- Email: [email protected]
Detection of antibodies to myelin basic protein by solid-phase radioimmunoassay with [125I]protein A
Journal of Neuro~mmu~olo~. I (1981) 17-26
17
Elstwier/No~h.HollandBiomedicalPre~.
Detection of antibodies to myelin basic protein by solid-phase radioimmunoassay with []2Si]protein A D.S. Linthicum *, S. Jones, L. Horvath and P.R. Carnegie The C/i~ical Re~t.¢rch Unit of The Walter and Eliza Hall l~zizweo/MMcal R~mro~ Royal Metbow.nt Hozp{ta~ P.O.. Victoria 3050, and $choo~of/l~cldture. L~ Trobe U~iwr~ity, Bundoora, Virroria 3083 (~4ustraha) (Received24 :,~vember,1980) (Ac~p~ed 26 Novcmbei, 1980}
S~mry A solid phase radi~immunoassay (RIA) for the detection of antibodies to myelin basic protein (MBP) in sera has been developed employing MBP-coated flexible polyvinykhloride microtiter trays and |tZ~l~rotein-A as the radiolabeL [12sl]ProteinA directly binds to the Fc region of serum I g ~ from several animal species and s e ~ e t as an excellent reagent for detecting antibody. It can also be used to bind to a second :mtibody l i ~ d , thereby making it useful even when it does not bind directly to the ~.rin~ry antibody Fc region. The use of one preparation of [nSl]protein-A label allows sera from sev-.xal species to be tesled for antibodies to MBP simultaneoedy, thereby makin~ this RIA technica~y simple, re-~id and ecoaomical. T I ~ assay hat, been particularly useful in examining serum sample~ from anim~!~ with experimental autoimmunc encephalomyelitis and hypomycli~osentsis con~c~ta, In patient~ with multiple scleroeis low levels of antibody to MBP can be detected in cerebrospinal fluid (CSF) and acid-eluted extracts from brain plaque material; detection of low levels of antibody in human serum has not b ~ n po:k~ibl¢ due to non.zpecific binding of human serum lg in this RIA.
C.ccmtpandee~~ be sent to: Dr. D.:;. Linthlcttm,Departmentof M~embiolc~y.US'~Schoolof Mcdic/ae, 2025Ze,ud Arena. l.m Anl~h:*,CA 90033. U.S.A. I).KL. w~ ~ by ~ p~-,loc~aJ rei~,h;p from the Nafoeal Mul~pk Sck.r,~ Society (U.S.A.). 1"~ work was wCpor~l by Itt~tt [o¢ Dr. l u Vatckayfrom the NaUoeal Health a~d Mt4ical ltlJesr¢~~ M ,,U~jCs11~=,~Ik Ibm~urdlGfsat~ Committee o{ ,k* Au~mdJan ~ TI~ work w m ~ by Grants 887-C-4 ~ I0022-A-7[n:~ t~ Nat~0Al M,,1~pk Sckro~ Sa:kW (USA) m d gnats from the N l ~ M ~ ~ d M ~ d Rt~ea~.-hCoJ~/! ~ the Nafio,ad
* l'e*t4ecto~ R_,v~.._..~F,:13owc4 , ~ Nmloe.,l Multiple Sderom Society (USA). thin,ere ~ : l)q,mmw~ of Micmbk,.qy, USC Scrod of Medki~, 2073 Zonal Avew~, Lo~.~ CA ~03~, U.S.A. 0165-~72' ~I/~(~O-CO00/$02.SO © F-J~/Nonh-HoU~l Biof~'dical Presa
18
Introduction Although ;mtoimman¢. dernyelinating diseases of the central nervous system (CNS) appear to be primarily due to cell-mediated mechanisms, the ancillary pathogenic role of humoral antibody to CNS antigcms rfmaln~ to be elucidated. ',n experimemal autoimmune encephalomyelitis (EAE), both cellular and hnmoral .~mmunity to myelin and myelin basic protein (MBP) ~re evoked but the detection of low levels of antibody to MBP usually requires sensitive techniques. A number of radioimmunoassays (RIA) have been devised for this purpose. Whitaker and McFarlin (1977) compared several and discussed their advantages and disadvantages. Randolph et al. (1977), devised a solid phase RIA employing glass tubes coated with MBP. The a~T~,,mt vf anti-MBP antibody in sera from rats with EAE was quantitared using [t2~i]rabbit anti-rat IgG (affinity purified) as the radiolabeL While this assay is an intprovement over previous methods for the detection of antibodies to MBP, it involves the manipulation of glass tubes and preparation of affinity-purified second antibody for radiolabelling. Obviously. if large ntlmbers of sera from several species are to be assayed for the presence of antibodies to MBP this would Lnvolve labor-inlensive~ costly and time-consuming procedures. n~l-radiolabelled protein-A ([nSl|ProA) from Staphylococcus aureus has been proved to be an excellent tracer for general use in immunoassays (Goding 1978; Langone 1978). Radio-lodination of ProA is easily accomplished and the labelled product remains stable for several weeks. The ability of [~251]ProAto react directly with the F¢ region lgG antibody from several species in a solid phase assay has been convincingly demonstrated (Langone 1978). Furthermore, [nSI]ProA can react with the Fe portion of an "indirect" or "second" antibody ligand; thereby making it a useful tracer ~ven when it does not bind directly to the primary antibody Fc i~giou. The method is similar to that used by Marier et al. (1979) for the assay of antibody to fungal ant~.~3~ns. Rec~.ntly several groups have employed the Pro-A technique to assay human sera and CSF for antibodies to myelin basic protein (Bashir and Whitaker 1980; Panitch et al. 1980). In this report we describe the development and use of the solid-phase RIA using 96-weli mierotiter plates coated with MBP and [IZsl]ProA to detect at~ti-MBP antibodies in the sera from animals hyperimmunized against MBP. This assay has also been employed to detect low levels of anti-MBP antibody in sera from guinea pigs with EAE and naturally ocet,rring mti-MBP antibody in lambs with Border disease aod in extracts from multil:le sclerosis brain. Materials Sheep, guinea pigs and rabbits were obtained from local couunercial sources. SJLxBalb/cFt h)brid female mice were bred from The Walter and Eliza Hall Institute (WF.HI) specific p~thogen-free stocks and used at an age of 6 - 8 wits; this hybrid is fully susce~ti~le :,, EAE (Bernard and Carnegie 1975). Purified protein-A (ProA) from Staphyl~cocsus aur,,us was obtained from Pharmacla, Uppsala, Sweden, Disposable polyviny!~:htoride "U" microti ter plates (cat, no. 1-22024) were obtained from C¢~oke Laboratb~, Products, Alexandria, VA.
