A modified rapid enzyme immunoassay for the detection of rabies and rabies-related viruses: RREID-lyssa

Biologicals (1992) 20, 51-58

A Modified Rapid Enzyme Immunoassay for the Detection of Rabies and Rabies-Related Viruses: RREID-lyssa Pierre Perrin,* Corinne Gontier, Eric Lecocq and Herve Bourhy Rabies Unit, National Reference Centre for Rabies, Institut Pasteur, 25 rue du Docteur Roux, 75724 Paris Cedex 15, France

Abstract. This paper presents a modification of the previously described Rapid Rabies Enzyme Immuno-Diagnosis test (RREID) by using biotinylated antibodies, streptavidin conjugate and a mixture of monospecific polyclonal antibodies against several lyssaviruses. In the modified technique (RREIDlyssa), microplates were sensitized with a mixture of purified antibodies against ribonucleoprotein (RNP) from Pasteur virus (Lyssavirus serotype 1), European Bat Lyssavirus (EBL, unclassified) and Mokola virus (Lyssavirus serotype 3). Bound RNP was detected by the same antibodies labelled with biotin and peroxidase-strepavidin conjugate. These techniques were used for the detection of RNP of different Lyssavirus serotypes (rabies and rabies-related viruses). For lyssavirus specimens of serotype 1, the threshold of detection of RREID and RREID-lyssa were similar. However, a smaller amount of labelled antibodies was needed when biotinylated antibodies were used. For specimens infected by rabies-related strains (serotypes 2, 3, 4 and EBL), the threshold of detection of the RREIDlyssa was between two and 512 times lower than with the RREID. The sensitivity and the specificity of the RREID-lyssa for rabies virus (serotype 1) when tested on a small field trial (53 specimens) were found to be identical to the RREID. Consequently, RREID-lyssa can be a useful tool for diagnostic laboratories that receive specimens infected by rabies-related viruses.

Introduction The principal procedures for the routine diagnosis of rabies consists of immunofluorescent s t a i n i n g (fluorescent antibody test: FAT) 1 of rabies antigen in brain tissues, inoculation of the specimen into mice (mouse inoculation test: MIT), 2 or isolation of the virus in cell cultures (rabies tissue culture infection test: RTCIT)2 However, not all laboratories possess equipment for immunofluoresence, cell culture or animal inoculation. We have recently developed an enzyme-linked immunosorbent assay (Rapid Rabies Enzyme Immuno-Diagnosis: RREID) for routine diagnosis and for epidemiological studies of large numbers of specimens? -6 Based on the immunocapture of rabies ribonucleoprotein (RNP), this procedure is easy to carry out and does not require complex and costly equipment. Moreover, the results obtained with RREID correlate well with those obtained with RTCIT and FAT.7Nevertheless, RREID shows a low sensitivity for rabies-related viruses 5 despite common antigenic properties shared by their * Corresponding author. 1045-1056/92/010051 +08 $03.00/0

ribonucleoprotein, s-l° This problem could be solved by using conjugates prepared either from serum of hyperimmunized animals 1~or from monoclonal antibodies (MAb) directed a g a i n s t r a b i e s - r e l a t e d viruses.I° MAb are commonly used in epizootiological studies for the precise identification of field isolates but they cannot be prepared in all laboratories and the results vary with the experimental conditions and antibody affinity. Moreover, one of the major disadvantages of using MAb would be the possibility of not detecting a strain which is an antigenic variant. In contrast, polyclonal antibodies (PAb) are easier to obtain and less likely to give variable results. We have prepared PAb against two of the most antigenically-distant Lyssavirus serotypes (Serotype 1: PV strain and serotype 3: Mokola strain) and also the European Bat Lyssavirus type 1 EBL12 -15 All these viruses are responsible for infections of animals and humans. In this report we discuss results obtained for purified RNP and field specimens with the RREIDlyssa using PAb against PV-, EBL- or Mok-RNP in association with peroxidase or biotin-labelled antibodies. © 1992 The International Associationof BiologicalStandardization

52

P. Perrin et al.

Table 1. Comparison of the sensitivity of RREID, RREID-biot and RREID-lyssa in the diagnosis of rabies and rabies-related strains Ratio of threshold of detection Serotype

