General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

Users may download and print one copy of any publication from the public portal for the purpose of private study or research.

You may not further distribute the material or use it for any profit-making activity or commercial gain

You may freely distribute the URL identifying the publication in the public portal If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from orbit.dtu.dk on: Oct 14, 2021

Microbial interactions in aquaculture - Probiotic roseobacters as a sustainable meansto control fish pathogens

Rasmussen, Bastian Barker

Publication date:2019

Document VersionPublisher's PDF, also known as Version of record

Link back to DTU Orbit

Citation (APA):Rasmussen, B. B. (2019). Microbial interactions in aquaculture - Probiotic roseobacters as a sustainable meansto control fish pathogens. Technical University of Denmark.

Microbial interactions in

aquaculture - Probiotic

roseobacters as a sustainable

means to control fish pathogens

Bastian Barker Rasmussen

PhD Thesis

Technical University of Denmark

Department of Biotechnology and Biomedicine

Section for Microbial and Chemical Ecology

Bacterial Ecophysiology and Biotechnology Group

March 2019

COVER ILLUSTRATION

Microscopy of Acartia tonsa copepod nauplii colonized by GFP-tagged Vibrio

anguillarum. Red fluorescent cells are red algae Rhodomonas salina.

Magnification x 100. Left) Phase-contrast microscopy.

Right) Fluorescence microscopy.

i

PREFACE

This PhD study was conducted at Department of Biotechnology and Biomedicine,

Technical University of Denmark from December 2015 to March 2019 under the

supervision of Professor Lone Gram and Assistant Professor Mikkel Bentzon-Tilia

A part of the study was carried out in the laboratory of Professor Heidrun

Wergeland at the University of Bergen, Norway.

The study was a part of the ProAqua project, which was funded by a grant from

The Danish Council for Strategic Research | Programme Commission on Health,

Food and Welfare (12-132390; ProAqua).

Bastian Barker Rasmussen

March 2019

ii

SUMMARY

Aquaculture is the fastest growing protein production sector in the world. A

major bottleneck in aquaculture is bacterial infections, which can lead to great

losses. Fish larvae of several commercially important finfish are fed live feed due

to the lack of appropriate artificial feed formulations. Live feed cultures are high

in nutrition, which promote growth of bacteria including fast growing

opportunistic pathogens. Live feed can function as infection vectors for

pathogenic bacteria. Thus, fish larvae are especially susceptible to bacterial

infections. Antibiotics are still used to combat bacterial infections in fish larvae

cultures as the larvae cannot be efficiently vaccinated, however, the risk of

resistance development and the associated health risk have led to the search for

sustainable alternatives. Two types of promising alternative biocontrol are the

use of probiotic bacteria and bacteriophages (virus that infect bacteria) the so-

called phage therapy. The probiotic Phaeobacter spp. are able to inhibit the

growth of a range of pathogens including Vibrio spp. due to the production of the

antimicrobial compound tropodithietic acid (TDA). Phaeobacter spp. have

successfully been shown to inhibit Vibrio spp. in axenic, i.e., without background

microbiota, as well as few non-axenic live feed systems. Furthermore,

Phaeobacter spp. are able to protect the live feed Artemia as well as fish larvae of

cod and turbot against the Vibrio spp. induce disease vibriosis.

The purpose of this PhD project was to investigate if probiotic Phaeobacter spp.

were able to inhibit fish pathogenic bacteria in non-axenic live feed cultures either

alone or in combination with bacteriophages (the latter in collaboration with

researchers from aquaculture and in the area of phage therapy). Furthermore,

this PhD sought to examine the possible practical application of Phaeobacter

based probiotics in commercial aquaculture via upscaling in live feed microalgae.

The present PhD study has shown that Phaeobacter spp. are able to antagonise

pathogenic vibrios in a range of non-axenic live feed systems, i.e., the microalgae

Rhodomonas salina and Tetraselmis suecica and the zooplankton Artemia and

copepod species Acartia tonsa with two log units or more. Furthermore,

Phaeobacter spp. significantly increased survival of vibrio-challenged, as well as,

iii

non-challenged Artemia in non-axenic systems. We found that inhibition of

pathogens by Phaeobacter spp. varied between different non-axenic systems and

we speculated that this is due to the composition of microbiota in the non-axenic

systems, which may or may not protect the pathogens.

In collaboration with researchers from aquaculture and the phage therapy field,

we tested the broad host-range bacteriophage KVP40s abilities to inhibit four

different Vibrio anguillarum strains and protect turbot and cod against vibriosis.

Mortality of Vibrio-challenged turbot and cod larvae were decreased or delayed

when treated with KVP40. Growth of indigenous pathogenic bacteria resulted in

high mortality in the non-challenged larvae controls. However, treatment with

KVP40 reduced larvae mortality in these cultures. KVP40 was also able to inhibit V.

anguillarum in live feed cultures T. suecica and Artemia, however, regrowth of V.

anguillarum was seen in the Artemia cultures.

By combining the probiotic Phaeobacter inhibens with KVP40 as treatment in T.

suecica and Artemia cultures, faster inhibition of V. anguillarum was seen

compared to P. inhibens only and lower V. anguillarum counts compared to KVP40

only. A marginal improvement was observed in the protection of Artemia against

vibriosis compared to P. inhibens only.

We were to some extend able to upscale Phaeobacter spp. in T. suecica culture,

however, as the Phaeobacter spp. did not grow well when changing from non-

aerated cultures, with large surface area to volume ratio, to aerated cultures, with

small surface area to volume ratio. In order to test if the upscaled Phaeobacter

spp. retained their probiotic abilities, Artemia were fed the Phaeobacter infused

algae and challenged with V. anguillarum. Only Phaeobacter piscinae S26 was able

to inhibit V. anguillarum and no protection of Artemia was observed, however, we

ascribe this to the low levels of probiont, not the loss of inhibitory ability.

Summed up, this study demonstrates that Phaeobacter spp. can inhibit

pathogenic vibrios in a range of non-axenic live feed systems, where they will

establish themselves and can be scaled up if conditions favour it. Furthermore, if

combined with phage therapy slightly improved inhibition profiles are seen.

iv

DANSK RESUMÉ

Akvakultur er den hurtigst voksende proteinproduktionssektor i verden.

Bakterielle infektioner er et stort problem i akvakulturer, som kan medføre store

tab. Fiskelarver af flere kommercielt vigtige arter fodres med levende

foderorganismer på grund af manglen på passende foderblandinger. Kulturer med

foderorganismer har højt næringsindhold, hvilket fremmer vækst af bakterier,

herunder hurtigt voksende opportunistiske patogene. Foderorganismer kan

fungere som infektionsvektorer for patogene bakterier og fiskelarver er således

særligt udsatte for bakterielle infektioner. Antibiotika bruges stadig til at

bekæmpe bakterielle infektioner i fiskelarvekulturer, da larverne ikke kan

vaccineres effektivt. Antibiotika medfører risiko for resistensudvikling og de

forbundne sundhedsrisici har ført til søgen efter bæredygtige alternativer. To

alternative typer af biokontrol er brugen af probiotiske bakterier og bakteriofager

(virus der inficerer bakterier) også kendt som fagterapi. De probiotiske

Phaeobacter spp. er i stand til at hæmme væksten af en række patogene

bakterier, herunder Vibrio spp. grundet produktionen af det antimikrobielle stof

tropodithietic acid (TDA). Phaeobacter spp. er i stand til at hæmme Vibrio spp. i

axeniske, dvs. uden baggrundsmikrobiota, såvel som få ikke-axeniske systemer

med foderorganismer. Desuden er Phaeobacter spp. i stand til at beskytte

foderorganismen Artemia samt torsk og pighvar fiskelarver mod Vibrio spp..

Formålet med dette ph.d. studie var at undersøge, om probiotiske Phaeobacter

spp. var i stand til at hæmme fiskepatogene bakterier i ikke-axeniske systemer

med foderorganismer enten alene eller i kombination med bakteriofager

(sidstnævnte i samarbejde med forskere fra akvakultur og inden for fagterapi).

Desuden har dette studie undersøgt den praktiske anvendelse af Phaeobacter

baserede probiotika i kommerciel akvakultur via opskalering i mikroalger.

Dette PhD studie har vist, at Phaeobacter spp. er i stand til at hæmme patogene

vibrioer med to logaritmiske enheder eller mere, i en række ikke-axeniske

systemer med foderorganismer, nærmere bestemt mikroalgerne Rhodomonas

salina og Tetraselmis suecica samt dyreplanktonen Artemia og copepod arten

Acartia tonsa. Desuden øgede Phaeobacter spp. signifikant overlevelse af Artemia

v

i både vibrio-inficerede og ikke-inficerede ikke-axeniske systemer. Phaeobacter

spp. hæmning af patogene varierer mellem forskellige ikke-axeniske systemer og

vi spekulerede på om dette skyldes sammensætningen af mikrobiotaen i de ikke-

axeniske systemer, der muligvis kan beskytte de patogene.

I samarbejde med forskere fra akvakultur og fagterapiområderne testede vi

bakteriofag KVP40s evne til at hæmme fire forskellige Vibrio anguillarum stammer

samt beskytte pighvar- og torskelarver mod vibriose. Dødeligheden af vibrio-

inficered pighvar- og torskelarver blev sænket eller forsinket, når de blev

behandlet med KVP40. Væksten af naturligt forekommende patogene bakterier

resulterede i høj dødelighed i de ikke-inficerede fiskelarvekontroller. Behandling

med KVP40 reducerede imidlertid dødeligheden af larverne i disse kulturer. KVP40

kunne også hæmme V. anguillarum i kulturer af foderorganismerne T. suecica og

Artemia, også selv om V. anguillarum voksede frem igen i Artemia kulturene.

Ved at kombinere probiotisk Phaeobacter inhibens med KVP40 som behandling i

T. suecica og Artemia kulturer, blev V. anguillarum hurtigere hæmmet

sammenligning med P. inhibens alene og kraftigere sammenlignet med KVP40

alene. En marginal forbedring blev observeret i beskyttelsen af Artemia mod

vibriose sammenlignet med P. inhibens alene.

Vi var delvist i stand til at skalere Phaeobacter spp. op i T. suecica kultur, selvom

Phaeobacter spp. ikke voksede godt, når der blev skiftet fra ikke-beluftede

kulturer, med stort overfladeareal til volumen ratio, til beluftede kulturer med lille

overflade til volumen ratio. For at teste om de opskalerede Phaeobacter spp.

beholdt deres probiotiske evner, blev Artemia fodret med alger med Phaeobacter

og inficeret med V. anguillarum. Kun Phaeobacter piscinae S26 var i stand til at

hæmme V. anguillarum og ingen beskyttelse af Artemia blev observeret. Dette

tilskriver vi de lave niveauer af probiont, ikke tabet af deres hæmmende evner.

Sammenfattet viser dette studie at Phaeobacter spp. kan hæmme patogene

vibrioer i en rækkevidde på ikke-axeniske foderorganisme systemer, hvor de kan

etablere sig og skaleres op, hvis forholdene favoriserer det. Desuden ses i

kombination med fagterapi lidt forbedrede hæmningsprofiler.

vi

TABLE OF CONTENTS

PREFACE ..................................................................................................................... i

SUMMARY ................................................................................................................. ii

DANSK RESUMÉ ........................................................................................................ iv

PUBLICATIONS INCLUDED IN THIS THESIS ............................................................... vii

1 BACKGROUND AND INTRODUCTION ................................................................ 1

2 MARINE AQUACULTURE ................................................................................... 6

2.1 MARINE FINFISH LARVICULTURE AND LIVE FEED ..................................... 8

3 DISEASES IN AQUACULTURE AND OVERVIEW OF TELEOSTS FISH IMMUNITY 16

3.1 BACTERIAL DISEASES IN AQUACULTURE ................................................ 18

3.2 THE VIBRIONACEAE FAMILY ................................................................... 20

3.2.1 VIBRIO ANGUILLARUM ................................................................... 22

4 DISEASE CONTROL IN AQUACULTURE ............................................................ 25

4.1 SUSTANIABLE BIOCONTROL OF FISH DISEASES ...................................... 26

4.1.1 PHAGE THERAPY IN AQUACULTURE ............................................... 28

4.1.2 PROBIOTICS IN AQUACULTURE ...................................................... 33

5 THE ROSEOBACTER GROUP ............................................................................ 37

5.1 THE PHAEOBACTER GENUS .................................................................... 38

5.1.1 TROPODITHIETIC ACID (TDA) .......................................................... 42

5.1.2 PHAEOBACTER SPP. AS PROBIOTICS ............................................... 47

6 CONCLUDING REMARKS AND FURTHER PERSPECTIVES ................................. 54

7 ACKNOWLEDGEMENTS ................................................................................... 57

8 REFERENCES .................................................................................................... 58

vii

PUBLICATIONS INCLUDED IN THIS THESIS

1. Dittmann KK, Rasmussen BB, Castex M, Gram L, Bentzon-Tilia M. 2017.

(Editorial) The aquaculture microbiome at the centre of business creation.

Microb Biotechnol 10: 1279–1282.

2. Rasmussen BB, Erner KE, Bentzon-Tilia M, Gram L. 2018. Effect of TDA-

producing Phaeobacter inhibens on the fish pathogen Vibrio anguillarum in

non-axenic algae and copepod systems. Microb Biotechnol 11:1070–1079.

3. Rørbo N, Rønneseth A, Kalatzis PG, Rasmussen BB, Engell-Sørensen K,

Kleppen HP, Wergeland HI, Gram L, Middelboe M. 2018. Exploring the effect

of phage therapy in preventing Vibrio anguillarum infections in cod and turbot

larvae. Antibiotics 7(2), 42

4. Rasmussen BB, Kalatzis PG, Middelboe M, Gram L. 2019. Combining probiotic

Phaeobacter inhibens DSM17395 and broad-host-range vibriophage KVP40

against fish pathogenic vibrios. Submitted manuscript

5. Rasmussen BB, Dittmann KK, Gram L, Bentzon-Tilia M. 2019. Upscaling

probiotic Phaeobacter spp. in Tetraselmis suecica algae cultures. Manuscript

in preparation

PUBLICATIONS NOT INCLUDED IN THIS THESIS

1. Dittmann KK, Rasmussen BB, J. Melchiorsen, Gram L, Bentzon-Tilia M. 2019

Probiotic Phaeobacter sp. affect the associated microbiotas of all trophic

levels in aquaculture. Manuscript in preparation

1

1 BACKGROUND AND INTRODUCTION

The United Nations estimate that human population will reach an all-time high 9.7

billion people in the year 2050 (1). With the growing population follows a growing

need for food, including high quality protein, in order not only to feed the

population but also to avoid malnutrition. The Food and Agriculture Organization

of the United Nations (FAO) estimates that 3.2 billion people receive almost 20 %

of their intake of animal protein from seafood, and that 12 % of the world’s

population is fully or partly dependent on seafood production for their income,

the majority of which is involved in wild fish capture (2). A shift in seafood

production is in progress and has been since the mid-1980s. While wild fish

capture has been stagnant around 90 million tonnes a year since 1985 (Figure 1A),

the aquaculture sector has gone from producing approximately 10 million tonnes

in 1985 to 80 million tons in 2016. The aquaculture sector increased its production

with 29.4 % from 2011 to 2016, making it the fastest growing protein production

sector in the world (2). At the same time, the total number of people depending

on seafood production for their income has decreased, while the number

employed in aquaculture has increased (2).

Figure 1 – (A) World capture fisheries and aquaculture production. Excludes aquatic

mammals, crocodiles, alligators and caimans, seaweeds and other aquatic plants. (B)

Global trends in the state of the world’s marine fish stocks, 1974–2015. FAO 2018.

2

With the growing demand for seafood, this increase in aquaculture production is

unlikely to stagnate in the near future, especially since fish stocks are becoming

depleted, with 90 % of fish stocks fully or overexploited (Figure 1B) (2).

In order for humanity to overcome the enormous challenge of having to provide

food and livelihood for the growing population, while coping with the

consequences of climate change and preventing the devastation of Earth, the

United Nations have developed their 2030 Agenda for Sustainable Development

and its 17 Sustainable Development Goals (SDGs) (3). Included in the SDGs are

SDG 2 (ending world hunger and undernourishment), SDG 12 (ensure sustainable

consumption and production patterns) and SDG 14 (conserve and sustainably use

the oceans, seas and marine resources) (3) and if these SDGs are to be met, a

more productive and sustainable aquaculture sector is needed.

A major bottleneck in the production of fish is bacterial diseases, which account

for 54.9 % of all disease outbreaks (4). Bacterial diseases are not only a restriction

on aquaculture productivity but also a cause of major financial losses for the

aquaculture sector. Fish larvae are especially prone to bacterial infection due to

immature immune systems and the use of live feed (5, 6). Live feed including

microalgae, rotifer and Artemia are used to feed larvae of several commercially

important finfish, such as cod, halibut and turbot in the early feeding stage due to

lack of appropriate feed formula (7). Live feed cultures are high in nutrients and a

haven for fast growing opportunistic pathogenic bacteria and live feed can act as

infection vectors and a potential point of entry for pathogenic bacteria (6).

Several methods are available to combat bacterial diseases, including disinfection

of eggs, tanks, etc., vaccination of fish and treatment of diseased fish with

antibiotics (8). Vaccination of fish has been a crucial tool in reducing infection

related mortality in fish farms, however, the selection and efficiency of vaccines

vary from disease to disease, between fish species and depend on the

developmental stage of the fish (9–11). Efficient vaccination of fish larvae is not

possible due to their immature immune systems and therefore antibiotics are still

3

used in larvae cultures (8, 10). However, due to potential development of

antibiotic resistance, which can render the antibiotics useless, sustainable

alternatives to antibiotics are highly sought after (8).

Two types of sustainable biocontrol, which are currently being developed for

aquaculture, are the use of bacteriophages (viruses that infect bacteria) and the

use of probiotic bacteria. Bacteriophages, or phages in short, play an important

role in shaping microbial communities, as well as influencing nutrients fluxes (12).

Using lytic phages to control pathogenic bacteria, the so-called phage therapy, can

be used to reduce pathogen density and/or mortality of live feed and fish (13).

Phage specificity enables phage therapy to target pathogens while leaving the rest

of the microbial community intact, however, phage specificity can also limit the

efficiency of a treatment due to non-targeted strains. In order to circumvent the

problem, several different phages, known as a “phage cocktails” or broad host

range phage, can be used (13).

The use of probiotics, i.e., live microorganisms which confer beneficial effects on

the host health (14), originally focused on intestinal bacteria isolated from

mammals, such as lactic acid bacteria known to provide health benefits in humans

and domesticated land animals (15, 16). However, studies on probiotics for

aquaculture have diversified with more focus on bacteria isolated from

aquaculture environments, e.g., rearing water, live feed or larvae cultures. Some

probiotic preparations exert their health beneficial effects by improving water

quality (e.g. removal of ammonia) whilst others may stimulate the immune

system or antagonise pathogenic agents (15, 17). Probiotics added to feed or

water can decrease deformities, increase fish growth rates and their resistance

towards pathogens (15, 17). Several studies have focused on antagonizing the

opportunistic pathogenic bacteria in the water using probiotic bacteria. Since

aquatic animals are constantly surrounded by water, maintaining high water

quality with low levels of opportunistic pathogenic bacteria is an issue in intensive

aquacultures (18–20).

Marine bacteria of the genus Phaeobacter can have probiotic activity in fish

larvae. They belong to the Roseobacter group (Alphaproteobacteria). Phaeobacter

4

spp., are indigenous to aquacultures, prolific biofilm formers, produce the

antimicrobial compound tropodithietic acid (TDA) and are able to inhibit a range

of fish pathogens in vitro (21–23). Phaeobacter spp. can also inhibit the fish

pathogen V. anguillarum in live feed cultures including several species of

microalgae, rotifer, and Artemia and can protect turbot and cod larvae against

vibriosis (24–28). Most studies conducted with Phaeobacter spp. have been

performed in axenix systems, i.e., systems without background microbiota,

however, one study by Grotkjær et al. showed that Phaeobacter spp. are also

antagonistic towards V. anguillarum in non-axenic systems (24).

In order for Phaeobacter based probiotics to work on a commercial scale, easy

deployment is crucial. Different methods of deployment have been tested, such as

bioencapsulation (rotifers enrichment with Phaeobacter spp.) and biofilters

incorporating Phaeobacter spp. (28, 29). However, since the introduction of the

Phaeobacter spp. before the pathogen results in the best protection of cod larvae,

(25) and since microalgae is the first step in the live feed food chain and the first

potential point of entry for pathogenic bacteria, the upscaling of Phaeobacter spp.

in microalgae cultures should be explored.

Based on the of inhibitory effect of Phaeobacter spp. against Vibrio in axenic (26,

27), as well as non-axenic (24) live feed systems, we hypothesised that

Phaeobacter spp. can be used as a probiotic in non-axenic live feed systems.

Furthermore, given the rapid working lytic abilities of phages and the stabile

inhibitory effect Phaeobacter spp., we hypothesised that combining Phaeobacter

spp. and phages would result in enhanced inhibitory effect against fish pathogens

and enhanced protection of live feed against such pathogens. Therefore, the

purpose of this PhD project was (i) to investigate the use of probiotic Phaeobacter

spp., alone or in combination with bacteriophages, as means of controlling

bacterial infections in non-axenic live feed cultures, and (ii) to examine the

possible practical application of probiotic Phaeobacter spp. in commercial

aquaculture via upscaling in live feed microalgae. This PhD was part of the

ProAqua project (funded by the Innovation Foundation), with the overall aim to

develop a more sustainable aquaculture sector by providing sustainable

5

alternatives to antibiotics as methods for inhibiting pathogenic bacteria in live

feed and larvicultures. Thus, several studies of this PhD project were carried out in

collaboration with researchers from the fields of aquaculture and in phage

therapy.

The thesis consists of a review, describing the background and results of the PhD

study; two published articles, a submitted manuscript, a published editorial and a

manuscript in preparation. The first article (See “Publications included in this

thesis”) focuses on the application of Phaeobacter inhibens as probiotics in non-

axenic live feed cultures, the second explores the use of phage therapy in

larvicultures and the third focuses on combining probiotics and phage therapy in

non-axenic live feed cultures. The editorial explores the potential of microbiome-

based products for aquaculture, while the fourth manuscript concerns the

upscaling of Phaeobacter spp. in live feed algae, and will contribute to developing

processes for applying Phaeobacter spp. in commercial aquaculture.

6

2 MARINE AQUACULTURE

In 2016, the global aquaculture industry produced 80 million tonnes fish at an

estimated value of USD 231.8 billion. Of the total production 28.7 million tonnes

equivalent to 35.9 % stemmed from marine and coastal aquaculture. The marine

and coastal aquaculture production was distributed as follows: finfish 6.6 million

tonnes, crustacean 4.8 million tonnes, molluscs 16.7 million tonnes and 0.4 other

aquatic animals (2).

According to the United Nations Food and Agricultural Organization (FAO), it can

be difficult to distinguish between marine and coastal aquaculture, especially for

finfish aquaculture in East and Southeast Asia, as some species are produced in

both marine cages and coastal ponds. However, the majority of finfish production

in Africa, the Americas, Europe, and Oceania aquaculture is marine (2). Marine

aquaculture is, as the name implies, practised in the marine environment, such as

fjords, however, marine finfish cultures are often land based in early stages, such

as hatching of eggs and larvae cultures.

The marine and coastal finfish production of 6.6 million tonnes is equivalent to 8.3

% of the total fish production in 2016. 56.9 % of the production came from Asia,

27.8 % from Europe, 13.8 % from the Americas, 1.2 % from Oceania, and 0.3 %

from Africa. Atlantic salmon is by far the most commercial important species in

marine finfish aquaculture and in 2016, it accounted for 34.1 % of the total

production. Other globally important species include milkfish (18.2 %), marine

fishes nei (12.8 %), wuchang bream (12.5 %), and rainbow trout (12.3 %) (2).

The most commercial important marine finfish species in Europe are Atlantic

salmon (mainly farmed in Norway, United Kingdom, Faroe Islands and Ireland),

rainbow trout (mainly farmed in Norway, Finland, France, Denmark, Italy and

Turkey), seabass, gilthead seabream and turbot (mainly farmed in Greece, Italy,

France, Spain, Turkey, and Portugal) (30). Cod has previously been the third and

second most produced fish species in Norway and Iceland, respectively, with a

7

total of 26.3 tonnes produced in 2010 (30). Several factors, including low prices,

poor growth rates and high mortality, resulted in the termination of cod

production in 2014 in Norway and 2015 in Iceland. The practice is, however, re-

emerging in Norway (31, 32). All in all, The Federation of European Aquaculture

Producers lists ten different species of marine finfish, which were produced by its

members in its “European Aquaculture Production Report 2008-2016” (30).

The exact production methods and systems used for marine finfish vary from

species to species and occasionally from producer to producer. However, overall

floating net cages are used for finfish once they reach a certain size while onshore

tanks or ponds are used for eggs, larvae, juvenile, and broodstock. Onshore tanks

or ponds are more expensive than net cages, however, they are more easily

accessible, which is needed for broodstock and early development stages.

Onshore tanks or ponds can be divided into Flow-Through Systems (FTS) and

Recirculating Aquaculture Systems (RAS). In FTS a continuous flow-through of

water treated to optimise quality for the farmed species and life stage.

Treatments may include temperature regulation, aeration, disinfection and

filtration in order to remove particles and potential pathogens (33). As FTS need

constant and reliable water supply, they are often found close to shore. RAS are

closed systems, which treat and reuse the water in the system. In RAS water is

treated with several mechanical and biological processes, which include filtration,

removal of organic and inorganic matter and aeration. Due to the extensive

treatment of water in RAS, it is possible to get high water quality, while reducing

water consumption and risk of contamination from external sources (34). The

reduced water usage means that RAS do not need a constant water supply and

thus aquacultures using RAS can be placed further inland compared to those using

FTS.

Environmental pollution from poorly managed aquacultures is a major problem in

several parts of the world. Net cages along coastlines and in areas, such as fjords

and estuaries should be spread out and placed such that the organic waste from

the fish does not sink and impact the sea floor ecosystem. Uncontrolled expansion

of aquaculture leading to greater density than the environment can sustain, may

8

lead to algal blooms of potential toxic algae or hypoxia, with devastating effects

on wildlife. Furthermore, countries, where aquaculture expansion goes

uncontrolled by the authorities, will often also have a high use of antibiotics in

their aquaculture production. One such example is salmon farming in Chile (35,

36). Aquaculture units using closed or semi-closed systems, such as RAS or FTS

make it possible to minimise the risk of environmental pollution, as it is possible

to treat outlet water in order to remove pathogens and pollutants, such as organic

and inorganic matter. Alternatively, it is possible to use co-culturing of other

organisms, such as molluscs, which clean the water by consuming excess feed and

organic waste particles (37).

Another problem facing the marine aquaculture industry, and an obstacle in its

sustainable development, is the dependency of capture fish for feed production.

Most species of marine finfish are carnivorous and thus need a specialised

composition of amino acids, lipids and high concentrations of highly unsaturated

fatty acids (HUFA). In order to provide these nutritional essentials, fishmeal and

fish oil are used, which are mainly produced of Peruvian anchoveta, which are

exploited to their maximum capacity (38). In order to overcome these problems

several different alternatives are currently being explored. These include using by-

catch in fishmeal production and exchanging the main animal ingredients of the

feed with plant based products (39). However, appropriate alternatives to

fishmeal and fish oil have yet to be found.

2.1 MARINE FINFISH LARVICULTURE AND LIVE FEED

Marine finfish larviculture is the hatching and production of finfish larvae and

include the live stages from eggs to juveniles, when the fish assume adult traits

and are ready to digest dry feed. Larvae cultures are usually produced in onshore

tanks for easy access due, to extensive handling in this early phase. Many finfish

start their early life just after hatching as planktonic yolk sac larvae, which do not

share the physiological and morphological traits of adults. At this stage, which is

9

an extension of the embryonic development, the larvae are still developing their

organs, and for several species the mouth is not yet open at the early yolk sac

stage (40). At the end of the yolk sac stage, the larvae's mouths open and they

start feeding, and are now termed feeding larvae. Most fish larvae are carnivorous

and feed on different types of zooplankton (mainly copepods) in the wild (41). In

larvicultures many marine finfish species, including cod, halibut, turbot, seabass

and gilthead seabream, require live feed, i.e., microalgae and zooplankton (rotifer,

brine shrimp Artemia and copepods), due to lack of appropriate feed formula (7).

Suitable and nourishing feed is one of the most important aspects of successful

larvae rearing, as it affects not only larvae growth, but also the development of

deformities, pigmentation, as well as overall health and thus survival (7, 42). The

period the larvae spend growing from newly hatched to juvenile is their most

vulnerable period, as their immune system is immature (see section 3). Thus,

great care concerning maintaining water quality and good hygiene is of utmost

importance at this stage.

Several studies have explored the microbiota of fish larvae including its

development over time and in context with the feed of the fish, as live feed can

have a great impact on the diversity of the larvae microbiota. Although the studies

had different aims, their findings are similar; marine finfish larvae harbour a great

diversity of bacteria species including Vibrio spp., Aremonas spp.,

Pseudoaltermonas spp., Pseudomonas spp., and Tenacibaculum spp. (originally

Flexibacter spp.) (Table 1), the majority without any indication of pathogenic

properties (5). It should be stated that most studies on microbiota of fish larvae

focus on intestinal microbiota. The microbiota of fish larvae have in general been

found to increase in number over time and especially in the period when they are

first introduced to live feed (43–45). Furthermore, the diversity of the larvae

microbiota is known to change over time and with changes in diet (44–46).

10

Table 1 – Examples of major bacterial groups identified in non-diseased marine finfish larvae

Bacteria Fish species Larvae stage Reference

Vibrio spp.

Atlantic halibut,

Atlantic cod, Red

seabream, Black

seabream, Turbot

Yolk-sac Larvae,

feeding larvae

(43, 44, 46–

51)

Photobacterium spp. Atlantic halibut Yolk-sac Larvae

feeding larvae

(43)

Tenacibaculum spp.

(Flexibacter spp.)

Atlantic halibut Yolk-sac Larvae,

feeding larvae

(43)

Pseudoaltermonas spp. Atlantic halibut,

Atlantic cod,

Turbot

Yolk-sac Larvae,

feeding larvae

(43, 45, 47,

50)

Cytophaga spp. Atlantic halibut,

Black seabream,

Turbot

Yolk-sac Larvae,

feeding larvae

(43, 44, 49)

Pseudomonas spp. Atlantic halibut, Red

seabream, Black

seabream, Turbot

Feeding larvae (43, 44, 48,

49)

Marinomonas spp. Atlantic cod Feeding larvae (45, 47)

Ruegeria spp./

Roseobacter spp.

Atlantic cod Feeding larvae (45, 47)

Aeromonas spp. Turbot Feeding larvae (48, 49)

Microbacterium spp. Atlantic cod Feeding larvae (45)

Acinetobacter spp. Turbot,

Black seabream

Feeding larvae (44, 48, 49)

Moraxella spp. Red seabream,

Black seabream

(44)

Enterobacteriaceae Turbot Feeding larvae (49)

11

Microalgae are the most important

live feed for finfish, as it is essential as

food in the production of the

zooplankton live feed and for larvae

of, e.g., cod and halibut (52, 53).

Microalgae can also be used to enrich

zooplankton life feed, in order to

optimise their nutritional value.

Several different species of microalgae

are used commercially as live feed in

aquaculture. The species differ in their

composition of lipids, pigment, vitamins

and minerals etc., and thus the

nutritional value of the different

species varies (52, 53). As a result,

different species of microalgae are

best suited for different species of fish

and shellfish. Noteworthy “aquaculture” species include Tetraselmis suecica,

Rhodomonas salina, Isochrysis galbana, Chlorella vulgaris, Nannochloropsis

oculata, and N. salina (52, 53). Microalgae require a light source and are grown in

specialised algae media, often made within sterilised seawater. Mass-scale

microalgae production is generated via upscaling, in which a starter culture is

grown to a desired density and used to inoculate a larger volume of sterile growth

medium, normally in the range of 1:5 to 1:100. When outgrown, the larger

volume is used to inoculate an even greater volume and so on. The process is then

continued until the desired batch size is reached (52, 53). Alternatively, a

continuous culture, where 20 - 50 % of the volume is frequently harvested and

replaced with fresh medium, can be used (52, 53). While most scientific studies on

microalgae have been conducted in axenic cultures, for example, in order to

minimise variability in study setup, microalgae are not axenic in nature or in

aquaculture settings. Bacteria can have positive effects on algal growth as non-

axenic algae or “bacteria-polluted” axenic algae have been shown to grow faster

and to higher density than axenic algae (Figure 2, upper panel) (54, 55). The

Figure 2 – Algae counts of living and dead

cell in axenic cultures vs. non-axenic

cultures of Thalassiosira rotula. Modified

from Grossart and Simon 2007.

12

enhanced growth is attributed to a release of re-mineralised nutrients, vitamins

and other growth factors by the bacteria, that promote algal growth, while living

or dead algae provide organic substances to the bacteria. Bacteria can be

pathogenic to the algae, and the decline in non-axenic algae after the growth

phase is more rapid compared to axenic algae (Figure 2, lower panel) (54).

Bacterial loads in cultures of aquaculture relevant algae were found to vary

between 5.7-16.4 CFU/algal cell and 0.2-4.3 CFU/algal cell depending on algae

species (56). Algae cultures can contain opportunistic bacteria, such as Vibrio spp.,

however, concentrations and occurrence are generally lower than in zooplankton

based live feed cultures (52). The dominating species in aquaculture relevant

microalgae, have generally been found to be Alpha-proteobacteria and of the

Cytophaga-Flavobacterium cluster (56, 57).

Rotifers, also known as wheel animals, are small filter feeders, that as live feed are

fed on yeast, bacteria or microalgae, such as Tetraselmis spp. In a live feed

context, rotifers most commonly refer to members of the genus Brachionus.

Rotifers are commonly used as live feed for species, such as cod, halibut, turbot,

sea brass and seabream in the early larval stage, as their size are optimal at this

stage. Rotifers have many advantages: they can be easily cultivated in high

densities, swim slowly, have high saline and temperature tolerance and have a

high reproductive rate (58, 59). However, using rotifers also has several

disadvantages. Rotifer cultures are prone to diseases and sudden crashes in

populations, thus several separate tanks are required. Since cultures are often

continuously grown throughout the year, maintenance is an expense, even in off

seasons (58). The nutritional value of rotifers, which vary depending on their diet

(yeast or algae), is not optimal and is lower than the nutritional value of

copepods. This issue can to a certain extent, be overcome by enrichment of the

rotifers (59). Organic loads in rotifer cultures are very high, promoting high

numbers of bacteria and numbers above 3.5 x 109 CFU/g rotifer wet weight have

been reported for enriched rotifers (60). Bacterial communities of rotifers seem to

be dominated by Vibrio spp., Pseudomonas spp., Morella spp., and

Cytophaga/Flavobacterium spp. when analysed using culture-based methods (51,

61–63). The microbiota community composition has been reported to be different

13

between different B. plicatilis strain cultures, and community composition appear

to influence growth rates of the rotifers (64). Although opportunistic pathogens

are readily found in rotifer cultures, rotifers are mostly unaffected by them, and

thus the pathogens become a problem only when the rotifers are fed to fish

larvae.

Brine shrimp Artemia is a genus of aquatic crustaceans native to hyper saline

water, the only genus in the family Artemiidae. Of the seven species of Artemia

currently known, A. salina and A. franciscana are the best studied (58). In the

wild, Artemia feed on microalgae, however, as live feed they can also be fed

particulate foods, such as yeast (58). Artemia produce resting eggs (cysts), making

them easy to use in aquaculture. In addition, they can be easily stored, have long

shelf life and batches of Artemia can be produced when needed, resulting in low

cost. Artemia are large compared to other types of live feed and thus they will

often not be used until more advanced life stages. However, if large live feed is

needed, Artemia can be grown up to one cm in length. As with rotifers, the

nutritional value of Artemia is not optimal and enrichment is often used (65). In

order to minimise initial bacterial loads, cysts can be disinfected or decapsulated

prior to incubation, which is important as bacterial loads increase dramatically

during hatching, and then potentially again during enrichment (43, 58). As for

rotifers, bacterial loads can be very high and 1010 CFU/g Artemia wet weight have

been reported for enriched Artemia (60). Although newly hatched Artemia nauplii

have been found to predominantly be colonized by uncultured members of

Gammaproteobacteria and Planctomycetales, the microbiota of feeding Artemia

seem to be similar to that of rotifer and dominated by Vibrio spp., Pseudomonas

spp. and Cytophaga/Flavobacterium spp. (62, 66, 67). Although several Vibrio spp.

have been reported as pathogenic towards Artemia they relatively robust against

most pathogens (27, 68).

Copepods are the most abundant type of zooplankton in the oceans and an

important food source for many higher organisms. Copepods are thought to be

the natural live feed for fish larvae, and have been shown to be nutritionally

optimal for several marine fish species (41, 65). There are three major orders of

14

copepods which have been considered as live feed in aquaculture; Calanoida,

Harpacticoida, and Cyclopoida, with the best studied genera the Calanoida order

members Centropages, Eurytemore and Acartia (58). Copepods have several

development stages thus, even copepod species, which are large as adults, start

out as small nauplii, which allow for copepods to be used for a longer period of

time than traditional live feed (69). Calanoida copepods, such as Acartia tonsa

nauplii and adults mainly feed on phytoplankton as suspension feeders and

therefore copepods are often fed microalgae in aquaculture settings (69). There

are several advantages in using copepods as live feed for fish larvae. Copepods

have desirable amino acid profiles and a high fatty acid content, eliminating the

need for enrichment of the live feed (65, 69, 70). Furthermore, copepods swim in

a zig-zag pattern which promotes feeding in fish larvae (71). Several studies have

shown that copepods are superior to other types of live feed. For example,

substituting rotifers or enriched rotifers with copepods (Acartia tonsa) in the first

feeding stage of Atlantic cod and ballan wrasse larvae, have a long term positive

effect on survival, growth, and viability (72). Despite all the above mentioned

advantages, copepods are only seldom used as live feed in aquaculture, mainly

becausecopepod production methods have been inefficient and too costly to

compete with rotifer and Artemia (58). However, as production methods are

further developed and made more cost-effective, a shift from rotifer and Artemia

to copepods is to be expected. Recently Abate et al. showed that using the

copepod A. tonsa in semi-intensive juvenile turbot farming was a cost-efficient

alternative to traditional live feed (Artemia) (73). Since copepods are seldom used

as live feed in aquaculture, the majority of studies on their associated microbiota

focus on the natural environment. Copepods are well known to function as

natural reservoirs for Vibrio spp., including the human pathogen Vibrio cholera

(74–76). In accordance Gugliandolo et al. predominantly found Vibrio spp.,

including V. parahaemoliticus, V. vulnificus and V. alginolyticus, when surveying

the natural microbiota of copepods in the Mediterranean Sea, however, they also

found Aeromonas spp. including A. sobria and A. hydrophila (77). Like rotifers,

copepods seem to be mostly unaffected by the associated opportunistic

pathogens. In this study we found that nauplii of the copepod A. tonas was

unaffected by the two serotype O1 strains (90-11-286 and NB10) of the fish

15

pathogen Vibrio anguillarum used in the experiments, with no noticeable

differences in mortality between control cultures and cultures with V. anguillarum

(78).

Since microalgae are the first step of the live feed food chain, they are also a

critical point of entry for pathogenic bacteria, as the bacteria can potentially

spread throughout the chain and eventually end up in the fish larvae cultures.

Furthermore, due to heavy loads of organic material, live feed cultures (especially

zooplankton cultures) are havens for fast growing opportunistic bacteria including

pathogens, which in many cases can use the live feed as infection vectors and

spread from live feed to the fish larvae (5, 6).

Several studies have shown that Artemia can function as infection vectors for

pathogenic bacteria, such as Vibrio spp., and although bacterial loads can be

lowered dramatically using washing steps, these do not remove bacteria situated

in the gut of the Artemia (66, 79, 80). While studying the gut-associated

microbiota during the development of Atlantic halibut larvae, Verner-Jeffreys et

al. found that the first feeding larvae fed enriched Artemia were generally

colonised by live feed associated bacteria, predominantly Vibrio spp. (43).

Furthermore, the study reported that bacterial loads, measured as CFU per

Artemia nauplii or copepod, were highest for enriched Artemia (> 104 CFU

nauplius-1), followed by unenriched Artemia (< 103 CFU nauplius-1), while the

lowest bacterial loads were found for copepods (5.36 x 102 CFU copepod-1). While

most studies seem to explore or exploit Artemia as infection vectors, rotifers can

also function as infection vectors (51, 61) exemplified by Munro et al., who

effectively used rotifers exposed to the opportunistic pathogen V. anguillarum to

infect turbot larvae, resulting in increased mortality (81).

As previously mentioned, copepods are known to be hosts for a natural

population of Vibrio spp., including the human pathogen V. cholerae and can

potentially function as infection vectors for fish pathogens. However, not all vibrio

species can colonise all copepod species, as shown by Dumontet et al. who found

that V. parahaemolyticus, V. alginolyticus and V. mimicus were unable to colonise

the copepod species Temora stylifera, Acartia clausi, Centropages typicus and

16

Paracalanus parvus (82). In this study, using a GFP tagged V. anguillarum (strain

NB10, serotype O1), we found that the fish pathogen V. anguillarum will colonise

the surface and gut of the copepod A. tonsa, thus making it a potential infection

vector (Figure 3) (78).

Figure 3 - Microscopy of A. tonsa nauplii colonized by GFP-tagged V. anguillarum

NB10. Red fluorescent cells are Rhodomonas salina microalga. Magnification 100 X. A)

Phase-contrast microscopy. B) Fluorescence microscopy (WIB excitation 460–490 nm,

emission > 515 nm). Rasmussen et al. 2018.

3 DISEASES IN AQUACULTURE AND OVERVIEW OF TELEOSTS FISH

IMMUNITY

Intensive fish farming is prone to outbreaks of infectious diseases, and once

diseases get a foothold they will often spread rapidly due to the high density of

animals. Outbreaks of infectious diseases can affect aquacultures in numerous of

ways. Diseases can result in mortality or make products unsellable. Diseases can

also potentially be a health risk for employees and their prevention and control is

time consuming and costly. Thus, disease outbreaks can have great economic

consequences for the aquaculture producers and can result in lost crops, lost jobs

and export revenue. In one of the worst cases described in the literature, Asian

17

shrimp farms were ravaged by white spot syndrome virus in 1992 and 1993, which

ended up costing the industry approx. $6 billion. In 1999 a similar outbreak in the

Americas caused a loss of $1-2 billion (83). In both Europe and the Americas, sea

lice, ectoparasitic copepods of the genera Caligus and Lepeophtheirus, are the

most damaging parasites in salmonid farming. The cost of preventing sea lice in

Norway, Scotland, Canada, USA, and Chile has been estimated to comprise an

average of 6 % the value of the total production of salmonids (84).

Although both parasites and viral diseases can cause major problems in

aquacultures, the most common cause of infectious diseases in aquaculture is

bacteria (54.9 %), followed by viruses (22.6 %), parasites (19.4 %) and fungi (3.1 %)

(4).

Both in the wild and in aquaculture, fish are constantly exposed to pathogens.

Teleosts fish have two immune systems for defense; the innate system and the

adaptive system. The innate system consists of physical barriers, as well as cellular

and humoral immune response. Examples of the innate immune system are skin

mucus, which harbours antimicrobial compounds, such as antimicrobial peptides

and lytic enzymes, and the low pH in the gastrointestinal tract which makes it a

very hostile place for pathogens (85–87). Another important part of the innate

immune system is the complement system which helps clear the fish of pathogens

via mechanisms, such as phagocytosis (87, 88). However, limited research on the

innate immune system suggests that it also plays an important role in the

activation of the adaptive system (87). The adaptive immune system includes

“memory” cells and receptors that allow the system to remember infections and

to react faster and more efficiently to recurring infections. For example, by

binding antibodies to the surface of bacteria which activates the complement

system. The “memory” of the adaptive immune system is what makes it possible

to efficiently vaccinate fish (86). The immune system of fish is influenced by

various internal and external factors. Factors, such as water temperature,

pollution, stress, food additives and immunostimulants can all affect the immune

response (85–87). Thus, the conditions in which fish are farmed, including stock

density and water quality, are important for the fish's susceptibility to infection.

18

When fish larvae hatch, they are solely relying on their innate immune system, as

the adaptive system is not yet developed. Furthermore, the innate immune

system is not fully developed at this stage, and the pH in the gastrointestinal tract

is not as low as in adults (89, 90). The time it takes for fish to fully develop their

immune system varies from species to species, for example, Atlantic cod larvae

take between two and three months to fully develop their immune system (91).

This results in the early larval stages being the most critical stages, as the larvae

are much more susceptible to infections in this period. In addition, several fish

species are fed live feed that may carry pathogenic bacteria that could potentially

induce mass mortality.

3.1 BACTERIAL DISEASES IN AQUACULTURE

As mentioned previously more than half of the infectious diseases in aquaculture

are caused by bacteria (4). The bacterial pathogens, infect different species and

different live stages, some are normally only found as pathogens of eggs and

larvae, while others only of grown fish. It should be noted that infections in larvae

cultures are commonly not identified in commercial hatcheries, as infections can

be difficult to identify in the larvae and can result in slow and continuous

mortalities, which are easily overlooked. However, infection in fish larvae cultures

can also result in population crashes as a result of mass mortality (Figure 4) (45).

Although a large number of bacterial species have been described as fish

pathogens in the literature, relatively few bacterial species are responsible for the

majority of outbreaks and mortalities.

19

Figure 4 - Viable cod larvae in production tanks over time. Note population crash in

Tank N8 at 73 days post hatching, due to bacterial infections. Reid et al. 2009.

Vibrio spp., such as V. anguillarum and V. splendidus (see sections 3.2 and 3.2.1)

are opportunistic pathogens that cause the disease vibriosis in a wide range of fish

species, including cod, turbot and salmonids (92, 93). Aeromonas salmonicida of

which there are several subspecies cause furunculosis in salmonids and non-

salmonids. A. salmonicida can be found in both fresh-, brackish- and seawater,

and A. salmonicida subsp. salmonicida can infect turbot and halibut yolk sac

larvae in challenge trails (94–96). Other pathogens of salmonids are Piscirickettsia

salmonis, which causes piscirickettsiosis in, among others, coho- and Atlantic

salmon and rainbow trout, Pasteurella skyensis the causing agent of pasteurellosis

in Atlantic salmon, and Renibacterium salmoninarum, an important Gram-positive

pathogen of salmonids, which causes bacterial kidney disease (BKD) (97, 98).

Flavobacterium psychrophilum, which thrives in brackish water can, among

others, infect coho salmon and rainbow trout. F. psychrophilum infection are

referred to as both bacterial coldwater disease (CWD) and Rainbow Trout Fry

Syndrome (RTFS), depending on the infected host (99). Francisella philomiragia

has been a problem for Norwegian cod farmers. F. philomiragia is the causing

20

agent of francisellosis which was first described in Norwegian farmed cod in 2004

and since in Chilean farmed Atlantic salmon (100–103). Lactococcus garvieae

(originally Enterococcus seriolicida) and L. piscium, along with several species of

Streptococcus can cause lactococcosis or streptococcosis in both freshwater and

marine fish species including eel, seabass, Atlantic salmon, and turbot (104).

Pseudomonas anguilliseptica causes red spot disease or winter disease in species,

such as Seabream, eel, turbot, and ayu (105, 106). Tenacibaculum spp., including

T. maritimum (originally Flexibacter maritimus), is the causing agent of marine

flexibacteriosis, which affect several species including turbot, salmonids, sole,

seabass, gilthead seabream, red seabream and flounder (107–109). The closely

related egg pathogen T. ovolyticum (originally F. ovolyticum) has been shown to

infect larvae of Atlantic halibut (110).

A large range of different bacteria have been isolated from marine finfish eggs

and larvae, including species related to the above mentioned pathogens, although

the majority have not shown pathogenic properties (5). One group of bacteria ,

which have continuously been coupled to infections and mortality in both larvae

and grown fish, are members of the family Vibrionaceae of the

Gammaproteobacteria (5, 111–113).

3.2 THE VIBRIONACEAE FAMILY

As indicated above, species of the Vibrionaceae family are among the most

prominent fish pathogenic bacteria. The Vibrionaceae family of the

Gammaproteobacteria consists of seven genera of which the most important in

an aquaculture and disease context, in decreasing order, are Vibrio, Aliivibrio and

Photobacterium, (105, 106, 114–116). Vibrionaceae members are facultative

anaerobic Gram-negative bacteria and the family is widely known for its human,

as well as fish pathogens (114–116).

21

In the genus Photobacterium, P. damselae subsp. piscicida (Pasteurella piscicida)

and P. damselae subsp. damselae can cause photobacteriosis, also known as

pasteurellosis, in several finfish species including seabream, seabass, sole, striped

bass, yellowtail, rainbow trout and Atlantic salmon (104, 106, 117).

