Download Rapid Detection of Infectious Pancreatic Necrosis Virus (IPNV) by

J.
321
gen. ViroL (1983), 64, 321-330• Printed in Great Britain
Key words: ELISA/IPNV/rapid detection~fish viruses
Rapid Detection of Infectious Pancreatic Necrosis Virus (IPNV) by the
Enzyme-linked lmmunosorbent Assay (ELISA)
By P. F. D I X O N * AND B. J. H I L L
Ministry of Agriculture, Fisheries and Food, Directorate of Fisheries Research, Fish Diseases
Laboratory, Weymouth, Dorset DT4 8UB, U.K.
(Accepted 24 September 1982)
SUMMARY
The enzyme-linked immunosorbent assay (ELISA) was used to demonstrate the
presence of infectious pancreatic necrosis virus (IPNV) antigen in cell cultures and in
fish. Virus antigen could be detected in infected cell cultures before visible cytopathic
effect (c.p.e.) was evident and cell cultures showing complete viral c.p.e, produced
intense colour reactions. Virus antigen was detected in infected fry during and
immediately after an epizootic, but although IPNV-carrier fish could be detected by
the ELISA technique, the sensitivity of detection was not as great as that of isolation of
the infectious virus in cell culture. The major IPNV serotypes, Sp, Ab and Vr, crossreacted at only a low level and it was shown that the ELISA technique could be used to
serotype IPNV strains rapidly. None of 10 other fish pathogenic viruses reacted with
plates sensitized for IPNV detection. The time taken to perform the technique was
reduced to 1 h 35 rain at room temperature and this still allowed the results to be readily
assessed visually as antigen-positive or antigen-negative.
INTRODUCTION
Infectious pancreatic necrosis (IPN) is an acute, highly contagious virus disease of young
salmonid fish reared under intensive farming conditions. The virus belongs to a group of bisegmented double-stranded RNA viruses which include the virus of infectious bursal disease of
chickens and virus X of Drosophila, for which the name birnavirus has been proposed (Dobos et
al., 1979). Epizootics of IPN may persist for several weeks if left unchecked and the resulting
cumulative mortalities may be very high. Survivors of an outbreak become persistent
asymptomatic carriers in which the levels of infectious virus vary widely throughout their life.
1PN can have serious economic consequences for commercial trout farms and in Great Britain
the disease is notifiable.
In suspected cases of IPN disease it is most important that diagnosis be carried out quickly
and accurately so that appropriate control measures may be introduced with the minimum of
delay. Although clinical signs and histopathological features of IPN may permit a presumptive
diagnosis, confirmation by the demonstration of the specific virus in the diseased tissues is
invariably required. By far the most widely used method for this is the isolation of the virus in
fish cell cultures followed by neutralization with specific antiserum, a procedure which requires
several days to complete. More rapid serological tests have been used for the identification of
IPNV isolated in cell culture: the fluorescent antibody test (FAT) (Piper et al., 1'973;
Jorgensen, 1974), immunoperoxidase staining (Nicholson & Henchal, 1978), haemagglu;tination (Cleator & Burney, 1980), and the complement-fixation test (Finlay & Hill, 1975). Even
greater speed of confirmatory diagnosis can be achieved by the direct serological demonstration
of viral antigen in the tissues of diseased fish, but reports of this approach are limited to the",use
of FAT (Swanson & Gillespie, 1981). However, for whatever reasons, such rapid ldentlficatl0n
methods have not been adopted for routine use in the majority of fish disease laboratorieS.
As an alternative to FAT and immunoperoxidase methods, the enzyme-linked lmmunosorbent assay (ELISA) technique has recently gained widespread acceptance as a rapid and
•
0022-1317/83/0000-5295
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 13:07:53
.
.
LI
,~, 6
322
P. E. DIXON AND B. J. HILL
sensitive m e a n s o f detecting and identifying a w i d e variety o f viruses in plants (Clark & A d a m s ,
1977), insects (Kelly et al., 1978) and m a m m a l s (see Voller et al., 1979), but to d a t e has not b e e n
r e p o r t e d for use in fish virology. W e report here an i n v e s t i g a t i o n into the speed a n d specificity o f
the E L I S A t e c h n i q u e for the identification o f I P N V isolated in cell cultures a n d the direct
detection o f I P N V antigen in fish tissues w i t h a v i e w to its a d o p t i o n for routine d i a g n o s t i c
purposes. O u r i n t e n t i o n was to e v e n t u a l l y d e v e l o p a test that could be assessed visually as virus
antigen-positive or -negative and our m e t h o d o l o g y is d i r e c t e d to t h a t end. A b s o r b a n c e values
(see M e t h o d s ) a b o v e a p p r o x i m a t e l y 0-2 i n d i c a t e a discernible colour reaction.
