Differential recognition of free and covalently bound polyamines by the monoclonal anti-spermine antibody SPM8-2

Differential recognition of free and covalently bound polyamines by the monoclonal anti-spermine antibody SPM8-2 Jean-Guy

Delcros ab**, Sophie Clement ‘, Vincent Thomas a, Vbronique Jacques-Philippe Moulinoux b

Quemener

b,

The reactivity of an anti-spenttine MAb GPMS-2) toward polyamines either free or bonnd to a solid surface was investigated using equilibrium dialysis and ELBA me(hods. When wlvamines were covalently linked to hy. drophi&d mic&e~plates using c&hodiimide, the MAb SPMS-2 reactkd both with spemdne and apernddine, with a higher affinity for the latter, but did not show any reactivity towards bound putrescine. In contrast, the MAb SPMS-2 reacted with all three p+uttines bound to the microtiter plates with glutaraldehyde, with an affinity in the order: putrcscine > spermidine > spermine. Equilibrium dialysis and competitive ELlSA tests showed that the MAb SPMS-2 exhibited high affinity for free spurnine and 50% and 5% cross-reactivity with free spermidine and putrescine rcspcctively. The affinity of the UAb SPMB-2 for putrescine, spermidinc and spenttine appears to depend on whether the polyaminc is free or bound. Tbe antigenicity of the polyamines differs according to the nature of their link to the solid phase. These observationsare discussedin the light of the st~ctur.4 modification produced by covalent binding of the polyamines. It is also concluded that when antibodies are used, due care has to be exercised in choosing the appropriate immunoassay for determining the specificity of antibodies directed against small haptens such as the polyaminrs. Keywor&: Polyamine: Monwlonal

antibody; ELISA: Covalent coupling; Gluuraldehyde; Carbodiimide

Abbr&r&ns: BSA, bovine scmm albumin: EDC. I-ethyl-3_(3-dimethylaminoproWI) cartodiimide: ELKA. enzyme-linked immunowrbentassay,1C,, eoncenlrationgiving50% inhibition: M.&b. monuclonalantibody;OD, optical density;TBST. Tris bulfcr salinewith Twecn 20. author. At: Froupe de Rechcrcbeen TbhirapeutiqucAmiCancireuse. URA CNRS 1529~DRED 1266, Laboratoirede ktiologieCeBulaire,Faculty dc Wdedne, 2 Avenuedu Protesseurtin Bernard,F.35043 RennesCedex,France. Tel.: (33) 99.33.69.91;Fax:(33) 9933.6S.R; e-mail: delcrosOdsprcnmsiv-iennesl.lr.

’ Corresponding

@X22-1739/95/Sw.500 1995Hsevier ScienceBY. AS rightsreserved SSDlOO22-t759~95~00114-x

1.

latmduetion

The polyamines, putrescine, spermidine and spemtine (Fig. 1) are ubiquitous low molecular weight polycations that are important for cell growth, differentiation and numerous biological events (see reviews of Heby and Persson, 1990; Seiler, 1991.1994). Antipolyamine specific antibodies provide twls for the study of their cellular and s&cellular localisation and their bioloaical functions (Quash, 1%9; Quash et al., 1973$78; Niveleau and Quash, 1979; Hougaard, 1#2). Since polyamines are very small haptens, their covalent binding to macromolecules (carriers) is necessary for eliciting an immune response, for screening the activity of specitic antibodies in solid-phase immunoassays, and also for immunohistochemical analyses. Numerous immunoassarj have been developed in the last three decades for measuring the activity and determining the specificity of antipolyamine antibodies. In those assays, the polyamines were used either in free form (radioimmunoassay, equilibrium dialysis) or covalently bound to protein or synthetic polypeptides (immunoprecipitation, enzyme immunoassay, ELISA) (Jonard et al., 1%7; Bartos et al., 1978; Fujiwara et al., 1983; Garthwaite et al., 1989: Roth et al., 1990; Schipper et al., 1991). Classically, carbodiimide was used for the synthesis of polyamine-protein conjugates in immunoassays. Recently, Fujiwara et al. (1993) reported an ELISA procedure using polyamine-protein conjugates formed by reaction with glutaraldehyde. The chemical link between small haptens and the macromolecular carrier is responsible for major changes in the structure of small haptens and their conjugation is likely to affect their immunogenicity and antigenicity (Abdullah et al., 1992: Danilova, 1994). Antibodies raised against the polyamines, spermidine and spermine showed a higher affinity for the polyamines when these made an isopeptide bond in the immunogen (Fujiwara et al., 1993; Fujiwara, 1994). For the screening of antibodies against small haptens, the nature of the carrier protein is important as well as the chemical link between the hapten and the carrier protein (Danilova, 1994). lo this report, we have compared the speci-