Of the Aliivibrio genus, especially three species; A. logei, A. salmonicida and A.

wodanis can cause problems in aquaculture. A. logei has been reported as

pathogen of cod larvae and Atlantic salmon (45, 118, 119). A. salmonicida can

cause the disease cold-water vibriosis in salmonids including Atlantic salmon and

rainbow trout and has also been found in non-salmonids, such as cod (120). A.

wodanis have been found to infect salmonid fish with winter ulcer, caused by

Moritella viscosa (originally Vibrio viscosus), thus worsening the condition (121). It

should be mentioned that the roles of M. viscosa and A. wodanis in winter ulcers

are not fully understood and it has been suggested that the two species work

together synergistically, resulting in the single disease condition (106).

The Vibrio genus, which is the largest in the Vibrionaceae family include several

species, which are both pathogenic towards humans and/or fish (106, 114). The

most famous and the first to be described is the waterborne human pathogen

Vibrio cholerae, which is the causative agent of the cholera disease (114, 122).

Vibrio spp. can be isolated as part of the aquaculture microbiota, where they can

become a problem if conditions favours it (114, 123). V. parahemolyticus and V.

vulnificus are both fish and shellfish pathogens, commonly isolated from seafood

and can cause food-borne diseases in humans usually after ingestion of raw or

undercooked shellfish like oysters (124–128). Several Vibrio species including V.

splendidus, V. alginolyticus, V. harveyi, V. ordalii and V. anguillarum infect

commercially important fish species, such as turbot, seabass, striped bass,

flounder, eel, ayu, cod, salmonids and seabream. Several of these species have

also been associated with mortality in fish larvae (45, 93, 114, 129–133). V.

splendidus has been identified as the causing agent of mortalities in hatchery-

reared turbot larval and has been shown to infect and kill halibut larvae in

challenge trails (130, 132). V. alginolyticus is able to infect and kill challenged

larvae of cod and halibut and has been associated with mass mortality in gilthead

22

seabream larvae (132, 134, 135). One of the most important Vibrio spp. in an

aquaculture context is the fish pathogen V. anguillarum (92, 93).

3.2.1 VIBRIO ANGUILLARUM

V. anguillarum, also known as Listonella anguillarum, is a highly infectious fish

pathogen and can cause disease in more than 50 fish species. It is an opportunistic

pathogenic bacteria found in salt or brackish water, where it causes the fatal

haemorrhagic septicaemic disease vibriosis (92, 93, 114, 136). V. anguillarum is a

comma-shaped motile rod, Gram-negative, non-spore-forming, halophile and

facultative anaerobic (92, 136, 137). V. anguillarum strains have been isolated

from a wide range of fish species including cod, rainbow trout, sea trout, brown

trout, pike, Atlantic salmon, sockeye salmon, coho salmon, turbot, Ayu, European

plaice, sea brass, halibut, European eel and Japanese eel (93, 132, 137–139). A

total of 23 serotypes have been reported for V. anguillarum with serotypes O1,

O2 and to some extend O3 being the major causative agents of vibriosis (93, 132,

138, 140). Due to their prevalence in isolates from diseased fish O1 and O2 are the

best studied serotypes.

Vibriosis, which has also been known as

salt-water furunculosis (141) and ulcer or

boil disease (142, 143), can affect fish,

bivalves and crustaceans, including

several aquaculture relevant once (92,

93, 106, 114, 144). Fish infected with

vibriosis will typically lose weight and get

skin discoloration, including red spots

and ulcers on the side and underside and

eventually die (Figure 5). Besides vibriosis

V. anguillarum has also been linked to fin

rod in Chinese turbot (145).

Figure 5 – Tail ulcers on sea bass

infected with V. anguillarum.

Haenen et al., 2014.

23

Although debated, several potential points of entry have been identified for V.

anguillarum infections in finfish, including skin, fins, gills and oral intake of

contaminated feed and water (146, 147). Oral intake in adult , although of minor

importance, might be of considerable importance in fish larvae as their

gastrointestinal track is not yet fully matured and thus its pH is not as low as that

of the adult fish (148). Grisez et al. reported extremely high mortalities in turbot

larvae feed V. anguillarum infected Artemia (79). Furthermore, using

immunohistochemical examinations Grisez et al. showed that the intestine was

indeed the point of entry of V. anguillarum in the challenged larvae, as it was

taken up by the intestinal epithelium by endocytosis and then via the mucosa

spread to the blood and on to the internal organs.

V. anguillarum has several virulence factors, which are thought to be important

for a successful infection. V. anguillarum uses chemotaxis, the movement of a cell

in response to chemical stimuli and motility, to locate and move towards potential

hosts by sensing a range of components, including amino acids and carbohydrates

in the mucus layer on the skin and gut of the host (149, 150). Once in contact,

adhesion plays an important role in colonisation and invasion of host tissue. V.

anguillarum is thought to use several cell-surface components, including fimbriae,

pili, lipopolysaccharide, outer membrane proteins and extracellular

polysaccharides when adhering to its host (92). V. anguillarum is thought to

produce several compounds, which are important for virulence, including the

extracellular zinc metalloprotease EmpA, which is important for degradation and

penetration of skin and intestinal mucus (151, 152). A range of haemolytic

exotoxins have been genetically identified in V. anguillarum, which, if produced,

are thought to lysis different host cells, including red blood cells, mast cells and

neutrocytes thus enabling colonisation and destruction of host tissue causing a

systemic infection (153–156). The virulence of V. anguillarum is regulated by

quorum sensing (QS), i.e., population density dependent communication and

sigma factors (157, 158). Recently Rønneseth et al. compared the virulence of

twenty-nine V. anguillarum strains against cod, turbot and halibut larvae at two

different doses and found that virulence between V. anguillarum strains differs

greatly and that virulence for some strains are extremely dose depended, which

24

have implications for the prevention of V. anguillarum mediated infection (132).

The mechanism behind the differences in virulence of the V. anguillarum strains

was investigated by Castillo et al. in a parallel study to Rønneseth et al. using

genome comparison of the same strains (156). Six out of nine highly virulent

strains contained unique dispensable genomic material, including pathogenic

genomic islands, prophage-like elements and virulence factors, while medium to

non-virulent strains had a high degree of homogeneity in their genomic makeup

(Figure 6). The genomic islands of the highly virulent V. anguillarum strains

encoded among other for a range of toxins, including zonula occludens toxins

(Zots) previously described in V. cholerae, which suggests that the highly virulent

V. anguillarum has received virulence related genes from other pathogens.

Figure 6 - The pan-genome of V. anguillarum. The flower plots represent the number

of shared (core) and specific (dispensable) genes based on cluster orthologs for each

chromosome. Petals display numbers of strain-specific genes found in each genome of

V. anguillarum strains with core gene numbers in the center. The gray colors indicate

the virulence category as found in three fish larva model systems by Rønneseth et al.

2017. Castillo et al. 2017.

25

4 DISEASE CONTROL IN AQUACULTURE

Bacterial diseases in aquaculture are controlled by several methods and new and

improved methods are constantly being researched. As no method is perfect,

several different methods are often used in combination to create the best

possible protection against diseases. It should be noted that the following will

focus on control of pathogenic bacteria and mainly on fish larvae cultures since

this is the most vulnerable period in the production.

The first and primary “line of defence” when it comes to disease control in

aquaculture is proper hygiene and good water quality. A high level of hygiene

minimises the risk of carryover from one batch to another via employees or

equipment and thus disinfection of equipment between batches is important.

Furthermore, area specific equipment and clothing for employees helps prevent

spreading of bacteria from one part of a fish farm to another. Good water quality

is important but can be difficult to achieve due to the often intensive scale of

commercial fish farms and as a consequence high nutrient levels will cause

proliferation of opportunistic bacteria, including pathogens (159). In a closed

system, such as a larvae rearing tank, water can be treated in a number of ways in

order to maintain water quality and avoid pathogens, including sterile filtration,

bobbling with ozone, heating, or electrolysis (160).

Surface disinfection of fish eggs with, for example, iodine, glutaraldehyde or H2O2

can be performed in order to prevent carryover from broodstock to larvae.

Surface disinfection of fish eggs has shown to increase survival and improve

hatching and development of larvae of different marine finfish (161–163).

Antibiotics have been and in some parts of the world still are, used extensively to

treat bacterial infection, including prophylactically. However, there are several

disadvantages with using antibiotics. Antibiotics kill all susceptible bacteria

without discriminate between pathogenic and non-pathogenic, which lead to an

environment that favours fast growing opportunistic bacteria, such as V.

anguillarum, which further can lead to the need for new antibiotic treatment (8,

159). Heavy use of antibiotics, including prophylactic use can lead to development

26

of antibiotic resistance, which not only can render an antibiotic useless, but

resistance genes and antibiotic resistant bacteria can spread from aquacultures to

the environment and further to human pathogens (8, 164). Due to antibiotic

resistance, the associated health risks and the introduction of vaccines, the use of

antibiotics have dropped dramatically in European and North American

aquaculture. An example of this can be seen in Norway, where the amount of

antibiotics used was reduced from 887 mg antibiotics per kg fish produced in 1987

to 0.4 mg antibiotics per kg fish produced in 2014 (165). In other parts of the

world, such as Chile and South East Asia, large quantities of antibiotic are still used

(8, 166, 167). Furthermore, as the fish cannot be efficiently vaccinated, antibiotics

are still used in fish larvae culture. Thus, alternative methods for sustainable

disease- and biocontrol are needed.

4.1 SUSTANIABLE BIOCONTROL OF FISH DISEASES

Two major groups of sustainable aquaculture biocontrol technics and methods

exist, some more developed than others; one group focuses on utilising and

stimulating the immune systems of the fish while the other group focuses on

inhibiting the growth (or killing) of opportunistic pathogenic bacteria.

As mentioned vaccination of fish has been

one of the main drivers in replacing antibiotics

with sustainable alternatives. Once a fish has

reached a certain size and has a sufficiently

developed immune system, it can be

vaccinated and vaccination of juvenile and

adult fish is common practise against many

diseases. Vaccines are normally made from

parts of or whole inactivated pathogens and

fish can be vaccinated via oral administration,

injection, or immersion (Figure 7). The idea

behind vaccinations is that by introducing, for

example, an inactivated bacterial pathogen,

Figure 7 - Manual vaccination of Atlantic salmon via injection. Sommerset et al. 2005.

27

the immune system is stimulated and will at a later stage be able to recognise and

more efficiently deal with the pathogen (9–11). Since 1976 when the first

commercial fish vaccines became available a large range of vaccines have been

developed, spanning a range of pathogens and fish species (9, 168). Although

vaccines are an efficient type of biocontrol, there are still pathogens for which

there are not effective vaccines and only fish with sufficiently developed immune

systems can be efficiently vaccinated.

As fish larvae cannot be vaccinated due to their immature immune systems,

alternative methods, which boost their immune systems have been invented,

such as immunostimulants. By adding non-specific immunostimulants, such as β-

glucans from yeast to live feed or water of the fish larvae, it is possible to activate

and boost the innate immune system of the fish larvae and commercial

immunostimulants are available (169, 170).

Quorum sensing (QS) is important for the regulation of virulence factors in many

pathogenic bacteria, including V. anguillarum. Quorum sensing inhibitors (QSI)

have been tested in attempts to make pathogens less- or avirulent. QSI are small

molecules, which interfere with QS without affecting bacteria growth and thus,

QSI have the advantage that the method does not kill any bacteria, such as

antibiotics, which leaves room for growth of new opportunistic pathogens (8,

171). The QSI furanone C-30 were shown to reduce mortality in rainbow trout

challenged with V. anguillarum 90-11-287, however, at higher concentrations the

compound were toxic rainbow trout and thus, the authors concluded that less

toxic QSI is needed for the method to have practical application in aquaculture

(171). The fact that QS controlled virulence and QSI are poorly understood was

shown by Tinh et al. who found that N-acyl homoserine lactones disruption by

lactonase enzyme had no effect on the virulence of V. anguillarum NB10 and on

mortality of sea bass and Artemia challenged with the bacteria (172).

Microbial mature water is a term, which describes water that has been filtered in

order to remove the majority of bacteria and nutrition and then re-colonised by

non-opportunistic bacteria via biofilters. The idea behind microbial mature water

is that by promoting growth of slow growing non-opportunistic bacteria over fast

28

growing opportunistic bacteria, the non-opportunistic bacteria will protect the

fish larvae from colonisation of fast growing opportunistic pathogens, such as V.

anguillarum (159, 173). Several studies exploring that effect of microbial matured

water have found enhanced survival and growth of marine fish larvae, including

Atlantic halibut, turbot and cod (173–175).

Other sustainable types of biocontrol are the use of probiotic bacteria and

bacteriophages to inhibit pathogenic bacteria. Probiotic bacteria and

bacteriophages do not require the immune systems of the fish to be at a certain

developmental stage and thus the methods are better suited for fish larvae then

some of the above mentioned methods.

4.1.1 PHAGE THERAPY IN AQUACULTURE

One emerging type of sustainable biocontrol for aquaculture is the use of

bacteriophages, or phages in short, to control bacterial growth, the so-called

phage therapy. The most abundant entity in the marine environment is viruses

and phages, which are virus that target, infect and kill bacteria, play an important

role in shaping the microbial community and driving its evolution (12, 176, 177).

Phages are divided into two main groups; lytic phages and lysogenic phages (also

known as temperate phages) based on their two different lifecycles (Figure 8).

Figure 8 - Overview of the two main life cycles of bacteriophages. Modified from

Oliveira et al. 2012.

29

It should be noted that some phages are capable of carrying out both and that

only lytic phages are relevant in a biocontrol perspective (13).

Phages need to infect a host in order to replicate and proliferate. A lytic phage

infection is carried out in the following steps: attachment, infection, replication,

assembly, lysis and burst. In the first stage the phage attach to a specific receptor

on the surface of its host, for example, an outer membrane protein. The phage

then injects its genetic material into the host, which can consist of doublet or

single stranded DNA or RNA. The genetic material of the phage is then replicated

and new phages are assembled. The new phages lyse the bacteria cell and the

phages are released when the bacterial cell wall breaks, which is referred to as the

burst. The new phages are now ready to infect a new host and the cycle can start

all over (13).

When phages bind to their host, binding proteins of the phage interact with very

specific proteins or lipopolysaccharides, the so-called phage host receptors on

bacterial cell surface. The phage and the phage host receptor need to match in

order for infection to occur, which makes phages very specific in their host. Often

phage are so specific that they will only infect certain strains within the same

species of bacteria, however, broad host range phages, which are able to infect

several different species also exists (178). The specificity of phages is one of the

“selling points” of phage therapy as, unlike antibiotics, it is possible to target

specific pathogenic bacteria without disrupting the overall microbial community.

Phage specificity also makes it possible to combine phage therapy with bacterial

based biocontrol, such as probiotics. The specificity of phages can also be a

problem. If a phage is too specific it might be able to infect and kill the bacteria,

which cause problems. This, however, can to a large extent, be overcome by using

broad host range phages or “phage cocktails”, where several phages with

different host ranges are combined.

Phage therapy is a relatively new discipline in an aquaculture context, with some

of the first experiment publications on the subject only two decades ago. Nakai et

al. showed that phage therapy could protect yellowtail against Lactococcus

garvieae using intraperitoneally or orally administrated anti-Lactococcus garvieae

30

phage PLgY-16 (179). The biggest effect was seen in injected yellowtails, where

survival rates of the challenged yellowtails increased from 10 % to 100 %, 80 %

and 50 % for the groups injected 0, 1 and 24 h after bacterial challenge,

respectively. Furthermore, high numbers of plaque-forming units (indicating

active phages) were isolated from the intestine and spleens of the yellowtails until

48 h after treatment, indicating that phages can be effective in treating diseased

fish. In a later study Park and Nakai showed that feeding of phages impregnated

pellets decreased mortality of ayu infected with Pseudomonas plecoglossicida

(180).

Due to their importance in aquaculture, several studies have focused on inhibition

of Vibrio spp. by vibriophages. Using V. parahaemolyticus as model, Mateus et al.

showed that the phages alone or in cocktails of two or three could inactivate the

bacteria with the cocktail being the most efficient (181). Furthermore, they

observed that the phages remained viable for several months in marine water.

Studies have also shown that phages can decrease vibrio induced mortality in

Atlantic salmon, zebrafish and Asian tiger shrimp (182–185). Especially one very

successful study found that the phage CHOED could completely eliminate

mortality in Atlantic salmon caused by the highly virulent V. anguillarum PF4 when

treated with high multiplicity of infection (MOI; number of phages per bacterial

host) (183).

In this PhD study (in collaboration with researchers from among other

Copenhagen University, Denmark and University of Bergen, Norway), we found

that the broad host range vibriophage KVP40 was able to reduce and/or delay the

mortality of cod and turbot larvae challenged with four different V. anguillarum

strains (186). We further found that the phage was able to reduce mortality

induced by natural pathogens, originating from the fish eggs. The natural

pathogens most likely included Vibrio spp. based on colony counts and

morphology on thiosulfate-citrate-bile salts-sucrose (TCBS) agar plates and the

known host range of KVP40. In the combination study, where we combined P.

inhibens and KVP40, we found that KVP40 was able to effectively reduce V.

anguillarum counts in T. suecica and Artemia cultures, although regrowth was

31

seen in the Artemia cultures (187). Bacterial phage tolerance and resistance can

become a problem when using phage therapy. Phage resistance or tolerance can

be non-mutational, as well as mutational (188–190) the latter of which may

develop within hours of treatment (8, 191). In order to investigate the regrowth in

the Artemia cultures, we isolated and tested phage susceptibility of thirty-two V.

anguillarum colonies. All colonies were susceptible to KVP40 and thus, the

regrowth of V. anguillarum observed in the Artemia cultures is likely due to non-

mutational phage resistance/tolerance, which V. anguillarum is known to be

capable of (188, 190). Despite the observed regrowth of V. anguillarum in the

Artemia cultures, KVP40 decreased mortality of non-challenged Artemia

compared to the control and delayed mortality in Vibrio-challenged Artemia

(Figure 9) (187).

Figure 9 - Mortality of non-challenged and vibrio-challenged Artemia in non-axenic

systems. Non-challenged control (●, solid line), Non-challenged; KVP40 treated (■,

solid line), V. anguillarum challenged control (▲, dashed line), V. anguillarum

challenged; KVP40 treated (▼, dashed line). Error bars show the standard deviation

of three biological replicates. Modified from Rasmussen et al. 2019.

32

Several technics, such as using phage cocktails or combining phage therapy with

other types of biocontrol have been suggested as means to overcome phage

tolerance and resistance. In this study, we combined the vibriophage KVP40 with

the probiotic Phaeobacter inhibens DSM17395 against V. anguillarum in non-

axenic T. suecica and Artemia systems. Combining the treatments resulted in

earlier inhibition compared to probiont only and lower vibrio counts than phage

only in both the T. suecica systems (Figure 10A) and the Artemia systems (Figure

10B). Regrowth of V. anguillarum was only observed in the phage only treated

Artemia and not in Artemia systems treated with the phage probiont

combination, which indicates that the probiont prevented the regrowth of V.

anguillarum (187).

Figure 10 – A) Vibrio anguillarum counts in non-axenic Tetraselmis suecica systems. B)

V. anguillarum counts in non-axenic Artemia systems. V. anguillarum alone (●), V.

anguillarum with P. inhibens (■), V. anguillarum with P. inhibens and KVP40 (▲), V.

anguillarum with KVP40 (▼). Arrows on panel B indicate vibrio counts below

detection level (10 CFU ml-1) as seen for treatments with P. inhibens and with P.

inhibens and KVP40 at time points 3 days and 4 days. Error bars show the standard

deviation of three biological replicates. Rasmussen et al. 2019.

33

4.1.2 PROBIOTICS IN AQUACULTURE

Probiotics, defined by the Food and Agriculture Organization (FAO) and the World

Health Organization (WHO) as ‘live micro-organisms which when administered in

adequate amounts confer a health benefit on the host’ (14), is another type of

sustainable biocontrol for aquaculture, which is currently being developed and is

the main focus of this PhD thesis.

The term probiotics cover a large range of different beneficial mechanisms. Some

probiotics exercise their effect directly on the host, for example, by aiding

digestion or stimulating host immune system. Some probiotics work directly

against undesirable microorganisms, for example, by competing for nutrition or

space against detrimental bacteria, antagonising pathogenic bacteria by

production of antimicrobial compounds, or disruption of quorum sensing. While

other types of probiotic improve water quality by decomposition of organic

matter or reduction of ammonia or nitrite (15, 17, 192). Although the beneficial

mechanisms and the mode of action behind these mechanisms often are hinted

at, many studies on probiotics do not actually investigate the mode of action

behind the probiotics and thus it can be difficult to determine why a particular

study fails or succeeds. As part of this study we surveyed the market for probiotic

products sold for aquaculture and found two types: one that is directed at

improving water quality and one that aims at changing the microbiota of the host

gut (192). The majority of the products were based on Bacilli and lactic acid

bacteria (LAB).

The first studies on probiotic bacteria in aquaculture focused on intestinal

bacterial, such as LAB isolated from mammals as they were known to provide

health benefits in humans and domesticated land animals. In a pioneering study

Gatesoupe tested the effect of three different LAB strains on the production rate

and dietary value of rotifer (193). He found that two of the strains increased

rotifer population density and that all three strains significantly increased the

dietary value of the rotifer as measured in increased weight of turbot larvae after

20 days. Furthermore, Gatesoupe found that the strain Lactobacillus plantarum

significantly inhibited the colonisation of rotifer by A. salmonicida. Using a LAB of

34

human origin, Lactobacillus rhamnosus ATCC 53103 originally isolated from

human faeces, Nikoskelainen et al. reported significantly reduced A. salmonicida

subsp. salmonicida induced mortality in challenged rainbow trout fed probiont

infused feed (16). Several other studies exploring the use of LAB probiotics on fish

have reported increased weight gain and reduced susceptibility to infectious

diseases (194–196), although not all studies have been successful. While

investigating the probiotic effect of Lactobacillus plantarum on growth and

susceptibility of Atlantic salmon towards A. salmonicida infections, Gildberg et al.

reported higher mortalities amongst the Atlantic salmon fry feed with the LAB

infused feed than in the control during co-inhabitation experiments (Figure 11)

(197).

Figure 11 - Average cumulative mortality as a function of time after infection for fish

given Feed A (△), B (⚪) and C (⚫). Standard deviation bars are included only for final

values. The death rate of fish fed the diet supplemented with LAB (C) was significantly

higher (p < 0.0002) than the death rate of the control group (A). No significant

difference was obtained between death rates of fish fed feed supplemented with fish

protein hydrolysate (B) and the control group (A). Gildberg et al. 1995.

35

In the search for new and more efficient probiotics, several studies have taken a

broader ecological approach realising that fish are constantly surrounded by

water and that some pathogenic bacteria are transferred via water and therefore

included bacteria isolated from aquaculture environments, such as cultures of live

feed or fish larvae. Furthermore, several studies have changed focus to inhibiting

pathogens in the water of larvae and live feed rather than in the intestine of the

animals.

Some of the first non-LAB used in probiotic studies have used non-pathogenic

Vibrio spp. with antagonistic properties since these have been a dominating group

of bacteria isolated from healthy animals, including fish larvae (47, 198). Austin et

al. found that a probiotic V. alginolyticus strain could lower mortality in Atlantic

salmon challenged with A. salmonicida and to a lesser extend mortality in Atlantic

salmon challenged with V. anguillarum and V. ordalii (199). Similarly probiotic

Vibrio spp. was found to protect challenged turbot larvae against V. splendidus

due to siderophore production (20). Siderophore, which are small iron-chelating

molecules used by many microorganisms and plats to scavenge for iron in iron

depleted environments, have frequently been described as the mode of action for

bacteria with probiotic potential. Although non-pathogenic vibrios have been

investigated frequently as probiotics in scientific studies, other genera have also

been diligently studied and while members of the Vibrionaceae family dominated

the bacterial isolates from the intestine of Atlantic cod larvae, capable of

inhibiting V. anguillarum, bacteria from the genera Hyphomicrobium,

Marinomonas, Pseudoalteromonas, Roseobacter and Shewanella were also

represented (47).

García De La Banda et al. found that using either Shewanella putrefaciens Pdp11

or S. baltica Pdp13 in feed for juvenile Senegalese sole protected the fish against

photobacteriosis induced by P. damselae subsp. piscicida, while increased growth

was only seen for fish feed Pdp11 (200). S. putrefaciens Pdp11 has also been

shown to have immunostimulating effects on gilthead seabream reared at high

stocking density (201).

36

Another genus, which has been investigated for potential probiotic properties, is

Pseudomonas. Two separate studies using different Pseudomonas sp. strains

namely M174 and M162, both isolated from rainbow trout eggs, showed that the

strains were able to inhibit F. psychrophilum in vitro and decrease mortality of F.

psychrophilum challenged rainbow trout (19, 202). Mode of action for the two

strains were suggested to be immunostimulation of the rainbow trout for M162

and immunostimulation and siderophore production for M174 (19, 202).

Siderophore production was determined as the main antagonistic activity of

Paeudomonas fluorescens AH2 against V. anguillarum in co-cultures. AH2 was

further able to decrease mortality of V. anguillarum challenged rainbow trout

(203). AH2 was also shown to antagonize A. salmonicida in vitro, however, AH2

was unable to protect Atlantic salmon against A. salmonicida induced furunculosis

in co-habitant infection trails (18), illustrating the importance of in-depth studies

of potential probionts.

When considering and evaluating bacteria with antagonistic properties for use as

probiotics in aquaculture, the following should be taken into account: a “good”

probiotic bacteria should be able to antagonise a range of pathogens, it should be

able to establish itself and survive on or in the environment of the host, the

bacteria must not be harmful to the host and should not “disrupt” the

microbiome of the host and its environment as this could leave an entry for

pathogens outside the probionts range of antagonise. Furthermore, if a probiont

is to make it to a commercial product, choosing a strain without closely related

pathogenic strains will make safety documentations for the use of the strain less

complicated.

A group of bacteria, which so far seem to fulfil the requirements of good

probionts and, which have shown great potential as probiotics in aquaculture, are

members of the Roseobacter group that produce the antimicrobial compound

tropodithietic acid (TDA).

37

5 THE ROSEOBACTER GROUP

The Roseobacter group, until recently known as the Roseobacter clade (204), has

its origin with the combined reclassification of Erythrobacter sp. OCh114 (205) to

Roseobacter denitrificans and description Roseobacter litoralis in 1991 (206). The

term “Roseobacter group” as suggested by Simon et al. covers the marine

members of the Rhodobacteraceae family (Alphaproteobacteria) of which more

than 70 genera and more than 160 species have been described (204, 207). The

group is dominated by chemoorganoheterotrophic non-fermentative bacteria and

its members inhabit a vast array of different habitats, including polar, temperate

and tropical waters, deep-sea sediment, open ocean, as well as coastal waters

(207–209). Roseobacter group members represent up to 7 % of bacteria in the

ocean surface waters and 30 % in blooms of microalgae (210, 211). Besides

association with microalgae, members of the group have also been associated

with other eukaryotes, such as sea grass and macroalgae (206, 210–216), cod, sea

bass and turbot larva (27, 47, 217) and several marine invertebrates, including

corals, molluscs, marine sponges, barnacles, starfish, cuttlefish and squid (21, 215,

218–227). The large range of habitats and symbiotic relations the Roseobacter

group members are found in, is mirrored by the equally large range of

metabolisms in the group. Some members of the group are key players in the

biogeochemical cycles and are among other able to oxidate the greenhouse gas

carbon monoxide and several organic and inorganic sulphur containing

compounds, including dimethylsulfoniopropionate (DMSP) (210, 220, 228, 229).

Degradation of aromatic compounds, denitrification and aerobic anoxygenic

photosynthesis is also seen in the group (206, 230, 231). A range of secondary

metabolites is produced by the Roseobacter group members as well, including

quorum sensing molecules, shellfish toxins, siderophores, anti-microbial

compounds and algaecides (232–236). A genus within the group, which is

particularly potent when it comes to production of secondary metabolites, is the

Phaeobacter genus.

38

5.1 THE PHAEOBACTER GENUS

The Phaeobacter genus of the Roseobacter group first came into taxonomic

existence with the description of P. inhibens and the reclassification of

Roseobacter gallaeciencis as P. gallaeciencis in 2006 (237). Since then the

Phaeobacter genus has grown and shrunk with new descriptions and

reclassifications. Most notably is the reclassifications suggested by Breider et al.,

which shrunk the genus from seven to two species by reclassifying P. aquaemixtae

(238), P. daeponensis (239) and P. caeruleus (240) as Leisingera spp. and P.

arcticus (241) and P. leonis (242) as Pseudophaeobacter with P. arcticus as type

species (243). Since this reclassification, three new species have been described,

bringing the total number of species in the Phaeobacter genus to five. These

species are; P. gallaeciencis, P. inhibens, P. piscinae, P. porticola and P.

marinintestinus.

Figure 12 - Phylogenetic tree based on 16S rRNA gene sequences of the Phaeobacter

species and their closest relatives, obtained using the neighbour-joining method and

compared with the maximum-likelihood method. Sonnenschein et al. 2017.

It is important to note that the paper describing P. marinintestinus was in revision

at the same time as the Breider reclassification paper and that its name has not

been validly published (223, 243). Had P. marinintestinus been described after the

39

reclassification it would likely have been described as a Pseudophaeobacter as it

clusters with P. arcticus and P. leonis (Figure 12) and thus P. marinintestinus will

not be regarded as a Phaeobacter in this thesis, which brings the total number of

genuine Phaeobacter species in the genus to four.

P. gallaeciencis, originally isolated from a scallop (Pecten maximus) larval culture,

is the type species and has BS107T = DSM 26640T = CIP 105210T as its type strain

(21, 237). P. inhibens has the type strain T5 T = DSM 16374 T = LMG 22475 T, which

was isolated from tidal mudflat in the German Wadden Sea (237, 245, 246). The

species has continuously been isolated in several Danish harbours (247) and

includes the best studied Phaeobacter strain P. inhibens DSM17395. DSM17395

was originally thought to be identical to CIP 105210, however, Buddruhs et al.

showed that this was not the case and thus DSM17395 was reclassified from P.

gallaeciensis to P. inhibens (248). P. piscinae has 27-4 T = DSM 103509 T = LMG

29708 T as its type strain (244). 27-4 was isolated from a Spanish turbot rearing

unit in 2001 (217) and to the best of the authors knowledge all P. piscinae strains

described have been isolated from aquaculture sites (244). P. porticola with type

strain P97 T = DSM 103148 T = LMG 29594 T was isolated from a sea harbour in the

German Wadden Sea (219).

Members of the Phaeobacter genus are Gram-negative facultative anaerobic rods.

Phaeobacter spp. are easily identifiable on marine agar (Difco 2216), where the

colonies are brown to dark brown and something, which resemble a diffusible

brown pigment is produced (see section 5.1.1) (237, 243). In liquid cultures

Phaeobacter spp. form star-shaped (rosette) aggregates formed by the

attachment of individual cells via N-acetylglucosamine containing polysaccharide

(249). Phaeobacter spp. are known to produce several secondary metabolites,

including aslgaecides (the roseobacticides), quorum sensing molecules (acylated

homoserine lactones), siderophores and the antimicrobial tropodithietic acid

(TDA) (233, 250–252). TDA production is a genus trait (after the Breider

reclassification), while not all strains and species have been found to produce

Roseobacticides (Figure 13) (243, 253).

40

Phaeobacter spp. are prolific biofilm producers and most isolates originate from

biofilms on biotic or abiotic surfaces, such as macro algae, barnacles, rotifers,

turbot larval tank surfaces and harbour jetty’s (216, 219, 244, 247). Indeed Rao et

al. found that P. gallaeciensis strain 2.10, isolated from the green alga Ulva

lactuca, was able to outgrow another aggressive biofilm former

Pseudoalteromonas tunicata when grown in a mixed biofilm (216).

Figure 13 - Neighbour-joining tree of 16S rRNA gene sequences, compared with the

maximum-likelihood method, of Roseobacter group strains tested by Sonnenschein et

al. (2018) for production of roseobacticide B. Roseobacticide B producers are marked

in green, non-producer in red. TDA-producers are marked with a black circle (⚫) and

the strain carrying the disrupted TDA biosynthetic gene cluster with an empty circle

(⚪). Sonneschein et al. 2018.

41

It has been suggested that P. inhibens has a naturally occurring algal–bacterial

symbiotic relationship with the microalgae Emiliania huxlieyi. The suggested

relationship is divided into a mutualistic phase and a parasitic phase. In the

mutualistic phase E. huxleyi provides a carbon and sulphur, while P. inhibens

promotes algae growth with phenylacetic acid and protects the algae with TDA.

When E. huxleyi senesce p-coumaric acid is released and a mutualist-to-parasite

switch takes place as p-coumaric acid induce roseobacticide production in P.

inhibens, which kills the algal host (233, 254). Several P. inhibens, P. gallaeciensis

and P. piscinae strains have been shown to produce roseobacticides (Figure 14)

and have been isolated in association microalgae. Thus, it has been a concern that

Phaeobacter spp. could potentially harm live feed algae in aquaculture settings.

Sonnenschein et al. recently tested the effect of extract from roseobacticides

producing and non-roseobacticides producing Phaeobacter spp. strains on four

different microalgae; R. salina, Thalassiosira pseudonana, T. suecica and E. huxleyi

(253). The study found different effects on the different algae, as the aquaculture

“workhorse” T. suecica was unaffected while the extract had negative effects with

varying intensity on R. salina, T. pseudonana and E. huxleyi. The authors note that

the quantities of the roseobacticides in the study are likely to be much higher

than in situ as roseobacticides cannot be detected in situ with current techniques.

In this study we have co-cultured Phaeobacter spp. DSM17395 and S26 with R.

salina and T. suecica and only T. suecica, respectively. We did not observe any

negative effect on the growth of T. suecica by the presence of either DSM17395

or S26 (187, 255), however, unlike Sonnenschein et al. we didn’t observe any

negative effect on the growth of R. salina when cultured with DSM17395 either

(78). This finding supports the suggestion that roseobacticide concentration in situ

is very low.

Several studies have demonstrated (using Tn-mutants) that TDA is a key

component in the ability of Phaeobacter spp. to antagonise other bacteria, such

as fish pathogens and to protect challenged live feed and fish larvae against

vibriosis (see section 5.1.2) (26, 27, 256–258). As an example D’Alvise et al.

compared the antagonistic abilities of the wild type and a TDA-negative mutant of

42

P. inhibens DSM17395 (then called P. gallaeciensis DSM17395) against V.

anguillarum HI610 in among other axenic cultures of T. suecica (26). They found

that V. anguillarum grow to significant higher numbers in the presence of the

TDA-negative mutant compared to the wild type (Figure 14).

Figure 14 - Reduction of V. anguillarum in cultures of axenic T. suecica by P. inhibens.

Colony-forming units of V. anguillarum inoculated at (A) 101 cfu/ml and at (B) 104

cfu/ml in presence of P. inhibens wild type (⬛), in presence of the P. gallaeciensis

TDA-negative mutant (▲) and in the control (▼). D’alvise et al. 2012.

5.1.1 TROPODITHIETIC ACID (TDA)

The broad spectrum antimicrobial compound tropodithietic acid (TDA) is a

disulfide tropone that was first chemically described in 2003 (259). TDA has a

tautomer thiotropocin (Figure 15), an antimicrobial compound produced by a

Pseudomonas sp. strain CB-104 isolated from a Japanese soil sample in 1984 (260,

261). Thiotropocin has since been described in a Caulobacter sp. strain KP654

isolated from marine microalgae (262, 263), however, its classification was

questioned by Porsby who suggested that KP654 might be a Phaeobacter sp. or a

Ruegeria sp. based on the description of KP654 (264). That TDA and thiotropocin

43

are tantomers was first suggested by Liang and later backed up by Greer et al. in a

computation study on thiotropocin, TDA and troposulfenin (259, 265). Production

of TDA have besides the Phaeobacter genus been described in other genera of the

Roseobacter group, more precisely in Ruegeria, Silicibacter and Pseudovibrio (21,

237, 259, 266–269).

Figure 15 - Tropodithietic acid (left) and thiotropocin (right). Modified from Thiel et al.

2010.

The bioactivity of TDA has been explored by Porsby et al. who found it to be

bactericidal, killing both growing and non-growing cells (23). TDA is active against

both Gram-negative and Gram-positive bacteria, including several human and fish

pathogens, such as Salmonella enterica, Pseudomonas aeruginosa, Staphylococcus

aureus, Escherichia coli, V. anguillarum, V. vulnificus and V. parahaemolyticus (23,

258, 270). TDA tolerance has been observed in microbiomes containing TDA-

producing Pseudovibrio sp., where 126 out of 136 (93 %) non-TDA producing

bacteria tested showed TDA tolerance, however, the mechanisms behind are

unknown (271). As with all other antimicrobial compounds, concerns regarding

development of TDA resistance have been raised. To date two studies have

addressed this issue and both concluded that TDA resistance or tolerance is

difficult to induce in in vitro experiments (23, 270). Porsby et al. exposed several

non-TDA producing bacteria, such as E. coli, P. aeruginosa and S. aureus to sub-

44

MIC and MIC concentrations of TDA both in liquid and on solid medium, however,

nothing more than a doubling in MIC was seen and this disappeared after growth

on media without TDA (23). Furthermore, Salmonella enterica serovar

Typhimurium was exposed to gradually increasing concentrations of TDA extract

in an adaptive evolution experiment. However, after 40 transfers MIC was only

doubled wild-type MIC, which disappeared after growth on media without TDA

(23). Likewise, no TDA resistance or tolerance were seen after V. anguillarum

when exposed to gradually increasing concentrations of pure TDA in up to 280

generations of adaptive evolution experiment (Figure 16) (270). Adaptive

evolution experiments have previously been used to successfully provoke

resistance development towards a range of other antimicrobial compounds, such

as colistin (272–274). TDA producers have themselves increased tolerance

towards TDA, which might be due to the TDA ”resistance” genes tdaR1-tdaR3,

which if cloned into E. coli decrease their TDA sensitivity (275). The fact that TDA

resistance or tolerance do not develop easily supports the use of TDA producing

Phaeobacter spp. as probiotics in aquaculture as it seem unlikely that the

pathogens will become resistance towards the probiotic and render it useless as

has been observed for some antibiotics.

45

Figure 16 - Attempt to select for TDA-tolerant or TDA-resistant V. anguillarum strains

in an adaptive laboratory evolution experiment. Solid lines, individual lineages, of

which 3 of 7 reached 1.5 times the wild-type MIC (18.75 µg/ml) and the rest reached

1.75 times the wild-type MIC (21.9 µg/ml). Dashed line, wild-type MIC (12.5 µg/ml

TDA). Rasmussen et al. 2016.

A mode of action (MoA) for TDA was purposed by Wilson et al. based on

cytological profiling, such as comparing cell cytology of E. coli treated with TDA

and compounds with known MoAs (275). It was suggested that TDA functions as

an electroneutral proton antiporter targeting the proton motive force (PMF) and

disrupting transmembrane proton gradient. If the purposed MoA is accurate it

would to some degree explain the lack of resistance and tolerance as the PMF is

essential and highly conserved and thus mutational changes in the PMF is likely to

kill the bacteria.

TDA has been shown to have dual concentration depending functions; as an

antimicrobial compound at high concentrations and as a quorum sensing (QS)

molecule at low concentrations (100-fold lower than needed for bacterial

inhibition). As a QS molecule TDA was shown to have a significantly impacts on

46

around 10 % of P. inhibens total genes and it regulates several phenotypes,

including motility, biofilm formation and antibiotic production all important for

the Phaeobacter spp. host and surface associated lifestyle (276). The dual function

of TDA and its potential importance for a host and surface associated lifestyle

might explain why TDA producers keep their TDA producing abilities even though

TDA production causes decreased fitness by limiting growth rate and biomass

production (277).

Nervous system cells N2a from mice and OLN-93 from rats have been shown to be

sensitive to TDA, which activated stress response resulting in cell death (278). TDA

is also toxic to several cancer cell lines (275). In contrast when tested in the

eukaryotic models Caenorhabditis elegance and Artemia TDA showed very low

toxicity (279), which support the use of TDA based probiotics in live feed.

The biosynthesis of TDA have to some extent been described in the closely related

strains P. inhibens DSM17395, P. piscinae 27-4 and Ruegeria sp. TM1040

(originally Silicibacter sp. TM1040) (251, 267, 280–284). DSM17395 contains three

plasmids of which the largest pPGA1_262 harbours the TDA biosynthetic gene

cluster (tdaA-tdaF). The gene cluster is found in all TDA producing bacteria and is

crucial in the later steps of the TDA biosynthesis (267, 271, 283, 284). In

DSM17395 several genes involved in the catabolism of phenylacetic acid and in

the primary sulfur metabolism are also thought to be important for TDA

production, including the aromatic ring oxidation complex PaaABCDE (267, 283,

284). Regulation of TDA production varies among TDA producers as QS mediated

by N-(C10) acyl homoserine lactone and TDA (as an autoinducer) regulates TDA

production in DSM17395. TDA also functions as an autoinducer in TM1040 while

neither QS nor TDA seem to regulate TDA production in Pseudovibrio sp. W74

(251, 271, 282)

Despite the existing knowledge about the compound, the role of TDA in the

environment is still poorly understood. TDA production in iron rich cultures

coincide with a brown substance. The brown substance, which is not produced by

TDA-negative mutants, have via chemical analysis been shown to be a complex or

47

polymer of TDA and Fe3+ [FeIII(TDA)2]x (Figure 17). The complex is thought to be

the diffusible brown “pigment” produced by Phaeobacter spp. on marine agar as

it also coincides with TDA production and as with the brown substance is not

produced by TDA deficient mutants (285, 286). The iron binding ability of TDA

resembles that of a siderophore, however, unlike classical siderophores, which are

produced and used for iron scavenging in iron-limited conditions, the presence of

iron, such as ferric citrate is a prerequisite for production of active TDA in P.

inhibens. If iron is limited an analogue of TDA (“pre-TDA”), which is non-inhibitory,

is produced. This has led to the speculation that TDA could potentially function as

iron reservoir utilised by the producer or symbiotic algae (285).

Figure 17 - Direct MS analysis (Bruker QTOF) of the suspended dark brown precipitate

of TDA and Fe3+, revealing complexes formed from one iron(III) ion and two or three

TDA molecules. Modified from D’Alvise et al. (2016).

5.1.2 PHAEOBACTER SPP. AS PROBIOTICS

Several experiments exploring the probiotic properties of Phaeobacter spp. have

been conducted during the last two decades (24–28, 217, 287). In fact all the P.

piscinae strains 27-4 , 8-1, S26 and M6-4.2 used in the description of the species

were isolated from aquacultures in an attempt to isolate bacteria, which could be

used as probiotics in aquaculture (27, 217, 244, 288). Phaeobacter spp. can inhibit

V. anguillarum in live feed and protect turbot and cod larvae challenged with V.

anguillarum (24–28). Planas et al. (2006) tested the effect of feeding V.

anguillarum challenged turbot larvae with P. piscinae 27-4 enriched rotifers and

found that the accumulated mortality decreased significantly from 80–90 % to

60–70 % after 10 days equal to that of the non-challenged controls. The choice of

48

using enriched rotifers was in accordance with the knowledge of that time namely

that probiotics needed to work in the intestines of the host animals it was

protecting. Phaeobacter spp. are strict aerobic, which, combined with the notion

that probiotic can exert their effect in the water, led D’Alvise et al. to choose a

different approach six years later. By adding P. inhibens DSM17395 to the water of

live feed and cod larvae D’Alvise et al. showed that P. inhibens was able to reduce

V. anguillarum in cultures of the microalgae T. suecica and Nannochloropsis

oculata, as well as in rotifer (Brachionus plicatilis) cultures (26). P. inhibens was

also able to protect cod larvae from V. anguillarum induced vibriosis when

inoculated directly into the water of the cultures. Furthermore, using a TDA-

negative DSM17395 mutant they showed that TDA production was important for

DSM17395 ability to protect cod larvae against vibriosis. DSM17395 wild type

reduced mortality of the V. anguillarum challenged cod larvae from 100 % to 12.5

% ± 2.0 % after 10 days while the TDA-negative mutant only reduced mortality to

68.8 % ± 30.4 % after 10 days (Figure 18). Later it was shown that the potency of

the probiotic properties of Phaeobacter spp. in V. anguillarum challenged cod

larvae correlated with their capability of producing TDA as tested in V.

anguillarum inhibition assays (25). D’Alvise et al. also showed that timing of the

addition of Phaeobacter spp. was not trivial and that the most efficient treatment

was seen when Phaeobacter spp. were introduced before the pathogen (25).

49

Figure 18 - Mortality of cod larvae during challenge trials. Mean values of two

independent triplicate experiments. The single-larvae cultures were simultaneously

inoculated with P. gallaeciensis wild type and V. anguillarum (T5, •), or with the TDA-

negative mutant of P. gallaeciensis and V. anguillarum (T6, □). Unexposed larvae and

larvae exposed to single bacterial strains acted as controls: Negative Control (T1, ▪),

only V. anguillarum (T2, ▴), only P. gallaeciensis wild type (T3, ▾) and only P.

gallaeciensis TDA-negative mutant (T4, ♦). D’Alvise et al. (2012).

Phaeobacter spp. protect Artemia against V. anguillarum and V. harveyi an

opportunistic pathogen of marine vertebrates and invertebrates (27). In the

present study we added the copepod species A. tonsa to the list of live feed

cultures, where Phaeobacter spp. are able to inhibit V. anguillarum (78). This is

especially important since A. tonsa was colonised but not killed by V. anguillarum.

Thus, infected A. tonsa could function as an efficient infection vector if fed to fish

larvae especially as larvae are more likely to feed on living than dead live feed,

which will sink uneaten to the bottom of the tank.

50

Most studies on the probiotic properties of Phaeobacter spp. have been

conducted in axenic cultures, i.e., without culturable background microbiota (25–

28, 217, 287). However, a tank with live feed or fish larvae is far from axenic. In

the contrary, a tank with live feed is loaded with organic matter supporting high

numbers of bacteria, such as more than 3.5 x 109 CFU/g rotifer wet weight and

1010 CFU/g Artemia wet weight for enriched rotifer and Artemia, respectively (60).

Thus, the logical step was to test Phaeobacter spp. in non-axenic systems.

Grotkjær et al. was the first study to test the probiotic properties of P. inhibens in

non-axenic live feed cultures (24). The study found that not only was P. inhibens

able to establish itself in the systems but a strong probiotic effect was also seen in

the non-axenic cultures of microalgae and Artemia. Likewise we tested the

abilities of P. inhibens in non-axenic cultures in the present study both in the

“copepod” study and in the “Phaeobacter phage combo” study (78, 187). Here we

also showed that P. inhibens is able to inhibit V. anguillarum in microalgae

cultures (R. salina and T. suecica) and zooplankton cultures (A. tonsa and Artemia)

furthermore, it protected Vibrio-challenged, as well as non-challenged Artemia.

The inhibition of V. anguillarum in the non-axenic systems of Grotkjær et al. and

in the copepod study in this PhD were less efficient (no growth) than what have

previously been observed in axenic systems (reduction in V. anguillarum counts),

which led us to speculate that the microbiota of the non-axenic systems were in

some manner protecting V. anguillarum against P. inhibens (24, 78). In order to

obtain an enhanced effect against V. anguillarum in non-axenic systems, we

combined a Phaeobacter strain with phage therapy using the broad range vibrio

phage KVP40. Although the results demonstrate the potential of P. inhibens, as

well as KVP40 on their own in non-axenic live feed systems, the combination of P.

inhibens and KVP40 only resulted in a marginal improved pathogen reduction and

protection of Vibrio-challenged Artemia (187). However, there is still potential in

the combination of Phaeobacter based probiotics and phage therapy and the

subject should be further studied. As reductions in V. anguillarum counts in the

combination study were seen in both treatments with only Phaeobacter and in

combination with KVP40 it seems that V. anguillarum is not necessarily better

51

protected against Phaeobacter in non-axenic than in axenic systems as first

speculated.

To date, most studies concerning probiotic Phaeobacter spp. have used V.

anguillarum as “model” pathogen, however, Phaeobacter spp. have been shown

to inhibit a large range of aquaculture relevant pathogens from both within the

Vibrionaceae family and others. The pathogens include V. harveyi, V. alginolyticus,

V. parahaemolyticus, V. vulnificus, V. splendidus, V. logei, V. proteolyticus, V.

tapetis, Photobacterium damselae ssp. piscicida, Moritella viscosa, Tenacibaculum

maritimum, Lactococcus piscium, Acinetobacter spp., Aeromonas spp.,

Pseudomonas spp., Micrococcus spp. and others (21, 22, 288).

Due to their broad spectrum of inhibition, concerns have been raised regarding

the potential consequences application of Phaeobacter spp. could have on the

natural host microbiota of the organisms in the treated systems and whether

Phaeobacter spp. potentially would provoke an imbalance in the community.

