initiation of FFN and ERY oral treatment (Day 8 p.i.).
80 Cumulative survival of three groups infected with FPG105 is shown in Figure 3.3.
Similar to fish infected with FPG101, the groups that received ERY had the highest cumulative survival (87%), followed by groups of fish treated with FFN (83%) while the untreated groups had the lowest survival (77%). No significance difference was observed between any treated groups (p > 0.05). There was no mortality observed in fish from the uninfected control groups
(C1 & C2) in either trial.
81
Figure 3.3 Mean cumulative survival triplicate groups of fish infected with F. psychrophilum FPG105 and untreated (UT2) or treated orally with erythromycin (ERY2) or florfenicol (FFN2). No significant difference (p > 0.05) was observed in fish survival rate between groups. Vertical dashed line indicates initiation of FFN and ERY oral treatment (Day 8 p.i.).
82 3.4.3 Splenic bacterial load
Flavobacterium psychrophilum was not detected using the rpoC qPCR assay from splenic tissues collected on Day 0 or from the control group (C). Splenic samples collected throughout the trial (sampled and moribund fish) that were detected positive for F. psychrophilum using the rpoC qPCR assay are summarized (Table 3.3). On Day 5 p.i., all splenic samples from infected groups were tested positive and at this time point, the groups started to exhibit a reduction in appetite (not shown). On Day 9 p.i., one day after initiation of antibiotic treatment, the number of fish splenic F. psychrophilum positive significantly decreased (p <
0.05) in all treated groups (except for FFN1). Fish groups treated with ERY (ERY1 and ERY2) had the greatest reduction (p = 0.007) in the number of F. psychrophilum positive fish between
Days 5 and 9 p.i. The untreated groups (UT1 and UT2) had the most F. psychrophilum positive fish at Day 9 p.i.; however, the difference from Day 5 p.i. was not significant. By Day 12 p.i., the number of F. psychrophilum positive fish in all groups was further reduced (except for ERY1); however, this difference was also not significant compared to Day 9 p.i. When compared to Day
5 p.i., the number of fish detected F. psychrophilum positive were significantly decreased in all fish groups on Day 12 p.i. No F. psychrophilum positive fish were detected in any of the samples collected from Day 15 p.i. onwards (data not shown).
83 Table 3.3 Number of fish detected positive with F. psychrophilum as determined using qPCR.
F. psychrophilum a Day p.i. Isolate Group Tested (n) positive (n) (%) 5 FPG101 UT1 6 6 100 FFN1 6 6 100 ERY1 6 6 100 FPG105 UT2 6 6 100 FFN2 6 6 100
ERY2 6 6 100 9 FPG101 UT1 14 7 50 FFN1 10 5 50
ERY1 10 2 20
FPG105 UT2 13 8 62 FFN2 13 6 46 ERY2 10 2 20 12 FPG101 UT1 11 3 27 FFN1 9 1 11
ERY1 9 2 22
FPG105 UT2 9 4 44
FFN2 9 2 22
ERY2 9 1 11
UT: Untreated FFN: Florfenicol-treated ERY: Erythromycin-treated
a Flavobacterium psychrophilum was not detected using the rpoC qPCR assay from splenic tissues collected on Day 0
84 The bacterial load of all F. psychrophilum positive spleens infected with FPG101 (Table
3.3) was quantified using rpoC qPCR (Figure 3.4). All FPG101 infected fish groups harbored an
-1 average of 2.20 ± 0.03 log10 gene copies reaction on Day 5 p.i. at which time there was no significant difference (p > 0.05) in the splenic bacterial load between the three infected groups.
On Day 9 p.i., there was an increase in the splenic bacterial load in all fish groups; however, a significant increase (p < 0.05) was only observed in groups of fish treated with ERY (ERY1;
-1 3.00 ± 0.12 log10 gene copies reaction ). The bacterial loads between groups treated with FFN
-1 (FFN1; 2.40 ± 0.20 log10 gene copies reaction ) were significantly lower (p < 0.05) than those treated with erythromycin. On Day 12 p.i., an insignificant increase in the bacterial load was
-1 detected in untreated (UT1; 3.18 ± 0.28 log10 gene copies reaction ) and FFN-treated (FFN1;
-1 2.50 ± 0.38 log10 gene copies reaction ) groups compared to the previous time points.
Meanwhile, a significant reduction in the bacterial load was detected in the ERY-treated group
-1 (ERY1; 1.51 ± 0.05 log10 gene copies reaction ) compared to bacterial loads from the previous time points. Significant differences (p < 0.05) in splenic bacterial loads were observed between groups UT1 and ERY1 as well as groups FFN1 and ERY1 on Day 12 p.i.
85
-1 Figure 3.4 Splenic F. psychrophilum FPG101 load (mean ± SEM log10 rpoC copies reaction ) detected from positive fish. Increase in the bacterial loads in untreated (UT1) and florfenicol- treated (FFN1) group from Day 5 p.i. to Day 12 p.i. were observed; however, there were statistically insignificant. There was a significant increase in the bacterial load in the erythromycin-treated (ERY1) group from Day 5 to Day 9 p.i. and a significant reduction observed from Day 9 to Day 12 p.i. At Day 12 p.i., erythromycin-treated fish had significantly less F. psychrophilum than both untreated or florfenicol-treated groups of fish. Significance for all comparisons was p < 0.05. Error bars denote standard error mean from samples in triplicate.
86 The splenic F. psychrophilum load from all positive samples infected with FPG105
(Table 3.3) was also quantified by rpoC qPCR (Figure 3.5). On Day 5 p.i., all fish groups
-1 harbored an average of 1.84 ± 0.02 log10 gene copies reaction and there was no significance difference (p > 0.05) amongst the three infected groups. On Day 9 p.i., a significant increase (p <
0.05) in the splenic bacterial load was detected in untreated fish (UT2; 3.54 ± 0.28 log10 gene
-1 -1 copies reaction ) and FFN-treated fish (FFN2; 3.36 ± 0.16 log10 gene copies reaction ) but not in
-1 groups of fish treated with ERY (ERY2; 1.78 ± 0.14 log10 gene copies reaction ). There was a significant difference in bacterial loads of groups UT2 and ERY2 as well as groups FFN2 and
ERY2 on Day 9 p.i. On Day 12 p.i., a significant reduction (p < 0.05) in splenic bacterial load
-1 was observed in fish groups UT2 (1.73 ± 0.02 log10 gene copies reaction ) and FFN2 (1.68 ±
-1 0.03 log10 gene copies reaction ) while groups treated with ERY (ER2; 1.60 ± 0.04 log10 gene copies reaction-1) had reduced splenic bacterial loads; however, the difference was not statistically significant compared to the previous time point.
87
-1 Figure 3.5 Splenic F. psychrophilum FPG105 load (mean ± SEM log10 rpoC copies reaction ) detected positive from infected fish groups. There was a significant increase in bacterial load in untreated (UT2) and florfenicol-treated (FFN2) group and on Day 9 p.i. compared to Day 5 p.i.
On Day 12 p.i., both of these groups had significant reductions in splenic bacterial count compared to Day 9 p.i. There was no significance change in splenic bacterial loads of fish treated with erythromycin (ERY2) at any time. Significance for all comparisons was p < 0.05. Error bars denote standard error mean from samples in triplicate.
88 3.4.4 Antibiotic concentration analyses
Analysis of the two medicated fish feeds indicated the presence of FFN and ERY at 638 ppm and 4510 ppm, respectively. The FFN- and ERY-medicated feed contained 95.2% and
89.8%, respectively, of the target concentrations. For control feed, less than 0.1 mg kg-1 of ERY and OTC were detected (FFN was not assayed). Oral delivery of the medicated diets resulted in
2.9 ppm of FFN and 38 ppm of ERY in fish muscle one day after the last oral treatment.
3.5 Discussion
In rainbow trout culture, control of BCWD outbreaks caused by F. psychrophilum has relied on farm management and chemotherapeutics (Lumsden et al., 2004). Florfenicol, first approved in Canada in 1996, and oxytetracycline are the two food-animal approved antibiotics that have some clinical impact for treatment of BCWD. However, relatively high MICs of OTC
(compared to FFN) have been reported for F. psychrophilum isolates from Ontario (Hesami et al., 2010), thus, FFN is currently the first choice of BCWD treatment in the province. Florfenicol has also been effective against other fish pathogens infecting cultured freshwater fish including
Aeromonas salmonicida subsp. salmonicida, the cause of furunculosis in salmonids (Inglis et al.,
1991; Samuelsen et al., 1998); F. columnare, the etiology of columnaris disease (Matthews et al.,
2013); Vibrio anguillarum, a cause of vibriosis in cod (Samuelsen and Bergh, 2004) and
Edwardsiella ictaluri, the cause of enteric septicemia in catfish (Gaunt et al., 2013). Previous in vitro studies suggested the potential of erythromycin for treating F. psychrophilum infections
(Hesami et al., 2010). In aquaculture, erythromycin has been used as part of broodstock management strategies against bacterial kidney disease (BKD), a disease caused by
Renibacterium salmoninarum in salmonids (Elliot et al., 1989; Lee and Evelyn, 1994; Weins,
89 2011; Elliot, 2014). Erythromycin has also been used for control of BCWD in Canada; however, its application is currently under an emergency drug release program under sections C.08.010 and C.08.011 of the Food and Drug Regulations outlined by the Canadian Food Inspection
Agency (CFIA).
The need for judicious use of antimicrobials but also a broader range of choice for rainbow trout producers is highlighted by shifts of F. psychrophilum susceptibility to antimicrobials over time (Bruun et al., 2000; Bruun, Madsen and Dalsgaard, 2003). In Denmark, oxolinic acid was initially used to control BCWD from 1986, but by 2000 most F. psychrophilum strains were resistant to the antibiotic and so, the use of OTC started to increase
(Bruun et al., 2000). Subsequently, OTC was described as being rarely used for treatment of
BCWD in Denmark due to antibiotic resistance (Brunn et al., 2003). Recently, using broth microdilution test, reduced susceptibility to OTC among F. psychrophilum isolates recovered from salmonid populations in Michigan, USA was noted which possibly due to repeated use of
OTC in fish farms (Vliet et al., 2017). In contrast, it was reported that 118 F. psychrophilum isolates from the United Kingdom showed no significant reduction in susceptibility to FFN when tested using broth microdilution and disk dilution tests, although this antimicrobial has been used its introduction in aquaculture industry over 25 years ago (Ngo et al., 2018).
In the present study, there was little variation in MIC of FFN and ERY in the six virulent
F. psychrophilum isolates tested. The MIC values for FFN and ERY were moderate in all of the isolates except FPG105, which had a low MIC value to ERY. This contrasts somewhat with a previous study in which 53% of Ontario F. psychrophilum isolates had high MIC values for FFN while 44% and 3% of the isolates were inhibited by the medium and low concentrations, respectively (Hesami et al., 2010). In studies by Kum et al. (2008), 25% of Turkish F.
90 psychrophilum isolates were resistant to FFN (MIC ranging from 0.5–16.0 µg mL-1) and variable susceptibility was observed in Chile (Henríquez-Núñez et al., 2012). The lack of variation in
ERY and FFN MIC in the current study is most likely due to the relatively small number of isolates tested and the fact that the majority had a similar antimicrobial exposure history at the hatchery level prior to their isolation.
Fish groups treated with ERY had the highest survival compared to FFN-treated or untreated control groups; however, the greater survival was not consistently statistically significant. As expected, the cumulative survival rates of the untreated control fish group in both trials were the lowest of all the groups; however, the overall mortality was less than expected.
Very similar experimental infection doses of FPG101 and FPG105 were used in a previous study produced mortalities in rainbow trout of 62% and 50%, respectively (Chapter 2; Table 2.2).
Since the fish used in this study were obtained from a farm that was involved in a family-based selective breeding program to increase genetic resistance against BCWD, it is possible that the fish used in this study were more resistant to F. psychrophilum experimental infection than those used in the previous study. However, further details following family selection of fish used in this study were not known. It is important to note that infected fish were treated only after inappetence was detected and mortality had begun. This mirrors what occurs routinely on rainbow trout farms in mortality outbreaks, but this also likely contributed to the relative lack of survival difference seen between treated groups relative to controls. Relative survival following erythromycin treatment for isolate FPG105 was also less than predicted, despite the slightly lower MIC value for this isolate compared to FPG101, although the difference in MIC was only a single concentration. Treatment of fish with FFN in both trials reduced mortality; however, this was not significant compared with untreated controls under the present experimental conditions.
91 While the dynamics of splenic bacterial loads differed between isolates, FPG101- and
FPG105-infected, ERY-treated fish had reduced loads compared to FFN-treated fish. In
FPG101-infected fish, ERY treatment significantly reduced splenic F. psychrophilum loads compared with FFN-treated fish on Day 12 p.i. In FPG105-infected fish, the splenic bacterial load was significantly lower than that of the FFN-treated and untreated control groups on Day 9 p.i. Thus, despite the delayed initiation of oral ERY treatment, it was able to contain the splenic bacterial loads during the treatment period.
Therapeutic levels of FFN and ERY (2.9 ppm and 38 ppm, respectively) were detected in fish muscle one day after the last oral treatment. Antimicrobial tissue concentrations achieved reflect uptake across the intestinal tract (Hunter and Hirst, 1997; Hatton et al., 2017; Lin et al.,
2017) and distribution to muscle (Moffitt and Schreck, 1988; Bowser et al., 2009; Wang et al.,
2015). Depletion studies of erythromycin A in rainbow trout after oral treatment at 100 mg kg-1 fish body weight day-1 for 21 days (Esposito et al., 2007) reported a concentration of 0.33 ppm in fish muscle and skin 10 days after the last ERY feeding. Florfenicol residues in rainbow trout treated per os with 20 mg kg-1 of fish body weight day-1 for 10 d (Meinertz et al., 2014) resulted in 11 ppm FFN in skin-on fillets at 6-12 h post-treatment and 0.25 ppm after 10 d.
Concentrations of FFN varied in different tissues of large yellow croaker Pseudosciaena
[Larimichthys] crocea (Richardson) following oral treatment (liver > kidney > serum > muscle > skin) (Wang et al., 2015), in agreement with other studies (Bowser et al., 2009; Sun et al., 2010).
Since the spleen is the target organ for F. psychrophilum infection, future work should examine the concentrations of antibiotics in the spleen at different time points to correlate with bacterial load.
92 In conclusion, ERY was as effective as FFN even when oral treatment was administered after mortality had begun (similar in farming practices), thus, it has a potential for treatment of
BCWD. Quantification of the splenic bacterial load indicated that ERY was also more effective than FFN at bacterial clearance, in agreement with the respective mortality produced.
Flavobacterium psychrophilum isolates with a greater range of MIC values are needed to further evaluate antimicrobial resistance, treatment success and potentially antimicrobial break-points.
Finally, for ERY to be eventually approved for use by fish farmers to treat BCWD, experimental trials examining appropriate dosage on mortality is needed.
93 CHAPTER 4: Effect of Prophylactic Treatment with Erythromycin and Florfenicol on the Intestinal Microbiota of Rainbow Trout (Oncorhynchus mykiss) Juveniles
4.1 Abstract
The diversity and function of rainbow trout (Oncorhynchus mykiss) intestinal microbiota have not been studied extensively as those of other farmed species, and few studies have examined the effects of antimicrobials on fish intestinal microbiota. To evaluate the effect of prophylactic oral antibiotic treatment with florfenicol and erythromycin on the diversity of intestinal microbiota in rainbow trout juveniles, high throughput sequencing technology was used. Prophylactic treatments were administered for 10 consecutive days before groups of fish were injected intraperitoneally with a virulent F. psychrophilum strain FPG101. Fish intestinal content sampling was performed throughout the trial to study the intestinal microbiota. The V3 to V4 region of the 16S rRNA gene was sequenced using Illumina MiSeq technology. Before prophylactic oral treatment, Tenericutes and Proteobacteria were the most abundant phyla identified and Mycoplasma was the most abundant genus; however, marked differences in the intestinal microbiota of the control groups prior to treatment were noted. Oral treatment with
ERY for 10 days did not significantly change the abundance of bacterial phyla in the intestine of rainbow trout juveniles. On the other hand, oral treatment with FFN resulted in a significant increase in Proteobacteria; Sphingomonas was the most abundant bacterial genus. The ERY- treated group had the highest survival rate followed by FFN-treated and the untreated control group following challenge; however, there was no significant difference (p > 0.05) observed amongst the three infected groups.
94 4.2 Introduction
Like other vertebrates, the normal gastrointestinal microflora plays important roles in the metabolism, nutrition, and immunity of the fish host. Early studies of fish microbiota used culture-dependent methods (Cahill, 1990; Onarheim and Raa, 1990; Ringø et al., 1995). These studies were limited in that not all species can be cultured and those that can be are often overgrown by fast-growing bacteria. In recent years, most microbiota studies have been done using culture-independent methods such as next generation sequencing (NGS) technologies such as Illumina technologies (Ingerslev et al., 2014a, 2014b; Lyons et al., 2016, 2017), pyrosequencing (Larsen et al., 2014; Givens et al., 2015; Li et al., 2015; Fan et al., 2017, Ion
Torrent (Standen et al., 2015; Gajardo et al., 2016), etc.
Rainbow trout (Oncorhynchus mykiss) are the most popular freshwater finfish farmed in
Canada and pursued in the sport fishing industry worldwide. Despite its popularity, relatively little is known about the composition of intestinal microbiota of rainbow trout or its influence on the fish health and well-being. In the studies of the intestinal microbiota of rainbow trout to date, the ß and γ subclasses of Proteobacteria, especially members of the families Enterobacteriaceae,
Pseudomonadaceae, and Vibrionaceae (Huber et al., 2004; Kim et al., 2007) are reported to be the dominant members of the intestinal microbiota.
Rainbow trout are highly susceptible to bacterial cold-water disease (BCWD), a disease that is caused by the Gram-negative filamentous bacterium Flavobacterium psychrophilum.
BCWD has been associated with high economic and production losses in aquaculture industry due to high mortality and reduction in fish growth (Holt, 1987; Cipriano and Holt, 2005).
Antimicrobial therapy has been commonly used to control and treat BCWD outbreaks; however, its application may not only promote the emergence of antimicrobial resistant pathogens, it may
95 also affect the balance of endogenous microbes that resides particularly in the gastrointestinal tract.
To date, the effects of antimicrobial agents on the intestinal microbiota of rainbow trout have received limited attention. Understanding the impact of antimicrobial agents on the intestinal microbiota composition of rainbow trout could aid in improving the farm management and lead to reduced antimicrobial use and thus enhanced the safety of rainbow trout for human consumption. Thus, the objectives of this study were i) to evaluate the effect of prophylactic oral florfenicol (FFN) and erythromycin (ERY) treatment on the diversity of intestinal microbiota in rainbow trout juveniles and ii) to investigate susceptibility of prophylactically treated fish to F. psychrophilum infection.
4.3 Materials and Methods
4.3.1 Fish and rearing conditions
Rainbow trout juveniles weighing approximately 30-40 g were obtained from Lyndon
Fish Hatcheries Inc. (New Dundee, ON, Canada) and housed in the Hagen Aqualab facility,
University of Guelph, ON, Canada. Fish (n = 50 tank-1) were kept in circular fiberglass tanks
(125 L) with continuous aeration and flow-through ground water at a rate of 2 L min-1 with an adjusted temperature of 10.0 ± 1.0 ˚C. The photoperiod was 12 h dark:light and fish were fed once daily with a commercial trout feed pellet (Profishent, Martin Mills Inc., ON, Canada) at a rate of 1.5% of biomass day-1 for the duration of the trial. Daily observations were conducted to evaluate fish behavior as well as morbidity and mortality. Prior to the start of the experiment, the fish were acclimated for at least one week in the infection trial room. Once acclimated, five fish were euthanized using benzocaine solution (ethyl-4-aminobenzoate; 50 mg L-1) for general
96 health screening by gross examination and tested for the presence of F. psychrophilum in spleen and kidney by bacterial culture. Identification of isolates was confirmed using a 16S rRNA PCR assay (Orieux et al., 2011). All animal procedures and protocols were approved by the University of Guelph Animal Care Committee and were in accordance with the guidelines of the Canadian
Council of Animal Care.
4.3.2 Preparation of medicated feed
The commercial powdered antibiotic preparations Aquaflor® (Merck, NJ, USA) and
Gallimycin®-50 (Vetoquinol N.A Inc, QC, Canada) were used to prepare the FFN- and ERY- medicated diets, respectively. The medicated diet for each treatment group was prepared based on the treatment dose of each antibiotic. FFN was administrated at 10 mg kg-1 fish body weight
(BW) day-1 and ERY at 75 mg kg-1 fish BW day-1. The medicated feeds were prepared by surface coating the antibiotics on the commercial fish feed pellet (Profishent 3.0 mm) following the manufacturer’s instructions. For the FFN-medicated diet preparation, Aquaflor® was mixed with
5% vegetable oil prior to mixing with the feed. For ERY-medicated diet preparation,
Gallimycin®-50 was dissolved in absolute ethanol prior to mixing with vegetable oil. The antibiotic + oil suspension was mixed thoroughly with the feed pellets to ensure even distribution then spread on a flat tray and air-dried overnight. Dried coated feed was dispensed into small bags (1 bag tank-1) and stored at -20 ˚C until use. Control feed was surface coated with oil without any antibiotics. Separate samples (300 g per sample) of control and medicated feed were kept in -20 ˚C until submission to an independent laboratory (Eurofins Analytical Laboratories
Inc., LA, USA) for antimicrobial concentration analysis by high performance liquid
97 chromatography (HPLC). The limit of detection for florfenicol and erythromycin was 0.1 g Mg-1 and 1 mg kg-1, respectively.
