Three-site sandwich radioimmunoassay with monoclonal antibodies for a sensitive determination of human alpha-fetoprotein

Journal of lmmunologtcal Methods, 58 (I 983) 293-300 Elsevier Bmmedlcal Press

293

Three-Site Sandwich Radioimmunoassay with Monoclonal Antibodies for a Sensitive Determination of Human Alpha-Fetoprotein Motozuml Nomura, Mitsunobu Imm, Kazualo Takahashl, Tomoko Kumakura, Katsuml Taclubana, Shoko Aoyagt, Sadakazu Usuda, Tetsuo Nakamura, Yuzo Mlyakawa and Makoto Mayurm * Immunology Dtvtsmn, Jtcht Medwal School, Tochzgt -Ken 329-04, the Second Department of Internal Medwme, Shmsyu Umverssty, Matsumoto City 390, Japan Red Cross Blood Center at Saltama, Sattama-ken 362, Institute of Immunology, Tokyo 113, and the Thtrd Department of Internal Medwme, Umverstty of Tokyo, Tokyo 113, Japan (Received 13 August, 1982, accepted 15 November 1982)

Utlhrang monoclonal antthodms agmnst human alpha-fetoprotem, 3 distract antigemc deternunants were identified These anugemc deterrmnants, provaslonally designated a, b and c, were arranged m such a manner that the binding of one deternunant with the corresponchng antibody did not mlublt, or only barely lrdublted the bmdmg of antibodies directed to the other 2 deternunants Monoclonal antthodms with 3 different speofmmes were, therefore, apphed to develop a sandwich-type sohd-phase radmlmmunoassay of the antigen m wluch wells were coated w~th ant~-a, and radmlabeled antt-b together w~th radmlabeled anti-c was employed to detect the bound anUgen The 3-s~te sandwmh radmlmmunoassay revolving 3 different deternunants gave a lugher sens~tivatythan 2-s~te assays m wluch only anu-b or anti-c was employed as a radmlabeled reagent, because the radioactiv~ty of the 2 labeled antthodms was added on the antigen bound to ~mmob~hzed anti-a Key words human alpha-fetoprotem - - monoclonal anttbodles - - sohd-phase tmmunoassay

Introduction Human alpha-fetoprotem (AFP) has attracted an increasing amount of attennon because of its undisputed clnucal lmphcatlons lughhghted by the assocmtlon with hepatocellular carcinoma and developmental errbrs (Ruoslahtl and Seppala, 1979) All currently avmlable methods for the measurement of AFP rely on its anUgemclty,

* Please address correspondence to Makoto Mayurm, M D , Immunology Dwtsion, Jiclu Medmal School, Mmanukawaclu-Maclu, Toclugl-Ken 329-04, Japan 0022-1759/83/0000-0000/$03 00 © 1983 Elsevier Scmnce Pubhshers B V, Amsterdam

294 such as immunodlffusion (Avelev et al, 1963), agglutination inhibition (Cahill et al, 1974) and radlolmmunoassay (Ruoslahtl and Seppala, 1971b) Hybridoma technique for the production of antibody of predefined specificity represents a recent major breakthrough in the field of immunological research (Kohler and Mllsteln, 1975) It has been succcessfully applied for the determination and analysis of a number of proteins and polypeptides with biological significance Antigenic deternunants recognized by monoclonal antibodies may be used for an elegant determination of polypeptldes Uotila et al (1981) produced monoclonal antibodies directed to 2 different antigenic determinants on AFP, and developed a 2-site sandwich immunoassay We identified 3 distinct antigenic sites on AFP and used these in the development of a '3-site' radlolmmunoassay Because 2 distinct antigenic sites were still available for the binding with labeled antibodies, the 3-site assay gave a higher sensitivity than 2-site assays in which only 1 site is detected by the labeled antibody Materials and Methods