Dittmann et al. recently investigated the effect of P. inhibens on the diversity and

composition of the host microbiota associated with the microalga Emiliania

huxleyi and European flat oyster, using amplicon sequencing of the V4 region of

16s rRNA gene (289). Here they found that P. inhibens changed the structure of

the E. huxleyi microbiome substantially, while the structure of the oyster

microbiome only saw minor changes. Thus, indicating that the effect of

Phaeobacter spp. on microbiomes depends on the structure, diversity and

complexity of the individual microbiome.

Easy application is important for the use of Phaeobacter spp. to become

widespread in commercial aquaculture as probiotics. Introduction of the probiotic

Phaeobacter spp. before the pathogen have been shown to be the most efficient

application (25). Since microalgae are the first step in the live feed food chain and

the first potential point of entry for pathogenic bacteria upscaling of Phaeobacter

spp. in Tetraselmis, algae cultures were explored in this study (Figure 19). We

found that both P. inhibens DSM17395 and P. piscinae S26 would establish

themselves in the algae cultures and upscale along with the algae (255). However,

when conditions were changed at time point four weeks from growth in flasks

52

with big surface area without aeration to small surface area with aeration the

Phaeobacter strains were unable to progress at the same rate as the rest of the

microbiota. The mechanisms behind these findings should be further explored

and different length of growth period along with inoculation ratios should be

tested.

Figure 19 – Growth of (A) microbiota and (B) T. suecica during upscaling of non-axenic

cultures. (A) CFU counts of total microbiota in T. suecica control cultures (⚫, dashed

line); CFU counts of P. piscinae S26 (⬛, solid line) and total microbiota (▲, dashed

line) in T. suecica cultures with S26: CFU counts of DSM17395 (▼, solid line) and total

microbiota (◆, dashed line) in T. suecica cultures with DSM17395. (B) Algae counts of

T. suecica in control cultures (⚫); S26 added cultures (⬛); DSM17395 added cultures

(▲). Every two weeks the cultures would be inoculated in to a 10 x larger volume of

fresh medium, hence the 1 log drop in bacterial and algae counts at time points 2 and

4. Error bars show the standard deviation of three biological replicates. Rasmussen et

al. 2019.

In order to test if the Phaeobacter spp. retained their probiotic abilities after

upscaling, the Phaeobacter infused algae were inoculated in to Artemia cultures

at approximately 105 algae mL-1 and half of the Artemia were challenged with V.

anguillarum (Strain 90-11-286, O1). No differences in Artemia mortality were seen

between cultures whether or not they were challenged with V. anguillarum or fed

T. suecica with or without Phaeobacter. However, V. anguillarum growth was

53

inhibited in P. piscinae S26 containing cultures with significant lower counts at

time points 3 and 4 days (Figure 20) indicating that S26 had kept its ability to

inhibit V. anguillarum even when inoculated at levels less than two log higher

than Vibrio. Grotkjær et al. tested TDA concentration after 72 h static growth in

marine broth at 25 °C for a range of Phaeobacter spp., including S26 and

DSM17395 and found that S26 had produced 0.92 ± 0.07 µM TDA while

DSM17395 had produced 0.54 ± 0.15 µM TDA, thus the fact that S26 was able to

inhibit V. anguillarum and DSM17395 was not might result from s26 being a more

productive TDA producer (27). In order to further explore Phaeobacter spp.'s

effect on microbiomes, DNA was extracted during the upscaling experiment and

will be subjected to amplicon sequencing for further analysis.

Figure 20 - Vibrio anguillarum counts in non-axenic systems of Artemia feed

Tetraselmis suecica. V. anguillarum alone (●), V. anguillarum with P. piscinae S26 (■),

V. anguillarum with P. inhibens DSM17395 (▲). Error bars show the standard

deviation of three biological replicates. Rasmussen et al. 2019.

In line with the potential commercialisation of Phaeobacter probiotics we

explored the current market for microbe-based products for aquaculture in the

editorial. Many additional steps are needed to be overcome in order for

Phaeobacter spp. to go from a success in scientific publications to a finished

product, which the customers can use to seed their algae cultures, these include

assessments of safety, production procedures, packaging and storage (192).

54

6 CONCLUDING REMARKS AND FURTHER PERSPECTIVES

Aquaculture is the fastest growing protein production sector in the world. With

the rapid growth, several problems and challenges follow, including how to secure

sufficient feed and high quality fish juveniles. These challenges must be met if the

industry is to keep its growth and be part of the solution for the United Nations

Sustainable Development Goals for a sustainable future. The challenge to secure

sufficient quantities of high quality juveniles is currently relying on the use of

antibiotics. The use of antibiotics to treat infections of vulnerable fish larvae

needs to be at a minimum and sustainable alternatives of which probiotics have

great potential must be utilised until the fish reach a maturity, where they can be

vaccinated.

The use of TDA producing Roseobacter spp. as probiotics in aquaculture has been

studied for two decades and this PhD explored several aspects contributing to

their potential use. Previous studies have shown that Phaeobacter spp. are able to

inhibit the growth of a range of pathogens, including Vibrio spp. due to the

production of tropodithietic acid (TDA). The inhibition of Vibrio spp. has

successfully been tested in axenic, as well as some non-axenic live feed cultures

and it has successfully protected Artemia, as well as fish larvae of cod and turbot

against vibriosis. Previous studies have mostly revolved around Phaeobacter spp.

probiotic abilities in axenic systems. The present PhD study has shown that

Phaeobacter spp. are also able to antagonise pathogenic vibrios in a range of non-

axenic systems of microalgae and zooplankton used as live feed in commercial fish

larvae cultures along with protecting Artemia against vibriosis. This is important

since previous studies have shown that probiotic functioning in one host–

pathogen combination might not necessarily function in other combinations. The

composition of microbiota in the non-axenic systems seems to influence the

ability of Phaeobacter spp. to inhibit pathogens. As this is a potential constraint on

the usability of Phaeobacter based probiotics, it should be tested whether this

reoccurs in special types of live feed or represents random incidences.

55

As part of the ProAqua project parts of this PhD was carried out together with

researchers from aquaculture and in the area of phage therapy. We tested the

broad host-range bacteriophage KVP40's abilities to antagonise four different V.

anguillarum strains and protect fish larvae of turbot and cod against vibriosis.

KVP40 was able to decrease or delay mortality in turbot and cod larvae challenged

with four different V. anguillarum strains. Due to growth of other pathogenic

bacteria, properly, including Vibrio spp. based in counts on selective TCBS plates

high mortality was seen in the non-challenged controls. However, KVP40 was able

to reduce mortality in these non-challenged systems compared to non-challenged

systems not treated with KVP40. In the combination study of this PhD we found

that KVP40 was able to inhibit V. anguillarum in live feed cultures of the

microalgae T. suecica and zooplankton Artemia and although regrowth of V.

anguillarum was seen in the Artemia cultures, V. anguillarum did not grow to the

same level as in the vibrio control. As in the fish larvae cultures high mortality was

observed in the non-challenge Artemia controls probably due to among other

Vibrio spp. and as with the fish larvae KVP40 were able to reduce mortality in

these non-challenged cultures. Thus, although the use of phage therapy require

optimising and further research is needed, it is a promising sustainable method of

biocontrol for aquaculture.

This study combined the probiotic P. inhibens with the phage KVP40 and tested it

in live feed microalgae T. suecica and zooplankton Artemia cultures along with fish

cell lines and evaluated the results against the individual treatments, as we

hypothesised that the combination would result in enhanced inhibition of V.

anguillarum and protection of Artemia and fish cell lines. The combination of P.

inhibens and KVP40 resulted in faster inhibition of V. anguillarum compared with

P. inhibens on its own and lower V. anguillarum counts compared to KVP40 on its

own in live feed systems. A marginal improvement was observed in the protection

of Artemia against vibriosis compared to P. inhibens only treated, while no

differences were seen between treatments in the cell line assay. Although the

effect of combining P. inhibens with KVP40 was not as pronounced as anticipated

for the aspects of combining Phaeobacter spp. with phage therapy or other types

of biocontrol should be further studied. It would be interesting to expand the

56

Phaeobacter phage combination study to fish larvae as the longer experiment

period might reveal larger differences in larvae survival between Phaeobacter

only treatments and Phaeobacter phage combination treatments. Likewise, it

would be relevant to combine Phaeobacter spp. with other types of biocontrol,

such as microbial matured water as this might increase fish larvae survival further.

This thesis has also investigated how Phaeobacter spp. are to be used in practise

in aquaculture. This is important as the ease of which Phaeobacter spp. can be

applied will impact their success in commercial aquaculture. We were to some

extend able to upscale Phaeobacter spp. in the microalgae T. suecica, however, as

the Phaeobacter spp. did not grow well in the last aerated cultures further

experiments tackling different dilution ratios and growth periods should be

conducted in order to optimise this process. In order to test if the upscaled

Phaeobacter spp. retained their probiotic abilities, Artemia were fed the

Phaeobacter infused algae and challenged with V. anguillarum. Only P. piscinae

S26 was able to inhibit V. anguillarum and no protection of Artemia was observed

by any of the Phaeobacter spp., however, we ascribe this to the low levels of

probiont in the cultures, not the loss of inhibitory ability. Once the upscaling

process has been optimised Phaeobacter infused microalgae should be further

tested by feeding to both live feed and fish larvae. In addition, mortality of

pathogen challenged live feed and larvae fed the upscaled Phaeobacter infused

algae, amplicon sequencing should be used to monitor potential effects of the

Phaeobacter spp. on the microbiome of the treated cultures.

Lastly, pilot or full scale trials of Phaeobacter based probiotics are needed in order

to move closer to a commercial Phaeobacter probiotic. These pilot or full scale

trials should preferably use Phaeobacter upscaled in microalgae and be conducted

in both live feed and larvae cultures.

57

7 ACKNOWLEDGEMENTS

First of all I would like to thank my supervisor Professor Lone Gram for her great

supervision and insightful discussions during my PhD but also during my master

and bachelor studies. Your passion and enthusiasm for scientific research showed

me that science is fun and worth fighting for, even when you have to redo

experiments the third time.

Thank you to my co-supervisor Assistant Professor Mikkel Bentzon-Tilia for your

infectious good mood, guidance and inspiring discussions over countless cops of

coffee. It has all contributed in making my PhD so much more fun and rewarding.

Thank you to all of the collaborators from the ProAqua project. I really feel that I

have been part of something greater. Special thanks to Panos G. Kalatzis for

excellent work on the Phaeobacter phage combi paper and fun during long nights

in the lab. Nanna Rørbo thank you for the great work on the phage larvae trail

paper and to Professor Mathias Middelboe for the collaboration and supervision

during my time at Elsinore.

To past and present members of the Bacterial Ecophysiology and Biotechnology

Group it has always been fun and pleasant for me to be in the group. Special

thanks to Jette Melchiorsen. Without you the lab would fall. To Karen Kiesbye

Dittmann, thank you for supporting me throughout my PhD and for being a good

friend. Thank you to Eva Sonnenschein, you have always been ready with

guidance or consolation for a sleep deprived parent.

To my family and friends thank you for your great support. Special thanks to my

wife Susanne and son Louis for your support and love, you make bad days better

and good days great. Even Phaeobacter inhibens cannot compare to you.

58

8 REFERENCES

1. United Nations, Department of Economic and Social Affairs, Population Division. 2015. World Porpulation Prospects: The 2015 Revision, Key Findings and Advance TablesWorking Paper No. ESA/P/WP.241.

2. FAO. 2018. The State of World Fisheries and Aquaculture 2018. Rome, Italy.

3. United Nations. 2019. About the Sustainable Development Goals.

4. Kibenge FSB, Godoy MG, Fast M, Workenhe S, Kibenge MJT. 2012. Countermeasures against viral diseases of farmed fish. Antiviral Res 95:257–281.

5. Hansen GH, Olafsen JA. 1999. Bacterial Interactions in Early Life Stages of Marine Cold Water Fish. Microb Ecol 38:1–26.

6. Olafsen JA. 2001. Interactions between fish larvae and bacteria in marine aquaculture. Aquaculture 200:223–247.

7. Shields RJ. 2001. Larviculture of marine finfish in Europe. Aquaculture 200:55–88.

8. Defoirdt T, Sorgeloos P, Bossier P. 2011. Alternatives to antibiotics for the control of bacterial disease in aquaculture. Curr Opin Microbiol 14:251–8.

9. Ringø E, Olsen RE, Jensen I, Romero J, Lauzon HL. 2014. Application of vaccines and dietary supplements in aquaculture: possibilities and challenges. Rev Fish Biol Fish 1005–1032.

10. Sommerset I, Krossøy B, Biering E, Frost P. 2005. Vaccines for fish in aquaculture. Expert Rev Vaccines 4:89–101.

11. Brudeseth BE, Wiulsrød R, Fredriksen BN, Lindmo K, Løkling KE, Bordevik M, Steine N, Klevan A, Gravningen K. 2013. Status and future perspectives of vaccines for industrialised fin-fish farming. Fish Shellfish Immunol 35:1759–1768.

12. Breitbart M. 2012. Marine Viruses: Truth or Dare. Ann Rev Mar Sci 4:425–448.

59

13. Oliveira J, Castilho F, Cunha A, Pereira MJ. 2012. Bacteriophage therapy as a bacterial control strategy in aquaculture. Aquac Int 20:879–910.

14. FAO, WHO. 2001. Report of a joint FAO/WHO expert consultation on evaluation of health and nutritional properties of probiotics in food including powder milk with live lactic acid bacteria.

15. Gatesoupe FJ. 1999. The use of probiotics in aquaculture. Aquaculture 180:147–165.

16. Nikoskelainen S, Ouwehand A, Salminen S, Bylund G. 2001. Protection of rainbow trout (Oncorhynchus mykiss) from furunculosis by Lactobacillus rhamnosus. Aquaculture 198:229–236.

17. Balcázar JL, Blas I De, Ruiz-Zarzuela I, Cunningham D, Vendrell D, Múzquiz JL. 2006. The role of probiotics in aquaculture. Vet Microbiol 114:173–186.

18. Gram L, Løvold T, Nielsen J, Melchiorsen J, Spanggaard B. 2001. In vitro antagonism of the probiont Pseudomonas fluorescens strain AH2 against Aeromonas salmonicida does not confer protection of salmon against furunculosis. Aquaculture 199:1–11.

19. Korkea-aho TL, Heikkinen J, Thompson KD, von Wright A, Austin B. 2011. Pseudomonas sp. M174 inhibits the fish pathogen Flavobacterium psychrophilum. J Appl Microbiol 111:266–277.

20. Gatesoupe F. 1997. Siderophore production and probiotic effect of Vibrio sp. associated with turbot larvae, Scophthalmus maximus. Aquat Living Resour 239–246.

21. Ruiz-Ponte C, Cilia V, Lambert C, Nicolas JL. 1998. Roseobacter gallaeciensis sp. nov., a new marine bacterium isolated from rearings and collectors of the scallop Pecten maximus. Int J Syst Bacteriol 48:537–542.

22. Prado S, Montes J, Romalde JL, Barja JL. 2009. Inhibitory activity of Phaeobacter strains against aquaculture pathogenic bacteria. Int Microbiol 12:107–114.

23. Porsby CH, Webber M a, Nielsen KF, Piddock LJ V, Gram L. 2011. Resistance and tolerance to tropodithietic acid, an antimicrobial in aquaculture, is hard to select. Antimicrob Agents Chemother 55:1332–7.

24. Grotkjær T, Bentzon-Tilia M, D’Alvise P, Dierckens K, Bossier P, Gram L. 2016. Phaeobacter inhibens as probiotic bacteria in non-axenic Artemia

60

and algae cultures. Aquaculture 462:64–69.

25. D’Alvise PW, Lillebø S, Wergeland HI, Gram L, Bergh Ø. 2013. Protection of cod larvae from vibriosis by Phaeobacter spp.: A comparison of strains and introduction times. Aquaculture 384–387:82–86.

26. D’Alvise PW, Lillebø S, Prol-Garcia MJ, Wergeland HI, Nielsen KF, Bergh Ø, Gram L. 2012. Phaeobacter gallaeciensis reduces Vibrio anguillarum in cultures of microalgae and rotifers, and prevents vibriosis in cod larvae. PLoS One 7:e43996.

27. Grotkjær T, Bentzon-Tilia M, D’Alvise P, Dourala N, Nielsen KF, Gram L. 2016. Isolation of TDA-producing Phaeobacter strains from sea bass larval rearing units and their probiotic effect against pathogenic Vibrio spp. in Artemia cultures. Syst Appl Microbiol 39:180–188.

28. Planas M, Pérez-Lorenzo M, Hjelm M, Gram L, Uglenes Fiksdal I, Bergh Ø, Pintado J. 2006. Probiotic effect in vivo of Roseobacter strain 27-4 against Vibrio (Listonella) anguillarum infections in turbot (Scophthalmus maximus L.) larvae. Aquaculture 255:323–333.

29. Prol Garcia MJ, Gómez M, Sánchez L, Pintado J. 2014. Phaeobacter grown in biofilters: a new strategy for the control of Vibrionaceae in aquaculture. Aquac Res 45:1012–1025.

30. FEAP. 2017. European Aquaculture Production Report 2008-2016aquamedia.org.

31. Walker T. 2019. Cod farming re-emerging in Norway with national program. hatcheryinternational.com.

32. The Fish Site. 2019. Is Norway set for a cod farming renaissance? Fish Site.

33. Lekang OI. 2007. Aquaculture Engineering. Blackwell Publsihing, Oxford.

34. Blancheton JP. 2000. Developments in recirculation systems for Mediterranean fish species. Aquac Eng 22:17–31.

35. Buschmann AH, Tomova A, López A, Maldonado MA, Henríquez LA, Ivanova L, Moy F, Godfrey HP, Cabello FC. 2012. Salmon Aquaculture and Antimicrobial Resistance in the Marine Environment. PLoS One 7:e42724.

36. Cole DW, Cole R, Gaydos SJ, Gray J, Hyland G, Jacques ML, Powell-dunford N, Sawhney C, Au WW. 2009. Aquaculture: Environmental,

61

toxicological, and health issues David. Int J Hyg Environ Heal 212:369–377.

37. Troell M, Joyce A, Chopin T, Neori A, Buschmann AH, Fang J. 2009. Ecological engineering in aquaculture — Potential for integrated multi-trophic aquaculture (IMTA) in marine offshore systems. Aquaculture 297:1–9.

38. Thiel M, Macaya E, Acuna E, Arntz W, Bastias H, Brokord K, Camus P, Castilla J, Castro L, Cortez M, Dumont C, Escribano R, Fernandez M, Garjardo J, Gaymer C, Gomez I, Gonzalez A, Gonzalez H, Haye P, Illanes J, Iriarte J, Lancelotti D, Luna‐Jorquera G, Luxoro C, Manriquez H, Marin V, Munoz P, Navarretes S, Perez E, Poulin E, Sellanes J, Sepulveda H, Stotz W, Tala F, Thomas A, Vargas C, Vasquez J, Vega J. 2012. The Humboldt Current System of Northern and Central Chile: oceanographic processes, ecological interactions and socioeconomic feedback. Oceanogr Mar Biol An Annu Rev Aberdeen Univ Press.

39. Oliva-Teles A, Enes P, Peres H. 2015. 8 - Replacing fishmeal and fish oil in industrial aquafeeds for carnivorous fishFeed and Feeding Practices in Aquaculture. Elsevier Ltd.

40. Moretti A, Pedini Fernandez-Criado M, Cittolin G, Guidastri R. 1999. Manual on Hatchery Production of Seabass and Gilthead Seabream - Volume 1. FAO, Rome, Italy.

41. Turner JT. 2004. The Importance of Small Pelagic Planktonic Copepods and Their Role in Pelagic Marine Food Webs. Zool Stud 43:255–266.

42. Tocher DR, Bendiksen EÅ, Campbell PJ, Bell JG. 2008. The role of phospholipids in nutrition and metabolism of teleost fish. Aquaculture 280:21–34.

43. Verner-Jeffreys DW, Shields RJ, Bricknell IR, Birkbeck TH. 2003. Changes in the gut-associated microflora during the development of Atlantic halibut (Hippoglossus hippoglossus L.) larvae in three British hatcheries. Aquaculture 219:21–42.

44. Muroga K, Higashi M, Keitoku H. 1987. The isolation of intestinal microflora of farmed red seabream (Pagrus major) and black seabream (Acanthopagrus schlegeli) at larval and juvenile stages. Aquaculture 65:79–88.

62

45. Reid HI, Treasurer JW, Adam B, Birkbeck TH. 2009. Analysis of bacterial populations in the gut of developing cod larvae and identification of Vibrio logei, Vibrio anguillarum and Vibrio splendidus as pathogens of cod larvae. Aquaculture 288:36–43.

46. Blanch AR, Alsina M, Simón M, Jofre J. 1997. Determination of bacteria associated with reared turbot (Scophthalmus maximus) larvae. J Appl Microbiol 82:729–734.

47. Fjellheim AJ, Playfoot KJ, Skjermo J, Vadstein O. 2007. Vibrionaceae dominates the microflora antagonistic towards Listonella anguillarum in the intestine of cultured Atlantic cod (Gadus morhua L.) larvae. Aquaculture 269:98–106.

48. Keskin M, Keskin M, Rosenthal H. 1994. Pathways of bacterial contamination during egg incubation and larval rearing of turbot, Scophthalmus maximus. J Appl Ichthyol 10:1–9.

49. Munro PD, Barbour A, Birkbeck TH. 1994. Comparison of the gut bacterial flora of start-feeding larval turbot reared under different conditions. J Appl Bacteriol 77:560–566.

50. Huys L, Dhert P, Robles R, Ollevier F, Sorgeloos P, Swings J. 2001. Search for beneficial bacterial strains for turbot (Scophthalmus maximus L.) larviculture. Aquaculture 193:25–37.

51. Nicolas JL, Robic E, Ansquer D. 1989. Bacterial flora associated with a trophic chain consisting of microalgae, rotifers and turbot larvae: Influence of bacteria on larval survival. Aquaculture 83:237–248.

52. Hemaiswarya S, Raja R, Kumar RR, Ganesan V, Anbazhagan C. 2011. Microalgae: A sustainable feed source for aquaculture. World J Microbiol Biotechnol 27:1737–1746.

53. Brown MR, Blackburn SI. 2013. Live microalgae as feeds in aquaculture hatcheries Advances in Aquaculture Hatchery Technology. Woodhead Publishing Limited.

54. Grossart H-P, Simon M. 2007. Interactions of planctonic algae and bacteria: Effects on algal growth and organic matter dynamics. Aquat Microb Ecol 47:163–176.

55. Berland BR, Bonin DJ, Maestrini SY. 1970. Study of bacteria associated

63

with marine algae in culture. Mar Biol 3:68–76.

56. Salvesen I, Reitan KI, Skjermo J, Øie G. 2000. Microbial environments in marine larviculture: Impacts of algal growth rates on the bacterial load in six microalgae. Aquac Int 8:275–287.

57. Nakase G, Eguchi M. 2007. Analysis of bacterial communities in Nannochloropsis sp. cultures used for larval fish production. Fish Sci 73:543–549.

58. Dhont J, Dierckens K, Støttrup J, Van Stappen G, Wille M, Sorgeloos P. 2013. Rotifers, Artemia and copepods as live feeds for fish larvae in aquaculture, p. 157–202. In Advances in Aquaculture Hatchery Technology. Elsevier.

59. Dhert P, Rombaut G, Suantika G, Sorgeloos P. 2001. Advancement of rotifer culture and manipulation techniques in Europe. Aquaculture 200:129–146.

60. Haché R, Plante S. 2011. The relationship between enrichment, fatty acid profiles and bacterial load in cultured rotifers (Brachionus plicatilis L-strain) and Artemia (Artemia salina strain Franciscana). Aquaculture 311:201–208.

61. Verdonck L, Grisez L, Sweetman E, Minkoff G, Sorgeloos P, Ollevier F, Swings J. 1997. Vibrios associated with routine productions of Brachionus plicatilis. Aquaculture 149:203–214.

62. Verdonck L, Swings J, Kersters K, Dehasque M, Sorgeloos P, Leger P. 1994. Variability of the Microbial Environment of Rotifer Brahionis plicatilis and Artemia Production Systems. J World Aquac Soc 25:55–59.

63. Skjermo J, Vadstein O. 1993. Characterization of the bacterial flora of mass cultivated Brachionus plicatilis. Hydrobiologia 255/256:185–191.

64. Qi Z, Dierckens K, Defoirdt T, Sorgeloos P, Boon N, Bao Z, Bossier P. 2009. Analysis of the evolution of microbial communities associated with different cultures of rotifer strains belonging to different cryptic species of the Brachionus plicatilis species complex. Aquaculture 292:23–29.

65. Bell JG, McEvoy LA, Estevez A, Shields RJ, Sargent JR. 2003. Optimising lipid nutrition in first-feeding flatfish larvae. Aquaculture 227:211–220.

66. Høj L, Bourne DG, Hall MR. 2009. Localization, abundance and community

64

structure of bacteria associated with Artemia: Effects of nauplii enrichment and antimicrobial treatment. Aquaculture 293:278–285.

67. Olsen AI, Olsen Y, Attramadal Y, Christie K, Birkbeck TH, Skjermo J, Vadstein O. 2000. Effects of short term feeding of microalgae on the bacterial flora associated with juvenile Artemia franciscana. Aquaculture 190:11–25.

68. Austin B, Austin D, Sutherland R, Thompson F, Swings J. 2005. Pathogenicity of vibrios to rainbow trout (Oncorhynchus mykiss, Walbaum) and Artemia nauplii. Environ Microbiol 7:1488–1495.

69. Støttrup JG, Richardson K, Kirkegaard E, Pihl NJ. 1986. The cultivation of Acartia tonsa Dana for use as a live food source for marine fish larvae. Aquaculture 52:87–96.

70. Støttrup JG, Norsker NH. 1997. Production and use of copepods in marine fish larviculture. Aquaculture 155:231–247.

71. Turingan RG, Beck JL, Krebs JM, Licamele JD. 2007. Development of Feeding Mechanics in Marine Fish Larvae and the Swimming Behavior of Zooplankton Prey: Implications for Rearing Marine Fishes, p. 119–132. In Copepods in Aquaculture. Blackwell Publishing Professional, Ames, Iowa, USA.

72. Øie G, Galloway T, Sørøy M, Holmvaag Hansen M, Norheim IA, Halseth CK, Almli M, Berg M, Gagnat MR, Wold PA, Attramadal K, Hagemann A, Evjemo JO, Kjørsvik E. 2017. Effect of cultivated copepods (Acartia tonsa) in first-feeding of Atlantic cod (Gadus morhua) and ballan wrasse (Labrus bergylta) larvae. Aquac Nutr 23:3–17.

73. Abate TG, Nielsen R, Nielsen M, Jepsen PM, Hansen BW. 2015. A cost-effectiveness analysis of live feeds in juvenile turbot Scophthalmus maximus (Linnaeus, 1758) farming: Copepods versus Artemia. Aquac Nutr.

74. Sochard MR, Wilson DF, Austin B, Colwell RR. 1979. Bacteria associated with the surface and gut of marine copepods. Appl Environ Microbiol 37:750–759.

75. Colwell RR, Huq A, Islam MS, Aziz KMA, Yunus M, Khan NH, Mahmud A, Sack RB, Nair GB, Chakraborty J, Sack DA, Russek-Cohen E. 2003. Reduction of cholera in Bangladeshi villages by simple filtration. Proc Natl

65

Acad Sci 100:1051–1055.

76. Heidelberg JF, Heidelberg KB, Colwell RR. 2002. Bacteria of the γ-subclass Proteobacteria associated with zooplankton in Chesapeake Bay. Appl Environ Microbiol 68:5498–5507.

77. Gugliandolo C, Irrera GP, Lentini V, Maugeri TL. 2008. Pathogenic Vibrio, Aeromonas and Arcobacter spp. associated with copepods in the Straits of Messina (Italy). Mar Pollut Bull 56:600–606.

78. Rasmussen BB, Erner KE, Bentzon-Tilia M, Gram L. 2018. Effect of TDA-producing Phaeobacter inhibens on the fish pathogen Vibrio anguillarum in non-axenic algae and copepod systems. Microb Biotechnol 11:1070–1079.

79. Grisez L, Chair M, Sorgeloos P, Ollevier F. 1996. Mode of infection and spread of Vibrio anguillarum in turbot Scophthalmus maximus larvae after oral challenge through live feed. Dis Aquat Organ 26:181–187.

80. Verner-Jeffreys D, Shields R, Birkbeck TH. 2003. Bacterial influences on Atlantic halibut Hippoglossus hippoglossus yolk sac larval survival and start feed response. Dis Aquat Organ 56:105–113.

81. Munro PD, Barbour A, Birkbeck TH. 1995. Comparison of the growth and survival of larval turbot in the absence of culturable bacteria with those in the presence of Vibrio anguillarum, Vibrio alginolyticus, or a marine Aeromonas sp. Appl Environ Microbiol 61:4425–4428.

82. Dumontet S, Krovacek K, Baloda SB, Grottoli R, Pasquale V, Vanucci S. 1996. Ecological relationship between Aeromonas and Vibrio spp. and planktonic copepods in the coastal marine environment in southern Italy. Comp Immunol Microbiol Infect Dis 19:245–54.

83. Lightner DV. 2011. Virus diseases of farmed shrimp in the Western Hemisphere (the Americas): A review. J Invertebr Pathol 106:110–130.

84. Costello MJ. 2009. The global economic cost of sea lice to the salmonid farming industry. J Fish Dis 32:115–118.

85. Ellis a. E. 2001. Innate host defense mechanisms of fish against viruses and bacteria. Dev Comp Immunol 25:827–839.

86. Ellis a. E. 1999. Immunity to bacteria in fish. Fish Shellfish Immunol 9:291–308.

66

87. Magnadóttir B. 2006. Innate immunity of fish (overview). Fish Shellfish Immunol 20:137–151.

88. Holland MCH, Lambris JD. 2002. The complement system in teleosts. Fish Shellfish Immunol 12:399–420.

89. Hoehne K, Kj E, Reitan I. 2001. Development of the pH in the intestinal tract of larval turbot. Mar Biol 139:1159–1164.

90. Yúfera M, Darías MJ. 2007. Changes in the gastrointestinal pH from larvae to adult in Senegal sole (Solea senegalensis). Aquaculture 267:94–99.

91. Schrøder MB, Villena AJ, Jørgensen TO. 1998. Ontogeny of lymphoid organs and immunoglobulin producing cells in Atlantic cod (Gadus morhua L.). Dev Comp Immunol 22:507–517.

92. Frans I, Michiels CW, Bossier P, Willems K a., Lievens B, Rediers H. 2011. Vibrio anguillarum as a fish pathogen: Virulence factors, diagnosis and prevention. J Fish Dis 34:643–661.

93. Larsen JL, Pedersen K, Dalsgaard I. 1994. Vibrio anguillarum serovars associated with vibriosis in fish. J Fish Dis 17:259–267.

94. Bergh Ø, Hjeltnes B, Skiftesvik A. 1997. Experimental infection of turbot Scophthalmus maximus and halibut Hippoglossus hippoglossus yolk sac larvae with Aeromonas salmonicida subsp. salmonicida. Dis Aquat Organ 29:13–20.

95. Wiklund T, Dalsgaard I. 1998. Occurrence and significance of atypical Aeromonas salmonicida in non-salmonid and salmonid fish species: A review. Dis Aquat Organ 32:49–69.

96. Austin DA, McIntosh D, Austin B. 1989. Taxonomy of Fish Associated Aeromonas spp., with the Description of Aeromonas salmonicida subsp. smithia subsp. nov. Syst Appl Microbiol 11:277–290.

97. Sanders E, Fryer JL. 1980. Renibacterium salmoninarum gen. nov., sp. nov., the Causative Agent of Bacterial Kidney Disease in Salmonid Fishes. Int J Syst Bacteriol 30:496–502.

98. Starliper CE, Smith DR, Shatzer T. 1997. Virulence of Renibacterium salmoninarum to Salmonids. J Aquat Anim Health 9:1–7.

99. Nematollahi A, Decostere A, Pasmans F, Haesebrouck F. 2003.

67

Flavobacterium psychrophilum infections in salmonid fish. J Fish Dis 26:563–574.

100. Nylund A, Ottem ÆKF, Watanabe ÆK. 2006. Francisella sp. (Family Francisellaceae) causing mortality in Norwegian cod (Gadus morhua) farming. Arch Microbiol 185:383–392.

101. Birkbeck TH, Bordevik M, Frøystad MK, Baklien Å. 2007. Identification of Francisella sp. from Atlantic salmon, Salmo salar L., in Chile. J Fish Dis 30:505–507.

102. Mikalsen J, Olsen AB, Tengs T, Colquhoun DJ. 2007. Francisella philomiragia subsp. noatunensis subsp. nov., isolated from farmed Atlantic cod (Gadus morhua L.). Int J Syst Evol Microbiol 57:1960–1965.

103. Olsen AB, Mikalsen J, Rode M, Alfjorden A, Hoel E, Straum-Lie K, Haldorsen R, Colquhoun DJ. 2006. A novel systemic granulomatous inflammatory disease in farmed Atlantic cod, Gadus morhua L., associated with a bacterium belonging to the genus Francisella. J Fish Dis 29:307–311.

104. Toranzo AE, Magariños B, Romalde JL. 2005. A review of the main bacterial fish diseases in mariculture systems. Aquaculture 246:37–61.

105. Austin B. 2011. Taxonomy of bacterial fish pathogens. Vet Res 42:20.

106. Austin B, Austin DA. 2016. Bacterial Fish Pathogens - Disease of Farmed and Wild FishSixth Edit. Springer International Publishing, Cham.

107. Eda Yardımcı R, Timur G. 2015. Isolation and identification of tenacibaculum maritimum, the causative agent of tenacibaculosis in farmed sea bass (dicentrarchus labrax) on the aegean sea coast of Turkey. Isr J Aquac - Bamidgeh 67.

108. Faílde LD, Losada AP, Bermúdez R, Santos Y, Quiroga MI. 2013. Tenacibaculum maritimum infection: Pathology and immunohistochemistry in experimentally challenged turbot (Psetta maxima L.). Microb Pathog 65:82–88.

109. Kolygas MN, Gourzioti E, Vatsos IN, Athanassopoulou F. 2012. Identification of Tenacibaculum maritimum strains from marine farmed fish in Greece. Vet Rec 170:623.1-623.

110. Hansen GH, Bergh O, Michaelsen J, Knappskog D. 1992. Flexibacter ovolyticus sp. nov., a Pathogen of Eggs and Larvae of Atlantic Halibut,

68

Hippoglossus hippoglossus L. Int J Syst Bacteriol 42:451–458.

111. Sandlund N, Bergh Ø. 2008. Screening and characterisation of potentially pathogenic bacteria associated with Atlantic cod Gadus morhua larvae: bath challenge trials using a multidish system. Dis Aquat Organ 81:203–217.

112. Sandlund N, Rødseth OM, Knappskog DH, Fiksdal IU, Bergh Ø. 2010. Comparative susceptibility of turbot, halibut, and cod yolk-sac larvae to challenge with Vibrio spp. Dis Aquat Organ 89:29–37.

113. Kim D, Han H, Kim S, Lee D, Park S. 2004. Bacterial enteritis and the development of the larval digestive tract in olive flounder, Paralichthys olivaceus (Temminck & Schlegel). J Fish Dis 27:497–505.

114. Thompson FL, Iida T, Swings J. 2004. Biodiversity of Vibrios. Microbiol Mol Biol Rev 68:403–431.

115. Nishiguchi MK. 2003. Evolution of symbiosis in the Vibrionaceae: a combined approach using molecules and physiology. Int J Syst Evol Microbiol 53:2019–2026.

116. Sawabe T, Ogura Y, Matsumura Y, Feng G, Amin AR, Mino S, Nakagawa S, Sawabe T, Kumar R, Fukui Y, Satomi M, Matsushima R, Thompson FL, Gomez-Gil B, Christen R, Maruyama F, Kurokawa K, Hayashi T. 2013. Updating the Vibrio clades defined by multilocus sequence phylogeny: proposal of eight new clades, and the description of Vibrio tritonius sp. nov. Front Microbiol 4:1–14.

117. Urbanczyk H, Ast JC, Dunlap P V. 2011. Phylogeny, genomics, and symbiosis of Photobacterium. FEMS Microbiol Rev 35:324–342.

118. Urbanczyk H, Ast JC, Higgins MJ, Carson J, Dunlap P V. 2007. Reclassification of Vibrio fischeri, Vibrio logei, Vibrio salmonicida and Vibrio wodanis as Aliivibrio fischeri gen. nov., comb. nov., Aliivibrio logei comb. nov., Aliivibrio salmonicida comb. nov. and Aliivibrio wodanis comb. nov. Int J Syst Evol Microbiol 57:2823–2829.

119. Benediktsdóttir E, Helgason S, Sigurjónsdóttir H. 1998. Vibrio spp. isolated from salmonids with shallow skin lesions and reared at low temperature. J Fish Dis 21:19–28.

120. Kashulin A, Seredkina N, Sørum H. 2017. Cold-water vibriosis. The current

69

status of knowledge. J Fish Dis 40:119–126.

121. Karlsen C, Vanberg C, Mikkelsen H, Sørum H. 2014. Co-infection of Atlantic salmon (Salmo salar), by Moritella viscosa and Aliivibrio wodanis, development of disease and host colonization. Vet Microbiol 171:112–121.

122. Reidl J, Klose KE. 2002. Vibrio cholerae and cholera: out of the water and into the host. FEMS Microbiol Rev 26:125–139.

123. Vandenberghe J, Thompson FL, Gomez-Gil B, Swings J. 2003. Phenotypic diversity amongst Vibrio isolates from marine aquaculture systems. Aquaculture 219:9–20.

124. Gordon K V., Vickery MC, DePaola A, Staley C, Harwood VJ. 2008. Real-time PCR assays for quantification and differentiation of Vibrio vulnificus strains in oysters and water. Appl Environ Microbiol 74:1704–1709.

125. Gugliandolo C, Lentini V, Spanò A, Maugeri TL. 2011. Conventional and molecular methods to detect bacterial pathogens in mussels. Lett Appl Microbiol 52:15–21.

126. McLaughlin JB, DePaola A, Bopp CA, Martinek KA, Napolilli NP, Allison CG, Murray SL, Thompson EC, Bird MM, Middaugh JP. 2005. Outbreak of Vibrio parahaemolyticus Gastroenteritis Associated with Alaskan Oysters. N Engl J Med 353:1463–1470.

127. Baker-Austin C, Stockley L, Rangdale R, Martinez-Urtaza J. 2010. Environmental occurrence and clinical impact of Vibrio vulnificus and Vibrio parahaemolyticus: A European perspective. Environ Microbiol Rep 2:7–18.

128. Drake SL, DePaola A, Jaykus L. 2007. An Overview of Vibrio vulnificus and Vibrio parahaemolyticus. Compr Rev Food Sci Food Saf 6:120–144.

129. Austin B, Zhang X-H. 2006. Vibrio harveyi: a significant pathogen of marine vertebrates and invertebrates. Lett Appl Microbiol 43:119–124.

130. Thomson R, Macpherson HL, Riaza A, Birkbeck TH. 2005. Vibrio splendidus biotype 1 as a cause of mortalities in hatchery-reared larval turbot, Scophthalmus maximus (L.). J Appl Microbiol 99:243–250.

131. Schiewe MH, Trust TJ, Crosa JH. 1981. Vibrio ordalii sp. nov.: A causative agent of vibriosis in fish. Curr Microbiol 6:343–348.

132. Rønneseth A, Castillo D, D’Alvise P, Tønnesen Ø, Haugland G, Grotkjaer T,

70

Engell-Sørensen K, Nørremark L, Bergh Ø, Wergeland HI, Gram L. 2017. Comparative assessment of Vibrio virulence in marine fish larvae. J Fish Dis.

133. Balebona MC, Andreu MJ, Bordas MA, Zorilla I, Moriñgo MA, Borrego JJ. 1998. Pathogenicity of Vibrio alginolyticus for cultured gilt-head sea bream (Sparus aurata L.). Appl Environ Microbiol 64:4269–4275.

134. Zorrilla I, Chabrillón M, Arijo S, Dıaz-Rosales P, Martınez-Manzanares E, Balebona M., Moriñigo M. 2003. Bacteria recovered from diseased cultured gilthead sea bream (Sparus aurata L.) in southwestern Spain. Aquaculture 218:11–20.

135. Varvarigos P. 2007. Vibrio alginolyticus combined with the motile Aeromonads Aeromonas sobria and Aeromonas hydrophila comprise the main bacterial taxa responsible for gilthead sea bream (Sparus aurata) larval enteropathy (LE). Eur Aquac Fish Pathol 13th Int Conf Fish Shellfish Dis O–126.

136. MacDonell MT, Colwell RR. 1985. Phylogeny of the Vibrionaceae, and Recommendation for Two New Genera, Listonella and Shewanella. Syst Appl Microbiol 6:171–182.

137. Hendrie MS, Hodgkiss W, Shewan JM. 1971. Proposal That the Species Vi brio anguillarum Bergman 1909, Vibriopiscium David 1927, and Vibrio ichthyodermis (Wells and ZoBell) Shewan, Hobbs, and Hodgkiss 1960 Be Combined as a Single Species, Vibrio anguillarum. Int J Syst Bacteriol 21:64–68.

138. Sørensen UB, Larsen JL. 1986. Serotyping of Vibrio anguillarum. Appl Environ Microbiol 51:593–7.

139. Pedersen K, Gram L, Austin D a, Austin B. 1997. Pathogenicity of Vibrio anguillarum serogroup O1 strains compared to plasmids, outer membrane protein profiles and siderophore production. J Appl Microbiol 82:365–371.

140. Pedersen K, Grisez L, Van Houdt R, Tiainen T, Ollevier F, Larsen JL. 1999. Extended serotyping scheme for Vibrio anguillarum with the definition and characterization of seven provisional O-serogroups. Curr Microbiol 38:183–189.

141. Rucker RR. 1963. Status of fish diseases and relation to production, p. 98–101. In Report of the Second Governor’s Conference on Pacific Salmon,

71

Seattle.

142. Kubota SS, Takakuwa M. 1963. Studies on the disease of marine culture fishes. 1. General description and preliminary discussion of fish diseases at Mie Prefecture. J Fac Fish Prefect Univ Mie 6:107–124.

143. Bagger J, Bagger O. 1956. Vibrio anguillarum som arsag til ulcussygdom has torsk (Gadus callarias L.) (Vibrio anguillarum as cause of the ulcer disease in cod). Nord Vet Med 8:481–492.

144. Haenen OLM, Fouz B, Amaro C, Isern MM, Mikkelsen H, Zrncic S, Travers MA, Renault T, Wardle R, Hellström A, Dalsgaard I. 2014. Vibriosis in aquaculture. 16th EAFP Conference, Tampere, Finland, 4th September 2013. Bull Eur Assoc Fish Pathol 34:138–148.

145. Lei Q, Yin Geng W, Zheng Z, Shao Li Y. 2006. The first report on fin rot disease of cultured turbot Scophthalmus maximus in China. J Aquat Anim Health 18:83–89.

146. Spanggaard B, Huber I, Nielsen J, Nielsen T, Gram L. 2000. Proliferation and location of Vibrio anguillarum during infection of rainbow trout, Oncorhynchus mykiss (Walbaum). J Fish Dis 23:423–427.

147. Weber B, Chen C, Milton DL. 2010. Colonization of fish skin is vital for Vibrio anguillarum to cause disease. Environ Microbiol Rep 2:133–139.

148. Larsen JL. 1984. Vibrio anguillavum: Influence of temperature, pH, NaCl concentration and incubation time on growth. J Appl Bacteriol 57:237–246.

149. O’Toole R, Milton DL, Wolf-Watz H. 1996. Chemotactic motility is required for invasion of the host by the fish pathogen Vibrio anguillarum. Mol Microbiol 19:625–637.

150. O’Toole R, Lundberg S, Fredriksson SÅ, Jansson A, Nilsson B, Wolf-Watz H. 1999. The chemotactic response of Vibrio anguillarum to fish intestinal mucus is mediated by a combination of multiple mucus components. J Bacteriol 181:4308–4317.

151. Norqvist a., Norrman B, Wolf-Watz H. 1990. Identification and characterization of a zinc metalloprotease associated with invasion by the fish pathogen Vibrio anguillarum. Infect Immun 58:3731–3736.

152. Milton DL, Norqvist a., Wolf-Watz H. 1992. Cloning of a metalloprotease gene involved in the virulence mechanism of Vibrio anguillarum. J Bacteriol

72

174:7235–7244.

153. Rodkhum C, Hirono I, Stork M, Di Lorenzo M, Crosa JH, Aoki T. 2006. Putative virulence-related genes in Vibrio anguillarum identified by random genome sequencing. J Fish Dis 29:157–166.

154. Hirono I, Masuda T, Aoki T. 1996. Cloning and detection of the hemolysin gene of Vibrio anguillarum. Microb Pathog 21:173–82.

155. Zhang X-H, Austin B. 2005. Haemolysins in Vibrio species. J Appl Microbiol 98:1011–1019.

156. Castillo D, Alvise PD, Xu R, Zhang F, Middelboe M. 2017. Comparative Genome Analyses of Vibrio anguillarum Strains Reveal a Link with pathogenicity traits. mSystems 2:1–14.

157. Milton DL. 2006. Quorum sensing in vibrios: Complexity for diversification. Int J Med Microbiol 296:61–71.

158. Weber B, Croxatto A, Chen C, Milton DL. 2008. RpoS induces expression of the Vibrio anguillarum quorum-sensing regulator VanT. Microbiology 154:767–780.

159. Skjermo J, Vadstein O. 1999. Techniques for microbial control in the intensive rearing of marine larvae. Aquaculture 177:333–343.

160. Jorquera M a., Valencia G, Eguchi M, Katayose M, Riquelme C. 2002. Disinfection of seawater for hatchery aquaculture systems using electrolytic water treatment. Aquaculture 207:213–224.

161. Salvesen I, Oie G, Vadstein O. 1997. Surface disinfection of Atlantic halibut and turbot eggs with glutaraldehyde: Evaluation of concentrations and contact times. Aquac Int 5:249–258.

162. Harboe T, Huse I, Oie G. 1994. Effects of egg disinfection on yolk sac and first feeding stages of halibut (Hippoglossus hippoglossus L.) larvae. Aquaculture 119:157–165.

163. De Swaef E, Van den Broeck W, Dierckens K, Decostere A. 2016. Disinfection of teleost eggs: a review. Rev Aquac 8:321–341.

164. Cabello FC. 2006. Heavy use of prophylactic antibiotics in aquaculture: A growing problem for human and animal health and for the environment. Environ Microbiol 8:1137–1144.

73

165. Norwegian Veterinary Institute. 2016. Use of Antibiotics in Norwegian Aquaculture. Rep from Nor Vet Inst Report 22.

166. Holmstrom K, Graslund S, Wahlstrom A, Poungshompoo S, Bengtsson B-E, Kautsky N. 2003. Antibiotic use in shrimp farming and implications for environmental impacts and human health. Int J Food Sci Technol 38:255–266.

167. Higuera-Llantén S, Vásquez-Ponce F, Barrientos-Espinoza B, Mardones FO, Marshall SH, Olivares-Pacheco J. 2018. Extended antibiotic treatment in salmon farms select multiresistant gut bacteria with a high prevalence of antibiotic resistance genes. PLoS One 13:1–22.

168. Evelyn TP. 1997. A historical review of fish vaccinology. Dev Biol Stand 90:3–12.

169. Bricknell I, Dalmo R. 2005. The use of immunostimulants in fish larval aquaculture. Fish Shellfish Immunol 19:457–472.

170. Vadstein O. 1997. The use of immunostimulation in marine larviculture: possibilities and challenges. Aquaculture 155:401–417.

171. Rasch M, Buch C, Austin B, Slierendrecht WJ, Ekmann KS, Larsen JL, Johansen C, Riedel K, Eberl L, Givskov M, Gram L. 2004. An inhibitor of bacterial quorum sensing reduces mortalities caused by Vibriosis in rainbow trout (Oncorhynchus mykiss, Walbaum). Syst Appl Microbiol 27:350–359.

172. Tinh NTN, Das M, Bossier P, Li X, Dung NV. 2017. No effect of N-acyl homoserine lactones disruption by lactonase enzyme on the virulence of Vibrio anguillarum towards sea bass (Dicentrarchus labrax) and brine shrimp (Artemia franciscana). Aquac Int 26:363–374.

173. Skjermo J, Saivesen I, Øie G, Olsen Y, Vadstein O. 1997. Microbially matured water: a technique for selection of a non-opportunistic bacterial flora in water that may improve performance of marine larvae. Aquac Int 5:13–28.

174. Salvesen I, Skjermo J, Vadstein O. 1999. Growth of turbot (Scophthalmus maximus L.) during first feeding in relation to the proportion of r/K-strategists in the bacterial community of the rearing water. Aquaculture 175:337–350.