Virus

Country

Serotype 1

Rabies virus

Africa Cameroon Ethiopia Ivory Coast Madagascar Namibia Niger Morocco Chad

Species

RREID-biot]RREID

RREID-lyssa/RREID

Dog Hyena Dog Dog Kudu Dog Dog Dog

0-5 1 0-5 1 1 1 0.5 1

1 1 2 2 1 2 1 2

Bovine Unknown Dog Dog Human Cat Bat Skunk

0.5 1 1 1 1 0.5 0.5 1

1 0.5 1 1 2 1 1 1

Deer Wolf Human Fox Human

0.5 1 1 0.5 0-5

0.5 1 1 1 1

France France France Poland Yugoslavia

Sheep Cat Horse Raccoon Dog Bovine

1 2 0.5 1 2

1 1 0.5 1 2

Nigeria Central African Republic

Eidolon helvum Micropterus pusillus

0.25

2

0-5

64

Nigeria Cameroon Central African Republic Republic of South Africa

Crocidura sp. C. occidentalis Lophuromys sikapusi

0.25 0.25

64 512

1

256

Cat

1

128

Human

0.062

4

Miniopterus sp.

0.125

4

Eptesicus serotinus

0.25 0-5

8 2

America Brazil Colombia French Guyana Haiti Mexico U.S.A. U.S.A. U.S.A. Asia China Iran Malaya Saudi Arabia Thailand Europe

Serotype 2

Serotype 3

Lagos bat Lagos bat Mokola 1 Mokola 2 Mokola 3 Umlanga

Serotype 4

Unclassified

Duvenhage virus 1 Duvenhage virus 2

Republic of South Africa Republic of South Africa

EBL 1 EBL 2

France Finland

Human

All these specimens of different serotypes were obtained from the collection of the French National Reference Centre for Rabies. Brain specimens were homogenized as a 1:3 (vol/vol) suspension in Dulbecco modified Eagle medium adjusted to pH 7 with Na(HCO.~)e and serial two-fold dilutions were prepared. The cut-offOD value of the tests using peroxidase- (RREID) and biotinlabelled antibodies (RREID-biot and RREID-lyssa) was determined by adding 0.1 and 0-2 OD units respectively to the negative control OD values. The lowest concentration of brain s u p e r n a t a n t t h a t gave a positive reaction was determined for each of the three types of immunosorbent and conjugate. The results correspond to the ratio between the highest positive dilution obtained in RREID-biotin and RREID-lyssa and t h a t obtained in RREID.

Modified rapid rabies enzyme immunoassay

Materials and methods

Animals, cells and virus strains New Zealand rabbits (male, 3 kg: CEGAV, 61 350 Passais, France) were used for the production ofantiRNP sera. The PV strain (PV-Paris/BHK) '6 was propagated on BHK-21 cells. 17 Mokola 's and EBL 14 virus strains were propagated in a BSR cell line, a clone of BHK cells 2° and purified according to a modification of the technique described by Wiktor e t a l 2 ° The origin of the isolates used for the field evaluation of the RREID-lyssa is described in Tables 1 and 2. T a b l e 2. Distribution of the field specimens used for the evaluation of RREID-biot and RREID-lyssa in routine conditions Results in FAT, RREID, RREID-biot and RREID-lyssa Species

No. of positive

No. of negative

Total

Bovine Fox Sheep Dog Goat Cat Horse Badger Weasel Deer Rat

2 11 6 2 2 0 0 0 1 0 0

0 2 4 4 1 14 1 1 0 1 1

2 13 10 6 3 14 1 1 1 1 1

Total

24

29

53

Cell infection and ribonucleoprotein purification Cells were grown in monolayers with Dulbecco minimal essential medium (MEM) containing glutamine (300 pg/ml) and supplemented with 5% foetal calf serum (FCS) with 5% newborn bovine serum for BHK cells and with 8% FCS for BSR cells. Infection by the PV-Paris/BHK strain was carried out in suspension without serum by adding 0.1 PFU per cell. After incubation for 1 h at 34°C under gentle agitation, the cell concentration was adjusted to 2.7 x 106 cells/ml and the FCS concentration to 10%. Some 150 ml of the infected cell suspension were added to roller bottles (850 cm 2) and incubated at 37°C for 24 h. The cell culture medium was then replaced with fresh medium (70 ml/bottle) supplemented with 0.3% bovine serum albumin (BSA) instead of FCS. Infected cells were incubated at 34°C for 24 h.