METHODS
Antigens. IPNV reference serotypes Sp, Ab and Vr 299 (ATCC) and Tellina virus (TV-1) (for virus origins, see
Underwood et aL, 1977) were grown and assayed for infectivity at 15 °C in bluegill fry (BF-2) cells as described by
Underwood et al. (1977) except that, to reveal plaques, cells were fixed with 10~ neutral buffered formalin (NBF)
for 30 min and then stained with 0.1 ~ aqueous methylene blue. In addition, IPNV serotype Sp was grown under
the same conditions in rainbow trout gonad (RTG-2) and chinook salmon embryo (CHSE-214) cells.
The origins, growth and assay of infectivity of the rhabdoviruses used in this study, viral haemorrhagic
septicaemia virus (VHSV), infectious haematopoietic necrosis virus (IHNV), spring viraemia of carp virus
(SVCV), pike fry disease virus (PFDV), American eel virus (EVA) and European eel virus (EVEX), were as
described previously (Hill et al., 1975, 1980). Eel rhabdovirus strain C30 (provided by Mme J. Castric, Laboratoire
Nationale de Pathologie des Animaux Aquatiques, Brest, France) was grown and assayed as described by Hill et
al. (1980).
The herpesvirus, channel catfish virus (CCV) (provided by Dr J. Plumb, Auburn University, U.S.A.), was
grown at 22 °C in brown bullhead (BB) cells and the herpesvirus, Oncorhynchus masou virus (OMV) (provided by
Professor T. Kimura, Hokkaido University, Japan), was grown at 15 °C in RTG-2 cells.
IPNV-infected rainbow trout (Salmo gairdneri) and brown trout (S. trutta) were either naturally or
experimentally infected.
Extraction of antigen from infected fish. Unless otherwise specified, whole fry, or combined liver, kidney and
spleen from larger fish, were homogenized in either an equal volume of maintenance medium or phosphatebuffered saline containing 0-05 ~ Tween 20 (PBST) and clarified at 1500 g for 30 rain. For some tests, fish extracts
were treated twice with Arklone-P (1,1,2-trichloro-1,1,2-trifluoroethane) after clarification as described by Dixon
& Hill (1983) to remove further host material. Extracts in maintenance medium were further diluted to 1:50 or
1 : 100 and added to BF-2 cell cultures to test for infectious virus. If no cytopathic effect (c.p.e.) was evident after 5
to 8 days, cultures were frozen at - 2 0 °C, thawed and passed at a 1:10 dilution; two such blind passages were
done. For the ELISA technique, any further dilution of the initial extract (in maintenance medium or PBST) was
done in PBST.
Purification o f l P N V . The virus was purified as described by Dixon & Hill (1983).
Production of antisera, ~,-globulin purification and enzyme conjugation. Antisera were produced against purified
preparations of IPNV serotypes Sp, Ab and Vr, using the procedure described for fish rhabdoviruses by Hill et al.
(1975). The ~-globulin was extracted and purified by two cycles of ammonium sulphate precipitation and washing
followed by extensive dialysis against PBS, and then diluted to 1 mg/ml. Alkaline phosphatase EC 3.1.3.1 (Type
VII-S, Sigma) was conjugated to the 3,-globulin as described by Clark & Adams (1977).
ELISA. The procedure was done basically as described by Clark & Adams (1977) using buffers prepared
according to Voller et al. (1979). The test was done in polystyrene microtitre plates (M 29 AR, Dynatech) or cuvette
racks (1414 × 9, Gilford Instruments) and the absorbance of the enzyme-substrate hydrolysis product was read at
405 nm in an SP-6 spectrophotometer (Pye Unicam) or an EIA reader (Gilford Instruments), the light path in both
cases being 1 cm. Figures for the absorbance at 405 nm (A4o5) quoted in the text are means of duplicate samples.
