Monoclonal antibodies against the free subunits of human chorionic gonadotrophin P. Berger, R. Klieber, W. Panmoung, S. Madersbacher, H. Wolf and G. Wick Immunoendocrinology Research Unit of the Austrian Academy of Sciences and ""Institute for General and Experimental Pathology, Medical School, University of Innsbruck, 6020 Innsbruck, Austria revised manuscript received
24 October 1989
ABSTRACT
Discordant results on body fluid levels of human chorionic gonadotrophin (hCG) free \g=a\-and \g=b\-subunits under physiological and pathophysiological conditions, prompted us to raise a total of 260 monoclonal antibodies (MCA) against free hCG-\g=a\,free hCG-\g=b\, holo-hCG, human follicle-stimulating hormone and bovine luteinizing hormone; 153 MCA recognizing the human \g=a\-subunit and 28 reacting with hCG-\g=b\ were extensively analysed for their intra- and interspecies cross-reactivity with homologous hormones, and for the compatibility of epitopes recognized by them. The immunological topography of free hCG-\g=a\ and free hCG-\g=b\ was resolved by these MCA, and epitope maps were designed. Six antigenic determinants on the free \g=a\-chain (\g=a\1\p=n-\ g=a\6), clustered in three spatially distinct domains, and seven epitopes on the surface of free hCG-\g=b\(\g=b\1\p=n-\ g=b\7), could be distinguished.
INTRODUCTION
phylogenetically, biochemically and immunologically closely related human glycoprotein hormones: chorionic gonadotrophin (hCG), luteiniz¬ Each of the four
ing hormone (LH), follicle-stimulating hormone (FSH) and thyrotrophin (TSH), is composed of two noncovalently linked subunits designated and ß. These are separately encoded by at least ten genes (Boothby, Ruddon, Anderson et al. 1981; Hayashizaki, Miyai,
Kato & Matsubara, 1985; Policastro, DanielsMcQueen, Carle & Boime, 1986; Watkins, Eddy, Beck et al. 1987). Within a given species, only a single gene exists for the a-subunit which all members of this family have in common. It is therefore the hormonespecific ß-chain that determines the biological and antigenic specificity (for review see Ryan, Keutmann, Charlesworth et al. 1987). The trophic roles of the holohormones have long been recognized, whereas the physiological importance and even levels in body fluids of their free subunits are still a matter of debate (Hay,
three \g=a\-chain epitopes (\g=a\4, \g=a\5 and \g=a\6) shared between various species, which is in contradiction to the concept of immunological species\x=req-\ specificity of \g=a\-subunits. Three determinants were found to be present only on the free subunits but not on holo-hCG (\g=a\6, \g=b\6 and \g=b\7), and only two determinants (\g=b\1 and \g=b\7) were hormone-specific for hCG. Based on this information, an immunoenzymometric assay for the free \g=a\-subunit of human glycoprotein hormones was established, with a sensitivity of 1 \m=.\3pg/ well and a cross-reactivity with holo-hCG of less than 0\m=.\005%Thus this assay provides the basis for detecting free \g=a\-subunitsin the presence of extremely high levels of holo-hormones, which may assist in elucidating the role of free \g=a\-subunitsin man. Journal of Endocrinology (1990) 125, 301\p=n-\309
Strikingly,
were
et al. 1987; Whitcomb, & Sangha, Schneyer Crowley, 1988). This is mainly due to the incomplete knowledge of their antigenic proper¬ ties and, in consequence, lack of specific and sensitive assay systems. Based on antigenic surface descriptions of the four holo-hormones (Schwarz, Berger & Wick, 1986; Berger, Panmoung, Khaschabi et al. 1988; Schwarz, Berger, Nelböck et al. 1988), the present study aimed to define the immunological maps of hCG-a and hCG-ß by monoclonal antibodies (MCA), providing a basis for investigations on hormonal structure-function relationships and the establishment of improved epitope-selective immunoenzymometric assays (IEMA) for the free hCG subunits.
1985; Ozturk, Bellet, Manil
MATERIALS AND METHODS
Monoclonal antibodies Monoclonal antibodies against bovine (b) LH (NIHLH-B9), hCG and free hCG-ß (Pregnyl; Organon, Oss,
The Netherlands), free hCG-a (PAA-Laborgesellschaft m.b.H., Gallneukirchen, Austria) and hFSH (Boehringer Mannheim GmbH, F.R.G. andlmmunex, San Diego, CA, U.S.A.) were produced as described in detail elsewhere (Kofler, Kalchschmid, Berger & Wick, 1981; Kofler, Berger & Wick, 1982; Bergere? al.
1988).