ficity of a monoclonal antibody (MAb) raised against spermine, as determined by equilibrium dialysis and ELISA methods. For the ELISA tests used in this study, polyamines were covalently coupled to hydrophiliied microtiter plates either with carbodiimide (ELISA-CD) or glutaraldehyde (ELISAGA).

2. Materials and methods 2.1.

Mamials

Polyaminostyrene Primaria microtiter plates were from Becton Dickinson (Grenoble, France). [l 4-‘4Clputrescine dihydrochloride, [l’&wmidine trihydrochloride and YClspermine tetrahydrochloride were purchased from Amersham (Bucks, England). Glotaraldehyde came from Fluka (Buchs, Switzerland). Unless specified, all other reagents were from Sigma (Saint-Quentin Fallavier, France).

BALB/c mice (Iffa Credo, L’Arbresle, France) were immuniscd with spermine conjugates to thyroglobulin or fetuin that were prepared using I-ethyl-3X3-dimethylaminopropyl) carbodiimide (EDC) as previously described (Bartos and Bartos, 1978). Every 10 days, the mice received one injection of the protein-spemtine conjugate: a first in the presence of Freud’s complete adju“ant, second in the presence of incomplete adjuvant and a third in the absence of adjuvant. Bled was collected by retroorbital puocture. All animal procedures were performed strictly according to institutional guidelines. Hybridoma cells were prepared by immortalising splenic lymphocytes from mice inoculated with thyroglobulin-spemdne (prepared as de.scribed above) by fusion with SpZ/OAg14 myeloma cells (American Tissue Culture Collection, Rockville, Mk USA), according to the procedures of Kiihler and Milstein (197.5). The selected clones were grown in Ultraculture serum-free culture medium from Whittaker (Walkersville, MD, USA). The MAb was purified on protein A-Sepharose using

the lmmunopure IgG purification kit from Pierce (Rockford, IL, USA). 2.3. Equilibrium

dialysb

Equilibrium dialysis was carried out using 14Clabelled haptens as described by Pinckard and Weir (1967). Aliquots of 250 ~1 of appropriately diluted MAb SPh48-2 were placed inside dialysis bags in triplicate and dialysed against 15 ml of polyamine solution in 0.14 M NaCl, 50 mM TrisHCl pH 75. After 2 days of rotation at 4°C. 100 gl aliquots from inside and outside the bags were counted. Appropriate controls for non-specific binding were included. Affinities and antibody valence were calculated from the data using the Langmuir equation as described by Eisen and Karush (1949). 2.4. Preparation

ofpolyamine-coated

ELBA

plates

Tne surface of the wells oi the microtiter plates were first covalently coated with gelatin as previously described (Thomas et al., 1993). Then, polyamines were covalently bound to the gelatin using either EDC or glutaraldehyde (Fig. 1). ELBA-CD

plates

The plates were prepared as described by Atanassov et al. (1992). Briefly, N-succinimidyl ester groups were generated on the carboxyl groups of the gelatin by treatment with 53 mM EDC, 50 mM N-hydroxysuccinimide in 10 mM NaH,PO,. A solution of 10 mM polyamine in 25 mM borate buffer pH 8 was introduced into the wells of plates with N-succinimidyl-activated ester groups. The plates were incubated overnight at 4”C, then extensively washed in Tris saline buffer supplemented with Tween 20 (0.14 M NaCI; 10 mM Tris pH 8.0; 0.1% Tween 20 (TBST)).

Gelatin coated plates were treated for 30 min at room temperature with 0.5% glutaraldehyde in 50 mM borate pH 8. After two washes in distilled water, a solution containing 10 mM polyamine, 15 mM sodium cyanoborohydride and 50 mM borate

pH 8 was then introduced into each well. The plates were incubated overnight at room temperature, then extensively washed in TBST.