Basic proteins of myelin (MBP) were prepared from human (H), chicken (C), mouse (M) and sheep (S) brains (Dankley and Carnegie 1974). A mixture of the large and small mouse MBP's was used in the experiments employing M-MBP. Methods Immunizations and antisera
Antisera to MBP were raised in sheep and rabbits by hyperimmenization over a period of 6-7 months. Sheep antisera to HMBP were raised by intramuscular injection of 2 or 3 mg MBP in Freund's complete adjuvant (FCA). supplemented with 0.75 mg M. tuberculosis. Fifty days later a booster injection o[ 1 mg MBP in FCA was given and similar booster injections were repeated every 30-50 days. Sheep were test-bled 10-14 days after eech booster injection. Rabbit antiserum to CMBP was raised acx:ording to the method of Cohen et al. (1975), by s.c. injection with 10 rag MBP in FCA, followed by booster injection of 5 mg every 30-50 days thereafter; Me f~rst in FCA and subsequently in Freund's incomplete adjuvant (FIA). Rabbit antisera to HMBP were raised by an initial injection of 250/~g HMBP in FCA; booster injections of 0.5-1.0 mg were given in FCA at 30-40 da3, intervals. Rabbits were test-bled 10-14 days after each booster immunization. Rabbit antiserum to sheep lgG was raised by injecting 1 mg purified sheep igG in FCA; a booster injection of 0.5 mg was given in FIA 30 days later and the antiserum w ~ obtained 10 days following the booster injection. Antibody from Ihis serum was affinity purified on a Sepharose-shi:ep IgG column and used for an indirect RIA technique (see below). The gamma-giobulin fraction of rabbit anti-sheep 18(3 was also purchased from Cal Biochem (California). Guinea pigs and n~3ee were immunized for the induction of EAE by standard procedures developed in our laboratory (Bernard and Carnegie 1975). Sera from newborn lambs afflicted with Border disease were kindly provided by Dr. H. Westhury OVesthary et al. 1979).
Optimization o/ the radioimmunoassay (RIA) Several parameters of the solid-phase RIA were extensively tested to establish the optimal conditions: (I) concentration and time required for antigen-coating of the Microtiter wells, including shelf-life stability of preconted trays, (2) anti-serum incubation temperature and time, (3) sp,¢ific activity and stability of the radiolabelled [t~l|ProA, (4) incubation temper#tare and time for the [*251]ProAlabelling, and (5) reproducibility o! titration end-points for specific antisera. This report describes optimal conditions which were establish,,,d for this RIA.
Labelling o/proteins Pro~inA and HMBP were radioiodinated with 12Sl using the c.'doramlne-T method (Hunter and Greenwood 1962) by Mr. 3. Pye and Mr. D. Quillci of the WEHI radioisotope laboratory. Several radioiodinations were performed on 25 p.lt ProA reaultinll in specific activities of I0, 30 and 50 ~Ci//~8; after several trials it was found that a qsecifie activity of 30 #Ci/~tg was optimal and the majority of the
2O data presented in this paper are based on a [t25I]ProA label with a specific activity of 30 /tCi/#tg" Storage at 4°C of [t2~l]ProA for gre, tcr than 6 weeks resulted in apparent radiation dam~ge and denaturation of ProA. especially at specific activities 30 ~Ci/~tg, hence only i:=l]ProA which was less than three weeks old was used. To compare with [:2eI}ProA radiolabelling, in one experiment an affinity purified sheep anti-mouse Ig antibody (antisera prepared by Dr. J. Goding) was used which was t251-radiolabelled with specific activity of 10 ltCi//tg"
Antigen cvati~,gof wells [t25I]HMBP (! ngi was added to varying amounts of unlabelled HMBP (1-10 pg) and the amount o¢ MBP hound in each well (4 h incubation at room temp) was calculated to be 0.65/~g/10 Itg MBP. Since no saturation of the MBP binding was observed., an arbitrary concentration of 10 itg/well (in 100 ~1) was chosen for the coating pro~Mure. That this concentration was effective was confirmed by using a rabbit antiserum to HMBP. Different dilutions of this antiserum, between 1 : 10 and 1 : 10,000, were tested with wells coated with HMBP between I ng/well and 100 /~g/welI. With wells coated with 5-25 ttg there was no increase in the amount of rabbit antibody which was bound from each antiserum dilution.
Recommendedprocedurefor RIA The RIA was performed in 2 stages: (1) coating of MBP to the 96-well microtiter plates and (2) titration of antisera and addition of the [~251lProA radiolabeL To coat the trays with antigen, 100 ~1 of borate-buffered saline (BBS), pH 8.5, cont~ning MBP (100/~g/ml) was added to each well of a microtiter tray and incubated at room temperature for 4 h. The trays were then washed 5 times with a gently flowing stream of distilled water. Bovine serum albumin (BSA), 1% ~n 200 itl BSS, was added to each weil and incubated for 30 min at room temperature; this step was found to be necessary to "saturate" any sites which may lead to non-specific binding of immunoglobulins. After this step, the trays were rinsed in distilled H20 as above and stored dry until used (plates may be stored dry at room temperature for up to 8 wks). To perform the serial dilutions of the sera to be tested, 100 Itl of diluent (1% BSA-BSS plus 0.05% Tween 20) was added to each well of the tray. In preliminary experiments it was determined that the presenee of Tween 20 increased antibody binding to MBP and greatly reduced the non-.~pecific binding of normal Ig to the well surface. Serial dilutions (log 3 or logs) using 50 ~tl or 25 #1 of sera were made. The trays were usually incubated at room temperature for 4-6 h and washed 5 times in distilled water. With most sera tested (human, guinea pig, mouse, sheep and rabbit) incubation at room temperature for 1-2 h proved to be sufficient. Recent experiments have proved that washing the plates in B~S with 0.5% Twecn 20 is more effective in r~:lucing the non-specific binding. [ml]ProA (30,000 cpm) was added in 100 ~tl diluent to each well and incubated overnight at room temperature. After removal of the excess [~2sl]ProA solution by suction and 5 washes in distilled water, the individual wells were cut off with a hot wire, placed in the plastio Wasserman tubes and assayed for bound |25I by gamma sp~trosoopy. After counting the wells were discarded and most of the plastic tubes can be reused after checking that all the
125I has been removed. With each serum tested, a similarly trea~.ed but MBP-free plate was rm: in parallel with the MBP coated plate to determine die non-specific binding. The amount of anti-MBP antibody in the test serum i~ exprebsed as the amount of I~l epm bound in each well. The data pre.zented in the figuxes are expressed as the mean cpm + standard deviation fo: each set of triplicate serint dilutions (inverse log 3 or logs) of serum t~ted. The calculated dilution which expressed 335 of the maximum cpl~l bound was chosen as the titration end-point (Zollinger et al. 1976).
Rendts Rabbita;ffiseratoMBP Sera from 3 rabbits hyperimmun~.zed with HMBP were tested for anti-HMBP antibodies on HMBP-coated microtiter trays. The binding of antibodies from rabbit N o . 15 anti-HMBP and normal rtdibit sera on I'~MBP
i : t ~ l - ,ta,tTl-~cj~ t , ~
}
Fig. I. 'riu'atioe of ~
from a nocmalrabbit (O) ~
from • rabbit immunizedwith HMBP(0) o~
sadte~-fme adcmftcr p l * m ( . . . . . . ) ~ tll~aP.emted p l a ~ ( ). Amoeat of l m l ] ~ o A iJ u ¢pm boead to the wells of the microtlte¢ platel ~,tota) (J~lllh'aA added was 30,000 ¢pm).