53

Rabies-related EBL1- and Mok-RNP were produced in the same conditions as PV-RNP with the following modifications. BSR cells were infected in monolayers (in flatbottom 150-cm 2 flasks) at a multiplicity of infection of 0-3. After incubation for 1 h at 34°C, 26 ml of MEM and 4 ml of MEM supplemented with 3% BSA and adjusted to pH 7 with NaHCO3 were added. Infected cells were then incubated at 34°C for 3 days. The rabies RNP was extracted from cells and purified by modifications to the methods described by Compans e t al., 2~ Soko122 and Atanasiu e t a l . 23" 24 The infected cell s u p e r n a t a n t was discarded and cell monolayers were washed with PBS (pH = 7.3, containing 0-68 mM CaC12 and 0-4 mM MgSO4. Cells were then scraped, washed twice in 40 ml of NT buffer (0.13 M NaC1, 0-05 M Tris(hydroxymethyl)aminomethan, pH 7.6) and centrifuged (900 g, 10 min). All steps of RNP extraction and purification were carried out at 4°C. The cell pellet (about 5 ml) was suspended in 5 ml of distilled water and incubated for 1 h at 4°C. Cells were disrupted by eight strokes in a glass Dounce homogenizer and centrifuged (900 g, 10 min). The supernatant was collected and kept at 4°C until use. The pellet was dispersed in 5 ml of NT buffer. The cells were further disrupted in the homogenizer (six strokes) and cell fragments were sedimented by centrifugation (900 g, l0 min). The two supernatants from cell disruption were mixed together and centrifuged (9000g, 10 min) to sediment nuclei. Some 3 ml of supernatant were then distributed in SW50 polycarbonate tubes (Beckman) containing 2 g of CsCl. After dissolution of the CsC1, the volume was made up to 4.5 ml with NT buffer. The samples were centrifuged (40 000 rpm for 22 h in an SWb0 rotor at 4°C) and RNP, which formed a band in the middle part of the gradient, were collected. The RNP were purified by another centrifugation in CsCl: 1-8 ml of the RNP band was added to 3 ml of NT containing 1.3 g of CsC1 and centrifuged as above. The band corresponding to RNP was collected and exhaustively dialysed against NT buffer and stored at-20°C. Protein content was determined by the Lowry e t al. technique using BSA as the reference. 25

Production and purification of rabbit IgG anti-rabies RNP At 2-week intervals, rabbits received 1 ml intramuscular inoculations of an emulsion containing 400 pg of RNP and 500 pl of completeFreund's adjuvant. After three injections, rabbits were boosted at 1-month intervals with the same emulsion where

54

P. Perrin et al.

Freund's adjuvant was replaced by incomplete adjuvant. The rabbit IgGs were chromatographically purified on QAE-Sephadex according to Joustra and Longren. 26 Fractions with an optical density (OD) at 280 nm at least more t h a n 1 were pooled. Purified IgGs were concentrated by ammonium sulfate precipitation and sterilized by filtration through a 0.22 nm filter and stored at 4°C.

IgG conjugation with peroxidase The IgGs were conjugated with peroxidase according to the two-step conjugation technique described by Avrameas and Ternynck 27 with the modifications previously reported. 23, 24, 2s IgGs (5 mg) were conjugated with 10 mg of h o r s e r a d i s h peroxidase (Boehringer), dialysed against PBS, diluted in 50% (pH 7) buffered glycerol and stored at - 2 0 °.

IgG conjugation with biotin Biotin-IgG conjugate was prepared according to Ternynck 29with modifications as follows: 1 ml of IgGs (5 mg) were dialysed against 0.1 M carbonate buffer (pH 8.5). The IgG concentration was then adjusted to 2 mg/ml. Then, 75 pl of D-biotin-N-hydroxy-succinimidester (BNHS: Boehringer) solution (1.5 mg dissolved in 300 pl of anhydrous dimethyl-formamide) were added to 2.5 ml of IgGs. The sample was mixed and agitated at room temperature for 1 h. Unbound BNHS was eliminated by gel filtration (G-25 Sephadex-Pharmacia-equilibrated with PBS) and 2 ml fractions were collected. Protein content was determined by measuring OD at 280 nm. Fractions containing biotin-IgG conjugate (OD > 0.8) were pooled. BSA (final concentration of 1 mg/ml) and glycerol (an equal volume) were added. The sample was stored at 4°C.

