Competitive Luminex immunoassays for detection of antibodies to foot-and-mouth disease and vesicular stomatitis viruses in multiple susceptible hosts

Abstract

Foot-and-mouth disease (FMD) and vesicular stomatitis (VS) cause such similar clinical signs and lesions that laboratory tests are required to distinguish between infections caused by each virus. Using mouse anti-foot-and-mouth disease virus (FMDV) 3B monoclonal or polyclonal anti-vesicular stomatitis virus-New Jersey (VSV-NJ) antibodies and recombinant FMDV 3ABC or VSV-NJ glycoprotein (G) antigens coated to MagPlex beads, competitive Luminex immunoassays (cLIAs) were developed for FMDV and VSV-NJ, respectively. The cLIAs successfully detected antibodies to FMDV 3ABC and VSV-NJ G in sera from infected animals. The diagnostic sensitivity and specificity were 93% and 98%, respectively for FMDV and 93% and 95.4%, respectively for VSV-NJ. These cLIAs are potential alternatives for competitive enzyme-linked immunosorbent assays (cELISAs) and provide the opportunity for multiplexing to reduce time and the amount of serum required for testing.

Foot-and-mouth disease (FMD) caused by FMD virus (FMDV) affects cloven-hoofed animals. This virus belongs to the family Picornaviridae, genus Aphthovirus. The 7 serotypes O, A, C, Asia 1, and Southern African Territories (SAT) 1, 2, and 3 are immunologically distinct. Foot-and-mouth disease virus (FMDV) causes vesicles in the epithelium of the mouth, feet, and mammary gland that eventually rupture, resulting in erosions. The open wounds on the feet with the concomitant irritation lead to lameness. Profuse salivation is also common, with the saliva, vesicular fluid, and vesicular epithelium all being highly infectious (1).

Vesicular stomatitis (VS) affects horses, mules, cattle, and swine and is caused by either vesicular stomatitis virus New Jersey (VSV-NJ) or VSV Indiana (VSV-IN), which belong to the family Rhabdoviridae, genus Vesiculovirus. Both viruses are antigenically distinct serotypes and exist in the Americas. The clinical signs and lesions seen in cattle and swine are identical to those of FMD and other vesicular diseases (2,3). Vesicular stomatitis virus (VSV) has also been detected in wildlife.

As FMD, VS, and other vesicular diseases produce similar clinical signs, laboratory tests are required for a definitive diagnosis. Sensitive assays exist for detecting viral genomes, antigen, and live virus in clinical material from acutely infected animals. Antibodies to FMDV or VSV in sera from convalescent animals are also detectable by available serological assays (4,5). Recombinant FMDV 3ABC protein-based competitive enzyme-linked immunosorbent assay (cELISA) is used to detect antibodies to FMDV nonstructural protein (NSP) regardless of infecting serotype (4). A competitive ELISA is also available for detecting antibodies to VSV-NJ and VSV-IN (5). The reagents in these cELISAs can be adapted for use in competitive Luminex immunoassays (cLIAs), which use less reagents and samples and are potentially cheaper, more sensitive, and more amenable to multiplexing than cELISAs.

The aim of this study was to use these readily available cELISA reagents for FMDV and VS-NJ to develop cLIAs for detecting antibodies to these viruses in bovine, porcine, and ovine sera.

The FMDV O1 Campos 3ABC gene sequence (GenBank accession number: AJ320488.1) was codon-optimized for expression in insect cells, cloned into a pAB-bee-FH transfer vector containing 8x His tag by NotI and XbaI (GenScript Biotech, Piscataway, New Jersey, USA) and co-transfected into Spodoptera frugiperda (Sf-9) cells with linearized baculovirus deoxyribonucleic acid (DNA) according to the manufacturer’s protocols (AB Vector, San Diego, California, USA). The VSV-NJ glycoprotein (G) gene was amplified with primers based on VSV-NJ G gene sequence from a patient [49-UT-B1 (Ogden), GenBank accession number: M21416.1.], cloned in-frame into Not1 and EcoR1 restriction enzyme sites of pAB-bee-FH transfer vector (AB Vector), and then co-transfected into Sf-9 cells with linearized baculovirus DNA according to the manufacturer’s protocols (AB Vector). Amplification of baculovirus, infection of Sf-9 cells for protein expression, and recovery of recombinant proteins from cell pellets were conducted as described in a previous study (6) and purified by batch procedure using Ni-NTA agarose according to manufacturer’s protocol (Qiagen, Maryland, USA). Protein expression was confirmed by Western blot using anti-His tag antibody, showing recombinant proteins of the expected size for FMDV 3ABC (24 kD) and VSV-NJ G (57 kD). After Ni-NTA agarose purification, unwanted proteins were eliminated, which resulted in purified FMDV 3ABC and VSV-NJ G recombinant proteins. The antigenicity of each protein was confirmed by indirect ELISA using strong positive and negative control sera.