4.3.3 Preparation of bacterial inoculum
A virulent strain of F. psychrophilum, FPG101, was used in this study. The bacterial maintenance and preparation of the F. psychrophilum inoculum were done as described previously (Chapter 2; Sections 2.3.1 & 2.3.2). Briefly, F. psychrophilum FPG101 was maintained in Cytophaga broth (CB) containing 15% glycerol at -80 ˚C until used. For preparation of challenge material, frozen stocks were first streaked on Cytophaga agar (CA) plates and incubated at 15 ˚C for 96 h to obtain single colonies. Five single colonies were then sub-cultured on CA plates and incubated at 15 ˚C for 72 h and cells were harvested and suspended in CB. The cell suspensions were adjusted using Novaspec Plus spectrophotometer
(GE Healthcare Life Sciences, ON, Canada) to an optical density of 0.6 at 600 nm which corresponded to ~ 2.24 x 109 cfu mL-1. Then, the suspension was diluted to obtain a concentration of ~ 1 x 109 cfu mL-1. To confirm the cell numbers, suspensions were enumerated by plating 0.1 mL of 10-fold serial dilutions on CA plates (in triplicates per dilution). After incubation at 15 ˚C for 4 d, visible colonies were counted on replicate plates containing 25-250 colonies and the number of cfu mL-1 calculated (Maturin and Peeler, 2001).
4.3.4 Experimental design
Figure 4.1 shows the experimental design for this study. The details of fish groups and samples of intestinal contents used in this study are shown in Table 4.1. The experimental trial involved two treatment periods-- prophylactic oral treatment (10 days of oral antibiotic
98 treatment) followed by infection. Fish were divided into four groups and each group was placed in triplicate tanks with 50 fish (30-40 g) stocked in each tank. During the 10-day prophylactic oral treatment period, two groups were treated with either FFN (group F) or ERY (group E), and the remaining groups were not treated with any antimicrobials (Ca & Cb). Following treatment period, three groups; two treated groups (F & E) and an untreated group (Cb), were injected with
F. psychrophilum FPG101 whereas untreated group Ca was mock-infected with CB and served as control.
4.3.4.1 Prophylactic oral antibiotic treatment
Treated fish groups (n = 2) were fed medicated feed prepared as described above.
Medicated feed was given at rate of 1.5% of fish body weight per day for 10 consecutive days.
Non-treated (n = 1) and control fish (n = 1) group were fed with control feed at the same rate.
The response to feeding was recorded and uneaten feed was removed after 15 min of feeding.
4.3.4.2 Flavobacterium psychrophilum infection
Prior to infection, fish were anaesthetized with a high dose of benzocaine solution (50 mg
L-1). Infected groups were injected intraperitoneally with F. psychrophilum FPG101 at a dose of
1 x 108 cfu fish-1 and a control group was mock infected with 0.1 mL of sterile CB. Following infection, fish were observed three times per day for any signs of morbidity and mortality. Dead and moribund fish with signs characteristic of BCWD infection were removed from the tank, humanely euthanized, and recorded as mortalities. Spleen samples were cultured on CB to look for the presence of F. psychrophilum. The identity of putative F. psychrophilum isolates was confirmed using a F. psychrophilum specific PCR assay targeting the 16S rRNA gene (Orieux et al., 2011).
99 Infection day
Figure 4.1 Experimental design.
100 Table 4.1 Fish groups and intestinal contents sampling.
Fish Intestinal contents Day of group Group details sample group sampling (p.i.) Ca_n11a -11 Ca_0 a 0 Ca Control Ca_12 b 12 Ca_24 b 24 F_n11a -11 FFN-treated F_0 d 0 F + FPG101-infected F_I12 12 F_I24 24 E_n11a -11 ERY-treated E_0 e 0 E + FPG101-infected E_I12 12 E_I24 24 Cb_n11a -11 Untreated Cb_0c 0 Cb + FPG101-infected Cb_I12 12 Cb_I24 24
Note: Intestinal contents sample group is following the group details unless otherwise stated. a Untreated + uninfected b Untreated + mock-infected c Untreated, uninfected d Florfenicol-treated + uninfected e Erythromycin-treated + uninfected
101 4.3.5 Sample collections
The intestinal contents (n fish = 3 tank-1) were sampled at four time points (-D11, D0,
D12 & D24 p.i.) as outlined in Figure 4.1. Spleen samples (n fish = 3 tank-1) were collected on
D3, D6, D9 and D12 p.i. One day prior to sampling, all fish groups were fasted. Sampled fish were euthanized using benzocaine and moved to the laboratory for dissection. Prior to dissection, the ventral surface of the fish was sprayed with 70% ethanol. Intestinal contents and spleen collection were done as described below.
4.3.5.1 Intestinal contents
For the microbiota study, samples of intestinal contents were collected from individual fish. The abdominal body wall and the tissues surrounding digestive tract were removed aseptically. The intestinal contents were carefully removed from the midgut and hindgut of rainbow trout by squeezing the intestine with sterile forceps into a sterile 1.5 mL microcentrifuge tube. The tubes containing intestinal contents samples were snap frozen in liquid nitrogen and immediately stored at -80 ˚C until use.
4.3.5.2 Spleen
To determine the bacterial load at different time points, spleens were collected and F. psychrophilum was quantified using an rpoC qPCR assay (Strepparava et al., 2014). Spleens were aseptically removed and placed in sterile 1.5 mL microcentrifuge tubes containing
RNAlater. The tubes were then incubated for 24 h at room temperature and stored at -80 ˚C until used.
102 4.3.6 Microbiota study
4.3.6.1 DNA extraction of intestinal contents
Total genomic DNA from samples of intestinal contents was extracted using the QIAmp
DNA Stool Mini Kit (Qiagen, ON, Canada) following the manufacturer’s spin column protocol for pathogen detection. DNA was extracted from ~180 to 220 mg of fecal matter. The protocol included lysis of the sample using buffer ASL, heat-lysis at 95 ˚C for 5 min, and treatment of impurities using the InhibitEx matrix before purification of the DNA by spin column. The purity and concentration of the extracted DNA was determined using a Nanodrop™ 2000 spectrophotometer (Thermo Fisher Scientific, ON, Canada). The extracted DNA was stored at -
20 ˚C until use.
4.3.6.2 16S rRNA PCR amplification and purification
The 16S rRNA gene sequencing library was prepared following the “16S Metagenomics
Sequencing Library Preparation” protocol (Illumina, San Diego, CA, USA) targeting the V3 to
V4 region of the 16S rRNA gene as described by Klindworth et al. (2013). In the initial PCR step, the V3 to V4 region was amplified using specific primers with attached overhang adapters.
In this study, the V3 to V4 region were amplified using the 341F (5´–
CCTACGGGNGGCWGCAG – 3´) forward primer and the 805R (5´–
GACTACHVGGGTATCTAATCC – 3´) reverse primer which produced an expected amplicon size of ~460 bp. PCR reactions were performed in a total volume of 25 µL and included 12.5 µL
2x KAPA HiFi HotStart ReadyMix (KAPA Biosystems, Wilmington, MA, USA), 3 µL of both forward and reverse primers (10 µM each), 2.5 µL DNA template (50 ng µL-1) and 4 µL PCR grade water. A reaction mixture containing PCR grade water (but no template DNA) was included as a negative control. The amplifications were performed in a Biometra TGradient
103 thermocycler (Biometra GmbH, Göttingen, Germany) using the following conditions: 95 ˚C for 3 min for initial denaturation, 25 cycles of denaturation at 95 ˚C for 30 s, annealing at 55 ˚C for 30 s and extension at 72 ˚C for 30 s; the final extension was carried out at 72 ˚C for 5 min.
Following amplification, 5 µL of the amplified product of each sample was resolved by electrophoresis in a 1.5% agarose gel containing 10% SYBR® Safe DNA gel stain (Life
Technologies Inc., ON, Canada) and visualized using a ChemiDocTM XRS+ imager (Biorad, ON,
Canada). The remaining 20 µL of amplified product was then purified using Agencourt AMPure
XP magnetic beads (Beckman Coulter Inc., Mississauga, ON, Canada) following the manufacturer’s instructions. Briefly, 20 µL of magnetic beads were added to the PCR product followed by two washing steps using 200 µL of 80% ethanol for 30 s. The ethanol was carefully discarded using a pipette and the magnetic beads were air-dried at room temperature for 10 min to remove any residual ethanol. A total of 52.5 µL of 10 mM Tris buffer (pH 8.5) was added to the beads and after centrifugation, 50 µL of clear supernatant was transferred into a sterile 0.2 mL PCR tube. An aliquot of the purified product (~ 5 µL amplicon-1) was subsequently evaluated by agarose gel electrophoresis as described above and the purified amplicons were stored at -20 ˚C until use.
4.3.6.3 Index PCR, purification and sequencing
A second PCR amplification step was performed to attach dual indexing primers and
Illumina sequencing adapters to the purified PCR products. These PCR reaction mixtures consisted of 5 µL of purified amplicon, 5 µL each of Illumina forward index primer (S5XX) and
Illumina reverse index primer 1 (N7XX), 25 µL of 2x KAPA HiFi HotStart ReadyMix and 10
µL of PCR grade water for a final reaction volume of 50 µL. The PCR amplification steps were performed as described above; however; the number of cycles was reduced to 8. Following
104 amplification, 5 µL of each PCR amplicon was separated on a 1.5% agarose gel and visualized to check for bands of the predicted size (~550 bp). For the final purification step, 56 µL of AMPure
XP magnetic beads were added to the remaining PCR product, incubated for 2 min then washed twice with 200 µL of 80% ethanol. After the second washing step, the beads were air dried at room temperature for 5 min. The beads were then resuspended in 32.5 µL of 10 mM Tris buffer
(pH 8.5) and a total volume of 30 µL of clear supernatant was transferred into a sterile 96-well
PCR plate (Roche Diagnostics, QC, Canada). Purified amplicons were quantified using a
Nanodrop spectrophotometer and normalized to a final concentration of 10 ng µL-1 using 10 mM
Tris buffer (pH 8.5). The libraries were then submitted for sequencing at the University of
Guelph’s Advanced Analysis Centre using the Illumina MiSeq system (San Diego, CA, USA) with 2 x 250 chemistry.
4.3.6.4 Bioinformatics and statistical analysis
Once the sequencing was complete, the output sequence read files (fastq) were obtained from the BaseSpace® Sequence Hub (Illumina cloud computing software). The sequences generated using MiSeq technology were further analyzed using the mothur v.1.38.1 software package (Schloss et al., 2009). Paired-end reads were aligned using the SILVA reference database (v123). Sequences which aligned to regions outside of the targeted V3 to V4 regions of the 16S rRNA gene or which were less than 440 bp and greater than 466 bp in length were removed from the dataset. Sequences that contained ambiguous bases as well as those with homopolymers longer than 8 bp in length and those identified as chloroplast/ mitochondria/ unknown/ archaea/ eukaryota were also removed. Chimera filtering was performed using the
UCHIME (Edgar et al., 2011) within mothur. The resulting sequences were binned into operational taxonomic units (OTUs) at a 3% dissimilarity level using the Ribosomal Database
105 Project (RDP) Naïve Bayesian Classifier. Based on the smallest number of sequences found in a particular sample from the dataset, subsampling was performed to normalize the sequences of the
OTUs. To assess the completeness of sampling, Good’s coverage was used and rarefaction curves were set to a 0.03 cutoff for each sample. Population richness and community diversity were assessed using Chao1 and Inverse Simpson indexes, respectively. Relative abundance in each sample were assessed using both the Jaccard index and the Yue and Clayton measure. To compare the relative abundance of groups at phylum level, a 100% stacked column chart was generated using Microsoft Excel. The R software package was used to analyze clustering of individual samples by 2-dimensions (2D) non-metric multidimentional scaling (NMDS).
4.3.7 Enumeration of splenic bacterial load
4.3.7.1 Splenic DNA extraction
Genomic DNA from spleen samples was extracted using the DNeasy Blood and Tissue® extraction kit (Qiagen, ON, Canada) according to the manufacturer’s instructions for purification of total DNA from animal tissue using spin-columns. Approximately, 0.1 g of spleen tissue was used for each DNA extraction. The concentration and purity of the DNA sample was estimated twice using a Nanodrop spectrophotometer.
4.3.7.2 qPCR assay
An rpoC qPCR assay based the method of Strepparava et al. (2014) was done as described previously (Chapter 2; Section 3.3.2.7). All samples were tested in triplicate. A known positive biological control from a clinical case of F. psychrophilum infection was included with each experiment. A no DNA template control and a dilution series of F. psychrophilum FPG101
DNA were included in each test as standards in every experiment.
106 4.3.8 Statistical analysis
Survival curves were calculated using the Kaplan-Meier Prism v7.0 software package
(GraphPad Software, Inc., San Diego, CA, USA). Survivorship curves were compared using the log-rank test. A two-way ANOVA was used, and if the overall two-way ANOVA had a p value
< 0.05, Tukey’s multiple comparison was used as a post hoc test. Calculations were performed using GraphPad v7.0 software. The bacterial load data were log-transformed prior to performing statistical analysis.
4.4 Results
4.4.1 Characterization of intestinal microbiota
DNA was extracted from 132 samples of intestinal contents collected during the trial.
Following the 16S rRNA amplification and purification steps, most of the DNA templates from the FFN- and ERY- treated groups produced faint bands when evaluated by agarose gel and 119 amplicon libraries were found to be suitable for sequencing. The remaining 13 samples were not included due to the absence of a band of desired size or insufficient DNA concentration (< 10 ng
µL-1). A total of 15,422,986 sequences generated from the 119 libraries following Illumina
MiSeq sequencing were analyzed using the mothur v.1.38.1 software package. A total of
3,866,812 of high quality sequences passed all quality control filters (mean: 32,494; median:
24,630; range: 2,962–109,335). Sequences were clustered into 447 OTUs with ≥ 97% nucleotide identity.
Based on the sample with smallest number of reads generated, 2,962 sequences were randomly subsampled within individual samples for normalization. Figure 4.2 shows rarefaction curves generated by each group of the subsampled population. Based on the curves, the greatest
107 and the least bacterial diversity were observed from groups E_I12 and Cb_I12, respectively.
Sample coverage of all groups was above 97% (mean: 99.6%), which indicated good community sampling for this analysis.
108 120 Ca_n11 F_n11 100 E_n11 Cb_n11 Ca_0 80 F_0 E_0 60 Cb_0 Ca_12
Number of OTUs F_I12 40 E_I12 Cb_I12
20 Ca_24 F_I24 E_I24 0 0 500 1000 1500 2000 2500 3000 Cb_I24 Number of sequences
Figure 4.2 Rarefaction curves comparing the number of OTUs (clustered at 97% similarity) with the number of sequences found in the intestinal contents of fish groups sampled throughout the trial.
109 4.4.2 Diversity of rainbow trout intestinal microbiota
The relative abundance of bacterial phyla (≥ 0.1% OTUs) found in groups are presented in Table 4.2. Five phyla (Proteobacteria, Tenericutes, Spirochaetes, Fusobacteria and Firmicutes) as well as unclassified Bacteria were identified in this study. Overall, Proteobacteria was the most prevalent phylum followed by Tenericutes and Spirochaetes.
Relative abundance of top 18 bacterial genera found within each treated and untreated group before and after challenge with F. psychrophilum are presented in Table 4.3. Overall,
Mycoplasma was the most abundant genus followed by Brevinema and Acinetobacter.
110 Table 4.2 Relative abundances (≥ 0.1%) of bacterial phyla found in rainbow trout intestinal microbiota in antibiotic-treated and untreated groups before and after challenge with F. psychrophilum.
Ca: Untreated + mock-infected (Control) Cb: Untreated + infected F: Florfenicol-treated + infected E: Erythromycin-treated + infected
111 Table 4.3 Relative abundances of the 18 most abundant bacterial genera found in rainbow trout pintestinal microbiota in antibiotic-treated and untreated groups before (D-11 and D0) and after (D12 and D24) challenge with F. psychrophilum.
Ca: Control Cb: Untreated, infected F: Florfenicol-treated, infected E: Erythromycin-treated, infected
112 4.4.2.1 Microbiota before and after oral antibiotic treatment
There were some differences in the relative abundance of the phyla identified in the different groups before (Figure 4.3a) and after (Figure 4.3b) antibiotic treatment. For example, group Ca possessed all five of the most abundant bacteria phyla (i.e., Proteobacteria, Tenericutes,
Spirochaetes, Fusobacteria, Firmicutes) and an unclassified bacterial phylum prior to treatment (-
D11 p.i.) whereas only three of the most abundant phyla were identified in the remaining three groups (Cb, F and E). Overall, the most abundant phylum prior to antibiotic treatment was the
Tenericutes (mean: 57.3%, range: 34.6-80.4%), followed by Proteobacteria (mean: 32.1%, range:
17.6-44.4%) and Spirochaetes (mean: 8%, range: 0.6-16%). There was a significant difference (p
= 0.0012) in the relative abundances of the bacteria within each fish group. In general, no significant difference observed in the relative abundance of bacterial phyla between groups.
In both groups F (florfenicol) and E (erythromycin), of the five most abundant phyla, only Proteobacteria and Tenericutes were identified after the treatment (Figure 4.3b). In group
Cb (untreated), three phyla (Proteobacteria, Spirochaetes and Firmicutes) and an unclassified bacterial phylum were identified. All five most abundant phyla (Proteobacteria, Tenericutes,
Spirochaetes, Fusobacteria and Firmicutes) were identified in group Ca (control). Based on the overall relative abundances of the four groups, the top three phyla identified were Proteobacteria
(mean: 63.5%; range: 30.7-90.4%), followed by Tenericutes (mean: 23.9%; range: 0.3-50.4%) and Spirochaetes (mean: 6.9%; range: 0.2-27.2%). There was a significant (p < 0.05) shift in the relative abundance of Proteobacteria in group F while bacterial abundance in group E remained unchanged.
113 a) 100% 90% 80%
70% Unclassified bacteria 60% Firmicutes 50% Fusobacteria 40% Spirochaetes
Relative abundance Relative 30% Tenericutes 20% Proteobacteria 10% 0% Ca_n11 Cb_n11 F_n11 E_n11 Sample group b) 100% 90% 80%
70% Unclassified bacteria 60% Firmicutes 50% Fusobacteria 40% Spirochaetes
Relative abundance Relative 30% Tenericutes 20% Proteobacteria 10% 0% Ca_0 Cb_0 F_0 E_0 Sample group Figure 4.3 Relative abundance of the top phyla (≥ 0.1%) found in samples of intestinal contents collected from groups before (-D11 p.i.) and after prophylactic oral antibiotic treatment (D0 p.i.). a) Before oral treatment, Tenericutes and Proteobacteria were the most abundant phyla in all groups. b) After oral treatment, Proteobacteria was the most abundant phylum in all groups except for group Cb (untreated & infected). Tenericutes was the most predominant phyla in group Cb. Groups Ca & Cb = untreated, F = florfenicol-treated, E = erythromycin-treated.
114 The relative abundance of the top 10 most abundant bacterial genera collected from intestinal samples of four fish groups before (-D11 p.i.) and after (D0 p.i.) prophylactic oral antibiotic treatment are shown in Figure 4.4a and Figure 4.4b, respectively. Before oral treatment, Mycoplasma was the most abundant bacterial genus with an average of 57.3% relative abundance (range: 34.6-80.4%; Figure 4.4a) in the overall bacterial community of the four groups. This was followed by unclassified Proteobacteria (mean: 13.6%; range: 8.3-24.8%) and
Brevinema (mean: 8%; range: 0.6-16%). Fish group Ca had the most diverse bacterial community compared to groups Cb, F and E. The relative abundance of Mycoplasma in group
Cb was significantly greater (p < 0.05) than in group Ca.
Mycoplasma remained the most abundant genus observed (mean: 23.9%; range: 0.3-
50.4%), followed by Sphingomonas (mean: 20.7%; range: 0.4-71.9%) and Methylobacterium
(mean: 10.5%; range 0-41.9%) identified as the three most abundant bacterial genera collected from intestinal contents of all fish groups sampled on D0 p.i. (Figure 4.4b). Group Ca had the most diverse bacterial community followed by groups F, Cb and E. After FFN treatment,
Sphingomonas (71.5%) was the most abundant genus identified in group F, followed by
Bradyrhizobium (6.2%) and Escherichia/Shigella (5.0%) whereas the three most abundant bacterial genera present in group E were Mycoplasma (50.4%), Methylobacterium (41.9%) and
Sphingomonas (1.8%). For the untreated groups, the three most abundant bacterial genera identified in group Ca were unclassified Proteobacteria (35.9%), Acinetobacter (21.5%) and
Sphingomonas (8.6%) while the top three genera identified in group Cb, Mycoplasma (41.1%),
Brevinema (27.2%) and Rhizobium (16.4%). In group E, a significant (p = 0.0023) increase in the genus Methylobacterium observed on D0 p.i. There was an insignificant reduction in the relative abundance of Mycoplasma compared to before oral treatment. Mycoplasma was significantly (p
115 < 0.05) more abundant in group E compared to groups Ca and F while a significant (p = 0.0009) reduction in the relative abundance of the Mycoplasma was observed after treatment in group F.