74

175. Attramadal KJK, Truong TMH, Bakke I, Skjermo J, Olsen Y, Vadstein O. 2014. RAS and microbial maturation as tools for K-selection of microbial communities improve survival in cod larvae. Aquaculture 432:483–490.

176. Suttle CA. 2005. Viruses in the sea. Nature 437:356–361.

177. Rohwer F, Thurber RV. 2009. Viruses manipulate the marine environment. Nature 459:207–212.

178. Matsuzaki S, Tanaka S, Koga T, Kawata T. 1992. A Broad-Host-Range Vibriophage, KVP40, Isolated from Sea Water. Microbiol Immunol 36:93–97.

179. Nakai T, Sugimoto R, Park KH, Matsuoka S, Mori K, Nishioka T, Maruyama K. 1999. Protective effects of bacteriophage on experimental Lactococcus garvieae infection in yellowtail. Dis Aquat Organ 37:33–41.

180. Park SC, Nakai T. 2003. Bacteriophage control of Pseudomonas plecoglossicida infection in ayu Plecoglossus altivelis 53:33–39.

181. Mateus L, Costa L, Silva YJ, Pereira C, Cunha A, Almeida A. 2014. Efficiency of phage cocktails in the inactivation of Vibrio in aquaculture. Aquaculture 424–425:167–173.

182. Silva YJ, Costa L, Pereira C, Mateus C, Cunha Â, Calado R, Gomes NCM, Pardo MA, Hernandez I, Almeida A. 2014. Phage therapy as an approach to prevent Vibrio anguillarum infections in fish larvae production. PLoS One 9:1–23.

183. Higuera G, Bastías R, Tsertsvadze G, Romero J, Espejo RT. 2013. Recently discovered Vibrio anguillarum phages can protect against experimentally induced vibriosis in Atlantic salmon, Salmo salar. Aquaculture 392–395:128–133.

184. Karunasagar I, Shivu MM, Girisha SK, Krohne G, Karunasagar I. 2007. Biocontrol of pathogens in shrimp hatcheries using bacteriophages. Aquaculture 268:288–292.

185. Karunasagar I, Vinod MG, Kennedy B, Vijay A, Deepanjali A, Umesh KR, Karunasagar I. 2005. Biocontrol of Bacterial Pathogens in Aquaculture with Emphasis on Phage Therapy. Dis Asian Aquac V 535–542.

186. Rørbo N, Rønneseth A, Kalatzis PG, Rasmussen BB, Engell-Sørensen K, Kleppen HP, Wergeland HI, Gram L, Middelboe M. 2018. Exploring the

75

Effect of Phage Therapy in Preventing Vibrio anguillarum Infections in Cod and Turbot Larvae. Antibiotics 7:42.

187. Rasmussen BB, Kalatzis PG, Middelboe M, Gram L. 2019. Combining probiotic Phaeobacter inhibens DSM17395 and broad-host-range vibriophage KVP40 against fish pathogenic vibrios. Submitt Manuscr.

188. Castillo D, Rørbo N, Jørgensen J, Lange J, Tan D, Kalatzis PG, Svenningsen S Lo, Middelboe M. 2019. Phage defense mechanisms and their genomic and phenotypic implications in the fish pathogen Vibrio anguillarum. FEMS Microbiol Ecol 95.

189. Castillo D, Christiansen RH, 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.

190. Tan D, Svenningsen S Lo, Middelboe M. 2015. Quorum Sensing Determines the Choice of Antiphage Defense Strategy in Vibrio anguillarum. MBio 6:1–10.

191. Tan D, Gram L, Middelboe M. 2014. Vibriophages and their interactions with the fish pathogen Vibrio anguillarum. Appl Environ Microbiol 80:3128–40.

192. Dittmann KKKKK, Rasmussen BBBB, Castex M, Gram L, Bentzon-Tilia M. 2017. The aquaculture microbiome at the centre of business creation. Microb Biotechnol 10:1279–1282.

193. Gatesoupe F-J. 1991. The effect of three strains of lactic bacteria on the production rate of rotifers, Brachionus plicatilis, and their dietary value for larval turbot, Scophthalmus maximus. Aquaculture 96:335–342.

194. Gatesoupe F. 1994. Lactic acid bacteria increase the resistance of turbot larvae , Scophthalmus maximus , against pathogenic vibrio. Aquat Living Resour 7:277–282.

195. Noh S., Han K, Won T., Choi Y. 1994. Effect of antibiotics, enzyme, yeast culture and probiotics on the growth performance of Israeli carp. Korean J Anim Sci 36:480–486.

196. Gildberg A, Mikkelsen H, Sandaker E, Ringø E. 1997. Probiotic effect of lactic acid bacteria in the feed on growth and survival of fry of Atlantic cod

76

(Gadus morhua). Hydrobiologia 352:279–285.

197. Gildberg A, Johansen A, Bøgwald J. 1995. Growth and survival of Atlantic salmon (Salmo salar) fry given diets supplemented with fish protein hydrolysate and lactic acid bacteria during a challenge trial with Aeromonas salmonicida. Aquaculture 138:23–34.

198. Bergh Ø. 1995. Bacteria associated with early life stages of halibut, Hippoglossus hippoglossus L., inhibit growth of a pathogenic Vibrio sp. J Fish Dis 18:31–40.

199. Austin B, Stuckey L, Robertson P, Effendi I, Griffith D. 1995. A probiotic strain of Vibrio alginolyticus effective in reducing diseases caused by Aeromonas salmonicida, Vibrio anguillarum and Vibrio ordalii. J Fish Dis 18:93–96.

200. García De La Banda I, Lobo C, León-rubio JM, Tapia-paniagua S, Balebona MC, Moriñigo MA, Moreno-ventas X, Lucas LM, Linares F, Arce F, Arijo S. 2010. Influence of two closely related probiotics on juvenile Senegalese sole (Solea senegalensis, Kaup 1858) performance and protection against Photobacterium damselae subsp. piscicida. Aquaculture 306:281–288.

201. Cordero H, Morcillo P, Meseguer J, Cuesta A, Esteban MA. 2016. Effects of Shewanella putrefaciens on innate immunity and cytokine expression profile upon high stocking density of gilthead seabream specimens. Fish Shellfish Immunol 51:33–40.

202. Korkea-aho TL, Papadopoulou A, Heikkinen J, von Wright A, Adams A, Austin B, Thompson KD. 2012. Pseudomonas M162 confers protection against rainbow trout fry syndrome by stimulating immunity. J Appl Microbiol 113:24–35.

203. Gram L, Melchiorsen J, Spanggaard B, Huber I, Al GET, Icrobiol APPLENM. 1999. Inhibition of Vibrio anguillarum by Pseudomonas fluorescens AH2, a Possible Probiotic Treatment of Fish. Appl Environ Microbiol 65:969–973.

204. Simon M, Scheuner C, Meier-Kolthoff JP, Brinkhoff T, Wagner-Döbler I, Ulbrich M, Klenk H-P, Schomburg D, Petersen J, Göker M. 2017. Phylogenomics of Rhodobacteraceae reveals evolutionary adaptation to marine and non-marine habitats. ISME J 11:1483–1499.

205. Shiba T, Simidu U, Taga N. 1979. Distribution of Aerobic-Bacteria Which

77

Contain Bacteriochlorophyll-A. Appl Environ Microbiol 38:43–45.

206. Shiba T. 1991. Roseobacter litoralis gen. nov., sp. nov., and Roseobacter denitrificans sp. nov., Aerobic Pink-Pigmented Bacteria which Contain Bacteriochlorophyll a. Syst Appl Microbiol 14:140–145.

207. Pujalte MJ, Lucena T, Ruvira MA, Arahal DR, Macian M. 2014. The Prokaryotes – Alphaproteobacteria and Betaproteobacteria.

208. Buchan A, Gonzalez JM, Moran MA. 2005. Overview of the Marine Roseobacter Lineage. Appl Environ Microbiol 71:5665–5677.

209. Brinkhoff T, Giebel H-A, Simon M. 2008. Diversity, ecology, and genomics of the Roseobacter clade: a short overview. Arch Microbiol 189:531–9.

210. Gonzalez JM, Simo R, Massana R, Covert JS, Casamayor EO, Pedros-Alio C, Moran MA. 2000. Bacterial Community Structure Associated with a Dimethylsulfoniopropionate-Producing North Atlantic Algal Bloom. Appl Environ Microbiol 66:4237–4246.

211. Wietz M, Gram L, Jørgensen B, Schramm a. 2010. Latitudinal patterns in the abundance of major marine bacterioplankton groups. Aquat Microb Ecol 61:179–189.

212. Alavi M, Miller T, Erlandson K, Schneider R, Belas R. 2001. Bacterial community associated with Pfesteria-like dinoflagellate cultures. Environ Microbiol 3:380–396.

213. Geng H, Belas R. 2010. Molecular mechanisms underlying roseobacter-phytoplankton symbioses. Curr Opin Biotechnol 21:332–8.

214. Penesyan A, Breider S, Schumann P, Tindall BJ, Egan S, Brinkhoff T. 2013. Epibacterium ulvae gen. nov., sp. nov., epibiotic bacteria isolated from the surface of a marine alga. Int J Syst Evol Microbiol 63:1589–1596.

215. Ivanova EP, Gorshkova NM, Sawabe T, Zhukova N V., Hayashi K, Kurinlenko V, Alexeeva Y, Buljan V, Nicolau D V., Mikhailov V V., Christen R. 2004. Sulfitobacter delicatus sp. nov. and Sulfitobacter dubius sp. nov., respectively from a starfish (Stellaster equestris) and sea grass (Zostera marina). Int J Syst Evol Microbiol 54:475–480.

216. Rao D, Webb JS, Kjelleberg S. 2005. Competitive interactions in mixed-species biofilms containing the marine bacterium Pseudoalteromonas tunicata. Appl Environ Microbiol 71:1729–36.

78

217. Hjelm M, Bergh O, Riaza A, Nielsen J, Melchiorsen J, Jensen S, Duncan H, Ahrens P, Birkbeck H, Gram L. 2004. Selection and identification of autochthonous potential probiotic bacteria from turbot larvae (Scophthalmus maximus) rearing units. Syst Appl Microbiol 27:360–71.

218. Chen MH, Sheu S-Y, Chen CA, Wang J-T, Chen W-M. 2011. Shimia isoporae sp. nov., isolated from the reef-building coral Isopora palifera. Int J Syst Evol Microbiol 61:823–827.

219. Breider S, Freese HM, Spröer C, Simon M, Overmann J, Brinkhoff T. 2017. Phaeobacter porticola sp. nov., an antibiotic-producing bacterium isolated from a sea harbour. Int J Syst Evol Microbiol 67:2153–2159.

220. King GM. 2003. Molecular and Culture-Based Analyses of Aerobic Carbon Monoxide Oxidizer Diversity? Appl Environ Microbiol 69:7257–7265.

221. Collins AJ, Nyholm S V. 2011. Draft genome of Phaeobacter gallaeciensis ANG1, a dominant member of the accessory nidamental gland of Euprymna scolopes. J Bacteriol 193:3397–8.

222. Kim YO, Park S, Nam BH, Kang SJ, Hur YB, Kim DG, Oh TK, Yoon JH. 2012. Description of litoreibacter meonggei sp. nov., isolated from the sea squirt halocynthia roretzi, reclassification of thalassobacter arenae as litoreibacter arenae comb. nov. and emended description of the genus litoreibacter romanenko et al. 2011. Int J Syst Evol Microbiol 62:1825–1831.

223. Lee M-H, Song E-J, Seo M-J, Hyun D-W, Bae J-W, Lee S-Y, Roh SW, Nam Y-D. 2015. Phaeobacter marinintestinus sp . nov ., isolated from the intestine of a sea cucumber (Apostichopus japonicus). Antonie Van Leeuwenhoek 107:209–216.

224. Lee OO, Tsoi MMY, Li X, Wong PK, Qian PY. 2007. Thalassococcus halodurans gen. nov., sp. nov., a novel halotolerant member of the Roseobacter clade isolated from the marine sponge Halichondria panicea at Friday Harbor, USA. Int J Syst Evol Microbiol 57:1919–1924.

225. Lee J, Whon TW, Shin NR, Roh SW, Kim J, Park SK, Kim MS, Shin KS, Lee JS, Lee KC, Kim YO, Bae JW. 2012. Ruegeria conchae sp. nov., isolated from the ark clam Scapharca broughtonii. Int J Syst Evol Microbiol 62:2851–2857.

79

226. Boettcher KJ, Geaghan KK, Maloy AP, Barber BJ. 2005. Roseovarius crassostreae sp. nov., a member of the Roseobacter clade and the apparent cause of juvenile oyster disease (JOD) in cultured Eastern oysters. Int J Syst Evol Microbiol 55:1531–1537.

227. Chen MH, Sheu SY, Chen CA, Wang JT, Chen WM. 2011. Paracoccus isoporae sp. nov., isolated from the reef-building coral isopora palifera. Int J Syst Evol Microbiol 61:1138–1143.

228. Moran MA, González JM, Kiene RP. 2003. Linking a Bacterial Taxon to Sulfur Cycling in the Sea : Studies of the Marine Roseobacter Group. Geomicrobilogy J 20:375–388.

229. Gonzalez JM, Kiene RP, Moran MA. 1999. Tranformation of Sulfur Compounds by an Abundant Lineage of Marine Bacteria in the a-Subclass of the Class Proteobacteria. Appl Environ Microbiol 65:3810–3819.

230. Allgaier M, Felske A, Wagner-do I. 2003. Aerobic Anoxygenic Photosynthesis in Roseobacter Clade Bacteria from Diverse Marine Habitats. Appl Environ Microbiol 69:5051–5059.

231. Buchan A, Collier LS, Neidle EL, Moran MA. 2000. Key Aromatic-Ring-Cleaving Enzyme, Protocatechuate 3,4-Dioxygenase, in the Ecologically Important Marine Roseobacter Lineage. Appl Environ Microbiol 66:4662–4672.

232. Gallacher S, Gallacher S, Flynn KJ, Flynn KJ, Franco JM, Franco JM, Brueggemann EE, Brueggemann EE, Hines HB, Hines HB. 1997. Evidence for production of paralytic shellfish toxins by bacteria associated with Alexandrium spp. (Dinophyta) in culture. Appl Environ Microbiol 63:239–245.

233. Seyedsayamdost MR, Case RJ, Kolter R, Clardy J. 2011. The Jekyll-and-Hyde chemistry of Phaeobacter gallaeciensis. Nat Chem 3:331–5.

234. Cude W, Buchan A. 2013. Acyl-homoserine lactone-based quorum sensing in the Roseobacter clade: complex cell-to-cell communication controls multiple physiologies. Front Microbiol 4:1–12.

235. Brinkhoff T, Bach G, Heidorn T, Liang L, Schlingloff A, Simon M. 2004. Antibiotic production by a Roseobacter clade-affiliated species from the German Wadden Sea and its antagonistic effects on indigenous isolates.

80

Appl Environ Microbiol 70:2560–5.

236. Cude WN, Mooney J, Tavanaei A a., Hadden MK, Frank AM, Gulvik C a., May AL, Buchan A. 2012. Production of the antimicrobial secondary metabolite indigoidine contributes to competitive surface colonization by the marine roseobacter Phaeobacter sp. strain Y4I. Appl Environ Microbiol 78:4771–4780.

237. Martens T, Heidorn T, Pukall R, Simon M, Tindall BJ, Brinkhoff T. 2006. Reclassification of Roseobacter gallaeciensis Ruiz-Ponte et al. 1998 as Phaeobacter gallaeciensis gen. nov., comb. nov., description of Phaeobacter inhibens sp. nov., reclassification of Ruegeria algicola (Lafay et al. 1995) Uchino et al. 1999 as Marinovu. Int J Syst Evol Microbiol 56:1293–304.

238. Park S, Park DS, Bae KS, Yoon JH. 2014. Phaeobacter aquaemixtae sp. nov., isolated from the junction between the ocean and a freshwater spring. Int J Syst Evol Microbiol 64:1378–1383.

239. Yoon J-H, Kang S-J, Lee S-Y, Oh T-K. 2007. Phaeobacter daeponensis sp. nov., isolated from a tidal flat of the Yellow Sea in Korea. Int J Syst Evol Microbiol 57:856–61.

240. Vandecandelaere I, Segaert E, Mollica A, Faimali M, Vandamme P. 2009. Phaeobacter caeruleus sp. nov., a blue-coloured, colony-forming bacterium isolated from a marine electroactive biofilm. Int J Syst Evol Microbiol 59:1209–14.

241. Zhang D-C, Li H-R, Xin Y-H, Liu H-C, Chi Z-M, Zhou P-J, Yu Y. 2008. Phaeobacter arcticus sp. nov., a psychrophilic bacterium isolated from the Arctic. Int J Syst Evol Microbiol 58:1384–7.

242. Gaboyer F, Tindall BJ, Ciobanu MC, Duthoit F, Le Romancer M, Alain K. 2013. Phaeobacter leonis sp. nov., an alphaproteobacterium from Mediterranean Sea sediments. Int J Syst Evol Microbiol 63:3301–3306.

243. Breider S, Scheuner C, Schumann P, Fiebig A, Petersen J, Pradella S, Klenk H-P, Brinkhoff T, Göker M. 2014. Genome-scale data suggest reclassifications in the Leisingera-Phaeobacter cluster including proposals for Sedimentitalea gen. nov. and Pseudophaeobacter gen. nov. Front Microbiol 5:416.

81

244. Sonnenschein EC, Phippen CBW, Nielsen KF, Mateiu RV, Melchiorsen J, Gram L, Overmann J, Freese HM. 2017. Phaeobacter piscinae sp. nov., a species of the Roseobacter group and potential aquaculture probiont. Int J Syst Evol Microbiol 67:4559–4564.

245. Dogs M, Vog S, Teshima H, Petersen J, Davenport K, Woyke T, Simon M, Klenk H, Göker M. 2013. Genome sequence of Phaeobacter inhibens type strain (T5T), a secondary metabolite producing representative of the marine Roseobacter clade, and emendation of the species description of Phaeobacter inhibens. Stand Genomic Sci 9:334–350.

246. Vandecandelaere I, Segaert E, Mollica A, Faimali M, Vandamme P. 2008. Leisingera aquimarina sp. nov., isolated from a marine electroactive biofilm, and emended descriptions of Leisingera methylohalidivorans Schaefer et al. 2002, Phaeobacter daeponensis Yoon et al. 2007 and Phaeobacter inhibens Martens et al. 2006. Int J Syst Evol Microbiol 58:2788–93.

247. Gram L, Rasmussen BB, Wemheuer B, Bernbom N, Ng YY, Porsby CH, Breider S, Brinkhoff T. 2015. Phaeobacter inhibens from the Roseobacter clade has an environmental niche as a surface colonizer in harbors. Syst Appl Microbiol 38:483–493.

248. Buddruhs N, Pradella S, Göker M, Päuker O, Pukall R, Spröer C, Schumann P, Petersen J, Brinkhoff T. 2013. Molecular and phenotypic analyses reveal the non-identity of the Phaeobacter gallaeciensis type strain deposits CIP 105210T and DSM 17395. Int J Syst Evol Microbiol 63:4340–9.

249. Segev E, Tellez A, Vlamakis H, Kolter R. 2015. Morphological Heterogeneity and Attachment of Phaeobacter inhibens. PLoS One 10:e0141300.

250. Martens T, Gram L, Grossart H-P, Kessler D, Müller R, Simon M, Wenzel SC, Brinkhoff T. 2007. Bacteria of the Roseobacter clade show potential for secondary metabolite production. Microb Ecol 54:31–42.

251. Berger M, Neumann A, Schulz S, Simon M, Brinkhoff T. 2011. Tropodithietic acid production in Phaeobacter gallaeciensis is regulated by N-acyl homoserine lactone-mediated quorum sensing. J Bacteriol 193:6576–6585.

252. Thole S, Kalhoefer D, Voget S, Berger M, Engelhardt T, Liesegang H,

82

Wollherr A, Kjelleberg S, Daniel R, Simon M, Thomas T, Brinkhoff T. 2012. Phaeobacter gallaeciensis genomes from globally opposite locations reveal high similarity of adaptation to surface life. ISME J 6:2229–44.

253. Sonnenschein EC, Phippen CBW, Bentzon-Tilia M, Rasmussen SA, Nielsen KF, Gram L. 2018. Phylogenetic distribution of roseobacticides in the Roseobacter group and their effect on microalgae. Environ Microbiol Rep 10:383–393.

254. Segev E, Wyche TP, Kim KH, Petersen J, Ellebrandt C, Vlamakis H, Barteneva N, Paulson JN, Chai L, Clardy J, Kolter R. 2016. Dynamic metabolic exchange governs a marine algal-bacterial interaction. Elife 5:e17473.

255. Rasmussen BB, Dittmann KK, Gram L, Bentzon-Tilia M. 2019. Upscaling probiotic Phaeobacter spp. in Tetraselmis suecica algae cultures. Manuscr Prep.

256. D’Alvise PW, Melchiorsen J, Porsby CH, Nielsen KF, Gram L. 2010. Inactivation of Vibrio anguillarum by attached and planktonic Roseobacter cells. Appl Environ Microbiol 76:2366–70.

257. Prol García MJ, D’Alvise PW, Gram L. 2013. Disruption of cell-to-cell signaling does not abolish the antagonism of Phaeobacter gallaeciensis toward the fish pathogen Vibrio anguillarum in algal systems. Appl Environ Microbiol 79:5414–5417.

258. Porsby CH, Gram L. 2016. Phaeobacter inhibens as biocontrol agent against Vibrio vulnificus in oyster models. Food Microbiol 57:63–70.

259. Liang L. 2003. Investigation of Secondary Metabolites of North Sea Bacteria : Fermentation, Isolation, Structure Elucidation and Bioactivity. PhD Thesis.

260. Kintaka K, Ono H, Tsubotani S, Harada S, Okazaki H. 1984. Thiotropocin, a new sulfur-containing 7-membered-ring antibiotic produced by a Pseudomonas sp. J Antibiot (Tokyo) 37:1294–1300.

261. Tsubotani S, Wada Y, Kamiya K, Okazaki H, Harada S. 1984. Structure of thiotropocin, a new sulfur-containing 7-membered antibiotic. Tetrahedron Lett 25:419–422.

262. Kawano Y, Nagawa Y, Nakanishi H, Nakajima H, Matsuo M, Higashihara T.

83

1997. Production of thiotropocin by a marine bacterium, Caulobacter sp. and its antimicroalgal activities. J Mar Biotechnol 5:225–229.

263. Kawano Y, Asada M, Inoue M, Nakagomi K, Oka S, Higashihara T. 1998. Biological activity of thiotropocin produced by marine bacterium, Caulobacter sp. PK654. J Mar Biotechnol 6:49–52.

264. Porsby CH. 2010. Antagonism of Roseobacter clade bacteria against pathogenic bacteria.

265. Greer EM, Aebisher D, Greer A, Bentley R. 2008. Computational studies of the tropone natural products, thiotropocin, tropodithietic acid, and troposulfenin. Significance of thiocarbonyl-enol tautomerism. J Org Chem 73:280–283.

266. Hjelm M, Riaza A, Formoso F, Melchiorsen J, Gram L. 2004. Seasonal Incidence of Autochthonous Antagonistic Roseobacter spp. and Vibrionaceae Strains in a Turbot Larva (Scophthalmus maximus) Rearing System. Appl Environ Microbiol 70:7288–7294.

267. Geng H, Bruhn JB, Nielsen KF, Gram L, Belas R. 2008. Genetic dissection of tropodithietic acid biosynthesis by marine roseobacters. Appl Environ Microbiol 74:1535–1545.

268. Penesyan A, Tebben J, Lee M, Thomas T, Kjelleberg S, Harder T, Egan S. 2011. Identification of the antibacterial compound produced by the marine epiphytic bacterium Pseudovibrio sp. D323 and related sponge-associated bacteria. Mar Drugs 9:1391–1402.

269. Muramatsu Y, Uchino Y, Kasai H, Suzuki K, Nakagawa Y. 2007. Ruegeria mobilis sp. nov., a member of the Alphaproteobacteria isolated in Japan and Palau. Int J Syst Evol Microbiol 57:1304–1309.

270. Rasmussen BB, Grotkjær T, D’Alvise PW, Yin G, Zhang F, Bunk B, Spröer C, Bentzon-Tilia M, Gram L. 2016. Vibrio anguillarum is genetically and phenotypically unaffected by long-term continuous exposure to the antibacterial compound tropodithietic acid. Appl Environ Microbiol 82:4802–4810.

271. Harrington C, Reen F, Mooij M, Stewart F, Chabot J-B, Guerra A, Glöckner F, Nielsen K, Gram L, Dobson A, Adams C, O’Gara F. 2014. Characterisation of Non-Autoinducing Tropodithietic Acid (TDA) Production from Marine

84

Sponge Pseudovibrio Species. Mar Drugs 12:5960–5978.

272. Perron GG, Zasloff M, Bell G. 2006. Experimental evolution of resistance to an antimicrobial peptide. Proc R Soc B Biol Sci 273:251–256.

273. Hein-Kristensen L, Franzyk H, Holch A, Gram L. 2013. Adaptive Evolution of Escherichia coli to an α-Peptide/β-Peptoid Peptidomimetic Induces Stable Resistance. PLoS One 8.

274. Jochumsen N, Marvig RL, Damkiær S, Jensen RL, Paulander W, Molin S, Jelsbak L, Folkesson A. 2016. The evolution of antimicrobial peptide resistance in Pseudomonas aeruginosa is shaped by strong epistatic interactions. Nat Commun 7:13002.

275. Wilson MZ, Wang R, Gitai Z, Seyedsayamdost MR. 2016. Mode of action and resistance studies unveil new roles for tropodithietic acid as an anticancer agent and the γ-glutamyl cycle as a proton sink. Proc Natl Acad Sci 113:1630–1635.

276. Beyersmann PG, Tomasch J, Son K, Stocker R, Göker M, Wagner-Döbler I, Simon M, Brinkhoff T. 2017. Dual function of tropodithietic acid as antibiotic and signaling molecule in global gene regulation of the probiotic bacterium Phaeobacter inhibens. Sci Rep 7:730.

277. Will SE, Neumann-Schaal M, Heydorn RL, Bartling P, Petersen J, Schomburg D. 2017. The limits to growth – energetic burden of the endogenous antibiotic tropodithietic acid in Phaeobacter inhibens DSM 17395. PLoS One 12:e0177295.

278. Wichmann H, Vocke F, Brinkhoff T, Simon M, Richter-Landsberg C. 2015. Cytotoxic Effects of Tropodithietic Acid on Mammalian Clonal Cell Lines of Neuronal and Glial Origin. Mar Drugs 13:7113–7123.

279. Neu AK, Månsson M, Gram L, Prol-García MJ. 2014. Toxicity of bioactive and probiotic marine bacteria and their secondary metabolites in Artemia sp. and Caenorhabditis elegans as eukaryotic model organisms. Appl Environ Microbiol 80:146–153.

280. Bruhn JB, Gram L, Belas R. 2007. Production of antibacterial compounds and biofilm formation by Roseobacter species are influenced by culture conditions. Appl Environ Microbiol 73:442–50.

281. Geng H, Belas R. 2011. TdaA regulates Tropodithietic acid synthesis by

85

binding to the tdaC promoter region. J Bacteriol 193:4002–5.

282. Geng H, Belas R. 2010. Expression of tropodithietic acid biosynthesis iscontrolled by a novel autoinducer. J Bacteriol 192:4377–87.

283. Berger M, Brock NL, Liesegang H, Dogs M, Preuth I, Simon M, DickschatJS, Brinkhoff T. 2012. Genetic analysis of the upper phenylacetatecatabolic pathway in the production of tropodithietic acid by Phaeobactergallaeciensis. Appl Environ Microbiol 78:3539–3551.

284. Brock NL, Nikolay A, Dickschat JS. 2014. Biosynthesis of the antibiotictropodithietic acid by the marine bacterium Phaeobacter inhibens. ChemCommun (Camb) 50:5487–9.

285. D’Alvise PW, Phippen CBW, Nielsen KF, Gram L. 2016. Influence of Iron onProduction of the Antibacterial Compound Tropodithietic Acid and ItsNoninhibitory Analog in Phaeobacter inhibens. Appl Environ Microbiol82:502–509.

286. Prol Garcia MJ, D’Alvise PW, Rygaard AM, Gram L. 2014. Biofilmformation is not a prerequisite for production of the antibacterialcompound tropodithietic acid in Phaeobacter inhibens DSM17395. J ApplMicrobiol 117:1592–1600.

287. Ruiz-Ponte C, Samain JJF, Sánchez JLJ, Nicolas JJL. 1999. The Benefit of aRoseobacter Species on the Survival of Scallop Larvae. Mar Biotechnol1:52–59.

288. Porsby CH, Nielsen KF, Gram L. 2008. Phaeobacter and Ruegeria species ofthe Roseobacter clade colonize separate niches in a Danish Turbot(Scophthalmus maximus)-rearing farm and antagonize Vibrio anguillarumunder different growth conditions. Appl Environ Microbiol 74:7356–64.

289. Dittmann KK, Sonnenschein EC, Egan S, Gram L, Bentzon-Tilia M. 2018.Impact of Phaeobacter inhibens on marine eukaryote-associated microbialcommunities. Environ Microbiol Rep.

Paper 1

Dittmann KK, Rasmussen BB, Castex M, Gram L, Bentzon-Tilia M. 2017.

(Editorial)

The aquaculture microbiome at the centre of business creation

Microbial Biotechnology 10: 1279–1282.

doi:10.1111/1751-7915.12877

Editorial: The microbiome as a source of new enterprises and job creation

The aquaculture microbiome at the centre of businesscreation

Karen K. Dittmann,1 Bastian B. Rasmussen,1 MathieuCastex,2 Lone Gram1 and Mikkel Bentzon-Tilia11Department of Biotechnology and Biomedicine,Technical University of Denmark, Kgs. Lyngby,Denmark.2Lallemand SAS, Blagnac Cedex, France.

Twelve per cent of the world’s population is currentlysecuring their livelihood partly, or fully, through the fish-eries and aquaculture sector (FAO Fisheries and Aqua-culture Department, 2016). Most people occupied in thissector rely on wild catches; however, fish stocks arebecoming depleted with 90% of stocks being fully oroverexploited (FAO Fisheries and Aquaculture Depart-ment, 2016). A more productive and sustainable aquacul-ture sector is needed to meet the sustainabledevelopment goals (SDGs) of the UN number 2, 12 and14 and supply a growing world population, which isexpected to reach 1010 individuals in approximately30 years (United Nations, Department of Economic andSocial Affairs, Population Division, 2015), with high-qual-ity protein. The aquaculture sector has, within the pastfew years, surpassed wild catches in the production ofseafood (fish and plants combined; Bentzon-Tilia et al.,2016), and overall employment in the fisheries sector hasdecreased by approximately one million individuals from2010 to 2014, while the aquaculture sector saw anincrease of 0.1 million individuals. In general, a shift hasbeen seen from 1990, where 83% were employed in fish-eries and 17% in aquaculture, to 2014 where 67% wereemployed in fisheries and 33% in aquaculture (FAO Fish-eries and Aquaculture Department, 2016). The sector isprojected to increase its output from 74 million tons in2014 to 102 million tons by 2025, and up to 121 milliontons by 2030 (FAO Fisheries and Aquaculture Depart-ment, 2016). Furthermore, it was recently suggested thatthe global biological production potential for marine aqua-culture is more than 100 times the current global seafoodconsumption, thus suitable habitats do not seem to be alimiting factor in the growth of the sector (Gentry et al.,2017). Consequently, the industry is faced with a need tosignificantly increase productivity while at the same timesecuring both livelihoods and sustainability.

Controlling the microorganisms that are associatedwith aquaculture systems (i.e. the aquaculture micro-biome) has always been essential in high-intensity rear-ing of fish. Disease outbreaks caused by pathogenicbacteria are believed to be one of the most serious chal-lenges faced by the aquaculture industry (Meyer, 1991),and consequently, extensive measures are taken to limitthe introduction and proliferation of such bacteria in theaquaculture systems. Furthermore, microbial activity inthese naturally eutrophied systems may produceunwanted toxic metabolites such as hydrogen sulphide(H2S), which is formed when microorganisms reduce sul-phate (SO4

�) in anaerobic respiration and which inter-feres with mammalian respiration. However, microbesmay also serve as a solution to an array of these verychallenges. In the agriculture industry, microbiome-basedproducts such as seed coatings that increase nutrientuptake in crops, and which antagonize plant pathogenicsoil organisms, are becoming increasingly popular toolsto improve productivity in a sustainable manner, andmicrobiome-based products may reach a market sizecomparable to that of chemical agro-chemicals within afew years (Singh, 2017). The very same technologiesthat have facilitated this development, for exampleadvances in high-throughput sequencing and syntheticbiology, have been proposed to be key in the sustain-able development of the aquaculture industry in the com-ing years as well (Bentzon-Tilia et al., 2016). However,with a few exceptions, such as studies on recirculatingaquaculture systems and fish-associated microbial com-munities (van Kessel et al., 2011; Llewellyn et al., 2014),the aquaculture microbiome has not been characterizedto the same degree as its terrestrial counterpart. In con-trast, most studies concerning the aquaculture micro-biome relies on bacterial isolation and PCR-basedapproaches. Hence, the implementation of microbiome-based products is in its infancy and many practices arestill of a ‘hope for the best’ fertilization-based nature(Moriarty, 1997), where specific functional groups of theaquaculture microbiome are enriched for by adding, forexample carbon-rich substrates. This is the case formost ‘biofloc’ approaches where molasses or an equiva-lent C-rich fertilizer is added as a means to increase the

ª 2017 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution andreproduction in any medium, provided the original work is properly cited.

bs_bs_banner

C:N ratio and induce the growth of the C-limited hetero-trophic fraction of the aquaculture microbiome, which inturn will remove toxic ammonia (NH3) from the rearingwater and form bioflocs (Bossier and Ekasari, 2017).Recirculated aquaculture systems (RAS) and biofiltershave facilitated the rearing of fish in closed systems witha minimum of water being exchanged with the surround-ing environment. This relies on the successful coloniza-tion of large-surface area structures by bacteria such asNitrosomonas spp. and Nitrospira spp. that in combina-tion convert NH3 to nitrate (NO3

�). Common for theseapproaches is that they in most cases have relied onmodulation of the existing microbiome in the system.However, applications of targeted microbiome-basedproducts containing a seeding microbial assemblage to

aid the heterotrophic assimilation of inorganic nitrogenand/or the nitrification process are now a common prac-tice in intensive tropical pond-based aquaculture sys-tems (Castex et al., 2014). In the case of RAStechnology, a similar approach to aid in the colonizationof biofilters is highly desirable as it may take up to sev-eral months to obtain an efficient microbiome, specifi-cally in marine biofilters (Manthe and Malone, 1987;Gutierrez-Wing and Malone, 2006). Seeding communi-ties of nitrifiers for pond systems are already available,for example Pond Protect by Novozymes (Table 1), andthese have been shown to mitigate increased NH3 andnitrite (NO2

�) levels in RAS systems as well (Kuhn et al.,2010). Furthermore, nitrification can be coupled with anefficient microbial denitrification process as a powerful

Table 1. Microbiome-based products for conditioning of water and pond as well as promotion of a healthy production animal microbiome (feedand feed additives).

Targetenvironment Company Product Purpose Composition Reference

Water and pond AquaInTech PRO4000X,AquaPro B,AquaPro EZ

Degrade organic matter, reduceammonia, Vibrio reduction

2 Strains of Bacillus – Bacillussubtilis, Bacillus licheniformis

1, 2, 3

Biomin Aquastar Stabilize water quality, improvepond bottom quality andsupport the gut health of fishand shrimp

Formula not publicly available 4

KeetonIndustries

Waste & SludgeReducer

Improve water and bottomquality, pathogen control

Bacillus cereus RRRL B-30535 5, 6

KeetonProbiotics

ShrimpShield,PondToss

Degrade organic sludge,improve feed efficiency

Formula not publicly available 7, 8

Lallemand Lalsea Biorem Degrade organic matter, reduceammonia, pathogen control,stabilize pH

7 specific bacterial strains 9

Novozymes Pond Plus Pathogen control,decomposition of organicsubstances

Spore forming bacteria 10

Novozymes Pond Dtox Hydrogen sulphide control Paracoccus pantotrophus 11Novozymes Pond Protect Ammonia and nitrite reduction Nitrosomonas eutropha,

Nitrobacter winogradskyi12

Gut microbiome(feed, feedadditive)

AquaInTech AquaPro F Organic matter degradation,improved digestion of feed

Five strains of bacillus combined 13

Evonik EcoBiol Improve gut health Bacillus amyloliquefaciensCECT 5940

14

Keeton Probiotics FeedTreat Degrade organic sludge andimprove feed efficiency

Formula not publicly available 15

Lallemand Bactocell� Reduce deformities across fishspecies, improve gut healthacross a range of fish andshrimp species

Pediococcus acidilactici (MA18/5M) 16, 17

Rubinum TOYOCERIN� Promote growth, increasespecimen homogeneity,improve intestinal mucosa

Bacillus cereus var. toyoi 18, 19

References: (1) http://www.aqua-in-tech.com/pro4000x.html; (2) http://www.aqua-in-tech.com/aquapro-b.html; (3) http://www.aqua-in-tech.com/aquapro-ez.html; (4) http://www.biomin.net/en/products/aquastar/; (5) http://keetonaquatics.com/beneficial-microbes/waste-and-sludge-reducer/;(6) Patent ‘US 6878373 B2’; (7) http://keetonaqua.com/products/beneficial-microbes/shrimpshield/; (8) http://keetonaqua.com/products/beneficial-microbes/pondtoss/; (9) http://lallemandanimalnutrition.com/en/asia/products/lalsea-biorem-aquaculture/; (10) http://www.syndelasia.com/aquaculture-probiotics/pond-aquaculture-probiotics-amp-water-manage-26/pond-plus_38; (11) http://ponddtox.com/; (12) http://www.syndelasia.com/aquaculture-probiotics/pond-aquaculture-probiotics-amp-water-manage-26/pond-protect_40; (13) http://www.aqua-in-tech.com/aquapro-f.html;(14) http://animal-nutrition.evonik.com/product/feed-additives/en/products/probiotics/ecobiol/pages/default.aspx; (15) http://keetonaqua.com/products/beneficial-microbes/feedtreat/; (16) http://lallemandanimalnutrition.com/en/asia/products/bactocell-2/; (17) http://www.biomar.com/en/denmark/product-and-species/pike-perch/fry_feeds/; (18) http://www.rubinum.es/en/especies/#acuicultura; (19) http://www.rubinum.es/en/productos/.

ª 2017 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology., MicrobialBiotechnology, 10, 1279–1282

1280 Editorial

tool in the complete removal of nitrogenous compoundsfrom the system, and the development and application ofa joined nitrification and denitrification approach for recir-culated aquaculture systems, similar to the Aqua Scien-ce� concept from Camanor, likely represents an area ofpotential business development. The commercializationof targeted microbiome-based products containing livingmicroorganisms, such as seeding microbial assemblagesthat improve water quality, has been seen for use inaquaria for decades, for example the BIO-Spira productfrom MarineLand Labs and its predecessors, which likePond Protect and similar microbiome-based products foraquaculture systems contain bacterial assemblages thatremove ammonia and nitrite. Similar microbiome-basedproducts for use in conjunction with biofloc technologyare also available now. One such product is Shrimp-Shield by Keeton Probiotics, which facilitates biofloc for-mation, degradation of sludge as well as microbialremoval of NH3 and NO2

� (Table 1). Hence, such micro-biome-based products aim to improve water quality andin some cases remove potential pathogens through, forexample, competitive exclusion (Table 1).Another category of microbiome-based products that

is being developed for the aquaculture industry targetsthe gut of the animal directly (Table 1), equivalent to themore conventional probiotics for livestock and humanconsumption. Microbial strains evaluated as probioticsfor aquaculture are from many phylogenetic lineages;however, most of them belong to two bacterial phyla, theFirmicutes (e.g. Bacillus spp., Lactobacillus spp., Lacto-coccus spp. and Carnobacterium spp.) and the Pro-teobacteria (e.g. Vibrio spp., Pseudomonas spp. andShewanella spp.), while yeasts are rarely studied (Gate-soupe, 2007). The majority of the commercially availableprobiotic feed and feed additives for aquaculture arebased on pure or mixed cultures of lactic acid bacteriaand Bacilli (Merrifield et al., 2010; Castex et al., 2014).This includes Bactocell� (Lallemand; Table 1), which isbased on a Pediococcus acidilactici strain and is, to thebest of our knowledge, the only probiotic registered inEurope for use in aquaculture feed. These bacteria areusually well studied and well known for their positiveeffect on the human and animal gut microbiome (Cutting,2011). Furthermore, they are Generally Regarded AsSafe (GRAS) or Qualified Presumption of Safety (QPS),which makes it easier to obtain authorization for theiruse in food and feed products. A natural extension ofthis type of microbiome-based products, and a potentialnew avenue to be explored in aquaculture microbiomebusiness creation, is the controlled colonization of thereared fish from larvae to adult by a microbiome that hasthe desired functional traits and can act as an infectionbarrier against pathogens and prevent major economic

losses by crashes in the population (De Schryver andVadstein, 2014).The successful application of probiotic Firmicutes,

originally applied as probiotics for humans or livestock,in aquaculture is fortunate considering the divergentniches in which these probiotics need to establish them-selves and function. An avenue of potential new enter-prises is to develop similar products based on bacteriaof marine origin instead. Marine bacteria including mem-bers of the Roseobacter group and the Vibrio and She-wanella genera have been studied extensively for theirprobiotic potential (Austin et al., 1995; Ringø and Vad-stein, 1998; D�ıaz-Rosales et al., 2009; D’Alvise et al.,2012; Lobo et al., 2014; Grotkjær et al., 2016; Bentzon-Tilia and Gram, 2017). Furthermore, these are oftenfound as part of the indigenous microbiome of marineeukaryotes, and although their application as probioticshas been proposed, they have not yet reached a com-mercialization stage. To succeed with this approach,much more thorough characterizations of aquacultureand marine host microbiomes are needed. Furthermore,in most cases, the putative probiotic candidates reportedin scientific publications do not go on to commercializa-tion and industrial application. Getting a probiont to thecommercial market requires many additional stepsincluding assessments of safety, scale-up efficacy, pro-duction scale-up and pre-market registration. Consis-tency, efficiency and most importantly safety are keypoints in all large-scale productions, and they should beconsidered from the early stages of the discovery phaseto the final application in feed products. Thus, not onlydoes the development of a commercial product rely onsubstantial financial investments, but also on the contri-bution from a multidisciplinary team encompassing closecollaborations between scientists, aquaculture experts,fermentation engineers and regulatory personnel. Thelatter part of the team is important for success in a regu-latory landscape which varies from an absence of regu-lation in certain countries to a rigid regulatory frameworknot always adapted to the effect a probiotic can display.Despite these challenges, the aquaculture industry hasalready embraced the industrial application of micro-biome-based products for the last two decades, and thishas truly created a vast range of new enterprises espe-cially in South East Asia, Central and South Americaand more recently in Europe.Using microbiome-based products also requires devel-

opments of production, packaging and distribution tech-nology. One must consider that the efficiency of suchproducts only in part depends on the choice of themicrobial strains that compose it (selection), but also onthe way the product is produced, conditioned and finallypackaged to withstand a variety of storage conditions.

ª 2017 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology., MicrobialBiotechnology, 10, 1279–1282

Editorial 1281

In conclusion, the aquaculture industry is one of thefastest growing food producing sectors in the world andthe increased productivity of this sector is essential forthe fulfilment of the sustainable development goals ofthe UN. Microbiome-based products for application inindustrial aquaculture are today a reality, but the fullpotential is far from exploited. Despite decades of experi-ence and an increasing number of microbial biotechno-logical products, there is a large innovation potential;from the discovery of new probionts of marine origin andlarge-scale cultivation strategies to manoeuvering thepolitical, regulatory landscape and disseminating the useof probiotics to ensure future, sustainable technologiesfor high-quality protein production.

References

Austin, B., Stuckey, L. F., Robertson, P. A. W., Effendi, I.,and Griffith, D. R. W. (1995) A probiotic strain of Vibrioalginolyticus effective in reducing diseases caused byAeromonas salmonicida, Vibrio anguillarum and Vibrioordalii. J Fish Dis 18: 93–96.

Bentzon-Tilia, M., and Gram, L. (2017) BiotechnologicalApplications of the Roseobacter Clade. In Biotechnologi-cal Applications of the Roseobacter Clade. Paterson, R.,and Lima, N. (eds). Basel: Springer International Publish-ing, pp. 137–166.

Bentzon-Tilia, M., Sonnenschein, E. C., and Gram, L. (2016)Monitoring and managing microbes in aquaculture – towardsa sustainable industry.Microb Biotechnol 9: 576–584.

Bossier, P., and Ekasari, J. (2017) Biofloc technology appli-cation in aquaculture to support sustainable developmentgoals. Microb Biotechnol 10: 1012–1016.

Castex, M., Durand, H., and Okeke, B. (2014) Issues withindustrial probiotics scale-up. Aquacult Nutr Gut HealthProbiot Prebiot. Merrifield, D. L., and Ringo, E. (Eds).Wiley-Blackwell. 347–359.

Cutting, S. M. (2011) Bacillus probiotics. Food Microbiol 28:214–220.

D’Alvise, P. W., Lillebø, S., Prol-Garcia, M. J., Wergeland,H. I., Nielsen, K. F., Bergh, Ø., and Gram, L. (2012)Phaeobacter gallaeciensis reduces Vibrio anguillarum incultures of microalgae and rotifers, and prevents vibriosisin cod larvae. PLoS ONE 7: e43996.

De Schryver, P., and Vadstein, O. (2014) Ecological theoryas a foundation to control pathogenic invasion in aquacul-ture. ISME J 8: 2360–2368.

D�ıaz-Rosales, P., Arijo, S., Chabrill�on, M., Alarc�on, F. J., Tapia-Paniagua, S. T., Mart�ınez-Manzanares, E., et al. (2009)Effects of two closely related probiotics on respiratory burstactivity of Senegalese sole (Solea senegalensis, Kaup)phagocytes, and protection against Photobacterium damse-lae subsp. piscicida. Aquaculture 293: 16–21.

Food and Agriculture Organization of the United Nations(FAO), Fisheries and Aquaculture Department (2016) TheState of World Fisheries and Aquaculture 2016.

Gatesoupe, F. J. (2007) Live yeasts in the gut: naturaloccurrence, dietary introduction, and their effects on fishhealth and development. Aquaculture 267: 20–30.

Gentry, R. R., Froehlich, H. E., Grimm, D., Kareiva, P.,Parke, M., Rust, M., et al. (2017) Mapping the globalpotential for marine aquaculture. Nat Ecol Evol 1: 1317–1324.

Grotkjær, T., Bentzon-Tilia, M., D’Alvise, P. W., Dourala, N.,Nielsen, K. F., and Gram, L. (2016) Isolation of TDA-pro-ducing Phaeobacter strains from sea bass larval rearingunits and their probiotic effect against pathogenic Vibriospp. in Artemia cultures. Syst Appl Microbiol 39: 180–188.

Gutierrez-Wing, M. T., and Malone, R. F. (2006) Biological filtersin aquaculture: trends and research directions for freshwaterand marine applications. Aquacult Eng 34: 163–171.

van Kessel, M. A., Dutilh, B. E., Neveling, K., Kwint, M. P.,Veltman, J. A., Flik, G., et al. (2011) Pyrosequencing of16S rRNA gene amplicons to study the microbiota in thegastrointestinal tract of carp (Cyprinus carpio L.). AMBExpress 1: 41.

Kuhn, D. D., Drahos, D. D., Marsh, L., and Flick, G. J.(2010) Evaluation of nitrifying bacteria product to improvenitrification efficacy in recirculating aquaculture systems.Aquacult Eng 43: 78–82.

Llewellyn, M. S., Boutin, S., Hoseinifar, S. H., and Derome,N. (2014) Teleost microbiomes: the state of the art in theircharacterization, manipulation and importance in aquacul-ture and fisheries. Front Microbiol 5: 207.

Lobo, C., Moreno-Ventas, X., Tapia-Paniagua, S.,Rodr�ıguez, C., Mori~nigo, M. A., and de La Banda, I. G.(2014) Dietary probiotic supplementation (Shewanellaputrefaciens Pdp11) modulates gut microbiota and pro-motes growth and condition in Senegalese sole larvicul-ture. Fish Physiol Biochem 40: 295–309.

Manthe, D. P., and Malone, R. F. (1987) Chemical additionfor accelerated biological filter acclimation in closed bluecrab shedding systems. Aquacult Eng 6: 227–236.

Merrifield, D. L., Dimitroglou, A., Foey, A., Davies, S. J.,Baker, R. T. M., Bøgwald, J., et al. (2010) The currentstatus and future focus of probiotic and prebiotic applica-tions for salmonids. Aquaculture 302: 1–18.

Meyer, F. P. (1991) Aquaculture disease and health man-agement. J Anim Sci 69: 4201.

