Serum neutralization test for epidemiological studies
of salmonid rhabdoviroses in France
AM
M Danton
P de Kinkelin
Hattenberger-Baudouy
G Merle
1
2
CNEVA, laboratoire central de recherches vétérinaires, 94703 Maisons-Alfort cedex;
INRA, laboratoire de virologie et immunologie moléculaires, 78352 Jouy-en-Josas cedex, France
Summary ― Serological examination is not yet accepted as being a suitable diagnostic method for fish
are asymptomatic virus carriers. Nevertheless, encouraging preliminary results using an endpoint serum neutralization test (SNT) in several French trout farm populations have demonstrated an
excellent correlation between the SNT and the previously established virus histories of the tested
populations. Following the isolation of infectious haematopoietic necrosis virus (IHNV) in France, serological screening of fish for a neutralizing antibody (NAb) to IHN was conducted on a national scale. This
survey confirmed the relationship between the serum-neutralizing immune response of the fish and the
presence of IHNV in a given trout farm population. Insofar as many trout populations underwent dual
rhabdovirus infections with both IHNV and viral haemorrhagic septicemia virus (VHSV), NAbs to both
viruses were also detected in the fish from such populations, often in distinct individuals. NAb-responding fish became detectable 2-3 months post-infection (pi). The number of responding fish reached a
mean prevalence of 20% between 3 and 6 months pi and disappeared after 8 months. The neutralizing serum titres (NST) were considered positive at > 32 and 64 for VHSV and IHNV, respectively. Both
the NST results and the prevalence varied greatly according to individuals, populations and the number of repeated stimuli involved in a given serum sampling series. Conversely, the thousands of sera
collected from trout in virus-free farms did not display any neutralizing activity against either VHSV or
that
IHNV. The SNT thus seems to constitute a reliable tool for the assessment of the health status of
trout farm populations for rhabdoviruses, and would be useful in the implementation of fish health
surveillance programmes.
rhabdovirose / salmonid / serology / serum neutralization / epidemiology
Résumé ― Épreuve de séroneutralisation pour l’étude épidémiologique des rhabdoviroses
chez les salmonidés. La sérologie n’est pas encore une méthode diagnostique reconnue en ichtyopathologie. Cependant, nous avions enregistré, dans le passé, des résultats très encourageants dans
le sérodiagnostic des rhabdoviroses par une microtechnique de séroneutralisation (SN). C’est pourquoi,
après que la nécrose hématopoïétique infectieuse (NHI) eut été identifiée en France, une enquête
épidémiologique fondée sur la SN a été conduite au niveau national. Elle a confirmé une bonne corrélation entre la détection des anticorps neutralisants (AcN) et la présence des virus de la NHI dans les
truites d’une population donnée. Comme de nombreuses populations de truites connaissaient une
*
Correspondence
and
reprints
double infection rhabdovirale par les virus de la NHI et de la septicémie hémorragique virale, l’épreuve
de SN a révélé des AcN spécifiques de chacun des virus. Inversement, aucune réponse positive n’a
été obtenue à partir des milliers de sérums prélevés chez des truites originaires de piscicultures reconnues indemnes de rhabdoviroses pendant au moins les 4 années précédant l’enquête. La SN apparaît donc comme une technique de choix dans la mise en service des programmes de surveillance sanitaire des piscicultures vis-à-vis des rhabdoviroses, maladies réputées contagieuses de la législation
vétérinaire française.
rhabdo viroses / salmonidés /
sérologie l
séroneutralisation l
INTRODUCTION
Serological examination is widely used as
diagnostic method for both human and
animal infectious and parasitic diseases. It
is not yet used for fish disease diagnosis.
Antibodies (Abs) to various fish pathogens
and especially to salmonid rhabdoviruses
a
discovered over 20 years ago, under
both natural and experimental conditions
were
(Vestergaard-Jorgensen, 1971, 1974;
Amend and Smith, 1974; de Kinkelin et al,
1977a, b; Bernard etal, 1983; HattenbergerBaudouy ef al, 1989; Olesen et al, 1991).
