Enzyme-linked immunosorbent assay for Brucellaovis specific antibody in ram sera

ence 1984, 3(

Enzyme-linked immunosorbent assay for Brucella ovk specific antiboc am ser NCER, G . W. BUR.bbbb, Keglonal Veterrnary Laboratory, Victorian Department of ?, Hume l figh way, A Benalla, Victoria, 3672, Australia

AIIr r r ~ ~ ~ ~ ~ r l immunosorb inked Ir bent assay the detection of Brucella ovis specific a n ... rne ram serum was compared wlrn y mployed complement fixation test (CFT). Rabbiit nti-sheep IgG coupled to horseradish peroxidase as used as the antibody-enzyme conjugate and 2,2' ,. tino-di[3-ethylbenzthiazolinsulphonate (6)l as thC substrate. The ratio of the optical density at 414 nnn for positive and negative control sera (P/N ratio) waS used to optimise the parameters of the test. Ran serum samples (16,527) were tested using ELSA ancj FT (warm and cold) over a one year period. Thte e LISA was more sensitive and provided a mor~ liable measure of B ovis specific antibody than dic1 le CFT. Implications of emplojing ELrsA as the sole st in an eradication scheme are discussec

..

-

'

iitted for r.outine dia gnosis 16,527' clotted bl of Agricu Iture's as par t of the Vi ?partment .. .. .-. ~ c c r e a l ~ a t r obcneme n us11~g the Ovine tlruce~~osis methods outlined by Burgess and Norris (1982). Fixatic)n of complement was at 37°C for 40 minutes using t he warm procedure and overnight at 4OC using t h o A d procedure. The lowest measurable titre in these s,ystems wa

- ..

"r\

Enzyn te linked i .

. . .

Coatmg. 'U' bottom polystyrene microtitre plates were CIoated with 50 pl of B ovis complement fixation antiger Serum Laboratories; -Melbourne) . .. i (Commonwealth usually at a dilution of 1:800. rhic antiger1 is a heat soluble extract of B ovis con: mainly of lipopolysaccharide (Myers and Varela 1 1979). The coating buffer was 0.1 M sc VINE br ucellosis has been recognis . . ..--,.^I- ------.- . .....+, I . , . , .I . , . . . rulirlrca ... Wllclc >limp ale u1 C C U l I~I I I~~ U~~ L~X I~C T~~ ~a l ~ ~ lpH ~ a9.5 t e and the incubation period UVTIith the notable exception of Britain. Reduced night at room temperature. The coated plates, were rtility in affected rams as a result of infection with washed either with a Dynawasher (Dynatech La~borae causative organism, Brucella ovis, may lead to tories) or Multiwash (Flow Laboratories) and a wash .. mM re1duced lambing percentages or an increased lambing fluid consisting of 150 mM sodium chloride, 1 -0 ...... Pe riod. This has prompted the introduction of sodium phosphate pH 7.2 plus 0.05 per cent 1ween 20 (pes/Tween). No significant differences were CO ntrol or eradication schemes, especially in stud , based on complement fixation testing observed in the results obtained with the numlerous flc~ c k s usually brands of microtitre plates which were tested. FT)and removal of reactors (Burgess 1982). It has been demonstrated that the co nventionalI T may fail to detect some infecteci animals dilutior1s. All serum sampl es were dsiluted . - -.' iberstein and McGowan 1958, G o r r ~ e 1959, ,..&"" ,.-~Seru,;,m nnr --I-. - > > .> - lT.... -....I .-.,..I.:-Iea to Bi berstein et al 1962, Ris 1964. Gradwell and Van Zyl Ites the plz whicf I were tlien incublated at Iroom 1975). Alternative diagnostic procedures have been in1~estigated including fluorescent antibody, gel temper;iture for t wo hours before wa:shing as ailove. dii-fusion, indirect haemagglutination and allereic rests as reviewed by Burgess (19Ei2). Conjugate. Antlsera to purrt~ed ovlne IgG was The present study describes an ELISA pro1cedure for produced in rabbits using established procedures the serological diagnosis of B ov,is infectio n in rams. (Butler and Maxwell 1972) and conjugated to horse.-- .- .. peroxidase, Sigma type V1 by the method n f Substrate incubation times, serum, antigen and radish njugate di lutions were investig ated. ELISA and CFT Wilson and Nakane (1978) except that the chrornato,ults were also com pared for a large nlumber of graphy step and sodium borohydride treatment were omittedI. The conjugate was mixed with an Icqual .um samples. vnltarno of glycerol, aliquotted and stored at - ? n o r LU L Materials and methods in the liquid state. Activity was ma intained t~ n d e r these c~onditions for at least 20 mo nths, how ever, Complement fixation rests deterior'ation occurred after storage for more than .. . .n Warm and cold CFT titres were determined GI, ~nthsat 4°C. Fifty microlitre volumes or" rnls : a ,

