See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/261997302

Development of a bacterial challenge test for gnotobiotic Nile tilapia

Oreochromis niloticus larvae

Article  in  Diseases of Aquatic Organisms · April 2014

DOI: 10.3354/dao02721 · Source: PubMed

CITATIONS READS

32 2,133

5 authors, including:

Magdalena Lenny Situmorang Kristof Dierckens

Bandung Institute of Technology Ghent University
22 PUBLICATIONS   189 CITATIONS    83 PUBLICATIONS   1,966 CITATIONS   

SEE PROFILE SEE PROFILE

Frank Thomas Mlingi Bart Van Delsen

Norwegian University of Science and Technology Research Centre for Ornamental Plants
6 PUBLICATIONS   53 CITATIONS    10 PUBLICATIONS   120 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Improving the sturgeon fish feed quality: Effect of survival, growth, digestive enzymes, immune system View project

BOF-GOA View project

All content following this page was uploaded by Frank Thomas Mlingi on 10 September 2019.

The user has requested enhancement of the downloaded file.

 Vol. 109: 23–34, 2014 DISEASES OF AQUATIC ORGANISMS
Published April 23
doi: 10.3354/dao02721 Dis Aquat Org

Development of a bacterial challenge test for

gnotobiotic Nile tilapia Oreochromis niloticus larvae
Magdalena Lenny Situmorang, Kristof Dierckens, Frank Thomas Mlingi,
Bart Van Delsen, Peter Bossier*
Laboratory of Aquaculture & Artemia Reference Center (ARC), Department of Animal Production, Ghent University,
Rozier 44, 9000 Ghent, Belgium

ABSTRACT: Gastrointestinal microbiota have an important impact on fish health and disease,
stimulating interest in a better understanding of how these gastrointestinal microbial communities
are composed and consequently affect host fitness. In this respect, probiotic microorganisms have
been extensively used in recent aquaculture production. To study the use of probiotics in the treat-
ment of infectious diseases, the establishment of a method of experimental infection to obtain con-
sistent results for mortality and infection in challenge tests is important. In pathogen-screening
tests, 4 candidate pathogenic bacteria strains (Edwardsiella ictaluri gly09, E. ictaluri gly10, E.
tarda LMG2793 and Streptococcus agalactiae LMG15977) were individually tested on xenic Nile
tilapia larvae. Only Edwardsiella strains delivered via Artemia nauplii, with or without additional
pathogen delivery via the culture water, led to increased mortality in fish larvae. A gnotobiotic
Nile tilapia larvae model system was developed to provide a research tool to investigate the
effects and modes-of-action of probiotics under controlled conditions. A double disinfection pro-
cedure using hydrogen peroxide and sodium hypochlorite solution was applied to the fish eggs,
which were subsequently incubated in a cocktail of antibiotic and antifungal agents. In the gnoto-
biotic challenge test, E. ictaluri gly09R was added to the model system via Artemia nauplii and
culture water, resulting in a significant mortality of the gnotobiotic fish larvae. The developed
gnotobiotic Nile tilapia model can be used as a tool to extend understanding of the mechanisms
involved in host−microbe interactions and to evaluate new methods of disease control.

KEY WORDS: Gnotobiotic · Bacterial challenge test · Oreochromis niloticus · Edwardsiellosis ·

Edwardsiella ictaluri · Host –microbe interactions
Resale or republication not permitted without written consent of the publisher

INTRODUCTION very important bacterial pathogens that cause severe

economic losses in both freshwater and marine aqua-
Tilapia Oreochromis spp. are an important group in culture in many countries (Ewing et al. 1965, Hawke
freshwater fish aquaculture due in part to their ease et al. 1981, Evans et al. 2011). E. ictaluri, the causative
of culture under a wide range of environmental con- agent of enteric septicaemia of catfish (ESC), also in-
ditions and their relative resistance to environmental fects and causes mortality in Nile tilapia fingerlings
stressors when compared to other cultured finfish (Hawke et al. 1981, Plumb & Sanchez 1983, Keskin et
species (Welker & Lim 2011). However, significant al. 2004, Nagai et al. 2008, Soto et al. 2012). E. tarda,
losses due to disease still occur, especially under in- which is a common bacterium in freshwater environ-
tensive culture where tilapia can become infected ments, infects and causes disease in red tilapia
with a wide range of viral, parasitic, fungal and, par- (Iregui et al. 2012).
ticularly, bacterial pathogens. Bacteria of the genus In some cases, the use of antibiotics for the control of
Edwardsiella, including E. ictaluri and E. tarda, are bacterial diseases in aquaculture has resulted in the

*Corresponding author: [email protected] © Inter-Research 2014 · www.int-res.com

24 Dis Aquat Org 109: 23–34, 2014

development and spread of antibiotic resistance, lead- MATERIALS AND METHODS

ing to ineffective treatment for some diseases (Defoirdt
et al. 2011). This has lead researchers to examine other Experimental set-up for pathogen-screening test
mechanisms for the control of bacterial diseases in fish. with xenic Nile tilapia larvae
As gastrointestinal (GI) bacteria play important roles
in the nutrition and health of the host organism, Nile tilapia Oreochromis niloticus that ranged from
various means of altering the intestinal bacteria to 120 to 150 g wet weight and 21 to 23 cm total length
achieve favourable effects such as better resistance to were naturally bred in our laboratory. Larvae were
pathogens, enhanced growth and immune stimulation collected from the mouths of brooding females 3 d
of the host have been investigated in various species of after hatching (3 DAH) and pooled. Upon collection,
fish and shrimp (Yousefian & Amiri 2009, Sihag & the pooled larvae were acclimatized in a 30 l aquar-
Sharma 2012). In this respect, the use of probiotics and ium for 6 d at a mean (± SD) temperature of 27 ± 1°C
prebiotics is considered to be a sustainable and effec- prior to bacterial challenge. Candidate pathogenic
tive alternative to antibiotics for disease control in bacteria strains were tested individually in triplicate
aquaculture production (Gatesoupe 1999, Verschuere 1.5 l aquaria, each of which contained 10 larvae. All
et al. 2000). In order to be able to discover more effec- aquaria were provided with gentle aeration and kept
tive probiotic bacteria, a better understanding of the in a heated room (constant air temperature of 29 ±
host−microbe interactions of several putative probiotic 1°C), and the water temperature was maintained at
microorganisms through in vivo experiments is essen- 26 ± 1°C. Filtered standard synthetic freshwater con-
tial and still needed (Tinh et al. 2008). taining 96 mg l−1 NaHCO3, 60 mg l−1 CaSO4 · 2H2O,
The composition of the GI microbial community 60 mg l−1 MgSO4 and 4 mg l−1 KCl (USEPA 2002) was
can be influenced by host genotype (Spor et al. 2011, used as fish culture water. To maintain the NH4+ and
Kostic et al. 2013), environmental conditions and sto- NO2− levels below 0.5 and 0.2 mg l−1, respectively,
chastic factors (Verschuere et al. 1999, Fjellheim et 25% of the culture water was renewed every 3 d dur-
al. 2012). This complicates the study of host−microbe ing the 12 d experimental period.
interactions, as the structure of the GI microbial com-
munity of fish reared in conventional rearing systems
is very dynamic, often causing problems in experi- Bacterial strains and culture conditions
ments such as lack of repeatability and reproducibil-
ity. Thus, a key strategy in studying host−microbe Streptococcus agalactiae LMG15977 and 2 strains
interactions is to first determine interactions under of Edwardsiella ictaluri, referred to as gly09 and
axenic conditions then further evaluate the effects of gly10, were tested in the first bacterial pathogen-
a single or defined populations of microbes or speci- screening test (Expt 1a). Another Edwardsiella spe-
fied compounds added under gnotobiotic conditions cies, E. tarda LMG2793, was tested along with the
(Gordon & Pesti 1971). The use of gnotobiotic organ- 2 E. ictaluri strains in the second pathogen-screening
isms leads to an increased control of variables, test (Expt 1b). Both E. ictaluri strains were obtained
enhanced reproducibility of results and more accu- from the Laboratory of Aquaculture & Artemia Refer-
rate experimental designs (Coates 1975, Marques et ence Center (Ghent University). These strains were
al. 2005) and thus can be an excellent tool to extend isolated from ichthyophthriasis (white spot disease)-
the understanding of the mechanisms involved in infected striped catfish Pangasius hypophthalmus.
host−microbe interactions (Marques et al. 2006). S. agalactiae LMG15977, which was previously iso-
The aims of the present work were to (1) screen can- lated from Nile tilapia brain, was obtained from
didate pathogenic bacteria strains on xenic/conven- BCCM/LMG (Belgian Co-ordinated Collections of
tional Nile tilapia Oreochromis niloticus larvae, (2) de- Micro-organisms, Laboratory of Microbiology, Ghent
velop a standardized gnotobiotic Nile tilapia larvae University). An isolate of E. tarda, LMG2793 was pro-
culture system to facilitate the study of host−microbe vided by the Laboratory of Veterinary Bacteriology
interactions, and (3) develop a standardized bacterial and Mycology (Ghent University). This isolate was
challenge test for gnotobiotic Nile tilapia larvae en- obtained from human faeces.
abling further studies on the mode-of-action of putative The strains were stored in brain heart infusion
pre- and probiotics as new ways of disease control, es- (BHI) broth (FLUKA, Sigma-Aldrich) supplemented
pecially in freshwater aquaculture. To our knowledge, with 20% (v/v) glycerol at −80°C. Edwardsiella
this is the first study of a gnotobiotic food chain consist- strains were grown in BHI broth and incubated on a
ing of Nile tilapia larvae and Artemia nauplii. horizontal shaker at 160 rpm and 27°C. Streptococ-
 Situmorang et al.: Challenge test in gnotobiotic Nile tilapia larvae 25