Initially, chequerboard titrations of serial dilutions of purified homologous antigen and enzyme-3,-globulin
conjugate (hereafter referred to solely as conjugate) were done on microtitre plates coated with different
concentrations of ~-globulin from which the following ELISA procedure was evolved. The solid phase was coated
with 5 pg/ml ~-globulin (200 ~tl/well)for 6 h at 37 °C, antigens (duplicate samples; 200 pl/well) were incubated for 1
h at 37 °C or 16 to 17 h at 4 °C, conjugate at a 1:600 dilution (200 ~tl/well)was incubated for t h at 37 °C, substrate
(p-nitrophenyl phosphate at 1 mg/ml in 10% diethanolamine, pH 9.8; 300 pl/well) was incubated for 1 h at room
temperature (20 to 23 °C) and the reaction was stopped by adding 50 ~tl 3 M-NaOH/well. Three 10 min washes in
PBST were done between each stage of the procedure. These conditions enabled 10 ~tg/ml of purified virus to be
detected photometrically.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 13:07:53
ELISA for detection of I P N V
323
Table 1. ELISA absorbance values of I P N V antigen in infected BF-2 cells harvested at 24-h
intervals post-infection
ELISA A405 of dilutions (log10)
Time postCells
infection
showing
(h)
c.p.e. (%)
Non-infected
0
cells
0
0
24
0
48
<10
72
50
Virus titre
(p.f.u./ml)
0
5.7
1-5
3.6
6.3
×
x
x
x
102
106
10v
108
-1
0.125
of cell harvest
~
-2
-3
-4
0.08
0.035
0.03
-5
0.02
0-115
0.43
1.43
1.775
0-065
0.115
0.68
1.12
0.015
0.01
0.02
0.015
c
0.025
0.03
0.105
0.29
0.02
0.03
0-03
0-04
RESULTS
Detection of virus antigen in infected cell cultures
An undiluted cell culture harvest of IPNV serotype Sp showing complete c.p.e, was tested by
the ELISA technique and a strong positive result (A,o5 > 2) was obtained, whereas a control cell
harvest produced a minimal A4o5 reading. An experiment was then done to determine whether
virus antigen could be detected in cell cultures showing less than complete c.p.e. IPNV serotype
Sp was adsorbed to cell cultures for 1 h, the adsorption volume was removed, the cells washed
with maintenance medium, then further maintenance medium was added. One flask (time 0)
was immediately stored at - 20 °C, and at 24 h intervals for 3 days further flasks were stored at
- 20 °C. The development of c.p.e., virus titre and the ELISA absorbances are shown in Table
1. IPNV antigen was detected in cultures before visible c.p.e, was evident and the dilutions of
the harvests showed that about 105 p.f.u./ml could easily be detected and about 104 p.f.u/ml was
approximately the limit of detection. However, it is recognized that the harvests would have
contained non-infectious IPNV antigens which would have reacted in the ELISA technique;
therefore, relating A4o5 values to virus titre is not strictly valid.
The ELISA technique also detected IPNV antigen in cell cultures that had been used for
isolating virus from infected fish when viral c.p.e, was not evident, when it was just evident, and
when toxicity from the fish extract masked the development of viral c.p.e.
All the experiments above were done using BF-2-grown virus, but similar strongly positive
results in the ELISA technique were obtained using IPNV grown in RTG-2 and CHSE-214
cells, both commonly used for IPNV isolation.
Cross-reactions with other viruses
The other representative members of IPNV serogroup 1 (serotypes Ab and Vr), TV-1
(representative of serogroup 2), VHSV, IHNV, SVCV, PFDV, EVEX, EVA, C30, CCV and
OMV, all cell culture-grown, were tested for cross-reactions in the ELISA technique with IPNV
serotype Sp y-globulin. There was a low-level cross-reaction with IPN serotypes Ab and Vr
reflecting virus neutralization data (Underwood et al., 1977), but no cross-reactions with any of
the other viruses.