The demonstration of trout complement
dependence of serum neutralization (SN)
of viral haemorrhagic septicaemia virus
(VHSV) (Dorson and Torchy, 1979) enabled
the routine use of an SN test (SNT). Nevertheless, the use of serological examination
as a diagnostic tool for determining the virus
status of trout populations was not included
in the guidelines for either the national and
international animal health authorities,
despite its recommendations in the International Animal Health Code of the Office
international des epizooties (1986). This
reluctance towards using serological diagnosis for certain fish diseases may be due in
part to the somewhat limited Ab repertoire
reported for fish (Du Pasquier, 1982). In
addition, the current dogma inherited from
the early fish pathologists is a belief that the
direct isolation of the virus is the best diagnostic method. There is also a real need for
more information about the immune
responses of fish to rhabdoviruses, based
on more field data.
épidémiologie
The preliminary results of an end-point
SNT used in field assays and experimental
conditions, demonstrated the utility of serological studies at the population level for the
detection of viral infections.
This work aims to demonstrate, using
thousands of SNT results from fish populations with previously known and unknown
virus histories, the correlation between the
results of virological and serological methods. The vast scope of the study also
allowed us to assess the virus status of trout
farm populations and thus provided a convenient tool for the descriptive epidemiology of trout rhabdoviroses.
MATERIALS AND METHODS
Cell lines
The fish cell line, epithelioma papulosum cyprini,
(EPC), derived from the common carp Cyprinus
carpio (Fijan et al, 1983) was used for the isolation, propagation and identification of the virus
isolates used in this study, as well as for the
serum-neutralization tests. The cell line, rainbow
trout gonad (ATCC CCL55) RTG-2 (Wolf and
Quimby, 1962), was used for certain virus isolation procedures. Both lines were propagated in
Stoker’s medium (Gibco Life Technologies Ltd,
Parsley, Scotland, UK), buffered to pH 7.45 with
0.16 M Tris-HCI (Sigma Chemical Company,
Saint Louis, MO, USA) and supplemented with
10% fetal bovine serum (FBS) (Gibco) and other
ingredients as previously described (Hattenberger-Baudouy etal, 1989). For the routine cell
propagation, the EPC and RTG-2 cells were incubated at 29 and
20°C, respectively.
Virus strains, virus propagation
and isolation
The virus strains used, and the techniques for
their propagation and isolation were identical to
those previously reported (Hattenberger et al,
1989). Briefly, the French isolate of VHSV, INRA
07.71 belonging to serotype 1 (Vestergaard-Jorgensen, 1972) and 2 IHNV isolates were used in
the SN tests. The IHNV isolates were the North
American virus isolate 1266 from rainbow trout
donated by D Amend (Seattle, WA, USA) in 1970
and the French isolate 32-87 (de Kinkelin et al,
1987).
All 3 virus isolates were propagated in EPC
cells incubated at 15°C in Stoker’s medium with
2% FBS. The supernatants from the infected cells
were centrifuged, filtered, aliquoted and frozen
at -70°C to await use as SNT antigens.
Virus isolation and identification from either
diseased fish or asymptomatic virus carriers were
performed according to the previously described
methods (de Kinkelin et al. 1985).
Serum neutralization tests
The SNT
were performed according to an endpoint technique conducted in cell culture
microplates. The technique was derived from one
previously described (Hattenberger-Baudouy et al,
1989) with modifications consisting of the use of
flat-bottomed microplate wells, immunized trout
sera as positive controls for the SN to IHNV
instead of rabbit serum, and the adjustment of
the virus titres of neutralization virus antigen in
order to generate a clearcut cytopathic effect in
the serum-negative control cells within 72-96 h.
These virus titres were usually lower than those
previously observed, ie 5 000 pfu and 30 000 pfu
per 25 pl.
