Molecular and Cellular Endocrinology, 80 (1991) 33-40
33
0 1991 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/91/$03.50 MOLCEL 02568
Number and topography of epitopes of human chorionic gonadotropin (hCG) are sh ared by desialylated and deglycosylated hCG S. Schwarz ‘, H. Krude I, R. Klieber 2, S. Dirnhofer 2, C. Lottersberger W.E. Merz 3, G. Wick 1 and P. Berger 2
*,
’ I~titute of General and ~~er~rne~tal Pat~oIo~~ Faculty of Medicine, ~‘m’versi~ of In~b~~k, A-6020 Innsb~c~ Austria, II, ’ Immunoendocrinolo~ Research Unit, Austrian Academy of Sciences, A-6020 Innsb~ck, Austria, ’ Institute of 3i~~e~t~ University of ~eideIberg,O-6900 Heidelberg I, F.R.G. (Received 25 March 1991; accepted 10 May 1991)
Key words: Monoclonal antibody; Epitope mapping; Microheterogeneity;
Carbohydrate moiety; Glycoprotein hormone
Summary A previously established map of the surface epitopes of human chorionic gonadotropin (hCG) served as template for the present study in which we investigated the antigenic surfaces of two glycosylation variants of hCG, i.e. desialylated hCG (asialo-hCG) and deglycosylated hCG (degly-hCG). This map allocates five epitopes to the a subunit, five to the p subunit and four +? epitopes to structures formed only by the a/P heterodimer holo-hCG (Schwarz et al. (1986) Endocrinology 118, 189-197; Berger et al. (1990) J. Endocrinol. 125, 301-309). Here it is described that both variants complied with this template: each of the 14 distinct monoclonal antibodies with which the epitopes of hCG were defined reacted with radiolabeled asialo-hCG and degly-hCG as well and generally bound degly-hCG with greater affinity than hCG. Moreover, every combination of capture and radiolabeled detection antibody that was either compatible or incompatible on unlabeled hCG was so also on unlabeled asialo-hCG and degly-hCG. It thus appears that alterations of the carbohydrate structure of hCG can be associated with a change in affinity between some antibodies and their respective epitopes but not with a loss of an epitope or with a change in the topographical relationships of the 14 epitopes.
Introduction Human chorionic gonadotropin (hCG) belongs to the glycoprotein hormone family which com-
Address for correspondence: S. Schwarz, Institute of Genera1 and Experimental Pathology, Faculty of Medicine, University of Innsbruck, A-6020 Innsbruck, Austria. Tel. 43 (512) 507-2266, Fax 43 (512) 507-2507.
prises in addition follicle-stimulating hormone (FSH), th~oid-stimulating hormone (TSH) and luteinizing hormone (III). They all resemble each other in molecular mass (M, 2: 301, overall conformation (elipsoid globular shape) and quaternary structure: they are non-covalently linked heterodimers consisting of an (Y subunit that is singular within a given species, and a p subunit that is unique for each hormone and that, by combination with the (Ysubunit, confers receptor
34
binding ability, receptor specificity and biological activity (for reviews, see Hussa, 1981; Pierce and Parsons, 1981; Ryan et al., 1987). The carbohydrate moieties (CHO) of hCG serve diverse functions. Intracellularly, they influence the folding, assembly competence, sorting and stability of the subunits, while extracelluIarly, the CHOs affect the circulatory half life and, more importantly, the biological activity of the secreted a/P hormone, i.e. the degree of hCG receptor activation (involving G, protein coupling, CAMP generation and steroidogenesis). Not only is the structural integrity of the CHOs important (Baenziger and Green, 1988; Sairam, 1989) but also their correct linkage to the protein backbone: while the metabolic fate of hCG through Ca2+-dependent binding to the asialogalactose receptor of hepatocytes (Ashwell and Harford, 1982) is determined by the sialic residues on all CHOs, activation of hCG receptors on target cells requires only the cYAsn-52linked CHO (Matzuk and Boime, 1989). Previously, we had probed the number and the topographical arrangement of the surface epitopes of hCG (Schwarz et al., 1985, 1986) and of the free subunits of hCG (Berger et al., 1990) by using a large panel of monocIona1 antibodies (MCA) prepared in our laboratory (Kofler et al., 1981, 1982; Berger et al., 1984,1988, 1990). Fourteen surface epitopes could thus be distinguished which are designated according to their localization as al to (~5, 81 to /35, and apl to a/34 (cw/?or c-epitopes are present only on the complete hormone; some may arise from two incomplete epitopes as shown by Bidart et al., 1990). Knowledge of the antigenic structure of hCG permitted us, first, to construct immunoassays of predictable specificity for holo-hCG or either of its subunits (Schwarz et al., 1985; Berger et al., 1990), and secondly, to study the relationship of these antibody epitopes with receptor-interaction domains (Schwarz et al., 1988). Here we asked whether or not desialylated hCG (asialo-hCG) and deglycosylated (degIy)hCG differ from native hCG (i) as to the number of epitopes, (ii) as to the affinities with which these epitopes are recognized by the MCAs, and (iii) as to the topographical relationships of these epitopes.