F/S. 2. T ~ ot oocaudrabbi (A, A) left~ a~l serumfromt r~Ibbit~ oa na/~t4~ ix~a c~mi ~,/daHM~P (@)c¢ CMBP ( O ).
w/th H ~ P (O. O)
22 CMBP, This difference in titer could be due to differences in the determinants. There are numerous changes in amino acid sequence when the CMBP and HMBP are co~/.pa.,c~ (Carnegie, unpublished). Rabbit No. 5 was immunized with CMBP and seca from serial bleedings were tested on CMBP- and HMBP-coated plates. The titration endpoints as determined by RIA on the test blecdings over a 400-day period are presented in Fig. 3. The peak riser on day 80 was i. 800,0t)0 for CMBP and 1:10,0/30 for MMBP. For some unexplained reason the antibody level fell dramatically after day 110 even though booster injections of CMBi' were still being administered. This decline in liter in this period was confirmed by a less sensitive precipitation assay (Biggins etal. 1978).
Sheep antisera to MBP Sheep immunized with HMBP were assayed for anti-MBP antibodies by 2 methods. The direct RIA technique employing solely [IzSI|ProA as the h'acer, yielded low levels (2,000-1,000 ¢pm) of binding because ProA only binds to sheep IgG2 (Ooudswaard et al. 1978). When an indirect technique was used employing a "sec~ id antibody" (rabbit anti-sheep lgG) incubated at room temperature for 4 h in RIA diluent, fo[!owed by |lZsl]ProA, high levels (9,000 cpm) of bi~ding were ~detectable (Fig. 4). However. the non-specific binding of normal sheep serum varied considerably from experiment to experiment and a titration of the rabbit anti-sheap IgG "second antibody" was performed to obtain the optimum dilution, with the highest concentration of the affinity purified rabbit anti-sheep IgG (10 p g / m l = I pg/well), there was a high non-specific binding on HMBP-coated plates. The lowest concentration of rabbit anti-sh¢~ lgG tested (0,1 ~g/ml -~ 10 ng/well) gave the best results" high binding of l'~iieroA and low levels of non-specific binding for normal serum. Using this indir¢~ technique, anti-MBP titets 30 days after the initial immanizations were 1 : 78,000 re: both sheep No. 2 and No. 12. Following 2 booster immunizations given at 30-day intervals, the titers imreased to 1:356,000 and 1:410,000 for sheep No. 2 and No, t2, respectively. Lambs with Border disease (hypom/elinogenisis congenita) induced experimentally, have been shown to have antimyalin antibody by a fluorescence assay (Patterson e t a l . 1977). In preliminary experiments with scra from lambs with naturally occurring Border disease (We.~tbury et at. 1979) binding of antibody to SMBP was observed in 2 lambs with typical symptoms (Fig. 5),
Sera from guinea pigs and mice with EAE Blood samples were obtained from guinea pigs and mice with clinical and histological EAE for measurement of anti-l~BP antibody levels. Since p2si]ProA binds IgG I a~d IgG 2 sub-classes of guinea pig immunogiobulins (Forsgsea 1968) high serum tite~ I'1 : 2,700- ! : 4,700) were easily detected in animals with EAE. As observed by previous investigators (Lennon et at. 1971; Berna.'d et at. 1976) there was no correl,~tion between serum antibody levels detected and the expn.~don of cfinical signs of disease, Detection , f serum anti-MMBP antibodies in mice proved to be a difficult problem. Using the direct [*2SllProA binding assay we dilaaJvered that tern from mice witl~ EAE and normal mice had high levels of non-specific binding to both
~-. to-3
/
'; 1oo
C~v5 ~TFm ~
;¢o
~oo
4oo
ft.~EC4K~
]:iS-3. Sequential mtmplcsof semm from • rabbit immt:+~.edv~th CMBP are assayedwith CMBP and MMBP. An end-point of 33~ maximumcour.i~ ~.unds was tak.~ in ca]cu]atingthe t~lerof antibody. MBP-coated and MBP-free plates. VeT.y tittle difference in the [~z~l]ProA cpm bound was observed between normal s e r u m , : d EAE serum. When an affinity-purified sheep anfi-mou~ I 8 (polyvalen 0 which ",/as radiolabelled with t~! (10 pCi/'pS ) was used again there was little difference between the normal serum (4,400 cpm) and EAE serum (4,500 cpm); both had high non-specific bindit~.~ to tmcoatad wells which was equal to ,.hat for MBP-coatecl wells. The use of an indirect rabbit anti-moase Ig antibody followed by [t~l]ProA tracer yielded sir,tim" results; normal mouse serum had high levels of n o n - s p e c i e binding thereby "masking" the detection of any specific binding due to antibody. In previous studies very low levels of antibody to MMBP were detected in mice injected with MMBP or mouse spinal cord ~ t e (Bernard and Carnegie 1975) but when rat ;viBP was used as the immunogen antibody was readily detected. Wbea samples of these sen~ .~':ra were t e l ~ d on wells coated with MMBP followed by ['~l]ProA simi]~ levels of antibody were detected. The indirect technique with mouse sera is advisable since only some of the IgG classm bind (KronvaU et al. 1970; Mackenzie et -~. 1978). Recent e x i t s , not described herein, have suggested that this high nonspecific binding is due to mouse lgM mid IgG (aggregates) which ate formed during
~sl. & P.llhlnccm~t o( b/nd/nl oi' [ '~IIPmA to dM~pIntibcdy to HIv~P (di~l) by the pr/or ~ a ~ of h ~ with a tecond m~body, viz. r~bi! anfi-,!~,~plgG (iadh'~t). Pil. 5. I M ~ d shmp Ii to SttlIP/n ICln~m~ I IL.'cp im.tmmiz~with ~ ~ P (&), ~ ~ ~ dlxm+ (O) md nonm/dk~p (41). Cpm tufa u dsc [ '~ IIProA stlw+bal to the microtiu~ wells ~ t ~ l~c.__.b-,_'_-~+"+/th r#~blt mini-sheepIs (indirect ~ I O C ) .
the plate washing phase; plates washed in BBS with Twce~ 20 have a significantly reduced "background" level of non-specific binding found in normal sera. Human sera, cerebrospinal fluid and brain extracts
With normal human scra in contrast to other species much higher levels of non-speclfic binding to MBP-free wells was found, Various alternatives to BSA coating were tried but as yet no satisfactory agent or method has been found to reduce the non-specific binding to the plates, which makes it impossible to detect low levels of antibody in sera from patients with neurological disease. In contrast, with cerebrospinal fluid and extracts of normal brain little difficulty was experienced in titering out the non-specific binding. The assay is proving particularly useful in studying the presence and purification of immunoglobulin which binds HMBP from brains of patients who have had multiple sclerosis. While it is possible that binding through the Fc region can account for some of the lgG bound in normal brain extracts it appears that the IgG extracted from multiple sclerosis brains is specific antibody to MBP (Bernard et at. 1981). Discussion
The development of a solid-phase radioimmunoassay on microtiter plates for the detection of antibody to MBP offers many advantages over previous assays. Plates can be prepared in advance and stored for long periods, washing of the 96-wells can be done much more rapidly than with assa3,s using centrifuge tubes, and the ¢~ostis considerably less. [t2~i]ProA has the advantage that it can be used as a reagent for detecting antibody from a number of species. Alternative techniques involving the use of a radiolabelled second antibody would require, considerable time and exposure to high levels of 1251in their preparation. The disadvantage that ProA only reacts with the Fc region of certain classes of IgG in some species (Kornvail et al. 1970) can be easily circumvented by the indirect procedure where a rabbit antibody to the primary antibody is used. Since ProA binds readily to rabbit IgG high levels of [t251]ProA can be bound to the wells. The indirect technique is recommended with species such as mice and sheep where" low levels of ProA bind to IgG. This solid-phase RIA could prove useful in studies on antibody to MBP in humans and animals witl~ neurological disease. Its use with human sera is at present limited by the failure to find a mixture of proteins to minimize the non-specific binding of normal human 18(3 to the polyvinyleldoride wells. When our method was developed reports from Panitch et ai. (1978, 1980). indicated that using a similar assay they were able to detect antibody to MBP in CSF in some patients with multiple sclerosis. With extracts from brains of ~adents who had multiple sclerosis the assay has demonstrated the presence of IgG which binds MBP in plaques and in normal appearing white matter (Bernard et el. 1981). In light of the work of SindJc et al. (1980), describing the binding of aggregated IgG and soluble immune complexes to MBP. caution must be exercised in evaluating the sigalflcance of IgG which binds to MBP. Indeed. the low level of non-specific binding we ohm.fred in normal le~a (e.g. see Fig. I) may be due to aggregates of lgGi ultrtceattitugation of sent before testing is helpful in removing IgG aggregates and thereby reducing the non-specific
25 b i n d i n g . I n a d d i t i o n , the u s e o f detergents s u c h as T w e e n 20 o r h i g h m o l a r s a l t c o n c e n t r a t i o n s (0.5 M N a C I ) h a v e p r o v e d useful in r e d u c i n g non-specific l g binding..