A significant (p < 0.0001) increase in relative abundance of Sphingomonas was noted after treatment in group F (p < 0.0001) compared to the other three groups. On the other hand, there was a reduction in relative abundance of Mycoplasma on D0 p.i.; however, this change was not statistically significant (p > 0.05).
116 a)
100% Bradyrhizobium 90% Unclassified Bacteria 80% Stenotrophomonas 70% Rhizobium 60% Acinetobacter 50% Unclassified Fusobacteriaceae 40% Deefgea 30% Sphingomonas Relative abundance 20% Pseudomonas 10% Brevinema 0% Unclassified Proteobacteria Ca_n11 Cb_n11 F_n11 E_n11 Aeromonas Sample group Mycoplasma
b)
100% Escherichia/Shigella
90% Aeromonas 80% Deefgea 70% 60% Bradyrhizobium 50% Acinetobacter 40% Rhizobium 30% Relative abundance Relative Brevinema 20% Unclassified Proteobacteria 10% 0% Methylobacterium Ca_0 Cb_0 F_0 E_0 Sphingomonas Sample group Mycoplasma
Figure 4.4 Relative abundance of the top 10 bacterial genera found in samples of intestinal contents collected from groups before (-D11 p.i.) and after prophylactic oral antibiotic treatment (D0 p.i.). a) Before oral treatment, Mycoplasma was the most abundant genus in all groups. Fish group Ca (untreated) had the most diverse community compared to groups Cb (untreated); F (florfenicol-treated); and E (erythromycin-treated). b) After oral treatment, group Ca had the most diverse community followed by group Cb; group E and F. Sphingomonas mostly occupying the intestines of group F while Mycoplasma and Methylobacterium were the most abundant genera in group E.
117 Figure 4.5 shows the numbers of OTUs identified in rainbow trout intestinal samples from fish groups F and E before (-D11 p.i.) and after (D0 p.i.) oral prophylactic treatment.
Sample groups F_n11 and F_0 had a total of 140 OTUs and 77 OTUs, respectively with 23
OTUs shared by the two groups (Figure 4.5a) while groups E_n11 and E_0 had a total of 315
OTUs and 47 OTUs, respectively. Sixteen OTUs were present in both groups (Figure 4.5b).
Table 4.4 shows the shared bacterial OTUs identified in the intestinal contents in both fish groups F and E. Of these shared OTUs, Mycoplasma, Aeromonas, Acinetobacter,
Pseudomonas, Sphingomonas, Pseudomonas, Bradyrhizobium, Rhizobium, Eschericia/Shigella,
Stenotrophomonas and Delftia were found in both groups F and E before and after the 10-day oral prophylactic antibiotic treatment period. These shared OTUs were among the top 20 OTUs
(range: 0.2-35% relative abundance).
118 a)
b)
Figure 4.5 Numbers of OTUs in rainbow trout intestinal microbiota before (-D11 p.i.) and after (D0 p.i.) prophylactic oral antibiotic treatment. a) Group F (florfenicol-treated): total OTUs richness was 194 and total number of OTUs shared by groups before (F_n11) and after (F_0) treatment was 23. b) Group E (erythromycin-treated): total OTUs richness is 346 and total number of OTUs shared by groups before (E_n11) and after (E_0) treatment was 16.
119 Table 4.4 Bacterial OTUs in the microbiota in both groups F (florfenicol-treated) and E (erythromycin-treated) before (-D11 p.i.) and after (D0 p.i.) prophylactic oral antibiotic treatment.
Group No. OTUs Label Genus 1 Otu000001* Mycoplasma 2 Otu000004* Aeromonas 3 Otu000005* Acinetobacter 4 Otu000007* Pseudomonas 5 Otu000009 Rhizobium 6 Otu000010* Sphingomonas 7 Otu000012* Pseudomonas 8 Otu000013* Bradyrhizobium 9 Otu000014* Rhizobium 10 Otu000018* Escherichia/Shigella 11 Otu000020 Stenotrophomonas F1,2 12 Otu000021 Sphingobium 13 Otu000026* Stenotrophomonas 14 Otu000031* Delftia 15 Otu000037 Lysinibacillus 16 Otu000041 Lactobacillus 17 Otu000048 Acinetobacter 18 Otu000056 Flavobacterium 19 Otu000072 Pseudomonas 20 Otu000087 Lactobacillus 21 Otu000090 Microbacterium 22 Otu000092 Ochrobactrum 23 Otu000101 Unclassified Candidatus Saccharibacteria 1 Otu000001* Mycoplasma 2 Otu000003 Mycoplasma E1,2 3 Otu000004* Aeromonas 4 Otu000005* Acinetobacter
120 5 Otu000007* Pseudomonas 6 Otu000010* Sphingomonas 7 Otu000012* Pseudomonas 8 Otu000013* Bradyrhizobium 9 Otu000014* Rhizobium 10 Otu000016 Gemmobacter 11 Otu000018* Escherichia/Shigella 12 Otu000019 Unclassified Gammaproteobacteria 13 Otu000026* Stenotrophomonas 14 Otu000031* Delftia 15 Otu000033 Unclassified Alphaproteobacteria 16 Otu000081 Mesorhizobium
* Shared OTUs observed in both groups F and E before and after prophylactic oral antibiotic treatments. 1: Before treatment (-D11 p.i.) 2: After treatment (D0 p.i.)
121 Figure 4.6 shows a 2-dimensional (2D) NMDS plot of the community similarity within each group a) F and b) E sampled at four time points. The lowest stress value was 0.18 with an r2 value of 0.77. Following analysis of molecular variance (AMOVA), the spatial separation between the centers of the clouds in the NMDS plot before (-D11 p.i.) and after (D0 p.i.) oral antibiotic treatment for individual samples of fish group F and E was statistically significant (p <
0.05). Based on homogeneity of molecular variance analysis (HOMOVA), the variation in samples from group E before ERY treatment (-D11 p.i.) was not statistically different (p > 0.05) from the variation in samples group E after ERY treatment (D0 p.i.). On the other hand, significant variation (p < 0.05) was observed in samples from group F before and after FFN treatment.
122 a)
b)
Figure 4.6 Non-metric dimensional scaling (NMDS) plots illustrating the intestinal microbiota beta-diversity of a) F (florfenicol-treated & infected) and b) E (erythromycin-treated & infected). Each circle represents an individual intestinal contents sample based on subsample of 2,962 OTUs.
123 4.4.2.2 Microbiota after F. psychrophilum infection (D12 & D24 p.i.)
Some changes in the diversity of the most abundant bacteria phyla present in the intestinal contents were detected in fish groups sampled on D12 p.i. (Figure 4.7a). Overall,
Proteobacteria was the most abundant phylum (mean relative abundance: 56.3%; range: 11.0-
93.9%) followed by Spirochaetes (mean relative abundance: 20.7%; range: 0-53.0%) and
Tenericutes (mean relative abundance: 16.2%; range: 0.1-35.0%). In group F (FFN-treated and infected), two phyla were identified: Proteobacteria (93.9% relative abundance) and Tenericutes
(0.1% relative abundance). In group E (ERY-treated and infected), three phyla were observed:
Proteobacteria (82.3% relative abundance), followed by Fusobacteria (8.5% relative abundance) and Firmicutes (1.5% relative abundance). In group Cb (untreated & infected), three bacterial phyla were identified; Spirochaetes (53% relative abundance) was the most abundant phylum followed by Tenericutes (35% relative abundance) and Proteobacteria (11% relative abundance).
In control group Ca (untreated & mock-infected), all of the most abundant phyla were observed except Fusobacteria. In this group, Proteobacteria was identified with the highest relative abundance (38%) followed by Spirochaetes (29.8%) and Tenericutes (29.6%). There was no significant (p < 0.05) difference observed in bacterial phyla in all groups compared to the previous (D0 p.i) time point.
124 a) 100% 90% 80% 70% Unclassified bacteria 60% Firmicutes 50% Fusobacteria 40% Spirochaetes
Relative abundance Relative 30% Tenericutes 20% Proteobacteria 10% 0% Ca_12 Cb_I12 F_I12 E_I12 Sample group
b) 100% 90% 80% 70% Unclassified bacteria 60% Firmicutes 50% Fusobacteria 40% Spirochaetes
Relative abundance Relative 30% Tenericutes 20% Proteobacteria 10% 0% Ca_24 Cb_I24 F_I24 E_I24 Sample group
Figure 4.7 Relative abundance of the top phyla (≥ 0.1%) found in samples of intestinal contents collected from groups after F. psychrophilum infection (D12 & D24 p.i.). a) On D12 p.i., Proteobacteria was predominant in groups F (florfenicol-treated & infected), E (erythromycin- treated & infected) and Ca (control). b) On D24 p.i., an increase of Tenericutes was observed in all groups. Proteobacteria remained the most abundant phyla in groups F, E and Ca.
125 On D24 p.i., changes in the relative abundances of intestinal microbiota in the four fish groups: Ca, Cb, F & E (Figure 4.7b) were detected. At this time point, the most abundant phylum was Tenericutes (mean: 48%; range: 14.9-92.8%), followed by Proteobacteria (mean: 43.7%; range: 5.7-80.9%) and Spirochaetes (mean: 6%; range: 0.2-23.7%). In group F, Proteobacteria remained the most abundant phylum (80.9%), although decreased in relative abundance compared with the D12 p.i. samples while an increase in the relative abundance of Tenericutes to
14.9% was observed. However, these changes were not significant (p > 0.05). In group E, the relative abundance of intestinal microbiota collected on D24 p.i. was similar to that of samples collected on D0 and –D11 p.i. with Proteobacteria (49%) being the most abundant followed by
Tenericutes (48.7%). An insignificant (p > 0.05) increase in the relative abundance of
Tenericutes was observed in group E on D24 p.i. compared to D12 p.i. In control group Ca, the three most abundant phyla collected on D24 p.i. were similar to those detected in samples collected on D12 p.i. At this time point, the relative abundances of Proteobacteria (39%) and
Tenericutes (35.4%) increased, while there was a decrease in the relative abundance of
Spirochaetes (23.7%). In group Cb, the relative abundance of the most abundant bacterial phyla collected on D24 p.i. was similar with samples collected on -D11 p.i. Tenericutes was identified as the top phylum, with relative abundance of 92.8%, followed by Proteobacteria (5.7%). An insignificant (p > 0.05) decrease was observed in the relative abundance of Proteobacteria in groups Cb, F and E on D24 p.i compared to D12 p.i. Increase in relative abundance was observed in control group Ca; however, it was not significant (p > 0.05). In group E, reduction in the relative abundance of Fusobacteria was observed from D0 p.i. to D12 p.i.; however, it was not significant (p > 0.05).
126 At the genus level, there were some changes in the three most abundant genera observed in samples collected on D12 p.i. (Figure 4.8a). Groups Ca and E had more diverse bacterial communities compared to groups Cb and F. Overall, the most abundant genera were Brevinema
(mean relative abundance: 20.7%; range: 0-53.0%), followed by Acinetobacteria (mean relative abundance: 18.0%; range: 0.2-50.1%) and Mycoplasma (mean relative abundance: 16.2%; range:
0.1-35.0%). The top three most abundant genera identified in group F were Acinetobacter
(50.1% relative abundance), Sphingomonas (13.0% relative abundance) and Bradyrhizobium
(6.7% relative abundance). Seven out of top 10 genera (except for Brevinema and Mycoplasma) were detected in group E with the three most abundant genera identified were Acinetobacter
(relative abundance: 19.8%), unclassified Neisseriaceae (relative abundance: 13.7%) and
Rhizobium (relative abundance: 12.7%). In group Cb, the most abundant bacterial genera were
Brevinema (53.0% relative abundance), Mycoplasma (35.0% relative abundance) and unclassified Proteobacteria (7.0% relative abundance). In control group Ca, Brevinema (29.8% relative abundance), followed by Mycoplasma (29.6% relative abundance) and Deefgea (10.1% relative abundance) were identified as the most abundant phyla on D12 p.i. A highly significant
(p = 0.0021) decrease in relative abundance of Sphingomonas was observed in group F on D12 p.i. compared to on D0 p.i. As well, a significant increase (p < 0.05) in the relative abundance of
Acinetobacter compared to D0 p.i. was observed. In group E, a significant decrease (p = 0.0002) in the relative abundance of Mycoplasma relative to samples collected on D0 p.i. was detected.
Compared to relative abundance from samples collected on D0 p.i., the genus Methylobacterium had significantly decreased (p = 0.0023) in group E. No significance change in the relative abundance of the top three bacterial genera was identified in group Cb between D12 p.i. and D0
127 p.i. In control group Ca, significant increase in the relative abundance of Mycoplasma (p =
0.0143) and Brevinema (p = 0.004) was observed on D12 p.i. compared to D0 p.i.
On D24 p.i., groups Ca and E had more diverse bacterial communities followed by groups F and Cb. Overall, the most abundant genus detected was Mycoplasma (mean relative abundance: 48.0%, range:14.9-92.8%), Acinetobacter (mean relative abundance: 12.5%; range:
0.3-37.9%) and Pseudomonas (mean relative abundance:10.2%; range: 0.3-32.9%) (Figure 4.8b).
The three most abundant bacterial genera identified in group F were Acinetobacter (relative abundance: 37.9%), Pseudomonas (32.9%) and Mycoplasma (relative abundance:14.9%). In group E, Mycoplasma (relative abundance: 48.7%) was the most abundant genus followed by
Aeromonas (relative abundance: 31.0%) and Pseudomonas (relative abundance: 4.8%). In group
Ca, Mycoplasma (relative abundance: 35.4%) was the most abundant genus followed by
Brevinema (relative abundance: 23.7%) and Deefgea (relative abundance: 13.7%). Mycoplasma was the most frequently identified genus in group Cb with relative abundance of 92.8% followed by unclassified Proteobacteria (3.5%). Significant increase (p < 0.05) was observed in relative abundance of Mycoplasma in groups E and Cb on D24 p.i. compared to D12 p.i. There were differences in the composition of the bacterial genera in all groups, compared to that was observed in D12 p.i.; however, they were not significant (p > 0.05).
128 a) 100% Gemmobacter 90% Unclassified Fusobacteriaceae 80% Bradyrhizobium 70% Deefgea 60% Pseudomonas 50% Unclassified Proteobacteria 40% Unclassified Neisseriaceae 30% Aeromonas Relative abbundanceRelative 20% Sphingomonas 10% Rhizobium 0% Mycoplasma Ca_12 Cb_I12 F_I12 E_I12 Acinetobacter Sample group Brevinema b) 100% Rhizobium 90% Stenotrophomonas 80% Sphingobium 70% Unclassified Neisseriaceae 60% Sphingomonas 50% Unclassified Proteobacteria 40% Deefgea
Relative abundance Relative 30% 20% Brevinema 10% Aeromonas 0% Pseudomonas Ca_24 Cb_I24 F_I24 E_I24 Acinetobacter Sample group Mycoplasma
Figure 4.8 Relative abundance of the top 10 bacterial genera found in rainbow trout samples of intestinal contents collected from groups after F. psychrophilum infection (D12 & D24 p.i.). a) On D12 p.i.: group Ca (control) and E (erythromycin-treated & infected) had more diverse bacterial community compared to group F (florfenicol-treated & infected) and group Cb (untreated & infected). b) On D24 p.i.: groups Ca and E had the most diverse community followed by group F and group Cb. Acinetobacter and Pseudomonas were the most abundant phyla in group F while Mycoplasma was the most abundant phyla occupying the intestines of groups Ca, Cb and E.
129 4.4.3 Flavobacterium psychrophilum infection
Following infection, some of the infected fish exhibited early signs of F. psychrophilum infection (i.e., lethargy, inappetence and darkening of skin color) before mortality started to occur. Fish in group Cb, followed by groups F and E, exhibited a marked reduction in appetite
(data not shown). The progress of the infection was as described in Chapter 2. Splenic bacterial culture was carried out from infected fish and yellow-pigmented bacteria (YPB) colonies with a phenotype characteristic of F. psychrophilum were confirmed as F. psychrophilum using the 16S rRNA PCR amplification test of Orieux et al. (2011). There was no detectable F. psychrophilum in the spleens of fish sampled on –D11 and D0 p.i.
Cumulative survival curves of the four infected groups (i.e., Ca (control; untreated and mock-infected), Cb (untreated & infected), F (FFN-treated and infected) and E (ERY-treated and infected)) are shown in Figure 4.9. Of the three groups of fish infected with F. psychrophilum
FPG101, group E had the highest cumulative survival (95%) followed by group F with a cumulative survival of 91% while group Cb had the lowest cumulative survival (90%) in this study. Neither morbidity nor mortality was observed in the control group Ca throughout the infection trial. No significant difference (p > 0.05) was observed in the relative survival amongst any of the infected groups (F, E & Cb) or between the infected groups and the mock-infected control (Ca).
130
Figure 4.9 Cumulative survival curves of groups following i.p. infection with F. psychrophilum after prophylactic oral treatment with ERY and FFN. Triplicate groups of 2 oral treated (F: florfenicol-treated; E: erythromycin-treated) and one untreated (Cb) rainbow trout were injected with FPG101 at 108 cfu fish-1 compared to triplicate mock-infected control group (Ca). Mortality was first observed on D6 p.i. There was no significant difference (p > 0.05) between survival curves of all fish groups. Group E had the highest percent survival followed by group F and group Cb. No mortalities occurred in the control fish group Ca.
131 To determine the F. psychrophilum load after infection, an rpoC qPCR assay was performed on samples collected on D3, D6, D9 and D12 p.i.. Figure 4.10 shows the splenic
-1 bacterial load (mean ± SEM log10 rpoC copies reaction ) of fish infected with F. psychrophilum
FPG101 sampled at these four time points. No F. psychrophilum was detected in the mock- infected, control group Ca (data not shown). Flavobacterium psychrophilum was first detected
-1 on D3 p.i. and ranged from 1.61-1.80 log10 rpoC copies reaction in the infected groups. Using
2-way ANOVA with Tukey’s multiple comparison, the bacterial load did not vary significantly
(p > 0.05) amongst the three groups. In group F, an insignificant reduction in bacterial load was observed on D6 p.i. and the count was no different on D9 p.i. On D12 p.i., a significant increase
-1 in the splenic bacterial load was detected (2.01 ± 0.22 log10 rpoC copies reaction ). Significant differences (p < 0.05) were observed in the bacterial loads of group F fish sampled on D6 p.i. and D12 p.i. and D9 p.i. and D12 p.i. In group E, an insignificant decrease in splenic bacterial load (1.28 ± 0.02 log10 rpoC copies reaction-1) was detected on D6 p.i. and D12 p.i. (1.20 ± 0.03
-1 log10 rpoC copies reaction ). On D12 p.i., an increase in bacterial load was observed (1.72 ±
-1 0.25 log10 rpoC copies reaction ); however, it was not significant. In group Cb, a small but not
-1 significant increase in bacterial load (1.63 ± 0.19 log10 gene copies reaction ) was detected on
D6 p.i. On D9 p.i., a non-significant reduction in bacterial load (1.23 ± 0.03 log10 rpoC copies reaction -1) was observed in this group and on D12 p.i., small, but not significant increase in
-1 bacterial load (1.70 ± 0.14 log10 rpoC copies reaction ) was detected.
132
-1 Figure 4.10 Splenic F. psychrophilum loads (mean ± SEM log10 rpoC copies reaction ) of infected fish randomly sampled at 4 time points (D3, D6, D9 & D12 p.i.) as determined using qPCR. Error bars denote the standard error of the mean. On D3 p.i., bacterial loads ranging from -1 1.61-1.80 log10 rpoC copies reaction were detected in all groups. There was a significant increase (p < 0.05) in bacterial loads in F (florfenicol-treated & infected) group on D6 to D12 p.i. and D9 p.i. to D12 p.i. There was no significance change (p > 0.05) in splenic bacterial loads of fish groups E (erythromycin-treated & infected) and Cb (untreated & infected) at any time. ** refers to statistically significant as p < 0.05.
133 4.4.4 Antimicrobial concentration analysis
Table 4.5 shows the antimicrobial residue analysis of medicated feed samples in this study. Results obtained from feed samples analysis of the two medicated feed pellets indicated the presence of FFN and ERY at 1,010 ppm and 4,970 ppm, respectively. We found that FFN concentration in FFN-medicated feed was higher (50.7%) than our target dose. Meanwhile, ERY concentration in ERY-medicated feed was 2% lower than our target dose. For control feed, less than 0.1 mg kg-1 of ERY and oxytetracycline were detected (FFN was not assayed).