To investigate further the ELISA cross-reactions of IPNV serotypes Sp, Ab and Vr, 7globulins from Ab and Vr antisera were conjugated with alkaline phosphatase and their
optimum dilutions in the ELISA technique determined using wells coated with anti-Ab or antiVr 7-globulins. The three virus serotypes were then adjusted to the same titre (although the total
virus antigenic mass may have varied) and the three conjugates were diluted to give homologous
A405 readings of between 1 and 2. The cross-reactions, together with the 50 % plaque-reduction
titres of the sera, are shown in Table 2. In both the ELISA technique and the neutralization test
there was a strong reaction with homologous antigen and only a low-level cross-reaction with
heterologous antigens.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 13:07:53
324
1,. F. DIXON AND B. J. HILL
Table 2. Cross-reactions of BF-2-grown I P N V serotypes Sp, Ab and Vr in the ELISA technique
and neutralization tests
Antiserum
A
f
Reciprocal of dilution giving
50~ plaque reduction
ELISA A4os
r
Antigen
Sp
Ab
Yr
BF-2 cells
~
Sp
1.852
0.215
0.181
0.042
Ab
0.177
1.798
0.181
0.040
Vr
0-165
0.198
1.458
0.009
g
Sp
2900000
57000
10500
Ab
10500
1260000
11000
Vr
6200
3500
400000
Table 3. ELISA absorbance values of purified I P N V diluted in P B S T or fry extract, with and
without subsequent Arklone-P treatment
ELISA A,*o5
&
(
Virus diluent
PBST
PBST
Fry extract
Fry extract
Arklone-P
treatment
+
+
rl0000
1.835
1.71
0.71
1.125
Virus concentration
~
1000
100
1-195
0.41
1-02
0.34
0-545
0.225
0.80
0.43
(ng/ml)
10
0.065
0.065
0.045
0.07
1
0.01
0-03
0.02
0.075
Background
0.03
0-025
0.12
0.06
Reduction o.1"background absorbance caused by fish extracts
Preliminary experiments using the standard test conditions showed that control (noninfected) fry homogenates or visceral extracts of fingerlings produced a high background A4os,
which would reduce the sensitivity of the test for detecting low levels of antigen and preclude
visual assessment of results. Control fry or visceral extracts were Arklone-P-treated to try to
reduce the background absorbance. Arklone-P treatment of fry extracts coupled with dilution
(>t 1:8) in PBST did help in reducing the non-specific A405, although Arklone-P treatment of
visceral extracts increased the A40s. However, dilution alone reduced the A4o 5 of the visceral
extracts.
To test the effect of Arklone-P treatment of fry extracts containing virus antigens, purified
virus was added to fry extracts such that the virus was diluted serially in tenfold steps from 10000
to 1 ng/ml in a 1 : 10 dilution of extract. Aliquots of each antigen dilution were Arklone-P-treated
and compared to non-treated aliquots (Table 3). Arklone-P treatment greatly improved the
confidence of detection of antigen, although the A40s readings of 10 000 and 1000 ng/ml antigen
in treated extracts were not as high as those in PBST. Curiously, the A4o5 values of non-treated
fry extracts containing 1 and 10 ng/ml virus were lower than the fry extract control, the reason
for which is unknown.
The type and concentration of wetting agent in the PBS used for washing the plates and
diluting the sample was varied to determine whether that would reduce the background to a
visual negative, whilst maintaining a high virus A4os. Fry were macerated in either Triton X-100
or Tween 20 at concentrations of 1.0, 0.5, 0-1 and 0-05 ~, clarified, and then further diluted in ~he
initial extraction buffer; purified virus was also diluted in PBS containing the different
concentrations of wetting agents. For each sample, all washings and dilutions of conjugate were
done in the same dilution and type of wetting agent. Tween 20 at a concentration of 0.05 ~o (the
previous standard concentration) gave one of the highest A405 readings for purified virus, but
also one of the highest A4o5 readings for fry extract (Fig. 1), although this was reduced to a visual
negative by dilution of the extract. Triton X-100 at a concentration of 1 "0~o gave a high A40s for
purified virus, and a low A405 for fry extract (Fig. 2). The other concentrations of Triton X-]~00
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 13:07:53
ELISA for detection of IPNV
0.3
(o)
I
I
I
I
I
2.0 -
(b)
325
I
I
I
I
I
I
B
1.5
0.2
0. I
0.5
i-6
32
64
Dilution of fry extract (reciprocal)
~i
10
1
0.1
Virus concentration (~tg/m[)
g
Fig. 1. ELISA absorbance of(a) fry extracts and (b) purified IPNV processed in PBS containing 0.05~
(O), 0.1% (A), 0-5~ (11) or 1-0~ (O) Tween 20.
0.2
I
!
I
I
I
2.0
(a)
(b)
1.5
"~0.1
-
-~
1.0
0.5
I
I
i
t
I
I
I
I
4
8
16
32
64
10
1
0.1
Dilution of fry extract (reciprocal)
Virus concentration (~tg/ml)
Fig. 2. ELISA absorbanceof(a)fryextractsand (b) purified lPNV processed in PBScontaining0.05~
(O), 0.1~ (A), 0.5~ ( n ) or 1.0~ (O) Triton X-100.
and Tween 20 were unsuitable as they reduced the A405 of purified virus. The two effective
concentrations of detergents were then c o m p a r e d for their efficacy when detecting antigen in
fish extracts (see below).