Fish, trout farms, sampling procedures
and objectives
Blood samples were collected from more than
14 000 fish. The majority of the fish were rainbow trout (Oncorhynchus mykiss) but 5% of the
samples were from brown trout (Salmo trutta)
and less than 1 % were from brook trout (Salvelinus fontinalis). The number of serum samples
varied from 18 to more than 180 per population.
At 196 of the sites, 60 samples were taken. In
most cases the blood samples were taken from
fish more than one year old and were generally
(60% of blood sample) from broodfish. When the
dual NAb screening for VHSV and IHNV is taken
into consideration, the final number of SNT performed reached slightly more than 16 500.
The materials sampled for virological examination were taken from the kidney, spleen and
brain or were sexual fluids. The organs from 5
fish were pooled before processing, the brains
being sometimes kept apart from the kidneys
and spleens. The ovarian fluids and milt were
always individually tested for the virus. This
resulted in more than 2 050 virological examinations.
The sampling campaign included trout populations from 229 farming sites located in all the
regions of France involved in salmonid fish culture. Prior to the beginning of the study, the virus
history of about 2/3 of the farming sites visited
had been previously established (within the last 4
years) on the basis of virological examinations
made either at the farmer’s request during periods
of disease outbreaks or as part of an official fish
health surveillance programme. This, therefore,
allowed comparisons between the data resulting
from the serological examination and the information provided by internationally agreed methods which utilized virus isolation as their central
criterion. Once the sampling for the national survey was completed, particular aspects of the serological response of trout populations to rhabdoviruses were reinvestigated. This was conducted
at several farming sites (table I) in order to gather
data that supported the use of the SNT as a diagnostic method for the implementation of fish health
surveillance programmes. At sites Ne and Lu,
the course of the Nab response in trout fingerlings to VHS following an outbreak of overt disease, was studied and at the site Lu again, the
SNT was used 3 years later for casual diagnosis of VHS.
Finally, at site Ro in 1993, simultaneous
screening for rhabdoviruses was conducted in
brood fish using both virological examination and
the SNT. This last study was aimed at demonstrating the efficacy of using SNT alone as a diagnostic method for evaluating the virus status of
a trout population.
RESULTS
Serological survey for IHN and
the detection of dual NAb responses
to rhabdoviruses
The screening of rainbow trout populations
from 229 farming sites for NAbs to the IHNV
provided positive results with groups of serum
batches from 35 of the sites (table II). Eighteen of these sites had between 5 and 20%
of the responding fish providing titres ranging
from 64 to512. In 6 of the responding trout
populations tested, IHNV was isolated prior to
the serological samplings. Conversely, due to
the constraints of French legislation, only 133
out of the 29 were checked for the virus during the months following the positive serological examination. Nevertheless, IHNV was
recovered from fish at all the sites, either during overt IHN outbreaks or during the screening of sexual fluids from broodfish. Ten of
the sites with IHN-responding fish were
checked for NAbs to VHSV. Five of these
sites were also found positive for this virus,
sometimes with a higher prevalence than that
found for IHNV. Indeed the example shown in
table III indicates prevalences of Nab
responses at 27.5 ± 10 and 8.5 ± 6% for VHS
and IHN, respectively.
In contrast, none of the 6 570 sera collected in trout populations from the 103 trout
farms operating in rhabdovirus-free zones
displayed neutralizing activity against IHNV.
Similarly, the virus screening of the ovarian
fluids from 1 721 rainbow and brown trout
broodfish from this zone was negative. Sampling for VHSV performed with 10 sera from
each of 60 trout farm populations was also
negative.
Response of naive rainbow trout
fingerlings to VHSV
In 1989, because of the restocking of VHSV
carriers upstream, overt disease occurred
in a formerly virus-free trout farm (Ne) that
had been supplying our laboratory with ’normal’ trout serum as the source of trout complement since 1975. VHSV was isolated
from the diseased trout about 8 d after the
onset of mortality. The periodic SNT tests
of 60 fish were negative both 1 month prior
to the virological diagnosis and 3 weeks
afterwards. Seroconversion was demonstrated at month 4 in 29/60 fish with titres
of 32-128, whereas the virological diagnosis was negative and dropped to 1 at
month 7.
