- Email: [email protected]
A latex agglutination test for the detection of canine parvovirus and corresponding antibodies
Journal of Vlrological Elsevier JVM 0067
Methods,
19 (lY88)
I-I?
I
A latex agglutination test for the detection of canine parvovirus and corresponding antibodies M. Bodeus’, ‘Unit of Microbiology,
C. Cambiaso’,
M. Surleraux’
and G. Burtonboy’
Universify of Louvain /U. C. L.). Ee~gium; ‘Unit of Experimental University of Louvain (U. C. L.), Belgium (Accepted
23 September
Medicine.
lY87)
Summary Latex particles coated with rat monoclonal antibodies directed against a canine parvovirus are agglutinated by the virus antigens. In this study, the reaction was quantitated by a technique which had been described as immunoassay by particles counting. The method is as sensitive as a radioimmunoassay with the same antibody, but has the advantages of simplicity and safety. It allows the detection of parvoviral antigens to 4 ngiml. Incubation of these antigens with specific antibodies resulted in inhibition of the agglutination reaction. By this procedure the presence of antibodies against canine parvovirus can be detected to a concentration of 150 ngiml. The system can be easily automated, thus increasing reproducibility. This latex agglutination technique can be performed as a slide test; and although this method is 4-5 times less sensitive, it could be useful in field studies, for the detection of the viral antigens and the corresponding antibodies. Latex
agglutination;
Canine
parvovirus;
Monoclonal
antibody
Introduction Polystyrene beads referred to as latex particles are agglutinated by antigen when coated with antibodies. The advantages of the system appear to be the ease with which immunoglobulins can be adsorbed onto the particles and the stability of the suspension compared to other particulate structures such as erythrocytes. For almost 30 years, the latex agglutination reaction has been used widely as a slide test Correspondence to: Dr. Brussels. Belgium. 0166-0934/X81$03.50
0
M. Bodeus,
1988 Elsevier
Microbiology
Science
Unit,
Publishers
U.C.L.-5490,
B.V. (Biomedical
avenue
Hippocrate
Division)
54, 1200
2
for detection of rheumatoid factor (Singer et al., 1956). Such a procedure is simple and inexpensive, but it is difficult to determine the end point of the reaction. More recently a technique has been developed for quantitating the degree of latex agglutination very accurately (Cambiaso et al., 1977). This immunoassay, based on instrumental particle counting, displays a sensitivity in the range of radio- or enzyme immunoassay. We have used this method for detection of a parvovirus namely the canine parvovirus (CPV) as well as the detection of antibodies against this virus.
Materials
and Methods
Antibodies
A panel of rat monoclonal antibodies against CPV antigens was produced and characterized (Burtonboy et al., 1982). The monoclonal antibodies designated as ‘RH’ followed by an identification number, were grown in roller bottles (Bodeus et al., 1984) and then purified by affinity chromatography (Bazin et al., 1984). For this purpose we used mouse monoclonal antibodies against rat IgG, either a mouse anti-rat kappa light chain (MARK 1) or a mouse anti-rat gamma 2b chain (MARG2b); both were kindly provided by Professor H. Bazin (Bazin et al., 1984). AcA22 coated with the mouse monoclonal antibody by glutaraldehyde (Avrameas and Ternynck, 1967) was packed in small columns (20-30 ml). Culture supernatant of hybridoma secreting rat monoclonal antibodies against CPV was applied to the affinity column. After 3 washing steps (with PBS, PBS containing 2.5 M sodium chloride and again PBS) the pH was decreased by running glycine-HCl 0.1 M buffer, pH 2.8, with 0.15 M sodium chloride. The pH of the eluate was adjusted to neutrality by adding 1 M Tris to each tube of the fraction collector. The eluate was dialysed against PBS with 0.1% sodium azide and concentrated with CX-10 (Millipore). Polyclonal antisera of dogs recovering from haemorrhagic gastroenteritis and of dogs, cats, rats or rabbits immunized with CPV were also used in this study. Rheumatoid factor was inactivated in sera by reduction; in practice, 50 ~1 serum were mixed with 170 ~1 GBS (glycine-HCl 0.1 M buffer, pH 9.2, 0.17 M NaCl) containing 50 mM EDTA and 15 ~1 di-thiothreitol (5 mgiml). After 30 min incubation at 37”C, 15 ~1 of a 0.2% hydrogen peroxide were added to stop the reaction (Cambiaso et al., 1978). Virus
The canine parvovirus used in this study was isolated from the faeces of a dog with haemorrhagic enteritis (Burtonboy et al., 1979). It was grown in feline kidney cells as previously described (Burtonboy et al., 1982). Concentrated virus was obtained by precipitation with polyethylene glycol (PEG): the infected cell culture medium clarified by low speed centrifugation and filtered through 0.45 p_ (Millipore) was treated at 4°C according to Paradiso (1981), by adding 0.3 M sodium chloride and PEG 6,000 (Merck) to a final concentration of 8.5% (w/v). The viral