Moriarty, D. J. W. (1997) The role of microorganisms inaquaculture ponds. Aquaculture 151: 333–349.

Ringø, E., and Vadstein, O. (1998) Colonization of Vibriopelagius and Aeromonas caviae in early developing turbot(Scophthalmus maximus L.) larvae. J Appl Microbiol 84:227–233.

Singh, B. K. (2017) Creating new business, economicgrowth and regional prosperity through microbiome-basedproducts in the agriculture industry. Microb Biotechnol 10:224–227.

United Nations, Department of Economic and Social Affairs,Population Division (2015) World Population Prospects:The 2015 Revision, Key Findings and Advance Tables.ESA/P/WP.241.

ª 2017 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology., MicrobialBiotechnology, 10, 1279–1282

1282 Editorial

Paper 2

Rasmussen BB, Erner KE, Bentzon-Tilia M, Gram L. 2018.

Effect of TDA-producing Phaeobacter inhibens on the fish pathogen Vibrio

anguillarum in non-axenic algae and copepod systems.

Microbial Biotechnology 11: 1070-1079.

Effect of TDA-producing Phaeobacter inhibens on thefish pathogen Vibrio anguillarum in non-axenic algaeand copepod systems

Bastian Barker Rasmussen, Katrine Ege Erner,Mikkel Bentzon-Tilia and Lone Gram*Department of Biotechnology and Biomedicine,Technical University of Denmark, Anker Engelundsvejbldg. 301, DK-2800 Kgs. Lyngby, Denmark.

Summary

The expanding aquaculture industry plays an impor-tant role in feeding the growing human populationand with the expansion, sustainable bacterial dis-ease control, such as probiotics, becomes increas-ingly important. Tropodithietic acid (TDA)-producingPhaeobacter spp. can protect live feed, for examplerotifers and Artemia as well as larvae of turbot andcod against pathogenic vibrios. Here, we show thatthe emerging live feed, copepods, is unaffected bycolonization of the fish pathogen Vibrio anguillarum,making them potential infection vectors. However,TDA-producing Phaeobacter inhibens was able tosignificantly inhibit V. anguillarum in non-axenic cul-tures of copepod Acartia tonsa and the copepodfeed Rhodomonas salina. Vibrio grew to106 CFU ml�1 and 107 CFU ml�1 in copepod andR. salina cultures, respectively. However, vibriocounts remained at the inoculum level(104 CFU ml�1) when P. inhibens was also added.We further developed a semi-strain-specific qPCRfor V. anguillarum to detect and quantify the patho-gen in non-axenic systems. In conclusion, P. in-hibens efficiently inhibits the fish larval pathogenV. anguillarum in the emerging live feed, copepods,supporting its use as a probiotic in aquaculture. Fur-thermore, qPCR provides an effective method for

detecting vibrio pathogens in complex non-axeniclive feed systems.

Introduction

Aquaculture provides high-quality protein for approxi-mately three billion people worldwide and with the esti-mated increase in the human population combined withoverfishing, this number is likely to increase (WWF andZSL, 2015; FAO, 2016). Larvae of several commerciallyimportant species including cod, halibut and turbot arefed using live feed due to the lack of suited artificial feedformulations. Traditionally, marine fish larvae have beenfed using rotifers and Artemia, although copepods aremore representative of the natural diet (Turner, 2004).Copepods have several advantages over traditionalaquaculture live feed including desirable amino acid pro-files, a high fatty acid content and swimming patternsthat promote feeding in fish larvae (Støttrup et al., 1986;Støttrup and Norsker, 1997; Bell et al., 2003; Turinganet al., 2007). Furthermore, substituting rotifers with cope-pods (Acartia tonsa) in the initial larval feeding stage hasa long-term positive effect on survival, growth and viabil-ity of Atlantic cod and ballan wrasse fish larvae (Øieet al., 2017). So far, the use of copepods as live feed inlarval rearing has not been cost-effective; however, newapproaches to copepod breeding are improving the costbalance (Abate et al., 2016), thus making copepods arelevant live feed in commercial aquaculture.A major bottleneck in aquaculture is disease outbreaks

where bacterial infections account for approx. 55%(Kibenge et al., 2012). One of the most prominent bacte-rial genera causing disease in aquaculture is Vibrioincluding Vibrio anguillarum, V. harveyi, V. vulnificus andV. splendidus (Thompson et al., 2004; Toranzo et al.,2005). Live feed used in aquaculture is a potential pointof entry and vector for pathogenic bacteria (Olafsen,2001), and several Vibrio species, including the humanand fish pathogens V. cholerae, V. vulnificus andV. splendidus, are found in association with zooplankton,for example copepods, which in the marine environmentseem to serve as natural Vibrio reservoirs (Sochardet al., 1979; Heidelberg et al., 2002; Colwell et al., 2003;Vezzulli et al., 2015). The role of copepods as hosts forV. cholerae has especially been studied extensively, and

Received 23 February, 2018; revised 27 March, 2018; accepted 8April, 2018.*For correspondence. E-mail gram@bio.dtu.dk; Tel. +45 4525 2586;Fax: +45 4525 2501.Microbial Biotechnology (2018) 11(6), 1070–1079doi:10.1111/1751-7915.13275Funding InformationThis study was funded by grants from The Danish Council forStrategic Research|Programme Commission on Health, Food andWelfare (12-132390; ProAqua) and by grant VKR023285 from theVillum Fonden.

ª 2018 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution andreproduction in any medium, provided the original work is properly cited.

bs_bs_banner

V. cholerae has repeatedly been isolated from the sur-face, gut and exuviae of these small crustaceans (Huqet al., 1983; Tamplin et al., 1990; Gugliandolo et al.,2008). Thus, although vibrios such as Vibrio para-haemolyticus, V. alginolyticus and V. mimicus have beenreported to be unable to colonize the copepod speciesTemora stylifera, Acartia clausi, Centropages typicusand Paracalanus parvus (Dumontet et al., 1996), it is avalid concern that Acartia tonsa may function as infec-tion vectors for other Vibrio species including the fishpathogenic V. anguillarum in aquaculture systems.Historically, antibiotics have been used to control bac-

terial infections in aquaculture. However, risk of antibioticresistance development and the associated health riskhave led to a search for sustainable alternatives(Skjermo and Vadstein, 1999; Sommerset et al., 2005;Defoirdt et al., 2011). The deployment of vaccines haslimited the use of antibiotics especially in Europe andNorth America (Defoirdt et al., 2011; Ringø et al., 2014).However, antibiotics are still used in many countries andfish larviculture as their undeveloped immune systemdoes not allow for vaccination (Sommerset et al., 2005;Defoirdt et al., 2011). An alternative to antibiotics in suchsystems is probiotics, that is live microorganisms whichexert beneficial effects on the host health (FAO andWHO, 2001), for example by inhibiting growth of patho-genic bacteria (Gatesoupe, 1999; Skjermo and Vadstein,1999; Kesarcodi-Watson et al., 2008). Members of thegenera Phaeobacter and Ruegeria which produce thebroad-spectrum antimicrobial compound tropodithieticacid (TDA) have proven probiotic properties (Ruiz-Ponteet al., 1999; Hjelm et al., 2004; Planas et al., 2006; D’Al-vise et al., 2012). Phaeobacter spp. inhibit the growth ofpathogenic vibrios and protect live feed such as rotifersand Artemia (D’Alvise et al., 2012; Grotkjær et al.,2016a,b). Furthermore, Phaeobacter spp. can protect lar-vae of turbot and cod in challenge trials (Planas et al.,2006; D’Alvise et al., 2012, 2013). Phaeobacter spp.have repeatedly been isolated from different aquaculturefacilities (Ruiz-Ponte et al., 1998; Hjelm et al., 2004;Porsby et al., 2008; Gram et al., 2015; Grotkjær et al.,2016b), which indicate that they are part of the naturalmicrobiota in these aquaculture units. Most studies onthe probiotic effect of Phaeobacter spp. and Ruegeriaspp. have been carried out in axenic live feed or fish lar-vae systems (Planas et al., 2006; D’Alvise et al., 2012,2013; Grotkjær et al., 2016b). Although non-axenic algaeand Artemia systems have been tested (Grotkjær et al.,2016a), the probiotic effect of Phaeobacter spp. andRuegeria spp. in these more ‘natural’ systems requiresfurther studies.The purpose of this study was to investigate whether

the copepod species Acartia tonsa would potentiallyfunction as an infection vector for fish pathogenic vibrio,

and whether the TDA-producing P. inhibens could antag-onize V. anguillarum in non-axenic copepod systems aswell as in algae used as copepod feed. Furthermore, weaimed at developing a strain-specific quantitative PCR(qPCR) protocol for detection and quantification of patho-gens in complex systems without the dependency onselection markers, such as antibiotic resistance thatpotentially could affect pathogen behaviour.

Results

Invasion of copepods by GFP-tagged V. anguillarum

To test whether V. anguillarum could invade the cope-pod Acartia tonsa and make the crustacean a potentialinfection vector in larviculture, a GFP-tagged V. anguil-larum strain NB10 was added to non-axenic A. tonsafeed, that is Rhodomonas salina. After 4 days of incuba-tion, the GFP signal of NB10 was detected on the sur-face and inside the copepods (Fig. 1). Specifically, thevibrios appeared to be concentrated within the intestinalsystem of the crustacean. No GFP signal was detectedin non-inoculated copepods. Throughout the experiment,Vibrio abundances in the surrounding medium werestable around the inoculum level of 105 CFU ml�1 (datanot shown). Furthermore, A. tonsa seemed unaffectedby the addition of vibrios, exhibiting mortalities after4 days of 1.47 � 0.72%.

Co-culture in non-axenic Rhodomonas salina

As A. tonsa was found to be a potential vector for V. an-guillarum, we investigated whether P. inhibens couldantagonize vibrios in cultures of copepod and in thecopepod live feed R. salina. We conducted a co-cultureexperiment in non-axenic R. salina cultures. Counts ofR. salina were stable during the co-culture experiment,and no significant difference was seen between the dif-ferently treated cultures (data not shown). Phaeobacterinhibens established itself and remained at approx.107 CFU ml�1 in all set-ups, independently of addition ofV. anguillarum strain 90-11-286 (Fig. 2). TSA countsshowed that V. anguillarum was significantly reducedwhen grown in the presence of P. inhibens. WithoutP. inhibens, V. anguillarum grew to approx. 107 CFUml�1, while in the presence of P. inhibens, V. anguil-larum numbers were significantly reduced with approxi-mately three orders of magnitude (P < 0.0001) andnever exceeded the inoculum level of 104 CFU ml�1.

Co-culture in non-axenic A. tonsa, and detection andquantification of V. anguillarum with qPCR

Non-axenic A. tonsa were not affected by the presenceof either the probiont or the pathogen resulting in

ª 2018 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology, MicrobialBiotechnology, 11, 1070–1079

Phaeobacter–Vibrio interactions in copepods 1071

mortalities after 3 days of 5.7 � 2.52%, 7.3 � 1.15%,5.7 � 2.08% and 6.3 � 3.51% in cultures with P. in-hibens, V. anguillarum, a combination of P. inhibens andV. anguillarum, or no addition (control), respectively. Asfor the R. salina co-culture experiment, no differenceswere observed between P. inhibens abundances in thepresence or absence of V. anguillarum (Fig. 3) withcounts being stable around the inoculum level of 106–107 CFU ml�1. The abundance of V. anguillarum wassignificantly reduced by P. inhibens (P = 0.0004). Withthe addition of P. inhibens, Vibrio counts remained atthe inoculum level around 104 CFU ml�1, approx. twoorders of magnitude below the abundances observed incultures without P. inhibens. Numbers of V. anguillarumwere estimated both by counts on TSA plates and byqPCR resulting in similar abundances (P > 0.05;Table 1).

Verification of primer specificity and standard curve

Primer-BLAST was used to design and test the speci-ficity of the V. anguillarum 90-11-286 primers in silico,and this indicated that the primers were strain-specific.The primers were subsequently tested in vitro against 11V. anguillarum strains and one Vibrio harveyi strain(Table 2). The primers amplified the DNA of three strainsresulting in cycle threshold (Ct) values within the limits ofdetection (Ct ≤ 30). Furthermore, one strain, NB10, wasclose to the limit of detection with a Ct value of30.29 � 0.13. Strains 90-11-286 and PF7 showed simi-lar Ct values of 19.56 � 0.05 and 19.16 � 0.11, respec-tively, and a higher Ct value of 23.53 � 0.08 was seenfor PF4. For quantification, a standard curve was madewith strain 90-11-286 overnight culture that was 10-foldserial diluted from approx. 107 to 103 (Fig. 4). The

(A) (B)

Fig. 1. Microscopy of A. tonsa nauplii colonized by GFP-tagged V. anguillarum NB10. Red fluorescent cells are R. salina. Magnification 100 9

A. Phase-contrast microscopy. B. Fluorescence microscopy (WIB excitation 460–490 nm, emission > 515 nm).

Fig. 2. Colony counts of V. anguillarum 90-11-286 and P. inhibensDSM 17395 during a co-culture experiment in non-axenic R. salinacultures. V. anguillarum without P. inhibens (■, solid line), V. anguil-larum with P. inhibens (▼, solid line), P. inhibens without V. anguil-larum (●, dashed line) and P. inhibens with V. anguillarum (▲,dashed line). The two P. inhibens lines are situated on top of oneanother. Error bars show the standard deviation of three biologicalreplicates.

Fig. 3. Colony counts of V. anguillarum 90-11-286 and P. inhibensDSM 17395 during co-culture in non-axenic A. tonsa cultures.V. anguillarum without P. inhibens (■, solid line), V. anguillarumwith P. inhibens (▼, solid line), P. inhibens without V. anguillarum(●, dashed line) and P. inhibens with V. anguillarum (▲, dashedline). Error bars show the standard deviation of three biologicalreplicates.

ª 2018 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology, MicrobialBiotechnology, 11, 1070–1079

1072 B. B. Rasmussen, K. E. Erner, M. Bentzon-Tilia and L. Gram

standard curve was used to compare Ct values to bacte-rial density (CFU ml�1) for Vibrio samples from the non-axenic A. tonsa co-culture.

Discussion

Live feed is essential at the larval stage in several com-mercially important species of finfish, and controllingpathogenic bacteria in the live feed is important as thesemay act as vectors for the pathogens. Here, we demon-strate that P. inhibens can inhibit the highly virulentV. anguillarum strain 90-11-286 in two different non-axe-nic live feed systems. Specifically, we demonstrate theprobiotic potential in copepods as these are superior aslarval feed compared to Artemia and rotifers, and futureaquaculture is expected to rely increasingly on a cope-pod-based live feed.Live feed is known as a point of entry and vector for

pathogenic bacteria (Olafsen, 2001). Copepods havepreviously been shown to host some Vibrio spp. (Huqet al., 1983; Tamplin et al., 1990; Heidelberg et al.,2002; Gugliandolo et al., 2008; Vezzulli et al., 2015);however, some Vibrio spp. seem to be unable to colo-nize some copepod species (Dumontet et al., 1996).Using a GFP-tagged V. anguillarum NB10, we found thatV. anguillarum was able to colonize the outside and the

gut of the feeding copepods without inducing notablemortality (1.47 � 0.72% after 4 days) hence makingthem potential infection vectors in larviculture.Vibrio spp. are part of the established microbiota of

live feed used in larviculture (Berland et al., 1970; Austinand Allen, 1982; Salvesen et al., 2000; L�opez-Torresand Liz�arraga-Partida, 2001) and when introduced, these

Table 1. Quantification of V. anguillarum strain 90-11-286 using colony counts and qPCR. Values are presented as log10-transformed counts.Standard deviations are based on three biological replicates. P values were calculated using t-test (alpha = 0.05) and describe differencesbetween TSA plate counts and Ct based counts.

Time (days)

V. anguillarum counts without P. inhibens V. anguillarum counts with P. inhibens

TSA Ct based P value TSA Ct based P value

0 3.76 � 0.10 3.82 � 0.14 0.6114 3.83 � 0.20 3.70 � 0.16 0.44161 5.57 � 0.11 5.70 � 0.08 0.1657 3.98 � 0.20 3.93 � 0.04 0.70952 5.71 � 0.13 5.72 � 0.12 0.9621 4.06 � 0.03 3.83 � 0.18 0.10163 5.84 � 0.17 5.88 � 0.17 0.7626 4.28 � 0.18 3.85 � 0.21 0.0561

Table 2. Vibrio strains used to test the specificity of the V. anguillarum strain 90-11-286 primers. Standard deviations are based on technicalduplicates.

Strain Species

Isolation

Serotype Plasmid pJM1

Accession no.

Mean Ct -valueHost Country Genome Plasmid

4299 V. anguillarum Unknown Norway O2b � CP011458/CP011459 35.57 � 1.5190-11-286 Rainbow trout Denmark O1 � CP011460/CP011461 19.56 � 0.0590-11-287 Rainbow trout Denmark O1 + CP011475/CP011476 CP016254 32.31 � 0.02775 Coho salmon Norway O1 + CP002284/CP002285 AY312585 33.41 � 0.78H610 Atlantic cod USA O2a � CP011462/CP011463 37.06 � 2.11NB10 Unknown Sweden O1 + LK021130/LK021129 LK021128 30.29 � 0.13PF4 Atlantic salmon Chile O3 � CP010080/CP010081 23.53 � 0.08PF7 Atlantic salmon Chile O3 � CP011464/CP011465 19.16 � 0.11VIB243 Sockeye salmon USA O1/VaNT1 + CP010030/CP010031 CP016261 35.18 � 0.30DSM21597 Atlantic cod Norway O2 � CP010084/CP010085 34.31 � 0.76S2 2/9 Rainbow trout Denmark O1 � CP011472/CP011473 36.61 � 2.17DSM19623 V. harveyi Amphipod USA – � BAOD01000001 39.58

Fig. 4. qPCR standard curves for V. anguillarum 90-11-286 in non-axenic copepod cultures. Standard curve was obtained by running,in triplicates, DNA extracted from one sample.

ª 2018 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology, MicrobialBiotechnology, 11, 1070–1079

Phaeobacter–Vibrio interactions in copepods 1073

opportunistic pathogens can proliferate rapidly leading togreat losses (Reid et al., 2009). In the co-culture experi-ment, no differences were seen in mortality of the unfedcopepods regardless of the presence of P. inhibens orV. anguillarum (5.7 � 2.52% to 7.3 � 1.15% mortality).In contrast, V. anguillarum has been shown to causehigh mortality (92%) in non-axenic Artemia cultures.Here, the addition of P. inhibens reduced mortality of theinfected Artemia to 11% (Grotkjær et al., 2016b). Theobserved differences in mortality potentially give cope-pods an advantage over Artemia as live feed, as it hasbeen reported that homogenized Artemia or sea basslarval biomass, simulating the presence of dead Artemiaor sea bass larvae, increase the virulence of V. anguil-larum, resulting in a significant increase in mortality inchallenged sea bass larvae (Li et al., 2014). Thus, cope-pods with a higher survival rate than Artemia could resultin fewer and potentially less virulent vibrios. The differ-ences in mortality between Artemia and copepods canbe explained by several factors. A notable differencebetween the studies is the vibrio abundances observedat the end of the co-culture experiments. Although inocu-lated at same level (104 CFU ml�1), abundances ofV. anguillarum in cultures without P. inhibens were 10-fold higher in the Artemia experiment (107 CFU ml�1)(Grotkjær et al., 2016a) as compared to the copepodexperiment conducted in the present study (106 CFUml�1). This suggests that the presence of copepods, orthe background microbiota of the copepods, could havean inhibiting effect on vibrio growth, especially as vibrioabundances in the algae cultures of both studies, Tetra-selmis suecica and R. salina, respectively, were similarto that of the Artemia cultures, that is 107 CFU ml�1.The immune systems of invertebrates are poorly under-stood (Loker et al., 2004). However, there are differ-ences between copepods and Artemia, and for instance,the heat shock protein hsp70, which is associated withpathogen resistance, is induced in Artemia wheninfected with Vibrio (Norouzitallab et al., 2016). In con-trast, an analog of this stress responder was not inducedin the copepod Eurytemora affinis when exposed to Vib-rio spp. (Almada and Tarrant, 2016). Different copepodspecies are colonized by different bacteria, and someVibrio spp. are unable to colonize some copepod spe-cies (Dumontet et al., 1996). Vibrios colonizing cope-pods have been reported to trigger upregulation ofsaposin-like genes and C-type lectins (Almada and Tar-rant, 2016) that have antimicrobial properties (Hoeck-endorf and Leippe, 2012; Wang et al., 2014). Thissuggests that copepods have a potential inhibiting effecton the growth of V. anguillarum, which could explain thelower Vibrio abundance and mortality in copepod cul-tures relative to Artemia cultures. It is also possible thatthe V. anguillarum strain 90-11-286 is avirulent against

A. tonsa. The virulence of strains within the same spe-cies can vary immensely as seen for thirty V. anguil-larum strains tested in three different hosts (turbot,halibut and cod larvae), causing mortality from 100% to9.1% with the same infection dose (Rønneseth et al.,2017). Thus, if other V. anguillarum strains had beentested in the Artemia and A. tonsa systems, the outcomemight have been different.Several co-culture studies of P. inhibens and V. an-

guillarum in axenic systems have shown significant inhi-bition of vibrios, resulting in lowered vibrio countsrelative to the inoculum level (D’Alvise et al., 2010,2012; Grotkjær et al., 2016b). In contrast, P. inhibenskept V. anguillarum at the inoculum level in both thenon-axenic co-culture experiments conducted in the pre-sent study and in similar experiments with non-axenicTetraselmis and Artemia (Grotkjær et al., 2016a). Thissuggests that the microbiota of the non-axenic systemshave a protective effect on V. anguillarum; however, thisis not necessarily a problem in an aquaculture contextas it is not essential to completely eradicate all vibrios toget a protective effect. Rønneseth et al. (2017) found asignificant decrease in virulence for several V. anguil-larum strains tested in turbot, halibut and cod larvaechallenge trials when added at low density (104 CFUml�1) as compared to high density (106 CFU ml�1).Hence, although V. anguillarum was not reduced tobelow inoculum levels in the presence of P. inhibens,our findings support the use of Phaeobacter spp. as pro-biotics in copepod systems. In both co-culture experi-ments, P. inhibens was able to establish itself at aconstant level, which is consistent with its reported asso-ciation with micro- and macroalgae, zooplankton, squids,copepods and fish larvae (Hjelm et al., 2004; Rao et al.,2005; Venmathi Maran et al., 2007; Porsby et al., 2008;Collins et al., 2012; Gram et al., 2015; Grotkjær et al.,2016b). This supports the proposal that one way of utiliz-ing Phaeobacter spp. in aquaculture is by introducingthem directly into the live feed systems (Planas et al.,2006; Prol et al., 2009; Grotkjær et al., 2016b).Previously, non-axenic co-culture experiments have

used a chloramphenicol-resistant V. anguillarum as tar-get organism and estimated vibro abundances on solidsubstrates supplemented with antibiotics (Grotkjær et al.,2016a). However, using antibiotic as selection marker isnot without challenges. The natural microbiome may har-bour bacteria resistant to the antibiotic marker, the antibi-otic marker may be unstable, and the marker may affectthe fitness of the tagged strain, all of which can result inover- or underestimation of the actual abundances(Andersson and Levin, 1999; Allen et al., 2010). Severalnon-growth dependent methods have been developedfor detection and/or quantification of vibrios (Eiler andBertilsson, 2006; Prol et al., 2009; Saulnier et al., 2009;

ª 2018 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology, MicrobialBiotechnology, 11, 1070–1079

1074 B. B. Rasmussen, K. E. Erner, M. Bentzon-Tilia and L. Gram

Kim and Lee, 2014). However, as previously described,vibrios are part of the established microbiota of live feed,thus genus- or species-specific primers will potentiallyamplify vibrios from the background microbiota. In thisstudy, we attempted to develop strain-specific primersfor quantification of the pathogenic V. anguillarum 90-11-286. Three of the 12 tested vibrio strains, including thetarget strain, were amplified within the limits of detection(Ct ≤ 30). The two non-target strains were the closelyrelated V. anguillarum PF7 and PF4 (Table 2). All threestrains are highly virulent and cluster closely when com-pared by 163 concatenated virulence factors and by coregenome phylogeny (Castillo et al., 2017; Rønnesethet al., 2017). However, the semi-strain-specific primersdeveloped provide an effective method for detecting fewspecific vibrios in non-axenic systems. As they weremade for research and not diagnostic purposes, it ispossible to test the applicability of the primer set prior toexperimental use.In conclusion, our study shows that the emerging live

feed copepod, A. tonsa, potentially can function as aninfection vector for pathogenic V. anguillarum strains.However, Vibrio counts in the copepod cultures werelower than what has previously been seen for Artemiaand algae cultures, supporting the use of A. tonsa asfish larvae feed. P. inhibens efficiently inhibited V. an-guillarum in both non-axenic copepod and algae culturessupporting its use as a probiotic in live feed systems.Lastly, we have described direct qPCR methodology thatprovides an effective means of detecting vibrios in com-plex non-axenic live feed systems without the use ofselection markers, which potentially could affect patho-gen behaviour.

Experimental procedures

Bacterial strain and culture conditions

Two fish pathogenic V. anguillarum strains were used inthis study. Strain NB10, used for the adhesion/invasionexperiment, was isolated from the Gulf of Bothnia and ispathogenic towards turbot and halibut larvae (Rønnesethet al., 2017). The strain has been tagged by insertion ofplasmid pNQFlaC4-gfp27 (cat, gfp) into the chromosome(Croxatto et al., 2007) and was provided by Prof. DebraL. Milton (Department of Molecular Biology, Ume�aUniversity). The highly virulent V. anguillarum strain 90-11-286 (Rønneseth et al., 2017) was used in the co-cul-ture experiments and was isolated from diseased rain-bow trout (Oncorhynchus mykiss) from a Danishaquaculture (Pedersen and Larsen, 1993; Skov et al.,1995). P. inhibens DSM 17395 was used for the co-cul-ture experiments (Martens et al., 2006; Vandecandelaereet al., 2008; Buddruhs et al., 2013). NB10 and 90-11-286 were grown and counted on TSA (Tryptone Soy

Agar, Difco 212185) supplemented with 2.5 mg l�1 chlo-ramphenicol for NB10. The vibrios have very distinct col-ony morphology on TSA. Phaeobacter inhibens DSM17395 is unable to grow on TSA due to its low salinity,thus TSA functions as a semiselective medium.Phaeobacter inhibens was grown and counted on MA(Marine Agare, Difco 2216) where it grows as distinctbrown colonies due to Fe-TDA complex (D’Alvise et al.,2016). Bacterial stock cultures were stored at �80°C in20% (vol/vol) glycerol. Two to three days prior to use,stock cultures were streaked on agar plates and incu-bated at 25°C. The purity of the bacteria was checkedby colony morphology, and single colonies were usedfor inoculation of each preculture. All bacterial precul-tures were grown in 20 ml of ½YTSS (2 g Bacto Yeastextract, 1.25 g Bacto Tryptone, 20 g Sigma Sea Salts,1 L deionized water) (Sobecky et al., 1997) at 25°C and200 rpm for 24 h.

Preparation of non-axenic algae and copepods cultures

The non-axenic R. salina K-1487 originates from Den-mark and was provided by Prof. Thomas Kiørboe(National Institute of Aquatic Resources, TechnicalUniversity of Denmark) (Nielsen and Kiørboe, 2015). AR. salina K-1487 stock culture was maintained in f/2medium (Guillard and Ryther, 1962; Guillard, 1975) with-out Na2SiO3 but with 5 mM NH4Cl in 3% IO (InstantOcean salts, Aquarium Systems Inc.) (f/2-Si+NH4) at18°C and 24 lmol photons m�2 s�1, photosyntheticallyactive radiation (PAR). The algal density was determinedby counting in a Neubauer-improved counting chamber,and the culture was adjusted to approx. 100 000 cellml�1 in f/2-Si+NH4 and 25 ml was distributed into 50-mlfalcon tubes. A. tonsa eggs were provided by Prof.Benni W. Hansen (Department of Science and Environ-ment, Roskilde University) (Drillet et al., 2006; Hansenet al., 2010) and were kept at 5°C until use. Two daysbefore the experiment, eggs were inoculated in 3% IOand incubated at 18°C and 24 lmol photons m�2 s�1,photosynthetically active radiation (PAR). For bacterialmono- and co-culture experiments, 50-ml falcon tubeswere set up with 25 ml 3% IO with 2–3 Acartia tonsanauplii per millilitre. For the invasion experiment, 50-mlfalcon tubes containing 20 ml R. salina culture were sup-plemented with approx. 5 Acartia tonsa nauplii per millil-itre.

A. tonsa invasion experiment

Precultures of V. anguillarum NB10 were 10-fold seriallydiluted in 3% IO, and 250 ll dilutions were used to inoc-ulate non-axenic cultures of A. tonsa fed R. salina, aim-ing at an initial concentration of 105 CFU ml�1.

ª 2018 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology, MicrobialBiotechnology, 11, 1070–1079

Phaeobacter–Vibrio interactions in copepods 1075

Experiments were carried out over 96 h where bacterialconcentrations and copepod mortality were determinedevery 24 h. At the end of the experiment, A. tonsa fromcultures with and without NB10 was investigated forGFP signal using phase-contrast and fluorescencemicroscopy at 1009 magnification using an OlympusBX51 fluorescence microscope (WIB excitation 460–490 nm, emission >515 nm).

Co-culture experiment in non-axenic R. salina and unfedA. tonsa cultures

Precultures of V. anguillarum 90-11-286 were 10-foldserially diluted in 3% IO, and 250 ll diluted culture wasused to inoculate non-axenic R. salina and A. tonsa cul-tures, aiming at an initial concentration of 104 CFU ml�1.Precultures of P. inhibens DSM 17395 were added undi-luted to the non-axenic R. salina and A. tonsa cultures,aiming at an initial concentration of 107 CFU ml�1.250 ll ½YTSS was added to uninoculated controls andcultures where only 90-11-286 had been added. The cul-tures were incubated at 18°C, lying horizontally on arotary shaker at 60 rpm in an algae growth cabinet.Experiments were carried out over 72 h where bacterialand algal abundances, and copepod mortality weredetermined every 24 h. Samples for DNA extraction forqPCR-based quantification of 90-11-286 were takenevery 24 h (see below). Bacterial abundances weredetermined by CFU counts. 90-11-286 counts weredetermined on TSA, and DSM 17395 counts were deter-mined on MA. Plates were incubated at 25°C. TSAplates were counted after 1 and 2 days and MA platesafter 2-3 days. R. salina abundances were counted in aNeubauer-improved counting chamber. The number ofdead A. tonsa was determined using a Sedgewick-Raftercounting chamber. The number of surviving A. tonsawas determined after the experiment was terminated. Allcombinations were made in biological triplicates.

Primer design and quantification of V. anguillarum 90-11-286 by qPCR

The primers Fw_90-11-286 (50 - CAACTTAGGCGTGCAATGGG - 30) and Rev_90-11-286 (50- ACCGCTTTACTGGTGGTGG - 30) were designed using Primer-BLASTfrom NCBI (Ye et al., 2012). Standard settings wereused except for the ‘PCR product size’ which was set tomin 75 bp and max 200 bp and the database specifica-tions which was set to Genomes and Bacteria (taxid:2).V. anguillarum 90-11-286 chromosome I, completesequence (NZ_CP011460.1) was used as template. ForqPCR detection and quantification of the pathogen,genomic DNA was extracted from 1 ml samples usingthe NucleoSpin Tissue kit (M740952; Macherey-Nagel,

D€uren, Germany) as described by the manufacturer.Extracted DNA was stored at -20° C until use. Amplifica-tion reaction mixtures contained 12.5 ll 2 9 SYBR�

Green PCR Master Mix (4309155; Applied Biosystems),1 ll (10 lM) Fw primer, 1 ll (10 lM) Rev primer, 2 lltemplate DNA and 8.5 ll H2O (DNA grade). Reactionswere run on a Mx3000P (Stratagene) qPCR System,using the program 1 cycle at 95°C for 10 min, 40 cyclesat 95°C for 15 s, 56°C for 1 min and 72°C for 1 min fol-lowed by a dissociation curve 95°C for 1 min, 55°C for30 s and 95°C for 30 s. DNA grade water was includedas non-template controls (NTC), and positive controlsconsisted of genomic DNA from the target strain. Forin vivo testing of primer specificity against 11 V. anguil-larum strains and one V. harveyi strain (Table 2), precul-tures were diluted to 107 CFU ml�1 in 3% IO from whichTSA plate counts were made and DNA was extracted.For quantification, a standard curve relating Ct (cyclethreshold) values to bacterial density (CFU ml�1) wasmade with 90-11-286 preculture. The preculture was 10-fold serial diluted in 3% IO and inoculated into non-axe-nic A. tonsa cultures aiming for 107 to 103 CFU ml�1.Vibrio counts of the cultures were made using TSAplates, and DNA was extracted from the samples. Themeasurements were analysed using GraphPad Prism7.04 (GraphPad Software, San Diego CA). The limit ofdetection was estimated based on Ct values of the con-trol samples with only A. tonsa and background micro-biota.

Statistical analysis

CFU ml�1 values and CFU ml�1 of the corresponding Ct

values for each biological replicate were log10-trans-formed prior to statistical analysis. Statistical analyses ofdifferences were performed using t-test (alpha = 0.05) inGraphPad Prism 7.04 (GraphPad Software, San DiegoCA).

Acknowledgements

This study was funded by grants from The Danish Coun-cil for Strategic Research|Programme Commission onHealth, Food and Welfare (12-132390; ProAqua) and bygrant VKR023285 from the Villum Fonden.

Conflict of interest

The authors declare no conflict of interest.

References

Abate, T.G., Nielsen, R., Nielsen, M., Jepsen, P.M., andHansen, B.W. (2016) A cost-effectiveness analysis of live

ª 2018 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology, MicrobialBiotechnology, 11, 1070–1079

1076 B. B. Rasmussen, K. E. Erner, M. Bentzon-Tilia and L. Gram

feeds in juvenile turbot Scophthalmus maximus (Lin-naeus, 1758) farming: copepods versus Artemia. AquacNutr 22: 899–910.

Allen, H.K., Donato, J., Wang, H.H., Cloud-Hansen, K.A.,Davies, J., and Handelsman, J. (2010) Call of the wild:antibiotic resistance genes in natural environments. NatRev Microbiol 8: 251–259.

Almada, A.A., and Tarrant, A.M. (2016) Vibrio elicits tar-geted transcriptional responses from copepod hosts.FEMS Microbiol Ecol 92: 1–11.

Andersson, D.I., and Levin, B.R. (1999) The biological costof antibiotic resistance. Curr Opin Microbiol 2: 489–493.

Austin, B., and Allen, D.A. (1982) Microbiology of labora-tory-hatched brine shrimp (Artemia). Aquaculture 26:369–383.

Bell, J.G., McEvoy, L.A., Estevez, A., Shields, R.J., andSargent, J.R. (2003) Optimising lipid nutrition in first-feed-ing flatfish larvae. Aquaculture 227: 211–220.

Berland, B.R., Bonin, D.J., and Maestrini, S.Y. (1970) Studyof bacteria associated with marine algae in culture. MarBiol 3: 68–76.

Buddruhs, N., Pradella, S., G€oker, M., P€auker, O., Pukall,R., Spr€oer, C., et al. (2013) Molecular and phenotypicanalyses reveal the non-identity of the Phaeobacter gal-laeciensis type strain deposits CIP 105210T and DSM17395. Int J Syst Evol Microbiol 63: 4340–4349.

Castillo, D., Alvise, P.D., Xu, R., Zhang, F., and Middelboe,M. (2017) Comparative genome analyses of Vibrio anguil-larum strains reveal a link with pathogenicity traits. mSys-tems 2: 1–14.

Collins, A.J., LaBarre, B.A., Won, B.S.W., Shah, M.V.,Heng, S., Choudhury, M.H., et al. (2012) Diversity andpartitioning of bacterial populations within the accessorynidamental gland of the squid Euprymna scolopes. ApplEnviron Microbiol 78: 4200–4208.

Colwell, R.R., Huq, A., Islam, M.S., Aziz, K.M.A., Yunus, M.,Khan, N.H., et al. (2003) Reduction of cholera in Bangla-deshi villages by simple filtration. Proc Natl Acad Sci USA100: 1051–1055.

Croxatto, A., Lauritz, J., Chen, C., and Milton, D.L. (2007)Vibrio anguillarum colonization of rainbow trout integu-ment requires a DNA locus involved in exopolysaccharidetransport and biosynthesis. Environ Microbiol 9: 370–382.

D’Alvise, P.W., Melchiorsen, J., Porsby, C.H., Nielsen, K.F.,and Gram, L. (2010) Inactivation of Vibrio anguillarum byattached and planktonic Roseobacter cells. Appl EnvironMicrobiol 76: 2366–2370.

D’Alvise, P.W., Lillebø, S., Prol-Garcia, M.J., Wergeland,H.I., Nielsen, K.F., Bergh, Ø., and Gram, L. (2012)Phaeobacter gallaeciensis reduces Vibrio anguillarum incultures of microalgae and rotifers, and prevents vibriosisin cod larvae. PLoS ONE 7: e43996.

D’Alvise, P.W., Lillebø, S., Wergeland, H.I., Gram, L., andBergh, Ø. (2013) Protection of cod larvae from vibriosisby Phaeobacter spp.: a comparison of strains and intro-duction times. Aquaculture 384–387: 82–86.

D’Alvise, P.W., Phippen, C.B.W., Nielsen, K.F., and Gram,L. (2016) Influence of iron on production of the antibacte-rial compound tropodithietic acid and its noninhibitoryanalog in Phaeobacter inhibens. Appl Environ Microbiol82: 502–509.

Defoirdt, T., Sorgeloos, P., and Bossier, P. (2011) Alterna-tives to antibiotics for the control of bacterial disease inaquaculture. Curr Opin Microbiol 14: 251–258.

Drillet, G., Iversen, M.H., Sørensen, T.F., Ramløv, H., Lund,T., and Hansen, B.W. (2006) Effect of cold storage uponeggs of a calanoid copepod, Acartia tonsa (Dana) andtheir offspring. Aquaculture 254: 714–729.

Dumontet, S., Krovacek, K., Baloda, S.B., Grottoli, R., Pas-quale, V., and Vanucci, S. (1996) Ecological relationshipbetween Aeromonas and Vibrio spp. and planktonic cope-pods in the coastal marine environment in southern Italy.Comp Immunol Microbiol Infect Dis 19: 245–254.

Eiler, A., and Bertilsson, S. (2006) Detection and quantifica-tion of Vibrio populations using denaturant gradient gelelectrophoresis. J Microbiol Methods 67: 339–348.

FAO (2016) The State of World Fisheries and Aquaculture2016. Rome: FAO.

FAO and WHO (2001) Report of a joint FAO/WHO expertconsultation on evaluation of health and nutritional proper-ties of probiotics in food including powder milk with livelactic acid bacteria.

Gatesoupe, F.J. (1999) The use of probiotics in aquaculture.Aquaculture 180: 147–165.

Gram, L., Rasmussen, B.B., Wemheuer, B., Bernbom, N.,Ng, Y.Y., Porsby, C.H., et al. (2015) Phaeobacter inhi-bens from the Roseobacter clade have a natural environ-mental niche as surface colonizer in Danish harbors. SystAppl Microbiol 38: 483–493.

Grotkjær, T., Bentzon-Tilia, M., D’Alvise, P., Dierckens, K.,Bossier, P., and Gram, L. (2016a) Phaeobacter inhibensas probiotic bacteria in non-axenic Artemia and algae cul-tures. Aquaculture 462: 64–69.

Grotkjær, T., Bentzon-Tilia, M., D’Alvise, P., Dourala, N., Niel-sen, K.F., and Gram, L. (2016b) Isolation of TDA-producingPhaeobacter strains from sea bass larval rearing units andtheir probiotic effect against pathogenic Vibrio spp. in Arte-mia cultures. Syst Appl Microbiol 39: 180–188.

Gugliandolo, C., Irrera, G.P., Lentini, V., and Maugeri, T.L.(2008) Pathogenic Vibrio, Aeromonas and Arcobacterspp. associated with copepods in the Straits of Messina(Italy). Mar Pollut Bull 56: 600–606.

Guillard, R.R.L. (1975) Culture of phytoplankton for feedingmarine invertebrates. In Culture of marine invertebrateanimals. Smith, W.L. and Chanley, M.H. (ed.). Boston,MA: Springer, pp. 29–60.

Guillard, R.R.L., and Ryther, J.H. (1962) Studies of marineplanktonic diatoms: Cyclotella nana hustedt, and Detonulaconfervacea (cleve) gran. Can J Microbiol 8: 229–239.

Hansen, B.W., Drillet, G., Kozm�er, A., Madsen, K.V., Peder-sen, M.F., and Sørensen, T.F. (2010) Temperature effectson copepod egg hatching: does acclimatization matter? JPlankton Res 32: 305–315.

Heidelberg, J.F., Heidelberg, K.B., and Colwell, R.R. (2002)Bacteria of the c-subclass Proteobacteria associated withzooplankton in Chesapeake Bay. Appl Environ Microbiol68: 5498–5507.

Hjelm, M., Bergh, O., Riaza, A., Nielsen, J., Melchiorsen, J.,Jensen, S., et al. (2004) Selection and identification ofautochthonous potential probiotic bacteria from turbot lar-vae (Scophthalmus maximus) rearing units. Syst ApplMicrobiol 27: 360–371.

ª 2018 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology, MicrobialBiotechnology, 11, 1070–1079

Phaeobacter–Vibrio interactions in copepods 1077

Hoeckendorf, A., and Leippe, M. (2012) SPP-3, a saposin-like protein of Caenorhabditis elegans, displays antimicro-bial and pore-forming activity and is located in the intes-tine and in one head neuron. Dev Comp Immunol 38:181–186.

Huq, A., Small, E.B., West, P.A., Huq, M.I., and Colwell,R.R. (1983) Ecological relationships between Vibrio cho-lerae and planktonic crustacean copepods. Appl EnvironMicrobiol 45: 275–283.

Kesarcodi-Watson, A., Kaspar, H., Lategan, M.J., and Gib-son, L. (2008) Probiotics in aquaculture: the need, princi-ples and mechanisms of action and screening processes.Aquaculture 274: 1–14.

Kibenge, F.S.B., Godoy, M.G., Fast, M., Workenhe, S., andKibenge, M.J.T. (2012) Countermeasures against viraldiseases of farmed fish. Antiviral Res 95: 257–281.

Kim, J.Y., and Lee, J.L. (2014) Multipurpose assessment forthe quantification of Vibrio spp. and total bacteria in fishand seawater using multiplex real-time polymerase chainreaction. J Sci Food Agric 94: 2807–2817.

Li, X., Defoirdt, T., Yang, Q., Laureau, S., Bossier, P., andDierckens, K. (2014) Host-induced increase in larval seabass mortality in a gnotobiotic challenge test with Vibrioanguillarum. Dis Aquat Organ 108: 211–216.

Loker, E.S., Adema, C.M., Zhang, S.-M., and Kepler, T.B.(2004) Invertebrate immune systems - not homogeneous,not simple, not well understood. Immunol Rev 198: 10–24.

L�opez-Torres, M.A., and Liz�arraga-Partida, M.L. (2001) Bac-teria isolated on TCBS media associated with hatchedArtemia cysts of commercial brands. Aquaculture 194:11–20.

Martens, T., Heidorn, T., Pukall, R., Simon, M., Tindall, B.J.,and Brinkhoff, T. (2006) Reclassification of Roseobactergallaeciensis Ruiz-Ponte et al. 1998 as Phaeobacter gal-laeciensis gen. nov., comb. nov., description ofPhaeobacter inhibens sp. nov., reclassification of Ruege-ria algicola (Lafay et al. 1995) Uc. Int J Syst Evol Micro-biol 56: 1293–1304.

Nielsen, L.T., and Kiørboe, T. (2015) Feeding currents facili-tate a mixotrophic way of life. ISME J 9: 2117–2127.

Norouzitallab, P., Baruah, K., Biswas, P., Vanrompay, D.,and Bossier, P. (2016) Probing the phenomenon oftrained immunity in invertebrates during a transgenera-tional study, using brine shrimp Artemia as a model sys-tem. Sci Rep 6: 1–14.

Øie, G., Galloway, T., Sørøy, M., Holmvaag Hansen, M.,Norheim, I.A., Halseth, C.K., et al. (2017) Effect of culti-vated copepods (Acartia tonsa) in first-feeding of Atlanticcod (Gadus morhua) and ballan wrasse (Labrus bergylta)larvae. Aquac Nutr 23: 3–17.

Olafsen, J.A. (2001) Interactions between fish larvae andbacteria in marine aquaculture. Aquaculture 200: 223–247.

Pedersen, K., and Larsen, J. (1993) rRNA gene restrictionpatterns of Vibrio anguillarum serogroup O1. Dis AquatOrgan 16: 121–126.

Planas, M., P�erez-Lorenzo, M., Hjelm, M., Gram, L., Ugle-nes Fiksdal, I., Bergh, Ø., and Pintado, J. (2006) Probioticeffect in vivo of Roseobacter strain 27-4 against Vibrio(Listonella) anguillarum infections in turbot (Scophthalmusmaximus L.) larvae. Aquaculture 255: 323–333.

Porsby, C.H., Nielsen, K.F., and Gram, L. (2008) Phaeobac-ter and Ruegeria species of the Roseobacter clade colo-nize separate niches in a Danish Turbot (Scophthalmusmaximus)-rearing farm and antagonize Vibrio anguillarumunder different growth conditions. Appl Environ Microbiol74: 7356–7364.

Prol, M.J., Bruhn, J.B., Pintado, J., and Gram, L. (2009)Real-time PCR detection and quantification of fish probi-otic Phaeobacter strain 27-4 and fish pathogenic Vibrio inmicroalgae, rotifer, Artemia and first feeding turbot (Psettamaxima) larvae. J Appl Microbiol 106: 1292–1303.

Rao, D., Webb, J.S., and Kjelleberg, S. (2005) Competitiveinteractions in mixed-species biofilms containing the mar-ine bacterium Pseudoalteromonas tunicata. Appl EnvironMicrobiol 71: 1729–1736.

Reid, H.I., Treasurer, J.W., Adam, B., and Birkbeck, T.H.(2009) Analysis of bacterial populations in the gut ofdeveloping cod larvae and identification of Vibrio logei,Vibrio anguillarum and Vibrio splendidus as pathogens ofcod larvae. Aquaculture 288: 36–43.

Ringø, E., Olsen, R.E., Jensen, I., Romero, J., and Lauzon,H.L. (2014) Application of vaccines and dietary supple-ments in aquaculture: possibilities and challenges. RevFish Biol Fish 24: 1005–1032.

Rønneseth, A., Castillo, D., D’Alvise, P., Tønnesen, Ø.,Haugland, G., Grotkjaer, T., et al. (2017) Comparativeassessment of Vibrio virulence in marine fish larvae. JFish Dis 40: 1373–1385.

Ruiz-Ponte, C., Cilia, V., Lambert, C., and Nicolas, J.L.(1998) Roseobacter gallaeciensis sp. nov., a new marinebacterium isolated from rearings and collectors of thescallop Pecten maximus. Int J Syst Bacteriol 48: 537–542.

Ruiz-Ponte, C., Samain, J.J.F., S�anchez, J.L.J., and Nico-las, J.J.L. (1999) The benefit of a Roseobacter specieson the survival of scallop larvae. Mar Biotechnol 1: 52–59.

Salvesen, I., Reitan, K.I., Skjermo, J., and Øie, G. (2000)Microbial environments in marine larviculture: impacts ofalgal growth rates on the bacterial load in six microalgae.Aquac Int 8: 275–287.

Saulnier, D., De Decker, S., and Haffner, P. (2009) Real-time PCR assay for rapid detection and quantification ofVibrio aestuarianus in oyster and seawater: a useful toolfor epidemiologic studies. J Microbiol Methods 77: 191–197.

Skjermo, J., and Vadstein, O. (1999) Techniques for micro-bial control in the intensive rearing of marine larvae.Aquaculture 177: 333–343.

Skov, M.N., Pedersen, K., and Larsen, J.L. (1995) Compar-ison of pulsed-field gel electrophoresis, ribotyping, andplasmid profiling for typing of Vibrio anguillarum SerovarO1. Appl Environ Microbiol 61: 1540–1545.

Sobecky, P.A., Mincer, T.J., Chang, M.C., and Helinski,D.R. (1997) Plasmids isolated from marine sedimentmicrobial communities contain replication and incompati-bility regions unrelated to those of known plasmid groups.Appl Environ Microbiol 63: 888–895.

Sochard, M.R., Wilson, D.F., Austin, B., and Colwell, R.R.(1979) Bacteria associated with the surface and gut ofmarine copepods. Appl Environ Microbiol 37: 750–759.