Detection of virus antigen in infected fish
Seventy samples of rainbow trout and brown trout that had been collected from commercial
fish farms were tested by the E L I S A technique. The samples (whole fry or visceral extracts) had
previously been screened for I P N V by cell culture techniques, and 36 were IPNV-positive, The
samples were diluted to 1 : 30 in PBS plus wetting agent such that there was either 0.05 ~ Tween
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 13:07:53
326
P.
F.
DIXON
AND
B.
J.
HILL
Table 4. Detection of lPNV antigen in infected fry by the ELISA technique
Fish
Mode of
group
infection
1
Experimental
2
Experimental
3
Experimental
4
Experimental
5
Experimental
6a Experimental
6b Experimental
7
Natural epizootic
8
Non-infected
Sample
tested
Viscera*
Viscera*
Viscera*
Viscera*
Viscera*
Viscera*
Viscerat
Whole fryt
Viscera
Mortality
(~)
81
11
72-6
46.3
11.6
63
63
90:~
4
A4o5of sample Virus isolated
diluted 1:10
in cell culture
2-0
+
0.89
+
2.76
+
1.97
+
0.10
+
1.67
+
2.49
+
2.98
+
0.06
-
* Apparently healthy fry sampled when mortalities in the experiment were declining (day 30 post-infection).
t Moribund fry sampled during peak mortality period.
:~Visual assessment.
20 or 1% Triton X-100 in the sample, and then tested. Nine of the samples in PBS containing
Tween 20 were positive by the ELISA technique, with a further one doubtful positive, but only
one of the samples diluted in PBS containing Triton X-100 was positive; the A4o5 of the IPNVpositive control was slightly reduced by Triton X-100. Thus, in this instance, Triton X-100
decreased the sensitivity of the ELISA technique, which contradicts the data given above.
However, Triton X-100 was previously tested with either fry extract or purified virus, whereas
here the antigen was in the fish extract. Thus, there may have been an interaction with the virus
antigen plus extract plus Triton X-100 which reduced the sensitivity of virus antigen detection.
Consequently, in subsequent experiments, the original concentration of 0-05~ Tween 20 was
used.
The nine positive and one doubtful positive samples given by the ELISA technique were also
IPNV-positive by cell culture isolation, and came from fry in an I P N epizootic; the viruscontaining samples not detected by the ELISA technique came from virus carrier fish. The
extracts for this experiment had been stored for about 1 month at - 20 °C before being tested by
the ELISA technique, and a reason for the non-detection of some virus-containing samples may
have been that there was some loss of virus antigenicity during storage. However, a more likely
explanation is that infected fry in an epizootic contain high levels of virus antigen, whereas virus
carriers contain only low levels, below the sensitivity of the ELISA technique.
To elucidate this, freshly prepared extracts of fry from an I P N epizootic, fry experimentally
infected with different strains of IPNV, and virus carriers were tested by the ELISA technique.
Virus antigen was detected with confidence by the ELISA technique in all fry except those
infected with an avirulent strain of virus (group 5, Table 4) where there was an equivocal result
photometrically, but negative visually. As even fry infected with a low virulence strain of virus
(group 2, Table 4) contained sufficient antigen to be detected visually by the ELISA technique, it
appears that if fry are infected with a strain of virus that causes mortalities the ELISA technique
can be used to detect virus antigen in them. A much higher A4o5 reading was obtained from
moribund fry from group 6 (Table 4) compared with apparently healthy but infected fry from the
same group. Nevertheless, the lower A4os reading was still a strong positive, indicating that if fry
cannot be sampled during the peak mortality period, testing surviving fry as soon as possible
should still allow diagnosis of the infection by the ELISA technique. Only one of five fish
sampled from the carrier population was IPNV-positive by the ELISA technique at a 1:2
dilution of viscera (A4o5 was 1-20; control viscera A405 values ~<-0.30), whereas virus was
isolated from all carrier fish in cell cultures. Thus, the sensitivity of this ELISA technique is not
sufficient to detect virus antigen in carrier fish with certainty.