100 ~1 of diluted MAh SPM8-2 or antiserum in TBST-1% BSA were incubated for 1 h at 37’C in the polyamine-coated plate. After three washes in TBST, I00 cl of an alkaline phosphataselabelled goat anti-mouse IgCXH + L) (Biosys, CompiPgne, France) diluted 2OOlXfoldin TBST1% BSA were added for 30 min at 37°C. After three washes in TBST. alkaline phosphatase ac-

tivitv was measured by the addition of a solution containing 8 mM p-nitrophenyl phosphate, SO mM M&I, and 1 M diethanolamine pi-L 9.8. Gptica~deusity (OD) measurements were obtained at 405 nm using a Titertek Multiskan MCC/340 microreader. For inhibition tests with free polyamines, the test compounds were mixed with the MAb SPM& 2 in TBST-1% BSA prior to the addition to the titration plate. In these experiments, a dilution of the MAb SPM8-2 was chosen that gave 59% binding on the ELISA plate. After 1 h incubation at 37”C, the plates were processed as described above. The concentrations of the test polyamines inhibiting the binding of the MAb SPM8-2 by 50% (IC,) were determined from the inhibition curves aaalysed by mathematically fitting expertmental points to a four-parameter logistic equation using a commercial software package (ASOtT, BioMetallics, Princeton, USA). Crossreactivity was calculated as (IC, spermine/lC,, test polyamine)x1OO (Garthwaite et al., 1993). 3.

BesuIts

II. ELBA bindbzgtests The MAb SPM&Z (IgGl,rc) was initially selected for its ability to bind spermine in an ELISA-CD test. Fig. 2 shows that in the ELISACD assay, the binding curves of the MAb SPM&2 usinasoermine and spermidine were similar, while no binding was observed with putrescine. In the El&A-GA test (Fin. 3), the MAb SPM8-2 reatied similarly with all three polyamines. Murine polyclonal antisera taken from mice immunised against spermine expressed similar reactivities toward immobilized spermine and spermidine in both ELISA-CD and ELISA-GA assays (Fig. 4). The antisera showed very little reactivity against putrescine in the ELISA-CD assay while they reacted avidly with putrescine in the ELISAGA test. 3.2. IX&4 inhibiriontests The relative affinity of the MAb SPMS-2 for the different polyamines, either free or bound, was determined using competitive ELISA assays.

IN’

10+

10.’ *nt,bcdy ,,,I: o-3

Rg. 2. Reactivity al the monodonal antibody SPMS-2 as measured in the ELBA-CD binding test. Polyamines were bound to hydrophilized miccmiter plates using EX as a crosslinking agent (unsubstituted gelatin-coated plate, 0; putretine, n ; spermidine. l; spamine. 0). The weltsxvere incubated with various quantities of the antibody SPMIS followed by phosphatax alkaline-labelled anti-mo~use IgCXH +L) (l/2000). Error bars: standard error from triplicates (missing error bars are tired by the wnbols).

The ability of a given free polyamine to inhibit the binding of the MAb SPM8-2 on ELISA plates with different bound polyamines was first deter-

k

1.6 -

Fig. 3. Reactivity of the rnonoclonal antibody SPMB-2 as measured in the ELISAGA binding test. Folyamincs were bound to hydrophilized microtiter plates using glutamldehvde as a cross-linking agent (unsubstitutcd gelalincoated r.:stz, q:puirexine, n : spemddine, q spamine, 0). The wells were incubated with various ouantities of the antibcdy SPMB-2 followed by phosphatase alkaline-labelted anti-m&e IgCXH +L) 11/2C!40). Error bars: standard ermr from triplicates (missing error bars are obscured by the symbols).

2.5 1

and-rpmnine antisaa in

Fig 4. Reactivity of murine ELBACD and ELISA-GA binding fem. Putrerine.spermidineor

mined. Table 1 shows that 6-7 times more free spemkte, spermidine or putrescine were necessary to inhibit the binding of the MAb SPM8-2 to spermidine than to spermine bound to the ELISA-CD plate. Table 2 shows that, in the ELISA-GA test, 4.6 and 1.6 times more free spemke were necessary to inhibit the binding of the MAb SPMS-2 to bound putrescine and spermidine respectively than to bound spermine. Similar results were obtained when free spermidine or putrescine were used as competitors. The relative ability of the free plyamines to inhibit the binding of the MAb SPM8-2 to ELBA plates with polyamines bound using EDC or glutaraldehyde was then compared. Cross-reactivities of the MAb SPM&2 determined on ELISACD plates with bound spermine are shown in Table 1: the MAb SPMS-2 exhibited higher affinity for spemxine and showed 48% cross-reactivity with spermidine and 5% with putrescine. Similar patterns of cross-reactivity were determined when