Streptavidin conjugation with peroxidase Streptavidin (Sigma) was conjugated with peroxidase according to Nakane et al. 3° with the following modifications: 12.5 mg of peroxidase (Boehringer) were dissolved in 375 pl of 0.36 N HEPES (Sigma) and 300 pl of 0.1 M sodium periodate (Sigma) were added. After incubation for 30 min in the dark at room temperature, 75 pl of diluted ethylene-glycol (1:21 in water) were added. Unbound products were eliminated by gel filtration (G-25 Sephadex -Pharmaciaequilibrated with 0.01M carbonate buffer, pH 9.5). The brown-coloured fractions (2 ml) containing activated peroxidase were pooled and mixed with streptavidin (2.5 mg of streptavidin dissolved in 400 pl of PBS: 0.01 M phosphate buffer (pH 7.2) con-

taining 0.13 M NaC1). 200 pl of 0.1 M carbonate buffer (pH 9.5) were then added and the mixture was incubated for 3 h in the dark at room temperature and agitated every 30 min. Some 50 pl of borohydride (Sigma) solution (4 mg/ml in water) was added and the mixture was incubated for 3 h in the dark at 4°C and then dialysed against PBS. The streptavidin-peroxidase conjugate was diluted with an equal volume of glycerol and stored at -20°C until use.

Purified RNP titration Purified RNP was titrated with peroxidase 4.5 or biotin-labelled IgGs. When biotin-labelled IgGs were used, their detection was modified as follows: the appropriate dilution of peroxidase-streptavidin conjugate was added (200 pl/well) and microplate was incubated for 30 min at 37°C. Samples were washed, peroxidase activity was revealed with a chromogenesubstrate mixture (o-phenylenediamine-hydrogen peroxide) and then assayed by measuring the OD at 492 nm. 4 Immunodiagnosis of field specimens The evaluation of the immunodiagnosis of field isolates was carried out with three types ofimmunosorbents and two types of labelled IgGs: (i) RREID: IgGs anti-PV-RNP as the immunosorbent (3.5 pg/ml for coating) and the same IgGs conjugated with peroxidase (final concentration of conjugate 5-0 pg/ml); (ii) RREID-biot: IgG anti-PV-RNP as the immunosorbent (3.5 pg/ml for coating) and the same IgGs labelled with biotin (final concentration: 0.8 pg/ml); did RREID-lyssa: a mixture of IgGs anti-PV-, EBLand Mok-RNP as the immunosorbent (3-5 pg/ml for each) and a mixture of the same IgGs labelled with biotin (final concentration: 0-8, 0.15 and 0.15 pg/ml for antibodies against PV-, EBL- and Mok-RNP respectively). Brain specimens were homogenized and serial twofold dilutions were prepared as previously described. 7 The cut-off OD value of the tests using peroxidase and biotin labelled antibodies was obtained by adding 0.1 and 0.2 OD units respectively, to the negative control OD values. RREID, RREID-biot and RREIDlyssa were performed on 36 rabies isolates belonging to the five serotypes oflyssavirus (Table 1). The highest dilutions of brain supernatant t h a t gave a positive r e a c t i o n in each of the t h r e e types of immunosorbent and labelled IgGs were compared. The properties of RREID, RREID-biot and RREIDlyssa were also evaluated using brain supernatants from 53 specimens received for rabies diagnosis at the National Reference Centre for Rabies (Institut

Modified rapid rabies enzyme immunoassay

o

b 3"0,

3"0

2-4 2.2

Y~ >.

o

4"0 3-8

2-8 2-61 2.41~ 2.2~2.0~-

2.8 2"6

2.0 1-8

i-Sl

1.6 1o4

~.61 ].41

1"2

1.21

I'0

i'Ol

3"6 3.4 3-2 3'0 2-8 2-6 2"4 2.2 2-0

0-6

0-2 0-0 0'0

_~ 0.,

- - S ~ I

I0

..... I00

I000

1"6 1.4 1'2 I-0 I 0"8 0'6 E 0"4 0"2 f 1"8 I

//

0"8

0"4

55

0"6 0 . 0"2 0"0 0.01

4 0-I

~ I

I0

I00

I000

0.0

==--_

0-01 0"I

......