protein concentration of the samples was estimated from densitometric analysis at 600 nm (CDS-Beckman) of the viral peptides separated by polyacrylamide gel electrophoresis and stained with Coomassie blue, compared to known amounts of protein standards (Surleraux et al., 1984). Virus affinity chromatography purification was carried out on a column of RH 1, according to the method for the preparation of monoclonal antibodies. Faecal samples from puppies were obtained from a breeding kennel during an outbreak of haemorrhagic enteritis. The specimens were suspended at a 1:lO dilution in PBS containing 0.1% sodium azide and clarified by centrifugation. Preparation of RH 1 coated latex Hundred ~1 of a 0.2% (w/v) suspension of purified RH 1 and 50 t~_lof a 10% (w/v) latex suspension (Polystyrene particles, 0.8 TV,,Rhone Poulenc, France) were mixed in 400 u.1 of buffer (5-fold diluted GBS). After stirring for a few seconds and incubation at 37°C for 45 min, the suspension was centrifuged for 10 min at 10,000 rpm (Sorvall-Rotor, 7455798). Particles washed once with 1 ml diluted GBS and twice in GBS 0.27 M containing 1% (w/v) bovine serum albumin (BSA-Calbiochem) were resuspended in 1 ml GBS 0.27 M with BSA. The preparation of the latex conjugate results in the formation of a few spontaneous aggregates; a brief sonication (Branson sonifier B12, Danbury, Connecticut) was used to dissociate the agglutinated particles (Masson et al., 1981). The coated latex was distributed in 100 ~1 aliquots and kept at -2o”C, and diluted 1:8 in GBS 0.27 M with 1% BSA before use. Detection of viral antigens Immunoassay by particle counting. For detection of the virus 25 t~_lof viral sample mixed with 25 ~1 of the RH 1 coated latex suspension were incubated at room temperature for 30 min on a vibrating mixer (lc50 Hz with a vibration amplitude of 2-5 mm). After incubation, 2 ml of GBS 0.27 M were added to each tube; the sample was then passed through the flow cell Technicon autocounter to measure the number of non-agglutinated particles. The height of the control peak corresponded to the number of isolated latex particles when no viral antigen was present in the reaction. Slide test agglutination assay. 20 ~1 of the undiluted latex suspension, coated with RH 1, were mixed on a piece of black cardboard, with 20 ~1 of a dilution of the viral suspension being tested. The result was observed after 5 min. The control. 20 t~_lPBS, was added to the latex. A positive result was characterized by the presence of large aggregates, whereas in the negative test, the suspension appears homogeneous (Fig. 1.). Haemagglutination (HA). tination of rhesus monkey et al., 1982).
The amount of virus may be estimated by haemaggluerythrocytes at 4°C as described previously (Burtonboy
Fig, 1. On the right. latex particles coated with RH 1 are mixed with a CPV suspension: agglutination is obvious. On the left, a contol where latex is mixed with uninfected culture medium.
RIA. Polystyrene beads of 6 mm diameter (Abbott Laboratories) were coated with purified RH 1 (Delacroix et al., 1982). Coated beads were incubated with the test specimen. The viral antigens, if present in the sample were bound to the antibodies. During a second incubation, RH 1 tagged with 1251 (McConahey and Dixon, 1966) was added, creating an antibody-antigen-antibody sandwich. After washing, the amount of radioactivity bound was determined in a gamma scintillation well counter. Detection of antibodies Samples of 25 ~1 of the minimal concentraImmunoassay by particle counting. tion of virus giving the maximal agglutination were pre-incubated with 25 ~1 of the test antibodies, for 15 min at room temperature. The latex was then added and the agglutination measured as described previously. In this case we observed an inhibition of agglutination and the peak increased with the concentration of antibodies. Slide test agglutination maximal agglutination at room temperature. suspension. Inhibition
giving the assay. 20 ~1 of the minimal viral concentration were pre-incubated with 20 l.~l of the antibodies for 15 min The mixture was then added to 20 ~1 of the undiluted latex was characterized by the absence of visible aggregates.
NUMBER
Fig. 2. In the optical
system
OF LATEX
PARTlCLESx107ml-’
used for the detection, peak height of free latex particles.
is proportional
to the concentration
inhibition of agglutination (IHA). mated by inhibition boy et al., 19982).
The amount of antibodies anti-CPV of haemagglutination of rhesus monkey erythrocytes
was esti(Burton-
RZA.
Polystyrene beads (6 mm) coated with RH 1 were incubated with a known viral suspension and then used to detect the antibodies, the presence of which was shown by using a specific anti-IgG antibody labelled with lz51, kindly supplied by Professor J. P. Vaerman (1969).
Results
Detection of viral antigens Suspension of polystyrene beads coated with RH 1 was analysed in the particle counter. When antigen was not added to the latex, the peak height (PH) appeared to be proportional to the number of particles (Fig. 2). When viral antigen was mixed with the latex, some particles were agglutinated, the number of isolated beads diminished and the peak decreased accordingly. Various concentrations ranging from 10 kg/ml-1 ng/ml of PEG concentrated virus were tested by this method. Agglutination was still detected at a concentration of 4 ng/ml (Fig. 3a). The difference between the peak without and with antigen was used as an agglutination index (A.I.), the percentage of which was calculated as follows: A I ~ = PH without antigen - PH with antigen . . 0 PH without antigen
x IO0
LATEX PARTICLE COUNTING IMMUNOASSAY. Antigen detection
160
Fig. 3a. From right to left: the first 2 peaks measure the amount of particles in absence of antigen. In the following samples the latex coated with RH 1 has been incubated with concentrations of purified CPV from 10 pg/ml to I ngiml. LATEX
PARTICLE
COUNTING
IMMUNOASSAY
: Antigen
Detection
1
lOO[
Log VIRUS
CONCENTRATION
(nglml)
Fig. 3b. The diagram summarizes agglutination index for the various amounts of antigen. Each point is the mean value of 5 independent assays, the little bars indicate the range limits.