A ~ t s The authors are grateful to Dr. Robia Anders and his helpful ad~ce on the use cf radioimmunoassay techniques. They also thank Mrs. Anne Wilson, Mr. John Pye and Mr. Dennis Quflici for technical ~ss;stance.
IRefeeenees Bashir, R~M. ~ J.N. Whilaker, Mok'cul~urfe.atut~b of imnlunorea¢tive MBP in CSF in persons wlth MS, Ann. Neut,,, 7 (1980) 50. Bernard. C.C~A. an3 P.R. Carnegie, Experh~entai autoimmune enceph~omyelifis in mice-- lmro .~ologic response to n3ouse spinal cord and myeli.-abasic proteins, J. lmmunol.. 114 (1975) 1537-154~+. Bernard, C.C.A., LR. Mackay, S. Whittinghc,-'n and P. Brous, Durability of immune protection gain.,. expethnentul autoimmone encephalom?¢iit+s. CeIL Immuno!., 22 (1976) 297- 310. Bernard, C.C.A~. P.R. Carnegie, V. Randel] and I.P~ Macklt), Purification •f :t|,toanlibody to myelin basic protein from multiple sclerosis. Brain, (1981) Submitted for publicati~r~. Biggi~. J.A.. A. Taylor and EA. Caspa~j, Circulating antibody to myelin basic protein in relapsing. remitting multiple sclerosis, J. NeuroL Neurosurg. psychJaL. 41 (1978) 1131-1134. Cohen, S.I~, G.M. McKhalm and G. Guarnieri, A radioimmtmoe.~say for myelin basic protein and its use for quantitative mcamrenk~ts, J. Neutocbem., 25 (19751 3"~I- 376. Duakley, P.R. and P.R. Carnegie, Isohtion of myelin basic proteins. In: N. Marks and R. Rodkuighl (Eds.), Re~trch Methods in Ncurachemistry, VUI. 2, Plenum press. New York, 1974, pp. 219-246. Fotagren, A., protein A f¢om Sfaphylococcus aureu~ V. Reactio~l with guinea pig y-globulins. J. |mmunol., 100 (19~g)921-926. Godi~l, J.W., Use of s t a p h y ~ pl~[einA as an immunolol~calreagent..I.[mmuno]. Methods, 20 (1978)241-253. GoodtwlMd, J.,J.A. Van der Denk, A. N~rdzij, el aJ.,ProteinA reactivityof variousmammalian immunofJobulins,Scand.J. Immtmol.,8 (1978)21-28. Hunter,W . M and F.C, Grecnwoed, Pt~paratlonof iodine-f31labelledh u m ~ growth hormone of high specificacli,Aty,Nanm: (Lond.),194 (1962)495, Kronvall, G., ll.M. Grey and R.C. Williams, Protein A reactivity with mouse immunoglobulins--Str~turai relatiomhJp between son-,¢ mouse and human immunoglobulins. J. lmmunoL, 105 (1970) 1116-1123. Langone, JJ., [ I~sllProtcin A - - A tracer for genertl use in immunoassay, J. Immunol. Methods, 24 (1978) 269-2B5. V.. S. W'hloingham. P.R. Carnegie. T.A McPherson and I.R. Mackay, Detection of antibodies to the ~lt~e prote~ of human myelin by redioimmunoassay and immunofluorescenc¢. J. [mmtmol., 107 (1931) 56-g2. Mackew~, M.R., N.L W~ner and G.F. Mitcbell, Tli¢ binding of routine immuanglobulin,s to staphylococcul ~'otein .% 3. Immunol., 120 (197g) 1493-1496. Marier, R.. M. Jansen and '~.T. ,4mdfiule, A new method for mc~unng antibody using radlolabeled ptO~.n A in • sulid-phese r~oimmtmo~uty, J. immunol. Methods, 28 (1979) 41-49. Pa,~',,=h, H,S., D.A. Ha/tier tad g . E Jolunon, Antibodies to myelin basic protein in mu|tipk sclcrosis-C'Enkul comdatiom. Het~oloty (Minneap.). 25 (1978) 3 % Pamtdl, H.S., CJ. Hepp~ and K.P. Johnme, CSF tnsibody to myelin basic protein. Arch. NeuroL (Chic.). 37 (IS~O) 206-209, Pstt~oa, D.S.P,. D. SweakW and J.W. Harking, AJati-myelin arJlibodies in the serum of hunb~ w~th e ~ ihicder ~ , J. Neurod~m., 29 (1977) 394. I ~ D.H,, R.F. Kibler and R,IIL Fnte, Sulid-phas¢ radioimmtmo~tay for detection of antibodies to myqdia buic p¢oto~ J. Immm~l. Methods, Ig (1977) 215-224.
26 Si~li¢, C.J.M., C.L. Cambiaso, P.L. ICatt~ a~l E.C. La~.rre, The ~ of myelin basle prote~n to the Fc re~o~ of aiw'~la~d I~G and to unmone complexes, Cl~.Exp. lmmum~.. 4 (19~) I-7. Sindic, C.J.M., C.L. Cambilao, P,L, Mauon and E.C. IAtlerr~ The dution of iSG from sub~-.ute sckrosing panencephllilh And multiple sclerosis braim, Clin. Exp. Inmmn~., 41 (1980) 8-14. Westbury. HA,. D.V. N~o thJne and E. Su'sube, Bc~der dise.~¢-- l~-wisten t infection with the v~rus, Vet. Record, 104 (1979) 406-409. Whltaker, J.N. and D.F_ McFarlin, A comparison of immmaochemk~ incthod,vfo~ the detection of antibodies w myelin encephalitoge~Jc protein, Brain Re~. 129 (1977) 121-128. Zoningcr, W.D., J.M. Dltl~mple and M.S..Arte~in, .An~ysis of pmmacl¢~ affectln~, the ~ pha~ radiolmmunoa.~ay quarltitation of antibody to mcnln~ococca] antiltcns. J. [mmunoL, 117 (1976) 1788-1798.