I

I0

I00

--I 0 0 0

RNP concentration (ng/ml)

Figure 1. Titration of purified RNP prepared from PV-, EBL- and Mok- viruses with RREID using three monospecific antibodies and peroxidase-labelled antibodies. ELISA microplates were coated with 0-7 pg/well of purified antibodies against PV- (a), EBL- (b) and Mok-RNP (c). Bound RNP was revealed with 0.25, 0.12 and 0.12 pg/well of peroxidase-labelled antibodies against PV- (a), EBL- (b) and Mok-RNP (c) respectively. • = P V , • = M o k , • =EBL. Pasteur, Paris, France). The characteristics of the specimens are described in Table 2. Results

PV-, EBL- and Mok-RNP were used to test three types of antibodies as immunosorbent and peroxidase conjugates. When antibodies against RNP from PV were used, the lowest concentrations detected were about 2 ng/ml of PV-RNP and 10 and 40 ng/ml of EBL and Mok respectively (Fig. 1 (a)). The test was repeated using antibodies against RNP from EBL, and about 1 ng/ml of EBL-RNP and 5 ng/ml of PVand Mok-RNP were the lowest concentrations detected (Fig. l(b)). With antibodies against RNP from Mok, 1.0, 5.0 and 10.0 ng/ml of Mok-, EBL- and PV-RNP respectively, were detected (Fig. 1(c)). Antibodies against EBL-RNP (Fig. l(b)) allowed a better detection ofheterologous RNP than antibodies against PV- or Mok-RNP (Fig l(a), 1(c)). None of the monospecific immunosorbents and peroxidase conjugates allowed the detection of RNP from all three serotypes with a high sensitivity (Fig. 1). Anti-RNP IgGs were therefore labelled with biotin instead of peroxidase with a view to increasing the sensitivity. The titration of purified PV-RNP using antibodies against PV-RNP and the two types of conjugates is reported in Fig. 2. An optical density of 2

was obtained in the following conditions: (i) 20 ng/ml of RNP with 2.5 pg/ml of peroxidase-labelled IgGs; 40 ng/ml of RNP with 1.25 pg/ml of peroxidase-labelled IgG; (iii) 5 ng/ml of RNP with 0.2 pg/ml of biotinlabelled IgGs; (iv) 10 ng/ml of RNP with 0.1 pg/ml of biotin-labelled IgG. Seventy-five-fold more IgGs labelled with peroxidase than with IgGs labelled with biotin were required to obtain the same OD with a given PV-RNP concentration (not shown). It is clear that the biotin and peroxidase-labelled streptavidin complex had enhanced the threshold of detection of the RNP. To determine whether the use ofbiotinylated antibodies allowed the detection ofPV, EBL or Mok antigens, antibodies against each of the three serotypes were tested in homologous system. When using antibodies against PV-RNP, it was not possible to detect EBL- or Mok-RNP (Fig. 3(a)). In contrast, antibodies against EBL-RNP (Fig. 3(b)) detected about 2 ng/ml of PV- and Mok-RNP, whereas antibodies against Mok-RNP detected 5 ng/ml of PV- and EBL-RNP (Fig. 3(c)). As with IgG labelled with peroxidase (Fig. 1), the assay with antibodies against EBL-RNP was better for the detection of heterologous RNP. On the other hand, 0.2 ng/ml of homologous RNP could be detected by each antibody. A mixture of the three monospecific IgGs was also tested as immunosorbent and biotinylated conjugate

P. P e r r i n et aL

56

(RREID-lyssa). The results obtained for the titration of PV-, EBL and Mok-RNP in this assay are reported in Fig. 4. About 0.5 ng/ml of each of the three purified RNPs were detected and each serotype could be correctly titrated: there was no major difference between the three curves. This type of assay presents a threshold of detection as low as the monospecific and homologous assays (Fig. 3). RREID, RREID-biot and RREID-lyssa were tested with 36 rabies isolates representative of the diversity of the Lyssavirus genus (Table 1). First, we tested 26 strains belonging to the serotype 1 of the Lyssavirus genus and isolated in Africa, America, Asia and Europe. The results obtained with the three methods were nearly identical (Table l) but only one sixth the amount of labelled antibodies was needed with RREID-biot and RREID-lyssa t h a n with RREID. Secondly, we tested 10 rabies-related isolates obtained mainly from Africa (except EBL) (Table 1). RREID-lyssa detected between two and 64 times more antigen than RREID for isolates of serotype 2, 4 and EBL. The largest difference was evidenced with strains of serotype 3: RREID-lyssa detected between 64 and 512 times more antigen than the other methods (Table 1). The threshold of detection of RREIDbiot was lower than that of RREID for the detection of rabies-related viruses. This was probably due to

3.0 2.8 2.6 2.4 E

2.2 2.0 1-8

."z 1-6 ceu 1.4 -o 1.2 u I-0

II

0.8

0"6 0.4 0.2 0"0 0.01

0.I

I0

I

Concentrotion

I00

of P V - R N P

I000

(ncj/ml)

Figure 2. Comparison between peroxidase- and biotinlabelled antibodies for the detection of PV-RNP by RREID. Microplates were coated with 0.7 pg/well of purified antibodies against PV-RNP and bound RNP was revealed by: biotin-labelled antibodies against PV-RNP (0.1 pg/well (Q); 0.2 pg/well ( • ) or peroxidase-labelled antibodies against PV-RNP (1.25 pg/well (O); 2.5 pg/well (D). Biotin was then detected with peroxidase-labelled streptavidin.