134 Table 4.5 Antimicrobial concentration analysis of medicated feed samples.
Antimicrobial concentration
Final Expected Sample + or -(%) (ppm) (ppm) (%)
FFN 670 1010 95.2 + 50.7
ERY 5025 4970 98.0 - 2.0
FFN: Florfenicol ERY: Erythromycin
135 4.5 Discussion
The first objective of this study was to determine the effects of prophylactic oral antibiotic treatment on the microbial communities present in rainbow trout intestines in uninfected and F. psychrophilum infected fish. The second objective was to investigate the susceptibility of florfenicol-, erythromycin-treated and untreated fish to F. psychrophilum infection. For the first objective, samples collected on -D11 and D0 p.i. (before and after oral antibiotic treatment) and on D12 and D24 p.i. (after F. psychrophilum infection) were evaluated using culture-independent, high-throughput sequencing (Illumina MiSeq). One hundred and nineteen of the 132 samples prepared were suitable for sequencing. The samples with insufficient
DNA concentration that were excluded from sequencing from the FFN- and ERY-treated groups collected a day after 10 d (D0 p.i.) of prophylactic oral treatment.
Our results showed that Tenericutes, Proteobacteria and Spirochaetes were the most abundant phyla with Mycoplasma, Aeromonas and unclassified Proteobacteria the most abundant genera residing in the intestines of healthy and untreated rainbow trout juveniles. Fusobacteria,
Firmicutes as well as unclassified Bacteria were also identified, but were not detected in all groups. Our findings are consistent with the previous pyrosequencing study of Lowrey et al.
(2015) in which the posterior and anterior intestines of rainbow trout were dominated by
Tenericutes and further, Proteobacteria, Firmicutes, Cyanobacteria, Bacteroidetes, and
Actinobacteria could be detected. In other studies of intestinal microbiota of rainbow trout,
Proteobacteria, Firmicutes, and Actinobacteria were reported to be the most abundant phyla in intestinal microbiota of rainbow trout (Kim et al., 2007; Desai et al., 2012; Navarrete et al.,
2012). Proteobacteria has also been previously identified as the most abundant phylum in rainbow trout intestinal microbiota (Lyons et al., 2017). In the current study, we found
136 Mycoplasma was the predominant genus as did Lowrey et al. (2015). Mycoplasma has also been reported as the most dominant bacterial genera in the distal intestines of wild and farmed salmon
(Holben et al., 2002). Also, Mycoplasma has been reported to be the dominant genus in the intestines of female wild largemouth bronze gudgeon (Coreius guichenoti, 1874) whereas
Shewanella and unclassified Bacteria were more abundant in male fish (Li et al., 2016). Further, in this study, the second most abundant bacteria detected was Aeromonas, a member of the phylum Proteobacteria. Aeromonas is widespread in the environment and is thought to be very prevalent in the intestines of many species of freshwater fishes (Austin, 2006; Gómez and
Balcázar, 2008).
After prophylactic oral treatment, both treated groups showed an increase in the relative abundance of Proteobacteria. We found a statistically significant shift in the intestinal microbiota composition of the FFN-treated group. Sphingomonas, a member of the phylum Proteobacteria was the most abundant bacterial genera this group. On the other hand, the ERY-treated group did not show any significant changes in the intestinal bacterial composition with Tenericutes and
Proteobacteria remaining as the two most abundant bacterial phyla. As well, Mycoplasma remained the most abundant bacterial genus followed by Methylobacterium. An increase in
Proteobacteria was also noted in the untreated groups. Since increase of Proteobacteria occured in all groups on D0 p.i., it is possible that other factors involved in fish rearing environment (e.g., water conditions, tank, etc.) may have lead to the increase of Proteobacteria in all groups at this time point. The FFN-treated group has increased its diversity compared to before the treatment and was more diverse than the ERY-treated group. The FFN-treated group possessed higher
OTUs richness than that of the ERY-treated group after oral treatment. The three most abundant shared OTUs between the two treated groups after 10 d oral treatment were: Mycoplasma,
137 Aeromonas, and Acinetobacter. A previous study has suggested that Sphingomonas, a member of phylum Proteobacteria, is able to hydrolyze FFN (Tao et al., 2012). This may explain the abundance of this genus in the FFN-treated group. The ability of Sphingomonas to degrade FFN is consistent with the notion that it could be a potential source of (acquired) FFN-resistance in other pathogens and could affect the efficacy of FFN in treating BCWD outbreaks. A FFN- resistance gene was first detected from a fish pathogen, Photobacterium damselae ssp. piscicida
(Kim and Aoki, 1996) in 1996, but there is no evidence that this same gene is present in this study. It was observed that Mycoplasma remained the dominant genus in the ERY-treated group and an increase in the abundance of Aeromonas was also observed after the treatment. It is possible that ERY reduced the relative abundance of susceptible bacterial population and created a niche for greater proliferation by Methylobacterium. This bacterium is known to be residing in various natural environments i.e., plants, soil and freshwater aquatic environments (Sanders et al., 2000; Vaneechoutte et al., 2011; Kovaleva et al., 2014). Methylobacterium has been identified as part of skin microbiota in brook charr Salvelinus fontinalis (Boutin et al., 2014).
In contrast, following F. psychrophilum infection, we found that the infected ERY-treated group harbored a more diverse bacterial community (19 & 15 core OTUs on D12 & D24 p.i.) compared to the infected FFN-treated group (13 core OTUs on D12 & D24 p.i. In the infected
FFN-treated group, Acinetobacter (a member of the phylum Proteobacteria) became the most abundant bacterial genus and increased in relative abundance on D24 p.i. Meanwhile, in the
ERY-treated group, Acinetobacter, unclassified Neisseriasseae and Rhizobium were initially identified as the most abundant genera on D12 p.i., while on D24 p.i., Mycoplasma and
Aeromonas became the most abundant bacterial genera residing in the intestines of this group. In the untreated and infected group, Brevinemia and Mycoplasma were the most abundant genera
138 initially while at later time point, the microbiota was mainly dominated by Mycoplasma. These results are, in part, consistent with previous study in which Acinetobacter, Sphingomonas and
Escherichia/Shigella were found to be present in large numbers in the intestinal microbiota of rainbow trout (Wong et al., 2013).
As expected, the cumulative survival rate of the untreated fish group was the lowest of all the groups. However, similar to our findings in Chapter 3, the mortality was less than expected based on preliminary experiments with the same strain and infection dose where FPG101 produced 62% mortality (Chapter 2; Table 2.2). It is possible that the fish were more resistant to
F. psychrophilum experimental infection since the fish used in this study were obtained from a hatchery that was involved in a family-based selective breeding program to increase genetic resistance against BCWD. The relative survival of infected fish groups prophylactically treated with ERY was higher than that of the FFN-treated group which was unexpected. Compared to the untreated fish group, the prophylactically treated fish groups had higher survivability during
F. psychrophilum infection; however, the difference in survivability between all infected groups was insignificant. Our findings showed that prophylactic oral treatment of rainbow trout using
FFN and ERY did not significantly (p > 0.05) affect the fish susceptibility (Figure 4.9) to F. psychrophilum infection, but further work would be required to rigorously demonstrate this.
Determination of the splenic bacterial load quantified using rpoC qPCR assay showed the number of F. psychrophilum in infected fish groups changed over the course of this study. As expected, the bacterial load of untreated fish group increased over time although there was a decrease in bacterial load on D9 p.i. (Figure 4.10). In the prophylactically-treated groups, reductions in the splenic bacterial loads were observed; however, they increased on D12 p.i. The reduction of the bacterial loads in the prophylactically-treated groups may possibly be due to the
139 presence of antibiotic residues that were able to control the bacterial numbers. No significant difference observed in splenic bacterial load on D9 p.i. in any of the infected groups. It is possible the fish immune response has started to clear the bacterial load in the spleen tissue; however, higher splenic bacterial loads were detected in all infected group on D12 p.i. were possibly due to increasing number of F. psychrophilum residing the spleen during the infection period. Also, it is possible that the level of FFN and ERY present in the fish were not able to reduce the bacterial load in the spleen.
There are several factors that influence the diversity of intestinal microbiota. Intestinal microbiota has been reported to be influenced by the type of fish rearing system which may vary over time (Giatsis et al., 2015). We noticed that although the fish were stocked in the same rearing system and held under the same conditions, there were differences in the intestinal microbiota composition of fish groups. For example, there was a difference in the diversity of intestinal microbiota between fish groups prior to antibiotic treatment (-D11 p.i.). Influence of tank surface and tank biofilm has been previously reported (Lyons et al., 2016; Stephens et al.,
2016). Lyons et al. (2016) reported that tank biofilm also influenced the intestinal microbiota composition in trout. Previous studies reported that the intestinal microbiota composition varies over time (Ye et al., 2014; Narrowe et al., 2015; Zarkasi et al., 2016). Similar to previous studies, we observed that the intestinal microbiota of all groups including the control group, changed over time. Our findings reveal that prophylactic oral treatment with FFN and ERY did not significantly (p > 0.05) affect the susceptibility of treated fish to F. psychrophilum infection.
Negative impacts of antibiotic treatment to fish intestinal microbiota have been previously reported as Carlson et al. (2015) reported that rifampicin changed the intestinal microbiota in mosquitofish which increased the fish susceptibility to disease and osmotic stress.
140 Although antimicrobial treatment can reduce or eliminate susceptible pathogens from bacterial communities, it may have a significant negative impact on the composition and functionality of the host’s microbiota. It has been demonstrated to contribute to some climactic dysbiosis and long-term shifts of intestinal microbiota, emergence of antibiotic resistant strains, as well as up-regulation of antibiotic-resistance genes (ARGs) (Jernberg et al., 2007; Buffie et al., 2012). In addition, some opportunistic pathogens may be able to occupy previously unavailable ecological niches due to less competition. Previous studies have shown that genetic material including antimicrobial resistance genes can be transferred from remaining normal endogenous populations to opportunistic or potentially pathogenic allochthonous bacteria in the intestinal tract (Navarrete et al., 2008). Acccordingly, antimicrobial agents should be used with caution. Future studies should be done to characterize the abundance ARGs and intestinal microbial diversity in rainbow trout from different farms/locations to give greater insights into interaction between ARGs and fish intestinal microbiota. More research is also needed to understand the role that alternatives to antimicrobial agents such as probiotics, prebiotics and chemical compounds have in the composition of the intestinal microbiota and in health and disease resistance of rainbow trout.
In conclusion, we identified five phyla (Proteobacteria, Tenericutes, Spirochaetes,
Fusobacteria and Firmicutes) plus unclassified Bacteria were important constituents of the intestinal microbiota in rainbow trout juveniles. In the intestines of healthy and untreated fish,
Tenericutes, Proteobacteria and Spirochaetes were the most abundant phyla with Mycoplasma,
Aeromonas and unclassified Proteobacteria the most abundant genera. Prophylactic antibiotic treatments with ERY (erythromycin) and FFN (florfenicol) had different effects on the intestinal microbial communities of juvenile rainbow trout. ERY did not significantly change the
141 composition of the core bacteria in the intestinal microbiota in contrast with treatment with FFN, which resulted in greater numbers of Sphingomonas after treatment. This shift in the intestinal microbiota composition did not appear to increase the susceptibility of prophylactically treated fish groups to subsequent F. psychrophilum infection, but more studies are needed to ensure that there are no negative effects with FFN or other antimicrobials.
142 CHAPTER 5: General Discussion and Conclusions
The Fish Pathology Laboratory (FPL), Department of Pathobiology, University of
Guelph has been isolating Flavobacterium psychrophilum isolates from various mortality events in Ontario since 1994. A total of 75 of these isolates were phenotypically and genotypically characterized (Hesami et al., 2008); however, their virulence potential was unknown. Virulent and avirulent isolates are required for F. psychrophilum infection and pathogenesis studies. This thesis has identified knowledge gaps that need to be filled to better understand the status of bacterial cold-water disease (BCWD), the most important disease in farmed rainbow trout juveniles in Ontario. The research performed in this thesis provided insights into the a) virulence potential of F. psychrophilum isolates from Ontario in rainbow trout, b) efficacy of oral treatment in controlling F. psychrophilum infection and c) impact of prophylactic treatment on the intestinal microbiota and fish susceptibility to BCWD.
In Chapter 2, the virulence potential of 21 F. psychrophilum isolates from Ontario was investigated in rainbow trout juveniles and splenic bacterial load (determined by plate counting and quantitative polymerase chain reaction (qPCR) assay) was compared with fish mortality. My first hypothesis was that there will be differences in the virulence between F. psychrophilum isolates. To test this hypothesis, a pilot infection trial was conducted using a high dose (108 cfu fish-1) of F. psychrophilum isolates injected by the i.p. route into rainbow trout juveniles. In this trial, I observed that the virulence of 21 F. psychrophilum isolates ranged from 0 to 63% mortality despite the fact that all of the isolates were cultured from BCWD mortality events.
Therefore, the first hypothesis was not rejected. The breadth of the virulence range was surprising; however, particularly as so many isolates produced no mortality. The recent F.
143 psychrophilum strain, FPG101, was the most virulent isolate in rainbow trout juveniles and so, was selected for further studies of BCWD.
The variance in mortality yielded from different experimental trials among F. psychrophilum isolates has been reported to be dependent on a number of factors (i.e., the bacterial strain, route of infection and the fish host) (Madsen and Dalsgaard, 2000; Ekman and
Norrgren, 2003; Silverstein et al., 2008; Marancik et al., 2014; Fredriksen et al., 2016). The route of experimental infection has also been reported to affect mortality produced by F. psychrophilum isolates. Two Norwegian F. psychrophilum isolates produced 0% mortality and
75 - 100% mortality following an i.p. and an i.m. infection, respectively, in rainbow trout
(Fredriksen et al., 2013). I used i.p. injection as the route of infection in my study. Although this route does not mimic the bacterial infection in real life event, it does offer relative reproducibility in experimental infection studies. In a previous study, Decostere et al. (2000) reported the inability of other routes of infection (i.e., immersion, oral and anal challenges) to reproduce F. psychrophilum infection in experimental infection trials. Similarly, our laboratory had attempted to reproduce F. psychrophilum infection using an immersion and scarification method; however, this challenge method was not effective (unpublished data). In my preliminary infection trial, since F. psychrophilum bypassed host mucosal immunity when introduced via i.p. injection, the virulence variability observed was possibly due to the survival of F. psychrophilum in splenic macrophages. Nematollahi et al. (2005) proposed that spleen is as a ‘safer site’ for F. psychrophilum to reside compared to that of the head kidney due to the lower levels of ROS produced by the tissue’s macrophages. In this tissue, high virulence in F. psychrophilum has been reported to be associated with macrophage cytotoxicity and resistance to bactericidal activities by splenic macrophages (Nematollahi et al., 2005).
144 The severity of F. psychrophilum infection has also been linked to the fish host. Fish species, age and genetics are among major factors influencing the virulence of F. psychrophilum.
Several studies have compared the in vivo susceptibility of different fish species to F. psychrophilum (Ekman and Norrgren, 2003; Fredriksen et al., 2013). Rainbow trout had more severe visible lesions, typical of RTFS, compared to Atlantic salmon and sea trout Salmo trutta
L. following an i.p. injection (Ekman and Norrgren, 2003); however, there was no significant difference in mortality between species. In a separate study, higher mortality was observed in rainbow trout than in Atlantic salmon following an i.m. injection using the same F. psychrophilum isolates (Fredriksen et al., 2013). Interestingly, I found that F. psychrophilum
ATCC 49418 which originated from coho salmon and was considered to be avirulent (Cain et al.,
2002), produced relatively low mortality (15%) in the present i.p. infection studies. The developmental stage of fish used in experimental trial has influenced the outcome of infection trial. Since fish fry and fingerlings are more susceptible to BCWD (Holt et al., 1993;
Nematollahi et al., 2003b), they are commonly used to study the virulence of F. psychrophilum in vivo. Ekman and Norrgen (2003) reported a virulent isolate in their study produced 55 - 70% mortality of in three different species of fish fry (~ 0.7 g in weight). In addition, there is a moderate heritability to experimental infection with F. psychophilum (Silverstein et al., 2008).
Therefore, in order to reduce the influence of host in my research, I used genetically similar population of rainbow trout at the same age from the same source.
Maintenance of virulent bacterial strains is crucial when relying on experimental infection as a research tool. Frequent transfer and culture of bacterial isolates potentially increase the risk of changes in the bacterial genotypic, physiology and virulence characteristics (Martinez and Baquero, 2000). Moreover, aged bacterial cultures grown in culture media could contribute
145 to loss of bacterial virulence (Sekowska et al., 2016). In my study, all of the isolates used were kept in Cytophaga broth (CB) supplied with 15% glycerol stock for long-term storage at -80 ˚C since their first isolation from field samples. In order to minimize loss of virulence of these bacterial isolates, I passaged F. psychrophilum isolates no more than three times on CA plates from the bacterial glycerol stock culture. After 96 hr incubation at 15˚C, I harvested the bacterial cells in glycerol stock media and stored them in small aliquots (single use application) prior to -
80 ˚C storage.
My second hypothesis was that fish experiencing morbidity and mortality will harbor significantly higher splenic F. psychrophilum load compared to asymptomatic fish. Following the preliminary trial, a challenge study was done using three different doses (106, 107, and 108 cfu fish-1) of F. psychrophilum FPG101 injected by the i.p. route into rainbow trout juveniles.
The highest infection dose (108 cfu fish-1) produced significantly higher mortality in rainbow trout juveniles (versus the 106 cfu fish-1). The splenic bacterial load was also evaluated over time in the 108 cfu fish-1 group using plate counting and rpoC qPCR. Viable plate counting correlated well with rpoC qPCR results; however, the plate counting method was insignificantly more sensitive than the rpoC qPCR assay. As well, there was a relationship between the splenic bacterial load and mortality with asymptomatic and the morbid fish having splenic bacterial
-1 -1 -1 -1 loads of < 2.8 log10 rpoC copies reaction and > 3.0 log10 rpoC copies reaction , respectively following infection with 108 cfu fish-1. Therefore, the second hypothesis was not rejected.
In the future, studies should be done using different routes of infection such as immersion bath challenge (to better mimic natural infection conditions) using selected virulent
(systemic) and avirulent (superficial) isolates to confirm the virulence of these isolates. Since the avirulent F. psychrophilum ATCC 49418 strain produced some mortality in this study (and was
146 isolated from coho salmon), future research to compare F. psychrophilum survival in macrophages of different fish species should also be performed to determine the host specificity and virulence trait of different isolates. Future research should also include serotyping of F. psychrophilum isolates using multiplex PCR-based serotyping scheme (mPCR) (Rochat et al.,
2017) to study the diversity of isolates from Ontario and association between serotype and virulence. It has been reported that serotype Th and Fd stains are predominant in rainbow trout and associated with disease outbreaks in some countries (Lorenzen and Olesen, 1997; Dalsgaard and Madsen, 2000). They may possibly be more virulent than serotype FpT isolates as demonstrated in experimental infection trial using rainbow trout (Madsen and Dalsgaard, 2000;
Madetoja et al., 2002).
Future studies should also be undertaken to investigate the survival of virulent and avirulent F. psychrophilum isolates identified in the present study in macrophages from head kidney and spleen of rainbow trout fry and juveniles in vitro (Nematollahi et al., 2005). Also, the correlation between the survival of F. psychrophilum in splenic macrophages (in vitro) and fish mortality (in vivo) should be further investigated. If there is a correlation, then it might be possible that an in vitro test could be used as an alternative to infection trial to evaluate virulence potential. The rainbow trout monocyte/macrophage-like cell line (RTS11) would potentially a valuable tool in in vitro studies for host-pathogen interactions (Ganassin and Bols, 1998). Since splenic bacterial load gives information on the status of infection during infection period and the rpoC qPCR assay is not very sensitive, further optimization of the (more sensitive, less specific)
16S rRNA qPCR assay (Orieux et al., 2011) should be done. Finally, since the ability of F. psychrophilum to survive in splenic macrophages may contribute to its virulence potential
(Decostere et al., 2001; Nematollahi et al., 2005) during infection it would be worthwhile to use
147 genome-sequencing data to search for genes or proteins that could be involved in intracellular survival.
Once a BCWD outbreak is identified in the farm, antimicrobial therapy is the treatment and control method of choice for rainbow trout farmers. The antimicrobial susceptibility of 72 F. psychrophilum isolates from Ontario was previously reported (Hesami et al., 2010). Hesami et al.
(2010) found that more than half of the isolates tested had high resistance to FFN and further suggested that most of the isolates from Ontario are most likely susceptible to ERY. However, following their study, the efficacy of antimicrobial treatment of fish infected with F. psychrophilum isolates from Ontario has not been evaluated. Florfenicol is currently a popular antimicrobial agent used to treat BCWD in rainbow trout farms in Ontario. However, excessive use of FFN potentially reduces F. psychrophilum susceptibility to antimicrobial treatment. Since vaccines for BCWD is not available, ERY may offer as another potential antimicrobial agent that can be used to treat the disease and delay resistance to FFN. Accordingly, the third hypothesis of this thesis was that oral treatment with ERY will significantly reduce the mortality of rainbow trout infected with F. psychrophilum compared to FFN. In Chapter 3, the susceptibility of virulent F. psychrophilum isolates to with FFN and ERY was evaluated in vitro using an MIC assay and in vivo using an experimental infection trial. Little MIC value variation was observed with these isolates. It is likely that these recent isolates had similar antimicrobial exposure history at the hatchery level prior to their isolation and/or that some of the isolates were the same strain since they mostly were obtained from the same farm. Molecular strain typing such as whole genome sequencing (WGS) could be used to investigate the genetic relatedness between these isolates (Duchaud et al., 2007; Wiens et al., 2014; Wu et al., 2015; Castillo et al., 2016;).