cus agalactiae LMG15977 was grown in BHI broth Disinfection protocol to obtain axenic
and incubated on a horizontal shaker at 180 rpm and Nile tilapia larvae
37°C. The density of each bacterial culture was de-
termined by measuring turbidity with a spectropho- Nile tilapia were naturally bred in our laboratory
tometer (Genesys 20, Thermospectronic) at 550 nm and 3 d post fertilization (3 DPF) eggs were collected
and comparing to the McFarland standard (Bio- and pooled following the procedures described
Mérieux). above for the xenic pathogen-screening tests. Upon
collection, eggs were put on a sterile nylon sieve
(mesh size 300 µm) and washed 4 times with 250 ml
Bacterial challenge procedure with xenic 0.2 µm-filtered autoclaved standard synthetic fresh-
Nile tilapia larvae water (USEPA 2002) at 25 ± 1°C to remove loose bac-
teria. Unfertilized eggs, dead eggs or eggs with rup-
Fish were challenged daily over the experimental tured yolk were discarded prior to the disinfection
period using 3 challenge methods: (1) via culture procedure. A double disinfection procedure was ap-
water, (2) via axenic Artemia nauplii, and (3) via both plied for the remaining eggs at the eyed egg stage
culture water and axenic Artemia nauplii. For the (3 DPF or stage 14 to 15) (Fujimura & Okada 2007).
culture water challenges, bacterial suspensions were In the first disinfection procedure, eggs were immer-
harvested by centrifuging at 1000 × g for 10 min and sed in diluted 30% hydrogen peroxide (MERCK-
washed twice in their respective culture medium Schuchardt 386790) with a final active peroxide con-
then once in fish culture water. The density of bacte- centration of 2000 mg l−1 for 10 min at 25 ± 1°C.
rial suspension was determined by measuring turbid- During disinfection, eggs were gently agitated to
ity with a spectrophotometer at 550 nm and compari- ensure that all eggs had equal contact with the disin-
son to the McFarland standard. Sufficient bacteria fecting agent. Subsequently, the eggs were rinsed
were added to achieve a density of 106 colony form- 4 times with 250 ml 0.2 µm-filtered autoclaved syn-
ing units (CFU) ml−1 in the tank. thetic freshwater. Following the first disinfection,
For the challenge via axenic Artemia nauplii, axe- eggs were incubated in an axenic incubation me-
nic Artemia cysts were incubated and hatched fol- dium which consisted of 0.2 µm-filtered autoclaved
lowing the procedure of Marques et al. (2004). standard synthetic freshwater containing 10 mg l−1
Axenic Artemia nauplii were harvested after 24 h each of ampicillin (Sigma-A0166), rifampicin (Sigma-
incubation at 27 ± 1°C, washed using filtered auto- 83907), trimethoprim (Sigma-T7883) and gentamycin
claved synthetic freshwater and counted. Bacterial (Sigma-G1264), and the antifungal agents Ampho-
suspensions were harvested by centrifuging at 1000 tericin-B (Sigma-A9528) and Fluorescent Brightener
× g for 10 min, washed twice using their respective 28 (Sigma-F3543) at concentrations of 0.5 mg l−1 and
culture medium and added to axenic Artemia culture 25 mg l−1, respectively.
(100 Artemia ml−1) at a final bacteria density of 108 The second disinfection procedure was done 24 h
CFU ml−1. Bacteria-loaded Artemia nauplii were after the first. A 75 mg l−1 active chlorine solution was
harvested after 1 h incubation and washed twice prepared using sodium hypochlorite technical grade
using sterile saline solution (9 g l−1 NaCl), counted (Sigma Aldrich 425044) and 0.2 µm-filtered auto-
and added at a density of 20 Artemia per fish. The claved synthetic freshwater. Eggs were immersed in
bacterial load of the Artemia nauplii in each treat- this solution for 2 min at 25 ± 1°C. The concentration
ment was determined by the plate count method. of active chlorine in the sodium hypochlorite solution
After rinsing and counting, subsamples of bacteria- was determined 24 h prior to use by a standard
loaded Artemia were homogenized according to the iodine/thiosulfate titration method (Lavens & Sorge-
procedure described by Huys et al. (2001). Subse- loos 1996). Following disinfection, eggs were rinsed
quently, 50 µl of the homogenized Artemia suspen- 4 times with 250 ml 0.2 µm-filtered autoclaved syn-
sion was plated (Spiral platerTM, Spiral Systems) on thetic freshwater. All disinfection procedures were
Lysogeny broth (LB) agar. performed in a laminar flow hood.
For the challenge via both Artemia nauplii and cul- Disinfected eggs were aseptically distributed to
ture water, larvae received a waterborne challenge 500 ml sterile glass bottles containing 200 ml incu-
as well as a challenge via axenic Artemia adminis- bation medium and incubated at a density of 500
tered at the same time. These challenges were con- eggs l−1. Fish were maintained in the axenic incuba-
ducted as described above. Mortality observation tion medium throughout the non-feeding larval
and dead fish removal were done twice daily. stage. Eggs of the control group underwent the
26 Dis Aquat Org 109: 23–34, 2014