This value appeared to be related to the amount of antigen following curve (Fig. 3b); 10 ngiml of antigen resulted in 20% of agglutination,
a S-shaped 90% being
AFFINITY Haemagglutination
z F 5 5
CHROMATOGRAPHY:
2,000
-
z i=
@
CPVl2
@
f=Bs
@
PBS+NaCI
CPV/Z-RHt
IA
(HA) @
supcrnatant
PBS
@pH
28
@PBS
25t.i
3
1,000 -
8
4
0
: I
PP
1 ,5,000
o--
?I&?
b__.I” _ Radio
( RIA 1
immunoassay
h
10,000 -
h
0
f
0
-
5,000
@ I
@
00
1
II
1
n
.i
60-
$ a
z i=
40 -
3 20 -
OO
20
40 ELUTION
60
60
0
TUBE
Fig. 4. Each fraction (5 ml) eluted from a RH 1 affinity column has been tested by HA of rhesus monkey erythrocytes at 4”C, RIA in a sandwich assay using RH 1 and by latex immunoassay with beads coated with the same RH 1.
produced by 10 Fg/ml. Each antigen concentration was tested 5 times and the range of values obtained for a given concentration was always smaller than 10% in the agglutination index.
8 The latex coated with RH 1 was also used in a slide test. to assay various amounts of PEG concentrated virus. By this method, the antigen was detected at a concentration of 100 ngiml; although in this case the limit of detection was subjective. The presence of virus can be demonstrated in faecal samples from dogs with diarrhoea; 9 such samples in which parvovirus was detected by electron microscopy were tested by particle counting immunoassay; 7 gave a positive reaction. of which 6 were also positive in a slide test. Each positive reaction could be inhibited by an antiserum against CPV or by purified RH 1. Six faecal samples obtained from healthy animals and negative for viral particles by electron microscopy were also negative by the latex agglutination test by instrument counting as well as in slide test. The technique was also useful in following the different steps of affinity chromatography. For this purpose the collected fractions were tested in parallel by HA, RIA and latex particle counting (Fig. 4). During the first step, the sample containing the unbound material flowing through the column, although negative by HA, was shown to contain viral antigen by latex agglutination and RIA. During the washing steps all the methods failed to detect any virus. A peak of eluted virus was detected at pH 2.8 by the 3 methods, but only latex and RIA demonstrated the presence of an additional peak in the last washing with PBS. Detection of antibodies When antibodies were added
to the minimal
viral concentration
giving a maxi-
LATEX PARTICLE COUNTING IMMUNOASSAY:Antibody detection
Fig. 5a. From right to left: the first 2 peaks in absence of antigen and antibody; for the incubated with a standard concentration of been mixed with various concentrations of being incubated
were obtained by measuring the amount of free particles 2 following peaks, the latex coated with RH 1 has been CPV; in the other samples the same amount of virus has purified RH 1 ranging from 10 &ml to 40 ng/ml before with the RH 1 coated beads.
Y
LATEX PARTICLE
COUNTING
IMMUNOASSAY
: Antibody Detection
100
80 Z r 60
m if Z -
LO
i
20
OO
3
2
1
4
Log RHl CONCENTRATION Fig. 5h. The diagram
summarizes
3
(ng/ml)
the results obtained by measuring tibody concentration.
the inhibition
index for each an-
agglutination, the agglutination process diminished, the peak height increased accordingly. This technique was used to measure the amount of antibody. Various concentrations of purified monoclonal antibody ranging from 10 kg/ml to 40 ngiml have been analyzed by this method. Inhibition of agglutination was still detected at a concentration of 150 ngiml (Fig. 5a). An inhibition index (1.1.) was calculcated as follows: mal
PH (with antibody I.1.B
=
PH (without
antibody
+ virus)
- PH (with virus)
and virus)
- PH (with virus)
x 100
This value was related to the amount of antibodies following a sigmoid curve (Fig. 5b): 300 ngiml of antibodies resulted in 20% inhibition, 90% being produced by 2 kg/ml. Each antibody concentration was tested 5 times with always less than 10% variation in the inhibition index. In the slide test, inhibition was detected at a concentration as low as 600 ngiml. The kinetics of the antibody response in vaccinated dogs was tested by latex and RIA; the sensitivity was the same for both methods (Fig. 6). The presence of antibodies was also demonstrated by particle counting in sera of dogs suffering from gastroenteritis and from dogs, cats, rats and rabbits immunized against CPV. The results confirmed those obtained by RIA. The technique was also useful to test different fractions collected during affinity chromatography: antibodies were detected at pH 2.8, as shown in Fig. 7.
7
80
4
/I ’
,/ &‘\\,
70
6-
I Fig. 6. Detection
1 DAYS
’
t
52
of anti-CPV antibodies in the serum ol a dog at day 0. 21, 34. 47 and X2 after immunization with a vaccinal strain. RIA = 0-0; PACIA = o----o.