The predetermined optimal amount for each recombinant protein (12 μg for FMDV 3ABC and 36 μg for VSV-NJ G) was coupled to 1.25 × 10⁶ MagPlex Luminex magnetic beads according to the manufacturer’s instructions (Luminex, Austin, Texas, USA). The beads are carboxylated polystyrene microparticles with spectrally distinct dyes or “regions” identified by numbers so they can be individually recognized by the Luminex machine. Regions #044 and #046 were used for FMDV 3ABC and VSV-NJ G, respectively. The antigens were coupled to beads in 2 steps, starting with activation of carboxyl groups on the beads with 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) in the presence of sulfo-NHS (N-hydroxysulfosuccinimide) to form a sulfo-NHS-ester intermediate. This sulfo-NHS-ester intermediate reacted with the primary amine of the recombinant antigen to form a covalent amide bond.

Both the production and characterization of the mouse anti-FMDV 3B monoclonal antibody used for this cLIA and the production of the mouse polyclonal anti-VSV-NJ antibodies in ascites fluid have been described in previous studies (4,5).

Positive samples were obtained from cattle (25 sera), sheep (19 sera), and pigs (28 sera) experimentally infected with FMDV (72 sera in total), as described in previous studies (4,7–9). Twenty-one bovine and 65 porcine sera positive for antibodies to VSV-NJ were obtained from animals experimentally infected with VSV-NJ (86 sera in total). For evaluation of analytical specificity, sera positive for antibodies to VSV-IN (cattle and pigs) and swine vesicular disease virus (SVSD) (pigs only) were similarly obtained from experimentally infected animals (10,11). Negative sera were obtained from clinically healthy cattle, pigs, and sheep in Canada, which is recognized as being free from FMD and VS.

The antigen-coupled beads were thoroughly mixed and diluted at 40 000 beads/milliliters in blocking buffer (StabilGuard; Surmodics, Eden Prairie, Minnesota, USA) and 50 μL (2000 beads) dispensed per well of 96-well Luminex plate. Mouse anti-FMDV 3B Mab or polyclonal mouse anti-VSV-NJ antibody at optimal dilution in StabilGuard were added to predetermined wells at 25 μL/well (except the blank wells), followed immediately by the addition of 25 μL of 1/4 dilution in StabilGuard of heat-inactivated positive or negative bovine, porcine, or ovine serum to predetermined wells or diluent only to diluent control (DC) wells to obtain a final volume of 100 μL/well. The plates were protected from light and incubated at room temperature with shaking for 1 h, followed by 3 washes with phosphate-buffered saline (PBS, pH 7.2) using a preprogrammed BioRad plate washer with a magnet placed under the plate to retain beads in the wells. Goat anti-mouse IgG H&L Phycoerythrin (Abcam, Toronto, Ontario) diluted 1/200 in StabilGuard was added to each well at 50 μL/well. The plate was incubated at room temperature in the dark for 30 min with shaking and washed as previously described. Beads were then resuspended in 125 μL of PBS, analyzed on the MAGPIX System (Luminex), and data for each sample were expressed as median fluorescence intensity (MFI). Percent inhibition (PI) of the MFI in test (T) wells relative to the diluent control (DC) was calculated as (1-MFI T/MFI DC) × 100. A total of 737 negative sera (cattle, 176; sheep, 152; and pigs, 409) and 72 positive sera (cattle, 25; sheep, 19; and pigs, 28) for FMDV were tested; and 348 negative sera (cattle, 88 and pigs, 260) and 86 positive sera (cattle, 21 and pigs, 65) for VSV were tested.

Competitive ELISAs were conducted as described in previous studies for FMDV 3B (4) and for VSV-NJ (5).