Materials
and methods
Preparation of hCG, desialylated hCG and deglycosylated hCG and iodinated derivatives
Purified hCG was prepared from crude hCG (Schering, Berlin, F.R.G.) by a modification of the procedure of Canfield et al. (19’71), as described in detail elsewhere (Merz et al., 1974). From this preparation ~12,000 IU/mg protein), desialylated hCG was obtained by digestion (1618 h, 37°C) with neuraminidase (EC 3.2.1.18; from Clostridium perfringens) immobilized on agarose beads (type X-A, Sigma Chemical Co., Deisenhofen, F.R.G.) in a buffer containing 50 mM sodium phosphate and 50 mM citric acid pH 5.6 according to Cassidy et al. (1966). The thus obtained asialo-hCG was desalted on a Biogel P-30 column (Bio-Rad Labs., Munich, F.R.G.) equilibrated with 50 mM NH,HCO, and Iyophilized. Deglycosylated hCG was prepared (Merz, 1988) by treating purified hCG in a closed polyterafluoroethylene (Teflon) apparatus with anhydrous hydrogen fluoride (HF) for 1 h at 0°C. After the HF had been removed under vacuum, degly-hCG was recovered in 0.5 M NH,HCO,, and after an overnight incubation at 37°C subjected to gel filtration (Biogel P-60 column 1.5 X 66 cm eluted with SO mM NH~HCO~). Purity and preservation of receptor binding activity of these three preparations were assessed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with silver or immunostaining and by radioreceptor displacement assays, all done in comparison with CR-123 as reference standard (kindly provided by the National Pituitary Agency, Baltimore, MD, U.S.A.). lZSI-labeled derivatives of hCG and the variants were prepared as described (Schwarz et al., 1986, 1988). One-site binding protocol MCA titration assay. Each of the 14 different
MCAs (stock: l-10 mg protein/ml) was serially diluted with 0.05 M phosphate buffer containing 1% (v/v) bovine serum albumin (PBS/BSA). Aliquots of 0.1 ml were incubated in parallel with 25,000 cpm of ““I-hCG, ‘251-asialo-hCG or “‘Idegly-hCG, respectively (0.1 nM, in a total volume of 0.3 ml PBS/BSA), for 12-16 h at 22°C. Unbound iz51-antigen was separated by the dou-
35
ble antibody-immunosorbent procedure (for details see Berger et al., 1990) and bound ‘251-antigen was measured. Production, characterization, purification (by ammonium sulfate and/or highperformance liquid chromatography (HPLC)) and radiolabeling of these MCAs has been described previously (Kofler et al., 1981, 1982; Berger et al., 1984, 1988, 1990; Schwarz et al., 1985, 1986, 1988). MCA displacement assay. Aliquots of 0.1 ml of MCA, appropriately diluted so as to bind < 50% of ‘251-hCG added (as determined in the foregoing MCA titration assay), were incubated with 25,000 cpm “‘1-hCG in the absence or presence of increasing concentrations of unlabeled hCG (in 0.2 ml PBS/BSA) for 12-16 h at 4°C. Conditions for separation of bound and free ‘251hCG were as above. The inhibitory concentration 50% (IC,,) of a given MCA was determined using the computer program ALLFIT (DeLean et al., 1972). Two-site sandwich protocol