Preparatton of monoclonal anttbodws Monoclonal antibodies against AFP (anti-AFP) were produced as described elsewhere (Nomura et al, 1982) in accordance with the method of O1 and Herzenberg (1980) Briefly, AFP in the ascites fluid from patients with hepatocellular carcinoma was purified by the method of Nishl and Hlral (1972), emulsified in Freund's complete adjuvant, and gtven lntrapentoneally to B A L B / c mice The immune spleen cells were harvested and hybridized with mouse myeloma (NS-1) cells Hybrid cells producing antl-AFP were cloned and propagated in the peritoneal cavity of truce which had been made ascltlC by means of tetramethylpentadecane (Pnstane) Ascltes fluid was harvested and T-globuhn fractions were precipitated with 1 33 M (NH4)2SO 4

Spectfictty of monoclonal antl-AFP Specificity of monoclonal antl-AFP was deterrmned by the following procedure Wells of a polyvmyl naxcrotlter plate were coated with purified AFP (75 /~g/ml, 50 /~1) The plate was washed and flooded with 50% normal rabbit serum in order to prevent non-specific binding The plate was washed extensively with saline Serial 3-fold dilutions of culture media containing monoclonal antl-AFP with different speclflClties (2051, 2062 and 2063) were prepared in phosphate-buffered (0 015 M, p H 7 2) sahne (PBS) containing 1% bovine serum alburmn (BSA), and a 100 btl portion was delivered to the AFP well The plate was incubated at 37°C for 3 h, and then washed with saline Thereafter, 50/~1 of PBS were added, supplemented with 1% BSA and containing 5 ng of antl-AFP with different speclflctties, labeled with 125I (5 #Cl//~g) by the chlorarmne-T method (Greenwood et al, 1963) The plate was further incubated at 37°C for 3 h and washed with sahne Wells were cut individually and the radioactivity in each well was counted in a y-counter The specificity of monoclonal antl-AFP was determined by compeUtlve binding to the AFP well of non-labeled and radlolabeled antl-AFP

295

Sandwwh radwzmmunoassay of AFP Sandwich immunoassays employing combinations of monoclonal antl-AFP antibodies of 3 different speclhclUes were performed on a single concentration of A F P For flus purpose, wells of a nucrotlter plate were coated with monoclonal antl-AFP (OD2s 0 = 0 1) by incubating at 4°C overnight, and washed with saline Then, the plate was flooded with 50% normal rabbit serum and washed wtih saline A solution contmmng A F P (100 /~g/ml, 25 pl) was dehvered to the well The plate was incubated at 37°C for 4 h and then washed with saline Thereafter, 200 pl of PBS contmning 1% BSA and 5 ng of 125I-labeled anti-AFP were added, and the plate was incubated at 37°C for 3 h The plate was washed with saline, the wells cut out and the radioactivity in each well counted Two-site and 3-site sandwich lmmunoassays of A F P were carried out as follows Serial concentrations of AFP ranging from 0 1 n g / m l to 1 p g / m l were prepared in PBS contaimng 5% BSA, and a 100 pl port~on was dehvered to a well of a tmcrotlter plate coated with monoclonal anti-AFP After incubating the plate at 37°C for 3 h, the plate was washed with PBS containing 1% BSA Then, radlolabeled antl-AFP in 200 pl of PBS supplemented with 1% BSA was added to each well, and the plate was further incubated at 37°C for 3 h The plate was washed with PBS containing 1% BSA Individual wells were removed and the radloacUwty in each well was counted

Results Three spectftcltws of monoclonai antt-AFP anttbodws Monoclonal anU-AFP antibodies were classified into 3 groups based on the 3 different antlgemc specificities of A F P they recognized, wluch were designated a, b and c, respecuvely The results of 3 representative antibodies (monoclonal 2051, 2062 and 2063) in tests for the specificity are depicted in Fig 1 As can be seen m Fig 1A, the binding of radiolabeled antl-a (2051) was inlublted by premcubatlon of lmmobihzed A F P with non-labeled anU-a, but was not affected by that with non-labeled anti-b (2062) or antl-c (2063) Likewise, the binding of radiolabeled anti-b was not inlublted by the premcubation of A F P with anti-AFP antibodies directed to the other specificmes (Fig 1B) The binding of anti-c (2063) was completely mlubited by prelncubaUon with non-labeled anti-c In addlUon, it was blocked to a hrmted extent by prelncubation with anu-a and very slightly by that with anu-b (Fig 1C)