The box plots of negative and positive sera were plotted in Version 2017.1 of XLSTAT (Microsoft, Redmond, Washington, USA) (Figure 1A, B). The cutoff determined by receiver operating characteristics (ROC) curve analysis in XLSTAT with data from 737 negative sera and 72 positive sera was ≥ 58 PI for samples considered positive for anti-FMDV 3B antibodies. Diagnostic sensitivity and specificity for FMDV cLIA were 93% and 98%, respectively. Similarly, the cutoff for VSV-NJ cLIA was determined by ROC curve analysis in XLSTAT with data from 348 negative sera and 86 positive sera as ≥ 21 PI for samples considered positive for antibodies to VSV-NJ. The diagnostic sensitivity and specificity for VSV-NJ cLIA were 93% and 95.4%, respectively.

Figure 1.

Box plots of the distribution of A — negative sera (n = 737) and positive sera (n = 72) tested for antibodies to foot-and-mouth virus 3B on the competitive Luminex immunoassay (cLIA); B — negative sera (n = 348) and positive sera (n = 86) tested for antibodies to vesicular stomatitis virus-New Jersey serotype glycoprotein on the cLIA.

Analytical specificity determined by testing sera positive for antibodies to swine vesicular disease virus (SVDV) and vesicular stomatitis virus Indiana and New Jersey (VSV-IN and VSV-NJ) was 100% (no cross-reactivity) for FMDV 3ABC cLIA. Similarly, no cross-reactivity was detected when sera positive for antibodies to FMDV, SVDV, and VSV-IN were tested on the VSV-NJ cLIA (100% analytical specificity). Time to first detection of positive antibody response in animals experimentally infected with FMDV was days post-infection (DPI) 5 to 7 for cattle, DPI 7 to 9 for sheep, and variable for pigs (Table I). Similarly, time to first detection of positive antibody response in animals experimentally infected with VSV was DPI 9 for cattle and DPI 6 to 14 for pigs (Table II).

Table I.

Detection of antibodies to foot-and-mouth virus 3B in serial bleeds from experimentally infected cattle, sheep, and pigs

	DPI	0	3	4	5	7	14	21	28
Cattle ID
cLIA (cutoff = 58)	c236	24	14	23	15	82	98	94	97
	c237	23	15	3	8	61	97		91
	c141	24	27	25	41	76	99	99	97
	c138	49	36	45	63	95	99	98	95
	c146	17	16	23	30	75	99	99	99
	c134	36	18	41	44	64	99	99	97
	c136	49	47	37	44	61	95	94	80
	c242	25	50	16	73	73	97	99	97
cELISA (cutoff = 50)	c236	0	0	0	4	90	99	97	97
	c237	0	0	0	0	79	98		100
	c141	0	0	0	0	67	103	100	100
	c138	13	16	25	48	98	99	101	99
	c146	0	0	0	0	77	100	100	99
	c134	0	0	0	0	22	101	100	101
	c136	0	23	15	14	60	101	100	100
	c242	5	12	8	5	55	94	97	94
	DPI	0	1	2	3	4	7	9	13	19
Sheep ID
cLIA (cutoff = 58)	s991	12	31	29	24	35	38	66	71	61
	s268	33	29	23	39	38	65	98	98	96
	s804	28	22	22	15	0	43	86	96
	s68	48	56	50	46	45	87	87	85	89
	s205	20	23	28	33	11	24	68	88	86
cELISA (cutoff = 50)	s991	0	22	30	25	24	38	103	104	104
	S268	3	3	2	0	3	93	100	100	100
	s804	11	8	0	0	0	89	103	104
	S68	0	0	0	0	0	91	92	90	91
	s205	2	3	2	2	0	6	94	100	98
	DPI	0	2	3	4	5	7	9	10	18	20
Pigs ID
cLIA (cutoff = 58)	p42	22	29	49	26	31	50	97	99
	p3	28	26	14	27	83	95
	p6	22	14	19	22	31	45	72	59
	p71	45	45	42	34	32	36	37	71
	p13	25	34	25	24	27	41	31	95	93
	p45	26	20	22	25	33	21	16	24	91	84
	p20	14	14	15	27	26	23	24	24	83	94
	p7		55	37	34	31	78	93
cELISA (cut off = 50)	p42	0	0	0	0	0	5	79	98
	p3	0	0	0	0	0	64
	p6	0	0	0	0	0	0	15	17
	p71	0	1	3	1	0	1	18	9
	p13	1	0	0	0	0	3	0	3	95
	p45	0	0	0	0	0	0	0	0	6	87
	p20	0	0	0	0	0	0	0	0	0	0
	p7	0	0	0	0	0	42	87

cLIA — competitive Luminex immunoassay; cELISA — competitive enzyme-linked immunosorbent assay; DPI — days post-infection.