Tests with formalin-fixed antigens
If the ELISA technique is to be used as a field test on fish farms, the virus used as a positive
control would have to be inactivated for safety reasons. IPNV inactivated using 1:200 formalin
for 7 days at 20 °C (Dixon & Hill, 1983) produced a high Aao5 reading and was thus shown to be
suitable for use as such a positive control.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 13:07:53
ELISA for detection of IPNV
327
Table 5. Comparison of microtitre plates and cuvette racks under identical test conditions for
their use in the ELISA technique
ELISA A4o5 at sample dilution in PBST (loglo) of
A
F
Solid p h a s e
Absorbancereader
.~EIA analyser
Cuvette r a c k
I~Spectrophotometer
Microtitre plate Spectrophotometer
- 1
> 3.0
> 3.0
2.84
- 2
c 3.0
c 3.0
1.60
- 3
0-67
0.605
0-36
PBST only
0.096
0-083
0.025
Infected fry were also fixed in 10~ NBF for 7 days, then eviscerated, and the viscera were
washed in PBS containing 0.04 ~o sodium bisulphite followed by homogenization. The A~o5 of a
1 : 10 dilution of viscera was 0.32 compared with 1-37 for similar material processed freshly and
0-08 for NBF-fixed non-infected fry viscera. Thus, although the ELISA technique was not so
sensitive for detecting IPNV antigen in NBF-fixed fish tissues, it may be of value when only
such material is available and virus isolation in cell culture cannot be done.
Use of different receptacles for the ELISA technique
The previous experiments had been done using 96-well microtitre plates, but for some uses of
the technique in this laboratory a smaller well number was more convenient and this was given
by cuvette racks, each comprising a strip of 10 micro-cuvettes. These racks were compared for
sensitivity with the microtitre wells under the same conditions, using the same volumes of
reagents. The A4o5 levels of the samples from microtitre wells were read on a spectrophotometer
and the A405 levels of samples from the cuvette racks were read on the same spectrophotometer
and on a reader designed for use with the cuvette racks to ensure any differences were not caused
by instrument variation. The A4o5 readings of virus samples tested in cuvette racks were higher
than those tested in microtitre wells and the PBST controls in cuvette racks were still visually
negative (Table 5). Consequently, all further tests were done in cuvette racks.
Variation between and within cuvette racks was tested for on samples diluted to give A4os
readings between 1 and 2. Variations of virus A405 readings within racks ranged from 0.116 to
0.275 absorbance units and the greatest difference between racks was 0-315 absorbance units
(rack means were 1.794, 1-700 and 1.662). The variation of control cell A4o5 readings ranged
from 0-013 to 0.014 absorbance units and the greatest difference between racks was 0.021
absorbance units (rack means were 0.040 and 0.034). Care should therefore be taken when
using racks for quantitative analysis although the variations did not produce any false positives
or false negatives; additionally, all the virus samples were visually positive and the control cell
samples visually negative.
Cuvette racks were coated with v-globulin, washed with PBST, dried, sealed in polythene bags
and then stored at 37 °C, room temperature, 4 °C, - 20 °C and - 70 °C. The activity of the
coated racks was completely lost when stored at 37 °C or room temperature for 1 month, whereas
storage at 4 °C, - 20 °C and - 70 °C for 9 months did not appreciably affect the activity of the
coated racks compared to freshly coated racks.
Ultra-rapid ELISA
The time taken for the ELISA technique varied from 4 h if the antigen was incubated for 1 h,
to 19 to 20 h if the antigen was incubated overnight producing high A4o5 readings ( > 2 or > 3)
for virus antigen in some infected fry or in cell culture harvests tested in cuvette racks. It seemed
likely, therefore, that shortening the time of incubation would still produce high A4o5 readings.
To test this, the antigens, conjugate and substrate were incubated for 15, 30, 45 or 60 rain with
washes in PBST for 3 x 3 min after an initial quick rinse; additionally, the antigen and
conjugate were incubated for 15, 30 or 45 rain and the substrate was incubated for 60 min in each
case (Fig. 3). When all reagents were incubated for 15 min each, there was a positive result for
virus antigen, but the A4os was even higher when the substrate was incubated for 60 min rather
than 15 min. Incubations of reagents for 30 min or longer gave strong positive results for virus
antigen, although it should be noted that in this experiment the control cell cultures gave a
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 13:07:53
328
P. F. D I X O N A N D B. J. H I L L
>3
3
15
30
45
60
Incubation time (min)
Fig. 3. Effect of varying the time of incubation of reactants on ELISA absorbance. Antigens (~], BF-2
cell-grown IPNV; [], control BF-2 cells) or PBST (k~), conjugate and substrate were all incubated for
15, 30, 45 or 60 min. Additionally, antigens ([3, BF-2 cell-grown IPNV; II, control BF-2 cells) and
conjugate were both incubated for 15, 30 or 45 min and substrate was incubated for 60 min.
higher than expected A4o5 reading, in the visually positive region. Incubation of control cell
antigens for 15 min did not give this high reading.