Sandwwhmg of AFP between two monoclonal antt-AFP antlbodws with different speclficttws An attempt was made to sandwich A F P between 2 monoclonal anti-AFP antibodies Table I gjves the results of 9 combinations of 3 antl-AFP anubodles used for coating wells or as radlolabeled reagent It is apparent that A F P could be detected by a combination of 2 antl-AFP antibodies with different speclfiCitles A F P could, however, not be sandwiched between antl-AFP antibodies of homologous specificity, indicating that only one each of a, b and c sites was available on an intact A F P molecule

296 100

(A) 1251-2051

50

0 100

(B) f2gI-2062

0-0~ "0-

2062 "XQ\

5c

1251-2062 ",% r2~1-2062

100

z~ ,,,.

20/6?

(C] 1251-2063

/~/ ,," '2si-~oe2

zx

50

L I I I T I T 0 1 2 3 4 5 6 DILUTION OF CULTURE MEDIUM

I 7 3 -N

0f

03

f ng

3

10

30

100

300

I pg

CONCENTRATION OF ALPHA-FETOPROTE/N WEIOHTlrnl

Fig I Three speclficltles of monoclonal antl-AFP antibodies Wells of a polyvlnyl ImcroUter plate were coated with AFP Serial 3-fold dduttons of culture medmm containing monoclonal anubody (2051 2062 or 2063) were incubated m AFP wells Wells were washed wzth sahne and then tested for binding with (A) IZSl-2051 (ann-a), (B) 1251-2062 (anu-b) and (C) 12sI-2063 (anu-c) Sohd circles and sohd hne represent 2051, open orcles and dashed hne 2062, and open tnangles and dotted hne 2063 The binding of 12s1-2051 to AFP was inhabited by the premcubatlon of AFP with non-labeled 2051 but not by that with non-labeled 2062 or 2063 Llkewse, the binding of ~251-2062 to AFP was mh~b~ted only by the premcubaUon of AFP with non-labeled 2062 The binding of 1251-2063 was completely inhibited by non-labeled 2063, and only partially by non-labeled 2051 or 2062 Fig 2 Comparison of 2-rote and 3-stte radlolmmunoassays of AFP Senal concentrauons of AFP were added to wells coated with 2051 (anti-a) The bound AFP was detected by 5 ng (2 × 104 cpm) of 1251-2062 (anti-b, open circles and dashed hne), and by 5 ng (2 × 104 cpm) of 1251-2063 (anU-c, open tnangles and dotted line), as well as by the mtxture of 5 ng each of t251-2062 and t251-2063 (sohd circles and sohd hne)

Three-stte sandwwh tmmunoassay of A F P wtth monoclonal antt-AFP anttbodtes T h e wells o f a m l c r o t l t e r p l a t e w e r e c o a t e d w i t h a n U - a S e n a l c o n c e n t r a U o n s o f A F P r a n g i n g f r o m 0 1 n g / m l to 1 / ~ g / m l w e r e d e h v e r e d to wells, a n d the b o u n d A F P was d e t e c t e d b y r a & o l a b e l e d a n t l - b o r anU-c a l o n e (2-site l m m u n o a s s a y ) , as w e l l as b y the r m x t u r e o f a n t l - b a n d anU-c (3-site l m m u n o a s s a y ) T h e results are s h o w n in F~g 2 T h e u p t a k e o f r a & o a c t w ~ t y d r a m a u c a l l y i n c r e a s e d b y e m p l o y i n g t h e m i x t u r e o f r a & o l a b e l e d anU-b (2062) a n d a n u - c (2063), the b i n d i n g o f r a & o l a b e l e d antx-b w i t h A F P f i x e d o n wells b y m e a n s o f a n U - a d i d n o t i n t e r f e r e w i t h the b i n d i n g of r a d l o l a b e l e d a n t l - c F o r e a c h c o n c e n t r a U o n o f A F P tested, t h e r e f o r e , the w e l l p i c k e d u p t h e r a & o a c t l w t y o f a n U - b p l u s t h a t o f anti-c, so t h a t the c o u n t b e c a m e m u c h h i g h e r t h a n w h e n a n U - A F P o f o n l y o n e s p e c l f i o t y was e m p l o y e d T a b l e II c o m p a r e s the 2-site a n d 3-rote s a n d w i c h r a d ~ o ~ m m u n o a s s a y for the d e t e c U o n o f A F P at l o w c o n c e n t r a U o n s (0 1 - 3 0 n g / m l ) W h e n r a & o l a b e l e d a n t l - b