nnrl.,...

m

,

~

1

11

.ar

"..

."."...~

...,A

m,.

-

"A

ELISA for

Brucella olvis antibody in ram

. .

--

.

conjugate dlluted In PBS/lween were added to th ITlicrotitre pdates, inci bated at r.oom temp t\vo hours 2~ n dthen Hlashed.

Substrate. This consisted of 0.1 M citrate. phosphate buffer pH 4.2, containing 1 mM 2,2' azino-dil3-ethylbenzthiazolin sulphonate(6)J (ABTS and 2.5 mM hydrogen peroxide. The optical densit: as read after addition of 100 pl, usually after 6( linutes (Fig I), at 414 nm ( 0 ~ 4 1 ~on ) a MR59( l i c r o ~ ~ rMini s ~ Reader (Dynatech) or Titertel lvlultiskan MC Plate Reader (Flow) interfaced to a1 pple I i Cc Conrrols, ,its and tt?st sample --id negatlve control serum samples were al~quotted and stored at - 20°C. The positive control originatec from an artificially infected ram which had a CFT titrtof 64 while the negative control was a pool from 2i Lqown negative ram flock. Dilutions (1:200) of thesc lsitive anid negatbve contrc)Is were arbitraril! ,signed va lues of I( 124 and 0 ELISA units respec. ~ely.Furtller dilutiolns of the positive serum werc .. -... - - I - .:. .- -. ;ed to construct- a- standard curve relaung V U , , ~ tc .ISA units (0 to 1024). Test samples were dilutec 200 in ~ ~ s / T w e eand n 40 duplicate 50 pl sampler Ided to a microtitre plate in rows 1 to 10. Rows 11 )d 12 were reserved for a conjugate background, :gative control and six dilutions of the positivc rntrol in duplicate. The ODdl4 values of the tesi mples were compared to those of the standards anc :urve fittir~gproceduire was use Ice relative .ISAunits.

'

I

1

40

A

I

60 100 Time (min)

I

I

140

180

FIG 1: 1I Ime course of colour de,velopment. ' control serumwas used at a Idilution of 1 ::? Other & coIM)5. as per material!;and methods. Results are!expressed a!

eaction tir;ne-course

Serum Idilution

Data con tained in Fig 1 show the time-course of .--. *. lostrare nydrolysis expressed as ODql4. Colour .oduction is linear only over the first 10 minutes, aches a plateau at 60 minutes (1 a54 rt 0.10) and is aximal at approximately 180 minutes (1 -82 r 0.10). Extending the incubation time to 16 hours sulted in a 17 per cent reduction in the OD4,,, to 51 -c 0.07. Choosing an incubation time of 60 inutes or greater results in an OD,,, which is relatively constant at 1.69 r 0.13 ( r 8 per cent). TIiere is minimal non-enzymic hydrolysis of ABT: o\Ier this time span. Increasing the ABTS concentraticIn above the 1 mM used here, produced larger es accomr)anied, holwever, by increased colour Idue to tlhe intrins.ic optical nhydrolysc:d ABTS. , . ~ & ~

' 8

FIG 2:

Ettect ot serum allutlon. uoutmna anutlons or a ~ositive sted under standard serum 10)and a negative serum as describec1 in the conditions. The PIN ratio (A) w? text. Results areexpressed as meal

increasing dilution on t h ~c u a n tesults for both positive and negative serum samples. All 0 ~ 4 ~values 4 have been corrected for background binding of the conjugate to the coated plates (0.02 .+ 0.01). In the case o f the positive serum there is a slight prozone effect (approximately 20 per cent) at the lowest dilutio n tested (1:4) and the OD,,, values fall from a maxim ... . um of 1.31 5 0.04 to 0.02 r 0.01 at serum dllutlons of 1:64 and 1: 1,048,5;76 respec tively. Statistical analysis shows that the sta ndard dev iation (SD)varies from rt 0.01 to r 0.07. T he mean a.nd SD in the area for diagnostic purpose a,, . ~ cut-off ~ . 1:16,3t54 (see below), is 0.54 -c 0.02 (2 4 per cent). The ra,tio of the respective positiveand negative OD414 values (P/N ratio) increases from a minimum ()f 3.2 1

l'Id,

Serum titration The data presented in Fig 2 show the effect of

IP

k Y 1 Conjugate dilution

tiqen dilutior

.--

n Fig 2 except Flti 3: Ettect 01f antigen dilution. ConditicIns were as i~ that the positive control serlJrn (0)and rlegative contlrol serum ( 01 were used at di'lutionsof 1:2DO

above 40 at dilutions of 1:4 and 1: 1024 respecvely. Abatve a dilution of 1:16,384 the P/N ratio . to 8.0 at 1:262,144. 311sagaln, 3

1.

-.-

Positive set.urn dilution

--

FIG 4: Eiffect of conjugate dilution. A negativ~ . . .--. . . . .. ,.. . e control. serl . urn at a sinole dillutlon 01 1:ZW ana aoubllng allutlons ot a posltlve control serum were tested at conjugate dilutions of 1:100 (0)and 1:m ( 0 ) .00414 readings are those observed after incubationI of the substrate for 10 10)and 60 ( 0 )minutes respectively. Insc!rt: PIN ratio as a function of conjugate dilution. Conditionsas in Fig 3

the su bstrate for 10 minu tes and 60 minutes in the 1:100 and 1:80C conjugat e dilution systems respecto the right as the The effect of varylng tne antigen coating conI- tively. The respatnse.curve . is shifted -C.--*L:.- I L I ~ L I U I I13 :.. L . . . . . : -n .ca.,.,*. from 1:100 to l.QM I .OW, L U l l J U (sate diluti~lnr ,,,,,dased S I I U W I L in Fig 3. The sensitivity of th G qstem, as evidenced by the o chanlges for the resulti ng in enhanced sensitivity to low lev,els of ositive serum sample, is similar over t he dilutio n specifiic antibody, eg, 1:12,800 and 1:25,600 dil utions ange 1:25 to 1:800. There is a decrease aibove 1:80 0 of the: positive control serum produce an or)4l? of ..u~lution -.-_tested, the- ueru n r n ....- - with 1:100 and 1:800 conjugate diluuuns ~ c that h at 1:25,600, the highest :vel is less than 50 per cent of the maximum respectively. However increasing conjugate dilutions bserved at 1:100. Concomitantly, the selectivity, as also produce a marked decrease in the P/K I ratio of videnced by the P/N ratio, increases threefold from (insert to Fig 4). This ratio has an optimu m, - 4 to above 16.0. This increase is a reflection of the approximately 20, at a conjugate dilution of I:ZW ecrease in ODdi4 for the negative serum sample from and falls by 45 per cent to 11 at a conjugate dilution -27 +. 0.04 to 0.05 2 0.01 at antigen dilutions of of 1:3200. The optimum conjugate dilution is 1:800 :25 and 1:12,800 respectively. Background binding because acceptable optical densities are observed uf the horseradish peroxidase conjugate to coated after 60 minute!j incubaticIn, there is an acceptable plates also decreased over this range from 0.09 e sensitivity and acceptable P/N r,atio to Iwable 0.01 to 0.02 e 0.02. Maximum sensitivity and differentiation IJetween negative a1nd low pcositive selectivity is observed at an antigen coating dilution serum readings. f 1:800 and for these reasons this dilution waS nployed routinely. Normally plates are coated a YL'LISAana CFT Comp )om temperature (22°C). However, similar resultS Ran serum samples (16,527) were tested by warm re obtained when coating is performed at 4OC o.r and cold CFT and ELISA. The data comparing ELISA units and CFT values are given in Table 1. Mean ELISA units increase progressively with increasing CFT titres in both the warm and cold a Data tori~tainedin Fig 4 show the effect of two series. The SDS for the ELSA units are large, a f,inding c(mjugate d ilutions on the OD414values for increasing consistent with the observed ranges, eg, the ELISA d ilutions o,f the positive control serum. Note that units for the 256 sera having a warm cm titre of 16 these readlngs were obtained after incubation with vary from 4 to 1534. Overall there is an appro, Lntigen cotating para

e

,.-.a

A..