same incubation procedure; however, they were not resistant mutants that could be used for challenge.
surface disinfected or treated with antibiotic and We investigated Edwardsiella ictaluri gly09 resist-
antifungal agents. Each incubation bottle was ance to the different antibiotics by inoculating an
equipped with 2 sterile 0.2 µm filters (Sartorius) for overnight E. ictaluri gly09 culture into sterile tubes
the aeration in- and outlet to provide gentle sterile containing BHI broth and 1 of the 5 antibiotics at a
aeration during the egg incubation and larval stage. concentration of 10 mg l−1. These cultures were incu-
Eggs hatched after 3 d incubation following the dis- bated overnight on a horizontal shaker at 160 rpm at
infection procedure. The effect of the disinfection 27°C. Growth, determined by the presence of turbid-
procedure on egg hatching was evaluated using the ity, was obtained in the presence of 10 mg l−1 ampi-
egg hatching percentage, which was measured as cillin, trimethoprim and gentamycin, suggesting that
the proportion of the total number of eggs incubated E. ictaluri gly09 is intrinsically resistant to these
that hatched, regardless of their viability (Komar et antibiotics. No growth occurred in the presence of
al. 2004). The hatching percentage was determined 10 mg l−1 rifampicin or kanamycin.
1 DAH. In order to obtain spontaneous rifampicin-resistant
and kanamycin-resistant Edwardsiella ictaluri gly09
mutants, wild type E. ictaluri gly09 was cultured
Tests for axenity separately on BHI agar plates containing 2 mg l−1 of
each antibiotic. Colonies appearing on the BHI-
Axenity was checked at several crucial steps dur- rifampicin and BHI-kanamycin plates after an incu-
ing the experiments. After 24 h following the disin- bation period ranging from 48 to 72 h at 27°C were
fection procedure, 5 eggs were aseptically sampled harvested, mixed and transferred into several tubes
from each incubation bottle, individually homoge- containing BHI broth with 2 mg l−1 of both anti-
nized and plated on LB medium + 15 g l−1 agar (LB biotics. Cultures were incubated on a horizontal
agar; bacteriological grade, MP Biomedicals). In shaker at 160 rpm and 27°C overnight, after which
addition, 1 ml water from each incubation bottle was time 1 ml samples were transferred into new tubes
added into a sterile tube containing 9 ml LB broth containing 9 ml BHI broth with a higher antibiotic
(10%). After determination of the larval survival, the concentration of 5 mg l−1. Following overnight incu-
fish larvae from the axenic treatments were checked bation, selected grown cultures (cultures with an
for bacterial contamination using the plate culture absorbance level higher than 0.5 at a wavelength of
method. From each incubation bottle, 2 larvae were 600 nm) were then transferred into new tubes con-
euthanized using 1 g l−1 benzocaine and 1 g l−1 ben- taining 9 ml BHI broth with an even higher antibi-
zalkonium chloride before being rinsed and homoge- otic concentration of 10 mg l−1.
nized in sterile saline solution (9 g l−1 NaCl). Subse- Cultures of Edwardsiella ictaluri gly09 that were
quently, 50 µl of the homogenized fish suspension resistant to 10 mg l−1 rifampicin and kanamycin were
was plated (Spiral platerTM, Spiral Systems) on LB transferred and grown in BHI broth containing 10 mg
agar plates. The inoculated plates were incubated at l−1 ampicillin, rifampicin, kanamycin, trimethoprim
27 ± 1°C for 96 h. At stocking of the larvae (6 DAH) and gentamycin. The resulting E. ictaluri strain
and at the end of the gnotobiotic challenge test, the (referred to as E. ictaluri gly09R) with multiple anti-
fish larvae from the axenic treatments were checked biotic resistance was used in gnotobiotic challenge
for bacterial contamination using the above-men- tests. Axenic fish larvae were challenged with E.
tioned procedure. ictaluri gly09R via both axenic Artemia nauplii and
culture water, following the procedures applied in
the pathogen-screening test on xenic larvae.
Bacterial challenge procedure with gnotobiotic The gnotobiotic challenge tests were done in tripli-
Nile tilapia larvae cate (n = 3) in 500 ml sterile glass bottles containing
200 ml incubation medium with 10 fish per bottle.
Along with the antifungal agents (0.5 mg l−1 Each incubation bottle was equipped with 2 sterile
Amphotericin-B and 25 mg l−1 Fluorescent Bright- 0.2 mm air filters (Sartorius) to provide gentle sterile
ener 28), a mixture of antibiotics containing ampi- aeration and a sterile septum for aseptic larval feed-
cillin, rifampicin, kanamycin, trimethoprim and gen- ing during the experimental period. An unchallenged-
tamycin (each at 10 mg l−1) was used in the larvae axenic fish group was used as the control group for
medium throughout the gnotobiotic challenge tests. the gnotobiotic system. In order to verify that there
Therefore, it was necessary to identify antibiotic was no effect of the culture system set-up on patho-
 Situmorang et al.: Challenge test in gnotobiotic Nile tilapia larvae 27

gen virulence, in Expt 2a, the Edwardsiella ictaluri RESULTS AND DISCUSSION
gly09R strain was also tested on xenic larvae using
the same culture system set-up but without the air Pathogen-screening tests on xenic
0.2 µm filtration or the use of both antibiotics and anti- Nile tilapia larvae
fungal mixtures. To maintain conditions in the exper-
imental system, 25% of the incubation medium with Streptococcus sp. and Edwardsiella sp. are the
or without antimicrobials was replaced every 3 d. common groups of bacteria reported to infect wild
The gnotobiotic challenge test was repeated twice and farmed tilapia (Plumb & Hanson 2010). E. tarda
(Expts 2a and 2b). In Expt 2a, Edwardsiella ictaluri infects and causes significant pathology in red tilapia
gly09R was tested on both the xenic and gnotobiotic (Iregui et al. 2012). In the last few years, different
fish groups, while in Expt 2b, it was only tested on strains of E. ictaluri, which were already known as
the gnotobiotic group as there were insufficient num- etiological agents of ESC, have also been reported as
bers of xenic larvae due to a poor hatching of the causative agents of morbidity and mortality in Nile
xenic eggs. tilapia fingerlings (Soto et al. 2012). In this study, 4
In order to determine larval bacterial load at the candidate pathogenic bacteria strains representing
end of the challenge test, surviving larvae were the 2 groups were screened for the development of a
killed using an overdose of benzocaine, surface-dis- bacterial challenge test for gnotobiotic Nile tilapia
infected in a bath of benzalkonium chloride, rinsed larvae: E. ictaluri gly09, E. ictaluri gly10, E. tarda
and homogenized according to the procedure de- LMG2793 and S. agalactiae LMG15977.
scribed by Huys et al. (2001). Subsequently, 50 ml of In the first pathogen-screening test (Expt 1a), mor-
the homogenized larval suspension was plated (Spi- tality was first observed in the group challenged with
ral platerTM, Spiral Systems) on BHI agar containing Edwardsiella ictaluri gly09 via Artemia nauplii and
10 mg l−1 ampicillin, rifampicin, kanamycin, trimetho- in the group challenged with E. ictaluri gly10 via the
prim and gentamycin. The experimental designs for culture water 6 d after challenge (6 DAC). With
bacterial challenge tests on Nile tilapia larvae were respect to E. ictaluri gly09, there were no significant
approved by the ethical committee of Ghent Univer- differences between treatments (challenge routes
sity under file number EC2012/070 for the xenic chal- and control) until 11 DAC when significantly higher
lenge tests (Expts 1a and 1b) and EC2013/69 for the levels of mortality were seen in the groups chal-
gnotobiotic challenge tests (Expts 2a and 2b). lenged via Artemia and via both Artemia and culture
water when compared to control and the groups
challenged via culture water (Table 1). A similar
Statistical analysis result was observed for E. ictaluri gly10, with the
exception that the group challenged via both Arte-
A chi-square test was used to detect significant dif- mia and culture water had significantly higher
ferences in the hatching percentage between the mortalities than all other treatment groups and the
xenic and the disinfected eggs. For comparison of the control at 10 DAC. At 11 DAC, significantly higher
cumulative mortality of fish larvae, data were arcsine levels of mortality were observed in the groups chal-
transformed before a 1-way analysis of variance lenged via Artemia and via both Artemia and culture
using the general linear model of STATISTICA 7.0 water when compared to the control and the group
(StatSoft 2004) was performed. A Tukey test was per- challenged via culture water (Table 1).
formed on the transformed data for multiple compar- The pathogenicity of both Edwardsiella ictaluri
isons among means (Sokal & Rohlf 1955). All analy- strains for tilapia larvae was confirmed in Expt 1b
ses were run at a minimum level of significance (Table 2), where oral and waterborne/immersion
of 5%. Results are reported as mean ± standard exposure of larvae to either E. ictaluri gly09 or gly10
deviation. resulted in a significant mortalities of 93 ± 11 to 100 ±
The relative percentage of survival (RPS), which is 0%, respectively when compared to the 40 ± 17%
the larval survival after challenge when compared to mortality seen in the control group at 9 DAC. The
control fishes (Amend 1981), was calculated to evalu- results of both pathogen-screening tests suggest that
ate the efficacy of xenic and gnotobiotic bacterial both E. ictaluri strains can cause significant larval
challenge tests as RPS = [1 − (% mortality in chal- mortalities when delivered orally through the feed-
lenged group/% mortality in control group)] × 100. ing of pathogen-loaded Artemia nauplii, with or
The RPS values obtained in the challenge studies without additional pathogen delivery via the culture
were analysed using a chi-square test. water; while lower or no significant mortality was
 28 Dis Aquat Org 109: 23–34, 2014