Polystyrol tubes were coated with = 10 pg capture MCA (= 300 nM IgG), and after washing, incubated for 2 h at 37°C with 1 nM hCG or asialo-hCG or degly-hCG, respectively (1 nM was found optimal in preliminary dose-response experiments). Control tubes received only PBS/ BSA. After washing, = 0.5 nM (= 200,000 cpm) of ‘2”I-labeled detection MCA (from which nonspecific binding material was removed by ultracentrifugation) was added. After an incubation of 4 h at 37°C and washing, radioactivity bound to tubes was measured. This approach provides an all-or-none type of response: a positive reaction signals sterical compatibility between two epitopes, a negative result (mutual exclusion of binding) signals identity or steric proximity, as outlined in detail previously (Schwarz et al., 1985, 1986). Results
At low dilution (1: loo), a given MCA could bind 100% of each of the added 1251-antigens, i.e. native, asialo- and degly-hCG. This binding was titratable as shown by diluting the MCA further up to 1: l,OOO,OOO. As can be seen in Fig. 1, the titer (dilution of MCA that would yield 50% of
111000 1110,0001/100,0001/1000,000
dilutionof antibody Fig. 1. Binding of ‘“I-hCG (Xl, “‘1-asialo-hCG (squares) and ‘251-degly-hCG (+I as a function of dilution of MCA (dilution being indicated on the abscissa). As an example, INN-hCG-2, a MCA specific for the pl-epitope, is shown. Note the rank order in titer: degly-hCG > asialo-hCG > hCG. Cpm bound (B) are expressed as a percentage of total (T) radioactivity added. Each point is the mean of duplicates of a typical experiment repeated once. Curves were generated by the computer program ALLFIT (DeLean et al., 1972).
tracer binding) of this particular PI-MCA was = 25 or = 20 times higher for ‘Z51-degly-hCG than for ‘251-hCG or ‘251-asialo-hCG, respectively. All 14 MCAs were tested in the same way and found to be able to bind ‘*‘I-asialo- and ‘251-degly-hCG as well. In addition, each of the 14 MCAs was analyzed by a displacement assay using 1251-hCG in the presence of unlabeled hCG
0
-10
-9 -8 log [hCG], Ii4
-7
-6
Fig. 2. Binding of ‘%hCG to five different P-epitope-specific MCAs (Bl, /32, p3, /34, @S) as a function of increasing concentrations of unlabeled hCG. Cpm bound (BI are expressed as a percentage of cpm bound in the absence of hCG (B,). Each point is the mean of triplicates of a typical experiment repeated once. Curves were generated by the computer program ALLFIT (DeLean et at., 1972).