Inhibition of the reactivitv of the antibodv SPMS-2 as measured in the ELBA-CD

inhibition tcsl

The ELBA-CD plates with bound polyaminer were incubated with the andbody (32 sg/l for ELISA-CD plates) in the presence of various concentrations of free poiyamincs. Thewells were then incubatedwithpbosphatase alkaline-lab&d anti-mouse k&H + L) ft/2wo). The concentrations giving Xl% inhibition UCSO) were determined from the inhibition CUIYCS.Values are the mean of ,riptica,e dererminadonr

Table 2 Inhibition of the reactwity of the antibody SPMS-2 as mearured in ELBA-CA Bound hapten

Putrescine

Putrescine Spennidine Spcrmine

inhibition test

Free competitor Spermidine

Spemdne

fC,,

%Crass-reactivity

IC,

%Cros-reactivity

IC,

%Cross-reacdvity

3.6 x 10-d 1.4x lo-’ 7.8X 10-s

5.6 4.9 5.5

4.0 x 10-5 1.5 x 10-s 1.1 x 10-S

50.0 46.0 39.1

z”xlo-5 6.9X IO-” 4.3 x 10-e

1w 1w 100

The ELISA-GA plates with bound polyaminer were incubated with the antibody 03 pg/lj in the presence of various concentrations of free potyamines. Tblhewells were then incubated with pbosphatase alkaline-labelled anti-mouse IgGIH + L) fL:ZOW. ‘The concentrations giving 50% inhibition UC,) were determined from the inhibition EWWB. Valuw IZ the mean of tri~liwte determinationr.

ELBA-CD plates with bound spermidinc (Table 1) or ELBA-GA plates with bound putrescine or spcnoidine (Table 2) were used in the assay. 3.3. Equilibriumdialysis Equilibrium dialysis experiments were finally conducted to detertnine the absolute affinity of the MAb SPM8-2 for the different free radiolabelled polyamincs. Table 3 shows that each molecule of the hUb SPMS-2 bound two molecules of polyamines with a high affinity for spennine. The MAb SPM&2 showed 50% and 5% cross-reactivity with spermidine and putrcscine respectively.

4. Discussion The affinity of the MAb SPM8-2 for irce polyamines was examined using two different techniques, equilibrium dialysis which measures the direct binding of free radiolabelled polyamines by the antibody and an ELISA inhibition test, which measures the ability of free polyamines to inhibit the binding of the MAb SPM8-2 to solid phase immobilized polyamines. Equilibrium dialysis showed that each molecule of the MAb SPM8-2 can hind two molecules of putrexine, spennidine or spennine. The MAb SPM8-2 had a higher affinity for spermine and showed 50% cross-reactivity with spermidine and only 5% with putrescine. The ELISA inhibition tests performed on both ELISA-CD and ELISAGA plates showed a pattern of cross-reactivity

similar to that observed with equilibrium dialysis. The nature of the chemical bond linking the polyamincs to the solid surface did not inter&ire in this assay since similar cross-rcactivitics were found using ELISA-CD plates or ELISA-GA plates. Liicwise, the nature of the polyamine imtnd to the ELISA elates did not affect the determination since similar cross-rcactivities were obscwcd when ELISA plates with bound putrescine. splrmidine or spcrmine were used. Thus although the ELISA inhibition test does not permit the dctertnittation of an absolute value for the affinity of the antibody for the different polyamines, it is reliable for the determination of the relative affinity of the MAL SPM8-2 for the different free polyamincs. In order to analyst the reactivity of the lvlAb SPM&2 toward bound polyamines, ELBA binding tests were conducted. To assess the importance of the nature of the chemical linkage of the polyamincs on the solid phase, the ELISA binding test %wasperformed on microtiter plates with polyamiws bound either with EDC (as used for the preparation of the immunogen) or with glutaraldehyde. The MAb SPM&2 expressed similar binding on spermine and spcrmidine bound using EDC whereas it did not show anv bindine to putrescine. In contrast, the MAb SPM8-2 e&bited similar binding to all three polyamincs when linked using glutaraldehyde. These results indicate the absence of any correlation between the affinity of the antibody SPM8-2 for the free polyantines and its ability to bind to immobilized polyamines. This fmding corroborates numerous studies aimed at determining the validity of solid-phase immunoassays for measuring antibody affinity for haptens. The binding of an antibody to a hapten linked to a solid phase depends on numerous factors such as the relative affinity of the antibody for each haptcn and on the epitape density. These factors arc often responsible for discrepancies between the antibody affinity determined with free haptens and the reactivity of the antibody in ELISA binding tests (Lew, 1984; Dcvey and Steward, 1988; Holland and Steward, 1991). Schipper et al. (1991) described an antibody tMAb MSPM/dBS) which resembled our antibody SPM8-2 and showed similar binding