17
Elstwier/No~h.HollandBiomedicalPre~.
Detection of antibodies to myelin basic protein by solid-phase radioimmunoassay with []2Si]protein A D.S. Linthicum *, S. Jones, L. Horvath and P.R. Carnegie The C/i~ical Re~t.¢rch Unit of The Walter and Eliza Hall l~zizweo/MMcal R~mro~ Royal Metbow.nt Hozp{ta~ P.O.. Victoria 3050, and $choo~of/l~cldture. L~ Trobe U~iwr~ity, Bundoora, Virroria 3083 (~4ustraha) (Received24 :,~vember,1980) (Ac~p~ed 26 Novcmbei, 1980}
S~mry A solid phase radi~immunoassay (RIA) for the detection of antibodies to myelin basic protein (MBP) in sera has been developed employing MBP-coated flexible polyvinykhloride microtiter trays and |tZ~l~rotein-A as the radiolabeL [12sl]ProteinA directly binds to the Fc region of serum I g ~ from several animal species and s e ~ e t as an excellent reagent for detecting antibody. It can also be used to bind to a second :mtibody l i ~ d , thereby making it useful even when it does not bind directly to the ~.rin~ry antibody Fc region. The use of one preparation of [nSl]protein-A label allows sera from sev-.xal species to be tesled for antibodies to MBP simultaneoedy, thereby makin~ this RIA technica~y simple, re-~id and ecoaomical. T I ~ assay hat, been particularly useful in examining serum sample~ from anim~!~ with experimental autoimmunc encephalomyelitis and hypomycli~osentsis con~c~ta, In patient~ with multiple scleroeis low levels of antibody to MBP can be detected in cerebrospinal fluid (CSF) and acid-eluted extracts from brain plaque material; detection of low levels of antibody in human serum has not b ~ n po:k~ibl¢ due to non.zpecific binding of human serum lg in this RIA.
C.ccmtpandee~~ be sent to: Dr. D.:;. Linthlcttm,Departmentof M~embiolc~y.US'~Schoolof Mcdic/ae, 2025Ze,ud Arena. l.m Anl~h:*,CA 90033. U.S.A. I).KL. w~ ~ by ~ p~-,loc~aJ rei~,h;p from the Nafoeal Mul~pk Sck.r,~ Society (U.S.A.). 1"~ work was wCpor~l by Itt~tt [o¢ Dr. l u Vatckayfrom the NaUoeal Health a~d Mt4ical ltlJesr¢~~ M ,,U~jCs11~=,~Ik Ibm~urdlGfsat~ Committee o{ ,k* Au~mdJan ~ TI~ work w m ~ by Grants 887-C-4 ~ I0022-A-7[n:~ t~ Nat~0Al M,,1~pk Sckro~ Sa:kW (USA) m d gnats from the N l ~ M ~ ~ d M ~ d Rt~ea~.-hCoJ~/! ~ the Nafio,ad
* l'e*t4ecto~ R_,v~.._..~F,:13owc4 , ~ Nmloe.,l Multiple Sderom Society (USA). thin,ere ~ : l)q,mmw~ of Micmbk,.qy, USC Scrod of Medki~, 2073 Zonal Avew~, Lo~.~ CA ~03~, U.S.A. 0165-~72' ~I/~(~O-CO00/$02.SO © F-J~/Nonh-HoU~l Biof~'dical Presa
18
Introduction Although ;mtoimman¢. dernyelinating diseases of the central nervous system (CNS) appear to be primarily due to cell-mediated mechanisms, the ancillary pathogenic role of humoral antibody to CNS antigcms rfmaln~ to be elucidated. ',n experimemal autoimmune encephalomyelitis (EAE), both cellular and hnmoral .~mmunity to myelin and myelin basic protein (MBP) ~re evoked but the detection of low levels of antibody to MBP usually requires sensitive techniques. A number of radioimmunoassays (RIA) have been devised for this purpose. Whitaker and McFarlin (1977) compared several and discussed their advantages and disadvantages. Randolph et al. (1977), devised a solid phase RIA employing glass tubes coated with MBP. The a~T~,,mt vf anti-MBP antibody in sera from rats with EAE was quantitared using [t2~i]rabbit anti-rat IgG (affinity purified) as the radiolabeL While this assay is an intprovement over previous methods for the detection of antibodies to MBP, it involves the manipulation of glass tubes and preparation of affinity-purified second antibody for radiolabelling. Obviously. if large ntlmbers of sera from several species are to be assayed for the presence of antibodies to MBP this would Lnvolve labor-inlensive~ costly and time-consuming procedures. n~l-radiolabelled protein-A ([nSl|ProA) from Staphylococcus aureus has been proved to be an excellent tracer for general use in immunoassays (Goding 1978; Langone 1978). Radio-lodination of ProA is easily accomplished and the labelled product remains stable for several weeks. The ability of [~251]ProAto react directly with the F¢ region lgG antibody from several species in a solid phase assay has been convincingly demonstrated (Langone 1978). Furthermore, [nSI]ProA can react with the Fe portion of an "indirect" or "second" antibody ligand; thereby making it a useful tracer ~ven when it does not bind directly to the primary antibody Fc i~giou. The method is similar to that used by Marier et al. (1979) for the assay of antibody to fungal ant~.~3~ns. Rec~.ntly several groups have employed the Pro-A technique to assay human sera and CSF for antibodies to myelin basic protein (Bashir and Whitaker 1980; Panitch et al. 1980). In this report we describe the development and use of the solid-phase RIA using 96-weli mierotiter plates coated with MBP and [IZsl]ProA to detect at~ti-MBP antibodies in the sera from animals hyperimmunized against MBP. This assay has also been employed to detect low levels of anti-MBP antibody in sera from guinea pigs with EAE and naturally ocet,rring mti-MBP antibody in lambs with Border disease aod in extracts from multil:le sclerosis brain. Materials Sheep, guinea pigs and rabbits were obtained from local couunercial sources. SJLxBalb/cFt h)brid female mice were bred from The Walter and Eliza Hall Institute (WF.HI) specific p~thogen-free stocks and used at an age of 6 - 8 wits; this hybrid is fully susce~ti~le :,, EAE (Bernard and Carnegie 1975). Purified protein-A (ProA) from Staphyl~cocsus aur,,us was obtained from Pharmacla, Uppsala, Sweden, Disposable polyviny!~:htoride "U" microti ter plates (cat, no. 1-22024) were obtained from C¢~oke Laboratb~, Products, Alexandria, VA.