Antibody characterization Direct binding radioimmunoassay (RIA) Radioiodination with Na125I (Amersham Inter¬ national pic, Amersham, Bucks, U.K.) of hCG (NIH, CR 125), hLH-I-1 (NIADDK, AFP 4345B), hFSH-I3 (NIADDK, AFP 4822B), hTSH-I-5 (NIADDK, AFP 4370B), hCG-a (NIH, CR 115A), hCG-ß (NIH, CR 115B) and hLH-ß (NIH, LER 1793B) was per¬ formed using the chloramine method of Hunter & Greenwood (1962). Working solutions of radiolabelled hormones were prepared immediately before use and rechromatographed on a 30 1 cm Sephadex G-75 superfine column (Pharmacia Fine Chemicals AB, Uppsala, Sweden). To assess MCA intraspecies cross-reactivity a double antibody RIA was applied. Dilutions of
MCA-containing
tissue culture super¬ undiluted to 1 : 1000 in plain (log10 steps; tissue culture medium; lOOpl/tube) were incubated for 20 h at 4 °C with radiolabelled hormones (100 pi; 25 000 c.p.m. in phosphate-buffered saline (PBS; pH 7-2) supplemented with 0-3% (w/v) bovine serum albu¬ min (BSA; Sigma, Deisenhofen, F.R.G.). Separation of bound from free tracer was achieved by addition of 100 pi immunoabsorbent consisting of sheep immunoglobulin G (IgG) to mouse IgG coupled to ActMagnogel AcA 44 (IBF, Villeneuve, France). This incubation (3 h on a shaking device) was terminated by three washes with 2 ml PBS/0-5% (v/v) Tween 80 natants
(Serva, Heidelberg, F.R.G.).
Competitive RIA a competitive RIA, the MCA (100 pi; diluted
In
in
tissue culture medium to 30% of maximum binding) were incubated for 20 h at 4 °C with radiolabelled hCG-a (NIH, CR 115A) or hCG-ß (NIH, CR 115B) and increasing concentrations (log10 increments) of hCG (NIH, CR 125), hLH-I-1 (NIADDK, AFP 4345B), hFSH-I-3 (NIADDK, AFP 4822B), hTSH-I5 (NIADDK, AFP4370B), hCG-a (NIH, CR 115A), hCG-ß (NIH, CR 115B) and hLH-ß (NIH, LER 1793B) diluted in 0-3% BSA/PBS. Separation of bound from free hormone was performed as described above. One-site
enzyme-linked immunosorbent assay
(ELISA)
Hormones were diluted in PBS to 1 pg/ml and adsorbed to flat-bottom 96-well ELISA-plates (Petra-
Plastic, Chur, Switzerland). Extensive washing of the plates with 0-05% Tween 20 (Serva) in PBS was fol¬ lowed by blocking of remaining binding sites with 1% (v/v) fetal calf serum (FCS) in PBS (200 pi, 30 min). MCA-containing tissue culture supernatants (diluted 1 :2 in tissue culture medium) were tested for their antigen-binding ability (90 min, 37 °C). After washing the plates with 0-05% Tween 20/PBS, bound MCA was detected by peroxidase-conjugated rabbit Ig to mouse Ig (DAKO, Copenhagen, Denmark), diluted 1 :500 in 1% FCS/PBS (30 min, 37 °C). The assay was developed by using 2,2'-azino-bis (3-ethyl)-benzthiazolinesulphonic acid (Sigma) as a substrate. Extinctions at 405 nm were recorded by an automatic ELISA reader (Dynatech, Santa Monica, CA, U.S.A.) and photographically documented (Agfa CT 64). The incubation volumes were 0-1 ml throughout the assay. Sandwich ELISA Purified MCA, prepared by ammonium sulphate pre¬ cipitation from ascitic fluid were adsorbed to the ELISA plates (50 pg MCA/ml PBS, 2 h, 37 °C). Sub¬ sequent blocking of remaining adsorption sites, extensive washing with 005% Tween 20/PBS, sub¬ strate development and recording of results was per¬ formed as described for the one-site ELISA. The antigen incubation (100 pi, 2h, 37 °C) with either hCG-a (NIH, CR 115A) or hCG-ß (NIH, CR 115B) diluted in 1 % FCS/PBS to 100 ng/ml was followed by a detection MCA (100 pi; 10pg/ml in 1% FCS/PBS, 60 min, 37 °C) that was coupled to horseradish peroxidase type VI (Sigma) according to the method of Wilson &Nakane( 1978).
Immunoenzymometric assay for the free a-subunit ELISA-plates were coated with a MCA, designated INN(sbruck)-hCG-72, which is specific for the free asubunit (50 pg MCA/ml PBS; 100 pi, 16h, 4°C). Blocking, extensive washing and the antigen incu¬ bation with hCG-a was performed as described above. The second peroxidase-labelled MCA (INNhFSH-158) was diluted to 10pg/ml in 1% BSA/PBS and incubated for 30 min at 37 °C (100 pi/ well). 3,3',5,5'-Tetramethylbenzidine (TMB; 6 mg; Boehringer Mannheim GmbH) was dissolved in 1 ml dimethylsulphoxide (Sigma). Immediately before use, this stock solution was diluted to 100 pg TMB/ml in 0-1 mol sodium acetate buffer/1 (ph 5-5) and 1-3 mmol
H202/1.