Table 1. Cumulative mortality (%, mean ± SD) of control and challenged (Edwardsiella ictaluri gly09, E. ictaluri gly10 and Strep-
tococcus agalactiae LMG15977) xenic Nile tilapia larvae Oreochromis niloticus using different challenge methods in Expt 1a. Dif-
ferent letters within the same row denote significant differences (p < 0.05). Number of replicates = 3; initial number of fish
larvae per replicate = 10. DAC: days after challenge

Time Control Edwardsiella ictaluri gly09 Edwardsiella ictaluri gly10 Streptococcus agalactiae LMG15977
(DAC) via via via via via via via via via
culture Artemia Artemia culture Artemia Artemia culture Artemia Artemia
water nauplii and water water nauplii and water water nauplii and water

6 0 ± 0a 0 ± 0a 4 ± 6a 0 ± 0a 4 ± 6a 0 ± 0a 0 ± 0a 0 ± 0a 0 ± 0a 0 ± 0a
7 0 ± 0a 4 ± 6a 7 ± 6a 0 ± 0a 4 ± 6a 4 ± 6a 0 ± 0a 0 ± 0a 0 ± 0a 4 ± 6a
8 0 ± 0a 33 ± 29a 30 ± 17a 18 ± 6a 18 ± 13a 41 ± 28a 33 ± 29a 18 ± 23a 4 ± 6a 15 ± 17a
9 26 ± 26a 44 ± 22a 56 ± 11a 52 ± 6a 41 ± 6a 63 ± 28a 74 ± 6a 22 ± 29a 26 ± 28a 37 ± 26a
10 37 ± 23a 59 ± 6a 74 ± 13a 81 ± 6a 48 ± 6a 74 ± 17a 85 ± 6b 44 ± 22a 48 ± 28a 63 ± 17a
11 48 ± 13a 67 ± 0a 81 ± 6b 81 ± 6b 70 ± 6a 81 ± 13b 89 ± 0b 67 ± 19a 70 ± 13a 74 ± 6a

Table 2. Cumulative mortality (%, mean ± SD) of control and kidney as the main targets of infection. Their obser-
challenged (Edwardsiella ictaluri gly09, E. ictaluri gly10 and vations support the study by Li et al. (2012), which
E. tarda LMG2793) xenic Nile tilapia larvae Oreochromis suggested that E. ictaluri gains entry through the
niloticus via both Artemia nauplii and culture water in
intestinal epithelium, possibly using actin polymer-
Expt 1b. Different letters within the same row denote signif-
icant differences (p < 0.05). Number of replicates = 3; initial ization and receptor-mediated endocytosis as mecha-
number of fish larvae per replicate = 10. DAC = days after nisms of invasion.
challenge Bullock & Herman (1985) suggested that E. ictaluri
is more pathogenic than E. tarda to channel catfish
Time Control E. ictaluri E. ictaluri E. tarda Ictalurus punctatus fingerlings. In this study, a signif-
(DAC) gly09 gly10 LMG2793 icantly higher larval mortality was observed in the
group challenged with E. tarda LMG2793 via
5 0 ± 0a 3 ± 6a 0 ± 0a 0 ± 0a
6 0 ± 0a 17 ± 6a 17 ± 21a 3 ± 6a Artemia and culture water when compared to the
7 0 ± 0a 37 ± 31ab 47 ± 25b 60 ± 17b control group, starting from 7 DAC and with a the
8 27 ± 23a 77 ± 6b 100 ± 0c 83 ± 15b final mortality of 93 ± 11% at 9 DAC (Table 2).
9 40 ± 17a 93 ± 11b 100 ± 0b 93 ± 11b In both of the pathogen-screening tests, the first
fish mortalities in the treatment group challenged
with Edwardsiella ictaluri occurred at 5 to 6 DAC.
observed when the fish were only exposed to the However, the mortality from the challenges via both
pathogens via the culture water. In the study by Soto Artemia and culture water in Expt 1a differed when
et al. (2012), tilapia fingerlings (~15 g) immersed in compared to Expt 1b, where significant differences
106 CFU ml−1 of tank water presented significant in mortality following challenge occurred as early as
100% mortality events at 8 d post challenge. The 7 DAC. The differences in the incubation period until
results of this current study suggest that bacterial significant mortalities were observed could be
infection via water culture (immersion challenge) caused by the presence of other microorganisms be-
in tilapia larvae culture is less likely than in adult sides the tested pathogens or a difference in patho-
culture. gen resistance between the different egg batches.
Due to the emergent nature of edwardsiellosis in With respect to Streptococcus agalactiae LMG
non-ictalurid fish, little is known regarding the 15977, which was the only isolate actually isolated
dynamics of Edwardsiella ictaluri infection in Nile from tilapia culture, there were no significant differ-
tilapia culture, especially during its larviculture ences between treatments (challenge routes) and the
stage. To our knowledge, this is the first report of E. control group at 11 DAC. However, significantly
ictaluri-induced mortality in Nile tilapia larviculture. lower mortalities were observed when compared to
In a further study, Soto et al. (2013) increased our the groups challenged with Edwardsiella ictaluri
understanding of the pathogenesis of E. ictaluri in strains via Artemia and via both Artemia and culture
Nile tilapia fingerlings, suggesting that the cuta- water (Table 1). It is possible that the pathogenicity of
neous and oral routes were the main ports of entry for S. agalactiae in this experiment was affected by the
the bacterium, which later spreads haematogenously water temperature, which was below the optimal
throughout the fish body, with the spleen and head growth temperature for S. agalactiae culture. A study
 Situmorang et al.: Challenge test in gnotobiotic Nile tilapia larvae 29