~

--A

."

L.

-

ELISA TABLE 1: Corn CFT titre

for Brucella (

~ d in y ram

LISA units an~dwarm and cold CFT titr

lean

+ So

Warm (n1

Mean + So

ange

1

I 11 1:

twotola altterence In tne warm ana cola cm rirres for a given ELISA unit value, eg, 120 +- 156 and 113 -+: 153 ELISA units at a warm CFT titre of 16 and cold CFT titre of 32 respectively. This is consistent with previously reported observations on the relative sensitivities of the cold and warm CFT (Allen et al 1976, Burgess and Norris 1982). Seven-hundred-andsixty-seven (4.6 per cent) and 1024 (6.2 per cent) of serum samples were positive to the warm and cold CFT respectively (CFT titre at least 8). In addition a number of sera exhibit ELISA reactions yet do not react in either warm or cold CFT. Assuming an arbitrary cut-off limit of 10 ELISA units, there are 222 mples which are cold CF-r and warm CFT negative ~dan extra 216 samples which are cold CFT positive ~d warm CFT negative. The former group represent a11extra 1.3 per cent of the total number of serurr samples tested, have tt mean + SD of 26 rk 30 ELlSP, units and range up 1to 262 EL ISA units. They arc: defined as being false negative Ireactors with respeclt the cold cm. L

'

Work reported In t h ~ spaper describes an ELISA system for performing B ovis specific serology in rams. Excellent differentiation between positive and negative serum samples was obtained as evidenced by calculated p0sitive:negative ratios. Using this ratio as the prime criterion differentiation was optimal at a serum dilution of approximately 1 :1024, antigen dilution of 1:800 and conjugate dilution of 1:800. As a compromise between ease of sample dilutionI and selectivity of the system, routine testing was1 performed using a 25 yl serum volume in 5 ml of PBS/' Tween (1:200). Several potential sources of error :re also investigated. There was very littlechange in e ELISA units of samples when the volume added to e plate was within 10 yl of the normally employed Air displacement pipettors with individual ,J clc:an tips wc:re necessary at this Istage to reduce carry ov.er from positive samples. Fiowever, the initial d i lution of s,erum could be carria:d out usir1g positive u1: A "-1: ~ p ~ a c cglass ~ ~ ~capillary r ~ ~ t pipettors as long as a rirising step in p~s/Tweenand wiping of the outside' 0 f the capillary with tissue was performed. Implemlentation of this step decreased the overall process'