The substrate reaction was terminated after 30 min by addition of 50 pi H2S04 (1 mol/1) per well and the results read at 450 nm. In order to achieve higher sensitivity and stability of the IEMA, the capture as well as the detection MCA were purified
by high-pressure liquid chromatography (HPLC) according to the manufacturer's recommendations (Biorad, Richmond, VA, U.S.A.), using a preparative
hydroxylapatite column. Purity was checked by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (Hoefer Scientific Instruments, CA, U.S.A.) on 14% gels (Laemmli, 1970). The detection MCA was labelled to a higher specific activity as originally described for polyclonal antibodies by Wilson & Nakane (1978) using ultracentrifuged MCA (100 000#, 4 mg MCA/ ml), 12 mg peroxidase/ml, 0-2 mol Nal04 and 0-4 mol carbonate buffer/1. Immunoafunity purification of hormones Hormone preparations, hCG (highly purified, kindly provided by V.C. Stevens, Ohio State University, Columbus, OH, U.S.A.), hLH-I-1 (NIADDK, AFP
4345B), hFSH-I-3 (NIADDK, AFP 4822B), hTSH-I5 (NIADDK, AFP 4370B), hCG-a (NIH, CR 115A) and hCG-ß (NIH, CR 115B) were depleted from con¬ taminating hormones by appropriate immunoaffinity purification with MCA against hCG (Kofler, Berger 6 Wick, 1982), free hCG-a, hLH, hTSH, hFSH (Berger et al. 1988) and, for control purposes, with an antibody against human prolactin (Staindl, Berger, Kofler & Wick, 1987). HPLC-purified or ammonium sulphate-precipitated ascites-derived MCA (approxi¬
50 pg/ml) were incubated for 2 h on an automatic shaker with 1 ml goat IgG-anti-mouse Ig coupled to cyanogen bromide-activated Sepharose 6B beads (Pharmacia Fine Chemicals) prepared accord¬ ing to the method of Cuatrecasas (1970). After exten¬ sive washing with 1% BSA/PBS, the MCA-coated beads were resuspended in equal volumes of hormone solutions (up to 1 pg/ml) and incubated for 2 h. After centrifugation, the supernatant was withdrawn and the absorption procedure repeated twice with fresh MCA-coated beads. The purified standards were measured in the free a-subunit as well as in respective holo-hormone assays.
mately
RESULTS
Monoclonal antibodies A total of 260 MCA-secreting cloned hybridoma cell lines were established (designated INN-hCG-1 to 82,
INN-hFSH-1 to 181 and INN-bLH-1 to 5), some of which have been described previously (Kofler et al. 1981, 1982; Berger et al. 1988); 205 of these MCA reacted with holo-hCG and/or free hCG-ß and/or free a-subunit. Characterization of MCA Monoclonal antibody subunit specificities Three classes of MCA against any of the glycoprotein hormones could be distinguished in direct-binding RIA (Table 1): 153 MCA reacted with the radiolabelled a-subunit (a-MCA), 75 MCA reacted with the ß-subunit (ß-MCA) and 32 antibodies recognized only the conformationally (c) intact holo-hormone (c-MCA) but neither free a- nor free ß-subunits. Further subdivision of these MCA groups could be achieved on the basis of MCA intraspecies crossreactivities in direct-binding RIA (Fig. \a-c, g-f) as well as competitive RIA (Fig. \d-fi k-m). MCA were able either to distinguish free a-subunit (INN-hCG-72 and -80; Fig. \a and d) or free hCG-ß (INN-hCG-64 and -68; Fig. \b and e, and \c and/) from the holohormone, or specifically detected holo-hormones only (e.g. INN-hCG-45 and INN-hFSH-3; data not shown) or holo-hormones and their free ß-subunits (e.g. INN-hCG-2 and -22; Fig. \h and / and \j and m respectively) or all human glycoprotein hormones and their free a-subunits (e.g. INN-hCG-5, -17 and -43, and INN-hFSH-73, -132 and -158; Fig. \g and k). One-site ELISA Similar to the one-site RIA, three classes of MCA could be demonstrated with solid phase-bound unlabelled antigens: -, ß- and c-MCA (Fig. 2). MCA against the ß-subunit were specific for hCG and hCG-ß (group ß,), or cross-reacted with hLH (ßj), or additionally recognized LH of various species (ß3, ß4 and ß5) or specifically distinguished free hCG-ß-subunits from holo-hCG and holo-hLH (ß6 and ß7). The antibodies against the -chain were classified as (i) reacting with all human a-subunits and holo-hCG (a,, a2 and a3), (ii) displaying interspecies cross-reactivity (a4, a5 and a6) and (iii) recognizing only free a-subunit but not holo-hCG (a6).
Intramolecular localization of the antigenic determinants recognized monoclonal antibodies (MCA) table 1.
MCA specificity
(numbers of MCA reacting)
Holo-hormone only MCA code INN-hCG-1 to 82 INN-hFSH-1 to 181 INN-bLH-1 to 5 Total numbers
by the
a-subunit
ß-subunit
Total
19
33
13
120
25 47
3
180 3
32
153
75
260
hCG, Human chorionic gonadotrophin; b, bovine.