with bound spemddine

and spemdne

and nune

with bound putrescine in an ELBA-CD binding assay. In view of our results, the specificity of this antibody needs to be reinvestigated. The relative

affinity

of the MAb

SPMS-2

for

different bound pulyamines may, however, be evaluated hy comparing the concentrations of a

tures resemble those of bound pulyamines are in progress to understand how N-alkylation of polyamines affect their affinity for the antibody SPMS-2. When specific antibodies for the immunostaining of aldehyde-fued material (cells and tissue)

given free polyamine necessary to inhibit the bindiie of the antibodv to different bound

are required, the ELBA-GA test appears tno~e appropriate since the hapten will exhibit strong similarities with the hapten bound in fixed cells.

polyat&s. Such tests h&e shown that the affinity of the MAb SPMS-2 for bound pulyamincs

The determination of the affinity of an antibody specific for free polyamines cannot be used to

could be ranked as follows: spetmidine > spemdne when bound using carbudiimide putrescine > spermidine > spertnine when bound using glutaraldehyde.

determine the specificity of an antibody toward aldehyde-fried pulyamines. Although the anti-

. .

The affinity of the MAb SPMS-2 for bound polyamines differs from that for free polyamines.

spermine antibody expresses only 50% and 5% cross-reactivity with spennidine and putrescine respectively in the free form, it is obvious from our results that their immobilization using an aldehyde modifies them to ‘polyamine-like’

sttw-

The MAb SPMS-2 showed a higher affimity for bound spetidine than spermine while it expressed a higher affinity for spennine in a free form. Moreover, the MAh SPMS-2 did not show any affinity for putrescine bound via an amide bond, while its affinity for putrescine bound with

tures for which the antibody has differential afftnity. For the interpretation of immunostaining resuits, immunoassays with either free ur bound pulyamines must be performed to determine the exact specificity of the antibody towards the dif-

glutaraldehyde was higher than that for bound spemddine or spemdne. This difference in binding to putrescine was not a reflection of the specificity of the monoclonal antibody used in thii study since anti-spermine polyclonal antisera

counter in fiied

had the same hehaviout (Fig. 4). Polyaminer :rre charged molecules. The use of EDC to cross-link polyamines with the carboxyl groups of aspartic or glutamic acid residues of a protein carrier produces a loss of one positive charge (Fig. 1): GX charged terminal amino group is transformed into an amide bond. The loss of one positive charge on putrescine when bound through an isopeptide bond transforms the diamine to a monoamine. This may he responsible for the absence of antibodv binding. On the other hand, when p&mines -we au&d with glutaraldehyde under reducing conditions, they are bound by N-alkylation (Fig. 1). puiyamines bound by reaction with

Thus the glutaralde-

hyde/sodium cyanobomhydride are N-substituted ‘polyamine-like’ structures, for which the MAb SPMS-2 has differential affinity. Studies using synthetic polyamine analogs whose strut-

ferent

structures

which

the

antibody

may en-

cells or tissues.

Acknowledgements The authors gratefully acknowledge Garcia-Delcros

Mrs. M.R.

for help with the manuscript,

Dr.

N. Seiler and Dr. G. Quash for critical reading of the manuscript, and Dr. A. Niveleau and Dr. R. Dante for helpful discussions. This work was supported by grants from the Direction GCn6rale de I’Alimentation the Institut National de la Sant6 Recherche MBdicale and the Centre la Recherche Scientifique (France).