Cold (n1

Range

(155031 12121 12581 (2781 11771

lng tlme by approx~matelyone minute per pla samples), a n important factor when I arge numt samples are being processe,d, u the An important feature of the s +;.,a inclusion of a series of posiLgVL "13 u I I Lac)I plate to compensate for interplate varia tions associated with variable incubation times, ternperaturer;, etc. Results are then calculated and expressed relal:ive to Ilse nf these controls not as a function of on,.. -,,. . The ...- ---O D ~is, meaningless ~ unless conditions are very strictly controlled and does not allow for direct day t o day comparison of individual samples. Colour pr,oduction was slightly increased in a random manner in t h e outside wells, the so called 'edge-effect'. Fortui tously the slope of the standard curve was such that these small variations were within acceptable limits. for B ovis has advantages when cornIpal FU C L "A IJ with CIFT methods for the following reasons. First, it i s a ciirect method of identification of sl)ecific antibsdy and not an indirect method of assessment . , sucn as complement induced haemolysis of red ,DlOOa re cells with its attendant complications. Therefo~ it is not prone t o false positive reactions associated with, eg, anticomplementary activity, as is the cold IIFT in particular (Ris 1 1 et al 1976, Burgesis and Norris 1982). d thus Secondly, it is more sens itive . -than .f has the potential to detect lntected animals which otherwise d o not react in the CFT. A limited nllmber of CFT n e g a t i v e / ~ ~ lpositive s~ anilnals have been tested and shown to be excretors of B ovis (SyJencer and Burgess, unpublished data). Thir~ dly, the antibody-enzyme conjugate employed has ligl-11 chain reactivity and thus is able to detect all classes o f antibody. As a comparison, in sheep and o.t h- e r ruminants the CFT only detects antibody nf h ~ -. t...IgG, subclass and to a lesser extent IgM (Fei nstein and Hobart 1969, Allen et al 1976). This, fact probably explains the large variations in the ELISA units observed for a given CFT titre (Table 1) a:. InP. comprises a large proport ion of to1:al ovine ! antibody (Heremans 1974). Fourthly, ELSA results p rovide an -C U I 1 1 ogical l--l.tool for investigating the ~,c--.:..I I I C L L ~ V>Laws IUCKb. Exclud ing known reactors, recently infected 1flocks invaria bly produce 'background' ELISA uni ts of betwee n 3 and 9. This would appear to be d ue to ,.A"+,-,

buIII1

..a

p1

La.-

'abortive' infection with B ovk and subseque recovery (Burgess et al 1982). Conversel!I flocks th .- - - ..-- . are known to have been free of infect~on over i :xtended period of time are less than 3 (usually 0 arld I ELISA Un a ent.. choicc ELISA is . : on which to base arge-scale B ovrs eradlcatlon campalgn In rams. I + las the necessary sensitivity yet does not produce glse positive reactions as does the cold CFT. Any endency to 'overkill' by removing serologically ~ositive/culturenegative animals from the flock :' nore tharI balance1d by the knowledge that :111 )n )otential excretors have been identified. Reinfectic ind consecluent revel-sion of the disease-free status (3f .. prevented. -. he flock is. rnus The method lends itsr''I,. o automaltion both of the test and the r.ecording I3f esults. ;A :ss to.. d .e ~ Further work is ,ystems for determining the contrlbutlon or Igb,, IgM to the B c - -

..

Acknowledgements

We thank Ms C. Stacpoole for capable technic assistance and Mr Norris for the CFT results. Th

Keterences ALLA.N, G. S., CHAPPhL, R. J., WILLIAMSON. P. & MclVAUGHT. D. J. (1976) Journal of Hygiene 76. 287-298 BlBEl*STEIN, E. L. & McGOWAN, B. (1958) CorneN Veterinarian 48, 31-44 BIBERSTEIN, E. L.. McGOWAN, B. Jr, ROBINSON, HAROLD, D. R. (1962) Cornell Veterinarian 52, 214BURGESS, G. W. (1982) Veterinary Microbiology 7, 551 BURGESS, G. W., McDONALD. J. W. & NORRIS, M. Aus~lralianVeterinacv Journa159.45-47 BURCiESS, G. W. & NORRIS, M. J. (1982)I Auslralian IVelerinary Jou.rnal59. 23-2 !5 BUTL ER, J. E. 8l MAXWEL L. C. F. ( I ' 972) Journal of Dairy Science 55, 151- 164 ..-,-. .. FElNISTEIN, A. & HOBART, M. J . ( IYJY) Narure rrd, GORA:IE, C. J. R. (1959) Australian Vecrinary Jo, 500--501 GRAI: )WELL, D. V. & VAN ZYL, F. E. (1975) Jourfiral of rhe

.. .

-a-

Po,,. ---lh African Veterinary Association 46. ?do <-,

HEREMANS. J. F. (1974) The Antigens. V'olume 2. Ed M. Sela. New York. Academic Press. pp 365-522 MYERS, D. M. & VARELA-DIAZ, V. Ibl. (1979) J( Clinical Microbiology 10, 45 1-453 DIP ,-..-..-, .* 73. R. (1%4) New Zealand Veterinary Jvurnut I L , r r :-,2 3N, M. B. & NAKANE. P. K. (1978 ) lmmunoflu orescence Related Staining Technicques. Eds W Knapp, K. Holubar G. Wick. Arnsrerdam. Elsevier. pp 21 5-224

.