77
! 50 5
PQ
o-oooi o-ooi o-oi
o-i
io ooooio-ooi
·
io o-oooi
o-i
o-ooi"ö'-oi" .i
'
10
Dilution
100
10
0-0001 0-001 0-01
0-1
1
10 0-0001 0-001 0-01 01 Dilution
10 0-0001 0-001 0-01
lOO-i
100-,
5
100 0-01
1 0-1 Hormone
10
(mg/1)
0-1
1
10
figure 2. Resolving epitopes on free -subunit (right panel) and free human chorionic gonadotrophin ß (hCG-ß) (left panel) by one- and two-site enzymelinked immunosorbent assay (ELISA). Intra- as well as interspecies crossreactivities of 13 groups of monoclonal antibody (MCA) against hCG, free subunits of hCG, bovine LH and hFSH (designated ß,-ß7 and ,- 6) were determined by an indirect one-site ELISA (lower panel). Each group of MCA (vertical rows) showed a characteristic reaction pattern with immobilized hor¬ mones or hormone subunits (horizontal rows). Not all groups could be resolved by this method (e.g. ß6 from ß7 and a2 from a3), which was possible by the sandwich ELISA approach (upper panel). Antibodies were tested pairwise in a chessboard-like fashion for complementation in antigen binding (free -subunit and free hCG-ß). oLH, ovine LH; cLH, cercopithecus LH.
Sandwich ELISA Pairs of solid phase-bound antibodies and peroxidaselabelled MCA were examined for their ability complement each other in detecting free subunits (Fig.
2). They revealed characteristic patterns of reactivity; five different patterns emerged with free hCG-ß and
three with free -subunit. Positive reactions reflected and epitope compatibility of
complementation antibody pairs.
1. (a-c, g-j) Direct-binding radioimmunoassay (RIA): titration curves of monoclonal antibodies (MCA) with radiolabelled antigens. The MCA shown are representatives for groups of antibodies (designated (a) a6, (b) ß6, (c) ß7, (g) 4, (h) ß,, (j) ß2) showing identical reaction patterns in all assay systems used (one- and two-site RIA and enzyme-linked immunosorbent assay), (a-c) Antibodies recognizing only free, i.e. non-assembled subunits, but not the holo-hormone. (g) All human glyco¬ protein hormones are recognized equally well by this a-MCA. (h) This MCA exclusively recognizes human chorionic gonado¬ trophin (hCG) and hCG-ß but neither holo-hLH nor hLH-ß. The antigenic determinant is not located on the C-terminal extension of hCG-ß (data not shown). This is in contrast to the MCA shown in (j) that also recognizes hLH and hLH-ß. (d-f, k-m) MCA specificities were also defined by a competitive RIA. Again, INN-hCG-72, -64 and -68 react with free subunits only, INN-hFSH-132 can be competed by hCG and free -subunit whereas INN-hCG-2 does not recognize hLH and free hLH-ß. B,hCG; ».hCG-a; D,hCG-ß; O, hLH; , hLH-ß; A,hFSH; A, human thyrotrophin. figure
3. Epitope maps of ( ) free human chorionic gonadotrophin ß (hCG-ß) and (b) free hCG-a. The two subunits were depicted as two globules in a cylindrical Mercatore' projection. Antigenic determinants, six on the -subunit (a,-a6) and seven on the ß-subunit (ß,-ß7), are represented by concentric circles, some of which conflue to larger domains. Adjacent structures with common outer circumference are not compatible for simultaneous antigen detection. Three epitopes ( 6, ß6 and ß7) disappear upon subunit assembly, whereas conformational epitopes of the holo-hormones emerge (Schwarz et al. 1986; Berger et al. 1988). W, west; E, east. figure
table
2.
Epitope characteristics of free -subunit and free human chorionic gonadotrophin-ß (hCG-ß) Epitope sharing
Epitope present on: Holo-hormone
Free a-subunit
+
+
+ + + + +
+
+ + +
Intraspecies
Interspecies
Epitope
+ + +
+
+
+ +
+ + +
+ +
Free
hCG-ß
+ +
+
+
+ + + +
+ + +
+
+
+
+t
+ +
only on free a-subunit; tWeak reaction with radiolabelled free hLH-ß (see Fig. \b). Epitopes with apparently identical qualities are discernible from each other either by the sandwich enzyme-linked immunosorbent assay approach (e.g. a2 and a,, a4 and a5, ß3 and ß4) and/or by differences in interspecies epitope sharing (e.g. ß, and ß4, ß3 and ß5) (see Fig. 2). *Present
Epitope maps for free -subunit and free hCG-ß Monoclonal antibodies with identical specificity in and two-site RIA and ELISA were assigned to
one-
designated ,- 6 and ß,-ß7, which were single epitopes. Topographical epitope arrangement was resolved by sandwich ELISA (Fig. 2). Spatial disparity of antigenic detergroups,
defined to represent
3. Levels of holo-human chorionic gonadotrophin (hCG), hCG-a and holo-hFSH in the serum of a male patient suffering from choriocarcinoma under polychemotherapy table
hCG (ng/1)
hCG-a (ng/1)*
hFSH (ng/1)
115 210 000 133 000 25
10 100 600 110
24 October 1989
ABSTRACT
Discordant results on body fluid levels of human chorionic gonadotrophin (hCG) free \g=a\-and \g=b\-subunits under physiological and pathophysiological conditions, prompted us to raise a total of 260 monoclonal antibodies (MCA) against free hCG-\g=a\,free hCG-\g=b\, holo-hCG, human follicle-stimulating hormone and bovine luteinizing hormone; 153 MCA recognizing the human \g=a\-subunit and 28 reacting with hCG-\g=b\ were extensively analysed for their intra- and interspecies cross-reactivity with homologous hormones, and for the compatibility of epitopes recognized by them. The immunological topography of free hCG-\g=a\ and free hCG-\g=b\ was resolved by these MCA, and epitope maps were designed. Six antigenic determinants on the free \g=a\-chain (\g=a\1\p=n-\ g=a\6), clustered in three spatially distinct domains, and seven epitopes on the surface of free hCG-\g=b\(\g=b\1\p=n-\ g=b\7), could be distinguished.