ª 2018 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology, MicrobialBiotechnology, 11, 1070–1079

1078 B. B. Rasmussen, K. E. Erner, M. Bentzon-Tilia and L. Gram

Sommerset, I., Krossøy, B., Biering, E., and Frost, P. (2005)Vaccines for fish in aquaculture. Expert Rev Vaccines 4:89–101.

Støttrup, J.G., and Norsker, N.H. (1997) Production and useof copepods in marine fish larviculture. Aquaculture 155:231–247.

Støttrup, J.G., Richardson, K., Kirkegaard, E., and Pihl, N.J.(1986) The cultivation of Acartia tonsa Dana for use as alive food source for marine fish larvae. Aquaculture 52:87–96.

Tamplin, M.L., Gauzens, A.L.L., Huq, A., Sack, D.A.A., andColwell, R.R.R. (1990) Attachment of Vibrio cholerae ser-ogroup 01 to zooplankton of Bangladesh waters. ApplEnviron Microbiol 56: 1977–1980.

Thompson, F.L., Iida, T., and Swings, J. (2004) Biodiversityof Vibrios. Microbiol Mol Biol Rev 68: 403–431.

Toranzo, A.E., Magari~nos, B., and Romalde, J.L. (2005) Areview of the main bacterial fish diseases in mariculturesystems. Aquaculture 246: 37–61.

Turingan, R.G., Beck, J.L., Krebs, J.M., and Licamele, J.D.(2007) Development of feeding mechanics in marine fishlarvae and the swimming behavior of zooplankton prey:implications for rearing marine fishes. In Copepods inaquaculture. Lee, C.S., O’Bryen, P.J. and Marcus, N.H.(ed.). Ames, IA: Blackwell Publishing Professional, pp.119–132.

Turner, J.T. (2004) The importance of small pelagic plank-tonic copepods and their role in pelagic marine foodwebs. Zool Stud 43: 255–266.

Vandecandelaere, I., Segaert, E., Mollica, A., Faimali, M.,and Vandamme, P. (2008) Leisingera aquimarina sp.nov., isolated from a marine electroactive biofilm, andemended descriptions of Leisingera methylohalidivoransSchaefer et al. 2002, Phaeobacter daeponensis Yoonet al. 2007 and Phaeobacter inhibens Martens et al.2006. Int J Syst Evol Microbiol 58: 2788–2793.

Venmathi Maran, B.A., Iwamoto, E., Okuda, J., Matsuda,S., Taniyama, S., Shida, Y., et al. (2007) Isolation andcharacterization of bacteria from the copepod Pseudo-caligus fugu ectoparasitic on the panther puffer Takifugupardalis with the emphasis on TTX. Toxicon 50: 779–790.

Vezzulli, L., Pezzati, E., Stauder, M., Stagnaro, L., Venier,P., and Pruzzo, C. (2015) Aquatic ecology of the oysterpathogens Vibrio splendidus and Vibrio aestuarianus.Environ Microbiol 17: 1065–1080.

Wang, X.W., Xu, J.D., Zhao, X.F., Vasta, G.R., and Wang,J.X. (2014) A shrimp C-type lectin inhibits proliferationof the hemolymph microbiota by maintaining the expres-sion of antimicrobial peptides. J Biol Chem 289: 11779–11790.

WWF and ZSL (2015) Living Blue Planet Report – Species,habitats and human well-being.

Ye, J., Coulouris, G., Zaretskaya, I., Cutcutache, I., Rozen,S., and Madden, T.L. (2012) Primer-BLAST: a tool todesign target-specific primers for polymerase chain reac-tion. BMC Bioinformatics 13: 134.

ª 2018 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology, MicrobialBiotechnology, 11, 1070–1079

Phaeobacter–Vibrio interactions in copepods 1079

Paper 3

Rørbo N, Rønneseth A, Kalatzis PG, Rasmussen BB, Engell-Sørensen K,

Kleppen HP, Wergeland HI, Gram L, Middelboe M. 2018.

Exploring the effect of phage therapy in preventing Vibrio anguillarum

infections in cod and turbot larvae.

Antibiotics 7(2), 42

antibiotics

Article

Exploring the Effect of Phage Therapyin Preventing Vibrio anguillarum Infectionsin Cod and Turbot Larvae

Nanna Rørbo 1, Anita Rønneseth 2, Panos G. Kalatzis 1,3 ID , Bastian Barker Rasmussen 4 ID ,

Kirsten Engell-Sørensen 5, Hans Petter Kleppen 6, Heidrun Inger Wergeland 2, Lone Gram 4 ID

and Mathias Middelboe 1,* ID

1 Marine Biological Section, University of Copenhagen, 3000 Helsingør, Denmark;nanna_ir@hotmail.com (N.R.); panos.kalatzis@bio.ku.dk (P.G.K.)

2 Department of Biology, University of Bergen, 5020 Bergen, Norway; anita.ronneseth@uib.no (A.R.);heidrun.wergeland@uib.no (H.I.W.)

3 Institute of Marine Biology, Biotechnology and Aquaculture, Hellenic Centre for Marine Research,71003 Heraklion, Greece

4 Department of Biotechnology and Biomedicine, Technical University of Denmark,2800 Kongens Lyngby, Denmark; bbara@bio.dtu.dk (B.B.R.); gram@bio.dtu.dk (L.G.)

5 Fishlab, 8270 Højbjerg, Denmark; kes@fishlab.dk6 ACD Pharmaceuticals AS, 8376 Leknes, Norway; hans.kleppen@acdpharma.com* Correspondence: mmiddelboe@bio.ku.dk; Tel.: +45-3532-1991

Received: 30 December 2017; Accepted: 10 May 2018; Published: 16 May 2018���������������

Abstract: The aquaculture industry is suffering from losses associated with bacterial infections byopportunistic pathogens. Vibrio anguillarum is one of the most important pathogens, causing vibriosisin fish and shellfish cultures leading to high mortalities and economic losses. Bacterial resistanceto antibiotics and inefficient vaccination at the larval stage of fish emphasizes the need for novelapproaches, and phage therapy for controlling Vibrio pathogens has gained interest in the past fewyears. In this study, we examined the potential of the broad-host-range phage KVP40 to control fourdifferent V. anguillarum strains in Atlantic cod (Gadus morhua L.) and turbot (Scophthalmus maximus L.)larvae. We examined larval mortality and abundance of bacteria and phages. Phage KVP40 was able toreduce and/or delay the mortality of the cod and turbot larvae challenged with V. anguillarum. However,growth of other pathogenic bacteria naturally occurring on the fish eggs prior to our experiment causedmortality of the larvae in the unchallenged control groups. Interestingly, the broad-spectrum phageKVP40 was able to reduce mortality in these groups, compared to the nonchallenge control groups nottreated with phage KVP40, demonstrating that the phage could also reduce mortality imposed by thebackground population of pathogens. Overall, phage-mediated reduction in mortality of cod and turbotlarvae in experimental challenge assays with V. anguillarum pathogens suggested that application ofbroad-host-range phages can reduce Vibrio-induced mortality in turbot and cod larvae, emphasizingthat phage therapy is a promising alternative to traditional treatment of vibriosis in marine aquaculture.

Keywords: Vibrio anguillarum; phage therapy; aquaculture; fish larvae; challenge trials

1. Introduction

Vibrionaceae is a genetic and metabolic diverse family of heterotrophic bacteria which arewidespread in aquatic environments around the world [1]. Several vibrios are able to infect a widerange of aquatic animals and constitute therefore a large problem in aquaculture [2]. One of the mostimportant is Vibrio anguillarum, which causes the disease vibriosis and is responsible for large-scale

Antibiotics 2018, 7, 42; doi:10.3390/antibiotics7020042 www.mdpi.com/journal/antibiotics

Antibiotics 2018, 7, 42 2 of 15

losses in the aquaculture industry [3,4]. Chemotherapy against vibriosis is associated with a majorconcern due to the risk of antibiotic-resistance developing in the pathogenic bacteria [5]. Vaccinesagainst vibrio have been successful in preventing disease [6,7], however, they are often not useful at thelarval stage, as the immune system is not fully developed. Therefore, alternative methods for the controland treatment of V. anguillarum infections in fish larvae and fry are needed. The use of bacteriophages(phages) has been explored in several studies as a treatment of pathogens in aquaculture [4,8–13].Pereira et al. [4] and Mateus et al. [11] did in vitro assays with phages infecting different bacteriaresponsible for the diseases vibriosis and furunculosis and showed that both single-phage suspensionsand phage cocktails could inactivate the bacteria [4,11]. However, often regrowth of phage tolerantbacteria was observed within 24 h after phage treatment [11,13]. Phage addition to shrimp larvaeinfected with V. harveyi caused a reduction in the pathogen load and significantly increased shrimpsurvival compared to untreated controls groups as well as parallel treatments with antibiotics [8,9].Another study on zebrafish larvae infected with V. anguillarum [12] also found significantly enhancedlarvae survival after phage addition. Successful phage treatment in Atlantic salmon (Salmo salar L.)infected with V. anguillarum strain PF4 was found for phage CHOED, resulting in complete eliminationof pathogen-induced mortality when phages were added at a high multiplicity of infection [10].Together, the previous experimental approaches demonstrate that phage therapy can be a feasiblealternative method to control specific Vibrio pathogens in aquaculture. However, the use of phages iscomplicated by the fact that multiple strains of the Vibrio pathogens with different phage susceptibilitypatterns may coexist in aquaculture environments [14]. The implications of strain diversity for theefficiency of phage control may be overcome either by combining several phages which target a broadrange of pathogenic hosts, or to use a broad-host-range phage which can infect multiple strains withina given species or even multiple species [15]. The phage KVP40 represents a broad-host-range phagewhich infects at least eight species of Vibrio sp. (V. parahaemolyticus, V. alginolyticus, V. natriegens,V. cholerae, V. mimicus, V. anguillarum, V. splendidus, and V. fluvialis) and one Photobacterium sp.(P. leignathi) [16]. All of these species contain a 26-kDa outer membrane protein named OmpK,which is a receptor for phage KVP40 [17].

The application of phages for controlling pathogens may be hampered by the development ofphage resistance in the bacteria [18], and several mechanisms have been described in V. anguillarumwhich can eliminate or reduce bacterial sensitivity to phages and thus limit the efficiency and durationof phage control [19].

The aim of this study was to examine the effect of phage KVP40 on the survival of turbot and codlarvae challenged with four different V. anguillarum strains. Larval mortality and abundance of bacteriaand phages were quantified to determine the potential of using phage KVP40 to control V. anguillaruminfections during the early larval stage. In general, phage KVP40 was able to reduce or delay themortality of both turbot and cod larvae in all the challenge trials and reduce larval mortality imposedby the background population of pathogens.

The results demonstrated that phage KVP40 reduced the mortality imposed by the added pathogensas well as other Vibrio pathogens already present in the environment during the initial 1–4 days ofthe experiment, emphasizing the potential of using phages to reduce turbot and cod mortality at thelarval stage.

2. Results

2.1. Phage Effect on Turbot Mortality in Vibrio Challenge Trials

2.1.1. Turbot Challenge Trial 1

In general, larval mortality was high in all treatments, including the nonchallenged controls wherea maximum mortality of 86% (i.e., 103 dead larvae out of 120) was found (Figure 1), indicating that theeggs were associated with unknown bacterial pathogens prior to the challenge trial. Challenging theturbot eggs with V. anguillarum resulted in higher mortalities for all four strains (Figure 1), emphasizing

Antibiotics 2018, 7, 42 3 of 15

that the added V. anguillarum pathogens increased larval mortality. Strain PF430-3 was the mostvirulent of the four strains, with 100% larval mortality after 3 days, whereas strains PF7, 90-11-286,and 4299 caused 97%–100% mortality after 4 and 5 days of challenge. Subsequent quantification of theabundance of colony forming bacteria in the water used for transportation of the fish eggs confirmedthe presence of a microbial community associated with the eggs (see Section 2.5).

Despite the presence of other pathogen communities associated with the eggs/larvae, addition ofphage KVP40 had a significant positive effect on larval survival in all the challenge treatments during allor part of the trials. When challenged with strain PF430-3, the maximum relative reduction in mortalitywas 29% (p < 0.05) one day after phage addition (Figure 1a; Table 1). The delay in mortality only lastedfor 3 days, and the mortality reached almost 100% mortality at day 5 (Figure 1a). When challenged withstrain PF7 or strain 90-11-286 (Figure 1b,c), the maximum phage-induced reduction in mortality was47% obtained 1 and 2 days (Table 1), respectively, after addition of KVP40 and a significant effect of thephage on mortality was observed for 3–4 days (p < 0.05). The effect of phage addition was largest in thetreatment group with strain 4299, where the larval mortality remained below 66% throughout the 8-daytrial, corresponding to an average of 36% reduction in larval mortality compared to larvae challengedwith V. anguillarum (p < 0.05) (Figure 1d).

Interestingly, larval mortality in the KVP40 controls (addition of phage but not V. anguillarum) showedthe lowest larval mortality, reaching 65% at day 4 and remaining at that level (Figure 1). This significantreduction in mortality compared to the nonchallenged control (i.e., 86% mortality in larvae not exposedto V. anguillarum or phage) suggested that phage KVP40 was able to control part of the unknownpathogen community, thereby increasing the larval survival. This was later confirmed by analysis ofphage susceptibility of bacteria initially associated with the eggs (see Section 2.5 below).

Figure 1. Cumulative percent mortality over time in turbot challenge trial 1: (a) strain PF430-3;(b) strain PF7; (c) strain 90-11-286; (d) strain 4299. Significant difference in mortality between cultures“V. anguillarum” and “V. anguillarum + KVP40” for individual time points is indicated by *. Significantdifference in mortality between cultures “Nonchallenge control” and “KVP40 control” is indicated by c.

Antibiotics 2018, 7, 42 4 of 15

Table 1. Overview of the percent reduction in mortality caused by phage KVP40 addition in the fourexperiments. The maximum relative reduction and reduction at the end of the experiment (final)is shown.

Relative Reduction * in Larval Mortality in the Presence of Phages (%)

V. anguillarum Strains

Turbot Challenge Trial Cod Challenge Trial

1 2 1 2

Max. Final Max. Final Max. Final Max. Final

PF430-3 29 N/S 1 60 N/S 1 79 N/S 1 86 N/ 1

PF7 47 N/S 1 53 N/S 1 75 43 59 3290-11-286 47 N/S 1 92 N/S 1 −119 N/S 1 49 N/S 1

4299 48 33 45 N/S 1 N/D 2 N/D 2 82 72

* The relative reduction in mortality is calculated as difference in mortality between V. anguillarum and V. anguillarum+ phage treatment, divided by the mortality in the V. anguillarum treatment. 1 N/S: not significant, 2 N/D:not determined.

2.1.2. Turbot Challenge Trial 2

The relatively high fraction of low-quality eggs and high mortality in the control group led us torepeat the challenge experiments in an attempt to optimize the egg quality and in order to verify theindications of positive effects of phages for larval mortality in replicate experiments.

Also in the second challenge trial with turbot larvae, a high mortality (71%) was observed in thenonchallenged control groups after 5 days (Figure 2), indicating pathogenic effects of the bacterialbackground community in the turbot eggs. In contrast to turbot challenge trial 1, the mortalitycaused by the background bacteria was not observed immediately, and mortality in the control groupsgradually increased during the first 4 days, indicating growth of the pathogenic bacteria. Addition ofV. anguillarum strains increased larval mortality in all four treatments, resulting in mortalities between72% and 98% after 4–5 days of incubation. As in turbot challenge trial 1, addition of phage KVP40 hadsignificant positive effects on the larval survival. However, in this case, the phage addition delayed themortality by 2–4 days relative to the treatment with V. anguillarum alone.

Figure 2. Cumulative percent mortality over time in turbot challenge trial 2: (a) strain PF430-3;(b) strain PF7; (c) strain 90-11-286; (d) strain 4299. Significant difference in mortality between cultures“V. anguillarum” and “V. anguillarum + KVP40” for individual time points is indicated by *.

Antibiotics 2018, 7, 42 5 of 15

When challenged with strain PF430-3, the addition of phages reduced mortality from 29% to 11%2 days after phage addition (Figure 2a), corresponding to a maximum phage-mediated reduction inmortality of 60% (p < 0.05, Table 1). The delay in mortality lasted until day 4, where mortality approached100% mortality as in the treatment without phage (Figure 2a). Phage addition to the larvae challengedwith strain PF7 and strain 90-11-286 resulted in a significant 3-day delay in mortality with a maximumreduction in mortality of 53% and 92%, respectively, after 2–3 days relative to the larvae challenged withV. anguillarum alone (p < 0.05 (Figure 2b,c; Table 1). As in the turbot challenge trial 1, the larvae challengedwith strain 4299 were best protected by phage addition, with a maximum relative reduction in mortality of45% (p < 0.05) obtained 3 days after phage addition (Table 1), and a continued reduction in larval mortalityof 22% relative to the larvae challenged with bacteria alone throughout the experiment (Figure 2d).

2.2. Abundance of Bacteria and Phages in Turbot Challenge Trial 2

In all the treatments in turbot challenge trial 2, the total count of colony forming bacteria (CFU)increased exponentially over time for the first 2–4 days (Figure 3).

The number of infective KVP40 phages increased about 100-fold reaching 1–5 × 1010 PFU mL−1 inall the treatment groups where KVP40 was added, with no significant differences between cultures withand without the addition of Vibrio pathogens. This indicated that the background bacteria supportedphage proliferation and that addition of V. anguillarum only had a minor effect on phage production.

Figure 3. Bacterial abundance (CFU mL−1) and phage abundance (PFU mL−1) in turbot challenge trial 2:(a) strain PF430-3; (b) strain PF7; (c) strain 90-11-286; (d) strain 4299.

2.3. Phage Effect on Cod Mortality in Vibrio Challenge Trials

2.3.1. Cod Challenge Trial 1

The cod larvae mortality in the nonchallenged controls remained low throughout the trial (<10%)(Figure 4), and the addition of Vibrio anguillarum strains increased mortality significantly (Figure 4).

Antibiotics 2018, 7, 42 6 of 15

Strain PF430-3 and strain 90-11-286 increased mortality to 82% and 78%, respectively, after 11 days(Figure 4a,c), whereas the mortality was 41% in the treatment with strain PF7 (Figure 4b).

The addition of phage KVP40 had significant positive effects on larval survival in the larvaeexposed to strain PF430-3 and strain PF7. For strain PF430-3, the mortality was reduced from 24% to5% in the phage added cultures after 5 days, corresponding to maximal relative reduction in mortalityby phage KVP40 of 79% compared to the larvae only challenged with V. anguillarum (p < 0.05; Table 1).The significant phage-induced reduction in mortality lasted to day 8 (Figure 4a). Phage KVP40 additionto strain PF7 reduced relative larval mortality by 75% compared to the larvae only challenged withV. anguillarum (p < 0.05) after 8 days (Table 1), and the significant phage-mediated reduction in mortalityremained throughout the 11-day trial (Figure 4b). Surprisingly, the addition of phage KVP40 increasedlarval mortality significantly in the cultures challenged with strain 90-11-286 with a maximum increasein mortality of 119 (p < 0.05) reached at day 6 (Figure 4c; Table 1). The negative effect of phage additionwas significant from day 5 to day 10, with the mortality reaching 100% in the phage treated cultures atday 11.

Despite the low mortality in the nonchallenged control treatment, the reduced larval mortalityin the phage KVP40 controls (addition of phage but not V. anguillarum) (<7%) compared with thenonchallenged control group without phages again indicated a positive effect of the phages in reducingthe original pathogenic bacterial load in the trials.

Figure 4. Cumulative percent mortality over time in cod challenge trial 1: (a) strain PF430-3; (b) strainPF7; (c) strain 90-11-286. Significant difference in mortality between cultures “V. anguillarum” and“V. anguillarum + KVP40” for individual time points is indicated by *.

2.3.2. Cod Challenge Trial 2

As for the turbot experiments, the challenge trials with cod were repeated to examine the reproducibilityof the first results using a new batch of eggs. The second challenge trial with cod larvae confirmed the highvirulence of strains PF430-3 and 90-11-286 obtained in cod challenge trial 1, whereas strain PF7 caused lessmortality in cod challenge trial 2. Strain 4299 was not very virulent to the cod larvae (Figure 5). A gradualincrease in mortality was observed in larvae challenged with strains PF430-3, PF7, and 90-11-286, whichreached mortalities of 74% to 91% after 11 days post challenge (Figure 5a–c). Challenge with strain 4299 didnot increase mortality compared to the nonchallenged control level, suggesting that this strain had very low

Antibiotics 2018, 7, 42 7 of 15

virulence to cod (Figure 5d). The nonchallenged control showed an increase in mortality from 5% to 15%between days 2 and 3, followed by a more gradual increase to 35% mortality at day 11 (Figure 5).

Addition of phage KVP40 had a significant positive effect on cod larvae survival in all thetreatments (Table 1). In the larvae challenged with strain PF430-3, phage addition kept larval mortalitybelow 27% for 6 days, with a maximum reduction in mortality of 86% (p < 0.05) obtained 4 days afterphage addition (Table 1). The reduced mortality lasted from day 2 to day 9, and after day 10 themortality reached almost the same level as in the cultures without phages (Figure 5a). When challengedwith strain PF7, the maximal effect of phage addition was a reduction in mortality of 59% (p < 0.05)obtained 6 days after phage addition (Table 1). The delay in mortality lasted throughout the trial,with the difference being significant from day 5 and onwards (Figure 5b). In the treatments challengedwith strain 90-11-286, the maximal reduction in mortality was 49% (p < 0.05) obtained 6 days afterphage addition (Table 1). The mortality then increased but remained below the nonphage treatedgroup throughout the experiment (Figure 5c). Phage KVP40 very efficiently reduced mortality oflarvae challenged with strain 4299, with a maximum reduction of 82% (p < 0.05) after 5 days (Table 1),and a significant reduction in mortality (mortality always < 12%) throughout the trial (Figure 5d).

The relatively high initial mortality in the nonchallenged control from day 1 to day 3 compared withcorresponding nonchallenge control group in cod challenge trial 1, and compared with the lower andmore gradual increase in mortality in the group challenged with strain 4299, suggested the presence ofa high fraction of low-quality eggs in this specific control group. As in the previous trials, the phage-addedcontrols showed a lower mortality than in the nonchallenged controls, again suggesting a positive effectof phage KVP40 in controlling other pathogens growing up during the trials (Figure 5).

Figure 5. Cumulative percent mortality over time in cod challenge trial 2: (a) strain PF430-3;(b) strain PF7; (c) strain 90-11-286; (d) strain 4299. Significant difference in mortality between cultures“V. anguillarum” and “V. anguillarum + KVP40” for individual time points is indicated by *. Significantdifference in mortality between cultures “Nonchallenge control” and “KVP40 control” is indicated by c.

2.4. Abundance of Bacteria and Phages in Cod Challenge Trials

2.4.1. Cod Challenge Trial 1

The total abundance of colony forming microorganisms increased approximately 10-fold in allVibrio challenged larval groups from approx. 105 to 106 CFU mL−1 (Figure 6). Addition of phages

Antibiotics 2018, 7, 42 8 of 15

only reduced the bacterial load in the strain PF7 challenged larval group and only during the first2 days (Figure 6b). In contrast to this, total CFU counts increased after addition of phage KVP40in larval groups challenged with strain PF430-3 and strain 90-11-286. Especially in the challengewith strain 90-11-286, a > 10-fold increase in colony forming bacteria was observed (Figure 6c) inaccordance with the increased larval mortality in this treatment (Figure 4c). The phage abundance wasapproximately 107 PFU mL−1 in all phage-added treatments and remained stable during the 4 dayswhen PFU was measured.

Figure 6. Bacterial abundance (CFU mL−1) and phage abundance (PFU mL−1) in cod challenge trial 1:(a) strain PF430-3; (b) strain PF7; (c) strain 90-11-286.

2.4.2. Cod Challenge Trial 2

The V. anguillarum load was approximately 10-fold higher in the second than in the first cod challengetrial and the CFU counts were approximately 106 CFU mL−1 in the Vibrio challenged groups (Figure 7).

In all the groups, addition of phage KVP40 reduced the bacterial counts significantly from day 0.In the groups challenged with strain PF430-3 and strain PF7, a significant phage-mediated reduction(approximately 1 log reduction) in the V. anguillarum pathogens was maintained for the first 8–9 days,followed by an increase in total CFU which then reached values close to the bacteria-alone group at day 11(Figure 7a,b). For the group challenged with strain 90-11-286, phage reduction of the Vibrio pathogen wasrather short. After 3 days, the bacterial abundance had reached the same level as in the bacteria-only group(Figure 7c). In the group challenged with strain 4299, the addition of phage KVP40 caused a 100-foldreduction in total CFU counts, indicating a strong phage control of the pathogen. However, after day 8,total bacterial cell counts increased 100-fold and reached numbers similar to the group without phage(Figure 7d). Phages were added at an initial concentration of 1.75 × 109 PFU mL−1 and the abundance ofphage remained stable throughout the trial, both in the absence and presence of the Vibrio hosts (Figure 7).

Antibiotics 2018, 7, 42 9 of 15

Figure 7. Bacterial abundance (CFU mL−1) and phage abundance (PFU mL−1) in cod challenge trial 2:(a) strain PF430-3; (b) strain PF7; (c) strain 90-11-286; (d) strain 4299. Significant difference in CFUbetween cultures “CFU: V. anguillarum” and “CFU: V. anguillarum + KVP40” for individual time pointsis indicated by *.

2.5. Abundance and Phage KVP40 Susceptibility of Bacterial Background Communities Associated with theTurbot Eggs

During the second turbot trial, the abundance of colony-forming bacteria in water used fortransportation of the fish eggs was determined to shed light on the observed positive effect of phageKVP40 on unchallenged control groups. Different general and Vibrio-promoting growth media wereused. In all the experiments, there was a high load of bacteria associated with the eggs, and a generalincrease in their abundance over time was found (Table 2). The high abundance of colonies growing onTCBS plates (up to >108 CFU mL−1) indicated that a large fraction of these background communitieswere presumptive Vibrio or Vibrio-related species.

Table 2. Abundance of the bacterial background community (CFU mL−1) associated with the fish eggs,in turbot challenge trial 2, and cultured on different media. Day 0: water the eggs were transported infor 24 h; Day 11: water in the wells of the live nonchallenged larvae.

Growth Substrate Day 0 (CFU mL−1) Day 11 (CFU mL−1)

LB media 2 × 107 9.39 × 106

TCBS media 2 × 106 1.5 × 108

Marine agar N/D 1 2.89 × 108

1 N/D: not determined.

The susceptibility to phage KVP40 was tested in 40 isolates obtained from the water containing theturbot eggs during transportation used for challenge trial 1 by quantification of the growth reductionrelative to a control culture without phage KVP40 (Figure 8). The results showed that 35 out of 40 isolatedshowed a growth reduction, indicating that the majority of the colony-forming cells originating from thewater used for transporting the eggs were susceptible to phage KVP40.

Antibiotics 2018, 7, 42 10 of 15

Figure 8. Quantification of phage KVP40-induced inhibition/promotion of cell growth in cultures ofbacteria isolated from water used for transport of eggs used in turbot challenge trial 1. Phage-inducedgrowth inhibition/promotion was determined as the percent cell density in cultures added phageKVP40 relative to control cultures without phage KVP40 (100%) after 3 h incubation.

3. Discussion

In general, the addition of phage KVP40 reduced or delayed the mortality of turbot and cod larvaechallenged with V. anguillarum, with the largest effect observed for strain 4299, where the relative turbotand cod mortality was reduced by 22–33% and 72%, respectively, by the end of the experiment. In mostof the challenges, the positive effect of phage KVP40 addition on larval survival was maintainedthroughout the incubation period. However, incubation with strain PF430-3 showed a temporary effectof phage addition on mortality and larval mortality reached the same level as in the bacterial challenges(without phage) after 4–10 days. Since the phage was maintained in high concentrations throughoutthe experiment, it is likely that strain PF430-3 was protected against infection, which supports previousobservations that strain PF430-3 can reduce its susceptibility to phage KVP40 by forming aggregates orbiofilm, creating spatial refuges [20].

In addition to the specific V. anguillarum pathogens, other pathogens already associated withthe fish eggs prior to the experiments were present in the experiments. This allowed an assessmentof the effects of phages on both the mortality caused by the V. anguillarum strains and the mortalityimposed by the natural background pathogen communities. The decrease in mortality recorded for allthe phage controls (without V. anguillarum) compared to the nonchallenge controls (without phageand V. anguillarum) demonstrated a strong effect of phage KVP40 on the initial bacterial pathogencommunities associated with the eggs. This was supported by the observation that >85% of the isolatedcolonies originating from the background bacterial community were susceptible to phage KVP40.

Despite the large fraction of phage susceptible strains, the bacterial abundance increased in allthe incubations over time, and only in cod challenge trial 2 did addition of phage KVP40 reduce thebacterial abundance for multiple days. This suggested that during the experiment, pathogens thatwere not infected by KVP40 (i.e., non-Vibrio pathogens and possibly phage-resistant V. anguillarumstrains) replaced the phage susceptible strains, and thus were the main cause of mortality in theexperiments. This was supported by the increased effects of phages on mortality in cod challengetrial 2, where the eggs were pretreated with 25% glutaraldehyde. These results emphasized that thegrowth of other pathogens than V. anguillarum was the main cause of mortality in the experimentsthat were not pretreated with glutaraldehyde, and that phage KVP40 was able to significantly reducemortality imposed by the added V. anguillarum strains.

Consequently, even though the presence of a bacterial background pathogen community maskedthe effect of phage KVP40 on the added V. anguillarum strains, it at the same time provided a more

Antibiotics 2018, 7, 42 11 of 15

realistic demonstration of how the addition of phage KVP40 will affect an infected aquaculturesystem. These results emphasized the potential of phage KVP40 to control not only the added hoststrains but also a broader range of pathogens present in the rearing facilities. Similar results wereobtained for the two broad-host-range KVP40-like phages ϕSt2 and ϕGrn1 infecting the fish pathogensV. alginolyticus [21]. These phages were able to reduce the natural Vibrio load present in Artemialive feed cultures used in fish hatcheries. The current study is, however, the first demonstrationof a positive effect of phage application on larval survival by reducing the natural microbiota,rather than exclusively focusing on the effects of one added pathogen. While the composition ofthe background microbiota was not analyzed in the current study, previous studies have foundthat bacterial communities associated with cod and turbot eggs in rearing units were dominatedby Pseudomonas, Alteromonas, Aeromonas, and Flavobacterium [22], but also Vibrio has been shownto be prevalent in these environments [23]. In our study, the high fraction of bacteria growing onVibrio-selective TCBS medium combined with the high susceptibility to phage KVP40 suggested thatthe background bacterial community was dominated by Vibrio or Vibrio-related species, as the phageKVP40 has been shown to infect at least eight Vibrio species and one Photobacterium [16]. This wasalso supported by preliminary analysis of the microbiome associated with the turbot eggs usedin challenge trial 2, which showed dominance of Vibrio species (Dittmann, unpublished results).The differences in mortality in the control treatments (nonchallenged control and KVP40 control)between different experiments may therefore reflect differences in the composition of backgroundbacterial community, representing differences in virulence and KVP40 susceptibility. Further, higherincubation temperature of the turbot than cod eggs may also have increased bacteria-induced mortalityin the turbot experiments. In one of the treatments (cod challenge trial 1 with strain 90-11-286),addition of phage KVP40 increased larval mortality (Figure 4c). Specific secondary metabolites ortoxins released during cell lysis may potentially inhibit larval growth [24]. However, since this wasnot observed in any of the other treatments, it is not likely that the viral lysates affected the cod larvae.Alternatively, the viral lysates may have stimulated growth of other specific pathogens already presentin the experiment, as also indicated by the enhanced bacterial growth in the phage added culture(Figure 6c). Previous studies have shown that lysogenization of V. harveyi with phage VHS1 increasedthe virulence of the bacterium against black tiger shrimp (Penaus monodon) by the phage encoded toxinassociated with hemocyte agglutination ([25]). There has not been any indication of lysogenization ofVibrio pathogens with phage KVP40, and the production of a KVP40-encoded toxin is therefore nota likely explanation for the observed increase in larval mortality in this experiment.

Our results support previous attempts to control pathogens in aquaculture by use of phages.A challenge trial in Atlantic salmon using V. anguillarum strain PF4, a close relative to strain PF430-3used in the current study [13], showed 100% survival using the phage CHOED, independent of theoriginal multiplicity of infection (MOI) [10]. The efficiency of this phage on fish survival comparedto the current study most likely relates to the fact that larger fish are more robust against infectionsby co-occurring pathogens than larvae. A delay in mortality after phage addition was also observedby Imbeault et al. [26] and Verner-Jeffreys et al. [27] in brook trout and Atlantic salmon, respectively,infected with A. salmonicida using different phages. While Imbeault et al. [26] were able to delay theonset of disease and reduce the mortality to 10%, Verner-Jeffreys et al. [27] also demonstrated a delayin the mortality, but only observed a temporary effect of the phages in survival.

Previous in vivo challenge studies with a positive outcome of phage therapy were conductedon >5 day old larvae [12] or fish averaging 15–25 grams [10], while our study was conducted oneggs which hatched during the course of the challenge trials. Eggs and newly hatched larvae aremore sensitive to the infection by pathogenic V. anguillarum and other pathogens than late stagesdue to the inefficient protection provided by the intestinal microflora associated with their gutmucosa, which constitutes a primary barrier [28]. Despite the general frailty of newly hatchedlarvae, we demonstrated a significant phage-mediated reduction in mortality of cod and turbotlarvae in experimental challenge trials with V. anguillarum pathogens in combination with the natural

Antibiotics 2018, 7, 42 12 of 15

pathogenic bacteria associated with the incubated fish eggs. These results emphasize that phagetherapy is a promising approach to reduce pathogen load and mortality in marine larviculture.

4. Materials and Methods

4.1. Bacterial Strains and Growth Conditions

The four V. anguillarum strains—PF4303-3, PF7, 90-11-286, and 4299—used in this study wereisolated in Chile, Denmark, and Norway [10,13,29,30]. The bacteria were stored at −80 ◦C inLuria-Bertani (LB) medium with 15% glycerol. Before each assay, the strains were inoculated onLB plates and grown overnight at 24 ◦C. Then, one colony was transferred to 4 mL LB medium andgrown overnight at 24 ◦C with agitation (200 rpm).

4.2. Phage Infectivity and Production

The broad-host-range phage KVP40 [16], which previously has been shown to infect the V. anguillarumstrains PF430-3, 90-11-286, and 4299 [13], was tested on V. anguillarum strain PF7 using the double-layeragar assay [14] with minor modifications. The double-layer agar assay in brief: 100 μL phage lysate wasmixed with 300 μL bacterial cells and incubated for 30 min at 24 ◦C. The mixture was added to 4 mL of45 ◦C top agar (LB with 0.4% agar) and poured onto a LB 1.5% agar plate, which was placed for incubationat 24 ◦C overnight. The next day, the presence of phages in the form of clear plaques in the top agar wasdetected. KVP40 was produced and purified by ACD Pharmaceuticals AS (Leknes, Norway).

4.3. Eggs and Larvae

Eggs from turbot and cod were used in the challenge trials. The eggs for turbot challenge trial 1were obtained from Stolt Sea Farm (Galicia, Spain), with 48 h of transport before conducting thechallenge trial at the University of Bergen (Bergen, Norway). The eggs for turbot challenge trial 2 wereobtained from France Turbot, hatchery L’Epine (Noirmoutier Island, France), with 24 h of transportbefore conducting the challenge trial at the Technical University of Denmark (Lyngby, Denmark).The eggs for cod challenge trial 1 and cod challenge trial 2 were obtained from the Institute of MarineResearch, Austevoll Research Station (Storebø, Norway), with 1 hour of transport before conductingthe challenge trial at the University of Bergen (Bergen, Norway). The eggs in cod challenge trial 2 weredisinfected with 25% glutaraldehyde at the Institute of Marine Research, Austevoll Research Stationbefore being transported to the University of Bergen for the challenge trial.

4.4. Phage Therapy Assays

Challenge trials with turbot and cod larvae were established as outlined in Table 3. For each of theV. anguillarum strains tested, eggs were distributed in 10 24-well dishes with 2 mL sterile filtered (0.2 μm)and autoclaved, oxygenated 80% sea water and 1 egg well−1. In group 1 (V. anguillarum only), five 24-wellplates were inoculated with 100 μL V. anguillarum culture in each well. Prior to addition, the bacterialculture had been grown overnight, washed twice in sterile sea water (ssw), and resuspended in ssw toa final concentration of 0.5–1 × 106 CFU mL−1. In group 2 (V. anguillarum + phage KVP40), five 24-wellplates were inoculated with V. anguillarum as above and 50 μL of phage KVP40 was added to each wellto a final concentration of 0.5–8 × 108 PFU mL−1, resulting in a multiplicity of infection (MOI) of ~5–100.The five 24-well plates in group 3 (nonchallenged control) were only inoculated with 100 μL autoclaved,oxygenated 80% ssw, whereas in group 4 (phage KVP40 control), each well also contained 50 μL of phageKVP40. Plates were then incubated in an air-conditioned room of 15.5 ◦C and 5.5 ◦C for turbot and cod,respectively, which are optimal conditions for larval development in the two species. The eggs in groups 1,2, and 4 had bacteria and/or phages added to them immediately after their distribution in the wells (=day 0of the experiment). Due to large variation in the viability of the eggs used for the experiment, the challengetrials were done twice for both fish species in an attempt to confirm the results at different egg qualities.The challenge trials lasted for 8 days for turbot challenge trial 1, 5 days for turbot challenge trial 2, and for

Antibiotics 2018, 7, 42 13 of 15

11 days for cod. The mortality was monitored daily. The quality of the eggs varied considerably dependingon transportation time and handling, resulting in differences in egg mortality prior to hatching. The initialegg mortality was calculated for each 24-well plate and then averaged for all 50 24-well plates used in theindividual experiments. Of the 1200 eggs used in each experiment, the average fraction of eggs that diedprior to hatching amounted to 0% and 30.3% in turbot challenge trials 1 and 2, respectively, and 4.9% and23.2% in cod challenge trials 1 and 2, respectively. These eggs were excluded from the analysis. The effectof phage addition on larval mortality was calculated as a relative reduction [31], corresponding to thereduction in mortality in treatments to which both phage KVP40 and V. anguillarum were added relativeto the mortality in treatments with V. anguillarum alone (i.e., the difference in mortality between the twotreatments in percentage of the mortality in the incubations without phage.

Table 3. Experimental design and addition V. anguillarum and phage KVP40.

Group TreatmentV. anguillarum(CFU mL−1)

Phage KVP40(PFU mL−1)

Replicate Wells

1 V. anguillarum only 0.5–1 × 106 - 5 × 24 wells × 4 strains2 V. anguillarum + phage KVP40 0.5–1 × 106 0.5–12 × 108 5 × 24 wells × 4 strains3 Nonchallenge control - - 5 × 24 wells4 Phage KVP40 control - 0.5–12 × 108 5 × 24 wells

The concentration of bacteria and phages was monitored daily except in turbot challenge trial 1,where neither was monitored. In turbot challenge trial 2, the concentrations were only monitoredfor half of the experiment, while the phage concentration was only monitored for 3 days in codchallenge trial 1. To determine the bacterial concentration, dilutions were inoculated on LB agarplates (in cod challenge trial 2, the dilutions were inoculated on marine agar plates and on selectivethiosulfate-citrate-bile salts-sucrose (TCBS) plates), which incubated overnight at 24 ◦C. To determinethe phage concentration, the double-layer agar assay was used as described earlier. The culturemedium was LB, the host strain was V. anguillarum strain PF430-3 ΔvanT [19], and the plates wereincubated overnight at 24 ◦C.

4.5. Bacterial Background Community and Susceptibility Assays

In order to characterize the bacterial background, different media were used in the challengetrials. The water used for the transport of the eggs in turbot challenge trial 1 was spread on TCBSplates at day 4. A total of 40 colonies were picked and transferred to LB medium and grown overnightat 24 ◦C with agitation (200 rpm). The bacteria had their optical density at 600 nm (OD600), measuredusing Novaspec Plus Visible Spectrophotometer after 1 hour in the presence and in the absence ofKVP40. The sterile 80% sea water with the live nonchallenged control larvae in turbot challenge trial 2were inoculated on LB, TCBS, and marine agar plates at day 11. The plates incubated overnight at24 ◦C before determining the bacterial concentration. Throughout cod challenge trial 2, the bacterialconcentration was determined on both marine agar and TCBS plates.

4.6. Statistical Analysis

Differences between challenged larvae with and without phage therapy and between the controls(nonchallenge control and KVP40 control) for each time point were analyzed by chi-squared testsusing the software R (R foundation for statistical computing). A value of p < 0.05 were consideredstatistically significant.

5. Conclusions

The significant positive effect of phage KVP40 on larval survival during hatching and initialgrowth observed in the current experiment demonstrates the potential in using phages to reducepathogen load in cod and turbot hatcheries and may also be a strategy to improve egg quality and

Antibiotics 2018, 7, 42 14 of 15

survival during transport from egg producers to hatcheries. It is obvious, however, that the effect ofthe phage addition on mortality is temporary, and we suggest that a more efficient and long-termcontrol of the pathogens may be obtained using a cocktail of different phages that target a broaderrange of pathogens.

Author Contributions: N.R., A.R. and M.M. designed the experiments; N.R. and A.R. performed turbot challengetrial 1 and cod challenge trial 1, P.G.K. and B.B.R. performed turbot challenge trial 2, N.R., A.R., P.G.K. andB.B.R. performed cod challenge trial 2; N.R. and M.M. analyzed the data; K.E.-S., H.P.K., H.I.W., L.G. and M.M.contributed reagents/materials/analysis tools; N.R. and M.M. wrote the paper with contributions from all authors.

Acknowledgments: The study was supported by the Danish Council for Strategic Research (ProAqua project12-132390) and the Danish Research Council for Independent Research (Project # DFF-7014-00080).

Conflicts of Interest: The authors declare no conflict of interest.

References

1. Thompson, F.L.; Iida, T.; Swings, J. Biodiversity of Vibrios. Microbiol. Mol. Biol. Rev. 2004, 68, 403–431. [CrossRef][PubMed]

2. Actis, L.A.; Tolmasky, M.E.; Crosa, J.H. Vibriosis. In Fish Diseases and Disorders; Woo, P.T.K., Bruno, D.W., Eds.;CAB International: Oxfordshire, UK, 2011; pp. 570–605.

3. Frans, I.; Michiels, C.W.; Bossier, P.; Willems, K.A.; Lievens, B.; Rediers, H. Vibrio anguillarum as a fishpathogen: Virulence factors, diagnosis and prevention. J. Fish Dis. 2011, 34, 643–661. [CrossRef] [PubMed]

4. Pereira, C.; Silva, Y.J.; Santos, A.L.; Cunha, A.; Gomes, N.C.M.; Almeida, A. Bacteriophages with potentialfor inactivation of fish pathogenic bacteria: Survival, host specificity and effect on bacterial communitystructure. Mar. Drugs 2011, 9, 2236–2255. [CrossRef] [PubMed]

5. Karunasagar, I.; Pai, R.; Malathi, G.R.; Karunasagar, I. Mass mortality of Penaeus monodon larvae due toantibiotic-resistant Vibrio harveyi infection. Aquaculture 1994, 128, 203–209. [CrossRef]

6. Bricknell, I.R.; Bowden, T.J.; Verner-Jeffreys, D.W.; Bruno, D.W.; Shields, R.J.; Ellis, A.A.E. Susceptibility ofjuvenile and sub-adult Atlantic halibut (Hippoglossus hippoglossus L.) to infection by Vibrio anguillarum andefficacy of protection induced by vaccination. Fish Shellfish Immunol. 2000, 10, 319–327. [CrossRef] [PubMed]

7. Mikkelsen, H.; Lund, V.; Larsen, R.; Seppola, M. Vibriosis vaccines based on various sero-subgroups of Vibrioanguillarum O2 induce specific protection in Atlantic cod (Gadus morhua L.) juveniles. Fish Shellfish Immunol. 2011.[CrossRef] [PubMed]

8. Vinod, M.G.; Shivu, M.M.; Umesha, K.R.; Rajeeva, B.C.; Krohne, G.; Karunasagar, I.; Karunasagar, I.Isolation of Vibrio harveyi bacteriophage with a potential for biocontrol of luminous vibriosis in hatcheryenvironments. Aquaculture 2006. [CrossRef]

9. Karunasagar, I.; Shivu, M.M.; Girisha, S.K.; Krohne, G.; Karunasagar, I. Biocontrol of pathogens in shrimphatcheries using bacteriophages. Aquaculture 2007, 268, 288–292. [CrossRef]

10. Higuera, G.; Bastías, R.; Tsertsvadze, G.; Romero, J.; Espejo, R.T. Recently discovered Vibrio anguillarumphages can protect against experimentally induced vibriosis in Atlantic salmon, Salmo salar. Aquaculture2013, 392, 128–133. [CrossRef]

11. Mateus, L.; Costa, L.; Silva, Y.J.; Pereira, C.; Cunha, A.; Almeida, A. Efficiency of phage cocktails in theinactivation of Vibrio in aquaculture. Aquaculture 2014, 424, 167–173. [CrossRef]

12. Silva, Y.J.; Costa, L.; Pereira, C.; Mateus, C.; Cunha, Â.; Calado, R.; Gomes, N.C.M.; Pardo, M.A.; Hernandez, I.;Almeida, A. Phage therapy as an approach to prevent Vibrio anguillarum infections in fish larvae production.PLoS ONE 2014. [CrossRef] [PubMed]

13. Tan, D.; Gram, L.; Middelboe, M. Vibriophages and their interactions with the fish pathogen vibrioanguillarum. Appl. Environ. Microbiol. 2014, 80, 3128–3140. [CrossRef] [PubMed]

14. Stenholm, A.R.; Dalsgaard, I.; Middelboe, M. Isolation and characterization of bacteriophages infectingthe fish pathogen Flavobacterium psychrophilum. Appl. Environ. Microbiol. 2008, 74, 4070–4078. [CrossRef][PubMed]

15. Letchumanan, V.; Chan, K.G.; Pusparajah, P.; Saokaew, S.; Duangjai, A.; Goh, B.H.; Ab Mutalib, N.S.; Lee, L.H.Insights into bacteriophage application in controlling vibrio species. Front. Microbiol. 2016, 7. [CrossRef][PubMed]

Antibiotics 2018, 7, 42 15 of 15

16. Matsuzaki, S.; Tanaka, S.; Koga, T.; Kawata, T. A broad-host-range vibriophage, KVP40, isolated from seawater. Microbiol. Immunol. 1992. [CrossRef]

17. Inoue, T.; Matsuzaki, S.; Tanaka, S. A 26-kDa outer membrane protein, OmpK, common to Vibrio species isthe receptor for a broad-host-range vibriophage, KVP40. FEMS Microbiol. Lett. 1995, 125, 101–105. [CrossRef][PubMed]

18. Labrie, S.J.; Samson, J.E.; Moineau, S. Bacteriophage resistance mechanisms. Nat. Rev. Microbiol. 2010, 8, 317–327.[CrossRef] [PubMed]

19. Tan, D.; Svenningsen, S.L.; Middelboe, M. Quorum sensing determines the choice of antiphage defensestrategy in Vibrio anguillarum. MBio 2015, 6, e00627. [CrossRef] [PubMed]

20. Tan, D.; Dahl, A.; Middelboe, M. Vibriophages differentially influence biofilm formation by Vibrio anguillarumstrains. Appl. Environ. Microbiol. 2015, 81, 4489–4497. [CrossRef] [PubMed]

21. Kalatzis, P.G.; Bastías, R.; Kokkari, C.; Katharios, P. Isolation and characterization of two lytic bacteriophages,ϕst2 and ϕgrn1; Phage therapy application for biological control of vibrio alginolyticus in aquaculture livefeeds. PLoS ONE 2016, 11. [CrossRef] [PubMed]

22. Hansen, G.H.; Olafsen, J.A. Bacterial colonization of cod (Gadus morhua L.) and halibut (Hippoglossus hippoglossus)eggs in marine aquaculture. Appl. Environ. Microbiol. 1989, 55, 1435–1446. [PubMed]

23. Austin, B. Taxonomy of bacteria isolated from a coastal marine fish-rearing unit. J. Appl. Bacteriol. 1982, 53, 253–268.[CrossRef]

24. Goodridge, L.D. Designing phage therapeutics. Curr. Pharm. Biotechnol. 2010, 11, 15–27. [CrossRef] [PubMed]25. Khemayan, K.; Prachumwat, A.; Sonthayanon, B.; Intaraprasong, A.; Sriurairatana, S.; Flegel, T.W. Complete

genome sequence of virulence-enhancing siphophage VHS1 from Vibrio harveyi. Appl. Environ. Microbiol.2012, 78, 2790–2796. [CrossRef] [PubMed]

26. Imbeault, S.; Parent, S.; Lagacé, M.; Carl, F.; Blais, J. Using bacteriophages to prevent furunculosis caused byAeromonas salmonicida in farmed brook trout. J. Aquat. Anim. Health 2006, 18, 203–214. [CrossRef]

27. Verner–Jeffreys, D.W.; Algoet, M.; Pond, M.J.; Virdee, H.K.; Bagwell, N.J.; Roberts, E.G. Furunculosis in Atlanticsalmon (Salmo salar L.) is not readily controllable by bacteriophage therapy. Aquaculture 2007, 270, 475–484.[CrossRef]

28. Hansen, G.H.; Olafsen, J.A. Bacterial interactions in early life stages of marine cold water fish. Microb. Ecol.1999, 38, 1–26. [CrossRef] [PubMed]

29. Skov, M.N.; Pedersen, K.; Larsen, J.L. Comparison of pulsed-field gel electrophoresis, ribotyping, and plasmidprofiling for typing of Vibrio anguillarum serovar O1. Appl. Environ. Microbiol. 1995, 61, 1540–1545. [PubMed]

30. Mikkelsen, H.; Schrøder, M.B.; Lund, V. Vibriosis and atypical furunculosis vaccines; efficacy, specificity andside effects in Atlantic cod, Gadus morhua L. Aquaculture 2004. [CrossRef]

31. Ranganathan, P.; Pramesh, C.; Aggarwal, R. Common pitfalls in statistical analysis: Absolute risk reduction,relative risk reduction, and number needed to treat. Perspect. Clin. Res. 2016, 7. [CrossRef] [PubMed]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open accessarticle distributed under the terms and conditions of the Creative Commons Attribution(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Paper 4

Rasmussen BB, Kalatzis PG, Middelboe M, Gram L. 2019.