(R89/05), et de la National

de

Wder, G. and M&stein, C. (1975) Ca”~i”uour cabures of fused ce,b xneti”g antibody of predeti”ed specilcity. Nature 256,495. kw. A.M. (19841 The effect of witwe denrilv and antibe+ ier. J.‘lr”rn”““l.

Meth.5

72,171.

’

Nivcleau, A. and Ouash, G. (1979) The effect of anti-spermine antibodies on a uheat germ cell-free protein-spthesizing ~,a. FEBS Let,. 99.20. Pinckxd, R.N. and Weir, D.M. (I%71 Equilibrium Dialysis and preparatia of hapten conjugates. I”: D.M. Weir (Ed.), Handbook oi Experimental Immu”“&y. Blackwcll Scientific Publications. oxford. D. 493. sdid phase immunoassays. Wiley, Chichester. 9. 135. Eke”, H.N. and Karush F. (1939) The inlrracdo” al purified Indian Med. J. 18. 1. Quash. G., kkmnefw-Rwh. AM.. Gapolo. L. and Niveleau. A (1978) Cell membrane pahlamines: a qualitative and quanlitarive study using immunoktex and tritialed a”tipoiyamine antibodies. I”: R.A. Campbe,, et a,. (Eds.), Advancer in P&amine Research, Vol. 2. Rave” Press.

tmidasc &a I&l. J. bm,,u”ol:Melhodr 61, 21;. 101.287. Fujiwara, K. A&i, M.. Kitagawa, T. and Inwe, Y. (1993~ A “ew c-e-linked immunosorknr assay (ELKA) for studying immu~oekmka, prwedures using a” amiserum produced against spennidine as a m&l. Hisloehemisw 99,477. Garthwaite. 1.. Stead. A.D. md Rider, CC. (1989) A mono. clonal an&x&-based bnm~iluauav for the oalvamines

rhe palyamine spermine by a monoclonal antibady-based EUSA. 1. Immune,. Methods 162. 175-178. Heby, 0. and Persso”. L W-%, Molecular ge”e,,cs “I po,yami”e synthesis in euhryotic cells. Trends Biachem. Sci. 15. 153. Holland, G-P and S,rward, M.W. W91) The ,““ue”ce of @ape density on the estimation of the affinity of antibody for comp,er antigens. J. Immunal. Methods 138.245. Hougaard, D.M. tIW2) Pnlyamioe cytochemirtry: Localizado” and pasible fuactians of polyamines. I”,. Rev. cytolow 138.51. Jonard. J.. Quash. G. and Wall, R. (1967, La rp&iiicitC des anticorps a”tiw,yami”es. CR. Acad. Sci. Paris 265. ,099.

New York, 985. Ouash, G.. Grcsland, L, D&in. E. and Hwywrt, J. (1973) Antipolyamine anlibodier and gm~h. I”: D.H. Russell (Ed.). Polyaminsr in Normal and “eoplanic Grauth. Rave” Press, New York ~157. Rocb, A.M.. Delcros. J.G.. Ripe,,. J.P.. Thomas, V., Richard. J. and Quash. G. WWl A navel cwa,e”f enzyme-linked immumaavay (CELIA) for simultaneously measuring free and immune ampler bound antibodies of defined spaificily. 1. App,icatio” 10 naturally occurring anlipolyamine antibadier in human sera. J. Immune,. Methods 133, 1. Scbiwer, KG., Jonis, ,.A., Rutcen. R.G.J.. Teser, G.,. and “erha‘stad, AAJ. ,199,) Preparation and cbamc,erizaafi”” of polyclonal and monoclona, antihadies I” wlvamins. I. lmmuool. Mcliwds 136.23. Seiler. N. 099,) Pharmacologiil pmpenies of “aiunl mtyamines and lbrir depletion by h&y”fbesis iohibitors a a therapeutic approach. Pmg. Drug &a. 37. 107. Seiler. N. (1994) Formatian, catabolism and paperlies cd the natural palyamines. In: C. Caner (Ed.), The Nauropbar mamlogy of Polyamines. Academic Press, London. 9. 1. Thomas. Y.. Bergrtr&“, K., Quash. G. and Holmherg. K. W93) Hydrapbiliid and fu”ctiooa,ized microtiter plates for the site-rpecibc coupling of antigens and antibodies: application to fbe diapnorir of viral cardiac and autoimmunc diseaw. Callaids Surfacer 77. 125.