Green cells represent positive samples.

The numbers represent percent inhibition of median fluorescence intensity (MFI) of the diluent control for cLIA or optical density (OD) for diluent control for cELISA.

c, s, and p in front of animal identity (ID) stand for cattle, sheep, and pig, respectively.

Table II.

Detection of antibodies to vesicular stomatitis virus-New Jersey serotype in serial bleeds from experimentally infected cattle and pigs.

	DPI	0	1	2	3	4	5	7	9	11	14	21	28
Cattle ID
cLIA (cutoff = 21)	c342	0	0	0	0	0	0	0	11	40	49	57	45
	c343	0	0	0	0	0	0	0	58	62	53	65	49
	c854	0	0	0	0	0	0	0	45	67	66
	c855	0	0	0	0	0	0	0	26	44	57
cELISA (cutoff = 40)	c342	0	0	0	0	0	16	53	80	88	91	93	92
	c343	0	0	0	0	0	6	26	50	58	59	76	80
	c854	0	7	8	3	3	20	59	86	89	94
	c855	0	0	9	0	0	8	60	78	86	91
	DPI	0	1	2	3	4	5	6	7	14	21	28
Pigs ID
cLIA (cutoff = 21)	p41	0	0	0	0	0	0	3	0	28	55	60
	p42	0	0	14	0	0	0	0	0	70	68	68
	p43	0	0	0	0	0	0	0	17	42	64	57
	p44	0	0	0	0	0	0	45	5	65
	p45	0	0	0	0	0	0	0	0	3	8
	p46	2	0	0	0	0	0	0	4	21	26
	p47	12	0	0	1	0	13	8		24	23	29
	p48	0	0	10	20	14	0	50	28	69	69	69
	p49	0	7	0	2	0	12	2	6	50	66	13
	p50	0	0	0	0	0	0	0	0	12	3	9
	p51	5	2	4	13	0	0	15	32	64	70	61
	p52	0	0	0	0	6	0	0	34	58	67
	p53	0	16	0	0	0	8	43	48	63	68
	p54	0	0	20	0	0	0	0	0	24	14
	p55	0	0	0	0	0	0	14	25	52	44
	p56	0	0	0	0	0	0	14	0	63	64	52
	p57	0	0	6	0	0	0	0	15	67	76	68
	p58	0	0	0	10	0	0	28	5	25	51	26
	p59	0	0	9	0	3	9	0	0	55	59	3
	p60	11	8	0	0	0	0	15	18	31	54	45
cELISA (cutoff = 40)	p41	5	10	0	0	19	20	29	37	33	38	41
	p42	10	19	25	0	8	13	21	30	57	67	68
	p43	0	6	4	5	7	22	41	48	42	47	41
	p44	5	0	5	0	0	0	0	3	70
	p45	0	1	0	0	0	8	21	27	41	3
	p46	0	0	0	0	6	6	22	33	27	24
	p47	7	10	0	3	1	3	0	14	28	26	32
	p48	0	0	0	0	1	15	28	37	49	60	68
	p49	0	0	0	0	0	18	15	29	9	26	52
	p50	1	0	0	0	0	0	15	26	22	12	27
	p51	15	8	0	0	0	14	37	58	61	68	62
	p52	2	0	11	0	0	0	22	29	37	39
	p53	0	0	6	3	0	24	20	44	59	56
	p54	5	7	18	0	10	6	17	6	16	16
	p55	24	36	16	21	0	18	37	50	54	55
	p56	0	0	0	0	0	0	27	17	55	62	69
	p57	5	8	0	0	0	0	20	40	54	59	63
	DPI	0	1	2	3	4	5	6	7	14	21	28
	p58	7	0	0	0	0	0	23	14	20	5	26
	p59	0	0	0	13	0	9	10	20	15	21	7
	p60	0	0	0	0	0	0	0	0	0	2	6

cLIA — competitive Luminex immunoassay; cELISA — competitive enzyme-linked immunosorbent assay; DPI — days post-infection.

Green cells represent positive samples.