by Mereghetti et al. (2008) revealed that the tran- tles. As suggested by Rana (1988) and El-Sayed
scription of some important virulence factors by (2006), in cases where eggs are not suspended in the
human S. agalactiae increased at higher temperature water by a current, they quickly sink and clump. This
(40°C compared with 30°C). Our results are also sup- may be the reason for the relatively low egg hatching
ported by Rodkhum et al. (2011), who evaluated the of the non-disinfected eggs, ranging from 11 to 27%,
association between water temperature and suscep- compared to hatching percentages of > 60% using
tibility of Nile tilapia to Streptococcus agalactiae down-welling round-bottomed incubators or up-
infection. In their study, Nile tilapia (100 g) were bath welling conical containers reported previously (Rana
challenged with 106, 107 or 108 CFU ml−1 of S. agalac- 1988, Rana & Macintosh 1988). With respect to these
tiae serotype Ia and maintained at different water design and operational requirements, there were
temperatures (25, 30 or 33°C) for 1 wk. Cumulative several constraints on the egg incubation system de-
mortality of tilapia was positively correlated to higher sign for this research on the gnotobiotic Nile tilapia
temperature, while no clinical signs of disease were larviculture. These included the need for a gentle mo-
exhibited at 25°C. These results indicate that the sus- tion of the eggs and the need for an axenic closed sys-
ceptibility of Nile tilapia to S. agalactiae infection is tem with easy access for removal of non-developing
temperature dependent. embryos and also refreshment of sterile culture
To study the use of pro- and/or prebiotics in the medium. For both gnotobiotic challenge tests, a
prevention and treatment of infectious diseases, higher (p > 0.05) egg hatching was observed in the
establishing a method of experimental infection to disinfected egg group compared to the xenic group,
obtain steady results in mortality and infection in indicating that the disinfection procedure improved
challenge tests is important. The results of both egg hatching in our experiments.
pathogen-screening tests showed that all Edward- In gnotobiotic studies with fish, the common
siella strains tested in this study can cause significant method for obtaining axenic larvae is to collect fertil-
mortality in tilapia larvae culture and can be used in ized eggs, disinfect them and then incubate them in a
the development of bacterial challenge tests with cocktail of disinfectants and antibiotics. Egg surface
gnotobiotic Nile tilapia larvae, with the challenge disinfection protocols using glutaraldehyde, at differ-
route via both Artemia and culture water as an effec- ent concentrations and exposure times, have been
tive challenge procedure. used successfully to produce axenic/gnotobiotic lar-
val marine fish, including turbot Scophthalmus max-
imus, Atlantic halibut Hippoglossus hippoglossus,
Egg disinfection protocol for axenic European sea bass Dicentrarchus labrax and Atlantic
Nile tilapia larvae cod Gadus morhua L. larvae (Munro et al. 1995,
Verner-Jeffreys et al. 2003, Dierckens et al. 2009,
The effect of the disinfection procedure on egg Forberg et al. 2011). For the purposes of this study,
hatching was evaluated by measuring the hatching we chose not to use glutaraldehyde due to its high
percentage at 1 DAH for the xenic and the disinfec- diffusion rate over the range of water temperatures
ted eggs. There was no significant effect of the disin- used for tilapia culture (Salvesen et al. 1997).
fection procedure on the egg hatching in both gnoto- A protocol to obtain bacteria-free zebrafish Danio
biotic experiments (mean ± SD): 32 ± 14% or 23 ± 3% rerio for use in gnotobiotic studies has been estab-
and 20 ± 7% or 11 ± 6% for the disinfected or the lished, in which fertilized embryos are surface-
xenic eggs of Expts 2a and 2b, respectively. disinfected using 1000 mg l−l povidone-iodine (PVP-I)
In order to obtain axenic larvae, artificial incubation solution for 2 min and 30 mg l−l sodium hypochlorite
of surface-disinfected eggs in axenic conditions until solution for 20 min to 1 h (Pham et al. 2008). A higher
hatching is crucial. The efficiency of a tilapia egg in- sodium hypochlorite concentration of 100 mg l−l for
cubation system/incubator depends on its type, size the disinfection of tropical marine red drum Sci-
and shape, the developmental stages of the eggs, and aenops ocellatus eggs has been reported to increase
the water quality and flow (El-Sayed 2006). It is im- larval survival, while disinfection using a 5 min expo-
portant that eggs are kept in gentle motion imitating sure to 3% hydrogen peroxide resulted in bacteria-
the natural incubation inside the mouths of female free larvae (Douillet & Holt 1994). In terms of antifun-
broodstock (Rana & Beveridge 1989). In this study, gal treatment, the use of hydrogen peroxide at
unlike in xenic incubators, egg incubation was done various concentrations from 100 up to 6000 mg l−l has
under static conditions with periodic water renewal been shown to also effectively control saprolegniasis
using a series of autoclavable flat-bottom glass bot- in cultured fish eggs, including rainbow trout Onco-
30 Dis Aquat Org 109: 23–34, 2014

rhynchus mykiss, common carp Cyprinus carpio and challenge tests using the gnotobiotic system, the E.
channel catfish Ictalurus punctatus (Marking et al. ictaluri gly09R loaded onto the Artemia nauplii
1994, Schreier et al. 1996, Barnes et al. 1998, Rach et resulted in densities of 4.5 ± 0.9 × 103 and 8.8 ± 0.5 ×
al. 1998, Small & Wolters 2003, Mitchell et al. 2009). 102 CFU ind.−1 after 1 h incubation in Expts 2a
Samples of homogenized eggs that were taken 24 h and 2b, respectively.
after disinfection never resulted in growth on LB Bacterial colonization of Artemia nauplii could
plates after 96 h incubation, indicating egg/larvae occur externally, via attachment to the body surfaces
axenity following the egg disinfection procedure. or internally by ingestion (Grisez et al. 1996). In this
Samples of the incubation medium or homogenized experiment, however, only instar I Artemia nauplii
larvae at larval stocking also did not show any were used. The digestive system of instar I Artemia
colonies on LB plates after 96 h incubation. This indi- nauplii is not yet functional (mouth and anus are not
cated that the application of the double disinfection open) so they do not take up food and thrive com-
protocol using hydrogen peroxide (2000 mg l−1 for pletely on their yolk reserves (Lavens & Sorgeloos
10 min) and sodium hypochlorite (75 mg active chlo- 1996). Therefore, it is most likely that bacterial colo-
rine l−1 solution for 2 min), followed by incubation in nization occurred externally, via attachment to the
medium containing antibiotics (10 mg l−1 ampicillin, body surfaces. It has been suggested that bacterial
rifampicin, trimethoprim and gentamycin) and anti- content decreases rapidly after the nauplii are
fungal products (0.5 mg l−1 Amphotericin-B and removed from the bacterial suspension (Gomez-Gil
25 mg l−1 Fluorescent Brightener 28) was effective in et al. 1998). This decrease might be due to the re-
obtaining axenic Nile tilapia larvae. moval of the external bacteria after the nauplii are
To our knowledge, this is the first published proto- washed and placed in new sterile water. The bacte-
col for generating axenic Nile tilapia larvae with con- ria still detected are those colonizing the interior or
tinued exposure to antimicrobial agents throughout firmly attached to the external surfaces. The differ-
the gnotobiotic experiment. In the course of the ences in bacteria load of Artemia nauplii between
experiment, fish samples were taken from the axenic treatments and between experiments may be
control treatment for bacteriology. No bacteria were related to differences in bacterial attachment on
detected in the samples of the axenic fish at the end Artemia nauplii during incubation. Waltman et al.
of Expt 2a, as verified by fish homogenate plating on (1986a,b) suggested that although both Edward-
the LB plates. A very low level of bacterial contami- siella ictaluri and E. tarda lack chitinase and chitin-
nation was observed in the axenic group at the end of binding proteins, 100% and 56% of 100 strains of
Expt 2b, as bacteria colonies grew on the LB plates both E. ictaluri and E. tarda, respectively, degraded
after 96 h incubation at a density of < 30 CFU per fish. chondroitin sulphate, which is also an important
This bacterial contamination might be related to the structural component of Artemia nauplii (Forward &
technical complexity of the system (Marques et al. Rittschof 1999). At the end of the gnotobiotic chal-
2006) where handlings and manipulations, such as lenge test, the densities of E. ictaluri gly09R inside
the culture medium exchange and daily feeding dur- the surviving fish was 7.9 ± 0.9 × 104 and 4.8 ± 0.6 ×
ing the experimental period, might have resulted in 103 CFU per fish in Expts 2a and 2b, respectively.
contamination.