Discussion Immunoassay by latex particle counting was shown to be a suitable method for the detection and quantification of various antigens and antibodies. haptens, immune complexes and rheumatoid factor (Cambiaso et al.. 1977; Masson et al.. 1979; AFFINITY
z
140
z5
100
CHROMATOGRAPHY
: RH%-MARG
2blA
w =
60
3 a
20
Fig. 7. From right to left: the first 2 peaks were obtained by measuring the amount 01 free particles in absence of antigen and of antibody. The 10 following peaks were a curve of virus dilution; the viral concentration indicated by the arrow has been used in the inhibition test. The eluate of the affinit!, column was collected in 5 ml fractions: peak I (solution to purify); peaks 2-S (loading the column): peaks 613 (washing steps): peaks l-1 and IS (elution at pH 2.8): peaks l&l8 (final wash in PBS).
I1
Sindic et al., 1984), the advantages and possible pitfalls of the procedure being discussed by each group of authors. The procedure appears to be a practical and sensitive method using non-radioactive reagents; the system is homogeneous, in which separation is conducted electronically without any need for filtration or centrifugation. It can be easily automated, thereby increasing precision, speed and facility It has been reported that non specific agglutinators or inhibitors can interfere with the agglutination process and various solutions have been proposed (Masson et al., 1981). Prozone effect was also described in antigen or antibody excess (Cambiaso et al., 1977). We investigated the potential of the method in the field of virology. Canine parvovirus (CPV) was chosen as a model for this study. It was the subject of investigations in our laboratory; purified viral antigens could be obtained from infected cell culture medium and a panel of rat monoclonal antibodies directed against this virus was available (Burtonboy et al., 1982). Moreover, we used faeces from dogs with gastroenteritis as well as sera of different animals immunized against CPV. The agglutination test we used in detecting antigen in purified or biological samples appears to be specific as it was inhibited by antisera against CPV, and the correlation between latex immunoassay and RIA with the same reagents was high with the sensitivity of both tests reaching 4 ngiml of viral antigen. A prozone effect was not observed with the concentration of CPV used for the test. and the faecal samples did not yield non-specific reactions. In the inhibition test for antibodies. the serum samples were treated with dithiothreitol to avoid non-specific reactions; under these conditions the correlation with RIA was excellent with a sensitivity of 150 ngiml. The latex particle counting system is a practical method for the detection of antigen or antibody in a large number of samples; 60 specimens can be tested in 1 h; it can be easily used for systematic screening of blood donors for the detection of anti-HBs (Galanti et al., 1987), parvovirus B 19 and HIV. We used only polystyrene beads coated with a monoclonal antibody, either to detect the CPV or the antibodies directed against this virus. Of course, other possibilities exist, such as antigen or F(ab’)2 coated beads and the versatility of the system has been previously described (Masson et al.. 1981). A slide test assay was also performed for the detection of both CPV antigens or antibodies: it does not give quantitative information but is rapid, easy to carry out and may be useful for field studies.
Acknowledgements
We thank Prof. H. Bazin for helpful suggestions and Prof. J.P. Vaerman for critical and useful comments. We also thank N. Delferriere, J.L. Guarin, J.P. Kindts, J.M. Malache and F. Nisol for their skillful assistance. This work was supported by the ‘Fonds National de la Recherche Scientifique’ (credit aux chercheurs. 1985).
12
References Avrameas. S. and Ternynck, T. (lY6Y) Immunochemistry 6. 53-59. Bazin. H., Xhurdebise, L.M.. Burtonhoy. G.. Lebacq. A.M.. Declercq. L. and Cormont. F. (19X-l) .I. Immunol. Methods 66, 261-269. Bodeus. M.. Burtonboy. G. and Bazin, H. (lY85) J. Immunol. Methods 7Y. l-h. Burtonboy. G., Coignoul, F.. DelterriPre. N. and Pastoret, P.P. (1979) Arch. Virol. 61. I-00. Burtonboy. G., Bazin. H. and Delferrihre. N. (1982) Arch. Viral. 71. 291-000. Cambiaso, C.L.. Leek, A.E.. de Steenwinkel, F.. Billen. J. and Masson. P.L. (1977) J. Immunol. Methods 1X. 3433. Cambiaso, C.L.. Riccomi. H.. Sindic. C. and Masson. P.L. (1978) J. Immunol. Methods 23, 29-50. Delacroix. D.L.. Dchennin. J.P. and Vaerman. J.P. (1982) J. Immunol. Methods 48. 327-337. Galanti, L., Cornu, C.. Cambiaso. C.L.. Lamy. M.E. and Masson. P.L. (1987) J. Viral. Methods IX. 215-223. McConahey. P.J. and Dixon, F.J. (lY66) Int. Arch. Allergy 2Y. 185. Masson. P.L.. Cambiaso. C.L., Collet-Casaart. D. 1 Magnusson. C.G.M.. Richards. C.B. and Sindlc. C.J.M. (1981) Methods in Enzymology (Academic Press). 71. 10&13Y. Paradise. P.R. (1981) J. Viral. 39. X0&807. Sindic. C.J.M., Cambiaso. C.L.. Depre. A.. Laterrc. E.C. and Masson. P.L. ( 1984)J. Neuroimmunol. 6. Y-18. Singer. J.M. and Plotz, C.M. (1956) Am. J. Med. 21. XXX. Surleraux. M. and Burtonboy. G. (1983) Arch. Viral. X2, 233-240. Vaerman. J.P. and Herremans. J.F. (1969) Immunochemistry 6. 779-786.