INTRODUCTION
phylogenetically, biochemically and immunologically closely related human glycoprotein hormones: chorionic gonadotrophin (hCG), luteiniz¬ Each of the four
ing hormone (LH), follicle-stimulating hormone (FSH) and thyrotrophin (TSH), is composed of two noncovalently linked subunits designated and ß. These are separately encoded by at least ten genes (Boothby, Ruddon, Anderson et al. 1981; Hayashizaki, Miyai,
Kato & Matsubara, 1985; Policastro, DanielsMcQueen, Carle & Boime, 1986; Watkins, Eddy, Beck et al. 1987). Within a given species, only a single gene exists for the a-subunit which all members of this family have in common. It is therefore the hormonespecific ß-chain that determines the biological and antigenic specificity (for review see Ryan, Keutmann, Charlesworth et al. 1987). The trophic roles of the holohormones have long been recognized, whereas the physiological importance and even levels in body fluids of their free subunits are still a matter of debate (Hay,
three \g=a\-chain epitopes (\g=a\4, \g=a\5 and \g=a\6) shared between various species, which is in contradiction to the concept of immunological species\x=req-\ specificity of \g=a\-subunits. Three determinants were found to be present only on the free subunits but not on holo-hCG (\g=a\6, \g=b\6 and \g=b\7), and only two determinants (\g=b\1 and \g=b\7) were hormone-specific for hCG. Based on this information, an immunoenzymometric assay for the free \g=a\-subunit of human glycoprotein hormones was established, with a sensitivity of 1 \m=.\3pg/ well and a cross-reactivity with holo-hCG of less than 0\m=.\005%Thus this assay provides the basis for detecting free \g=a\-subunitsin the presence of extremely high levels of holo-hormones, which may assist in elucidating the role of free \g=a\-subunitsin man. Journal of Endocrinology (1990) 125, 301\p=n-\309
Strikingly,
were
et al. 1987; Whitcomb, & Sangha, Schneyer Crowley, 1988). This is mainly due to the incomplete knowledge of their antigenic proper¬ ties and, in consequence, lack of specific and sensitive assay systems. Based on antigenic surface descriptions of the four holo-hormones (Schwarz, Berger & Wick, 1986; Berger, Panmoung, Khaschabi et al. 1988; Schwarz, Berger, Nelböck et al. 1988), the present study aimed to define the immunological maps of hCG-a and hCG-ß by monoclonal antibodies (MCA), providing a basis for investigations on hormonal structure-function relationships and the establishment of improved epitope-selective immunoenzymometric assays (IEMA) for the free hCG subunits.
1985; Ozturk, Bellet, Manil
MATERIALS AND METHODS
Monoclonal antibodies Monoclonal antibodies against bovine (b) LH (NIHLH-B9), hCG and free hCG-ß (Pregnyl; Organon, Oss,
The Netherlands), free hCG-a (PAA-Laborgesellschaft m.b.H., Gallneukirchen, Austria) and hFSH (Boehringer Mannheim GmbH, F.R.G. andlmmunex, San Diego, CA, U.S.A.) were produced as described in detail elsewhere (Kofler, Kalchschmid, Berger & Wick, 1981; Kofler, Berger & Wick, 1982; Bergere? al.
1988).