36 TABLE
I
APPARENT DEGLY-hCG
AFFINITIES
Epitope specificity and designation of MCA (Yl cu2 a3 a4 ru5 a&2 a/33 QP4 Pl P2 P3 P4 P5
INN-hFSH-73 INN-hFSH-8 INN-hFSH-15 INN-hFSH-132 INN-hFSH 158 INN-hCG-10 INN-hCG-40 INN-hCG-45 INN-hCG-26 INN-hCG-2 INN-hCG-22 INN-bLH-1 INN-hCG-24 INN-hCG-58
(K,)
OF
MONOCLONAL
ANTIBODIES
(MCA)
FOR
NATIVE
K, (liter.mol~‘) Native hCG a
hCG,
ASIALO-hCG
Rank order Asialo-hCG
b
Degly-hCG
AND
’
b
( + SEM) 6.9 + 0.58 x 3.3 + 0.29 x n.d. 2.6 k 0.27 x 1.0+0.09x10s 1.3f0.11X10s 1.5~0.15x10’ 6.4 + 0.54 x 2.4+0.25x 2.5f0.26x107 1.1f0.09X10’ 8.6 f 0.70 x 1.4~0.13x10s 3.1 _to.31 x
10h 10” 10s
10h lox
10’ 10s
1.1 x 10h 1.5 x 109 0.2x 0.3 x 1.2x 1.4x 3.5 x 2.0 x 4.7 x 1.8x 10.0x 2.1 x 3.5 x
10s 10” 10s 10’ 10h 10s 10’ 10” 10’ 10s lox
1.6x lo6 3.3 x 10’ 1.0x 2.5 x 2.4 x 2.7 x 53.0x 6.1 x 57.0 x 3.2 x 72.0 x 7.8 x 39.0 x
10s 10s 10s 10’ 10h 10s 10’ 10’ 10’ 10s 10s
n>d>a n=d>a d>n>a n>d>a d>n>a d>n>a d>n>a d>n>a d>n>a d>a>n d>a>n d>a>n d>a>n d>a>n
aKaoc,;, = 1/Ics,y~,,~
h Kacvarlantj = Ka(hCC).[titer(,,,,,,,) /fiter,t,d ’ Native hCG, n; asialo-hCG,
a; degly-hCG,
d.
(Fig. 2). The reciprocal of the herefrom determined IC,,, was taken as an estimate of the affinity (K,) of a MCA for native hCG. The apparent K, of a MCA for the variants was calculated by multiplying KachCGj with the relative titer, i.e. titer(variant/ titer(,,,,. These values are summarized in Table 1. Three different patterns of affinities with respect to native (n), asialo(a) and degly-hCG (d) emerged: (i) n 2 d 2 a, shared by the crl-, a2- and a4-MCAs; (ii) d > n 2 a, shared by the (r3- and a5-MCAs and all a@-MCAs; and (iii) d > a > n, shared by all /3MCAs (as exemplified in Fig. 1). Since all 14 epitopes, previously detected on hCG, were thus found to be present also on asialo-hCG and degly-hCG (Table 11, we asked whether these variants would share also the topographical relationships of the epitopes with hCG. Therefore, each of the 14 MCAs was tested also in a two-site binding protocol in one or the other configuration (i.e. as capture or as detection MCA). Since the epitope map of hCG was already known from previously performed complete 14 x 14 chessboard matrix experiments (Schwarz et al., 1985, 1986; Berger et al., 19901,
the present question required retesting of only the 42 most critical and informative MCA combinations (omitting all self-by-self and all reciprocally identical ones). As shown in Table 2, every of the 26 pairs of MCAs predicted to be positive (compatible), turned out to be so on the variants. The same was true for the 16 pairs predicted to be negative (incompatible). These data are summarized as an epitope compatibility table in Fig. 3. Discussion Glycoprotein hormones are of similar general architecture (Pierce and Parsons, 1981). Different laboratories arrived at similar numbers and topographical distributions of surface epitopes when investigating hCG (Schwarz et al., 1985, 1986; Bidart et al., 1987; Berger et al., 1990), hLH (Alonso-Whipple et al., 19881, hTSH (Benkirane et al., 1987) or hFSH (Berger et al., 1988). The carbohydrate moieties (CHO), representing = 30% of the total molecular mass of hCG, constitute hydrophilic, hence surface exposed portions of the molecule. In order to investigate the possi-
37
bility whether CHOs influence the epitopes that also are on the surface of hCG, two glycosylation variants of hCG were mapped, i.e. asialo-hCG and degly-hCG. The former retains most of its potency at the hCG receptor (Moyle et al., 1975) whereas the latter is a hCG receptor-bindable ligand but, in terms of CAMP response, a competitive hCG antagonist (Moyle et al., 1975; Chen et al., 1982; Sairam and Bhargavi, 1985; Baenziger and Green, 1988; Matzuk and Boime, 1989; Sairam, 1989). In addition, while native hCG displays negligible crossreactivity, asialo-hCG considerably crossreacts (as an antagonist) with the TSH receptor (Carayon et al., 1981) and deglyhCG (as a partial agonist) with the FSH receptor (Ranta et al., 1985). The principal finding of