4"0 3"8 3"6 3-4 3"2 3'0 "~ 2-8 c c~

/

2-6

o~ 2'4 ~ 2'2 2" 0 •~

~.s: 1.6 o .u_ I- 4 o 1.2 1.0 O-S 0-6 0"4 0'2 0-0 0'01 _~

0"I

I

I0

I00 I000

4.0 3.8 3,6 3"4 3"2 :3.0 2"8 2"6 2-4 2-2 2"0 1'8 1"6 1"4 1"2 I'0 0.8 0"6 0.4 0"2 O.C 0.01

~

! ~

:==--0-I

" I

I0

RNP concenlrotion

....... I00

I000

4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2-4 2-2 2-0 1'8 1.6 1"4 1"2 1.0 0"8 0"6 0-4 0.2 0.0 0-01

-"--~:'=;.--0-I

I

" I0

.......... I00

I000

(ng/ml)

Figure 3. Titration of purified RNP prepared from PV-, EBL- and Mok- viruses with RREID using three monospecific antibodies and biotin-labelled antibodies. ELISA microplates were coated with 0.7 pg/well of purified antibodies against PV- (a), EBL- (b) and Mok-RNP (c). Bound RNP was revealed with 0-15, 0-05 and 0-05 pg/well of biotin-labelled IgGs against PV- (a), EBL- (b) and Mok-RNP (c) respectively. Biotin was then detected with peroxidase-labelled streptavidin. • = PV, • = Mok, • = EBL.

Modified rapid rabies enzyme immunoassay

4"0 3"8 3"6 3"4 3-2 3-0 2"8 2"6 2"4. 2-2 2"0 ® 1"8 1"6 3 1"4 8 1"2 1"0 0-8 0"6 0"4. 0-2 ; -.. -- _ _ ~ 0"0 0'01 0"1 I

l l .................. I0 I 0 0 1000

RNP concenlrotion (nq/ml)

Figure 4. Titration of purified RNP prepared from PV-, EBL- and Mok-viruses with RREID using a mixture of the three monospecific antibodies for coating and RNP revelation. Microplates were coated with 0-7 pg/well of each purified antibody against PV-, EBL- and Mok-RNP. Bound RNP was revealed with 0-16, 0.03 and 0.03 pg/well of a mixture of biotin-labelled antibodies against PV-, EBL- and Mok-RNP. Biotin was then detected with peroxidase labelled streptavidin. • = PV, • = Mok, • = EBL.

the lower concentration of labelled antibodies used in RREID-biot. RREID and RREID-lyssa were then compared on 53 field specimens from France received for rabies diagnosis (Table 2). There was a perfect concordance between the results obtained with RREID and RREID-lyssa and those obtained by FAT and RTCIT. Discussion

Replacement of immunofluorescence by immunoperoxidase for rabies diagnosis was suggested many years ago for laboratories which do not possess a fluorescence microscope. Several techniques have been proposed for the detection of rabies antigens in tissue impressions 23.24.2s or in ground tissues. 4Today, among the techniques using peroxidase, only the detection of rabies antigens in ground tissues by an enzyme immunoassay is routinely carried out. 4,6,~The Rapid Rabies Enzyme Immuno-Diagnosis (RREID) can be recommended for epidemiological studies, for l a b o r a t o r i e s w h i c h are not e q u i p p e d for immunofluorescence and as a back-up procedure to confirm FAT results. RREID has been shown to be slightly less sensitive t h a n FAT but as sensitive and