Antibody characterization Direct binding radioimmunoassay (RIA) Radioiodination with Na125I (Amersham Inter¬ national pic, Amersham, Bucks, U.K.) of hCG (NIH, CR 125), hLH-I-1 (NIADDK, AFP 4345B), hFSH-I3 (NIADDK, AFP 4822B), hTSH-I-5 (NIADDK, AFP 4370B), hCG-a (NIH, CR 115A), hCG-ß (NIH, CR 115B) and hLH-ß (NIH, LER 1793B) was per¬ formed using the chloramine method of Hunter & Greenwood (1962). Working solutions of radiolabelled hormones were prepared immediately before use and rechromatographed on a 30 1 cm Sephadex G-75 superfine column (Pharmacia Fine Chemicals AB, Uppsala, Sweden). To assess MCA intraspecies cross-reactivity a double antibody RIA was applied. Dilutions of
MCA-containing
tissue culture super¬ undiluted to 1 : 1000 in plain (log10 steps; tissue culture medium; lOOpl/tube) were incubated for 20 h at 4 °C with radiolabelled hormones (100 pi; 25 000 c.p.m. in phosphate-buffered saline (PBS; pH 7-2) supplemented with 0-3% (w/v) bovine serum albu¬ min (BSA; Sigma, Deisenhofen, F.R.G.). Separation of bound from free tracer was achieved by addition of 100 pi immunoabsorbent consisting of sheep immunoglobulin G (IgG) to mouse IgG coupled to ActMagnogel AcA 44 (IBF, Villeneuve, France). This incubation (3 h on a shaking device) was terminated by three washes with 2 ml PBS/0-5% (v/v) Tween 80 natants
(Serva, Heidelberg, F.R.G.).
Competitive RIA a competitive RIA, the MCA (100 pi; diluted
In
in
tissue culture medium to 30% of maximum binding) were incubated for 20 h at 4 °C with radiolabelled hCG-a (NIH, CR 115A) or hCG-ß (NIH, CR 115B) and increasing concentrations (log10 increments) of hCG (NIH, CR 125), hLH-I-1 (NIADDK, AFP 4345B), hFSH-I-3 (NIADDK, AFP 4822B), hTSH-I5 (NIADDK, AFP4370B), hCG-a (NIH, CR 115A), hCG-ß (NIH, CR 115B) and hLH-ß (NIH, LER 1793B) diluted in 0-3% BSA/PBS. Separation of bound from free hormone was performed as described above. One-site
enzyme-linked immunosorbent assay
(ELISA)
Hormones were diluted in PBS to 1 pg/ml and adsorbed to flat-bottom 96-well ELISA-plates (Petra-
Plastic, Chur, Switzerland). Extensive washing of the plates with 0-05% Tween 20 (Serva) in PBS was fol¬ lowed by blocking of remaining binding sites with 1% (v/v) fetal calf serum (FCS) in PBS (200 pi, 30 min). MCA-containing tissue culture supernatants (diluted 1 :2 in tissue culture medium) were tested for their antigen-binding ability (90 min, 37 °C). After washing the plates with 0-05% Tween 20/PBS, bound MCA was detected by peroxidase-conjugated rabbit Ig to mouse Ig (DAKO, Copenhagen, Denmark), diluted 1 :500 in 1% FCS/PBS (30 min, 37 °C). The assay was developed by using 2,2'-azino-bis (3-ethyl)-benzthiazolinesulphonic acid (Sigma) as a substrate. Extinctions at 405 nm were recorded by an automatic ELISA reader (Dynatech, Santa Monica, CA, U.S.A.) and photographically documented (Agfa CT 64). The incubation volumes were 0-1 ml throughout the assay. Sandwich ELISA Purified MCA, prepared by ammonium sulphate pre¬ cipitation from ascitic fluid were adsorbed to the ELISA plates (50 pg MCA/ml PBS, 2 h, 37 °C). Sub¬ sequent blocking of remaining adsorption sites, extensive washing with 005% Tween 20/PBS, sub¬ strate development and recording of results was per¬ formed as described for the one-site ELISA. The antigen incubation (100 pi, 2h, 37 °C) with either hCG-a (NIH, CR 115A) or hCG-ß (NIH, CR 115B) diluted in 1 % FCS/PBS to 100 ng/ml was followed by a detection MCA (100 pi; 10pg/ml in 1% FCS/PBS, 60 min, 37 °C) that was coupled to horseradish peroxidase type VI (Sigma) according to the method of Wilson &Nakane( 1978).
Immunoenzymometric assay for the free a-subunit ELISA-plates were coated with a MCA, designated INN(sbruck)-hCG-72, which is specific for the free asubunit (50 pg MCA/ml PBS; 100 pi, 16h, 4°C). Blocking, extensive washing and the antigen incu¬ bation with hCG-a was performed as described above. The second peroxidase-labelled MCA (INNhFSH-158) was diluted to 10pg/ml in 1% BSA/PBS and incubated for 30 min at 37 °C (100 pi/ well). 3,3',5,5'-Tetramethylbenzidine (TMB; 6 mg; Boehringer Mannheim GmbH) was dissolved in 1 ml dimethylsulphoxide (Sigma). Immediately before use, this stock solution was diluted to 100 pg TMB/ml in 0-1 mol sodium acetate buffer/1 (ph 5-5) and 1-3 mmol
H202/1.