Basic proteins of myelin (MBP) were prepared from human (H), chicken (C), mouse (M) and sheep (S) brains (Dankley and Carnegie 1974). A mixture of the large and small mouse MBP's was used in the experiments employing M-MBP. Methods Immunizations and antisera
Antisera to MBP were raised in sheep and rabbits by hyperimmenization over a period of 6-7 months. Sheep antisera to HMBP were raised by intramuscular injection of 2 or 3 mg MBP in Freund's complete adjuvant (FCA). supplemented with 0.75 mg M. tuberculosis. Fifty days later a booster injection o[ 1 mg MBP in FCA was given and similar booster injections were repeated every 30-50 days. Sheep were test-bled 10-14 days after eech booster injection. Rabbit antiserum to CMBP was raised acx:ording to the method of Cohen et al. (1975), by s.c. injection with 10 rag MBP in FCA, followed by booster injection of 5 mg every 30-50 days thereafter; Me f~rst in FCA and subsequently in Freund's incomplete adjuvant (FIA). Rabbit antisera to HMBP were raised by an initial injection of 250/~g HMBP in FCA; booster injections of 0.5-1.0 mg were given in FCA at 30-40 da3, intervals. Rabbits were test-bled 10-14 days after each booster immunization. Rabbit antiserum to sheep lgG was raised by injecting 1 mg purified sheep igG in FCA; a booster injection of 0.5 mg was given in FIA 30 days later and the antiserum w ~ obtained 10 days following the booster injection. Antibody from Ihis serum was affinity purified on a Sepharose-shi:ep IgG column and used for an indirect RIA technique (see below). The gamma-giobulin fraction of rabbit anti-sheep 18(3 was also purchased from Cal Biochem (California). Guinea pigs and n~3ee were immunized for the induction of EAE by standard procedures developed in our laboratory (Bernard and Carnegie 1975). Sera from newborn lambs afflicted with Border disease were kindly provided by Dr. H. Westhury OVesthary et al. 1979).
Optimization o/ the radioimmunoassay (RIA) Several parameters of the solid-phase RIA were extensively tested to establish the optimal conditions: (I) concentration and time required for antigen-coating of the Microtiter wells, including shelf-life stability of preconted trays, (2) anti-serum incubation temperature and time, (3) sp,¢ific activity and stability of the radiolabelled [t~l|ProA, (4) incubation temper#tare and time for the [*251]ProAlabelling, and (5) reproducibility o! titration end-points for specific antisera. This report describes optimal conditions which were establish,,,d for this RIA.
Labelling o/proteins Pro~inA and HMBP were radioiodinated with 12Sl using the c.'doramlne-T method (Hunter and Greenwood 1962) by Mr. 3. Pye and Mr. D. Quillci of the WEHI radioisotope laboratory. Several radioiodinations were performed on 25 p.lt ProA reaultinll in specific activities of I0, 30 and 50 ~Ci//~8; after several trials it was found that a qsecifie activity of 30 #Ci/~tg was optimal and the majority of the
2O data presented in this paper are based on a [t25I]ProA label with a specific activity of 30 /tCi/#tg" Storage at 4°C of [t2~l]ProA for gre, tcr than 6 weeks resulted in apparent radiation dam~ge and denaturation of ProA. especially at specific activities 30 ~Ci/~tg, hence only i:=l]ProA which was less than three weeks old was used. To compare with [:2eI}ProA radiolabelling, in one experiment an affinity purified sheep anti-mouse Ig antibody (antisera prepared by Dr. J. Goding) was used which was t251-radiolabelled with specific activity of 10 ltCi//tg"
Antigen cvati~,gof wells [t25I]HMBP (! ngi was added to varying amounts of unlabelled HMBP (1-10 pg) and the amount o¢ MBP hound in each well (4 h incubation at room temp) was calculated to be 0.65/~g/10 Itg MBP. Since no saturation of the MBP binding was observed., an arbitrary concentration of 10 itg/well (in 100 ~1) was chosen for the coating pro~Mure. That this concentration was effective was confirmed by using a rabbit antiserum to HMBP. Different dilutions of this antiserum, between 1 : 10 and 1 : 10,000, were tested with wells coated with HMBP between I ng/well and 100 /~g/welI. With wells coated with 5-25 ttg there was no increase in the amount of rabbit antibody which was bound from each antiserum dilution.
Recommendedprocedurefor RIA The RIA was performed in 2 stages: (1) coating of MBP to the 96-well microtiter plates and (2) titration of antisera and addition of the [~251lProA radiolabeL To coat the trays with antigen, 100 ~1 of borate-buffered saline (BBS), pH 8.5, cont~ning MBP (100/~g/ml) was added to each well of a microtiter tray and incubated at room temperature for 4 h. The trays were then washed 5 times with a gently flowing stream of distilled water. Bovine serum albumin (BSA), 1% ~n 200 itl BSS, was added to each weil and incubated for 30 min at room temperature; this step was found to be necessary to "saturate" any sites which may lead to non-specific binding of immunoglobulins. After this step, the trays were rinsed in distilled H20 as above and stored dry until used (plates may be stored dry at room temperature for up to 8 wks). To perform the serial dilutions of the sera to be tested, 100 Itl of diluent (1% BSA-BSS plus 0.05% Tween 20) was added to each well of the tray. In preliminary experiments it was determined that the presenee of Tween 20 increased antibody binding to MBP and greatly reduced the non-.~pecific binding of normal Ig to the well surface. Serial dilutions (log 3 or logs) using 50 ~tl or 25 #1 of sera were made. The trays were usually incubated at room temperature for 4-6 h and washed 5 times in distilled water. With most sera tested (human, guinea pig, mouse, sheep and rabbit) incubation at room temperature for 1-2 h proved to be sufficient. Recent experiments have proved that washing the plates in B~S with 0.5% Twecn 20 is more effective in r~:lucing the non-specific binding. [ml]ProA (30,000 cpm) was added in 100 ~tl diluent to each well and incubated overnight at room temperature. After removal of the excess [~2sl]ProA solution by suction and 5 washes in distilled water, the individual wells were cut off with a hot wire, placed in the plastio Wasserman tubes and assayed for bound |25I by gamma sp~trosoopy. After counting the wells were discarded and most of the plastic tubes can be reused after checking that all the
125I has been removed. With each serum tested, a similarly trea~.ed but MBP-free plate was rm: in parallel with the MBP coated plate to determine die non-specific binding. The amount of anti-MBP antibody in the test serum i~ exprebsed as the amount of I~l epm bound in each well. The data pre.zented in the figuxes are expressed as the mean cpm + standard deviation fo: each set of triplicate serint dilutions (inverse log 3 or logs) of serum t~ted. The calculated dilution which expressed 335 of the maximum cpl~l bound was chosen as the titration end-point (Zollinger et al. 1976).
Rendts Rabbita;ffiseratoMBP Sera from 3 rabbits hyperimmun~.zed with HMBP were tested for anti-HMBP antibodies on HMBP-coated microtiter trays. The binding of antibodies from rabbit N o . 15 anti-HMBP and normal rtdibit sera on I'~MBP
i : t ~ l - ,ta,tTl-~cj~ t , ~
}
Fig. I. 'riu'atioe of ~
from a nocmalrabbit (O) ~
from • rabbit immunizedwith HMBP(0) o~
sadte~-fme adcmftcr p l * m ( . . . . . . ) ~ tll~aP.emted p l a ~ ( ). Amoeat of l m l ] ~ o A iJ u ¢pm boead to the wells of the microtlte¢ platel ~,tota) (J~lllh'aA added was 30,000 ¢pm).