57

specific as RTCIT. 7 It allows the detection of antigens from all serotypes (about 2 ng/ml of RNP of serotype 1) but with a lower sensitivity for antigens from rabies-related viruses, s To enhance its sensitivity and to extend its specificity we have prepared biotinylated antibodies and polyclonal antibodies against RNP from the Lyssavirus serotypes 1, 3 and EBL. For the same concentration of PV-RNP, 25-75 times less biotin- t h a n peroxidase-labelled antibodies were needed to obtain the same optical density with the immunoassay. In these conditions, the minimum PV-RNP concentration detected was lowered to 0.1-0.2 ng/ml. When using monospecific antibodies and biotin-labelled IgGs, 0.2 ng/ml of each RNP from the three serotypes could be detected in a homologous assay. Despite the partial common antigenicity of rabies and rabies-related viruses at the RNP level, s.9no monospecific antibody allowed the detection of RNP from the other serotypes with a comparable threshold of detection. However, when using antibodies against EBL-RNP as immunosorbent and biotin-labelled IgGs, about 2 ng/ml of heterologous RNP (PV or Mok) can be detected. This result was comparable to t h a t obtained with RREID for the homologous assay (PV antigens). If a single monospecific immunosorbent and labelled IgGs must be prepared, antibodies against EBL-RNP which seem to better detect antigens from the other two serotypes should be used. The detection of the RNP from the three serotypes could be improved by using a mixture of the antibodies for the plate coating and by the revelation of bound antigens by biotin labelled IgGs (RREID-lyssa). With a mixture of antibodies and labelled IgGs against PV-, EBL- and Mok-RNP, about 0.2 ng/ml of RNP from each serotype could be detected. When compared with the RREID, the threshold of detection of the RREID-lyssa was lower for the purified RNP from all of the serotypes 1, 3 or EBL. This was confirmed by the comparison of both tests on 36 strains chosen to be as representative as possible of the antigenic variability of the lyssaviruses. The threshold of detection of RREIDlyssa was again lower than t h a t of RREID, especially for rabies-related strains, and despite the six-fold less amount of labelled antibody used. In view of this increased level of detection, a decrease of the specificity could have been suspected when used in field conditions. This was not confirmed in tests with 53 field specimens of various species received for rabies diagnosis. On the contrary, RREID-lyssa was in perfect agreement with FAT. RREID-lyssa has still to be further evaluated in field conditions but can be recommended for laboratories t h a t are faced with the

58

P. Perrin et al.

diagnosis of r a b i e s - r e l a t e d v i r u s e s or w a n t to i n c r e a s e the s e n s i t i v i t y of R R E I D for r e s e a r c h purposes.

Acknow/eclgements T h e a u t h o r s w a n t to t h a n k D i a g n o s t i c s P a s t e u r (92 430 M a r n e s - l a - C o q u e t t e , F r a n c e ) for t h e i r financial aid a n d D r J. F. D e l a g n e a u for his helpful suggestions. References

1. Dean D, Abelseth K. The fluorescent antibody test. In: Kaplan M, Koprowski H, eds. Laboratory Techniques in Rabies, Geneva: WHO, 1973: 73-84. 2. Webster L, Dawson J. Early diagnosis of rabies by mouse inoculation. Measurement of humoral immunity to rabies by mouse protection test. Proc Soc Exp Biol Med 1935; 32: 570-573. 3. Sureau P. Les techniques rapides de diagnostic de laboratoire de la rage. Arch Inst Pasteur Tunis 1986; 63: 183-197. 4. Perrin P, Rollin P, Sureau P. A rapid rabies enzyme immunodiagnosis (RREID): a useful and simple technique for the routine diagnosis of rabies. J Biol Standard 1986; 14: 217-222. 5. Perrin P, Sureau P. A collaborative study of an experimental kit for rabies enzyme immunodiagnosis. Bull WHO 1987; 65: 489-493. 6. Perrin P, Sureau P. A new immunoenzymatic test for the rapid laboratory diagnosis of rabies on field specimens (RREID). In: Thongchaoren P, Kurstak E, eds. Virus Diseases in Asia, Bangkok: Mahidol University, 1988: 345-348. 7. Bourhy H, Rollin P, Vincent J, Sureau P. Comparative field evaluation of the fluorescent-antibody test, virus isolation from tissue culture, and enzyme immunodiagnosis for rapid laboratory diagnosis of rabies. J Clin Microbiol 1989; 27: 519-523. 8. Schneider L, Dietzschold B, Dierks R, Mathaeus W, Enzman P, Strohmaier K. Rabies group-specific ribonucleoprotein antigen and test system for grouping and typing rhabdoviruses. J Virol 1973; 11: 748-755. 9. Flamand A, Wiktor T, Koprowski H. Use ofhybrydoma monoclonal antibodies in the detection of antigenic difference between rabies and rabies-related virus proteins. I. The nucleocapsid protein. J Gen Virol 1980; 48: 97-104. 10. Wiktor T, Flamand A, Koprowski H. Use ofmonoclonal antibodies in diagnosis or rabies virus infection and differentiation of rabies-related virus. J Virol Meth 1980; 1: 33-46. 11. Wiktor T, MacFarlan R, Foggin C, Koprowski H. Antigenic analysis of rabies and Mokola virus from Zimbabwe using monoclonal antibodies. Dev Biol Stand 1984; 57: 199-211. 12. Bussereau F, Vincent J, Coudrier D, S u r e a u P. Monoclonal antibodies to Mokola virus for identification of rabies and rabies-related viruses. J Clin Microbiol 1988; 26: 2488-2494. 13. Vincent J, Bussereau F, Sureau P. Immunological relationships between rabies virus and rabies-related virus studied with monoclonal antibodies to Mokola virus. Ann Inst Pasteur/Virol 1988; 139: 157-173.