Two F. psychrophilum isolates (FPG101 and FPG105) with slightly different MIC profiles were
148 selected to investigate the efficacy of FFN and ERY to treat experimentally infected rainbow trout. These in vivo experimental infection and treatment studies showed that although fish groups treated with ERY had the highest survival followed by the FFN-treated and the untreated control groups, the increased survival was not statistically significant. With that, the third hypothesis was rejected for both FPG101- and FPG105-infected fish groups.
In these antibiotic treatment trials, both FPG101 and FPG105 produced less mortality than observed during the preliminary trial (Chapter 2; Jarau et al., 2018). It is possible that the fish used in these experiments were more resistant to F. psychrophilum as they originated from a hatchery that was involved with a family-based selective breeding program (to increase genetic resistance against BCWD). Relative survival with ERY treatment following challenge with isolate FPG105 was also less than predicted, despite the slightly lower MIC values for this isolate compared to FPG101, although the difference in MIC was only a single 10-fold concentration. Treatment of fish with FFN reduced mortality in both trials; however, this was not significant compared to the untreated controls. Since bacterial load contributes to the progress of the infection, early detection and quick action are needed to control the disease in the farm environment. Splenic bacterial loads of FPG101- and FPG105-infected and ERY-treated fish were insignificantly (p > 0.05) reduced over time compared to FFN-treated fish. Although there was delay in the initiation of oral ERY treatment (two days after first mortality observed), it was able to control infection during the treatment period. The first day of treatment was planned to imitate the actual situation at the farm level.
In the future, evaluation of the antimicrobial susceptibility patterns of clinical isolates collected over time might help to gain insights into any changes in the susceptibility F. psychrophilum to antimicrobial agents. Also, more virulent F. psychrophilum isolates with
149 greater variability in MIC profiles are needed to further evaluate levels of antimicrobial resistance and treatment success. Further work to determine the basis of molecular resistance on
FPG101 and FPG105 may be able to fill molecular gaps on the virulence and antimicrobial resistance factors in addition to this study. A traditional experimental infection trial evaluating dose effects on mortality is needed to optimize the use of ERY by fish farmers in the treatment of
BCWD since the dose used was obtained from use in other fish species (i.e., sockeye salmon
Oncorhynchus nerka and Chinook salmon Oncorhynchus tshawytscha). In addition, future work to examine the concentrations of antibiotics in fish tissues (especially the spleen) in fish treated at different doses over time are needed to provide insights into how to reach optimal tissue concentrations as well as the association between tissue concentrations and antimicrobial effects.
Intestinal microbiota play significant roles in the host's development, immunity, nutrition and digestion (Austin, 2006; Moffitt and Mobin, 2006). Antimicrobial administration is one of the several factors that can cause changes in the composition of intestinal microbiota in the host.
Side effects of antibiotic treatments on fish intestinal microbiota (i.e., reduction total microbial population and alteration the bacterial diversity) and fish susceptibility to specific fish pathogen have been previously reported (Moffitt and Mobin, 2006; Navarrete et al., 2008; Carlson et al.,
2015). However, no studies have been reported on the effects of prophylactic oral antibiotic treatment on the intestinal microbiota of juvenile rainbow trout nor on the susceptibility of treated fish to BCWD. Thus, the first aim of Chapter 4 was to evaluate the impact of prophylactic oral antibiotic treatment of rainbow trout on the intestinal microbiome before and after F. psychrophilum infection, and the second aim was to investigate the susceptibility of prophylactically treated fish to F. psychrophilum infection. To address these two aims, the in vivo experimental trial was divided into two periods: i) an antibiotic treatment phase and ii) a F.
150 psychrophilum infection phase. Two antibiotics were used: a) FFN, an antibiotic of choice for treatment of BCWD in Ontario and b) ERY, a possible alternative to FFN. Rainbow trout juveniles were given prophylactic oral antibiotic treatment with either FFN or ERY at 10 mg kg-1 fish body weight day-1 and 75 mg kg-1 of fish body weight day-1, respectively, once daily for 10 consecutive days before being challenged with F. psychrophilum FPG101 via the i.p. route. The intestinal microbiota of rainbow trout was analyzed by 16S rRNA amplicon sequencing
(Klindworth et al., 2013) using Illumina MiSeq technology and the splenic bacterial load was estimated during the infection period using an rpoC qPCR assay (Strepparava et al., 2014). The fourth hypothesis of this thesis was that supplementation of FFN and ERY in the diet will alter the composition of bacterial phyla in the intestinal microbiota of rainbow trout. In this study, prophylactic treatment with FFN significantly affected the relative abundance of the bacterial intestinal microbiota (Tenericutes, Proteobacteria and Spirochaetes) in the treated group, therefore, this hypothesis was not rejected for FFN treatment. Meanwhile, ERY showed little effect on the composition of bacterial phyla in the intestinal microbiota of ERY-treated group, therefore, this hypothesis was was rejected. In addition, five phyla were identified during the study: Proteobacteria, Tenericutes, Spirochaetes, Fusobacteria and Firmicutes and unclassified
Bacteria constituted the majority of the intestinal microbiota in rainbow trout juveniles. I also found that the intestinal microbiota of healthy and untreated juvenile rainbow trout were mainly dominated by the phyla Tenericutes, Proteobacteria and Spirochaetes with Mycoplasma,
Aeromonas and unclassified Proteobacteria the most abundant genera.
The fifth hypothesis of this thesis was that prophylactic oral treatment with FFN or ERY will increase rainbow trout susceptibility to F. psychrophilum infection compared to untreated fish. I found that neither treatment with FFN nor ERY increased fish susceptibility to F.
151 psychrophilum infection and the differences between the survival of all treated and untreated groups were insignificant. With that, the fifth hypothesis was rejected. Little increase in relative survival of treated fish groups suggests minimal impact of the treatment on the intestinal microbiota given the significant change in the microbiota of FFN-treated fish was not correlated with increased susceptibility to F. psychrophilum infection. There may be other factors that potentially affecting the susceptibility of rainbow trout to F. psychrophilum such as host resistance to pathogen, pathogen resistance to treatment, antibiotic dispersion in tissues, etc. As was observed in Chapter 3, it is possible that the fish used in this study had high resistance to
BCWD and so high survival rates were observed even in an untreated fish group. Negative impacts of antibiotic treatment to hosts’ intestinal microbiota have been previously reported. For example, alteration of intestinal microbiota in mosquitofish after treatment with rifampicin increased the fish susceptibility to disease and osmotic stress (Carlson et al., 2015). Treatment with metronidazole has been reported to increase susceptibility of mice to oral Citrobacter rodentium infection (Wlodarska et al., 2011, 2014). It may appear that restoration of the rainbow trout intestinal microbiome to a composition similar to pre-treatment conditions may to take some time after infection with F. psychrophilum. Nonetheless, relying solely upon the use of antimicrobial agents to control BCWD is not desirable as it may potentially give negative impacts on the composition and functionality of the fish microbiota. Previous studies on antibiotic treatment in human and in mice showed community dysbiosis and long-term shifts of intestinal microbiota, emergence of antibiotic-resistant strains, as well as up-regulation of the expression of antibiotic resistance genes (ARGs) (Jernberg et al., 2007; Buffie et al., 2012).
Future studies that elucidate bacterial activities and metabolic profiles before and after treatment would be a great addition to this study. The methods employed in this study were
152 limited to the detection and relative quantitation of bacteria present in the intestinal contents and mucosa of rainbow trout, and the impact of those two antibiotics on bacterial gene expression and metabolic profiles were not investigated. Future study of the abundance of antibiotic- resistance genes (ARGs) and intestinal microbial diversity of rainbow trout from different farms/locations would help to give insights on the prevalence of ARGs in fish intestinal microbiota. Since there was uncertainty in this study as to whether prophylactic treatment with
FFN or ERY increased fish resistance to BCWD, future work should evaluate the intestinal microbiota and fish susceptibility to F. psychrophilum infection during i) prevention (similar to
Chapter 4) and ii) control (treatment after infection, similar to the experimental design used in
Chapter 3) of the BCWD while evaluating antibiotic concentrations in tissues over time. Our findings could be used as a baseline for further studies about the role of core bacterial communities before and after antimicrobial on the fish susceptibility to F. psychrophilum infection.
Like virtually all biological experiments, there a number of limitations to the studies described in this thesis. In order to have a successful infection, fish source and availability is crucial. Obtaining fry and juvenile (~10-40 g) fish which are known to be the most susceptible to
F. psychrophilum infection can be challenging, especially at some times of the year. Fish availability depends on the hatchery spawning schedules so experimental trials involving the desired size were dependent on fish availability. As well, unlike e.g., inbred mice, the resistance of fish to disease (even from the same hatchery) can vary markedly.
The following conclusions can be drawn from:
Chapter 2:
153 a) 21 F. psychrophilum isolates from Ontario produced a broad variability (0-63%) in
mortality in rainbow trout juveniles following an i.p. injection.
b) FPG101 was chosen as the most virulent isolate in our collection and a good candidate
for future experimental infection studies involving BCWD.
c) Isolates obtained originally from systemic tissues were more likely to produce higher
mortality compared to isolates obtained from superficial sites after i.p. injection.
d) Splenic bacterial loads were significantly higher in moribund fish than in those that were
asymptomatic
Chapter 3:
a) Little variance in the MIC profiles of six virulence isolates of F. psychrophilum of
Ontario was observed.
b) Relative to untreated controls, ERY treatment significantly (p < 0.05) reduced mortality
of rainbow trout infected with F. psychrophilum FPG101.
c) Erythromycin was more effective than FFN at splenic bacterial clearance.
d) Even when oral treatment was administered after mortality had begun (similar farming
practices), ERY was as effective as FFN and has a potential for treatment of BCWD
Chapter 4:
a) Proteobacteria, Tenericutes, Spirochaetes, Fusobacteria and Firmicutes and unclassified
Bacteria constituted the major constituents of the intestinal microbiota in rainbow trout
juveniles.
154 b) Tenericutes, Proteobacteria, and Spirochaetes were the three most abundant phyla with
Mycoplasma, Aeromonas and unclassified Proteobacteria being the most abundant
genera residing in the intestines of healthy rainbow trout juveniles.
c) After 10-d prophylactic oral treatment, ERY did not significantly change in the relative
abundance of bacterial phyla, whereas FFN significantly increased (p < 0.05) the relative
abundance of Proteobacteria in the intestines of the treated groups
d) Prophylactic oral treatment with ERY and FFN did not significantly (p > 0.05) affect the
susceptibility of juvenile rainbow trout to subsequent F. psychrophilum infection.
155 REFERENCES
Álvarez, B., Álvarez, J., Ménendez, A., Guijarro, J.A., 2008. A mutant in one of two exbD loci
of a TonB system in Flavobacterium psychrophilum shows attenuated virulence and confers
protection against cold water disease. Microbiology 154, 1144–1151.
Álvarez, B., Guijarro, J.A., 2007. Recovery of Flavobacterium psychrophilum viable cells using
a charcoal-based solid medium. Lett. Appl. Microbiol. 44, 569–572.
Álvarez, B., Secades, P., Prieto, M., McBride, M.J., Guijarro, J.A., 2006. A mutation in
Flavobacterium psychrophilum tlpB inhibits gliding motility and induces biofilm formation.
Appl. Environ. Microbiol. 72, 4044–4053.
Amita, K., Hoshino, M., Honma, T., Wakabayashi, H., 2000. An investigation on the distribution
of Flavobacterium psychrophilum in the Umikawa river. Fish Pathol. 35, 193–197.
Anacker, R.L., Ordal, E.J., 1959. Studies on the myxobacterium Chondrococcus columnaris. I.
Serological typing. J. Bacteriol. 78, 25–32.
Antaya, C., 2008. Current eco-economical impacts of Flavobacterium psychrophilum. MMG 445
Basic Biotechnol. eJournal 4, 16–21.
Aoki, M., Kondo, M., Kawai, K., Oshima, S., 2005. Experimental bath infection with
Flavobacterium psychrophilum, inducing typical signs of rainbow trout Oncorhynchus
mykiss fry syndrome. Dis. Aquat. Organ. 67, 73–79.
Arai, H., Morita, Y., Izumi, S., Katagiri, T., Kimura, H., 2007. Molecular typing by pulsed-field
gel electrophoresis of Flavobacterium psychrophilum isolates derived from Japanese fish. J.
Fish Dis. 30, 345–355.
Austin, B., Austin, D.A., 2007. Epizootiology: Gram-negative Bacteria, in: Bacterial Fish
Pathogens. Springer Netherlands, Dordrecht, pp. 245–282.
156 Austin, B., 2006. The bacterial microflora of fish, revised. Scientific World Journal. 6, 931–945.
Avendaño-Herrera, R., Houel, A., Irgang, R., Bernardet, J.-F., Godoy, M., Nicolas, P., Duchaud,
E., 2014. Introduction, expansion and coexistence of epidemic Flavobacterium
psychrophilum lineages in Chilean fish farms. Vet. Microbiol. 170, 298–306.
Bader, J. A., Shotts, E.B., 1998. Identification of Flavobacterium and Flexibacter species by
species-specific polymerase chain reaction primers to the 16S ribosomal RNA gene. J.
Aquat. Anim. Health 10, 311–319.
Baliarda, A., Faure, D., Urdaci, M.C., 2002. Development and application of a nested PCR to
monitor brood stock salmonid ovarian fluid and spleen for detection of the fish pathogen
Flavobacterium psychrophilum. J. Appl. Microbiol. 92, 510–516.
Bebak, J.A., Welch, T.J., Starliper, C.E., Baya, A.M., Garner, M.M., 2007. Improved husbandry
to control an outbreak of rainbow trout fry syndrome caused by infection with
Flavobacterium psychrophilum. J. Am. Vet. Med. Assoc. 231, 114–116.
Becattini, S., Taur, Y., Pamer, E.G., 2016. Antibiotic-induced changes in the intestinal
microbiota and disease. Trends Mol. Med. 22, 458–478.
Bernardet, J.-F., 2001. Comment: Phylogeny of Flavobacterium and Flexibacter. J. Aquat. Anim.
Health 13, 69–71.
Bernardet, J.-F., 2011. Family Flavobacteriaceae Reichenbach 1989. in: W. Whitman (Ed.).
Bergey's Manual of Systematic Bacteriology 2nd edition Volume 4. The Williams and
Wilkins Co., Baltimore, Maryland.
Bernardet, J.-F., Bowman, J.P., 2006. The Genus Flavobacterium, in: The Prokaryotes. Springer
New York, New York, NY, pp. 481–531.
157 Bernardet, J.-F., Grimont, P.A.D., 1989. Deoxyribonucleic acid relatedness and phenotypic
characterization of Flexibacter columnaris sp. nov., nom. rev., Flexibacter psychrophilus
sp. nov., nom. rev., and Flexibacter maritimus Wakabayashi, Hikida, and Masumura 1986.
Int. J. Syst. Bacteriol. 39, 346–354.
Bernardet, J.-F., Kerouault, B., 1989. Phenotypic and genomic studies of Cytophaga
psychrophila isolated from diseases rainbow trout (Oncorhynchus mykiss) in France. Appl.
Environ. Microbiol. 55, 1796–1800.
Bernardet, J.-F., Nakagawa, Y., 2006. An Introduction to the Family Flavobacteriaceae, in: The
Prokaryotes. Springer New York, New York, NY, pp. 455–480.
Bernardet, J.-F., Segers, P., Vancanneyt, M., Berthe, F., Kersters, K., Vandamme, P., 1996.
Cutting a Gordian Knot: Emended Classification and Description of the Genus
Flavobacterium, Emended Description of the Family Flavobacteriaceae, and Proposal of
Flavobacterium hydatis nom. nov. (Basonym, Cytophaga aquatilis Strohl and Tait 1978).
Int. J. Syst. Bacteriol. 46, 128–148.
Bernardet, J.-F., Baudin-Laurencin, F., Tizerant, G., 1988. First identification of “Cytophaga
psychrophila”; in France. Bull. Eur. Assoc. Fish Pathol. 8, 104–105.
Bertolini, J.M., Wakabayashi, H., Watral, V.G., Whipple, M.J., Rohovec, J.S., 1994.
Electrophoretic detection of proteases from selected strains of Flexibacter psychrophilus
and assessment of their variability. J. Aquat. Anim. Health 6, 224–233.
Blazer, V., Stark, K., Starliper, C., 1996. Unusual histologic manifestations of Flexibacter
psychrophila in hatchery salmonids. in: 21st Annual Eastern Fish Health Workshop,
Gloucester Point, VA, USA. pp. 10.
Borg, A.F.,1948. Studies on Myxobacteria Associated with Diseases in Salmonid Fishes.
158 Doctoral dissertation, University of Washington.
Borg, A.F., 1960. Studies on myxobacteria associated with diseases in salmonid fishes. J. Wildl.
Dis. 8, 1–85.
Boutin, S., Bernatchez, L., Audet, C., Derôme, N., 2013. Network analysis highlights complex
interactions between pathogen, host and commensal microbiota. PLoS One 8, e84772.
doi:10.1371/journal.pone.0084772
Boutin, S., Bernatchez, L., Audet, C., Derôme, N., 2012. Antagonistic effect of indigenous skin
bacteria of brook charr (Salvelinus fontinalis) against Flavobacterium columnare and F.
psychrophilum. Vet. Microbiol. 155, 355–361.
Boutin, S., Sauvage, C., Bernatchez, L., Audet, C., Derome, N., 2014. Inter individual variations
of the fish skin microbiota: host genetics basis of mutualism? PLoS One 9, e102649.
doi:10.1371/journal.pone.0102649
Bowser, P.R., Kosoff, R.E., Chen, C.-Y., Wooster, G.A., Getchell, R.G., Craig, J.L., Lim, P.,
Wetzlich, S.E., Craigmill, A.L., Tell, L.A., 2009. Florfenicol residues in Nile tilapia after
10-d oral dosing in feed: effect of fish size. J. Aquat. Anim. Health 14–17.
Branson, E.J., 1998. Rainbow trout fry syndrome: an update. Fish Vet. J. J. Fish Vet. Soc. 2, 63–
66.
Brown, L.L., Cox, W.T., Levine, R.P., 1997. Evidence that the causal agent of bacterial cold-
water disease Flavobacterium psychrophilum is transmitted within salmonid eggs. Dis.
Aquat. Org. 29, 213–218.
Bruno, D.W., 1992. Cytophaga psychrophila (= ‘Flexibacter psychrophilus’)(Borg),
histopathology associated with mortalities among farmed rainbow trout, Oncorhynchus
mykiss (Walbaum) in the UK. Bull. Eur. Assoc. Fish Pathol. 12, 215–216.
159 Bruun, M.S., Madsen, L., Dalsgaard, I., 2003. Efficiency of oxytetracycline treatment in rainbow
trout experimentally infected with Flavobacterium psychrophilum strains having different in
vitro antibiotic susceptibilities. Aquaculture 215, 11–20.
Bruun, M.S., Schmidt, A.S., Madsen, L., Dalsgaard, I., 2000. Antimicrobial resistance patterns in
Danish isolates of Flavobacterium psychrophilum. Aquaculture 187, 201–212.
Buffie, C.G., Jarchum, I., Equinda, M., Lipuma, L., Gobourne, A., Viale, A., Ubeda, C., Xavier,
J., Pamer, E.G., 2012. Profound alterations of intestinal microbiota following a single dose
of clindamycin results in sustained susceptibility to Clostridium difficile-induced colitis.
Infect. Immun. 80, 62–73.
Bullock, G.L., Hsu, T.C., Shotts, E.B. Jr., 1986. Columnaris disease of fishes. US Fish and
Wildlife Service, Fish Disease Leaflet 72. U.S. Department of the Interior, Fish and
Wildlife Division of Fisheries and Wetland Research, Washington, DC.
Bustos, P.A., Calbuyahue, J., Montaña, J., Opazo, B., Entrala, P., Solervicens, R., 1995. First
isolation of Flexibacter psychrophilus, as causative agent of rainbow trout fry syndrome
(RTFS), producing rainbow trout mortality in Chile. Bull. Eur. Assoc. Fish Pathol. 15, 162–
164.
Cahill, M.M., 1990. Bacterial Flora of Fishes: A Review. Microb. Ecol. 19, 21–41.
Cain, K.D., Grabowski, L., LaPatra, S.E., 2002. Separation and comparison of proteins from
virulent and nonvirulent strains of the fish pathogen Flavobacterium psychrophilum, using a
2-D electrophoretic approach. Bio-Rad Tech. note 2670. Bio-Rad.