Mortality of the gnotobiotic challenged

Quantification of the bacteria load in Artemia Nile tilapia larvae
nauplii and Nile tilapia larvae
In both gnotobiotic challenge tests, the mortality
In Expt 1a, incubation of axenic Artemia nauplii of the fish group after challenge with Edwardsiella
with Edwardsiella ictaluri gly09, E. ictaluri gly10 ictaluri gly09R was not significantly different (p >
and Streptococcus agalactiae LMG15977 resulted in 0.05) from the unchallenged (axenic or xenic) fish
densities of 7.8 ± 0.2, 5.2 ± 0.1 and 6.5 ± 0.1 × 102 up to and including 9 DAC (Table 3). From 10
CFU per Artemia nauplius (ind.−1) after 1 h incuba- DAC onwards, significantly higher mortalities were
tion, respectively. Similar results were observed in observed in this challenged group (p < 0.05) com-
Expt 1b, where incubation of axenic Artemia nauplii pared to the unchallenged (axenic) group in both
resulted in densities of 6.9 ± 0.7, 5.1 ± 0.4 and 7.7 ± gnotobiotic challenge tests, with a mortality at
0.2 × 102 CFU ind.−1 for E. ictaluri gly09, E. ictaluri 12 DAC of 57 ± 6% or 23 ± 6% (challenged versus
gly10 and E. tarda LMG2793, respectively. In the axenic, Expt 2a) and 63 ± 11% or 23 ± 6% (chal-
 Situmorang et al.: Challenge test in gnotobiotic Nile tilapia larvae 31

Table 3. Cumulative mortality (%, mean ± SD) of xenic (not challenged or an additional cost for the strain that
challenged), axenic and gnotobiotic Nile tilapia larvae Oreochromis niloticus can result in a longer generation time
challenged with antibiotic-resistant Edwardsiella ictaluri gly09R via Artemia
nauplii and culture water in Expts 2a and 2b. Different letters within the same and altered metabolic activity (Ander-
row within 1 experiment denote significant differences (p < 0.05). Number of sson & Levin 1999, Levin et al. 2000,
replicates = 3; initial number of fish larvae per replicate = 10. DAC: days after McDermott et al. 2006, Feng & Rise
challenge 2010). This implies that a bacterial
strain with an introduced resistance
Time Expt 2a Expt 2b may behave differently than the orig-
(DAC) Xenic Xenic Axenic Gnotobiotic Axenic Gnotobiotic inal strain. However, the Edward-
E. ictaluri E. ictaluri E. ictaluri
gly09R gly09R gly09R siella ictaluri gly09R strain used in
this study resulted in significant mor-
5 0 ± 0a 3 ± 6a 0 ± 0a 0 ± 0a 0 ± 0a 0 ± 0a talities in the gnotobiotic challenged
6 3 ± 6a 3 ± 6a 0 ± 0a 0 ± 0a 0 ± 0a 0 ± 0a groups compared to the unchal-
7 7 ± 6a 3 ± 6a 0 ± 0a 0 ± 0a 0 ± 0a 0 ± 0a
a a a a a a lenged/axenic groups, indicating that
8 7±6 10 ± 10 0±0 0±0 7±6 10 ± 10
9 7 ± 6a 10 ± 10a 0 ± 0a 3 ± 6a 13 ± 6a 23 ± 11a the bacterial pathogenicity was pre-
10 7 ± 6ab 13 ± 6b 0 ± 0a 13 ± 6b 20 ± 0a 53 ± 15b served after the acquisition of antibi-
11 7 ± 6a 23 ± 6ab 10 ± 10ab 33 ± 15b 20 ± 0a 60 ± 10b otic resistance.
12 13 ± 6a 40 ± 10bc 23 ± 6ab 57 ± 6c 23 ± 6a 63 ± 11b
There was no significant difference
(p > 0.05) in mortality of the gnotobi-
otic group challenged with Edward-
lenged versus axenic, Expt 2b) (Table 3). Challenge siella ictaluri gly09R when compared to the xenic
with E. ictaluri gly09R under xenic conditions challenged group at 12 DAC (57 ± 6% vs. 40 ± 10%;
resulted in a mortality (p < 0.05) of 40 ± 10%, sig- Table 3). Similar results were also observed with
nificantly higher than that of the xenic unchal- respect to RPS: values of 67 and 73% (p > 0.05) were
lenged group, but no significant differences were obtained for the gnotobiotic and xenic challenged
observed when compared to the gnotobiotic chal- groups, respectively. Several studies using gnoto-
lenged group (12 DAC; Expt 2a). biotic culture systems have suggested that the cellu-
In this laboratory test system, we relied on antibi- lar and humoral defence systems of axenic animals
otics to maintain gnotobiotic conditions. This resulted appear to be underdeveloped, hence making them
in some limitations, as resistant spontaneous mutants more susceptible to disease compared to xenic ani-
need to be isolated for use in challenge trials. Such mals (Coates 1975, Marques et al. 2006). Unlike these
spontaneous mutations, however, have been found in previous studies, the results of the current study did
several Gram-negative bacteria, including Edward- not show that axenic Nile tilapia larvae are more
siella species (Ingham & Furneaux 2000, Sikorski & susceptible to pathogen infection compared to xenic
Nevo 2005, Thavasi et al. 2007). E. ictaluri strains larvae, as indicated by the similar RPS values of the
have been reported which are naturally resistant gnotobiotic and xenic challenged groups. More
to rifampicin, macrolides, lincosamides, strepto- importantly, this study showed that the bacterial
gramins, glycopeptides, fusidic acid, oxacillin (Stock challenge test in a gnotobiotic system allows a
& Wiedemann 2001) and gentamycin (Reger et al. similar/comparable result with a xenic/conventional
1993); and also E. ictaluri isolates displaying ac- open system, but without uncontrolled interference.
quired resistance to trimethoprim, streptomycin, oxy- In conclusion, the Nile tilapia gnotobiotic system
tetracycline (Dung et al. 2008), kanamycin (Russo et developed here is a useful tool for extending under-
al. 2009) and ampicillin (Russo 2011). In this study, standing of the mechanisms involved in host−
the multiple antibiotic resistance of E. ictaluri gly09R microbe interactions and evaluating new methods of
resulted from both intrinsic mechanism (for ampi- disease control in aquaculture production.
cillin, trimethoprim and gentamycin) and sponta-
neous mutation (for rifampicin and kanamycin). This
E. ictaluri gly09R strain was used in both gnotobiotic Acknowledgements. This work was performed and funded
challenge tests (Expts 2a and 2b) and resulted in sig- under financial support from a Doctoral Grant to M.L.S. from
nificant larvae mortality. The World Bank, USA, under the Japan Indonesia Presiden-
tial Scholarship (JIPS) Program. Additional funding was
When introducing antibiotic resistance to a strain obtained from the frame of the European FP7 project ‘Promi-
of bacteria, either via plasmids or by selecting spon- crobe — Microbes as positive actors for more sustainable
taneous mutants, the acquired resistance represents aquaculture’ (project reference: 227197).
 32 Dis Aquat Org 109: 23–34, 2014