The rainbow trout fingerlings response
also studied in 1989 at another trout
farm that had a long history of VHS (Lu).
The fish became infected after they were
transferred from the virus-free indoor facilities to the outdoor ponds. Virological diagnosis performed under the same conditions
as those indicated above was initially positive and became negative at month 4 after
the onset of clinical signs, whereas NAbs
were detectable in 12/60 fish within titres
ranging from 32 to512. At month 8, only 1
fish serum reacted to VHSV with a titre of 64
was
(not shown).
Casual use of serological diagnosis
of VHS at a trout farm
During a visit to the above farm, Lu, in 1992,
the fish farmer randomly netted 12 ninemonth-old rainbow trout that had survived
a VHS attack 5 months earlier (lot 1) and
similarly captured 30 six-month-old trout
from another lot (2) in which some individuals had undergone overt VHS (table IV).
No virus was recovered from the fish in lot 1
but half of them responded serologically,
whereas in lot 2, VHSV was easily isolated
from the infected organ homogenates and
low titres of NAbs were detected in only 2
sera. Both virological and serological examinations were IHN negative.
Simultaneous use of virus isolation
and SNT in the diagnosis
of asymptomatic rhabdovirus carriers
A virus assay of the ovarian fluids and a
SNT were performed on 60 rainbow trout
broodfish at trout farm Ro, a site with a wellestablished history of rhabdoviroses. The
results revealed that 14 fish harboured
VHSV. Ten fish displayed high NAb titres
to VHSV. The fish 11 displayed a positive
response to both the serological and the
virological tests. Two fish, 47 and 48,
reacted serologically with IHNV with titre of
64 and >_ 512 respectively (table V). All
samples were negative for IHNV isolation.
DISCUSSION
The
serological screening of trout using the
end-point neutralization technique, which
was
the basis for the national survey for IHN
conducted in France in 1989-1990, confirmed the correlation between the NAb
response and the presence of IHNV in a
given trout farm stock as previously reported
(Hattenberger-Baudouy et al, 1989, 1995).
Insofar as several trout populations underwent dual rhabdoviroses (VHS and IHN),
NAbs to both viruses were also detected in
the fish from such populations. In contrast,
no neutralizing titres higher than 32 and 16
6
for IHNV and VHSV, respectively, were
demonstrated in sera collected from the
trout populations of virus-free origin. This
allowed us to consider the responses of
NAb titres 64 for IHNV and 32 for VHSV as
positive. Thus, each time the SNT and virus
assay in cell culture were used either simultaneously or separately with material from a
given trout population, the virological and
serological results correlated well, establishing the health status of the given population.
The scale of our field investigations constitutes one of the largest epidemiological
studies of rhabdoviroses ever undertaken.
These studies were made necessary by the
implementation of fish surveillance programmes at both the national and international levels. The data collected were both
informative and predictive. Indeed, in several
cases, the NAb response of one trout population, especially in the broodfish, preceded
the onset of overt rhabdovirus infection
some months later. An example of this
occurred at site Ro where the response of
asymptomatic fish to IHNV was followed by
several episodes of clinical infection among
the young trout that were the progeny of the
broodfish that had responded positively.
The results of the SNT used during the
of field VHSV infections in naive rainbow trout fingerlings confirmed the acquired
aspect of the NAb response to VHSV pre-
course
viously found during the national survey. At
temperatures of 9-11°C, the NAbs
that were detected 4 months post-infection
at sites Ne and Lu 89 likely appeared after
6-8 weeks as had been found at site Lu in
1992. These results are in agreement with
those previously reported under experimental and natural conditions (Bernard et
al, 1983; Olesen et al, 1991The data in
table IV show that the virus assay method
encounters limitations when used for the
water
causal diagnosis of VHS in an asymptomatic
fish population. Conversely, limitations to
STN appear if it is used too early in the
course of a natural infection. Nevertheless,
at the population level on trout farms there
was a good correlation between the virological and serological results which demonstrates the utility of the SNT as a screening
method for VHSV in surveillance programs.