Combining probiotic Phaeobacter inhibens DSM17395 and broad-host-

range vibriophage KVP40 against fish pathogenic vibrios.

Submitted manuscript

1

Combining probiotic Phaeobacter inhibens DSM17395 and broad-host-range

vibriophage KVP40 against fish pathogenic vibrios

Bastian Barker Rasmussena, Panos G. Kalatzisb,c, Mathias Middelboeb and Lone Grama,*

a. Department of Biotechnology and Biomedicine, Technical University of Denmark, Søltofts plads bldg. 221,

DK-2800 Kgs. Lyngby, Denmark

b. Marine Biological Section, University of Copenhagen, Strandpromenaden 5, DK-3000 Helsingør, Denmark

c. Institute of Marine Biology, Biotechnology and Aquaculture, Hellenic Centre for Marine Research, Former

American Base of Gournes, Heraklion 71003, Crete, Greece

* corresponding author: e-mail: gram@bio.dtu.dk, phone: +45 4525 2586

Keywords: Roseobacter, Phaeobacter, Probiotics, Phage therapy, Vibrio, Live feed

Running-title: Combining probiotic bacteria and phage therapy

24th April 2019

2

ABSTRACT 1

Opportunistic pathogenic bacteria readily proliferate in live feed cultures used to feed marine fin fish 2

larvae. As a consequence, larvae cultures fed infected live feed can suffer devastating larvae mortalities. 3

Antibiotics are still used to control these pathogenic bacteria, but due to the risk of antibiotic resistance, 4

sustainable alternatives are required. Phage therapy and probiotic bacteria are examples of alternatives 5

which are currently being developed. However, regrowth of phage tolerant bacteria or reduced inhibition 6

by probiotic bacteria in non-axenic systems, have been seen. The purpose of this study was to investigate if 7

combining the broad-host-range vibriophage KVP40 and probiotic bacteria Phaeobacter inhibens 8

DSM17395 would result in an enhanced inhibitory effect against the fish pathogenic Vibrio anguillarum and 9

protection of live feed and cell lines. The combined and individual treatments were tested in non-axenic 10

live feed systems (i.e. Tetraselmis suecica algae and Artemia salina brine shrimps) obtained from a 11

commercial larval rearing unit. Further, we determined if the probiont and/or phages could protect CHSE-12

214 fish cell line challenged with V. anguillarum. P. inhibens, KVP40 and their combination reduced vibrio 13

counts in the T. suecica cultures with one to two orders of magnitude below inoculum level. Combining 14

KVP40 and P. inhibens resulted in a lower level of V. anguillarum than using the probiont alone. In the 15

Artemia cultures, P. inhibens alone or in combination with KVP40 reduced vibrio counts and Artemia 16

mortality relative to the control, while KVP40 alone only delayed vibrio growth and Artemia mortality. 17

Combining KVP40 and P. inhibens caused an immediate killing of V. anguillarum, however, it did not 18

provide extra inhibitory potential and only little extra protection of Artemia, as compared to using only P. 19

inhibens.. Our results demonstrate that probiotic bacteria and phages in commercial, non-axenic live feed 20

can reduce levels of fish pathogenic Vibrio. The combinations of phage and probiont only resulted in a 21

marginal improved pathogen reduction, however, this could and should be further optimized. 22

Words = 320 Max 400 words 23

3

1. INTRODUCTION 24

Bacterial diseases are the most common infectious diseases in marine aquaculture and a major obstacle in 25

the farming of marine animals (Kibenge et al., 2012; Toranzo et al., 2005). The bacterial family Vibrionaceae 26

which includes several human pathogenic agents also includes many important pathogens causing disease 27

in commercially important crustacean, shellfish and fish (Thompson et al., 2004; Toranzo et al., 2005). 28

Vibrio anguillarum is the causative agent of the hemorrhagic septicemia disease vibriosis and can infect 29

more than 50 species of fish (Frans et al., 2011). It is a key challenge in the rearing of many marine fin fish 30

species when the larvae first start feeding (Reid et al., 2009). Several fin fish species are fed live feed, 31

including microalgae, rotifers and Artemia, due to lack of suitable artificial feed formula. Live feed 32

introduces large amounts of nutrients into the larvae rearing systems and this may promote proliferation of 33

fast-growing opportunistic bacteria such as V. anguillarum (Skjermo and Vadstein, 1999). Furthermore, live 34

feed can act as infection vectors for pathogenic bacteria (Olafsen, 2001; Rasmussen et al., 2018). 35

36

Therapeutic and prophylactic use of antibiotics has previously been used routinely to control pathogenic 37

bacteria in aquaculture (Defoirdt et al., 2011). Although the use of antibiotics has been heavily reduced in 38

Europe and North America mainly due to deployment of vaccines, it is still used in fish larvae cultures since 39

they cannot effectively be vaccinated due to an immature immune systems (Defoirdt et al., 2011; Ringø et 40

al., 2014; Sommerset et al., 2005). Bacterial antibiotic resistance is a major concern and has been reported 41

in aquaculture environments (Alderman and Hastings, 1998; Holmstrom et al., 2003; Spanggaard et al., 42

1993); thus, there is a great need for sustainable disease control measures. 43

44

Several biological means of disease control are being investigated, including the use of probiotics, i.e. live 45

microorganisms which exert beneficial effects on the host health (FAO and WHO, 2001). Tropodithietic acid 46

(TDA) producing members of the Roseobacter group such as Phaeobacter spp. inhibit fish pathogenic 47

vibrios (D’Alvise et al., 2012; Hjelm et al., 2004; Planas et al., 2006; Ruiz-Ponte et al., 1999) and can protect 48

4

vibrio challenged turbot and cod larvae (D’Alvise et al., 2013, 2012; Planas et al., 2006) as well as live feed 49

i.e. rotifers and Artemia (D’Alvise et al., 2012; Grotkjær et al., 2016b, 2016a). Phaeobacter spp. is part of 50

the indigenous aquaculture microbiota (Grotkjær et al., 2016b; Hjelm et al., 2004; Porsby et al., 2008; Ruiz-51

Ponte et al., 1998), and resistance to their antibacterial compound (TDA) does not develop easily (Porsby et 52

al., 2011; Rasmussen et al., 2016) supporting their use in aquaculture. In axenic live feed systems, 53

Phaeobacter spp. reduces the level of fish pathogens (D’Alvise et al., 2012, 2010; Grotkjær et al., 2016b), 54

however, in non-axenic live feed systems the organism only prevents growth of the pathogen (Grotkjær et 55

al., 2016a; Rasmussen et al., 2018). Whilst keeping numbers low (104 CFU mL-1) is sufficient to reduce 56

infection by many V. anguillarum strains (Rønneseth et al., 2017), a further reduction of the pathogen 57

would be preferable. 58

59

Another alternative to antibiotics is lytic bacteriophages, the so-called phage therapy (Almeida et al., 2009; 60

Castillo et al., 2012; Oliveira et al., 2012). Marine viruses are the most abundant biological entities in the 61

marine environment and bacteriophages (viruses that infect bacteria) play an important role in shaping the 62

microbiomes (Breitbart, 2012; Brussaard et al., 2008; Rohwer and Thurber, 2009; Suttle, 2007). Phage 63

therapy can be effective in reducing pathogen density and/or mortality of production animals (Karunasagar 64

et al., 2007; Nakai et al., 1999; Park and Nakai, 2003; Rørbo et al., 2018). Unlike antibiotics, bacteriophages 65

are highly host specific which enables targeting of specific pathogens, while leaving the overall microbial 66

community intact (Oliveira et al., 2012). The target of phage therapy can be broadened by using “phage 67

cocktails” with several different phages or using broad host range phages. The broad host range 68

vibriophage KVP40 can infect seven species of Vibrio and one species of Photobacterium (Matsuzaki et al., 69

1992) and can also reduce and/or delay the mortality of cod and turbot larvae challenged with V. 70

anguillarum (Rørbo et al., 2018). However, a challenge in phage therapy is non-mutational as well as 71

mutational bacterial phage resistance or tolerance (Castillo et al., 2019, 2015; Tan et al., 2015) which may 72

develop within hours of treatment (Defoirdt et al., 2011; Tan et al., 2014). 73

5

74

Given the lytic ability of phages, we hypothesized that combining bacteriophages and probiotic bacteria 75

could result in rapid killing and longer term inhibition of fish pathogens and potentially reduce problems 76

with phage resistant bacteria as they would be killed or reduced by the probiont. We here report a series of 77

experiments in non-axenic live feed systems (i.e. Tetraselmis suecica algae and Artemia salina brine 78

shrimps) investigating this proposal. 79

80

2. MATERIALS AND METHODS 81

2.1. Bacterial strains, bacteriophages and culture conditions 82

The highly virulent Vibrio anguillarum strain 90-11-286 (Rønneseth et al., 2017) was used as pathogen in 83

live feed systems and in the CHSE-214 fish cell protection assay (Lannan et al., 1984). Strain 90-11-286 was 84

isolated from diseased rainbow trout (Oncorhynchus mykiss) from a Danish fish farm (Pedersen and Larsen, 85

1993; Skov et al., 1995). P. inhibens DSM17395 (Buddruhs et al., 2013; Martens et al., 2006; 86

Vandecandelaere et al., 2008) was used as probiont, and the broad host range vibriophage KVP40 was used 87

for the phage therapy part (Matsuzaki et al., 1992). KVP40 was produced and purified by ACD 88

Pharmaceuticals AS (Leknes, Norway). Strain 90-11-286 was grown and enumerated on TSA (Tryptone Soy 89

Agar, Difco 212185) on which it grows very fast and has very distinct colony morphology. P. inhibens DSM 90

17395 does not grow on TSA due to the low salinity of the substrate and 90-11-286 is able to outgrow the 91

background microbiota of the non-axenic systems on the TSA plates. Thus, it is possible to enumerate vibrio 92

despite it being outnumbered by the background microbiota by two to three orders of magnitude 93

(Rasmussen et al., 2018). P. inhibens was grown and enumerated on MA (Marine Agar, Difco 2216) where it 94

forms distinct brown colonies due to Fe-TDA complex (D’Alvise et al., 2016). A Plaque Forming Unit (PFU) 95

assay using the double-layer agar method (Kutter, 2009) was performed in order to calculate the titer of 96

KVP40. Briefly, 10 μL drops of 10-fold serially diluted phage containing samples were spotted on V. 97

anguillarum strain PF430-3 ΔvanT (Tan et al., 2015) bacterial lawns, prepared on sea salt agar plates (1% 98

6

Tryptone, 0.5% Yeast extract, 2% Sigma Sea Salts, 1.5% agar). Following overnight incubation at room 99

temperature (~25° C), PFUs were enumerated on the bacterial lawn. Bacterial pre-cultures were grown for 100

24 h in 20 mL of half-strength YTSS (½YTSS) (2 g Bacto Yeast extract, 1.25 g Bacto Tryptone, 20 g Sigma Sea 101

Salts, 1 L deionized water) (Sobecky et al., 1997) at 25° C and 200 rpm. 102

103

2.2. Preparation of non-axenic Tetraselmis suecica and Artemia cultures 104

A non-axenic Tetraselmis suecica stock culture acquired from a commercial aquaculture unit was 105

maintained in f/2 medium (Guillard, 1975; Guillard and Ryther, 1962) without Na2SiO3 but with 5 mM NH4Cl 106

in 3 % IO (Instant Ocean salts, Aquarium Systems Inc.) (f/2-Si+NH4) at 18° C and 24 μmol photons m-2 s-1, 107

photosynthetically active radiation (PAR). Density of T. suecica was determined by counting in a Neubauer-108

improved counting chamber (Assistent). The cultures were adjusted to approximately 105 cell mL-1 in f/2-109

Si+NH4 and 25 mL distributed in to 50 mL falcon tubes. Artemia salina cysts were acquired from the same 110

commercial aquaculture unit as T. suecica and were stored at 5° C until use. The day before the experiment, 111

20 mg non-disinfected Artemia cysts were add to 20 ml 3 % IO and incubated at 20 °C and 200 rpm. The 112

experiment was conducted in 50 mL falcon tubes with 25 mL 3 % IO with approximately 1.25 Artemia mL-1. 113

114

2.3. Co-culture experiment in non-axenic Tetraselmis suecica and Artemia systems 115

250 μL of 10-fold serially diluted V. anguillarum 90-11-286 pre-cultures was used to inoculate non-axenic T. 116

suecica and Artemia cultures, aiming at an initial concentration of 104 CFU mL-1. Undiluted P. inhibens DSM 117

17395 preculture was added to the non-axenic T. suecica and Artemia cultures, aiming at an initial 118

concentration of 107 CFU mL-1. 250 μL ½YTSS was added to cultures without DSM 17395 and un-inoculated 119

controls. KVP40 were inoculated from the stock aiming at 107 PFU ml-1 equivalent to a multiplicity of 120

infection (MOI) of 1000. The cultures were incubated at 18° C, lying horizontally on a rotary shaker at 60 121

rpm. Experiments were carried out over 96 h and bacterial, phage, and algae abundances, and Artemia 122

mortality were determined every 24 h. Furthermore, bacterial and phage abundances were also 123

7

determined after 5 and 10 h. Bacterial abundances were determined by colony counts and phage 124

abundance was determined by PFU counts. TSA plates for vibrio counts and MA plates for P. inhibens 125

counts were incubated at 25° C and counted after 1-2 days and 2-3 days, respectively. The abundance of T. 126

suecica were counted in a Neubauer-improved counting chamber. The number of dead Artemia was 127

determined using a Sedgewick-Rafter counting chamber. Likewise, after the experiment was terminated, 128

the surviving Artemia were killed and counted in a Sedgewick-Rafter counting chamber. All experiments 129

were carried out simultaneously in biological triplicates. At the time points 2 days, V. anguillarum colonies 130

were picked from the TSA plates and streaked on to new TSA plates for phage susceptibility tests (see 2.4.). 131

132

2.4. Susceptibility of Vibrio anguillarum isolates to phage infection 133

Susceptibility of V. anguillarum colonies to KVP40 was evaluated after 2 days in all different vibrio-134

containing treatments. A total of eight different colonies were picked from each of the four different 135

treatments and phage susceptibility was tested using the double agar layer method (Kutter, 2009). Ten μL 136

drops of KVP40 stock were spotted on bacterial lawns of the bacterial isolates, and the lytic effect was 137

examined following overnight incubation at room temperature (~25° C). 138

139

2.5. Fish cell line protection assay 140

The possible ability of P. inhibens and KVP40 to protect against vibriosis, was investigated using the fish cell 141

line CHSE-214 (catalogue no. 91041114; Sigma) (Lannan et al., 1984). Experiments were carried out 142

simultaneously in biological triplicates. CHSE-214 cells were grown as previously described (Rasmussen et 143

al., 2016). In short, the cells were grown in Leibovitz L-15 medium with glutamine (catalogue no. L1518; 144

Sigma), 10% fetal bovine serum (FBS) (catalogue no. F4135; Sigma) and 1% antibiotic-antimycotic solution 145

(catalogue no. A5955; Sigma) in 6 well plates until a confluent cell layer was formed after which the cells 146

were washed with Hanks’ balanced salt solution (HBSS) (catalogue no. H4641; Sigma). The medium was 147

removed from precultures of P. inhibens and V. anguillarum which were then washed with HBSS-IO+ (10 x 148

8

Hanks’ balanced salt solution resuspend in 0.22 % Instant Ocean salts, resulting in HBSS with approximately 149

1 % salts) and suspensions of an optical density at 600nm (OD600) of 1 was prepared. Wells of three 6 well 150

plates were filled with three and a half ml HBSS-IO+ containing approximately 104 CFU ml-1 V. anguillarum 151

with or without 107 CFU ml-1 P. inhibens and/or 107 PFU ml-1 KVP40. Immediately after inoculation 500 μl 152

were sampled from each well and CFU and PFU counts were performed in order to check inoculum levels. 153

After 24 h at 20°C, the HBSS-IO+ suspensions was removed and the cells were trypsinated in 1 ml trypsin-154

EDTA for 10 min, after which 250μl FBS were added to inactivate the trypsin-EDTA. The dislodged cells were 155

then washed three times in phosphate-buffered saline (PBS) and spun down (400 x g for 1 min at 5°C) after 156

which pellets were resuspended in 500μl PBS. Fish cells were counted in a Neubauer-improved counting 157

chamber and live/dead stained using the Live/Dead Cell Double Staining Kit (04511-1KT-F; Sigma), as 158

described by the manufacturer. Survival of CHSE-214 was accessed in an Olympus BX51 fluorescence 159

microscope at a magnification of x 100. A Neubauer-improved counting chamber was used to define the 160

area. The protection assay was performed in triplicate, and 111 to 376 cells were assessed per sample. 161

162

2.6. Statistical analysis 163

CFU mL-1 values were log10-transformed prior to statistical analysis. Differences in vibrio counts were 164

analysed as a two-factor design for each time point and tested using a Two-way analysis of variance 165

(ANOVA) tests (α ≤ 0.05) in order to test for statistical interaction in the combined treatment i.e. if the 166

variance of one treatment (P. inhibens) have an effect on the variance of the other treatment (KVP40) and 167

vice versa. Furthermore, Tukey’s multiple comparisons test was performed for pairwise comparisons. For 168

vibrio counts below detection level (10 CFU ml-1) in the Artemia co-culture (Fig. 5), a number of 9 CFU ml-1 169

were used for statistical analysis. Differences in survival of non-challenged and vibrio-challenged Artemia 170

were analysed separately using Two-way ANOVA tests (α ≤ 0.05) and Tukey’s multiple comparisons test, 171

and differences in CHSE-214 cell survival were analysed using One-way ANOVA tests (α ≤ 0.05) and Tukey’s 172

9

multiple comparisons test. All statistical analyses were performed in GraphPad Prism 6.07 (GraphPad 173

Software, San Diego CA). 174

175

3. RESULTS 176

3.1. Non-axenic Tetraselmis suecica systems 177

T. suecica numbers were unaffected by the presence of P. inhibens or KVP40 and remained stable at 178

approximately 105 cells mL-1 (data not shown). Also, P. inhibens numbers remained at the inoculum of 107 179

CFU mL-1 throughout the experiment (Fig. 1A), regardless of the addition of vibrio and/or phages. The level 180

of KVP40 was reduced slightly, maximum of 1 log unit, over time (Fig. 1B). V. anguillarum grew to 106 CFU 181

mL-1 after one day in control cultures (Fig. 2) but were significant lowered by the addition of KVP40 and/or 182

P. inhibens. Numbers of vibrio were significantly reduced after 5 hours when KVP40 was added alone (P-183

value <0.0001). V. anguillarum regrew to 103 CFU mL-1 after one day and remained at this level throughout 184

the experiment. Growth of V. anguillarum was prevented when P. inhibens was added and numbers 185

gradually declined two orders of magnitude between day 2 and 4 reaching 102 CFU mL-1. Similarly, vibrio 186

numbers declined two orders of magnitude in the combined treatment (P. inhibens and KVP40). Addition of 187

P. inhibens plus KVP40 caused a significant lower numbers of V. anguillarum than the addition of only P. 188

inhibens at time point 5 hours through 2 days (P-values 0.0473, 0.0004. 0.0018 and 0.0056). There was a 189

significant statistical interaction between P. inhibens and KVP40, in the combined treatment throughout 190

the experiment (P-values <0.0001). The statistical interaction was negative meaning that the reduction in 191

vibrio abundance in the combined treatment was lower than the sum of the reductions in the individual 192

treatments together. 193

194

3.2. Non-axenic Artemia systems 195

P. inhibens remained stable at inoculum level 107 CFU mL-1 in the Artemia cultures (Fig. 3A) regardless of 196

other additions to the cultures. KVP40 counts (Fig. 3B) were significantly higher in the combination cultures 197

10

(P-values <0.0001) and KVP40 plus V. anguillarum cultures (P-values 0.0427, 0.0173 and 0.0133), relative to 198

the control from time points 10 hours and 2 days, respectively. V. anguillarum grew to approximately 5 x 199

106 CFU mL-1 when alone (Fig. 4). Vibrio counts were reduced to or below detection levels in all cultures to 200

which P. inhibens was added (with or without phages) and were significantly lower than vibrio counts in the 201

control cultures (P-values <0.0001 for both treatments). V. anguillarum counts were initially reduced by 2 202

logs in cultures to which only phages were added, but regrew to inoculum levels after 10 hours and 203

reached levels of 5 x 105 CFU mL-1 at the end of the experiment, yet significantly lower than the control (P-204

value 0.0284). Vibrio counts were only significantly lower in the combination compared to both P. inhibens 205

alone and KVP40 alone at time point 1 day (P-values 0.0185 and <0.0001), respectively. A significant 206

negative statistical interaction between P. inhibens and KVP40 in the combined treatment was seen for 207

time points 5 hours, 10 hours, 1 day and 3 days (P-values <0.0001, 0.0002, 0.0362 and 0.0208), respectively, 208

at time point 2 days and 4 days no statistical interactions were seen. Artemia survival in the non-challenged 209

systems (Fig. 5A) varied from 94±6 % to 64±13 % after three days and 47±12 % to 8±1 % after four days. 210

Survival was significantly higher in both non-challenged P. inhibens and KVP40 treated systems, relative to 211

the non-challenged control at day 3 (P-values <0.0001 and 0.0003) and day 4 (P-values <0.0001 and 212

0.0126). Furthermore, survival in the non-challenged P. inhibens treated system was significantly higher 213

that the KVP40 treated after four days (P-value <0.0001). In the vibrio challenged systems (Fig. 5B), survival 214

varied from 94±4 % to 30±6 % after three days and from 65±14 % to 9±4 % after four days. Survival in the 215

vibrio-challenged control was significant lower at 2 and 3 days (P-values 0.0060 and <0.0001) compared to 216

the non-challenged control. Survival was significantly lower in non-treated vibrio-challenged system 217

compared to the rest after two days (P-values 0.0024, 0.0044, 0.0044). After three and four days Artemia 218

survival was significant higher in the vibrio-challenged P. inhibens or P. inhibens plus KVP40 treated systems 219

as compared to non-treated vibrio-challenged and KVP40 treated (P-values <0.0001). After three days and 220

beyond no difference in survival was seen between vibrio-challenged non-treated and KVP40 treated and 221

between the vibrio-challenged P. inhibens only and P. inhibens and KVP40 combi treated systems except for 222

11

day four where survival was significant higher in the P. inhibens and KVP40 combi treated relative to the P. 223

inhibens only treated (P-value 0.0474). 224

225

3.3. Susceptibility of Vibrio anguillarum isolates to phage infection 226

Thirty-two V. anguillarum colonies were isolated from the different Artemia cultures after two days to test 227

the susceptibility of the isolates to phage infection. All isolates were susceptible to infection by KVP40 (data 228

not shown). 229

230

3.4. Fish cell line protection assay 231

CHSE-214 fish cell lines were challenged with V. anguillarum in the presence of P. inhibens and/or KVP40 to 232

determine if they would protect the cells from V. anguillarum. Survival of CHSE-214 cells varied from 96 % ± 233

0 % to 85 % ± 6 % (Suppl Figure 1) with the lowest survival found in the vibrio only challenged. All 234

treatments resulted in significant higher survival than the vibrio only challenged, with 94 % ± 1 % (P-value 235

0.0255) found in the vibrio + phage KVP40, 94 % ± 2 % (P-value 0.0255) in the vibrio + P. inhibens and 96 % 236

± 2 % (P-value 0.0086) in the vibrio + P. inhibens - KVP40 combination. No significant difference in survival 237

were seen between the treatments. Furthermore, no significant difference in survival were seen between 238

the HBSS-IO+ control 96 % ± 0 % and P. inhibens only control 95 % ± 1 %. 239

240

4. DISCUSSION 241

Several alternatives to antibiotic treatment in aquaculture have been or are currently under development. 242

Here we used commercial, non-axenic live feed cultures and demonstrate the efficiency of both phage 243

therapy and probiotic bacteria in inhibiting V. anguillarum and protecting the live feed Artemia. 244

245

Whilst the phage KVP40 caused an initial reduction of V. anguillarum (Fig. 2 and 4), this was followed by 246

regrowth of the pathogen. This could be due to selection for phage resistant vibrios (Tan et al., 2014). 247

12

However, no phage resistant vibrios were subsequently isolated from the Artemia cultures, suggesting that 248

bacterial protection against phage infection was not caused by mutational changes. However, V. 249

anguillarum has several non-mutational protection mechanisms against phage KVP40 (i.e. aggregation and 250

receptor down-regulation (Kalatzis et al., 2017; Tan et al., 2015)), and such mechanisms could explain the 251

observed regrowth. Vibrio regrowth was much more pronounced in Artemia than in T. suecica cultures (Fig 252

4 and 2). This could be due to the higher initial bacterial load in T. suecica cultures (107 cfu/ml) than in the 253

Artemia cultures (104 cfu/ml) competing with the Vibrio. Also, the dead Artemia in the cultures could have 254

supported vibrio growth in these experiment since Artemia are highly nutritious and supports bacterial 255

growth efficiently (López-Torres and Lizárraga-Partida, 2001). 256

257

Phaeobacter spp. can prevent V. anguillarum growth or reduce numbers in different types of larvae culture 258

live feed systems (D’Alvise et al., 2012; Grotkjær et al., 2016b, 2016a; Rasmussen et al., 2018). In previous 259

studies (Grotkjær et al., 2016a; Rasmussen et al., 2018), reduction of vibrio counts have been seen in axenic 260

systems whereas only growth prevention has been observed in non-axenic systems. In the current study, 261

however, we also found a significant reduction in vibrio abundance in non-axenic T. suecica (Fig. 2) and 262

non-axenic Artemia (Fig. 4) systems. The observed differences in inhibition between the different studies in 263

non-axenic systems could be due to differences in the composition of the background microbiota, which 264

may affect the inhibitory effect of P. inhibens. For instance; combining several bacteria from Ulva australis: 265

Microbacterium phyllosphaerae, Shewanella japonica, Dokdonia donghaensis, and Acinetobacter lwoffi 266

increased tolerance to invasion by Pseudoalteromonas tunicata (Burmølle et al., 2006). 267

268

Artemia cysts carry high loads of bacteria which can include pathogens such as Vibrio spp. (Austin and 269

Allen, 1982; Igarashi et al., 1989), and the low survival in the non-challenged controls (Fig. 5A) were thus 270

expected. The protection of the Artemia in the non-challenged systems by P. inhibens and KVP40 indeed 271

indicates the presence of pathogens, including KVP40 susceptible pathogenic vibrios. Addition of V. 272

13

anguillarum significantly accelerated Artemia mortality as seen in the non-treated control cultures. No 273

significant difference was seen in survival of V. anguillarum challenged Artemia to which KVP40 was added, 274

compared to V. anguillarum challenged controls, except for day 2 where survival was significantly higher in 275

Artemia with phage addition. This is consistent with the finding that turbot larvae mortality, and to some 276

extend cod larvae mortality, was delayed but not reduced in challenge trials with V. anguillarum 90-11-286 277

and KVP40 (Rørbo et al., 2018). The fact that KVP40 reduced mortality in non-challenged Artemia but not in 278

the vibrio challenged incubations is consistent with the finding that KVP40 has a low infection efficiency 279

against 90-11-286 (Castillo et al., 2019; Tan et al., 2014), which also explains why an increase in PFU counts 280

were not observed in the V. anguillarum infected T. suecica (Fig. 1B) and Artemia (Fig. 3B) systems. 281

However, the protection of the non-challenged Artemia and delayed mortality of V. anguillarum challenged 282

Artemia supports previous studies and emphasizes the potential of KVP40 to protect the Artemia 283

(Matsuzaki et al., 1992). 284

285

In this study we hypothesized that combining KVP40 and P. inhibens would result in an enhanced effect 286

against V. anguillarum and higher survival of Artemia and CHSE-214 cells, relative to the individual 287

treatments, however, little or no combined effect was seen. KVP40, P. inhibens and the combination, were 288

able to significantly inhibit V. anguillarum in the T. suecica cultures (Fig. 2) and although the combination 289

reduced V. anguillarum counts significantly more at than P. inhibens only until day 3, at no point did the 290

combination show any dramatic advantages over both of the two treatments on their own. A significant 291

negative statistical interaction were seen between P. inhibens and KVP40 in the combined treatments of 292

both experiments meaning that the combined treatment did not result in a Vibrio reduction larger than the 293

sum of the reductions in the individual treatments. In the combination treatments the presence of P. 294

inhibens seemed to reduce the inhibitory effect of KVP40 on V. anguillarum. It is possible that the reduced 295

effect is due to P. inhibens and TDA altering the sensitivity of V. anguillarum towards phage KVP40. 296

Antimicrobials especially in sub-inhibitory concentrations can alter bacterial behaviour including biofilm 297

14

formation, known to be involved in antiphage defence mechanisms in V. anguillarum (Romero et al., 2011; 298

Tan et al., 2015). It is a possibility that TDA may affect expression of the KVP40 phage receptor OmpK 299

(Inoue et al., 1995; Tan et al., 2015) since TDA has been shown to affect the gene encoding another outer 300

membrane protein namely OmpD in V. anguillarum after long term exposure (Rasmussen et al., 2016). If P. 301

inhibens and/or TDA do alter the sensitivity of V. anguillarum to KVP40, combining P. inhibens with other 302

phages or KVP40 with other probionts would potentially make it possible to circumvent the problem. 303

All treatments of Vibrio infected cell lines increased cell survival (Supplementary Fig. 1). No significant 304

differences were seen between cell survival in the P. inhibens only and the HBSS-IO+ control, indicating that 305

the presence of P. inhibens and TDA does not have a negative effect on CHSE-214 cells. No significant 306

differences in protection of the CHSE-214 cells were seen between the treatments, indicating that KVP40, 307

P. inhibens and the combination were equally effective in prevented V. anguillarum from killing CHSE-214. 308

It should be noted that the survival even in the Vibrio-only challenged cell lines was high. 309

310

In conclusion, combining KVP40 and P. inhibens resulted a marginal improved anti-vibrio effect and extra 311

protection of Artemia. Our results suggests that P. inhibens potentially could have a negative effect on the 312

ability of phage KVP40 to inhibit V. anguillarum, however, combining probiotics and phage therapy could 313

and should be further studied. This could for example be done by investigating possible interactions 314

between P. inhibens and KVP40 or combine other probiotic Phaeobacter spp. and/or phages. Although 315

KVP40 was able to initially lower V. anguillarum counts it, unlike P. inhibens, was unable to lower final 316

mortality of the V. anguillarum challenged Artemia, hence favouring the use of P. inhibens over KVP40 as 317

biocontrol against V. anguillarum in continuous live feed cultures. Thus, if the desired effect is a quick 318

reduction of vibrios in for example Artemia prior to feeding them to fish larvae, the phages are superior 319

over the probiont, whereas the probiont is the method of choice if low levels of vibrio during the cultivation 320

of the live feed is the desired goal. Further, the observed immediate (5 h) and efficient (2 logs) control of 321

the vibrio population after addition of KVP40, followed by regrowth of the pathogen, and the negative 322

15

effect of adding both phage and probiont together, suggest that the timing of the combinatory treatment 323

may be important. Consequently, an initial KVP40 treatment providing a fast reduction of the pathogen, 324

followed by a subsequent probiont addition to prevent regrowth of the same or other opportunistic 325

pathogens, may be a feasible strategy for controlling vibrios in live feed cultures. However, further studies 326

are required to evaluate the synergistic potential of such a two-step approach to utilize the different 327

inhibitory properties of phages and probionts, respectively. 328

329

ACKNOWLEDGEMENTS 330

The authors would like to thank Dr. Hans Petter Kleppen (ACD Pharmaceuticals, Norway) for providing 331

KVP40, D.V.M. Nancy Dourala (Selonda, Greece) for providing T. suecica cultures and Artemia cysts and Dr. 332

Mikael Lenz Strube (DTU Bioengineering, Denmark) for inputs regarding statistical analysis. This study was 333

funded by a grant from The Danish Council for Strategic Research | Programme Commission on Health, 334

Food and Welfare (12-132390; ProAqua). 335

336

REFERENCES 337

Alderman, D.J., Hastings, T.S., 1998. Antibiotic use in aquaculture: development of antibiotic resistance - 338

potential for consumer health risks. Int J food Sci Technol 33, 139–155. 339

Almeida, A., Cunha, Â., Gomes, N.C.M., Alves, E., Costa, L., Faustino, M.A.F., 2009. Phage therapy and 340

photodynamic therapy: Low environmental impact approaches to inactivate microorganisms in fish 341

farming plants. Mar Drugs 7, 268–313. doi:10.3390/md7030268 342

Austin, B., Allen, D.A., 1982. Microbiology of laboratory-hatched brine shrimp (Artemia). Aquaculture 26, 343

369–383. doi:10.1016/0044-8486(82)90170-3 344

Breitbart, M., 2012. Marine Viruses: Truth or Dare. Ann Rev Mar Sci 4, 425–448. doi:10.1146/annurev-345

marine-120709-142805 346

16

Brussaard, C.P.D., Wilhelm, S.W., Thingstad, F., Weinbauer, M.G., Bratbak, G., Heldal, M., Kimmance, S.A., 347

Middelboe, M., Nagasaki, K., Paul, J.H., Schroeder, D.C., Suttle, C.A., Vaqué, D., Wommack, K.E., 2008. 348

Global-scale processes with a nanoscale drive: The role of marine viruses. ISME J 2, 575–578. 349

doi:10.1038/ismej.2008.31 350

Buddruhs, N., Pradella, S., Göker, M., Päuker, O., Pukall, R., Spröer, C., Schumann, P., Petersen, J., Brinkhoff, 351

T., 2013. Molecular and phenotypic analyses reveal the non-identity of the Phaeobacter gallaeciensis 352

type strain deposits CIP 105210T and DSM 17395. Int J Syst Evol Microbiol 63, 4340–9. 353

doi:10.1099/ijs.0.053900-0 354

Burmølle, M., Webb, J.S., Rao, D., Hansen, L.H., Sørensen, S.J., Kjelleberg, S., 2006. Enhanced biofilm 355

formation and increased resistance to antimicrobial agents and bacterial invasion are caused by 356

synergistic interactions in multispecies biofilms. Appl Environ Microbiol 72, 3916–3923. 357

doi:10.1128/AEM.03022-05 358

Castillo, D., Christiansen, R.H., Dalsgaard, I., Madsen, L., Middelboe, M., 2015. Bacteriophage Resistance 359

Mechanisms in the Fish Pathogen Flavobacterium psychrophilum: Linking Genomic Mutations to 360

Changes in Bacterial Virulence Factors. Appl Environ Microbiol 81, 1157–1167. 361

doi:10.1128/AEM.03699-14 362

Castillo, D., Higuera, G., Villa, M., Middelboe, M., Dalsgaard, I., Madsen, L., Espejo, R.T., 2012. Diversity of 363

Flavobacterium psychrophilum and the potential use of its phages for protection against bacterial cold 364

water disease in salmonids. J Fish Dis 35, 193–201. doi:10.1111/j.1365-2761.2011.01336.x 365

Castillo, D., Rørbo, N., Jørgensen, J., Lange, J., Tan, D., Kalatzis, P.G., Svenningsen, S. Lo, Middelboe, M., 366

2019. Phage defense mechanisms and their genomic and phenotypic implications in the fish pathogen 367

Vibrio anguillarum. FEMS Microbiol Ecol in press. doi:10.1093/femsec/fiz004 368

D’Alvise, P.W., Lillebø, S., Prol-Garcia, M.J., Wergeland, H.I., Nielsen, K.F., Bergh, Ø., Gram, L., 2012. 369

17

Phaeobacter gallaeciensis reduces Vibrio anguillarum in cultures of microalgae and rotifers, and 370

prevents vibriosis in cod larvae. PLoS One 7, e43996. doi:10.1371/journal.pone.0043996 371

D’Alvise, P.W., Lillebø, S., Wergeland, H.I., Gram, L., Bergh, Ø., 2013. Protection of cod larvae from vibriosis 372

by Phaeobacter spp.: A comparison of strains and introduction times. Aquaculture 384–387, 82–86. 373

doi:10.1016/j.aquaculture.2012.12.013 374

D’Alvise, P.W., Melchiorsen, J., Porsby, C.H., Nielsen, K.F., Gram, L., 2010. Inactivation of Vibrio anguillarum 375

by attached and planktonic Roseobacter cells. Appl Environ Microbiol 76, 2366–70. 376

doi:10.1128/AEM.02717-09 377

D’Alvise, P.W., Phippen, C.B.W., Nielsen, K.F., Gram, L., 2016. Influence of Iron on Production of the 378

Antibacterial Compound Tropodithietic Acid and Its Noninhibitory Analog in Phaeobacter inhibens. 379

Appl Environ Microbiol 82, 502–509. doi:10.1128/AEM.02992-15 380

Defoirdt, T., Sorgeloos, P., Bossier, P., 2011. Alternatives to antibiotics for the control of bacterial disease in 381

aquaculture. Curr Opin Microbiol 14, 251–8. doi:10.1016/j.mib.2011.03.004 382

FAO, WHO, 2001. Report of a joint FAO/WHO expert consultation on evaluation of health and nutritional 383

properties of probiotics in food including powder milk with live lactic acid bacteria. 384

Frans, I., Michiels, C.W., Bossier, P., Willems, K. a., Lievens, B., Rediers, H., 2011. Vibrio anguillarum as a fish 385

pathogen: Virulence factors, diagnosis and prevention. J Fish Dis 34, 643–661. doi:10.1111/j.1365-386

2761.2011.01279.x 387

Grotkjær, T., Bentzon-Tilia, M., D’Alvise, P., Dierckens, K., Bossier, P., Gram, L., 2016a. Phaeobacter inhibens 388

as probiotic bacteria in non-axenic Artemia and algae cultures. Aquaculture 462, 64–69. 389

doi:10.1016/j.aquaculture.2016.05.001 390

Grotkjær, T., Bentzon-Tilia, M., D’Alvise, P., Dourala, N., Nielsen, K.F., Gram, L., 2016b. Isolation of TDA-391

18

producing Phaeobacter strains from sea bass larval rearing units and their probiotic effect against 392

pathogenic Vibrio spp. in Artemia cultures. Syst Appl Microbiol 39, 180–188. 393

doi:10.1016/j.syapm.2016.01.005 394

Guillard, R.R.L., 1975. Culture of phytoplankton for feeding marine invertebrates, in: Culture of Marine 395

Invertebrate Animals. Springer US, Boston, MA, pp. 29–60. doi:10.1007/978-1-4615-8714-9_3 396

Guillard, R.R.L., Ryther, J.H., 1962. Studies of marine planktonic diatoms: Cyclotella nana hustedt, and 397

Detonula confervacea (cleve) gran. Can J Microbiol 8, 229–239. doi:10.1139/m62-029 398

Hjelm, M., Bergh, O., Riaza, A., Nielsen, J., Melchiorsen, J., Jensen, S., Duncan, H., Ahrens, P., Birkbeck, H., 399

Gram, L., 2004. Selection and identification of autochthonous potential probiotic bacteria from turbot 400

larvae (Scophthalmus maximus) rearing units. Syst Appl Microbiol 27, 360–71. doi:10.1078/0723-401

2020-00256 402

Holmstrom, K., Graslund, S., Wahlstrom, A., Poungshompoo, S., Bengtsson, B.-E., Kautsky, N., 2003. 403

Antibiotic use in shrimp farming and implications for environmental impacts and human health. Int J 404

Food Sci Technol 38, 255–266. doi:10.1046/j.1365-2621.2003.00671.x 405

Igarashi, M.A., Segugita, H., Deguchi, Y., 1989. Microflora associated with eggs and nauplii of Artemia salina. 406

Nippon Suisan Gakkaishi 55, 2045. 407

Inoue, T., Matsuzaki, S., Tanaka, S., 1995. A 26-kDa outer membrane protein, OmpK, common to Vibrio 408

species is the receptor for a broad-host-range vibriophage, KVP40. FEMS Microbiol Lett 125, 101–105. 409

doi:10.1016/0378-1097(94)00480-F 410

Kalatzis, P.G., Rørbo, N., Castillo, D., Mauritzen, J.J., Jørgensen, J., Kokkari, C., Zhang, F., Katharios, P., 411

Middelboe, M., 2017. Stumbling across the same phage: Comparative genomics of widespread 412

temperate phages infecting the fish pathogen Vibrio anguillarum. Viruses 9, 1–19. 413

19

doi:10.3390/v9050122 414

Karunasagar, I., Shivu, M.M., Girisha, S.K., Krohne, G., Karunasagar, I., 2007. Biocontrol of pathogens in 415

shrimp hatcheries using bacteriophages. Aquaculture 268, 288–292. 416

doi:10.1016/j.aquaculture.2007.04.049 417

Kibenge, F.S.B., Godoy, M.G., Fast, M., Workenhe, S., Kibenge, M.J.T., 2012. Countermeasures against viral 418

diseases of farmed fish. Antiviral Res 95, 257–281. doi:10.1016/j.antiviral.2012.06.003 419

Kutter, E., 2009. Phage Host Range and Efficiency of Plating, in: Clokie, M.R., Kropinski, A.M. (Eds.), 420

Bacteriophages. Methods in Molecular Biology. pp. 141–149. doi:10.1007/978-1-60327-164-6_14 421

Lannan, C.N., Winton, J.R., Fryer, J.L., 1984. Fish Cell Lines: Establishment and Characterization of Nine Cell 422

Lines from Salmonids. In Vitro 20, 671–676. 423

López-Torres, M.A., Lizárraga-Partida, M.L., 2001. Bacteria isolated on TCBS media associated with hatched 424

Artemia cysts of commercial brands. Aquaculture 194, 11–20. doi:10.1016/S0044-8486(00)00505-6 425

Martens, T., Heidorn, T., Pukall, R., Simon, M., Tindall, B.J., Brinkhoff, T., 2006. Reclassification of 426

Roseobacter gallaeciensis Ruiz-Ponte et al. 1998 as Phaeobacter gallaeciensis gen. nov., comb. nov., 427

description of Phaeobacter inhibens sp. nov., reclassification of Ruegeria algicola (Lafay et al. 1995) 428

Uc. Int J Syst Evol Microbiol 56, 1293–304. doi:10.1099/ijs.0.63724-0 429

Matsuzaki, S., Tanaka, S., Koga, T., Kawata, T., 1992. A Broad-Host-Range Vibriophage, KVP40, Isolated from 430

Sea Water. Microbiol Immunol 36, 93–97. doi:10.1111/j.1348-0421.1992.tb01645.x 431

Nakai, T., Sugimoto, R., Park, K.H., Matsuoka, S., Mori, K., Nishioka, T., Maruyama, K., 1999. Protective 432

effects of bacteriophage on experimental Lactococcus garvieae infection in yellowtail. Dis Aquat 433

Organ 37, 33–41. doi:10.3354/dao037033 434

Olafsen, J.A., 2001. Interactions between fish larvae and bacteria in marine aquaculture. Aquaculture 200, 435

20

223–247. doi:10.1016/S0044-8486(01)00702-5 436

Oliveira, J., Castilho, F., Cunha, A., Pereira, M.J., 2012. Bacteriophage therapy as a bacterial control strategy 437

in aquaculture. Aquac Int 20, 879–910. doi:10.1007/s10499-012-9515-7 438

Park, S.C., Nakai, T., 2003. Bacteriophage control of Pseudomonas plecoglossicida infection in ayu 439

Plecoglossus altivelis 53, 33–39. 440

Pedersen, K., Larsen, J., 1993. rRNA gene restriction patterns of Vibrio anguillarum serogroup O1. Dis Aquat 441

Organ 16, 121–126. doi:10.3354/dao016121 442

Planas, M., Pérez-Lorenzo, M., Hjelm, M., Gram, L., Uglenes Fiksdal, I., Bergh, Ø., Pintado, J., 2006. Probiotic 443

effect in vivo of Roseobacter strain 27-4 against Vibrio (Listonella) anguillarum infections in turbot 444

(Scophthalmus maximus L.) larvae. Aquaculture 255, 323–333. doi:10.1016/j.aquaculture.2005.11.039 445

Porsby, C.H., Nielsen, K.F., Gram, L., 2008. Phaeobacter and Ruegeria species of the Roseobacter clade 446

colonize separate niches in a Danish Turbot (Scophthalmus maximus)-rearing farm and antagonize 447

Vibrio anguillarum under different growth conditions. Appl Environ Microbiol 74, 7356–64. 448

doi:10.1128/AEM.01738-08 449

Porsby, C.H., Webber, M.A., Nielsen, K.F., Piddock, L.J. V, Gram, L., 2011. Resistance and tolerance to 450

tropodithietic acid, an antimicrobial in aquaculture, is hard to select. Antimicrob Agents Chemother 451

55, 1332–7. doi:10.1128/AAC.01222-10 452

Rasmussen, B.B., Erner, K.E., Bentzon-Tilia, M., Gram, L., 2018. Effect of TDA-producing Phaeobacter 453

inhibens on the fish pathogen Vibrio anguillarum in non-axenic algae and copepod systems. Microb 454

Biotechnol 11, 1070–1079. doi:10.1111/1751-7915.13275 455

Rasmussen, B.B., Grotkjær, T., D’Alvise, P.W., Yin, G., Zhang, F., Bunk, B., Spröer, C., Bentzon-Tilia, M., 456

Gram, L., 2016. Vibrio anguillarum is genetically and phenotypically unaffected by long-term 457

21

continuous exposure to the antibacterial compound tropodithietic acid. Appl Environ Microbiol 82, 458

4802–4810. doi:10.1128/AEM.01047-16 459

Reid, H.I., Treasurer, J.W., Adam, B., Birkbeck, T.H., 2009. Analysis of bacterial populations in the gut of 460

developing cod larvae and identification of Vibrio logei, Vibrio anguillarum and Vibrio splendidus as 461

pathogens of cod larvae. Aquaculture 288, 36–43. doi:10.1016/j.aquaculture.2008.11.022 462

Ringø, E., Olsen, R.E., Jensen, I., Romero, J., Lauzon, H.L., 2014. Application of vaccines and dietary 463

supplements in aquaculture: possibilities and challenges. Rev Fish Biol Fish 24, 1005–1032. 464

doi:10.1007/s11160-014-9361-y 465

Rohwer, F., Thurber, R.V., 2009. Viruses manipulate the marine environment. Nature 459, 207–212. 466

doi:10.1038/nature08060 467

Romero, D., Traxler, M.F., López, D., Kolter, R., 2011. Antibiotics as signal molecules. Chem Rev 111, 5492–468

5505. doi:10.1021/cr2000509.Antibiotics 469

Rønneseth, A., Castillo, D., D’Alvise, P., Tønnesen, Ø., Haugland, G., Grotkjaer, T., Engell-Sørensen, K., 470

Nørremark, L., Bergh, Ø., Wergeland, H.I., Gram, L., 2017. Comparative assessment of Vibrio virulence 471

in marine fish larvae. J Fish Dis 40, 1373–1385. doi:10.1111/jfd.12612 472

Rørbo, N., Rønneseth, A., Kalatzis, P.G., Rasmussen, B.B., Engell-Sørensen, K., Kleppen, H.P., Wergeland, 473

H.I., Gram, L., Middelboe, M., 2018. Exploring the Effect of Phage Therapy in Preventing Vibrio 474

anguillarum Infections in Cod and Turbot Larvae. Antibiotics 7, 42. doi:10.3390/antibiotics7020042 475

Ruiz-Ponte, C., Cilia, V., Lambert, C., Nicolas, J.L., 1998. Roseobacter gallaeciensis sp. nov., a new marine 476

bacterium isolated from rearings and collectors of the scallop Pecten maximus. Int J Syst Bacteriol 48, 477