2051 2062 2063

AnU-AFP fixed o n wells

121 5 426 2 584

34 43 48

87 5 383 2 536

10448 104 2 743

52 24 38

AFP( - )

AFP( + )

A cpm

AFP( + )

AFP( - )

1251-2062 (anti-b)

1251-2051 (anti-a)

Radlolabeled antl-AFP

10396 80 2 705

A cpm

11 859 4236 88

AFP( + )

56 48 57

AFP( - )

1251-2063 (anti-c)

1 i 803 4188 31

A Clam

An attempt was made to detect AFP using the combination of 3 monoclonal ant~-AFP antibodies (2051, 2062 and 2063) which were used for coating cells of a nucrotlter plate and as radlolabeled reagent AFP (100/~g/ml, 25/~1) was dehvered to the well, and the bound AFP was detected by rad~olabeled antl-AFP (5 ng, 2 × 104 cpm)

SANDWICH R A D I O I M M U N O A S S A Y OF ALPHA-FETOPROTEIN WITH T H R E E M O N O C L O N A L A N T I - A F P ANTIBODIES

TABLE I

298 TABLE 11 COMPARISON OF 2-SITE A N D 3-SITE S A N D W I C H RADIOIMMUNOASSAYS WITH MONOCLONAL ANTIBODIES FOR THE D E T E R M I N A T I O N OF ALPHA-FETOPROTEIN Serml concentrations of AFP were incubated m wells coated with mono~.lonal antl-AFP (2051), and the bound AFP was detected by radtolabeled antl-AFP antibodies with different speclficmes (2062 and 2063) as well as with the mixture thereof Figures represent cpm 25 I-labeled antibodies

2062 (5 2062 (10 2063 (5 2063 (10 2062 (5 plus 2063 (5

ng) ng) ng) ng) ng)

AFP concentration ( n g / m l )

Buffer

0 1

03

1

3

10

30

93 I01 114 124

164 185 183 216

423 478 460 538

1084 1 154 1079 1230

2717 3081 2378 3260

7323 8 191 7009 8575

29 31 56 61

192

336

871

2 172

5 184

14227

41

ng)

(2062) or antl-c (2063) alone was used in 2-site assays, only a shght increase in the binding of rad~oactwlty was observed by doubhng ~ts amount from 5 to 10 ng In sharp contrast, the increase m the uptake of radloactwtty was much greater by employing the mixture of 5 ng each of radlolabeled 2062 and 2063 m the 3-site assay The background binding of the rmxture of radlolabeled anu-b and antl-C did not increase appreciably compared to that of ant~-b or antt-c alone

Discussion A F P is composed of a single polypeptlde with a molecular weight of approximately 70,000 daltons (RuoslahU and Seppala, 1971a) A number of epltopes is expeted to be present on AFP, winch may be most speofically detected by monoclonal antibodies We have reported that there are two anugemc s~tes on A F P recogmzable by monoclonal antxbodles winch we provisionally designated a and b (Nomura et al, 1983) On an intact A F P molecule, a and b sites are arranged in such a manner that the occupauon of a by antx-a proinblts a further binding of ant~-a wxth AFP, but does not interfere with an addluonal binding of antl-b The two different antlgemc sites may be used m the ~mmunoassay of A F P Uotda et al developed a 2-site lmmunoassay of A F P employing monoclonal antl-AFP wxth different speclfiotles (1981) In tins paper, we described a tturd anugemc site of A F P that could be recogmzed by the monoclonal antibody, and designated ~t ' c ' Just as for a and b sites, there ~s only one c site avadable for binding of monoclonal antibodies on an intact A F P molecule Consequently, AFP was sandwxched to allow Its detection between any two of anu-a, anu-b and anU-c, A F P was not, however, detectable by antl-AFP