The SNT method we used in this study was
specific but less sensitive than the SN
microplaque technique (data not shown)
described by Olesen and Vestergaard-Jorgensen (1986). The 2 tests provide similar
neutralization patterns although the titres
and number of positive are lower with our
end-point SNT.
However, the ability of the end-point SNT
to examine large numbers of samples
makes it highly informative at the population level. In addition, we have shown that
this test provides results that correlate well
with those generated by the virus isolation
technique (table V) particularly when the
sample sizes are sufficiently large. According to existing sampling charts, 60 is a suitable sample size (Ossiander and Wedemeyer, 1973). We found that the neutralizing
titres and the prevalence of positive
responses varied greatly with individuals,
the frequency of virus exposure and the timing of a given fish sample. The risk of obtaining false negative responses could be
reduced by selecting a specific class of fish
for the sampling from a population rather
than by testing randomly. Older fish, particularly broodfish, are the most suitable targets. If this age group does not exist at a
given farming site, an alternative is to sample those fish that have been transferred to
the outdoor ponds 3-6 months prior to the
serological sampling date. When comparing the serological versus virus isolation
approaches it is apparent that the SNT is
more effective with asymptomatic virus carriers unless the virus can be recovered from
the sexual products at the time of spawning. The virological examinations of (nonspawning) carriers are laborious, costly and
less reliable than the SNT to assess the
virus status of such populations.
Furthermore, the somewhat lower
sen-
of the
sitivity
end-point SNT diminishes the
risk of finding false positive responses compared with the plaque neutralization
approach.
Attempts
perhaps
at
more
substituting a quicker and
up-to-date serological diag-
nostic method for rhabdoviruses, such as
the ELISA test, have been made (Vester-
gaard-Jorgensen et al, 1991These tests,
however, lack antivirus specificity. This supports the continued
of the SNT
method.
as a
as an
indi-
use
serological diagnostic
SNT is gaining acceptance
cator of exposure to IHNV in the USA
(La-
Patra et al, 1993, 1994), but so far no offi-icial agreement has been made for its use
as a fish health assessment method for
rhabdoviruses, despite the utility of this
technique for examining the fish during periods when virus isolation would not be productive.
On the basis of
our
present work, SNT,
serological diagnostic method, should
be considered as an approach that is of
equal value and also complementary to the
as a
virus isolation method for
lance programmes.
use
in virus surveil-
ACKNOWLEDGMENTS
The collaboration and help of the following persons, for the collection and processing of the
samples is warmly acknowledged: J Castric and
staff CNEVA/Brest, J Catel, M Morand, P
Nougayrbde, A Vigouroux and the respective
staffs of the veterinary laboratories of Pas-deCalais, Jura, Landes and Finistere.I Clement
is also acknowledged for her excellent typing
work.
REFERENCES
Amend DF, Smith L (1974) Pathophysiology of infectious hematopoietic necrosis virus disease in rainbow trout (Salmo gairdnen): early changes in blood
and aspects of the immune response after injection
of IHN virus. J Fish Res Board Can 31, 1371-1378
Bernard J, de Kinkelin P, Bearzotti M (1983) Viral hemorrhagic septicemia of trout: relation between the G
polypeptide, antibodies production and protection of
the fish following infection with the F25 attenuated
4
variant strain. Infect and Immun 39, 7-14
Dorson M, Torchy C (1979) Complement dependent
neutralization of Egtved virus by trout antibodies.