Carlson, J.M., Hyde, E.R., Petrosino, J.F., Manage, A.B.W., Primm, T.P., 2015. The host effects
of Gambusia affinis with an antibiotic-disrupted microbiome. Comp. Biochem. Physiol. Part
C, 163–168.
160 Castillo, D., Christiansen, R.H., Dalsgaard, I., Madsen, L., Middelboe, M., 2015. Bacteriophage
resistance mechanisms in the fish pathogen Flavobacterium psychrophilum: linking
genomic mutations to changes in bacterial virulence factors. Appl. Environ. Microbiol. 81,
1157–1167.
Castillo, D., Christiansen, R.H., Espejo, R., Middelboe, M., 2014. Diversity and geographical
distribution of Flavobacterium psychrophilum isolates and their phages: patterns of
susceptibility to phage infection and phage host range. Microb. Ecol. 67, 748–757.
Castillo, D., Chritiansen, R.H., Dalsgaard, I., Madsen, L., Espejo, R., Middelboe, M., 2016.
Comparative genome analysis provides insights into the pathogenicity of Flavobacterium
psychrophilum. PLoS One 11. doi:10.1371/journal.pone.0152515
Castillo, D., Higuera, G., Villa, M., Middelboe, M., Dalsgaard, I., Madsen, L., Espejo, R.T.,
2012. Diversity of Flavobacterium psychrophilum and the potential use of its phages for
protection against bacterial cold water disease in salmonids. J. Fish Dis. 35, 193–201.
Cepeda, C., Garcı́a-Márquez, S., Santos, Y., 2004. Improved growth of Flavobacterium
psychrophilum using a new culture medium. Aquaculture 238, 75–82.
Chakroun, C., Grimont, F., Urdaci, M.C., Bernardet, J.-F., 1998. Fingerprinting of
Flavobacterium psychrophilum isolates by ribotyping and plasmid profiling. Dis. Aquat.
Organ. 33, 167–177.
Chakroun, C., Urdaci, M., Faure, D., Grimont, F., Bernardet, J.-F., 1997. Random amplified
polymorphic DNA analysis provides rapid differentiation among isolates of the fish
pathogen Flavobacterium psychrophilum and among Flavobacterium species. Dis. Aquat.
Organ. 31, 187–196.
161 Chen, Y.-C., Davis, M. a, Lapatra, S.E., Cain, K.D., Snekvik, K.R., Call, D.R., 2008. Genetic
diversity of Flavobacterium psychrophilum recovered from commercially raised rainbow
trout, Oncorhynchus mykiss (Walbaum), and spawning coho salmon, O. kisutch (Walbaum).
J. Fish Dis. 31, 765–773.
Cipriano, R.C., Holt, R.A., 2005. Flavobacterium psychrophilum, cause of Bacterial Cold-Water
Disease and Rainbow Trout Fry Syndrome. Fish Dis. Leafl. U.S. Dept. Inter. 1–44.
Cipriano, R.C., Schill, W.B., Teska, J.D., Ford, L.A., 1996. Epizootiological study of bacterial
cold-water disease in Pacific Salmon and further characterization of the etiologic agent,
Flexibacter psychrophila. J. Aquat. Anim. Health 8, 28–36.
CLSI, 2014. Methods for Broth Dilution Susceptibility Testing for Bacteria Isolated from
Aquatic Animals: Approved Guideline. Clinical and Laboratory Standards Institute,
Document VET04-A2. Clinical and Laboratory Standards Institute, Wayne, Pennsylvania.
Costerton, J.W., 1999. Bacterial biofilms: A common cause of persistent infections. Science.
284, 1318–1322.
Crumlish, M., Diab, A.M., George, S., Ferguson, H.W., 2007. Detection of the bacterium
Flavobacterium psychrophilum from a natural infection in rainbow trout, Oncorhynchus
mykiss (Walbaum), using formalin-fixed, wax-embedded fish tissues. J. Fish Dis. 30, 37–41.
Crump, E.M., Perry, M.B., Clouthier, S.C., Kay, W.W., 2001. Antigenic characterization of the
fish pathogen Flavobacterium psychrophilum. Appl. Environ. Microbiol. 67, 750–759.
Crump, E.M., Perry, M.B., Gale, S., Crawford, E., Kay, W.W., 2003. Lipopolysaccharide O-
antigen antibody-based detection of the fish pathogen Flavobacterium psychrophilum. J.
Mol. Microbiol. Biotechnol. 6, 182–190.
162 Dalsgaard, I., 1993. Virulence mechanisms in Cytophaga psychrophila and other Cytophaga-like
bacteria pathogenic for fish. Annu. Rev. Fish Dis. 3, 127–144.
Dalsgaard, I., Madsen, L., 2000. Bacterial pathogens in rainbow trout, Oncorhynchus mykiss
(Walbaum), reared at Danish freshwater farms. J. Fish Dis. 23, 199–209.
Daskalov, H., Austin, D. A., Austin, B., 1999. An improved growth medium for Flavobacterium
psychrophilum. Lett. Appl. Microbiol. 28, 297–299.
Davis, H.S., 1946. Care and Diseases of Trout. United States Fish and Wildlife Service Research
Report 12. Washington, DC.
Decostere, A., D’Haese, E., Lammens, M., Nelis, H., Haesebrouck, F., 2001. In vivo study of
phagocytosis, intracellular survival and multiplication of Flavobacterium psychrophilum in
rainbow trout, Oncorhynchus mykiss (Walbaum), spleen phagocytes. J. Fish Dis. 24, 481–
487.
Decostere, A., Lammens, M., Haesebrouck, F., 2000. Difficulties in experimental infection
studies with Flavobacterium psychrophilum in rainbow trout (Oncorhynchus mykiss) using
immersion, oral and anal challenges. Res. Vet. Sci. 69, 165–169.
Del Cerro, A., Marquez, I., Guijarro, J.A., 2002. Simultaneous detection of Aeromonas
salmonicida, Flavobacterium psychrophilum, and Yersinia ruckeri, three major fish
pathogens, by multiplex PCR. Appl. Environ. Microbiol. 68, 5177–5180.
Del Cerro, A., Márquez, I., Prieto, J.M., 2010. Genetic diversity and antimicrobial resistance of
Flavobacterium psychrophilum isolated from cultured rainbow trout, Onchorynchus mykiss
(Walbaum), in Spain. J. Fish Dis. 33, 285–291.
163 Del Cerro, A., Mendoza, M.C., Guijarro, J.A., 2002. Usefulness of a TaqMan-based polymerase
chain reaction assay for the detection of the fish pathogen Flavobacterium psychrophilum.
J. Appl. Microbiol. 93, 149–156.
Desai, A.R., Links, M.G., Collins, S.A., Mansfield, G.S., Drew, M.D., Van Kessel, A.G., Hill,
J.E., 2012. Effects of plant-based diets on the distal gut microbiome of rainbow trout
(Oncorhynchus mykiss). Aquaculture 350–353, 134–142.
Duchaud, E., Boussaha, M., Loux, V., Bernardet, J.-F., Michel, C., Kerouault, B., Mondot, S.,
Nicolas, P., Bossy, R., Caron, C., Bessières, P., Gibrat, J.-F., Claverol, S., Dumetz, F., Le
Hénaff, M., Benmansour, A., 2007. Complete genome sequence of the fish pathogen
Flavobacterium psychrophilum. Nat. Biotechnol. 25, 763–769.
Durmaz, Y., Onuk, E.E., Ciftci, A., 2012. Investigation of the presence and antibiotic
susceptibilities of Flavobacterium psychrophilum in rainbow trout farms (Oncorhynchus
mykiss Walbaum, 1792) in The Middle and Eastern Black Sea Regions of Turkey. Ankara
Univ. Vet. Fak. Derg. 59, 141–146.
Edgar, R.C., Haas, B.J., Clemente, J.C., Quince, C., Knight, R., 2011. UCHIME improves
sensitivity and speed of chimera detection. Bioinformatics 27, 2194–2200.
Ekman, E., 2003. Natural and experimental infections with Flavobacterium psychrophilum in
salmonid fish. PhD thesis. Swedish University of Agricultural Sciences.
Ekman, E., Åkerman, G., Balk, L., Norrgren, L., 2003. Nanoinjection as a tool to mimic vertical
transmission of Flavobacterium psychrophilum in rainbow trout Oncorhynchus mykiss. Dis.
Aquat. Organ. 55, 93–99.
164 Ekman, E., Norrgren, L., 2003. Pathology and immunohistochemistry in three species of
salmonids after experimental infection with Flavobacterium psychrophilum. J. Fish Dis. 26,
529–538.
Elliot, D.G., 2017. Fish Viruses and Bacteria: Pathobiology and Protection, in: Woo, P.T.K.,
Cipriano, R.C. (Eds.), Fish Viruses and Bacteria: Pathobiology and Protection. CABI,
Boston, MA, pp. 286–297.
Elliot, D.G., Pascho, R.J., Bullock, G.L., 1989. Developments in the control of bacterial kidney
disease of salmonid fishes. Dis. Aquat. Organ. 6, 201–215.
Esposito, A., Fabrizi, L., Lucchetti, D., Marvasi, L., Coni, E., Guandalini, E., 2007. Orally
administered erythromycin in rainbow trout (Oncorhynchus mykiss): residues in edible
tissues and withdrawal time. Antimicrob. Agents Chemother. 51, 1043–1047.
Evensen, Ø., Lorenzen, E., 1996. An immunohistochemical study of Flexibacter psychrophilus
infection in experimentally and naturally infected rainbow trout (Oncorhynchus mykiss) fry.
Dis. Aquat. Organ. 25, 53–61.
Evensen Ø., Lorenzen E., 1997. Simultaneous demonstration of infectious pancreatic necrosis
virus (IPNV) and Flavobacterium psychrophilum in paraffin-embedded specimens of
rainbow trout Oncorhynchus mykiss fry by use of paired immunohistochemistry. Dis Aquat
Organ. 29, 227–232.
Fan, L., Chen, J., Meng, S., Song, C., Qiu, L., Hu, G., Xu, P., 2017. Characterization of
microbial communities in intensive GIFT tilapia (Oreochromis niloticus) pond systems
during the peak period of breeding. Aquac. Res. 48, 459–472.
Faruk, M.A.R., 2002. A Review on Rainbow Trout Fry Syndrome (RTFS). Pakistan J. Biol. Sci.
5, 230–233.
165 Fredriksen, B.N., Furevik, A., Olsen, R.H., Gauthier, D., Mendoza, J., Norderhus, E.-A., 2016.
Virulence of Chilean field isolates of Flavobacterium psychrophilum in Atlantic salmon
(Salmo salar L .) parr. Bull. Eur. Assoc. Fish Pathol. 36, 67–75.
Gajardo, K., Rodiles, A., Kortner, T.M., Krogdahl, Å., Bakke, A.M., Merrifield, D.L., Sørum,
H., 2016. A high-resolution map of the gut microbiota in Atlantic salmon (Salmo salar): A
basis for comparative gut microbial research. Sci. Rep. 6, 30893. doi:10.1038/srep30893
Ganassin, R.C., Bols, N.C., 1998. Development of a monocyte/macrophage-like cell line,
RTS11, from rainbow trout spleen. Fish Shellfish Immunol. 8, 457–476.
Gaunt, P.S., Langston, C., Wrzesinski, C., Gao, D., Adams, P., Crouch, L., Sweeney, D., Endris,
R., 2013. Multidose pharmacokinetics of orally administered florfenicol in the channel
catfish (Ictalurus punctatus). J. Vet. Pharmacol. Ther. 36, 502–506.
Giatsis, C., Sipkema, D., Smidt, H., Heilig, H., Benvenuti, G., Verreth, J., Verdegem, M., 2015.
The impact of rearing environment on the development of gut microbiota in tilapia larvae.
Sci. Rep. 5, 18206. doi:10.1038/srep18206
Gieseker, C.M., Crosby, T.C., Mayer, T.D., Bodeis, S.M., Stine, C.B., 2016. Development of
similar broth microdilution methods to determine the antimicrobial susceptibility of
Flavobacterium columnare and F. psychrophilum. J. Aquat. Anim. Health 28, 27–38.
Gieseker, C.M., Mayer, T.D., Crosby, T.C., Carson, J., Dalsgaard, I., Darwish, A.M., Gaunt,
P.S., Gao, D.X., Hsu, H.-M., Lin, T.L., Oaks, J.L., Pyecroft, M., Teitzel, C., Somsiri, T.,
Wu, C.C., 2012. Quality control ranges for testing broth microdilution susceptibility of
Flavobacterium columnare and F. psychrophilum to nine antimicrobials. Dis. Aquat. Organ.
101, 207–215.
166 Gómez, G.D., Balcázar, J.L., 2008. A review on the interactions between gut microbiota and
innate immunity of fish. FEMS Immunol Med Microbiol 52, 145–154.
Green, T.J., Smullen, R., Barnes, A.C., 2013. Dietary soybean protein concentrate-induced
intestinal disorder in marine farmed Atlantic salmon, Salmo salar is associated with
alterations in gut microbiota. Vet. Microbiol. 166, 286–292.
Groff, J.M., Lapatra, S.E., 2000. Infectious diseases impacting the commercial culture of
salmonids. J. Appl. Aquac. 10, 17–90.
Gultepe, N., Tanrikul, T.T., 2006. Treatment methods of Flavobacterium psychrophilum: cause
of rainbow trout fry syndrome (RFTS) and bacterial cold-water disease (BCWD) in Turkey.
J. Fish. Int. 1, 102–105.
Hadidi, S., Glenney, G.W., Welch, T.J., Silverstein, J.T., Wiens, G.D., 2008. Spleen size predicts
resistance of rainbow trout to Flavobacterium psychrophilum challenge. J. Immunol. 180,
4156–4165.
Hatton, G.B., Madla, C.M., Rabbie, S.C., Basit A.W., 2018. All disease begins in the gut:
Influence of gastrointestinal disorders and surgery on oral drug performance. Int. J. Pharm.
548, 408–422. doi:10.1016/j.ijpharm.2018.06.054
He, S., Zhou, Z., Liu, Y., Cao, Y., Meng, K., Shi, P., Yao, B., Ringø, E., 2010. Effects of the
antibiotic growth promoters flavomycin and florfenicol on the autochthonous intestinal
microbiota of hybrid tilapia (Oreochromis niloticus ♀ × O. aureus ♂). Arch. Microbiol.
192, 985–994.
Henriksen, M.M.M., Dalsgaard, I., Lorenzen, N., Buchmann, N., Kania, P.W., 2013.
Flavobacterium psychrophilum -Experimental challenge and immune response. DTU,
Copenhagen.
167 Henríquez-Núñez, H., Evrard, O., Kronvall, G., Avendaño-Herrera, R., 2012. Antimicrobial
susceptibility and plasmid profiles of Flavobacterium psychrophilum strains isolated in
Chile. Aquaculture 354–355, 38–44.
Henryon, M., Berg, P., Olesen, N.J., Kjær, T.E., Slierendrecht, W.J., Jokumsen, A., Lund, I.,
2005. Selective breeding provides an approach to increase resistance of rainbow trout
(Onchorhynchus mykiss) to the diseases, enteric redmouth disease, rainbow trout fry
syndrome, and viral haemorrhagic septicaemia. Aquaculture 250, 621–636.
Hesami, S., Allen, K.J., Metcalf, D., Ostland, V.E., MacInnes, J.I., Lumsden, J.S., 2008.
Phenotypic and genotypic analysis of Flavobacterium psychrophilum isolates from Ontario
salmonids with bacterial coldwater disease. Can J Microbiol 54, 619–629.
Hesami, S., Parkman, J., MacInnes, J.I., Gray, J.T., Gyles, C.L., Lumsden, J.S., 2010.
Antimicrobial susceptibility of Flavobacterium psychrophilum isolates from Ontario. J.
Aquat. Anim. Health 22, 39–49.
Högfors-Rönnholm, E., Norrgård, J., Wiklund, T., 2015. Adhesion of smooth and rough
phenotypes of Flavobacterium psychrophilum to polystyrene surfaces. J. Fish Dis. 38, 429–
437.
Högfors-Rönnholm, E., Wiklund, T., 2010. Phase variation in Flavobacterium psychrophilum:
characterization of two distinct colony phenotypes. Dis. Aquat. Organ. 90, 43–53.
Holben, W.E., Williams, P., Saarinen, M., Särkilahti, L.K., Apajalahti, J.H.A., 2002.
Phylogenetic analysis of intestinal microfora indicates a novel Mycoplasma phylotype in
farmed and wild salmon. Microb. Ecol. 44, 175–185.
Holt, R.A., 1987. Cytophaga psychrophila, the causative agent of bacterial cold-water disease in
salmonid fish. PhD thesis. Oregon State University.
168 Holt, R.A., Amandi, A., Rohovec, J.S., Fryer, J.L., 1989. Relation of water temperature to
bacterial cold-water disease in coho salmon, chinook salmon, and rainbow trout. J. Aquat.
Anim. Health 1, 94–101.
Holt, R.A., Rohovec J.S., Fryer J.L. 1993. Bacterial Coldwater Disease. in: Inglis V., Roberts
R.J., Bromange N.R. (Eds.) Bacterial Disease of Fish. Blackwell Scientific, Oxford, pp 2–
23.
Huber, I., Spanggaard, B., Appel, K.F., Rossen, L., Nielsen, T., Gram, L., 2004. Phylogenetic
analysis and in situ identification of the intestinal microbial community of rainbow trout
(Oncorhynchus mykiss, Walbaum). J. Appl. Microbiol. 96, 117–132.
Hunnicutt, D.W., McBride, M.J., 2000. Cloning and characterization of the Flavobacterium
johnsoniae gliding-motility genes gldB and gldC. J. Bacteriol. 182, 911–918.
Hunter, J., Hirst, B.H., 1997. Intestinal secretion of drugs. The role of P-glycoprotein and related
drug efflux systems in limiting oral drug absorption. Adv. Drug Deliv. Rev. 25, 129–157.
Huq, A., Whitehouse, C.A., Grim, C.J., Alam, M., Colwell, R.R., 2008. Biofilms in water, its
role and impact in human disease transmission. Curr. Opin. Biotechnol. 19, 244–247.
Iida, Y., Mizokami, A., 1996. Outbreaks of coldwater disease in wild ayu and pale chub. Fish
Pathol. 31, 157–164.
Ikegami, A., Honma, K., Sharma, A., Kuramitsu, H.K., 2004. Multiple functions of the leucine-
rich repeat protein LrrA of Treponema denticola. Infect. Immun. 72, 4619–4627.
Ingerslev, H.-C., Strube, M.L., Jørgensen, L. von G., Dalsgaard, I., Boye, M., Madsen, L., 2014a.
Diet type dictates the gut microbiota and the immune response against Yersinia ruckeri in
rainbow trout (Oncorhynchus mykiss). Fish Shellfish Immunol. 40, 624–633.
Ingerslev, H.-C., von Gersdorff Jørgensen, L., Lenz Strube, M., Larsen, N., Dalsgaard, I., Boye,
169 M., Madsen, L., 2014b. The development of the gut microbiota in rainbow trout
(Oncorhynchus mykiss) is affected by first feeding and diet type. Aquaculture 424–425, 24–
34.
Inglis, V., Richards, R.H., Varma, K.J., Sutherland, I.H., Brokken, E.S., 1991. Florfenicol in
Atlantic salmon, Salmo salar L., parr: tolerance and assessment of efficacy against
furunculosis. J. Fish Dis. 14, 343–351.
Izumi, S., Aranishi, F., Wakabayashi, H., 2003. Genotyping of Flavobacterium psychrophilum
using PCR-RFLP analysis. Dis. Aquat. Organ. 56, 207–214.
Izumi, S., Fujii, H., Aranishi, F., 2005. Detection and identification of Flavobacterium
psychrophilum from gill washings and benthic diatoms by PCR-based sequencing analysis.
J. Fish Dis. 28, 559–564.
Izumi, S., Ouchi, S., Kuge, T., Arai, H., Mito, T., Fujii, H., Aranishi, F., Shimizu, A., 2007.
PCR-RFLP genotypes associated with quinolone resistance in isolates of Flavobacterium
psychrophilum. J. Fish Dis. 30, 141–147.
Jarau, M., Di Natale, A., Huber, P.E., MacInnes, J.I., Lumsden, J.S., 2018. in press. Virulence of
Flavobacterium psychrophilum isolates in rainbow trout Oncorhynchus mykiss
(Walbaum). J. Fish Dis. doi:10.1111/jfd.12861.
Jernberg, C., Löfmark, S., Edlund, C., Jansson, J.K., 2007. Long-term ecological impacts of
antibiotic administration on the human intestinal microbiota. ISME J. 1, 56–66.
Kawai, Y., Ishida Okawara, A., Okuyama, H., Kura, F., Suzuki, K., 2000. Modulation of
chemotaxis, O2− production and myeloperoxidase release from human polymorphonuclear
leukocytes by the ornithine-containing lipid and the serineglycine-containing lipid of
Flavobacterium. FEMS Immunol. Med. Microbiol. 28, 205–209.