299 antibodies of homologous speoficity used for coating wells or as radiolabeled reagents Although the binding of radlolabeled ana-a to A F P was not mlublted by the preincubatlon of the antigen with anti-c, the binding of radlolabeled antl-C was slgmficantly ininblted by the preincubatlon of A F P with anti-a The reason for tins difference was not clear The binding of anubody to one epltope of A F P appeared to induce its conformauonal change to block parually the binding of another epltope with corresponding antibody Taking advantage of 3 distinct antlgemc sites on AFP, a new type of sandwich lmmunoassay was developed for a sensitive determination of A F P winch we designated the '3-site' lmmunoassay The method involved the binding of A F P fixed on the well of a macrotiter plate coated with anti-a, and the detecuon by the mixture of radlolabeled anU-b and anti-c Since the binding of anti-b to A F P fixed on the well by means of anti-a did not interfere with the additional binding of anti-c, an A F P molecule picked up two radiolabeled antibodies instead of the only one in 2-site immunoassays, the radioactixnty of antl-b was added to that of anti-c on A F P in the 3-s~te lmmunoassay Hybrldomas may produce antibodies to a number of different epitopes on polypeptldes If the epitopes recognized by monoclonal antibodies are situated in such a fasinon that the occupation of one site by the corresponding antibody does not interfere w~th the binding of anubodies &rected to the other sites, a 'multiple-site' immunoassay can be worked out by the same pnnciple as described here for the deternunatlon of AFP, the sensitivity increasing with the number of labeled antibodies Such a 'multiple-site' lmmunoassay with monoclonal anubodies would be particularly useful for the deternunaUon of molecules with an extremely low concentration in the serum, such as hormones Another apphcation of monoclonal antibodies &rected to different antigenic sites of A F P would be their use for the analysis of antigenic structure We have found that the a site ~s much more resistant to reduction and alkylation than the b s~te Furthermore, although there is only one a site avmlable on an intact A F P molecule, there IS another a site buried inside, winch becomes exposed after reduction and alkylatlon (Nomura et al, 1983) With the advent of the monoclonal antibody directed to the c site, slmdar studies may be planned for an immunochenucal approach to tins polypeptide with a number of biological imphcat~ons

Acknowledgements We thank Dr Y Endo of the First Department of Internal Medicine, Umverslty of Tokyo, Dr T U n u m a of the Mltsui Memonal Hospital, and Drs Y Murakaml, A Machu and Y Mitta for ascltes fluid from patients with hepatocellular carcinoma contmmng A F P in a high concentration

300

References Avelev, G I, S D Perova, N I Khramkova, Z A Postmkova and I S Irlln, 1963, Transplantation 1 174 Cahdl, J , H Cohen and N Starkovsky, 1974, Am J Obstet Gynecol 119, 1095 Greenwood, F C W M Hunter and J S Glover, 1963, Blochem J 89, 114 Kohler, G and C Mdstem, 1975, Nature (London) 256, 495 Nlslu, S and H Hlral, 1972, Blochlm Blophys Acta 278, 293 Nomura, M, M Imal, T Nakamura, Y Mlyakawa and M Mayuml, 1983, Mol Immunol m press O1, V T and L A Herzenberg, 1980, m Selected Methods m Cellular Immunology, eds B Mishell and S M SImgl (Freeman and Co, San Francisco, CA) p 351 Ruoslahtl, E and M Seppala, 1971a, Int J Cancer 7, 218 Ruoslahtl, E and M Seppala, 1971b, Int J Cancer 8 374 Ruoslahtl, E and M Seppala, 1979, Adv Cancer Res 29 275 Uotda, M, E Ruoslahu and E Engvall, 1981, J Immunol Methods 42, 11