537–542. doi:10.1099/00207713-48-2-537 478

Ruiz-Ponte, C., Samain, J.J.F., Sánchez, J.L.J., Nicolas, J.J.L., 1999. The Benefit of a Roseobacter Species on 479

22

the Survival of Scallop Larvae. Mar Biotechnol 1, 52–59. doi:10.1007/PL00011751 480

Skjermo, J., Vadstein, O., 1999. Techniques for microbial control in the intensive rearing of marine larvae. 481

Aquaculture 177, 333–343. doi:10.1016/S0044-8486(99)00096-4 482

Skov, M.N., Pedersen, K., Larsen, J.L., 1995. Comparison of Pulsed-Field Gel Electrophoresis, Ribotyping, and 483

Plasmid Profiling for Typing of Vibrio anguillarum Serovar O1. Appl Environ Microbiol. 484

Sobecky, P. a., Mincer, T.J., Chang, M.C., Helinski, D.R., 1997. Plasmids isolated from marine sediment 485

microbial communities contain replication and incompatibility regions unrelated to those of known 486

plasmid groups. Appl Environ Microbiol 63, 888–895. 487

Sommerset, I., Krossøy, B., Biering, E., Frost, P., 2005. Vaccines for fish in aquaculture. Expert Rev Vaccines 488

4, 89–101. doi:10.1586/14760584.4.1.89 489

Spanggaard, B., Jørgensen, F., Gram, L., Huss, H.H., 1993. Antibiotic resistance in bacteria isolated from 490

three freshwater fish farms and an unpolluted stream in Denmark. Aquaculture 115, 195–207. 491

doi:10.1016/0044-8486(93)90136-M 492

Suttle, C.A., 2007. Marine viruses - Major players in the global ecosystem. Nat Rev Microbiol 5, 801–812. 493

doi:10.1038/nrmicro1750 494

Tan, D., Gram, L., Middelboe, M., 2014. Vibriophages and their interactions with the fish pathogen Vibrio 495

anguillarum. Appl Environ Microbiol 80, 3128–40. doi:10.1128/AEM.03544-13 496

Tan, D., Svenningsen, S. Lo, Middelboe, M., 2015. Quorum Sensing Determines the Choice of Antiphage 497

Defense Strategy in Vibrio anguillarum. MBio 6, 1–10. doi:10.1128/mBio.00627-15. 498

Thompson, F.L., Iida, T., Swings, J., 2004. Biodiversity of Vibrios. Microbiol Mol Biol Rev 68, 403–431. 499

doi:10.1128/MMBR.68.3.403 500

23

Toranzo, A.E., Magariños, B., Romalde, J.L., 2005. A review of the main bacterial fish diseases in mariculture 501

systems. Aquaculture 246, 37–61. doi:10.1016/j.aquaculture.2005.01.002 502

Vandecandelaere, I., Segaert, E., Mollica, A., Faimali, M., Vandamme, P., 2008. Leisingera aquimarina sp. 503

nov., isolated from a marine electroactive biofilm, and emended descriptions of Leisingera 504

methylohalidivorans Schaefer et al. 2002, Phaeobacter daeponensis Yoon et al. 2007 and Phaeobacter 505

inhibens Ma. Int J Syst Evol Microbiol 58, 2788–93. doi:10.1099/ijs.0.65844-0 506

507

24

LEGENDS FOR FIGURES AND TABLES 508

Fig. 1. (A) P. inhibens and (B) KVP40 counts in non-axenic T. suecica systems. (A) P. inhibens alone (●), P. 509

inhibens with V. anguillarum (■), P. inhibens with V. anguillarum and KVP40 (▲). (B) KVP40 alone (●), 510

KVP40 with V. anguillarum (■), KVP40 with V. anguillarum and P. inhibens (▲). Error bars show the 511

standard deviation of three biological replicates 512

513

Fig. 2. Vibrio anguillarum counts in non-axenic T. suecica systems. V. anguillarum alone (●), V. anguillarum 514

with P. inhibens (■), V. anguillarum with P. inhibens and KVP40 (▲), V. anguillarum with KVP40 (▼). Error 515

bars show the standard deviation of three biological replicates 516

517

Fig. 3. (A) P. inhibens and (B) KVP40 counts in non-axenic Artemia systems. (A) P. inhibens alone (●), P. 518

inhibens with V. anguillarum (■), P. inhibens with V. anguillarum and KVP40 (▲). (B) KVP40 alone (●), 519

KVP40 with V. anguillarum (■), KVP40 with V. anguillarum and P. inhibens (▲). Error bars show the 520

standard deviation of three biological replicates 521

522

Fig. 4. Vibrio anguillarum counts in non-axenic Artemia systems. V. anguillarum alone (●), V. anguillarum 523

with P. inhibens (■), V. anguillarum with P. inhibens and KVP40 (▲), V. anguillarum with KVP40 (▼). 524

Arrows indicate vibrio counts below detection level (10 CFU ml-1) as seen for treatments with P. inhibens 525

and with P. inhibens and KVP40 at time points 3 days and 4 days. Error bars show the standard deviation of 526

three biological replicates 527

528

Fig. 5. Mortality of (A) non-challenged and (B) vibrio-challenged Artemia in non-axenic systems. (A) Non-529

challenged control (●), Non-challenged; P. inhibens treated (■), Non-challenged; KVP40 treated (▼). (B) V. 530

anguillarum challenged (●), V. anguillarum challenged; P. inhibens treated (■), V. anguillarum challenged; 531

25

KVP40 treated (▼), V. anguillarum challenged; P. inhibens and KVP40 treated (▲). Error bars show the 532

standard deviation of three biological replicates 533

534

Supplementary Fig. 1. Survival of CHSE-214 fish cell lines challenged with V. anguillarum. Abbreviations: 535

Pro = P. inhibens; Pha = KVP40. Error bars show the standard deviation of three biological replicates 536

537

26

538

539

Fig. 1. (A) P. inhibens and (B) KVP40 counts in non-axenic T. suecica systems. (A) P. inhibens alone (●), P. 540

inhibens with V. anguillarum (■), P. inhibens with V. anguillarum and KVP40 (▲). (B) KVP40 alone (●), 541

KVP40 with V. anguillarum (■), KVP40 with V. anguillarum and P. inhibens (▲). Error bars show the 542

standard deviation of three biological replicates 543

27

544

545

Fig. 2. Vibrio anguillarum counts in non-axenic T. suecica systems. V. anguillarum alone (●), V. anguillarum 546

with P. inhibens (■), V. anguillarum with P. inhibens and KVP40 (▲), V. anguillarum with KVP40 (▼). Error 547

bars show the standard deviation of three biological replicates 548

549

550

551

552

553

554

555

556

28

557

Fig. 3. (A) P. inhibens and (B) KVP40 counts in non-axenic Artemia systems. (A) P. inhibens alone (●), P. 558

inhibens with V. anguillarum (■), P. inhibens with V. anguillarum and KVP40 (▲). (B) KVP40 alone (●), 559

KVP40 with V. anguillarum (■), KVP40 with V. anguillarum and P. inhibens (▲). Error bars show the 560

standard deviation of three biological replicates 561

562

563

564

565

566

29

567

Fig. 4. Vibrio anguillarum counts in non-axenic Artemia systems. V. anguillarum alone (●), V. anguillarum 568

with P. inhibens (■), V. anguillarum with P. inhibens and KVP40 (▲), V. anguillarum with KVP40 (▼). 569

Arrows indicate vibrio counts below detection level (10 CFU ml-1) as seen for treatments with P. inhibens 570

and with P. inhibens and KVP40 at time points 3 days and 4 days. Error bars show the standard deviation of 571

three biological replicates 572

573

574

575

576

30

577

Fig. 5. Mortality of (A) non-challenged and (B) vibrio-challenged Artemia in non-axenic systems. (A) Non-578

challenged control (●), Non-challenged; P. inhibens treated (■), Non-challenged; KVP40 treated (▼). (B) V. 579

anguillarum challenged (●), V. anguillarum challenged; P. inhibens treated (■), V. anguillarum challenged; 580

KVP40 treated (▼), V. anguillarum challenged; P. inhibens and KVP40 treated (▲). Error bars show the 581

standard deviation of three biological replicates 582

31

583

Supplementary Fig. 1. Survival of CHSE-214 fish cell lines challenged with V. anguillarum. Abbreviations: 584

Pro = P. inhibens; Pha = KVP40. Error bars show the standard deviation of three biological replicates 585

586

Paper 5

Rasmussen BB, Dittmann KK, Gram L, Bentzon-Tilia M. 2019.

Upscaling probiotic Phaeobacter spp. in Tetraselmis suecica algae cultures.

Manuscript in preparation

1

Upscaling probiotic Phaeobacter spp. in Tetraselmis suecica algae cultures

Bastian Barker Rasmussen1, Karen Kiesbye Dittmann1, Lone Gram1 and Mikkel Bentzon-Tilia1*

1. Department of Biotechnology and Biomedicine, Technical University of Denmark, Søltofts plads bldg.

221, DK-2800 Kgs. Lyngby, Denmark

* corresponding author: e-mail: mibti@bio.dtu.dk, phone: +45 4525 2518

Keywords: Aquaculture, Phaeobacter, roseobacters, Probiotics, Live feed, Vibrio

Running-title: Upscaling probiotic phaeobacters in live feed

29th March 2019

2

ABSTRACT 1

Bacteria cause more than half of all infections in aquaculture, and antibiotics are still used to treat 2

infections of finfish larvae. Tropodithietic acid (TDA) producing members of the marine Roseobacter group 3

are able to protect turbot and cod larvae as well as live feed, e.g. rotifers and Artemia, against pathogenic 4

vibrios and hence they represent a potential sustainable alternative to antibiotics.. It is however currently 5

unknown to what extent these potential probiotic bacteria will proliferate in an aquaculture setting. One 6

possible approach to establish probiotic roseobacters in aquaculture systems is to scale up the probionts in 7

algae cultures used as live feed. Here we show that the Roseobacter group bacteria Phaeobacter inhibens 8

DSM17395 and Phaeobacter piscinae S26 can indeed be seeded in small volumes of live feed algal cultures 9

(40 mL) and successfully scale up alongside the algae Tetraselmis suecica without the need for further 10

seeding. Afterwards, upscaled algae with the Phaeobacter strains were fed to Artemia challenged with the 11

fish pathogenic Vibrio anguillarum in order test if the six weeks of upscaling had any impact on the 12

probiotic abilities the two Phaeobacter strains. V. anguillarum was inhibited by P. piscinae S26 but not P. 13

inhibens DSM17395, however, we ascribe this to the low levels of Phaeobacter in the cultures rather than 14

loss of probiotic abilities. Amplicon based community analysis will show if and how the Phaeobacter spp. 15

change the community composition of algal microbiomes during the scale up. 16

17

18

19

20

Words = 239 21

22

23

3

INTRODUCTION 24

Live feed is important in the breeding of several commercially important marine finfish species such as cod, 25

turbot and halibut due to a lack of appropriate feed formulas at the larval stage. Cultures of live feed, such 26

as Artemia and rotifers, are rich in nutrients, which promote growth of opportunistic pathogenic bacteria, 27

and thus bacterial infections are a serious risk in such systems (Kibenge et al., 2012; Ringø et al., 2014). The 28

family of Vibrionacea include human and fish pathogens such as Vibrio cholera, V. vulnificus, V. 29

parahaemolyticus, V. splendidus, V. alginolyticus, and V. anguillarum, many of which are important 30

pathogens in aquaculture (Thompson et al., 2004; Toranzo et al., 2005). V. anguillarum is able to infect 31

more than 50 different fish species, including cod, halibut, and turbot where it causes the haemoreatic 32

septecimic disease vibriosis (Frans et al., 2011). Antibiotics have historically been the method of choice 33

when battling pathogenic bacteria, however, the development of antibiotic resistance, and the associated 34

health risks, have led to the search for sustainable alternatives (Defoirdt et al., 2011). Vaccines have 35

reduced the need for antibiotics in the industry dramatically, however, the small size and immature 36

immune systems of fish larvae make vaccination troublesome and inefficient (Defoirdt et al., 2011; 37

Sommerset et al., 2005). Thus other alternatives are sought after for use in larviculture. 38

39

Defined by FAO and WHO as “Live microorganisms which, when administered in adequate amounts, confer 40

a health benefit to the host” (FAO and WHO, 2001), probiotics can be used to protect fish larvae against 41

pathogenic bacteria by inhibiting growth of the pathogens (D’Alvise et al., 2013, 2012; Planas et al., 2006). 42

Several members of the genus Phaeobacter of the diverse marine Roseobacter group have shown 43

promising potential as probiotics in aquaculture. Phaeobacter gallaeciensis, P. inhibens and P. piscinae, 44

which produce the broad spectrum antimicrobial compound tropodithietic acid (TDA) (Martens et al., 2006; 45

Sonnenschein et al., 2017), have been shown to protect live feed (D’Alvise et al., 2012; Grotkjær et al., 46

2016a, 2016b) as well as turbot and cod larvae against vibriosis (D’Alvise et al., 2013, 2012; Planas et al., 47

2006). TDA resistance does not develop easily (Porsby et al., 2011; Rasmussen et al., 2016) and long-term 48

4

exposure of V. anguillarum to TDA does not result in increased mutations or increased virulence 49

(Rasmussen et al., 2016). Furthermore, Phaeobacter spp. have been observed repeatedly as part of the 50

indigenous aquaculture microbiota (Grotkjær et al., 2016b; Porsby et al., 2008), and considering all if the 51

above, the use of Phaeobacter spp. as probionts in aquaculture seems to be a viable alternative to the 52

administration of antibiotics. However, in order for Phaeobacter spp. to become a cost-efficient alternative 53

to antibiotics, effective and easy deployment in commercial aquaculture is crucial. 54

55

Several different methods of deployment of probiotic bacteria exist. Phaeobacter piscinae 27-4 have for 56

example been tested using a bioencapsulation-based approach where rotifers were enriched with the 57

probiotic strain resulting in a decrease in mortality of V. anguillarum-challenged turbot larvae when fed the 58

Phaeobacter enriched rotifers (Planas et al., 2006). Incorporation of P. piscinae in biofilters used to filter the 59

rearing water in larva tanks has also been tested showing a decrease in the abundance of V. anguillarum 60

and a simultaneous decrease in turbot larval mortality as a result (Prol-García and Pintado, 2013). Live feed 61

can function as a potential point of entry and as vectors for pathogens (Olafsen, 2001; Rasmussen et al., 62

2018), and since algae are the first step in the live feed food chain, we speculate that if the probiotic 63

bacteria are added to and establishes itself in the algae, it will spread through the system and inhibit 64

pathogenic growth at all levels of the food chain ultimately resulting in protection of the fish larvae. 65

66

The ability of Phaeobacter spp. to inhibit Vibrio growth in different axenic- and non-axenic alga systems has 67

been shown repeatedly along with Phaeobacter spp. ability to establish themselves at stable levels in these 68

systems, during the four to five days duration of the experiments (D’Alvise et al., 2012; Grotkjær et al., 69

2016a; Rasmussen et al., 2018). If Phaeobacter spp. are to successfully be scaled up in algae they need to 70

establish themselves at stable level and regrow back to that level after each upscaling step. If the 71

Phaeobacter spp. are unable to regrow they will eventually be deluded out of the cultures. Furthermore, 72

changes in gene expression and physiology due to long-term cultivation has been reported (Somerville et 73

5

al., 2002). Therefore, it is a genuine concern whether or not long-term upscaling of Phaeobacter spp. in 74

algae can have a negative impact on their probiotic capabilities. Furthermore, since only very limited data 75

exist describing the effect of Phaeobacter spp. on the microbiota of live feed, it is a concern to what extent 76

Phaeobacter spp. will affect the microbiota of the algae on a spatial scale. Thus, the purpose of this study 77

was to investigate if the potential probionts Phaeobacter inhibens DSM17395 and Phaeobacter piscinae S26 78

could be scaled up in Tetraselmis suecica algal cultures, how the long-term upscaling process with 79

Phaeobacter probiotics would affect the microbiota of T. suecica, and ultimately if Artemia fed with 80

Phaeobacter-infused algae would be protected protected against vibriosis. 81

82

MATERIALS AND METHODS 83

Bacterial strains and culture conditions. Phaeobacter inhibens DSM17395 (Buddruhs et al., 2013; Martens 84

et al., 2006; Vandecandelaere et al., 2008) and Phaeobacter piscinae S26 (Grotkjær et al., 2016b; 85

Sonnenschein et al., 2017) were used as probionts in the upscaling and subsequent Artemia experiments. 86

Vibrio anguillarum strain 90-11-286 is highly virulent against cod, turbot and halibut larvae and was used as 87

pathogen in the Artemia experiment (Rønneseth et al., 2017). 90-11-286 was isolated from diseased 88

rainbow trout (Oncorhynchus mykiss) from a Danish fish farm (Pedersen and Larsen, 1993; Skov et al., 89

1995). TSA (Tryptone Soy Agar, Difco 212185) was used to grow and enumerate 90-11-286. The 90

Phaeobacter spp. does not grow on TSA due to its low salinity and 90-11-286 have distinct colony 91

morphology on TSA and is able to outgrow the background microbiota of the non-axenic systems on the 92

TSA plates. Thus, accurate enumeration of 90-11-286 is possible despite it being outnumbered by the 93

background microbiota by two to three orders of magnitude (Rasmussen et al., 2018). The Phaeobacter 94

spp. were grown and enumerated on MA (Marine Agar, Difco 2216) where they form distinct brown 95

colonies due to production of a Fe-TDA complex (D’Alvise et al., 2016). Bacterial pre-cultures were grown in 96

20 mL of half-strength YTSS (½YTSS) (2 g Bacto Yeast extract, 1.25 g Bacto Tryptone, 20 g Sigma Sea Salts, 1 97

6

L deionized water) (Sobecky et al., 1997) for 1 day at 25° C and 200 rpm. 98

99

Preparation of Tetraselmis suecica and Artemia cultures. Non-axenic Tetraselmis suecica stock cultures 100

from a commercial aquaculture unit were maintained in f/2 medium (Guillard, 1975; Guillard and Ryther, 101

1962) without Na2SiO3, but supplemented with 5 mM NH4Cl in 3 % IO (Instant Ocean salts, Aquarium 102

Systems Inc.) (f/2-Si+NH4) at 18° C and 42 μmol photons m-2 s-1, photosynthetically active radiation (PAR). T. 103

suecica density was determined by counting in a Neubauer-improved counting chamber (Assistent). The T. 104

suecica cultures were adjusted to approximately 5 x 104 cell mL-1 in f/2-Si+NH4 and 40 mL distributed into 105

100 mL Erlenmeyer flasks. Artemia cysts obtained from the same commercial aquaculture unit as T. suecica 106

were stored at 5° C. A day prior to the experiment, 20 mg non-disinfected Artemia cysts were added to 20 107

ml 3 % IO and incubated at 20 °C and 200 rpm. 108

109

Upscaling of Phaeobacter spp. in Tetraselmis suecica cultures. 110

Aiming at concentrations of 106 CFU mL-1 40 μl P. piscinae S26 or P. inhibens DSM17395 pre-culture were 111

used to inoculate the 40 mL T. suecica cultures, except for control cultures where nothing was added. The 112

cultures were then incubated at 18° C and 42 μmol photons m-2 s-1 for 2 weeks after which the 40 mL 113

cultures were used to inoculate 360 mL fresh f/2-Si+NH4 in 1L Erlenmeyer flasks, which were incubated at 114

18° C and 42 μmol photons m-2 s-1 for 2 weeks after which the 400 mL were used to inoculate 3.6 L fresh 115

f/2-Si+NH4 in 5 L Erlenmeyer flasks. The 4 L flasks were then incubated at 21° C and 42 μmol photons m-2 s-1 116

with aeration for 2 weeks. Before and after each inoculation, and at the end of the experiment the cultures 117

were sampled, bacterial and algal abundances were determined, and 1 mL of each culture was centrifuged 118

at 8000 g x 5 for min and pellets were re-suspended in 1 mL lysis buffer (400 mM sodium chloride, 750 mM 119

sucrose, 20 mM EDTA, 50 mM Tris-HCl, 1 mg mL-1 lysozyme, pH 8.5) (Boström et al., 2004) and stored at -120

80° C until DNA was extracted. Bacterial abundances were determined using colony counts on MA, while 121

7

algae abundances were determined in a Neubauer-improved counting chamber. All experiments were 122

carried out in biological triplicates. 123

124

Co-culture experiment in Artemia fed Tetraselmis suecica. Artemia co-cultures were conducted in 50 mL 125

falcon tubes with 25 mL T. suecica culture diluted in 3 % IO to 105 algae mL-1 with or without P. piscinae S26 126

or P. inhibens DSM17395 from the 4 L upscaling cultures and approximately 2 Artemia mL-1. 125 μL of 10-127

fold serially diluted V. anguillarum 90-11-286 pre-cultures was used to inoculate the non-axenic Artemia 128

cultures, aiming at a concentration of 2 x 103 CFU mL-1. The Artemia cultures were incubated horizontally 129

on a rotary shaker at 27° C and 60 rpm. Experiments were carried out over 96 h and every 24 h bacterial 130

abundances and Artemia mortality were determined. Bacterial abundances were determined by colony 131

counts on TSA and MA plates for V. anguillarum and Phaeobacter spp., respectively. Plates were incubated 132

at 25° C. TSA plates were counted after 1-2 days and MA plates after 3 days. Dead Artemia was counted in a 133

Sedgewick-Rafter counting chamber. At experiment termination, surviving Artemia were killed and also 134

counted in a Sedgewick-Rafter counting chamber. All experiments were carried out in biological triplicates. 135

136

DNA extraction, PCR amplification, amplicon sequencing and bioinformatics data analysis. 137

(Steps no yet performed) 138

DNA extractions will be done using the phenol/chloroform-based protocol described by Dittmann et al. 139

(2018). 140

16S rRNA genes will be amplified using the universal primers 27F (AGAGTTTGATCMTGGCTCAG) and 1492R 141

(TACGGYTACCTTGTTACGACTT) (Lane, 1991) and the extracted genomic DNA as template. 142

The PCR products will be used as template in a subsequent PCR amplification of the V4 regions of the 16S 143

rRNA genes using the primers 515F-Y (GTGYCAGCMGCCGCGGTAA) (Parada et al., 2016) and 806R 144

(GGACTACNVGGGTWTCTAAT) (Apprill et al., 2015). Amplicons will be purified using the AMPure XP 145

purification beads (Beckman Coulter) according to manufacturer’s instructions. 146

8

Amplicons will be indexed and sequenced at the Novo Nordisk Foundation Center for Biosustainability, Kgs. 147

Lyngby, Denmark. 148

Raw reads will be processed and analyzed using mothur (Schloss et al., 2009) and the SILVA database 149

(Pruesse et al., 2007) 150

Bioinformatic analyses will be used to identify the community composition of T. suecica and any impact the 151

Phaeobacter spp. may have on community richness, diversity, structure or individual OTUs 152

153

Accession numbers. 154

Sequences will be deposited in the Sequencing Read Archive (SRA) at NCBI 155

156

Statistical analysis. CFU ml-1 values for each biological replicate were log10-transformed prior to statistical 157

analysis. Statistical analyses of differences in CFU ml-1 were performed using Two-way analysis of variance 158

(ANOVA) tests (α ≤ 0.05) and differences in Artemia survival were also analysed using Two-way ANOVA 159

tests (α ≤ 0.05). All analysis were performed in GraphPad Prism 6.07 (GraphPad Software, San Diego CA). 160

161

RESULTS 162

Upscaling of Phaeobacter spp. in Tetraselmis suecica cultures. T. suecica with or without Phaeobacter spp. 163

were scaled up from 40 mL to 400 mL, and from 400 mL to 4 L. At each step, cultures were inoculated to 164

approx. 5 x 104 algae mL-1 and grew to approx. 5 x 105 algae mL-1. No differences in T. suecica growth was 165

observed between the control cultures and culture with Phaeobacter spp. (Figure 1). The microbiota scaled 166

up along with the algae from approx. 106 CFU mL-1 to 107 CFU mL-1 at each step. No significant difference in 167

CFU counts were observed between the cultures (Figure 2). In the initial 40 mL cultures, the Phaeobacter 168

spp. were inoculated to approx. 106 CFU mL-1 at which they were stable until the first upscaling step from 169

40 mL to 400 mL (time point 2 weeks), where they were diluted to approx. 105 CFU mL-1. In the 400 mL 170

cultures both Phaeobacter spp. grew 1 log unit to approx. 106 CFU mL-1. After the second upscaling step 171

9

from 400 mL to 4 L (time point 4 weeks) where the Phaeobacter spp. were diluted to approx. 105 CFU mL-1 172

again, both Phaeobacter spp. were stable until the termination of the experiment. No significant 173

differences in CFU counts were observed between P. piscinae S26 or P. inhibens DSM17395 during the 174

upscaling experiment. 175

176

Co-culture experiment in Artemia fed Tetraselmis suecica. The microbiota grew from approx. 5 x 106 CFU 177

mL-1 to approx. 2 x 107 CFU mL-1 from day 0 to time day 4, with no significant differences between the 178

cultures (data not shown). V. anguillarum was inoculated at approx. 2 x 103 CFU mL-1 and grew with 3 log 179

units in the control and P. inhibens DSM17395 cultures and with 2 log units in the P. piscinae S26 cultures. 180

V. anguillarum counts were significantly lower in the S26 cultures compared to the control (P-values 0.0297 181

and 0.0052) and the P. inhibens DSM17395 cultures (P-values 0.0486 and 0.0099) at time points 3 and 4 days. 182

P. piscinae S26 abundances were at a stable level just above 104 CFU mL-1 in both cultures with non-183

challenged (Figure 4(A)) and Vibrio-challenged (Figure 4(B)) Artemia. P. inhibens DSM17395 was not 184

detected in any of the cultures at day 0 and 1 day where the limit of detection was 104 CFU mL-1. From day 185

2 to 4 the limit of detection was 103 CFU mL-1 and P. inhibens DSM17395 was detected and stable at 104 186

CFU mL-1 in the non-challenged cultures while it was only detected in one out of three cultures at day 2 and 187

4 at 103 CFU mL-1 in the Vibrio-challenged cultures. Artemia survival in the non-challenged cultures at day 188

were 66±6 %, 60±32 %, and 76±12 % (Figure 5 (A)) for the non-challenged control, the S26 treated and DSM 189

17395 treated cultures, respectively. In the Vibrio-challenged cultures Artemia survival at day 4 were 61±10 190

%, 60±5 %, and 54±35 % (Figure 5 (B)) for the Vibrio-challenged control, the S26 treated and DSM17395 191

treated, respectively. No significant differences were seen in Artemia survival between the treatments. 192

193

DISCUSSION 194

10

In order for Phaeobacter spp. based probiotics to become relevant in commercial aquaculture easy 195

deployment is crucial. In this study we showed that it is possible to scale up Phaeobacter spp. in live feed 196

algae cultures. 197

198

Both Phaeobacter strains established themselves around 106 CFU mL-1 after the first two weeks of 199

upscaling. After the second stage (1 : 10 inoculation ration) of the upscaling ended they had reached the 200

same level of approximately 106 CFU mL-1. A similar trend was seen for the total microbiota which started 201

at 106 CFU mL-1 and ended around 107 CFU mL-1. This suggest that the “natural” obtainable level for 202

Phaeobacter spp. are one log unit under the total microbiota if conditions favour it. Likewise, the algae 203

grew from approx. 5 x 104 algae mL-1 to 5 x 105 algae mL-1. In other words, both algae, the Phaeobacter 204

strains, and the total microbiota grew one log unit within a period of two weeks in the 400 mL culture. 205

Moving from 400 mL cultures with no aeration and large surface area to volume ratio to 4 L cultures with 206

aeration and small surface area to volume ratio, this apparently changed the conditions enough for the 207

Phaeobacter spp. to be unable to keep up with the growth of the algae and rest of the microbiota. 208

Which changes disfavoured the Phaeobacter spp. is hard to tell, however, the upscaling process should be 209

optimised. 210

211

The success criteria for upscaling Phaeobacter in live feed algae must be that Phaeobacter is able to grow at 212

least as fast as the algae it is upscaled in. Furthermore, Phaeobacter must be fast enough the compete with 213

the rest of the microbiota to avoid being outcompeted on nutrition or space. In this study we used an 214

inoculation ratio of 1 : 10, however, in industrial upscaling productions inoculation ratios range from 1 : 5 to 215

1 : 100 (Brown and Blackburn, 2013). It is possible that higher or lower inoculation ratio will favour 216

Phaeobacter growth and thus this should be tested. Another parameter which could be optimised is the 217

incubation period of the individual upscaling steps as this could potentially also impact growth of 218

Phaeobacter. In order to optimise the upscaling in favour of Phaeobacter relatively to the rest of the 219

11

microbiota or the algae, growth ratios between them should be used, in order to insure that conditions 220

favour Phaeobacter. 221

222

S26 inhibited the growth of V. anguillarum in the Artemia cultures, while DSM17395 did not. This could be 223

due to several things. DSM17395 grew at a lower level than S26 and thus probiont to pathogen ratio was 224

lower for DSM17395 than S26. It is also a possibility that S26 produced more TDA than DSM17395. 225

Grotkjær et al. (2016b) tested analysed TDA concentrations of cultures several Phaeobacter spp. after 3 226

days growth in marine broth and found that P. piscinae S26 produced approximately one and a half time as 227

much TDA as P. inhibens DSM17395, as S26 had produced 0.92 ± 0.07 μM TDA while DSM17395 only had 228

produced 0.54 ± 0.15 μM TDA. 229

230

No significant differences were seen in Artemia survival between cultures, which stands in contrast to 231

previous studies of Phaeobacter spp. in Artemia (Grotkjær et al., 2016b; Rasmussen et al., 2019), where 232

Phaeobacter spp. have been able to increase survival in both Vibrio-challenged and non-challenged 233

cultures. This could be due to the upscaling in T. suecica, however, it seems unlikely due to the inhibitin of 234

V. anguillarum by S26. More likely is the explanation that the differences in probiont and pathogen was too 235

low for Phaeobacter to significantly impact the cultures. Previous studies have used a minimum ratio of 100 236

: 1 probiont to pathogen, however, in this study we had a ration around 10:1 for S26 and below for one for 237

DSM17395, as it could not be enumerated at inoculation and after one day, DSM17395 levels must have 238

been below 104 CFU mL-1. 239

240

Dittmann et al. (2018) used amplicon sequencing to explore P. inhibens DSM17395 effect on the 241

microbiomes of two eukaryote systems, the microalga Emiliania huxleyi and European flat oyster. They 242

found that Phaeobacter had a larger impact on the community structure of the more simple microbiota of 243

E. huxleyi than on the more complex oyster microbiota. However, the experiments were only looking at a 244

12

short time frame 48-96 hour and thus, it is hard to use the data to predict how the T. suecica microbiomes 245

will be affected by the extended period of time used in the upscaling. 246

247

The results of this study indicate that Phaeobacter spp. are able to establish themselves in T. suecica 248

cultures and upscale along with the algae. However, optimisation of the upscaling process is needed in 249

order to secure the continuous regrowth of the Phaeobacter spp. in the algae cultures. Furthermore, our 250

results suggest that the Phaeobacter spp. do not lose their probiotic properties during the upscaling, this 251

should, however, be retested once the upscaling proses has been optimised. 252

253

ACKNOWLEDGEMENTS 254

This study was funded by grants from The Danish Council for Strategic Research | Programme Commission 255

on Health, Food and Welfare (12-132390; ProAqua) 256

257

REFERENCES 258

Apprill, A., Mcnally, S., Parsons, R., Weber, L., 2015. Minor revision to V4 region SSU rRNA 806R gene 259

primer greatly increases detection of SAR11 bacterioplankton. Aquat Microb Ecol 75, 129–137. 260

doi:10.3354/ame01753 261

Boström, K.H., Simu, K., Hagström, Å., Riemann, L., 2004. Optimization of DNA extraction for quantitative 262

marine bacterioplankton community analysis. Limnol Oceanogr Methods 2, 365–373. 263

doi:10.4319/lom.2004.2.365 264

Brown, M.R., Blackburn, S.I., 2013. Live microalgae as feeds in aquaculture hatcheries, Advances in 265

Aquaculture Hatchery Technology. Woodhead Publishing Limited. doi:10.1533/9780857097460.1.117 266

Buddruhs, N., Pradella, S., Göker, M., Päuker, O., Pukall, R., Spröer, C., Schumann, P., Petersen, J., Brinkhoff, 267

T., 2013. Molecular and phenotypic analyses reveal the non-identity of the Phaeobacter gallaeciensis 268

13

type strain deposits CIP 105210T and DSM 17395. Int J Syst Evol Microbiol 63, 4340–9. 269

doi:10.1099/ijs.0.053900-0 270

D’Alvise, P.W., Lillebø, S., Prol-Garcia, M.J., Wergeland, H.I., Nielsen, K.F., Bergh, Ø., Gram, L., 2012. 271

Phaeobacter gallaeciensis reduces Vibrio anguillarum in cultures of microalgae and rotifers, and 272

prevents vibriosis in cod larvae. PLoS One 7, e43996. doi:10.1371/journal.pone.0043996 273

D’Alvise, P.W., Lillebø, S., Wergeland, H.I., Gram, L., Bergh, Ø., 2013. Protection of cod larvae from vibriosis 274

by Phaeobacter spp.: A comparison of strains and introduction times. Aquaculture 384–387, 82–86. 275

doi:10.1016/j.aquaculture.2012.12.013 276

D’Alvise, P.W., Phippen, C.B.W., Nielsen, K.F., Gram, L., 2016. Influence of Iron on Production of the 277

Antibacterial Compound Tropodithietic Acid and Its Noninhibitory Analog in Phaeobacter inhibens. 278

Appl Environ Microbiol 82, 502–509. doi:10.1128/AEM.02992-15 279

Defoirdt, T., Sorgeloos, P., Bossier, P., 2011. Alternatives to antibiotics for the control of bacterial disease in 280

aquaculture. Curr Opin Microbiol 14, 251–8. doi:10.1016/j.mib.2011.03.004 281

Dittmann, K.K., Sonnenschein, E.C., Egan, S., Gram, L., Bentzon-Tilia, M., 2018. Impact of Phaeobacter 282

inhibens on marine eukaryote-associated microbial communities. Environ Microbiol Rep. 283

doi:10.1111/1758-2229.12698 284

FAO, WHO, 2001. Report of a joint FAO/WHO expert consultation on evaluation of health and nutritional 285

properties of probiotics in food including powder milk with live lactic acid bacteria. 286

Frans, I., Michiels, C.W., Bossier, P., Willems, K. a., Lievens, B., Rediers, H., 2011. Vibrio anguillarum as a fish 287

pathogen: Virulence factors, diagnosis and prevention. J Fish Dis 34, 643–661. doi:10.1111/j.1365-288

2761.2011.01279.x 289

Grotkjær, T., Bentzon-Tilia, M., D’Alvise, P., Dierckens, K., Bossier, P., Gram, L., 2016a. Phaeobacter inhibens 290

14

as probiotic bacteria in non-axenic Artemia and algae cultures. Aquaculture 462, 64–69. 291

doi:10.1016/j.aquaculture.2016.05.001 292

Grotkjær, T., Bentzon-Tilia, M., D’Alvise, P., Dourala, N., Nielsen, K.F., Gram, L., 2016b. Isolation of TDA-293

producing Phaeobacter strains from sea bass larval rearing units and their probiotic effect against 294

pathogenic Vibrio spp. in Artemia cultures. Syst Appl Microbiol 39, 180–188. 295

doi:10.1016/j.syapm.2016.01.005 296

Guillard, R.R.L., 1975. Culture of phytoplankton for feeding marine invertebrates, in: Culture of Marine 297

Invertebrate Animals. Springer US, Boston, MA, pp. 29–60. doi:10.1007/978-1-4615-8714-9_3 298

Guillard, R.R.L., Ryther, J.H., 1962. Studies of marine planktonic diatoms: Cyclotella nana hustedt, and 299

Detonula confervacea (cleve) gran. Can J Microbiol 8, 229–239. doi:10.1139/m62-029 300

Kibenge, F.S.B., Godoy, M.G., Fast, M., Workenhe, S., Kibenge, M.J.T., 2012. Countermeasures against viral 301

diseases of farmed fish. Antiviral Res 95, 257–281. doi:10.1016/j.antiviral.2012.06.003 302

Lane, D., 1991. 16S/23S rRNA sequencing, in: Stackebrandt, E., Goodfellow, M. (Eds.), Nucleic Acid 303

Techniques in Bacterial Systematics. J.Wiley and Sons, Amsterdam, pp. 115–175. 304

Levasseur, M., Thompson, P.A., Harrison, P.J., 1993. Physiological acclimation of marine phytoplankton to 305

different nitrogen sources. J Phycol 29, 587–595. doi:10.1111/j.0022-3646.1993.00587.x 306

Martens, T., Heidorn, T., Pukall, R., Simon, M., Tindall, B.J., Brinkhoff, T., 2006. Reclassification of 307

Roseobacter gallaeciensis Ruiz-Ponte et al. 1998 as Phaeobacter gallaeciensis gen. nov., comb. nov., 308

description of Phaeobacter inhibens sp. nov., reclassification of Ruegeria algicola (Lafay et al. 1995) 309

Uc. Int J Syst Evol Microbiol 56, 1293–304. doi:10.1099/ijs.0.63724-0 310

Olafsen, J.A., 2001. Interactions between fish larvae and bacteria in marine aquaculture. Aquaculture 200, 311

223–247. doi:10.1016/S0044-8486(01)00702-5 312

15

Parada, A.E., Needham, D.M., Fuhrman, J.A., 2016. Every base matters: assessing small subunit rRNA 313

primers for marine microbiomes with mock communities, time series and global field samples. 314

Environ Microbiol 18, 1403–1414. doi:10.1111/1462-2920.13023 315

Pedersen, K., Larsen, J., 1993. rRNA gene restriction patterns of Vibrio anguillarum serogroup O1. Dis Aquat 316

Organ 16, 121–126. doi:10.3354/dao016121 317

Planas, M., Pérez-Lorenzo, M., Hjelm, M., Gram, L., Uglenes Fiksdal, I., Bergh, Ø., Pintado, J., 2006. Probiotic 318

effect in vivo of Roseobacter strain 27-4 against Vibrio (Listonella) anguillarum infections in turbot 319

(Scophthalmus maximus L.) larvae. Aquaculture 255, 323–333. doi:10.1016/j.aquaculture.2005.11.039 320

Porsby, C.H., Nielsen, K.F., Gram, L., 2008. Phaeobacter and Ruegeria species of the Roseobacter clade 321

colonize separate niches in a Danish Turbot (Scophthalmus maximus)-rearing farm and antagonize 322

Vibrio anguillarum under different growth conditions. Appl Environ Microbiol 74, 7356–64. 323

doi:10.1128/AEM.01738-08 324

Porsby, C.H., Webber, M. a, Nielsen, K.F., Piddock, L.J. V, Gram, L., 2011. Resistance and tolerance to 325

tropodithietic acid, an antimicrobial in aquaculture, is hard to select. Antimicrob Agents Chemother 326

55, 1332–7. doi:10.1128/AAC.01222-10 327

Prol-García, M.J., Pintado, J., 2013. Effectiveness of Probiotic Phaeobacter Bacteria Grown in Biofilters 328

Against Vibrio anguillarum Infections in the Rearing of Turbot (Psetta maxima) Larvae. Mar Biotechnol 329

15, 726–738. doi:10.1007/s10126-013-9521-4 330

Pruesse, E., Quast, C., Knittel, K., Fuchs, B.M., Ludwig, W., Peplies, J., Glöckner, F.O., 2007. SILVA: a 331

comprehensive online resource for quality checked and aligned ribosomal RNA sequence data 332

compatible with ARB. Nucleic Acids Res 35, 7188–7196. doi:10.1093/nar/gkm864 333

Rasmussen, B.B., Erner, K.E., Bentzon-Tilia, M., Gram, L., 2018. Effect of TDA-producing Phaeobacter 334

16

inhibens on the fish pathogen Vibrio anguillarum in non-axenic algae and copepod systems. Microb 335

Biotechnol 11, 1070–1079. doi:10.1111/1751-7915.13275 336

Rasmussen, B.B., Grotkjær, T., D’Alvise, P.W., Yin, G., Zhang, F., Bunk, B., Spröer, C., Bentzon-Tilia, M., 337

Gram, L., 2016. Vibrio anguillarum is genetically and phenotypically unaffected by long-term 338

continuous exposure to the antibacterial compound tropodithietic acid. Appl Environ Microbiol 82, 339

4802–4810. doi:10.1128/AEM.01047-16 340

Rasmussen, B.B., Kalatzis, P.G., Middelboe, M., Gram, L., 2019. Combining probiotic Phaeobacter inhibens 341

DSM17395 and broad-host-range vibriophage KVP40 against fish pathogenic vibrios. Submitt 342

Manuscr. 343

Ringø, E., Olsen, R.E., Jensen, I., Romero, J., Lauzon, H.L., 2014. Application of vaccines and dietary 344

supplements in aquaculture: possibilities and challenges. Rev Fish Biol Fish 24, 1005–1032. 345

doi:10.1007/s11160-014-9361-y 346

Rønneseth, A., Castillo, D., D’Alvise, P., Tønnesen, Ø., Haugland, G., Grotkjaer, T., Engell-Sørensen, K., 347

Nørremark, L., Bergh, Ø., Wergeland, H.I., Gram, L., 2017. Comparative assessment of Vibrio virulence 348

in marine fish larvae. J Fish Dis 40, 1373–1385. doi:10.1111/jfd.12612 349

Schloss, P.D., Westcott, S.L., Ryabin, T., Hall, J.R., Hartmann, M., Hollister, E.B., Lesniewski, R.A., Oakley, 350

B.B., Parks, D.H., Robinson, C.J., Sahl, J.W., Stres, B., Thallinger, G.G., Horn, D.J. Van, Weber, C.F., 351

2009. Introducing mothur: Open-Source, Platform-Independent, Community-Supported Software for 352

Describing and Comparing Microbial Communities. Appl Environ Microbiol 75, 7537–7541. 353

doi:10.1128/AEM.01541-09 354

Skov, M.N., Pedersen, K., Larsen, J.L., 1995. Comparison of Pulsed-Field Gel Electrophoresis, Ribotyping, and 355

Plasmid Profiling for Typing of Vibrio anguillarum Serovar O1. Appl Environ Microbiol. 356

17

Sobecky, P. a., Mincer, T.J., Chang, M.C., Helinski, D.R., 1997. Plasmids isolated from marine sediment 357

microbial communities contain replication and incompatibility regions unrelated to those of known 358

plasmid groups. Appl Environ Microbiol 63, 888–895. 359

Somerville, G.A., Beres, S.B., Fitzgerald, J.R., DeLeo, F.R., Cole, R.L., Hoff, J.S., Musser, J.M., 2002. In vitro 360

serial passage of Staphylococcus aureus: changes in physiology, virulence factor production, and agr 361

nucleotide sequence. J Bacteriol 184, 1430–7. doi:10.1128/JB.184.5.1430 362

Sommerset, I., Krossøy, B., Biering, E., Frost, P., 2005. Vaccines for fish in aquaculture. Expert Rev Vaccines 363

4, 89–101. doi:10.1586/14760584.4.1.89 364

Sonnenschein, E.C., Phippen, C.B.W., Nielsen, K.F., Mateiu, R.V., Melchiorsen, J., Gram, L., Overmann, J., 365

Freese, H.M., 2017. Phaeobacter piscinae sp. nov., a species of the Roseobacter group and potential 366

aquaculture probiont. Int J Syst Evol Microbiol 67, 4559–4564. doi:10.1099/ijsem.0.002331 367

Thompson, F.L., Iida, T., Swings, J., 2004. Biodiversity of Vibrios. Microbiol Mol Biol Rev 68, 403–431. 368

doi:10.1128/MMBR.68.3.403 369

Toranzo, A.E., Magariños, B., Romalde, J.L., 2005. A review of the main bacterial fish diseases in mariculture 370

systems. Aquaculture 246, 37–61. doi:10.1016/j.aquaculture.2005.01.002 371

Vandecandelaere, I., Segaert, E., Mollica, A., Faimali, M., Vandamme, P., 2008. Leisingera aquimarina sp. 372

nov., isolated from a marine electroactive biofilm, and emended descriptions of Leisingera 373

methylohalidivorans Schaefer et al. 2002, Phaeobacter daeponensis Yoon et al. 2007 and Phaeobacter 374

inhibens Ma. Int J Syst Evol Microbiol 58, 2788–93. doi:10.1099/ijs.0.65844-0 375

376

377

18

LEGENDS FOR FIGURES AND TABLES 378

Figure 1. Tetraselmis suecica counts during upscaling experiment. T. suecica counts in control cultures ( ); 379

T. suecica counts in cultures with S26 ( ); T. suecica counts in cultures with DSM17395 (▲).Error bars 380

show the standard deviation of three biological replicates. 381

382

Figure 2. Bacterial growth in non-axenic Tetraselmis suecica cultures during upscaling experiment. CFU 383

counts of total microbiota in T. suecica control cultures ( , dashed line); CFU counts of P. piscinae S26 ( , 384

solid line) and total microbiota (▲, dashed line) in T. suecica cultures with S26: CFU counts of DSM17395 385

(▼, solid line) and total microbiota ( , dashed line) in T. suecica cultures with DSM17395. Error bars show 386

the standard deviation of three biological replicates. 387

388

Figure 3. Vibrio counts in co-culture system with Artemia fed Tetraselmis suecica. V. anguillarum alone (●), 389

V. anguillarum with P. piscinae S26 (■), V. anguillarum with P. inhibens DSM17395 (▲). Error bars show the 390

standard deviation of three biological replicates. 391

392

Figure 4. Phaeobacter counts in co-culture systems with (A) non-challenge and (B) Vibrio-challenged 393

Artemia fed Tetraselmis suecica. (A) P. piscinae S26 (●), P. inhibens DSM17395 (■). (B) P. piscinae S26 with 394

V. anguillarum (●), P. inhibens DSM17395 with V. anguillarum (■). For both systems detection limit at time 395

points 0 and 1 days = 104 (CFU mL-1) and at time points 2, 3, and 4 days = 103 (CFU mL-1). 396

397

Figure 5. Mortality of (A) non-challenged and (B) vibrio-challenged Artemia fed non-axenic Tetraselmis 398

suecica. (A) Non-challenged control (●), Non-challenged; P. piscinae S26 treated (■), Non-challenged; P. 399

inhibens DSM17395 treated (▲). (B) V. anguillarum challenged (●), V. anguillarum challenged; P. piscinae 400

S26 treated (■), V. anguillarum challenged; P. inhibens DSM17395 treated (▲). Error bars show the 401

standard deviation of three biological replicates 402

19

403

404

Figure 1. Tetraselmis suecica counts during upscaling experiment. T. suecica counts in control cultures ( ); 405

T. suecica counts in cultures with S26 ( ); T. suecica counts in cultures with DSM17395 (▲). Error bars 406

show the standard deviation of three biological replicates. 407

408

20

409

Figure 2. Bacterial growth in non-axenic T. suecica cultures during upscaling experiment. CFU counts of 410

total microbiota in T. suecica control cultures ( , dashed line); CFU counts of P. piscinae S26 ( , solid line) 411

and total microbiota (▲, dashed line) in T. suecica cultures with S26; CFU counts of P. inhibens DSM17395 412

(▼, solid line) and total microbiota ( , dashed line) in T. suecica cultures with DSM17395. Error bars show 413

the standard deviation of three biological replicates. 414

415

416

417

418

21

419

Figure 3. Vibrio counts in co-culture system with Artemia fed Tetraselmis suecica. V. anguillarum alone (●), 420

V. anguillarum with P. piscinae S26 (■), V. anguillarum with P. inhibens DSM17395 (▲). Error bars show the 421

standard deviation of three biological replicates. 422

423

424

425

22

426

Figure 4. Phaeobacter counts in co-culture systems with (A) non-challenge and (B) Vibrio-challenged 427

Artemia fed Tetraselmis suecica. (A) P. piscinae S26 (●), P. inhibens DSM17395 (■). (B) P. piscinae S26 with 428

V. anguillarum (●), P. inhibens DSM17395 with V. anguillarum (■). For both systems detection limit at time 429

points 0 and 1 days = 104 (CFU mL-1) and at time points 2, 3, and 4 days = 103 (CFU mL-1). 430

431

432

Figure 5. Mortality of (A) non-challenged and (B) vibrio-challenged Artemia fed non-axenic Tetraselmis 433

suecica. (A) Non-challenged control (●), Non-challenged; P. piscinae S26 treated (■), Non-challenged; P. 434

inhibens DSM17395 treated (▲). (B) V. anguillarum challenged (●), V. anguillarum challenged; P. piscinae 435

S26 treated (■), V. anguillarum challenged; P. inhibens DSM17395 treated (▲). Error bars show the 436

standard deviation of three biological replicates 437

438