Methods,
19 (lY88)
I-I?
I
A latex agglutination test for the detection of canine parvovirus and corresponding antibodies M. Bodeus’, ‘Unit of Microbiology,
C. Cambiaso’,
M. Surleraux’
and G. Burtonboy’
Universify of Louvain /U. C. L.). Ee~gium; ‘Unit of Experimental University of Louvain (U. C. L.), Belgium (Accepted
23 September
Medicine.
lY87)
Summary Latex particles coated with rat monoclonal antibodies directed against a canine parvovirus are agglutinated by the virus antigens. In this study, the reaction was quantitated by a technique which had been described as immunoassay by particles counting. The method is as sensitive as a radioimmunoassay with the same antibody, but has the advantages of simplicity and safety. It allows the detection of parvoviral antigens to 4 ngiml. Incubation of these antigens with specific antibodies resulted in inhibition of the agglutination reaction. By this procedure the presence of antibodies against canine parvovirus can be detected to a concentration of 150 ngiml. The system can be easily automated, thus increasing reproducibility. This latex agglutination technique can be performed as a slide test; and although this method is 4-5 times less sensitive, it could be useful in field studies, for the detection of the viral antigens and the corresponding antibodies. Latex
agglutination;
Canine
parvovirus;
Monoclonal
antibody
Introduction Polystyrene beads referred to as latex particles are agglutinated by antigen when coated with antibodies. The advantages of the system appear to be the ease with which immunoglobulins can be adsorbed onto the particles and the stability of the suspension compared to other particulate structures such as erythrocytes. For almost 30 years, the latex agglutination reaction has been used widely as a slide test Correspondence to: Dr. Brussels. Belgium. 0166-0934/X81$03.50
0
M. Bodeus,
1988 Elsevier
Microbiology
Science
Unit,
Publishers
U.C.L.-5490,
B.V. (Biomedical
avenue
Hippocrate
Division)
54, 1200
2
for detection of rheumatoid factor (Singer et al., 1956). Such a procedure is simple and inexpensive, but it is difficult to determine the end point of the reaction. More recently a technique has been developed for quantitating the degree of latex agglutination very accurately (Cambiaso et al., 1977). This immunoassay, based on instrumental particle counting, displays a sensitivity in the range of radio- or enzyme immunoassay. We have used this method for detection of a parvovirus namely the canine parvovirus (CPV) as well as the detection of antibodies against this virus.
Materials
and Methods
Antibodies
A panel of rat monoclonal antibodies against CPV antigens was produced and characterized (Burtonboy et al., 1982). The monoclonal antibodies designated as ‘RH’ followed by an identification number, were grown in roller bottles (Bodeus et al., 1984) and then purified by affinity chromatography (Bazin et al., 1984). For this purpose we used mouse monoclonal antibodies against rat IgG, either a mouse anti-rat kappa light chain (MARK 1) or a mouse anti-rat gamma 2b chain (MARG2b); both were kindly provided by Professor H. Bazin (Bazin et al., 1984). AcA22 coated with the mouse monoclonal antibody by glutaraldehyde (Avrameas and Ternynck, 1967) was packed in small columns (20-30 ml). Culture supernatant of hybridoma secreting rat monoclonal antibodies against CPV was applied to the affinity column. After 3 washing steps (with PBS, PBS containing 2.5 M sodium chloride and again PBS) the pH was decreased by running glycine-HCl 0.1 M buffer, pH 2.8, with 0.15 M sodium chloride. The pH of the eluate was adjusted to neutrality by adding 1 M Tris to each tube of the fraction collector. The eluate was dialysed against PBS with 0.1% sodium azide and concentrated with CX-10 (Millipore). Polyclonal antisera of dogs recovering from haemorrhagic gastroenteritis and of dogs, cats, rats or rabbits immunized with CPV were also used in this study. Rheumatoid factor was inactivated in sera by reduction; in practice, 50 ~1 serum were mixed with 170 ~1 GBS (glycine-HCl 0.1 M buffer, pH 9.2, 0.17 M NaCl) containing 50 mM EDTA and 15 ~1 di-thiothreitol (5 mgiml). After 30 min incubation at 37”C, 15 ~1 of a 0.2% hydrogen peroxide were added to stop the reaction (Cambiaso et al., 1978). Virus
The canine parvovirus used in this study was isolated from the faeces of a dog with haemorrhagic enteritis (Burtonboy et al., 1979). It was grown in feline kidney cells as previously described (Burtonboy et al., 1982). Concentrated virus was obtained by precipitation with polyethylene glycol (PEG): the infected cell culture medium clarified by low speed centrifugation and filtered through 0.45 p_ (Millipore) was treated at 4°C according to Paradiso (1981), by adding 0.3 M sodium chloride and PEG 6,000 (Merck) to a final concentration of 8.5% (w/v). The viral