F/S. 2. T ~ ot oocaudrabbi (A, A) left~ a~l serumfromt r~Ibbit~ oa na/~t4~ ix~a c~mi ~,/daHM~P (@)c¢ CMBP ( O ).
w/th H ~ P (O. O)
22 CMBP, This difference in titer could be due to differences in the determinants. There are numerous changes in amino acid sequence when the CMBP and HMBP are co~/.pa.,c~ (Carnegie, unpublished). Rabbit No. 5 was immunized with CMBP and seca from serial bleedings were tested on CMBP- and HMBP-coated plates. The titration endpoints as determined by RIA on the test blecdings over a 400-day period are presented in Fig. 3. The peak riser on day 80 was i. 800,0t)0 for CMBP and 1:10,0/30 for MMBP. For some unexplained reason the antibody level fell dramatically after day 110 even though booster injections of CMBi' were still being administered. This decline in liter in this period was confirmed by a less sensitive precipitation assay (Biggins etal. 1978).
Sheep antisera to MBP Sheep immunized with HMBP were assayed for anti-MBP antibodies by 2 methods. The direct RIA technique employing solely [IzSI|ProA as the h'acer, yielded low levels (2,000-1,000 ¢pm) of binding because ProA only binds to sheep IgG2 (Ooudswaard et al. 1978). When an indirect technique was used employing a "sec~ id antibody" (rabbit anti-sheep lgG) incubated at room temperature for 4 h in RIA diluent, fo[!owed by |lZsl]ProA, high levels (9,000 cpm) of bi~ding were ~detectable (Fig. 4). However. the non-specific binding of normal sheep serum varied considerably from experiment to experiment and a titration of the rabbit anti-sheap IgG "second antibody" was performed to obtain the optimum dilution, with the highest concentration of the affinity purified rabbit anti-sheep IgG (10 p g / m l = I pg/well), there was a high non-specific binding on HMBP-coated plates. The lowest concentration of rabbit anti-sh¢~ lgG tested (0,1 ~g/ml -~ 10 ng/well) gave the best results" high binding of l'~iieroA and low levels of non-specific binding for normal serum. Using this indir¢~ technique, anti-MBP titets 30 days after the initial immanizations were 1 : 78,000 re: both sheep No. 2 and No. 12. Following 2 booster immunizations given at 30-day intervals, the titers imreased to 1:356,000 and 1:410,000 for sheep No. 2 and No, t2, respectively. Lambs with Border disease (hypom/elinogenisis congenita) induced experimentally, have been shown to have antimyalin antibody by a fluorescence assay (Patterson e t a l . 1977). In preliminary experiments with scra from lambs with naturally occurring Border disease (We.~tbury et at. 1979) binding of antibody to SMBP was observed in 2 lambs with typical symptoms (Fig. 5),
Sera from guinea pigs and mice with EAE Blood samples were obtained from guinea pigs and mice with clinical and histological EAE for measurement of anti-l~BP antibody levels. Since p2si]ProA binds IgG I a~d IgG 2 sub-classes of guinea pig immunogiobulins (Forsgsea 1968) high serum tite~ I'1 : 2,700- ! : 4,700) were easily detected in animals with EAE. As observed by previous investigators (Lennon et at. 1971; Berna.'d et at. 1976) there was no correl,~tion between serum antibody levels detected and the expn.~don of cfinical signs of disease, Detection , f serum anti-MMBP antibodies in mice proved to be a difficult problem. Using the direct [*2SllProA binding assay we dilaaJvered that tern from mice witl~ EAE and normal mice had high levels of non-specific binding to both
~-. to-3
/
'; 1oo
C~v5 ~TFm ~
;¢o
~oo
4oo
ft.~EC4K~
]:iS-3. Sequential mtmplcsof semm from • rabbit immt:+~.edv~th CMBP are assayedwith CMBP and MMBP. An end-point of 33~ maximumcour.i~ ~.unds was tak.~ in ca]cu]atingthe t~lerof antibody. MBP-coated and MBP-free plates. VeT.y tittle difference in the [~z~l]ProA cpm bound was observed between normal s e r u m , : d EAE serum. When an affinity-purified sheep anfi-mou~ I 8 (polyvalen 0 which ",/as radiolabelled with t~! (10 pCi/'pS ) was used again there was little difference between the normal serum (4,400 cpm) and EAE serum (4,500 cpm); both had high non-specific bindit~.~ to tmcoatad wells which was equal to ,.hat for MBP-coatecl wells. The use of an indirect rabbit anti-moase Ig antibody followed by [t~l]ProA tracer yielded sir,tim" results; normal mouse serum had high levels of n o n - s p e c i e binding thereby "masking" the detection of any specific binding due to antibody. In previous studies very low levels of antibody to MMBP were detected in mice injected with MMBP or mouse spinal cord ~ t e (Bernard and Carnegie 1975) but when rat ;viBP was used as the immunogen antibody was readily detected. Wbea samples of these sen~ .~':ra were t e l ~ d on wells coated with MMBP followed by ['~l]ProA simi]~ levels of antibody were detected. The indirect technique with mouse sera is advisable since only some of the IgG classm bind (KronvaU et al. 1970; Mackenzie et -~. 1978). Recent e x i t s , not described herein, have suggested that this high nonspecific binding is due to mouse lgM mid IgG (aggregates) which ate formed during
~sl. & P.llhlnccm~t o( b/nd/nl oi' [ '~IIPmA to dM~pIntibcdy to HIv~P (di~l) by the pr/or ~ a ~ of h ~ with a tecond m~body, viz. r~bi! anfi-,!~,~plgG (iadh'~t). Pil. 5. I M ~ d shmp Ii to SttlIP/n ICln~m~ I IL.'cp im.tmmiz~with ~ ~ P (&), ~ ~ ~ dlxm+ (O) md nonm/dk~p (41). Cpm tufa u dsc [ '~ IIProA stlw+bal to the microtiu~ wells ~ t ~ l~c.__.b-,_'_-~+"+/th r#~blt mini-sheepIs (indirect ~ I O C ) .
the plate washing phase; plates washed in BBS with Twce~ 20 have a significantly reduced "background" level of non-specific binding found in normal sera. Human sera, cerebrospinal fluid and brain extracts
With normal human scra in contrast to other species much higher levels of non-speclfic binding to MBP-free wells was found, Various alternatives to BSA coating were tried but as yet no satisfactory agent or method has been found to reduce the non-specific binding to the plates, which makes it impossible to detect low levels of antibody in sera from patients with neurological disease. In contrast, with cerebrospinal fluid and extracts of normal brain little difficulty was experienced in titering out the non-specific binding. The assay is proving particularly useful in studying the presence and purification of immunoglobulin which binds HMBP from brains of patients who have had multiple sclerosis. While it is possible that binding through the Fc region can account for some of the lgG bound in normal brain extracts it appears that the IgG extracted from multiple sclerosis brains is specific antibody to MBP (Bernard et at. 1981). Discussion
The development of a solid-phase radioimmunoassay on microtiter plates for the detection of antibody to MBP offers many advantages over previous assays. Plates can be prepared in advance and stored for long periods, washing of the 96-wells can be done much more rapidly than with assa3,s using centrifuge tubes, and the ¢~ostis considerably less. [t2~i]ProA has the advantage that it can be used as a reagent for detecting antibody from a number of species. Alternative techniques involving the use of a radiolabelled second antibody would require, considerable time and exposure to high levels of 1251in their preparation. The disadvantage that ProA only reacts with the Fc region of certain classes of IgG in some species (Kornvail et al. 1970) can be easily circumvented by the indirect procedure where a rabbit antibody to the primary antibody is used. Since ProA binds readily to rabbit IgG high levels of [t251]ProA can be bound to the wells. The indirect technique is recommended with species such as mice and sheep where" low levels of ProA bind to IgG. This solid-phase RIA could prove useful in studies on antibody to MBP in humans and animals witl~ neurological disease. Its use with human sera is at present limited by the failure to find a mixture of proteins to minimize the non-specific binding of normal human 18(3 to the polyvinyleldoride wells. When our method was developed reports from Panitch et ai. (1978, 1980). indicated that using a similar assay they were able to detect antibody to MBP in CSF in some patients with multiple sclerosis. With extracts from brains of ~adents who had multiple sclerosis the assay has demonstrated the presence of IgG which binds MBP in plaques and in normal appearing white matter (Bernard et el. 1981). In light of the work of SindJc et al. (1980), describing the binding of aggregated IgG and soluble immune complexes to MBP. caution must be exercised in evaluating the sigalflcance of IgG which binds to MBP. Indeed. the low level of non-specific binding we ohm.fred in normal le~a (e.g. see Fig. I) may be due to aggregates of lgGi ultrtceattitugation of sent before testing is helpful in removing IgG aggregates and thereby reducing the non-specific
25 b i n d i n g . I n a d d i t i o n , the u s e o f detergents s u c h as T w e e n 20 o r h i g h m o l a r s a l t c o n c e n t r a t i o n s (0.5 M N a C I ) h a v e p r o v e d useful in r e d u c i n g non-specific l g binding..