170 leukocytes by the ornithine-containing lipid and the serineglycine-containing lipid of
Flavobacterium. FEMS Immunol. Med. Microbiol. 28, 205–209.
Kent, M.L., Groff, J.M., Yasutake, W.T., Holt, R.A., 1989. Spiral swimming behavior due to
cranial and vertebral lesions associated with Cytophaga psychrophila infections in salmonid
fishes. Dis. Aquat. Organ. 6, 11–16.
Kim, D.-H., Brunt, J., Austin, B., 2007. Microbial diversity of intestinal contents and mucus in
rainbow trout (Oncorhynchus mykiss). J. Appl. Microbiol. 102, 1654–1664.
Kim, E.-H., Aoki, T., 1996. Sequence analysis of the florfenicol resistance gene encoded in the
transferable R-plasmid of a fish pathogen, Pasteurella piscicida. Microbiol. Immunol. 40,
665–669.
Kim, J.H., Gomez, D.K., Nakai, T., Park, S.C., 2010. Isolation and identification of
bacteriophages infecting ayu Plecoglossus altivelis altivelis specific Flavobacterium
psychrophilum. Vet. Microbiol. 140, 109–115.
Klindworth, A., Pruesse, E., Schweer, T., Peplies, J., Quast, C., Horn, M., Glöckner, F.O., 2013.
Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-
generation sequencing-based diversity studies. Nucleic Acids Res. 41, e1.
doi:10.1093/nar/gks808
Kovaleva, J., Degener, J.E., van der Mei, H.C., 2014. Methylobacterium and its role in health
care-associated infection. J. Clin. Microbiol. 52, 1317–1321. doi:10.1128/JCM.03561-13
Kondo, M., Kawai, K., Kurohara, K., Oshima, S., 2002. Adherence of Flavobacterium
psychrophilum on the body surface of the ayu Plecoglossus altivelis. Microbes Infect. 4,
279–283.
171 Kühlwein, H., Emery, M.J., Rawling, M.D., Harper, G.M., Merrifield, D.L., Davies, S.J., 2013.
Effects of a dietary β-(1,3)(1,6)-D-glucan supplementation on intestinal microbial
communities and intestinal ultrastructure of mirror carp (Cyprinus carpio L.). J. Appl.
Microbiol. 115, 1091–1106.
Kum, C., Kirkan, S., Sekkin, S., Akar, F., Boyacioglu, M., 2008. Comparison of in vitro
antimicrobial susceptibility in Flavobacterium psychrophilum isolated from rainbow trout
fry. J. Aquat. Anim. Health 20, 245–251.
Kumagai, A., Nawata, A., Taniai, Y., 2010. Monitoring of outbreaks of bacterial cold water
disease among ayu in a river where asymptomatic carriers of Flavobacterium
psychrophilum were released. Fish Pathol. 45, 115–120.
LaFrentz, B.R., Cain, K.D., 2004. Bacterial Coldwater Disease An Extension Bulletin for the
Western Regional Aquaculture Center (WRAC). WRAC Publications. Moscow, Idaho.
LaFrentz, B.R., LaPatra, S.E., Call, D.R., Wiens, G.D., Cain, K.D., 2011. Identification of
immunogenic proteins within distinct molecular mass fractions of Flavobacterium
psychrophilum. J. Fish Dis. 34, 823–830.
LaFrentz, B.R., LaPatra, S.E., Call, D.R., Wiens, G.D., Cain, K.D., 2009. Proteomic analysis of
Flavobacterium psychrophilum cultured in vivo and in iron-limited media. Dis. Aquat.
Organ. 87, 171–182.
Lammens, M., Decostere, A., Haesebrouck, F., 2000. Effect of Flavobacterium psychrophilum
strains and their metabolites on the oxidative activity of rainbow trout Oncorhynchus mykiss
phagocytes. Dis. Aquat. Organ. 41, 173–179.
172 LaPatra, S.E., Fehringer, T.R., Cain, K.D., 2014. A probiotic Enterobacter sp. provides
significant protection against Flavobacterium psychrophilum in rainbow trout
(Oncorhynchus mykiss) after injection by two different routes. Aquaculture 433, 361–366.
Larsen, A.M., Mohammed, H.H., Arias, C.R., 2014. Characterization of the gut microbiota of
three commercially valuable warmwater fish species. J. Appl. Microbiol. 116, 1396–1404.
Lee, E.G.-H., Evelyn, T.P.T., 1994. Prevention of vertical transmission of the bacterial kidney
disease agent Renibacterium salmoninarum by broodstock injection with erythromycin. Dis.
Aquat. Organ. 18, 1–4.
Lee, K.-B., Heo, G.-J., 1998. First isolation and identification of Cytophaga psychrophila from
cultured ayu in Korea. Fish Pathol. 33, 37–38.
Lehmann, J., Mock, D., Stürenberg, F.-J., Bernardet, J.-F., 1991. First isolation of Cytophaga
psychrophila from a systemic disease in eel and cyprinids. Dis. Aquat. Org. 10, 217–220.
Levipan, H.A., Avendaño-Herrera, R., 2017. Different phenotypes of mature biofilm in
Flavobacterium psychrophilum share a potential for virulence that differs from planktonic
state. Front. Cell. Infect. Microbiol. 7, 76. doi:10.3389/fcimb.2017.00076
Levipan, H.A., Quezada, J., Avendaño-Herrera, R., 2018. Stress tolerance-related genetic traits
of fish pathogen Flavobacterium psychrophilum in a mature biofilm. Front. Microbiol. 9,
18. doi:10.3389/fmicb.2018.00018
Li, T., Long, M., Gatesoupe, & F.-J., Zhang, Q., Li, A., Gong, X., 2015. Comparative analysis of
the intestinal bacterial communities in different species of carp by pyrosequencing.
Microbiol. Aquat. Syst. 25–36.
173 Li, X., Yan, Q., Ringø, E., Wu, X., He, Y., Yang, D., 2016. The influence of weight and gender
on intestinal bacterial community of wild largemouth bronze gudgeon (Coreius guichenoti,
1874). BMC Microbiol. 16, 191. doi:10.1186/s12866-016-0809-1
Lin, C.-H., Chen, C.-H., Lin, Z.-C., Fang, J.-Y., 2017. Recent advances in oral delivery of drugs
and bioactive natural products using solid lipid nanoparticles as the carriers. J. Food Drug
Anal. 25, 219–234. doi:10.1016/j.jfda.2017.02.001
Lindstrom, N.M., Call, D.R., House, M.L., Moffitt, C.M., Cain, K.D., 2009. A quantitative
enzyme-linked immunosorbent assay and filtration-based fluorescent antibody test as
potential tools to screen broodstock for infection with Flavobacterium psychrophilum. J.
Aquat. Anim. Health 21, 43–56.
Liu, Y., Zhou, Z., Wu, N., Tao, Y., Xu, L., Cao, Y., Zhang, Y., Yao, B., 2012. Gibel carp
Carassius auratus gut microbiota after oral administration of trimethoprim/
sulfamethoxazole. Dis. Aquat. Organ. 99, 207–213.
Llewellyn, M.S., Boutin, S., Hoseinifar, S.H., Derome, N., 2014. Teleost microbiomes: the state
of the art in their characterization, manipulation and importance in aquaculture and
fisheries. Front. Microbiol. 5, 1–17.
Loch, T.P., Fujimoto, M., Woodiga, S.A., Walker, E.D., Marsh, T.L., Faisal, M., 2013. Diversity
of fish-associated Flavobacteria of Michigan. J. Aquat. Anim. Health 25, 149–164.
Long, A., Call, D.R., Cain, K.D., 2014. Investigation of the link between broodstock infection,
vertical transmission, and prevalence of Flavobacterium psychrophilum in eggs and
progeny of rainbow trout and coho salmon. J. Aquat. Anim. Health 26, 66–77.
doi:10.1080/08997659.2014.886632
174 Long, A., Polinski, M.P., Call, D.R., Cain, K.D., 2012. Validation of diagnostic assays to screen
broodstock for Flavobacterium psychrophilum infections. J. Fish Dis. 35, 407–419.
Lorenzen, E., 1994. Study on Flexibacter psychrophilus in Relation to Rainbow Trout Fry
Syndrome (RTFS). PhD thesis. Royal Veterinary and Agricultural University, Copenhagen.
Lorenzen, E., Dalsgaard, I., Bernardet, J.-F., 1997. Characterization of isolates of
Flavobacterium psychrophilum associated with coldwater disease or rainbow trout fry
syndrome I: phenotypic and genomic studies. Dis. Aquat. Organ. 31, 197–208.
Lorenzen, E., Dalsgaard, I., From, J., Hansen, E.M., Hørlyck, V., Korsholm, H., Mellergaard, S.,
Olesen, N.J., 1991. Preliminary investigations of fry mortality syndrome in rainbow trout.
Bull. Eur. Ass. Fish Pathol. 11, 77–79.
Lorenzen, E., Karas, N., 1992. Detection of Flexibacter psychrophilus by immunofluorescence
in fish suffering from fry mortality syndrome: a rapid diagnostic method. Dis. Aquat.
Organ. 13, 231–234.
Lorenzen, E., Olesen, N.J., 1997. Characterization of isolates of Flavobacterium psychrophilum
associated with coldwater disease or rainbow trout fry syndrome II: Serological studies. Dis.
Aquat. Organ. 31, 209–220.
Lorenzen, K., Vannini, A., Cramer, P., Heck, A.J.R., 2007. Structural biology of RNA
polymerase III: mass spectrometry elucidates subcomplex architecture. Struct. 15, 1237–
1245.
Lorenzen, L., van Deventer, J.S., Landi, W., 1995. Factors affecting the mechanism of the
adsorption of arsenic species on activated carbon. Miner. Eng. 8, 557–569.
175 Lowrey, L., Woodhams, D.C., Tacchi, L., Salinas, I., 2015. Topographical mapping of the
rainbow trout (Oncorhynchus mykiss) microbiome reveals a diverse bacterial community
with antifungal properties in the skin. Appl. Environ. Microbiol. 81, 6915–6925.
Lumsden, J.S., Ostland, V.E., Ferguson, H.W., 1996. Necrotic myositis in cage cultured rainbow
trout, Oncorhynchus mykiss (Walbaum), caused by Flexibacter psychrophilus. J. Fish Dis.
19, 113–119.
Lumsden, J.S., Young, K., MacInnes, J., Russell, S., Hesami, S., 2004. Management Approaches
for Coldwater DIsease Caused by Flavobacterium psychrophilum, in: Gilbert, É., Stechey,
D., Struthers, M. (Eds.), Proceedings of the Canadian Freshwater Aquaculture Symposium-
AAC Special Publication. Aquaculture Association of Canada, Quebec City, Quebec, pp.
111–117.
Lyons, P.P., Turnbull, J.F., Dawson, K.A., Crumlish, M., 2017. Exploring the microbial diversity
of the distal intestinal lumen and mucosa of farmed rainbow trout Oncorhynchus mykiss
(Walbaum) using next generation sequencing (NGS). Aquac. Res. 48, 77–91.
Lyons, P.P., Turnbull, J.F., Dawson, K.A., Crumlish, M., 2016. Phylogenetic and functional
characterization of the distal intestinal microbiome of rainbow trout Oncorhynchus mykiss
from both farm and aquarium settings. Appl. Microbiol. 122, 347–363.
Madetoja, J., Hänninen, M.-L., Koski, V.H., Dalsgaard, I., Wiklund, T., 2001. Phenotypic and
genotypic characterization of Flavobacterium psychrophilum from Finnish fish farms. J.
Fish Dis. 24, 469–479.
Madetoja, J., Nyman, P., Wiklund, T., 2000. Flavobacterium psychrophilum, invasion into and
shedding by rainbow trout Oncorhynchus mykiss. Dis. Aquat. Organ. 43, 27–38.
176 Madetoja, J., Wiklund, T., 2002. Detection of the fish pathogen Flavobacterium psychrophilum
in water from fish farms. Syst. Appl. Microbiol. 25, 259–266.
Madsen, L., Arnbjerg, J., Dalsgaard, I., 2001. Radiological examination of the spinal column in
farmed rainbow trout Oncorhynchus mykiss (Walbaum): experiments with Flavobacterium
psychrophilum and oxytetracycline. Aquac. Res. 32, 235–241.
Madsen, L., Dalsgaard, I., 2000. Comparative studies of Danish Flavobacterium psychrophilum
isolates: Ribotypes, plasmid profiles, serotypes and virulence. J. Fish Dis. 23, 211–218.
Madsen, L., Dalsgaard, I., 1999. Vertebral column deformities in farmed rainbow trout
(Oncorhynchus mykiss). Aquaculture 171, 41–48.
Mansfield, G.S., Desai, A.R., Nilson, S.A., Van Kessel, A.G., Drew, M.D., Hill, J.E., 2010.
Characterization of rainbow trout (Oncorhynchus mykiss) intestinal microbiota and
inflammatory marker gene expression in a recirculating aquaculture system. Aquaculture
307, 95–104.
Marancik, D.P., Wiens, G.D., 2013. A real-time polymerase chain reaction assay for
identification and quantification of Flavobacterium psychrophilum and application to
disease resistance studies in selectively bred rainbow trout Oncorhynchus mykiss. FEMS
Microbiol. Lett. 339, 122–129.
Martinez, J.L., Baquero, F., 2000. Mutation frequencies and antibiotic resistance. Antimicrob.
Agents Chemother. 44, 1771–1777. doi:10.1128/AAC.44.7.1771-1777.2000
Martin-Antonio, B., Manchado, M., Infante, C., Zerolo, R., Labella, A., Alonso, C., Borrego,
J.J., 2007. Intestinal microbiota variation in Senegalese sole (Solea senegalensis) under
different feeding regimes. Aquac. Res. 38, 1213–1222.
177 Martínez, J.L., Casado, A., Enríquez, R., 2004. Experimental infection of Flavobacterium
psychrophilum in fins of Atlantic salmon Salmo salar revealed by scanning electron
microscopy. Dis. Aquat. Organ. 59, 79–84.
Mata, M., Skarmeta, A., Santos, Y., 2002. A proposed serotyping system for Flavobacterium
psychrophilum. Lett. Appl. Microbiol. 35, 166–170.
Matthews, M.D., Bowker, J.D., Carty, D.G., Wandelear, N., Bowman, M.P., Sakmar, J.C.,
Childress, K., 2013. Efficacy of Aquaflor (50% florfenicol)–medicated feed to control
mortality associated with Flavobacterium columnare infection in Florida largemouth bass
and bluegill. N. Am. J. Aquac. 75, 385–392.
Maturin, L., Peeler J.P., 2001. Aerobic Plate Count. in: Bacteriological Analytical (Ed.) Manual,
8th ed., United States: AOAC International. Available at
https://www.fda.gov/Food/FoodScienceResearch/LaboratoryMethods/ucm063346.h
tm (accessed September 12, 2016).
McKeever, D., Rege, J.E., 1999. Vaccines and diagnostic tools for animal health: the influence
of biotechnology. Livest. Prod. Sci. 59, 257–264.
Meinertz, J.R., Hess, K.R., Bernardy, J.A., Gaikowski, M.P., Whitsel, M., Endris, R.G., 2014.
Florfenicol residues in rainbow trout after oral dosing in recirculating and flow-through
culture systems. J. Aquat. Anim. Health 26, 243–250.
Meyers, T.R., 1991. Apparent chronic bacterial myeloencephalitis in hatchery-reared juvenile
coho salmon Oncorhynchus kisutch in Alaska. Dis. Aquat. Organ. 6, 217–219.
Michel, C., Antonio, D., Hedrick, R.P., 1999. Production of viable cultures of Flavobacterium
psychrophilum: approach and control. Res. Microbiol. 351–358.
178 Michel, C., Kerouault, B., Martin, C., 2003. Chloramphenicol and florfenicol susceptibility of
fish-pathogenic bacteria isolated in France: Comparison of minimum inhibitory
concentration, using recommended provisory standards for fish bacteria. J. Appl. Microbiol.
95, 1008–1015.
Misaka, N., Nishizawa, T., Yoshimizu, M., 2008. Quantitative detection of viable
Flavobacterium psychrophilum in chum salmon Oncorhynchus keta by colony blotting and
immunostaining. Fish Pathol. 43, 117–123.
Miwa, S., Nakayasu, C., 2005. Pathogenesis of experimentally induced bacterial cold water
disease in ayu Plecoglossus altivelis. Dis. Aquat. Organ. 67, 93–104.
Moccia, R.D., Bevan, D.J., 2017. Ontario Aquacultural Production in 2016. University of
Guelph, Aquaculture Extension Centre Factsheet, Publ. No. 01-17. 2 pp.
Moffitt, C.M., Schreck, J.A., 1988. Accumulation and depletion of orally administered
erythromycin thiocyanate in tissues of Chinook salmon. Trans. Am. Fish. Soc. 117, 394–
400.
Narrowe, A.B., Albuthi-Lantz, M., Smith, E.P., Bower, K.J., Roane, T.M., Vajda, A.M., Miller,
C.S., 2015. Perturbation and restoration of the fathead minnow gut microbiome after low-
level triclosan exposure. Microbiome 6. doi:10.1186/s40168-015-0069-6
Navarrete, P., Magne, F., Araneda, C., Fuentes, P., Barros, L., Opazo, R., Espejo, R., Romero, J.,
2012. PCR-TTGE analysis of 16S rRNA from rainbow trout (Oncorhynchus mykiss) gut
microbiota reveals host-specific communities of active bacteria. PLoS One 7, e31335.
doi:10.1371/journal.pone.0031335
179 Navarrete, P., Magne, F., Mardones, P., Riveros, M., Opazo, R., Suau, A., Pochart, P., Romero,
J., 2009. Molecular analysis of intestinal microbiota of rainbow trout (Oncorhynchus
mykiss). FEMS Microbiol. Ecol. 71, 148–156.
Navarrete, P., Mardones, P., Opazo, R., Espejo, R., Romero, J., 2008. Oxytetracycline treatment
reduces bacterial diversity of intestinal microbiota of Atlantic salmon. J. Aquat. Anim.
Health 203, 177–183.
Neiger, R., Thomas, M., Das, S., Barnes, M., Fletcher, B., Snekvik, K., Thompson, J., Scaria, J.,
2016. Draft genome sequences of three Flavobacterium psychrophilum strains isolated from
coldwater disease outbreaks at three production hatcheries. Genome Announc. 4, e00035-
16. doi:10.1128/genomeA.00035-16
Nematollahi, A., Decostere, A., Pasmans, F., Ducatelle, R., Haesebrouck, F., 2003a. Adhesion of
high and low virulence Flavobacterium psychrophilum strains to isolated gill arches of
rainbow trout Oncorhynchus mykiss. Dis. Aquat. Organ. 55, 101–107.
Nematollahi, A., Decostere, A., Pasmans, F., Haesebrouck, F., 2003b. Flavobacterium
psychrophilum infections in salmonid fish. J. Fish Dis. 26, 563–574.
Nematollahi, A., Pasmans, F., Haesebrouck, F., Decostere, A., 2005. Early interactions of
Flavobacterium psychrophilum with macrophages of rainbow trout Oncorhynchus mykiss.
Dis. Aquat. Organ. 64, 23–28.
Ngo, T.P.H., Smith, P., Bartie, K.L., Thompson, K.D., Verner-Jeffreys, D.W., Hoare, R., Adams,
A., 2018. Antimicrobial susceptibility of Flavobacterium psychrophilum isolates from the
United Kingdom. J. Fish Dis. 41, 309–320. doi:10.1111/jfd.12730
180 Nicolas, P., Mondot, S., Achaz, G., Bouchenot, C., Bernardet, J.-F., Duchaud, E., 2008.
Population structure of the fish-pathogenic bacterium Flavobacterium psychrophilum. Appl.
Environ. Microbiol. 74, 3702–3709.
Nilsen, H., Sundell, K., Duchaud, E., Nicolas, P., Dalsgaard, I., Madsen, L., Aspán, A., Jansson,
E., Colquhoun, D.J., Wiklund, T., 2014. Multilocus sequence typing identifies epidemic
clones of ø in Nordic countries. Appl. Environ. Microbiol. 80, 2728–2736.
Onarheim, A.M., Raa, J., 1990. Characterization and possible biological significance of
autochthonous flora in the intestinal mucosa of seawater fish. in: Lesel, R. (Ed.),
Microbiology in Poikilotherms, Elsevier, Amsterdam,The Netherlands. pp.197-201.
Onuk, E.E., Ciftci, A., Findik, A., Durmaz, Y., 2010. Development and evaluation of a multiplex
PCR assay for simultaneous detection of Flavobacterium psychrophilum, Yersinia ruckeri
and Aeromonas salmonicida subsp. salmonicida in culture fisheries. J. Vet. Sci. 11, 235–
241.
Orieux, N., Bourdineaud, J.P., Douet, D.G., Daniel, P., Le Henaff, M., 2011. Quantification of
Flavobacterium psychrophilum in rainbow trout, Oncorhynchus mykiss (Walbaum), tissues
by qPCR. J. Fish Dis. 34, 811–821.