protein concentration of the samples was estimated from densitometric analysis at 600 nm (CDS-Beckman) of the viral peptides separated by polyacrylamide gel electrophoresis and stained with Coomassie blue, compared to known amounts of protein standards (Surleraux et al., 1984). Virus affinity chromatography purification was carried out on a column of RH 1, according to the method for the preparation of monoclonal antibodies. Faecal samples from puppies were obtained from a breeding kennel during an outbreak of haemorrhagic enteritis. The specimens were suspended at a 1:lO dilution in PBS containing 0.1% sodium azide and clarified by centrifugation. Preparation of RH 1 coated latex Hundred ~1 of a 0.2% (w/v) suspension of purified RH 1 and 50 t~_lof a 10% (w/v) latex suspension (Polystyrene particles, 0.8 TV,,Rhone Poulenc, France) were mixed in 400 u.1 of buffer (5-fold diluted GBS). After stirring for a few seconds and incubation at 37°C for 45 min, the suspension was centrifuged for 10 min at 10,000 rpm (Sorvall-Rotor, 7455798). Particles washed once with 1 ml diluted GBS and twice in GBS 0.27 M containing 1% (w/v) bovine serum albumin (BSA-Calbiochem) were resuspended in 1 ml GBS 0.27 M with BSA. The preparation of the latex conjugate results in the formation of a few spontaneous aggregates; a brief sonication (Branson sonifier B12, Danbury, Connecticut) was used to dissociate the agglutinated particles (Masson et al., 1981). The coated latex was distributed in 100 ~1 aliquots and kept at -2o”C, and diluted 1:8 in GBS 0.27 M with 1% BSA before use. Detection of viral antigens Immunoassay by particle counting. For detection of the virus 25 t~_lof viral sample mixed with 25 ~1 of the RH 1 coated latex suspension were incubated at room temperature for 30 min on a vibrating mixer (lc50 Hz with a vibration amplitude of 2-5 mm). After incubation, 2 ml of GBS 0.27 M were added to each tube; the sample was then passed through the flow cell Technicon autocounter to measure the number of non-agglutinated particles. The height of the control peak corresponded to the number of isolated latex particles when no viral antigen was present in the reaction. Slide test agglutination assay. 20 ~1 of the undiluted latex suspension, coated with RH 1, were mixed on a piece of black cardboard, with 20 ~1 of a dilution of the viral suspension being tested. The result was observed after 5 min. The control. 20 t~_lPBS, was added to the latex. A positive result was characterized by the presence of large aggregates, whereas in the negative test, the suspension appears homogeneous (Fig. 1.). Haemagglutination (HA). tination of rhesus monkey et al., 1982).
The amount of virus may be estimated by haemaggluerythrocytes at 4°C as described previously (Burtonboy
Fig, 1. On the right. latex particles coated with RH 1 are mixed with a CPV suspension: agglutination is obvious. On the left, a contol where latex is mixed with uninfected culture medium.
RIA. Polystyrene beads of 6 mm diameter (Abbott Laboratories) were coated with purified RH 1 (Delacroix et al., 1982). Coated beads were incubated with the test specimen. The viral antigens, if present in the sample were bound to the antibodies. During a second incubation, RH 1 tagged with 1251 (McConahey and Dixon, 1966) was added, creating an antibody-antigen-antibody sandwich. After washing, the amount of radioactivity bound was determined in a gamma scintillation well counter. Detection of antibodies Samples of 25 ~1 of the minimal concentraImmunoassay by particle counting. tion of virus giving the maximal agglutination were pre-incubated with 25 ~1 of the test antibodies, for 15 min at room temperature. The latex was then added and the agglutination measured as described previously. In this case we observed an inhibition of agglutination and the peak increased with the concentration of antibodies. Slide test agglutination maximal agglutination at room temperature. suspension. Inhibition
giving the assay. 20 ~1 of the minimal viral concentration were pre-incubated with 20 l.~l of the antibodies for 15 min The mixture was then added to 20 ~1 of the undiluted latex was characterized by the absence of visible aggregates.
NUMBER
Fig. 2. In the optical
system
OF LATEX
PARTlCLESx107ml-’
used for the detection, peak height of free latex particles.
is proportional
to the concentration
inhibition of agglutination (IHA). mated by inhibition boy et al., 19982).
The amount of antibodies anti-CPV of haemagglutination of rhesus monkey erythrocytes
was esti(Burton-
RZA.
Polystyrene beads (6 mm) coated with RH 1 were incubated with a known viral suspension and then used to detect the antibodies, the presence of which was shown by using a specific anti-IgG antibody labelled with lz51, kindly supplied by Professor J. P. Vaerman (1969).
Results
Detection of viral antigens Suspension of polystyrene beads coated with RH 1 was analysed in the particle counter. When antigen was not added to the latex, the peak height (PH) appeared to be proportional to the number of particles (Fig. 2). When viral antigen was mixed with the latex, some particles were agglutinated, the number of isolated beads diminished and the peak decreased accordingly. Various concentrations ranging from 10 kg/ml-1 ng/ml of PEG concentrated virus were tested by this method. Agglutination was still detected at a concentration of 4 ng/ml (Fig. 3a). The difference between the peak without and with antigen was used as an agglutination index (A.I.), the percentage of which was calculated as follows: A I ~ = PH without antigen - PH with antigen . . 0 PH without antigen
x IO0
LATEX PARTICLE COUNTING IMMUNOASSAY. Antigen detection
160
Fig. 3a. From right to left: the first 2 peaks measure the amount of particles in absence of antigen. In the following samples the latex coated with RH 1 has been incubated with concentrations of purified CPV from 10 pg/ml to I ngiml. LATEX
PARTICLE
COUNTING
IMMUNOASSAY
: Antigen
Detection
1
lOO[
Log VIRUS
CONCENTRATION
(nglml)
Fig. 3b. The diagram summarizes agglutination index for the various amounts of antigen. Each point is the mean value of 5 independent assays, the little bars indicate the range limits.