A ~ t s The authors are grateful to Dr. Robia Anders and his helpful ad~ce on the use cf radioimmunoassay techniques. They also thank Mrs. Anne Wilson, Mr. John Pye and Mr. Dennis Quflici for technical ~ss;stance.
IRefeeenees Bashir, R~M. ~ J.N. Whilaker, Mok'cul~urfe.atut~b of imnlunorea¢tive MBP in CSF in persons wlth MS, Ann. Neut,,, 7 (1980) 50. Bernard. C.C~A. an3 P.R. Carnegie, Experh~entai autoimmune enceph~omyelifis in mice-- lmro .~ologic response to n3ouse spinal cord and myeli.-abasic proteins, J. lmmunol.. 114 (1975) 1537-154~+. Bernard, C.C.A., LR. Mackay, S. Whittinghc,-'n and P. Brous, Durability of immune protection gain.,. expethnentul autoimmone encephalom?¢iit+s. CeIL Immuno!., 22 (1976) 297- 310. Bernard, C.C.A~. P.R. Carnegie, V. Randel] and I.P~ Macklt), Purification •f :t|,toanlibody to myelin basic protein from multiple sclerosis. Brain, (1981) Submitted for publicati~r~. Biggi~. J.A.. A. Taylor and EA. Caspa~j, Circulating antibody to myelin basic protein in relapsing. remitting multiple sclerosis, J. NeuroL Neurosurg. psychJaL. 41 (1978) 1131-1134. Cohen, S.I~, G.M. McKhalm and G. Guarnieri, A radioimmtmoe.~say for myelin basic protein and its use for quantitative mcamrenk~ts, J. Neutocbem., 25 (19751 3"~I- 376. Duakley, P.R. and P.R. Carnegie, Isohtion of myelin basic proteins. In: N. Marks and R. Rodkuighl (Eds.), Re~trch Methods in Ncurachemistry, VUI. 2, Plenum press. New York, 1974, pp. 219-246. Fotagren, A., protein A f¢om Sfaphylococcus aureu~ V. Reactio~l with guinea pig y-globulins. J. |mmunol., 100 (19~g)921-926. Godi~l, J.W., Use of s t a p h y ~ pl~[einA as an immunolol~calreagent..I.[mmuno]. Methods, 20 (1978)241-253. GoodtwlMd, J.,J.A. Van der Denk, A. N~rdzij, el aJ.,ProteinA reactivityof variousmammalian immunofJobulins,Scand.J. Immtmol.,8 (1978)21-28. Hunter,W . M and F.C, Grecnwoed, Pt~paratlonof iodine-f31labelledh u m ~ growth hormone of high specificacli,Aty,Nanm: (Lond.),194 (1962)495, Kronvall, G., ll.M. Grey and R.C. Williams, Protein A reactivity with mouse immunoglobulins--Str~turai relatiomhJp between son-,¢ mouse and human immunoglobulins. J. lmmunoL, 105 (1970) 1116-1123. Langone, JJ., [ I~sllProtcin A - - A tracer for genertl use in immunoassay, J. Immunol. Methods, 24 (1978) 269-2B5. V.. S. W'hloingham. P.R. Carnegie. T.A McPherson and I.R. Mackay, Detection of antibodies to the ~lt~e prote~ of human myelin by redioimmunoassay and immunofluorescenc¢. J. [mmtmol., 107 (1931) 56-g2. Mackew~, M.R., N.L W~ner and G.F. Mitcbell, Tli¢ binding of routine immuanglobulin,s to staphylococcul ~'otein .% 3. Immunol., 120 (197g) 1493-1496. Marier, R.. M. Jansen and '~.T. ,4mdfiule, A new method for mc~unng antibody using radlolabeled ptO~.n A in • sulid-phese r~oimmtmo~uty, J. immunol. Methods, 28 (1979) 41-49. Pa,~',,=h, H,S., D.A. Ha/tier tad g . E Jolunon, Antibodies to myelin basic protein in mu|tipk sclcrosis-C'Enkul comdatiom. Het~oloty (Minneap.). 25 (1978) 3 % Pamtdl, H.S., CJ. Hepp~ and K.P. Johnme, CSF tnsibody to myelin basic protein. Arch. NeuroL (Chic.). 37 (IS~O) 206-209, Pstt~oa, D.S.P,. D. SweakW and J.W. Harking, AJati-myelin arJlibodies in the serum of hunb~ w~th e ~ ihicder ~ , J. Neurod~m., 29 (1977) 394. I ~ D.H,, R.F. Kibler and R,IIL Fnte, Sulid-phas¢ radioimmtmo~tay for detection of antibodies to myqdia buic p¢oto~ J. Immm~l. Methods, Ig (1977) 215-224.
26 Si~li¢, C.J.M., C.L. Cambiaso, P.L. ICatt~ a~l E.C. La~.rre, The ~ of myelin basle prote~n to the Fc re~o~ of aiw'~la~d I~G and to unmone complexes, Cl~.Exp. lmmum~.. 4 (19~) I-7. Sindic, C.J.M., C.L. Cambilao, P,L, Mauon and E.C. IAtlerr~ The dution of iSG from sub~-.ute sckrosing panencephllilh And multiple sclerosis braim, Clin. Exp. Inmmn~., 41 (1980) 8-14. Westbury. HA,. D.V. N~o thJne and E. Su'sube, Bc~der dise.~¢-- l~-wisten t infection with the v~rus, Vet. Record, 104 (1979) 406-409. Whltaker, J.N. and D.F_ McFarlin, A comparison of immmaochemk~ incthod,vfo~ the detection of antibodies w myelin encephalitoge~Jc protein, Brain Re~. 129 (1977) 121-128. Zoningcr, W.D., J.M. Dltl~mple and M.S..Arte~in, .An~ysis of pmmacl¢~ affectln~, the ~ pha~ radiolmmunoa.~ay quarltitation of antibody to mcnln~ococca] antiltcns. J. [mmunoL, 117 (1976) 1788-1798.