Thereafter, tests were done using 15 min incubation periods for antigen and conjugate, 45 min
incubation with substrate, and washes of 3 × 3 min after a quick rinse, giving a total time for the
test of just over 1.5 h; incubation of substrate for 45 min was the maximum required and colour
development of strong positives could be seen within a short time after the addition of substrate.
The incubation of antigen and conjugate was done at 37 °C, whereas all the other stages were
done at room temperature. As it was more convenient to do the test at room temperature and
preferable for a potential field test, incubation of antigen and conjugate at room temperature
was compared with their incubation at 37 °C. The mean A4o5 readings of BF-2 cell-grown virus
antigen with 37 °C incubations or all incubation at room temperature were 2-33 and 1.93
respectively, and those of non-infected BF-2 cell cultures were 0-078 and 0.059 respectively. The
ultra-rapid ELISA technique with room temperature incubation would also still detect I P N
antigen before c.p.e, was evident in infected cell cultures. As we use the technique for detecting
I P N V antigen in cell cultures showing definite c.p.e, and as the reduction in A40s caused by
room temperature incubation was not sufficient to interfere with such detection, room
temperature incubation is now used routinely.
The ultra-rapid technique with room temperature incubation was used to detect I P N V
antigen in experimentally infected fry. Moribund fry were treated as follows: two whole fry were
homogenized, two had their heads, tails and musculature of the back removed and the
remainder of the abdomen homogenized, and the viscera only of two further fry were
homogenized. The whole fry and abdomen samples were both visually positive at a 1:10 dilution
and the viscera sample was visually positive at a 1:100 dilution. However, in further
experiments, virus antigen in fry infected with avirulent strains of I P N V was not detected under
these conditions, although virus was isolated in cell cultures. Non-specific A405 readings of
control fish extracts were low in this rapid method; thus, treatment of extracts with Arklone-P is
unnecessary providett a control extract is included on the cuvette rack.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 13:07:53
ELISA for detection of I P N V
329
DISCUSSION
The study reported here has shown the ELISA technique to be a valuable aid to the diagnosis
of IPN. The test was as quick or quicker to perform and obtain results with (under 2 h) than the
other 'rapid' diagnostic tests described for IPNV, and although we did not directly compare the
sensitivity of the ELISA technique with those other rapid methods, our results showed that the
technique has sufficient sensitivity for most diagnostic applications. The ELISA we developed
can be assessed visually as virus antigen-positive or -negative, and can be done at room
temperature, thus obviating the need for absorbance readers or incubators. As little other
equipment is required, the technique is suitable for use in most fish disease diagnostic
laboratories and it could be used for diagnostic tests on fish farm sites.
Using the technique we could detect virus antigen in cell cultures before c.p.e, was eVident, at
low levels of c.p.e, and in the presence of cytotoxic substances. For routine purposes, however,
we recommend that the ELISA technique is used to replace the confirmatory virus
neutralization test when at least minimal c.p.e, is observed. The technique was specific for
IPNV serogroup 1 as it was demonstrated that only I P N V serotypes Sp, Ab and Vr crossreacted, whereas 10 other viruses pathogenic for rainbow trout or other fish species did not; TV1 (IPNV serogroup 2) was one of the viruses that did not cross-react with I P N V serotype Sp
antiserum, in agreement with the neutralization data of Underwood et al. (1977) but at variance
with those of Macdonald & Gower (1981). It was possible to rapidly serotype I P N V using
coating ~-globulin and conjugates prepared from antiserum against each of the three serogroup 1
serotypes, and there was a good correlation between our virus-neutralization data and ELISA
results except with serotype Vr antiserum. With that, there was a strong homologous reaction
in both techniques, but the cross-reactions of the heterologous serotypes were not consistent
between techniques. The reason for this has not yet been determined.
The data presented here suggest that the tissues of moribund fry with IPN will contain
sufficient virus antigen to give a strong reaction in the ELISA technique. Thus, in cases of
suspected IPN outbreaks in fry the technique may be used for direct diagnosis from fry extracts.