J Fish Dis 2, 345-347
Du
Pasquier L (1982) Antibody diversity in lower vertebrates - Why is it so restricted? Nature (Lond) 296,
3111
Fijan N, Sulimanovic D, Bearzotti M et al (1983) Some
properties of the epithelioma papulosum cyprini
(EPC) cell line from carp (Cyprinus carpio). Ann Virol
(Institut Pasteur) 134 E, 207-220
Hattenberger-Baudouy AM, Danton M, Merle G, Torchy
C, de Kinkelin P (1989) Serological evidence of infectious hematopoietic necrosis in rainbow trout from
a French outbreak of disease. J Aquat Anim Health
1, 126-134
Hattenberger-Baudouy AM, Danton M, Merle G, de
Kinkelin P (1995) Epidemiologie de la necrose
hbmatopoibtique infectieuse (NHI) des salmonides
France : suivi de l’infection naturelle par des techet s6rologiques et tentatives
d’eradication. Vet Res 26, 256-275
en
niques virologiques
de Kinkelin P, Baudouy AM, Le Berre M (1977a) Reaction
de la truite Fario (salmo trutta) et arc-en-ciel (Salmo
gairdnen)a I’infection par un nouveau rhabdovirus.
CR Acad Sc Paris S6rie D, 284, 401-404
de Kinkelin P, Gerard JP, Dorson M, Le Berre M (1977b)
Viral haemorrhagic septicaemia: demonstration of
a protective immune response following natural infection. Fish Health News 6, 3-4
de Kinkelin P, Hattenberger AM, Torchy C, Lieffrig F
(1987) Infectious haematopoietic necrosis (IHN): first
report in Europe. European Association of Fish
Pathologists, Third international conference (Abstract
57)
de Kinkelin P, Michel C, Ghittino P (eds) (1985) Precis de
pathologie des poissons. INRA, Office international
des epizooties. Paris, 348 p
LaPatra SE, Turner T, Lauda KA, Walker SC, Jones GR
(1993) Characterization of the humoral response of
rainbow trout to infectious hematopoietic necrosis
virus. J Aquat Anim Health 5, 165-171
LaPatra SE, Lauda KA, Jones GR, Shewmaker WD,
Walker SC (1994) Development of passive
immunotherapy for control of infectious hematopoietic necrosis. Dis Aquat Org 20, 1-6
Office international des 6pizooties (1986) International
Animal Health Code, Recommended norms for the
pathogens of fish. Vth edition, 364, 1-23
Olesen NJ, Vestergaard-Jorgensen PE (1986) Detection of neutralizing antibodies to Egtved virus in rainbow trout (Salmo gairdnen) by plaque neutralization
test with complement additions. J/tpp/ /enrhyo<.
2, 3341
Olesen NJ, Lorenzen N, Vestergaard-Jorgensen PE
(1991) Detection of rainbow trout antibody to Egtved
virus by enzyme-linked immunosorbent assay
(ELISA), immunofluorescence (IF) and plaque neutralization tests (50% PNT). Dis Aquat Org 10, 3138
Ossiander JF, Wedemeyer G (1973) Computer program
for sampling sizes required to determine disease
incidence in fish populations. J Fish Res Board Can
30, 1383-1384
Vestergaard-Jorgensen PE (1971) Egtved virus: demonstration of neutralizing antibodies in serum from artificially infected rainbow trout. J Fish Res Board Can
28, 875-877
Vestergaard-Jorgensen PE (1972) Egtved virus: antigenic variations in 76 virus isolates examined on
neutralization tests and by the means of the fluorescent antibody technique. In: Symposia of the Zoological Society of London (LE Mawdeslex-Thomas,
ed), Academic Press, UK, 30, 333-339
Vestergaard-Jorgensen PE (1974) Indirect fluorescent
antibody techniques for demonstration of trout viruses
and corresponding antibodies. Acta Vet Scand 15,
198-205
Vestergaard-Jorgensen PE, Olesen NJ, Lorenzen N,
Winton JR, Ristow SS (1991) Infectious hematopoietic necrosis (IHN) and viral hemorrhagic septicemia
(VHS): detection of trout antibodies to the causative
viruses by the means of plaque neutralization,
immunofluorescence
and
enzyme-linked
immunosorbent assay. J Aquat Anim Health 3, 100108
Wolf K, Quimby MC (1962) Established eurythermic line
of fish cell in vitro. Science 135, 1065-1066