Ostland, V.E., Byrne, P.J., Hoover, G., Ferguson, H.W., 2000. Necrotic myositis of rainbow
trout, Oncorhynchus mykiss (Walbaum): proteolytic characteristics of a crude extracellular
preparation from Flavobacterium psychrophilum. J. Fish Dis. 23, 329–336.
Ostland, V.E., Byrne, P.J., Lumsden, J.S., MacPhee, D.D., Derksen, J.A., Haulena, M., Skar, K.,
Myhr, E., Ferguson, H.W., 1999. Atypical bacterial gill disease: a new form of bacterial gill
disease affecting intensively reared salmonids. J. Fish Dis. 22, 351–358.
181 Ostland, V.E., McGrogan, D.G., Ferguson, H.W., 1997. Cephalic osteochondritis and necrotic
scleritis in intensively reared salmonids associated with Flexibacter psychrophilus. J. Fish
Dis. 20, 443–451.
Otis, E.J. 1984. Lesions of Cold Water Disease in Steelhead Trout (Salmo gairdneri): The Role
of Cytophaga psychrophila Extracellular Products. MSc thesis. University of Rhode Island.
Pacha, R.E., 1968. Characteristics of Cytophaga psychrophila (Borg) isolated during outbreaks
of bacterial cold-water disease. Appl. Microbiol. 16, 97–101.
Papadopoulou, A., Dalsgaard, I., Lindén, A., Wiklund, T., 2017. In vivo adherence of
Flavobacterium psychrophilum to mucosal external surfaces of rainbow trout
(Oncorhynchus mykiss) fry. J. Fish Dis. 40, 1309–1320.
Pérez-Pascual, D., Gómez, E., Álvarez, B., Méndez, J., Reimundo, P., Navais, R., Duchaud, E.,
Guijarro, J.A., 2011. Comparative analysis and mutation effects of fpp2–fpp1 tandem genes
encoding proteolytic extracellular enzymes of Flavobacterium psychrophilum.
Microbiology 157, 1196–1204.
Pérez-Pascual, D., Rochat, T., Kerouault, B., Gómez, E., Neulat-Ripoll, F., Henry, C., Quillet,
E., Guijarro, J.A., Bernardet, J.F., Duchaud, E., 2017. More than gliding: Involvement of
GldD and GldG in the virulence of Flavobacterium psychrophilum. Front. Microbiol. 8,
2168. doi:10.3389/fmicb.2017.02168
Plant, K.P., LaPatra, S.E., Call, D.R., Cain, K.D., 2011. Immunization of rainbow trout,
Oncorhynchus mykiss (Walbaum), with Flavobacterium psychrophilum proteins elongation
factor-Tu, SufB Fe-S assembly protein and ATP synthaseβ. J. Fish Dis. 34, 247–250.
Post, G., 1987. Textbook of Fish Health. T.F.H. Publications.
182 Rangdale, R.E., 1995. Studies on Rainbow Trout Fry Syndrome (RTFS). PhD thesis. University
of Stirling.
Rangdale, R.E., Richards, R.E., Alderman, D.J., 1996. Isolation of Cytophaga psychrophila,
causal agent of rainbow trout fry syndrome (RTFS) from reproduction fluids and egg
surfaces of rainbow trout (Oncorhynchus mykiss). Bull. Eur. Ass. Fish Pathol. 16, 63–67.
Rangdale, R.E., Richards, R.H., Alderman, D.J., 1999. Histopathological and electron
microscopical observations on rainbow trout fry syndrome. Vet. Rec. 144, 251–254.
Rangdale, R.E., Richards, R.H., Alderman, D.J., 1997. Minimum inhibitory concentrations of
selected antimicrobial compounds against Flavobacterium psychrophilum the causal agent
of rainbow trout fry syndrome (RTFS). Aquaculture 158, 193–201.
Rangdale, R.E., Way, K., 1995. Rapid identification of C. psychrophila from infected spleen
tissue using an enzyme-linked-immunosorbent assay (ELISA). Bull. Euro. Ass. Fish Pathol.
15, 213–216.
Rawls, J.F., Samuel, B.S., Gordon, J.I., 2004. Gnotobiotic zebrafish reveal evolutionarily
conserved responses to the gut microbiota. Proc. Natl. Acad. Sci. U. S. A. 101, 4596–4601.
Ringø, E., Sperstad, S., Myklebust, R., Refstie, S., Krogdahl, Å., 2006. Characterisation of the
microbiota associated with intestine of Atlantic cod (Gadus morhua L.) The effect of fish
meal, standard soybean meal and a bioprocessed soybean meal. Aquaculture 261, 829–841.
Ringø, E., Strøm, E., Tabachek, J.-A., 1995. Intestinal microflora of salmonids: a review. Aquac.
Res. 26, 773–789.
Ringø, E., Zhou, Z., Vecino, J.L.G., Wadsworth, S., Romero, J., Krogdahl, Å., Olsen, R.E.,
Dimitroglou, A., Foey, A., Davies, S., Owen, M., Lauzon, H.L., Martinsen, L.L., De
183 Schryver, P., Bossier, P., Sperstad, S., Merrifield, D.L., 2016. Effect of dietary components
on the gut microbiota of aquatic animals. A never-ending story? Aquac. Nutr. 22, 219–282.
Robinson, C.J., Young, V.B., 2010. Antibiotic administration alters the community structure of
the gastrointestinal micobiota. PLoS One 1, 279–284. doi:10.4161/gmic.1.4.12614
Rochat, T., Fujiwara-Nagata, E., Calvez, S., Dalsgaard, I., Madsen, L., Calteau, A., Lunazzi, A.,
Nicolas, P., Wiklund, T., Bernardet, J.-F., Duchaud, E., 2017. Genomic characterization of
Flavobacterium psychrophilum serotypes and development of a multiplex PCR-based
serotyping scheme. Front. Microbiol. 8, 1752. doi:10.3389/fmicb.2017.01752
Romero, J., Navarrete, P., 2006. 16S rDNA-Based Analysis of dominant bacterial populations
associated with early life stages of coho salmon (Oncorhynchus kisutch). Microb. Ecol. 51,
422–430.
Samuelsen, O.B., Bergh, Ø., 2004. Efficacy of orally administered florfenicol and oxolinic acid
for the treatment of vibriosis in cod (Gadus morhua). Aquaculture 235, 27–35.
Samuelsen, O.B., Hjeltnes, B., Glette, J., 1998. Efficacy of orally administered florfenicol in the
treatment of furunculosis in Atlantic salmon. J. Aquat. Anim. Health 10, 56–61.
Sanchez-Moreno, M., Monteoliva-Sanchez, M., Ortega, F., Ramos-Cormenzana, A., Monteoliva,
M., 1989. Superoxide dismutase in strains of the genus Flavobacterium: isolation and
characterization. Arch. Microbiol. 152, 407–410.
Sanders, J.W., Martin, J.W., Hooke, M., Hooke, J., 2000. Methylobacterium mesophilicum
infection: case report and literature review of an unusual opportunistic pathogen. Clin.
Infect. Dis. 30:936–938. doi:10.1086/313815
184 Santos, Y., Huntly, P.J., Turnbull, A., Hastings, T.S., 1992. Isolation of Cytophaga psychorphila
(Flexibacter psychrophilus) in association with rainbow trout mortality in the United
Kingdom. Bull. Eur. Assoc. Fish Pathol. 12, 209–210.
Schachte, J.H., 1983. Coldwater Disease, in: Meyer, F.P., Warren, J., Carey, T. (Eds.), A Guide
to Integrated Fish Health Management in the Great Lakes Basin. Great Lakes Fishery
Commission, Ann Harbour, Michigan, USA, pp. 193–197.
Schloss, P.D., Westcott, S.L., Ryabin, T., Hall, J.R., Hartmann, M., Hollister, E.B., Lesniewski,
R.A., Oakley, B.B., Parks, D.H., Robinson, C.J., Sahl, J.W., Stres, B., Thallinger, G.G., Van
Horn, D.J., Weber, C.F., 2009. Introducing mothur: open-source, platform-independent,
community-supported software for describing and comparing microbial communities. Appl.
Environ. Microbiol. 75, 7537–7541.
Schmidt, A.S., Bruun, M.S., Dalsgaard, I., Pedersen, K., Larsen, J.L., 2000. Occurrence of
antimicrobial resistance in fish-pathogenic and environmental bacteria associated with four
Danish rainbow trout farms. Appl. Environ. Microbiol. 66, 4908–4915.
Schmidtke, L.M., Carson, J., 1995. Characteristics of Flexibacter psychrophilus isolated from
Atlantic salmon in Australia. Dis. Aquat. Organ. 21, 157–161.
Sealey, W.M., Conley, Z.B., Bensley, M., 2015. Prebiotic supplementation has only minimal
effects on growth efficiency, intestinal health and disease resistance of Westslope cutthroat
trout Oncorhynchus clarkii lewisi fed 30% soybean meal. Front. Immunol. 6, 396.
doi:10.3389/fimmu.2015.00396
Secades, P., Alvarez, B., Guijarro, J.A., 2003. Purication and properties of a new psychrophilic
metalloprotease (Fpp2) in the fish pathogen Flavobacterium psychrophilum. FEMS
Microbiol. Lett. 226, 273–279.
185 Secades, P., Alvarez, B., Guijarro, J.A., 2001. Purification and characterization of a
psychrophilic, calcium- induced, growth-phase-dependent metalloprotease from the fish
pathogen Flavobacterium psychrophilum. Appl. Environ. Microbiol. 67, 2436–2444.
Sekowska, A., Wendel, S., Fischer, E.C., Nørholm, M.H.H., Danchin, A., 2016. Generation of
mutation hotspots in ageing bacterial colonies. Sci. Rep. 6, 2. doi:10.1038/s41598-016-
0005-4
Semova, I., Carten, J.D., Stombaugh, J., Mackey, L.C., Knight, R., Farber, S.A., Rawls, J.F.,
2012. Microbiota regulate intestinal absorption and metabolism of fatty acids in the
zebrafish. Cell Host Microbe 12, 277–288.
Semple, S.L., Kellendonk, C.J., Al-Hussinee, L., Macinnes, J.I., Lumsden, J.S., Dixon, B., 2018.
Serum IgM, MH class IIβ genotype and respiratory burst activity do not differ between
rainbow trout families displaying resistance or susceptibility to the coldwater pathogen,
Flavobacterium psychrophilum. Aquaculture 483, 131–140.
Sharma, A., Sojar, H.T., Glurich, I., Honma, K., Kuramitsu, H.K., Genco, R.J., 1998. Cloning,
expression, and sequencing of a cell surface antigen containing a leucine-rich repeat motif
from Bacteroides forsythus ATCC 43037. Infect. Immun. 66, 5703–5710.
Shieh H.S. (1980) Studies on the nutrition of a fish pathogen, Flexibacter columnaris. Microbios
Letters 13, 129–133.
Shimizu, M., Goda, H., Yamasaki, K., Oshima, S.-I., Ohnishi, K., Osaki, Y., Kataoka, S.,
Imajoh, M., 2016. Draft genome sequence of Flavobacterium psychrophilum strain KTEN-
1510 with genotype A/G-C, isolated from an ayu (Plecoglossus altivelis altivelis) in the
Kagami River, Kochi, Japan. Genome Announc. 4, e01762-15.
doi:10.1128/genomeA.01762-15
186 Silverstein, J.T., Vallejo, R.L., Palti, Y., Leeds, T.D., Rexroad, C.E., Welch, T.J., Wiens, G.D.,
Ducrocq, V., 2008. Rainbow trout resistance to bacterial cold-water disease is moderately
heritable and is not adversely correlated with growth. J. Anim. Sci. 87, 860–867.
Skrodenyte-Arbaciauskiene, V., Sruoga, A., Butkauskas, D., 2006. Assessment of microbial
diversity in the river trout Salmo trutta fario L. intestinal tract identified by partial 16S
rRNA gene sequence analysis. Fish. Sci. 72, 597–602.
Smith, P.R., Le Breton, A., Horsberg, T.E., Corsin, F., 2009. Guidelines for Antimicrobial Use in
Aquaculture, in: Guardabassi, L., Jensen, L.B., Kruse, H. (Eds.), Guide to Antimicrobial
Use in Animals. Blackwell Publishing, Ltd, Oxford, UK, pp. 207–218.
Soule, M., LaFrentz, S., Cain, K., LaPatra, S., Call, D.R., 2005. Polymorphisms in 16S rRNA
genes of Flavobacterium psychrophilum correlate with elastin hydrolysis and tetracycline
resistance. Dis. Aquat. Organ. 65, 209–216.
Spor, A., Koren, O., Ley, R., 2011. Unravelling the effects of the environment and host genotype
on the gut microbiome. Nat. Rev. Microbiol. 9, 279–290.
Standen, B.T., Rodiles, & A., Peggs, & D.L., Davies, S.J., Santos, G.A., Merrifield, D.L., 2015.
Modulation of the intestinal microbiota and morphology of tilapia, Oreochromis niloticus,
following the application of a multi-species probiotic. Appl Microbiol Biotechnol 99, 8403–
8417.
Starliper, C.E., 2011. Bacterial coldwater disease of fishes caused by Flavobacterium
psychrophilum. J. Adv. Res. 2, 97–108.
Stenholm, A.R., Dalsgaard, I., Middelboe, M., 2008. Isolation and characterization of
bacteriophages infecting the fish pathogen Flavobacterium psychrophilum. Appl. Environ.
Microbiol. 74, 4070–4078.
187 Strepparava, N., Nicolas, P., Wahli, T., Segner, H., Petrini, O., 2013. Molecular epidemiology of
Flavobacterium psychrophilum from Swiss fish farms. Dis. Aquat. Organ. 105, 203–210.
Strepparava, N., Wahli, T., Segner, H., Petrini, O., 2014. Detection and quantification of
Flavobacterium psychrophilum in water and fish tissue samples by quantitative real time
PCR. BMC Microbiol. 14, 105. doi:10.1186/1471-2180-14-105
Strepparava, N., Wahli, T., Segner, H., Polli, B., Petrini, O., 2012. Fluorescent in situ
hybridization: a new tool for the direct identification and detection of F. psychrophilum.
PLoS One 7, e49280. doi:10.1371/journal.pone.0049280
Ström-Bestor, M., Wiklund, T., 2011. Inhibitory activity of Pseudomonas sp. on Flavobacterium
psychrophilum, in vitro. J. Fish Dis. 34, 255–264.
Sun, Y.-X., Zhao, H.-Y., Shan, Q., Zhu, S., Zeng, D.-P., Liu, Z.-C., 2010. Tissue distribution and
elimination of florfenicol in crucian carp (Carassius auratus cuvieri) after a single dose
intramuscular or oral administration. Aquaculture 309, 82–85.
Sundell, K., Wiklund, T., 2015. Characteristics of epidemic and sporadic Flavobacterium
psychrophilum sequence types. Aquaculture 441, 51–56.
Sundell, K., Wiklund, T., 2011. Effect of biofilm formation on antimicrobial tolerance of
Flavobacterium psychrophilum. J. Fish Dis. 34, 373–383.
Suzuki, K., Arai, H., 2008. Reliability of PCR methods for the detection of Flavobacterium
psychrophilum. Fish Pathol. 43, 124–127.
Tao, W., Lee, M.H., Wu, J., Kim, N.H., Kim, J.-C., Chung, E., Hwang, E.C., Lee, S.-W., 2012.
Inactivation of chloramphenicol and florfenicol by a novel chloramphenicol hydrolase.
Appl. Environ. Microbiol. 78, 6295–6301.
188 Taylor, P.W., 2004. Detection of Flavobacterium psychrophilum in eggs and sexual fluids of
pacific salmonids by a polymerase chain reaction assay: Implications for vertical
transmission of bacterial coldwater disease. J. Aquat. Anim. Health 16, 104–108.
Toranzo, A.E., Barja, J.L., 1993. Fry mortality syndrome (FMS) in Spain. Isolation of causative
bacterium Flexibacter psychrophilus. Bull. Eur. Assoc. Fish Pathol. 13, 30–32.
Uddin, M.N., Wakabayashi, H., 1997. Effects of temperature on growth and protease production
of Cytophaga psychrophila. Fish Pathol. 32, 225–226.
Valdebenito, S., Avendaño-Herrera, R., 2009. Phenotypic, serological and genetic
characterization of Flavobacterium psychrophilum strains isolated from salmonids in Chile.
J. Fish Dis. 32, 321–333.
Vaneechoutte, M., Dijkshoorn, L., Nemec, A., Kämpfer, P., Wauters G., 2011. Acinetobacter,
Chryseobacterium, Moraxella, and other nonfermentative Gram-negative rods, in:
Versalovic J, Carroll KC, Jorgensen JH, Funke G, Landry M.L., Warnock DW (Eds.),
Manual of Clinical Microbiology, 10th ed, vol 1. p 729–731. ASM Press, Washington, DC.
Vatsos, I.N., Thompson, K.D., Adams, A., 2002. Development of an immunofluorescent
antibody technique (IFAT) and in situ hybridization to detect Flavobacterium
psychrophilum in water samples. Aquac. Res. 33, 1087–1090.
Vliet, D. Van, Loch, T.P., Smith, P., Faisal, M., 2017. Antimicrobial susceptibilities of
Flavobacterium psychrophilum isolates from the Great Lakes Basin, Michigan. Microb.
Drug Resist. 23, 791–798. doi:10.1089/mdr.2016.0103
Wagener, C., 1997. Molecular Diagnostics. J. Mol. Med. 75, 728–744.
Wakabayashi, H., Toyama, T., Iida, T., 1994. A study on serotyping of Cytophaga psychrophila
isolated from fishes in Japan. Fish Pathol. 29, 101–104.
189 Wang, L., Han, Y.-N., Jin, S., Ma, Y., Wang, G.-L., Zhao, Q.-S., Chen, Y.-E., 2015.
Pharmacokinetic study of florfenicol in healthy and vibriosis-infected Pseudosciaena
crocea after oral administration. J. Appl. Biol. Chem. 58, 363–368.
Wiens, G.D., 2011. Bacterial kidney disease (Renibacterium salmoninarum). In: Fish Diseases
and Disorders, Volume 3: Viral, Bacterial, and Fungal Pathogens, 2nd edn. (ed. by P.T.K.
Woo & D.W. Bruno), pp. 338–374. CABI, Wallingford, UK.
Wiens, G.D., LaPatra, S.E., Welch, T.J., Rexroad, C., Call, D.R., Cain, K.D., LaFrentz, B.R.,
Vaisvil, B., Schmitt, D.P., Kapatral, V., 2014. Complete genome sequence of
Flavobacterium psychrophilum strain CSF259-93, used to select rainbow trout for increased
genetic resistance against bacterial cold water disease. Genome Announc. 2, e00889-14.
doi:10.1128/genomeA.00889-14
Wiklund, T., Madsen, L., Bruun, M.S., Dalsgaard, I., 2000. Detection of Flavobacterium
psychrophilum from fish tissue and water samples by PCR amplification. J. Appl.
Microbiol. 88, 299–307.
Wlodarska, M., Thaiss, C.A., Nowarski, R., Henao-Mejia, J., Zhang, J.-P., Brown, E.M.,
Frankel, G., Levy, M., Katz, M.N., Philbrick, W.M., Elinav, E., Finlay, B.B., Flavell, R.A.,
2014. NLRP6 inflammasome orchestrates the colonic host-microbial interface by regulating
goblet cell mucus secretion. Cell 156, 1045–1059.
Wlodarska, M., Willing, B., Keeney, K.M., Menendez, A., Bergstrom, K.S., Gill, N., Russell,
S.L., Vallance, B.A., Finlay, B.B., 2011. Antibiotic treatment alters the colonic mucus layer
and predisposes the host to exacerbated Citrobacter rodentium-induced colitis. Infect.
Immun. 79, 1536–1545.
190 Wong, S., Waldrop, T., Summerfelt, S., Davidson, J., Barrows, F., Kenney, P.B., Welch, T.,
Wiens, G.D., Snekvik, K., Rawls, J.F., Good, C., 2013. Aquacultured rainbow trout
(Oncorhynchus mykiss) possess a large core intestinal microbiota that is resistant to
variation in diet and rearing density. Appl. Environ. Microbiol. 79, 4974–4984.
Wood, E.M., Yasutake, W.T., 1956. Histopathology of fish: III. peduncle (cold-water) disease.
Progress. Fish-Culturist 18, 58–61.
Wu, A.K., Kropinski, A.M., Lumsden, J.S., Dixon, B., MacInnes, J.I., 2015. Complete genome
sequence of the fish pathogen Flavobacterium psychrophilum ATCC 49418(T.). Stand.
Genomic Sci. 10, 3. doi:10.1186/1944-3277-10-3
Ye, L., Amberg, J., Chapman, D., Gaikowski, M., Liu, W.-T., 2014. Fish gut microbiota analysis
differentiates physiology and behavior of invasive Asian carp and indigenous American
fish. ISME J. 8, 541–551.
Zarkasi, K.Z., Taylor, R.S., Abell, G.C.J., Tamplin, M.L., Glencross, B.D., Bowman, J.P., 2016.
Atlantic salmon (Salmo salar L.) gastrointestinal microbial community dynamics in relation
to digesta properties and diet. Microb. Ecol. 71, 589–603.
191