LITERATURE CITED in nauplii of the brine shrimp (Artemia franciscana). Appl

Environ Microbiol 64:2318−2322
Amend DF (1981) Potency testing of fish vaccines. Dev Biol ➤ Gordon HA, Pesti L (1971) The gnotobiotic animal as a tool
Stand 49:447−454 in the study of host microbial relationships. Bacteriol Rev
➤ Andersson DI, Levin BR (1999) The biological cost of anti- 35:390−429
biotic resistance. Curr Opin Microbiol 2:489−493 ➤ Grisez L, Chair M, Sorgeloos P, Ollevier F (1996) Mode of
➤ Barnes ME, Ewing DE, Cordes RJ, Young GL (1998) Obser- infection and spread of Vibrio anguillarum in turbot
vations on hydrogen peroxide control of Saprolegnia Scophthalmus maximus larvae after oral challenge
spp. during rainbow trout egg incubation. Prog Fish-Cult through live feed. Dis Aquat Org 26:181−187
60:67−70 ➤ Hawke JP, McWhorter AC, Stegerwalt AG, Brenner DJ
Bullock GL, Herman RL (1985) Edwardsiella infections of (1981) Edwardsiella ictaluri sp. nov., the causative agent
fishes. Fish Disease Leaflet 71. US Fish and Wildlife of enteric septicemia of catfish. Int J Syst Bacteriol 31:
Service, Division of Fishery Research 396−400
➤ Coates ME (1975) Gnotobiotic animals in research: their Huys L, Dhert P, Robles R, Ollevier F, Sorgeloos P, Swings J
uses and limitations. Lab Anim 9:275−282 (2001) Search for beneficial bacterial strains for turbot
➤ Defoirdt T, Sorgeloos P, Bossier P (2011) Alternatives to (Scophthalmus maximus L.) larviculture. Aquaculture
antibiotics for the control of bacterial disease in aquacul- 193:25–37
ture. Curr Opin Microbiol 14:251−258 Ingham CJ, Furneaux PA (2000) Mutations in the β subunit
➤ Dierckens K, Rekecki A, Laureau S, Sorgeloos P, Boon N, of the Bacillus subtilis RNA polymerase that confer both
Van den Broeck W, Bossier P (2009) Development of a rifampicin resistance and hypersensitivity to NusG.
bacterial challenge test for gnotobiotic sea bass (Dicen- Microbiology 146:3041−3049
trarchus labrax) larvae. Environ Microbiol 11:526−533 ➤ Iregui CA, Guarin M, Tibata VM, Ferguson HW (2012)
➤ Douillet PA, Holt GJ (1994) Surface disinfection of red drum Novel brain lesions caused by Edwardsiella tarda in a
(Sciaenops ocellatus Linnaeus) eggs leading to bacteria- red tilapia (Oreochromis spp.). J Vet Diagn Invest 24:
free larvae. J Exp Mar Biol Ecol 179:253−266 446−449
➤ Dung TT, Haesebrouck F, Tuan NA, Sorgeloos P, Baele M, Keskin O, Secer S, Izgur M, Turkyilmaz S, Mkakosya R
Decostere A (2008) Antimicrobial susceptibility pattern (2004) Edwardsiella ictaluri infection in rainbow trout
of Edwardsiella ictaluri isolates from natural outbreaks of (Oncorhynchus mykiss). Turk J Vet Anim Sci 28:649−653
bacillary necrosis of Pangasianodon hypophthalmus in ➤ Komar C, Turnbull JF, Roque A, Fajer E, Duncan NJ (2004)
Vietnam. Microb Drug Resist 14:311−316 Effect of water treatment and aeration on the percentage
El-Sayed AFM (2006) Tilapia culture. CABI Publishing, hatch of demersal, adhesive eggs of the bullseye puffer
Wallingford (Sphoeroides annulatus). Aquaculture 229:147−158
Evans JJ, Klesius PH, Plumb J, Shoemaker CA (2011) ➤ Kostic AD, Howitt MR, Garrett WS (2013) Exploring host−
Edwardsiella septicaemias. In: Woo PTK, Bruno DW microbiota interactions in animal models and humans.
(eds) Fish diseases and disorders, Vol 3: bacterial, viral Genes Dev 27:701−718
and fungal infections. CABI Publishing, Wallingford, Lavens P, Sorgeloos P (1996) Manual on the production and
p 512−534 use of live food for aquaculture. FAO Fish Tech Paper
Ewing WH, McWhorter AC, Escobar MR, Lubin AH (1965) No. 361. FAO, Rome
Edwardsiella, a new genus of Enterobacteriaceae based ➤ Levin BR, Perrot V, Walker N (2000) Compensatory muta-
on a new species, E. tarda. Int Bull Bacteriol Nomencl tions, antibiotic resistance and the population genetics of
Taxon 15:33−38 adaptive evolution in bacteria. Genetics 154:985−997
➤ Feng CY, Rise ML (2010) Characterization and expression ➤ Li C, Zhang Y, Wang R, Lu J, Nandi S (2012) RNA-seq analy-
analyses of anti-apoptotic Bcl-2-like genes NR-13, Mcl-1, sis of mucosal immune responses reveals signatures of
Bcl-X1, and Bcl-X2 in Atlantic cod (Gadus morhua). Mol intestinal barrier disruption and pathogen entry follow-
Immunol 47:763−784 ing Edwardsiella ictaluri infection in channel catfish,
➤ Fjellheim AJ, Playfoot KJ, Skjermo J, Vadstein O (2012) Ictalurus punctatus. Fish Shellfish Immunol 32:816−827
Inter-individual variation in the dominant intestinal ➤ Marking LL, Rach JJ, Schreier TM (1994) Evaluation of anti-
microbiota of reared Atlantic cod (Gadus morhua L.) lar- fungal agents for fish culture. Prog Fish-Cult 56:225−231
vae. Aquacult Res 43:1499−1508 ➤ Marques A, François JM, Dhont J, Bossier P, Sorgeloos P
➤ Forberg T, Arukwe A, Vadstein O (2011) A protocol and cul- (2004) Influence of yeast quality on performance of gno-
tivation system for gnotobiotic Atlantic cod larvae tobiotically grown Artemia. J Exp Mar Biol Ecol 310:
(Gadus morhua L.) as a tool to study host microbe inter- 247−264
actions. Aquaculture 315:222−227 ➤ Marques A, Dinh T, Ioakeimidis C, Huys G and others (2005)
➤ Forward RB, Rittschof D (1999) Brine shrimp larval photore- Effects of bacteria on Artemia franciscana cultured in dif-
sponses involved in diel vertical migration: Activation by ferent gnotobiotic environments. Appl Environ Microbiol
fish mucus and modified amino sugars. Limnol Oceanogr 71:4307−4317
44:1904−1916 ➤ Marques A, Ollevier F, Verstraete W, Sorgeloos P, Bossier P
➤ Fujimura K, Okada N (2007) Development of the embryo, (2006) Gnotobiotically grown aquatic animals: opportu-
larva and early juvenile of Nile tilapia Oreochromis nities to investigate host-microbe interactions. J Appl
niloticus (Pisces: Cichlidae) developmental staging sys- Microbiol 100:903−918
tem. Dev Growth Differ 49:301−324 ➤ McDermott JR, Leslie FC, D’Amato M, Thompson DG,
➤ Gatesoupe FJ (1999) The use of probiotics in aquaculture. Grencis RK, McLaughlin JT (2006) Immune control of
Aquaculture 180:147−165 food intake: enteroendocrine cells are regulated by
➤ Gomez-Gil B, Herrera-Vega MA, Abreu-Grobois FA, Roque CD4+ T lymphocytes during small intestinal inflamma-
A (1998) Bioencapsulation of two different Vibrio species tion. Gut 55:492−497
 Situmorang et al.: Challenge test in gnotobiotic Nile tilapia larvae 33