Initially, non-specific absorbance caused by extracts of non-infected fry reduced the
effectiveness of the technique for detecting low levels of virus antigen. However, the nonspecific absorbance could be reduced by treatment of the extract with Arklone-P, although this
treatment was not necessary when the samples were only incubated for 15 min in the ultra-rapid
test, as the short incubation time had the effect of reducing the non-specific absorbance to a very
low level. The ELISA technique was not as sensitive as isolation of the virus in cell cultures for
detecting IPNV carrier fish and, therefore, its use in routine screening of fish stocks is not to be
advocated. The use of the technique for I P N V carrier detection may ultimately be possible if its
sensitivity can be improved and if only virus antigen-rich fractions of viscera are tested, rather
than a combination of the liver, kidney and spleen as done here.
We are currently investigating the use of the ELISA technique to detect other viruses
pathogenic for fish. Initial results have shown that the technique can be used to detect I H N V
and eel rhabdoviruses isolated in cell cultures; we are currently assessing the feasibility of
detecting those viruses directly in infected fish tissues.
We acknowledge the excellent technical assistance of Mr K. Way, who provided the neutralization data on the
antisera.
REFERENCES
CLARK, M. F. & ADAMS,A. N. (1977). Characteristics of the microplate method of enzyme-linked i m m u n o s o r b e n t
assay for the detection of plant viruses. Journal of General Virology 34, 475-483.
CLEATOR, G. M. & BURNEY, L. A. (1980). The haemagglutinating properties of infectious pancreatic necrosis virus.
Archives of Virology 63, 81 85.
DIXON, P. F. & HILL, B. J. (i983). Inactivation of infectious pancreatic necrosis (IPN) virus for vaccine use. Journal
of Fish Diseases (in press).
DOBOS, P., HILL, B. l., HAELETT,R., KELLS, D. T. C., BECHT, H. & TENINGES, D. (1979). Biophysical and biochemical
characterization of five animal viruses with bisegmented double-stranded R N A genomes. Journalof Virology
32, 593M05.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 13:07:53
330
P. F. D I X O N A N D B. J. H I L L
FINLAY, J. & HILL, a. J. (1975). The use of the complement fixation test for the rapid typing of infectious pancreatic
necrosis virus. Aquaculture 5, 305 310.
HILL, B. J., UNDERWOOD,B. O., SMALL,C. J. & BROWN, F. (1975). Physicochemical and serological characterization of
five rhabdoviruses infecting fish. Journal of General Virology 27, 369-378.
HILL, a. J., WILLIAMS,R. F., SMALL,C. J., UNDERWOOD,B. O. & BROWN, F. (1980). Physicochemical and serological
characterization of two rhabdoviruses isolated from eels. Intervirology 14, 208-212.
JORGENSEN, P. E. V. (1974). Indirect fluorescent antibody techniques for demonstration of trout viruses and
corresponding antibody. Acta veterinaria scandinavica 15, 198-205.
KELLY, D. C., EDWARDS,M-L., EVANS,H. F. & ROBERTSON,J. S. 0978). The use of the enzyme linked immunosorbent
assay to detect a nuclear polyhedrosis virus in Heliothis armigera larvae. Journalof General Virology40, 465469.
MACDONALD,R. D. & GOWER, D. A. 0981). Genomic and phenotypic divergence a m o n g three serotypes of aquatic
birnaviruses (infectious pancreatic necrosis virus). Virology 114, 187-195.
NICHOLSON, B. L. & HENCHAL, E. A. (1978). Rapid identification of infectious pancreatic necrosis virus in infected
cell cultures by immunoperoxidase techniques. Journal of Wildlife Diseases 14, 465~,69.
PIPER, D., NICHOLSON,B. L. & DUNN, J. (1973). Immunofluorescent study of the replication of infectious pancreatic
necrosis virus in trout and Atlantic salmon cell cultures. Infection and Immunity 8, 249-254.
SWANSON,R. N. & GILLESPIE, J. H. (1981). An indirect fluorescent antibody test for the rapid detection of infectious
pancreatic necrosis virus in tissues. Journal ofFish Diseases 4, 309-315.
UNDERWOOD, B. O., SMALL, C. J., BROWN, F. & HILL, B. J. (1977). Relationship of a virus from Tellina tenuis to
infectious pancreatic necrosis virus. Journal of General Virology 36, 93-109.
VOLLER, A., BIDWELL, O. E. & BARTLETT,A. (1979). The Enzyme-linked Immunosorbent Assay (ELISA), pp. 125. St
Peter Port: Dynatech Europe.
(Received 14 June 1982)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 29 Apr 2017 13:07:53