14. Montano Hirose JA, Bourhy H, Lafon L. A reduced panel of antinucleocapsid monoclonal antibodies for bat rabies virus identification in Europe. Res Virol 1990; 141: 571-581. 15. WHO report. Report of the sixth WHO consultation on monoclonal antibodies in rabies diagnosis and research. 1990; WHO/Rab. Res./90.34. 16. Perrin P, Portnoi D, Sureau P. Etude de l'adsorption et de la p4n~tration du virus rabique: interactions avec les cellules BHK-21 et des membranes artificielles. Ann Virol (Inst Pasteur) 1982; 133E: 403-422. 17. MacPherson I, Stocker M. Polyoma transformation of hamster cell clones--an investigation of genetic factors affecting cell competence. Viroi 1962; 16: 147-151. 18. Bourhy H, Tordo N, Lafon M, Sureau P. Complete cloning and molecular organization of a rabies-related virus, Mokola virus. J Gen Virol 1989; 70: 2063-2074. 19. Sato M, Maeda N, Yoshida H, Urade M, Saito S, Miyazaki T, Shibata T, Watanabe M. Plaque formation of herpes virus hominis type 2 and rubella virus in variants isolated from the colonies of BHK-21/WI-2 cells formed in soft agar. Arch Virol 1977; 53: 269-273. 20. Wiktor T, Dietzschold B, Leamnson RN. Koprowski H. Induction and biological properties of defective interfering particles of rabies virus. J Virol 1977; 21: 626-635. 21. Compans R, Choppin. The length of the helical nucleocapsid of Newcastle disease virus. Virol 1967; 33: 344-346. 22. Sokol, F. Purification of rabies virus and isolation of its components. In: Kaplan, N, Koprowski H, eds. Laboratory Techniques in Rabies. Geneva: WHO, 1973: 165-178. 23. Atanasiu P, Perrin P, Favre S, Chevallier G, Tsiang H. Immunofluorescence and immunoperoxidase in the diagnosis of rabies. In: Kurstak E, Morisset R, eds. Viral immunodiagnosis, New York: Academic Press, 1974: 141-155. 24. Atanasiu P, Nozaki J, Sisman J, Perrin P. Identification des glycoprot~ine et nucl~oprot~ine rabiques des cellules infect~es par des immunoglobulines sp~cifiques marquees. CR Acad Sci Paris 1975; 281D: 1913-1915. 25. Lowry D, Rosenbrough N, Farr N, Randall R. Protein measurement with Folin phenol reagent. J Biol Chem 1951; 195: 265. 26. Joustra M, Lundgren H. Preparation of freeze-dried, monomeric and immunochemically pure IgG by a rapid and reproducible chromatographic technique. In: Protides of Biological Fluids 1969; 17: 511-515. 27. Avrameas S, Ternynck T. Peroxidase labelled antibody and Fab conjugates. Immunochem 1971; 8: 1175. 28. Atanasiu P, Tsiang H, Perrin P, Favre S. Les IgG antiglycoprot~ines rabiques marquees par la peroxydase et l'isothiocyanate de fluoresc~ine. R~sultats Ann Microbiol (Inst. Pasteur) 1975; 126B: 69-75. 29. Ternynck T, Avrameas S. Techniques Immuno-enzymatiques. Techniques en Immunologie. Les Editions INSERM 1987: 87-90. 30. Nakane P, Kawaoi A. Peroxidase-labelled antibody, a new method of conjugation. J Histochem Cytochem 1974; 22: 1084-1091.

Received for publication 20 November 1991; accepted 8 January 1991.