The numbers represent percent inhibition of median fluorescence intensity (MFI) of the diluent control for cLIA or optical density (OD) for diluent control for cELISA.

c and p in front of animal identity (ID) stand for cattle and pig, respectively.

The response pattern for serial bleeds from experimentally infected cattle and sheep was identical for FMDV cLIA and cELISA. However, positive antibody response was detected in more serial bleeds from experimentally infected pigs with cLIA than with cELISA (Table I). The response pattern for serial bleeds from experimentally infected cattle was identical for cLIA and cELISA. Similarly, positive antibody response to VSV was detected in more serial bleeds from experimentally infected pigs with cLIA than with cELISA (Table II). Indeed, sera from 10 of 20 pigs were negative for antibodies to VSV-NJ at DPI 14 by cELISA, while all but 1 were positive by cLIA. The differences between cLIA and cELISA for swine may be linked to differences in antigen-antibody binding affinities under the different assay conditions (buffers and serum dilutions). In addition, the fact that cLIA beads are in suspension could provide a larger antigen surface area compared to antigen coated on the plate and could therefore result in better contact between antigen and antibodies.

The cELISA using a recombinant FMDV 3ABC protein and anti-FMDV 3B-specific MAb (4) currently used in routine diagnosis of FMD at the National Centre for Foreign Animal Disease (NCFAD) is serotype-independent and detects antibodies against FMDV 3B in multiple species. Taking advantage of the reagents in this cELISA, we have now developed a cLIA with comparable performance characteristics. Luminex assays have previously been developed for FMDV structural and nonstructural proteins (7,8,12,13). In these assays, antibodies in bovine or porcine serum react with bead-bound antigen, followed by the addition of biotinylated anti-bovine or anti-porcine detection antibodies and then streptavidin-PE. Although this setup was successful at detecting FMDV-specific antibodies in sera, optimization can be challenging as nonspecific binding of serum to beads is a common problem. With a cLIA, the problem of nonspecific binding is minimal, especially with MAb because competition is for a specific epitope on the antigen. In addition, similar to the cELISA, the cLIA can detect antibodies in sera from multiple species. Our data indicate that the FMDV 3B cLIA is comparable to the FMDV 3B cELISA for bovine and ovine sera. Furthermore, the cLIA seems superior to the cELISA for porcine sera, detecting positive antibodies earlier and in more experimentally infected pigs compared to cELISA. This indicates that the cLIA can be exploited to generate even more sensitive tests for detecting antibodies to FMDV.

A cELISA for VSV-NJ is also routinely used at the NCFAD and relies on antibodies in serum competing with mouse anti-VSV-NJ polyclonal antibodies for epitopes on the whole virus antigen (5). The glycoprotein constitutes part of the VSV structure and is responsible for entry and release of virus from susceptible cells (14). Thus mouse anti-VSV-NJ polyclonal antibody is expected to react with recombinant VSV-NJ G protein. Consequently, we have used the mouse anti-VSV-NJ polyclonal antibody in combination with recombinant VSV-NJ G protein in a cLIA. The VSV-NJ G cLIA was comparable to the cELISA for detecting antibodies in bovine and porcine sera, which indicates that the cLIA is a viable alternative for cELISA. This was further supported by the identical antibody response dynamics in sera from experimentally infected cattle measured by cLIA and cELISA. On the other hand, cLIA was apparently better at detecting early antibody response in experimentally infected pigs, recording 95% positive results at DPI 14 compared to only 50% detected at the same time point by cELISA.

Competitive Luminex immunoassays (cLIAs) show great promise as proof of concept and could be suitable for diagnostic use with further validation, including samples from field outbreaks. Luminex assays can be multiplexed to detect multiple analytes at the same time. In fact, multiplex Luminex assays have already been developed for simultaneous detection of antibodies to multiple FMDV nonstructural proteins (NSPs) (7,8) and the detection of antibodies to both structural protein and NSP of FMDV (12,13). Multiplex Luminex assays can also be applied to the detection of antibodies to multiple viruses, especially those responsible for similar clinical signs as seen with FMDV and VSV in cattle and pigs. These cLIAs for FMDV and VSV, as well as other cLIAs for related viruses (VSV-IN and SVDV), could therefore be combined into a single assay for differential detection of antibodies to foot-and-mouth disease virus (FMDV) and vesicular stomatitis virus (VSV).