➤ Mereghetti L, Sitkiewicz I, Green NM, Musser JM (2008) hydrogen peroxide, and sodium chloride in fungal-
Remodeling of the Streptococcus agalactiae transcrip- infected rainbow trout eggs. Aquaculture 140:323−331
tome in response to growth temperature. PLoS ONE 3: Sihag RC, Sharma P (2012) Probiotics: the new ecofriendly
e2785 alternative measures of disease control for sustainable
➤ Mitchell AJ, Radomski AA, Straus DL, Carter R (2009) The aquaculture. J Fish Aquat Sci 7:72−103
effect of hydrogen peroxide on the hatch rate of channel ➤ Sikorski J, Nevo E (2005) Adaptation and incipient sym-
catfish eggs and on Saprolegnia spp. infesting channel patric speciation of Bacillus simplex under microclimatic
catfish eggs. N Am J Aquaculture 71:276−280 contrast at ‘Evolution Canyons’ I and II, Israel. Proc Natl
➤ Munro PD, Barbour A, Birkbeck TH (1995) Comparison of Acad Sci USA 102:15924−15929
the growth and survival of larval turbot in the absence of ➤ Small BC, Wolters WR (2003) Hydrogen peroxide treatment
culturable bacteria with those in the presence of Vibrio during egg incubation improves channel catfish hatch-
anguillarum, Vibrio alginolyticus, or a marine Aero- ing success. N Am J Aquaculture 65:314−317
monas sp. Appl Environ Microbiol 61:4425−4428 Sokal R, Rohlf F (1955) The principles and practice of statis-
➤ Nagai T, Iwamoto E, Sakai T, Arima T and others (2008) tics in biological research. Freeman & Company, New
Characterization of Edwardsiella ictaluri isolated from York, NY
wild ayu Plecoglossus altivelis in Japan. Fish Pathol 43: ➤ Soto E, Maziel A, Riofrio A, Martinez A, Cabrejos ME (2012)
158−163 Edwardsiella ictaluri as the causative agent of mortality
➤ Pham LN, Kanther M, Semova I, Rawls JF (2008) Methods in cultured Nile tilapia. J Aquat Anim Health 24:81−90
for generating and colonizing gnotobiotic zebrafish. Nat ➤ Soto E, Illanes O, Revan F, Griffin M, Riofrio A (2013) Bacte-
Protoc 3:1862−1875 rial distribution and tissue targets following experimen-
Plumb JA, Hanson LA (2010) Tilapia bacterial diseases. In: tal Edwardsiella ictaluri infection in Nile tilapia Oreo-
Health maintenance and principal microbial diseases of chromis niloticus. Dis Aquat Org 104:105−112
cultured fishes. Wiley-Blackwell, Oxford, p 445−463 ➤ Spor A, Koren O, Ley R (2011) Unravelling the effects of the
➤ Plumb JA, Sanchez DJ (1983) Susceptibility of five species of environment and host genotype on the gut microbiome.
fish to Edwardsiella ictaluri. J Fish Dis 6:261−266 Nat Rev Microbiol 9:279−290
➤ Rach JJ, Gaikowski MP, Howe GE, Schreier TM (1998) Eval- StatSoft (2004) Electronic statistics textbook. StatSoft, Tulsa,
uation of the toxicity and efficacy of hydrogen peroxide OK. www.statsoft.com/textbook/
treatments on eggs of warm- and coolwater fishes. Aqua- ➤ Stock I, Wiedemann B (2001) Natural antibiotic susceptibili-
culture 165:11−25 ties of Edwardsiella tarda, E. ictaluri, and E. hoshinae.
Rana KJ (1988) Reproductive biology and hatchery rearing Antimicrob Agents Chemother 45:2245−2255
of tilapia eggs and fry. In: Muir JF, Roberts RJ (eds) ➤ Thavasi R, Aparnadevi K, Jayalakshmi S, Balasubramanian
Recent advances in aquaculture. Croom Helm, London T (2007) Plasmid mediated antibiotic resistance in mar-
and Sydney and Timber press, Portland, OR, p 343−406 ine bacteria. J Environ Biol 28:617−621
Rana KJ, Beveridge MCM (1989) Tilapias and their cultures. ➤ Tinh NTN, Direckens K, Sorgeloos P, Bossier P (2008) A
In: Coche A, Edwards D (eds) Selected aspects of review of the functionality of probiotics in the larvicul-
warmwater fish culture. FAO, Rome, p 181 ture food chain. Mar Biotechnol 10:1−12
Rana KJ, Macintosh DJ (1988) A comparison of the quality USEPA (US Environmental Protection Agency) (2002) Meth-
of hatchery-reared Oreochromis niloticus and O. mos- ods for measuring the acute toxicity of effluents and
sambicus fry. In: Pullin RSV, Bhukaswan T, Tonguthai K, receiving waters to freshwater and marine organisms,
Maclean JL (eds) Tilapia in aquaculture. Proc 2nd 5th edn. EPA-821-R-02-012, USEPA, Office of Water,
Int Symp Tilapia in Aquaculture. ICLARM, Manila, Washington, DC
p 497−502 ➤ Verner-Jeffreys DW, Shields RJ, Birkbeck TH (2003) Bacter-
➤ Reger PJ, Mockler DF, Miller MA (1993) Comparison of anti- ial influences on Atlantic halibut Hippoglossus hippo-
microbial susceptibility, β-lactamase production, plasmid glossus yolk sac larval survival and start feed response.
analysis and serum bactericidal activity in Edwardsiella Dis Aquat Org 56:105−113
tarda, E. ictaluri and E. hoshinae. J Med Microbiol 39: ➤ Verschuere L, Rombaut G, Huys G, Dhont J, Sorgeloos P,
273−281 Verstraete W (1999) Microbial control of the culture of
Rodkhum C, Kayansamruaj P, Pirarat N (2011) Effect of Artemia juveniles through preemptive colonization by
water temperature on susceptibility to Streptococcus selected bacterial strains. Appl Environ Microbiol 65:
agalactiae serotype Ia infection in Nile tilapia (Ore- 2527−2533
ochromis niloticus). Thai J Vet Med 41:309−314 ➤ Verschuere L, Rombaut G, Sorgeloos P, Verstraete W (2000)
Russo R (2011) An attenuated strain of Edwardsiella ictaluri Probiotic bacteria as biological control agents in aqua-
is killed by channel catfish (Ictalurus punctatus) macro- culture. Microbiol Mol Biol Rev 64:655−671
phages and confers protection in few days. Dyn Biochem ➤ Waltman WD, Shotts EB, Hsu TC (1986a) Biochemical
Process Biotech Mol Biol 5:76−82 characteristics of Edwardsiella ictaluri. Appl Environ
➤ Russo R, Panangala VS, Wood RR, Klesius PH (2009) Chem- Microbiol 51:101−104
ical and electroporated transformation of Edwardsiella ➤ Waltman WD, Shotts EB, Hsu TC (1986b) Biochemical and
ictaluri using three different plasmids. FEMS Microbiol enzymatic characterization of Edwardsiella tarda from
Lett 298:105−110 the United States and Taiwan. Fish Pathol 21:1−8
➤ Salvesen I, Øie G, Vadstein O (1997) Surface disinfection of Welker TL, Lim CE (2011) Use of probiotics in diets of
Atlantic halibut and turbot eggs with glutaraldehyde: tilapia. J Aquac Res Dev S1:014
Evaluation of concentrations and contact times. Aquacult Yousefian M, Amiri MS (2009) A review of the use of pre-
Int 5:249−258 biotic in aquaculture for fish and shrimp. Afr J Biotechnol
➤ Schreier TM, Rach JJ, Howe GE (1996) Efficacy of formalin, 8:7313−7318

Editorial responsibility: Stewart Johnson, Submitted: June 10, 2013; Accepted: January 14, 2014
Nanaimo, British Columbia, Canada Proofs received from author(s): March 30, 2014

View publication stats