The substrate reaction was terminated after 30 min by addition of 50 pi H2S04 (1 mol/1) per well and the results read at 450 nm. In order to achieve higher sensitivity and stability of the IEMA, the capture as well as the detection MCA were purified
by high-pressure liquid chromatography (HPLC) according to the manufacturer's recommendations (Biorad, Richmond, VA, U.S.A.), using a preparative
hydroxylapatite column. Purity was checked by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (Hoefer Scientific Instruments, CA, U.S.A.) on 14% gels (Laemmli, 1970). The detection MCA was labelled to a higher specific activity as originally described for polyclonal antibodies by Wilson & Nakane (1978) using ultracentrifuged MCA (100 000#, 4 mg MCA/ ml), 12 mg peroxidase/ml, 0-2 mol Nal04 and 0-4 mol carbonate buffer/1. Immunoafunity purification of hormones Hormone preparations, hCG (highly purified, kindly provided by V.C. Stevens, Ohio State University, Columbus, OH, U.S.A.), hLH-I-1 (NIADDK, AFP
4345B), hFSH-I-3 (NIADDK, AFP 4822B), hTSH-I5 (NIADDK, AFP 4370B), hCG-a (NIH, CR 115A) and hCG-ß (NIH, CR 115B) were depleted from con¬ taminating hormones by appropriate immunoaffinity purification with MCA against hCG (Kofler, Berger 6 Wick, 1982), free hCG-a, hLH, hTSH, hFSH (Berger et al. 1988) and, for control purposes, with an antibody against human prolactin (Staindl, Berger, Kofler & Wick, 1987). HPLC-purified or ammonium sulphate-precipitated ascites-derived MCA (approxi¬
50 pg/ml) were incubated for 2 h on an automatic shaker with 1 ml goat IgG-anti-mouse Ig coupled to cyanogen bromide-activated Sepharose 6B beads (Pharmacia Fine Chemicals) prepared accord¬ ing to the method of Cuatrecasas (1970). After exten¬ sive washing with 1% BSA/PBS, the MCA-coated beads were resuspended in equal volumes of hormone solutions (up to 1 pg/ml) and incubated for 2 h. After centrifugation, the supernatant was withdrawn and the absorption procedure repeated twice with fresh MCA-coated beads. The purified standards were measured in the free a-subunit as well as in respective holo-hormone assays.
mately
RESULTS
Monoclonal antibodies A total of 260 MCA-secreting cloned hybridoma cell lines were established (designated INN-hCG-1 to 82,
INN-hFSH-1 to 181 and INN-bLH-1 to 5), some of which have been described previously (Kofler et al. 1981, 1982; Berger et al. 1988); 205 of these MCA reacted with holo-hCG and/or free hCG-ß and/or free a-subunit. Characterization of MCA Monoclonal antibody subunit specificities Three classes of MCA against any of the glycoprotein hormones could be distinguished in direct-binding RIA (Table 1): 153 MCA reacted with the radiolabelled a-subunit (a-MCA), 75 MCA reacted with the ß-subunit (ß-MCA) and 32 antibodies recognized only the conformationally (c) intact holo-hormone (c-MCA) but neither free a- nor free ß-subunits. Further subdivision of these MCA groups could be achieved on the basis of MCA intraspecies crossreactivities in direct-binding RIA (Fig. \a-c, g-f) as well as competitive RIA (Fig. \d-fi k-m). MCA were able either to distinguish free a-subunit (INN-hCG-72 and -80; Fig. \a and d) or free hCG-ß (INN-hCG-64 and -68; Fig. \b and e, and \c and/) from the holohormone, or specifically detected holo-hormones only (e.g. INN-hCG-45 and INN-hFSH-3; data not shown) or holo-hormones and their free ß-subunits (e.g. INN-hCG-2 and -22; Fig. \h and / and \j and m respectively) or all human glycoprotein hormones and their free a-subunits (e.g. INN-hCG-5, -17 and -43, and INN-hFSH-73, -132 and -158; Fig. \g and k). One-site ELISA Similar to the one-site RIA, three classes of MCA could be demonstrated with solid phase-bound unlabelled antigens: -, ß- and c-MCA (Fig. 2). MCA against the ß-subunit were specific for hCG and hCG-ß (group ß,), or cross-reacted with hLH (ßj), or additionally recognized LH of various species (ß3, ß4 and ß5) or specifically distinguished free hCG-ß-subunits from holo-hCG and holo-hLH (ß6 and ß7). The antibodies against the -chain were classified as (i) reacting with all human a-subunits and holo-hCG (a,, a2 and a3), (ii) displaying interspecies cross-reactivity (a4, a5 and a6) and (iii) recognizing only free a-subunit but not holo-hCG (a6).
Intramolecular localization of the antigenic determinants recognized monoclonal antibodies (MCA) table 1.
MCA specificity
(numbers of MCA reacting)
Holo-hormone only MCA code INN-hCG-1 to 82 INN-hFSH-1 to 181 INN-bLH-1 to 5 Total numbers
by the
a-subunit
ß-subunit
Total
19
33
13
120
25 47
3
180 3
32
153
75
260
hCG, Human chorionic gonadotrophin; b, bovine.