This value appeared to be related to the amount of antigen following curve (Fig. 3b); 10 ngiml of antigen resulted in 20% of agglutination,
a S-shaped 90% being
AFFINITY Haemagglutination
z F 5 5
CHROMATOGRAPHY:
2,000
-
z i=
@
CPVl2
@
f=Bs
@
PBS+NaCI
CPV/Z-RHt
IA
(HA) @
supcrnatant
PBS
@pH
28
@PBS
25t.i
3
1,000 -
8
4
0
: I
PP
1 ,5,000
o--
?I&?
b__.I” _ Radio
( RIA 1
immunoassay
h
10,000 -
h
0
f
0
-
5,000
@ I
@
00
1
II
1
n
.i
60-
$ a
z i=
40 -
3 20 -
OO
20
40 ELUTION
60
60
0
TUBE
Fig. 4. Each fraction (5 ml) eluted from a RH 1 affinity column has been tested by HA of rhesus monkey erythrocytes at 4”C, RIA in a sandwich assay using RH 1 and by latex immunoassay with beads coated with the same RH 1.
produced by 10 Fg/ml. Each antigen concentration was tested 5 times and the range of values obtained for a given concentration was always smaller than 10% in the agglutination index.
8 The latex coated with RH 1 was also used in a slide test. to assay various amounts of PEG concentrated virus. By this method, the antigen was detected at a concentration of 100 ngiml; although in this case the limit of detection was subjective. The presence of virus can be demonstrated in faecal samples from dogs with diarrhoea; 9 such samples in which parvovirus was detected by electron microscopy were tested by particle counting immunoassay; 7 gave a positive reaction. of which 6 were also positive in a slide test. Each positive reaction could be inhibited by an antiserum against CPV or by purified RH 1. Six faecal samples obtained from healthy animals and negative for viral particles by electron microscopy were also negative by the latex agglutination test by instrument counting as well as in slide test. The technique was also useful in following the different steps of affinity chromatography. For this purpose the collected fractions were tested in parallel by HA, RIA and latex particle counting (Fig. 4). During the first step, the sample containing the unbound material flowing through the column, although negative by HA, was shown to contain viral antigen by latex agglutination and RIA. During the washing steps all the methods failed to detect any virus. A peak of eluted virus was detected at pH 2.8 by the 3 methods, but only latex and RIA demonstrated the presence of an additional peak in the last washing with PBS. Detection of antibodies When antibodies were added
to the minimal
viral concentration
giving a maxi-
LATEX PARTICLE COUNTING IMMUNOASSAY:Antibody detection
Fig. 5a. From right to left: the first 2 peaks in absence of antigen and antibody; for the incubated with a standard concentration of been mixed with various concentrations of being incubated
were obtained by measuring the amount of free particles 2 following peaks, the latex coated with RH 1 has been CPV; in the other samples the same amount of virus has purified RH 1 ranging from 10 &ml to 40 ng/ml before with the RH 1 coated beads.
Y
LATEX PARTICLE
COUNTING
IMMUNOASSAY
: Antibody Detection
100
80 Z r 60
m if Z -
LO
i
20
OO
3
2
1
4
Log RHl CONCENTRATION Fig. 5h. The diagram
summarizes
3
(ng/ml)
the results obtained by measuring tibody concentration.
the inhibition
index for each an-
agglutination, the agglutination process diminished, the peak height increased accordingly. This technique was used to measure the amount of antibody. Various concentrations of purified monoclonal antibody ranging from 10 kg/ml to 40 ngiml have been analyzed by this method. Inhibition of agglutination was still detected at a concentration of 150 ngiml (Fig. 5a). An inhibition index (1.1.) was calculcated as follows: mal
PH (with antibody I.1.B
=
PH (without
antibody
+ virus)
- PH (with virus)
and virus)
- PH (with virus)
x 100
This value was related to the amount of antibodies following a sigmoid curve (Fig. 5b): 300 ngiml of antibodies resulted in 20% inhibition, 90% being produced by 2 kg/ml. Each antibody concentration was tested 5 times with always less than 10% variation in the inhibition index. In the slide test, inhibition was detected at a concentration as low as 600 ngiml. The kinetics of the antibody response in vaccinated dogs was tested by latex and RIA; the sensitivity was the same for both methods (Fig. 6). The presence of antibodies was also demonstrated by particle counting in sera of dogs suffering from gastroenteritis and from dogs, cats, rats and rabbits immunized against CPV. The results confirmed those obtained by RIA. The technique was also useful to test different fractions collected during affinity chromatography: antibodies were detected at pH 2.8, as shown in Fig. 7.
7
80
4
/I ’
,/ &‘\\,
70
6-
I Fig. 6. Detection
1 DAYS
’
t
52
of anti-CPV antibodies in the serum ol a dog at day 0. 21, 34. 47 and X2 after immunization with a vaccinal strain. RIA = 0-0; PACIA = o----o.