77
! 50 5
PQ
o-oooi o-ooi o-oi
o-i
io ooooio-ooi
·
io o-oooi
o-i
o-ooi"ö'-oi" .i
'
10
Dilution
100
10
0-0001 0-001 0-01
0-1
1
10 0-0001 0-001 0-01 01 Dilution
10 0-0001 0-001 0-01
lOO-i
100-,
5
100 0-01
1 0-1 Hormone
10
(mg/1)
0-1
1
10
figure 2. Resolving epitopes on free -subunit (right panel) and free human chorionic gonadotrophin ß (hCG-ß) (left panel) by one- and two-site enzymelinked immunosorbent assay (ELISA). Intra- as well as interspecies crossreactivities of 13 groups of monoclonal antibody (MCA) against hCG, free subunits of hCG, bovine LH and hFSH (designated ß,-ß7 and ,- 6) were determined by an indirect one-site ELISA (lower panel). Each group of MCA (vertical rows) showed a characteristic reaction pattern with immobilized hor¬ mones or hormone subunits (horizontal rows). Not all groups could be resolved by this method (e.g. ß6 from ß7 and a2 from a3), which was possible by the sandwich ELISA approach (upper panel). Antibodies were tested pairwise in a chessboard-like fashion for complementation in antigen binding (free -subunit and free hCG-ß). oLH, ovine LH; cLH, cercopithecus LH.
Sandwich ELISA Pairs of solid phase-bound antibodies and peroxidaselabelled MCA were examined for their ability complement each other in detecting free subunits (Fig.
2). They revealed characteristic patterns of reactivity; five different patterns emerged with free hCG-ß and
three with free -subunit. Positive reactions reflected and epitope compatibility of
complementation antibody pairs.
1. (a-c, g-j) Direct-binding radioimmunoassay (RIA): titration curves of monoclonal antibodies (MCA) with radiolabelled antigens. The MCA shown are representatives for groups of antibodies (designated (a) a6, (b) ß6, (c) ß7, (g) 4, (h) ß,, (j) ß2) showing identical reaction patterns in all assay systems used (one- and two-site RIA and enzyme-linked immunosorbent assay), (a-c) Antibodies recognizing only free, i.e. non-assembled subunits, but not the holo-hormone. (g) All human glyco¬ protein hormones are recognized equally well by this a-MCA. (h) This MCA exclusively recognizes human chorionic gonado¬ trophin (hCG) and hCG-ß but neither holo-hLH nor hLH-ß. The antigenic determinant is not located on the C-terminal extension of hCG-ß (data not shown). This is in contrast to the MCA shown in (j) that also recognizes hLH and hLH-ß. (d-f, k-m) MCA specificities were also defined by a competitive RIA. Again, INN-hCG-72, -64 and -68 react with free subunits only, INN-hFSH-132 can be competed by hCG and free -subunit whereas INN-hCG-2 does not recognize hLH and free hLH-ß. B,hCG; ».hCG-a; D,hCG-ß; O, hLH; , hLH-ß; A,hFSH; A, human thyrotrophin. figure
3. Epitope maps of ( ) free human chorionic gonadotrophin ß (hCG-ß) and (b) free hCG-a. The two subunits were depicted as two globules in a cylindrical Mercatore' projection. Antigenic determinants, six on the -subunit (a,-a6) and seven on the ß-subunit (ß,-ß7), are represented by concentric circles, some of which conflue to larger domains. Adjacent structures with common outer circumference are not compatible for simultaneous antigen detection. Three epitopes ( 6, ß6 and ß7) disappear upon subunit assembly, whereas conformational epitopes of the holo-hormones emerge (Schwarz et al. 1986; Berger et al. 1988). W, west; E, east. figure
table
2.
Epitope characteristics of free -subunit and free human chorionic gonadotrophin-ß (hCG-ß) Epitope sharing
Epitope present on: Holo-hormone
Free a-subunit
+
+
+ + + + +
+
+ + +
Intraspecies
Interspecies
Epitope
+ + +
+
+
+ +
+ + +
+ +
Free
hCG-ß
+ +
+
+
+ + + +
+ + +
+
+
+
+t
+ +
only on free a-subunit; tWeak reaction with radiolabelled free hLH-ß (see Fig. \b). Epitopes with apparently identical qualities are discernible from each other either by the sandwich enzyme-linked immunosorbent assay approach (e.g. a2 and a,, a4 and a5, ß3 and ß4) and/or by differences in interspecies epitope sharing (e.g. ß, and ß4, ß3 and ß5) (see Fig. 2). *Present
Epitope maps for free -subunit and free hCG-ß Monoclonal antibodies with identical specificity in and two-site RIA and ELISA were assigned to
one-
designated ,- 6 and ß,-ß7, which were single epitopes. Topographical epitope arrangement was resolved by sandwich ELISA (Fig. 2). Spatial disparity of antigenic detergroups,
defined to represent
3. Levels of holo-human chorionic gonadotrophin (hCG), hCG-a and holo-hFSH in the serum of a male patient suffering from choriocarcinoma under polychemotherapy table
hCG (ng/1)
hCG-a (ng/1)*
hFSH (ng/1)
115 210 000 133 000 25
10 100 600 110