Discussion Immunoassay by latex particle counting was shown to be a suitable method for the detection and quantification of various antigens and antibodies. haptens, immune complexes and rheumatoid factor (Cambiaso et al.. 1977; Masson et al.. 1979; AFFINITY
z
140
z5
100
CHROMATOGRAPHY
: RH%-MARG
2blA
w =
60
3 a
20
Fig. 7. From right to left: the first 2 peaks were obtained by measuring the amount 01 free particles in absence of antigen and of antibody. The 10 following peaks were a curve of virus dilution; the viral concentration indicated by the arrow has been used in the inhibition test. The eluate of the affinit!, column was collected in 5 ml fractions: peak I (solution to purify); peaks 2-S (loading the column): peaks 613 (washing steps): peaks l-1 and IS (elution at pH 2.8): peaks l&l8 (final wash in PBS).
I1
Sindic et al., 1984), the advantages and possible pitfalls of the procedure being discussed by each group of authors. The procedure appears to be a practical and sensitive method using non-radioactive reagents; the system is homogeneous, in which separation is conducted electronically without any need for filtration or centrifugation. It can be easily automated, thereby increasing precision, speed and facility It has been reported that non specific agglutinators or inhibitors can interfere with the agglutination process and various solutions have been proposed (Masson et al., 1981). Prozone effect was also described in antigen or antibody excess (Cambiaso et al., 1977). We investigated the potential of the method in the field of virology. Canine parvovirus (CPV) was chosen as a model for this study. It was the subject of investigations in our laboratory; purified viral antigens could be obtained from infected cell culture medium and a panel of rat monoclonal antibodies directed against this virus was available (Burtonboy et al., 1982). Moreover, we used faeces from dogs with gastroenteritis as well as sera of different animals immunized against CPV. The agglutination test we used in detecting antigen in purified or biological samples appears to be specific as it was inhibited by antisera against CPV, and the correlation between latex immunoassay and RIA with the same reagents was high with the sensitivity of both tests reaching 4 ngiml of viral antigen. A prozone effect was not observed with the concentration of CPV used for the test. and the faecal samples did not yield non-specific reactions. In the inhibition test for antibodies. the serum samples were treated with dithiothreitol to avoid non-specific reactions; under these conditions the correlation with RIA was excellent with a sensitivity of 150 ngiml. The latex particle counting system is a practical method for the detection of antigen or antibody in a large number of samples; 60 specimens can be tested in 1 h; it can be easily used for systematic screening of blood donors for the detection of anti-HBs (Galanti et al., 1987), parvovirus B 19 and HIV. We used only polystyrene beads coated with a monoclonal antibody, either to detect the CPV or the antibodies directed against this virus. Of course, other possibilities exist, such as antigen or F(ab’)2 coated beads and the versatility of the system has been previously described (Masson et al.. 1981). A slide test assay was also performed for the detection of both CPV antigens or antibodies: it does not give quantitative information but is rapid, easy to carry out and may be useful for field studies.
Acknowledgements
We thank Prof. H. Bazin for helpful suggestions and Prof. J.P. Vaerman for critical and useful comments. We also thank N. Delferriere, J.L. Guarin, J.P. Kindts, J.M. Malache and F. Nisol for their skillful assistance. This work was supported by the ‘Fonds National de la Recherche Scientifique’ (credit aux chercheurs. 1985).
12
References Avrameas. S. and Ternynck, T. (lY6Y) Immunochemistry 6. 53-59. Bazin. H., Xhurdebise, L.M.. Burtonhoy. G.. Lebacq. A.M.. Declercq. L. and Cormont. F. (19X-l) .I. Immunol. Methods 66, 261-269. Bodeus. M.. Burtonboy. G. and Bazin, H. (lY85) J. Immunol. Methods 7Y. l-h. Burtonboy. G., Coignoul, F.. DelterriPre. N. and Pastoret, P.P. (1979) Arch. Virol. 61. I-00. Burtonboy. G., Bazin. H. and Delferrihre. N. (1982) Arch. Viral. 71. 291-000. Cambiaso, C.L.. Leek, A.E.. de Steenwinkel, F.. Billen. J. and Masson. P.L. (1977) J. Immunol. Methods 1X. 3433. Cambiaso, C.L.. Riccomi. H.. Sindic. C. and Masson. P.L. (1978) J. Immunol. Methods 23, 29-50. Delacroix. D.L.. Dchennin. J.P. and Vaerman. J.P. (1982) J. Immunol. Methods 48. 327-337. Galanti, L., Cornu, C.. Cambiaso. C.L.. Lamy. M.E. and Masson. P.L. (1987) J. Viral. Methods IX. 215-223. McConahey. P.J. and Dixon, F.J. (lY66) Int. Arch. Allergy 2Y. 185. Masson. P.L.. Cambiaso. C.L., Collet-Casaart. D. 1 Magnusson. C.G.M.. Richards. C.B. and Sindlc. C.J.M. (1981) Methods in Enzymology (Academic Press). 71. 10&13Y. Paradise. P.R. (1981) J. Viral. 39. X0&807. Sindic. C.J.M., Cambiaso. C.L.. Depre. A.. Laterrc. E.C. and Masson. P.L. ( 1984)J. Neuroimmunol. 6. Y-18. Singer. J.M. and Plotz, C.M. (1956) Am. J. Med. 21. XXX. Surleraux. M. and Burtonboy. G. (1983) Arch. Viral. X2, 233-240. Vaerman. J.P. and Herremans. J.F. (1969) Immunochemistry 6. 779-786.