this study was that asialo-hCG and degly-hCG share with native hCG the same 14 epitopes previously identified on hCG. The very same MCAs could recognize also every band of different pl obtained by isoelectrofocussing of a microheterogenous preparation of hCG (Mann et al., 1989). Asialo-hCG and deglyhCG (and most, if not all, other glycosylation variants) are antigenically quite similar and thus targets for anti-hCG antibodies. This has several practical consequences: first, advantageous ones as to detection of hCG producing tumors by immunoassays. Secondly, disadvantageous ones as it becomes clear that immunoassays generally overestimate the actual bioactive concentration in a given sample, in contrast to bioassays and ‘bioimmunoassays’ (Moyle et al., 1988). The use of agonist-specific diagnostic assays becomes compelling in view of the recent detection of natural FSH ‘antihormones’, likely to be equivalents of degly-FSH (Dahl et al., 1988). Furthermore, one has to recognize that if antibodies occur in vivo, they are likely to react with all glycoforms of hCG. Such antibodies may be passively administered (treatment of ectopic pregnancy: Frydman et al., 1989) or may arise upon application of asialo-hCG and degly-hCG (Liu et al., 1989) or hCG-toxin conjugates (‘hormonotoxins’) (Singh et al., 1989). If one of the 14 epitopes is formed by a particular CHO, the absence of the CHOs should lead to abolished binding of the respective MCA. Since this was not the case, all the 14 epitopes of hCG
appear to be shaped predominantly by the protein backbone, as has been suggested also by others (Bidart et al., 1987). This explains on the single epitope level why polyclonal antisera, produced against native hCG, recognized degly-hCG with similar affinity as hCG (Chen et al., 1982; Manjunath and Sairam, 1982; Rebois and Fishman, 1984) and vice versa (Rebois and Liss, 1987). Thus, while degly-hCG is antigenically not different from hCG, it may still be immunogenically different. Rebois et al. (1984) and Sairam (1989) have discovered such degly-specific antibodies (of P-specificity) within an hCG-absorbed antiserum raised against degly-hCG suggesting that this protein backbone epitope may have been concealed by the CHOs on hCG. Yet per se, the CHOs seem to be of little immunogenicity in mice in contrast to CHO antigens in other contexts, e.g. blood group or tumor-associated antigens (Feizi, 1985). A second finding of this study was that each of the MCAs recognized its epitope with affinities that were often quite different depending on whether the epitope was measured on native hCG or on the variants. The prevailing pattern, however, was that degly-hCG was bound by 11 of 14 MCAs with greater affinity than native hCG. Similar was observed by Moyle et al. (1988) and Sairam et al. (1988) with a single MCA. We do as yet not know the relationship of those 14 epitopes to the primary structure of hCG and thus to the known attachment sites of the CHOs. Degly-hCG is devoid of all the complex-type N-linked sugar antennae except for the basal Gal-NAc residues (Chen et al., 1982; Manjunath and Sairam, 1982; Keutmann et al., 1983, 1985; Kalyan et al., 1985; Blithe, 1990) but still contains the O-linked CHO. These, however, reside on the C-terminal extension (Baenziger and Green, 1988) while all ‘our’ five p epitopes localize in the core of the p subunit (Kofler et al., 1981). The here observed behavior of degly-hCG relative to hCG suggests that the CHOs compromise as a whole the binding energy of a given MCA sterically (by partially shielding an epitope), i.e. by their size, length and hydration. The behavior of asialo-hCG (lacking only the terminal sialic acid residues) relative to hCG suggests that there may be also electrostatic repulsion for certain MCAs, e.g. the (Y-and (Y@-
38 TABLE
2
TWO-SITE
MONOCLONAL
ANTIBODY
BINDING
TO hCG, ASIALO-hCG
AND DEGLY-hCG
Results in each row represent ‘251-MCA binding to either hCG, asialo-hCG or degly-hCG whereby the cpm bound were expressed as A% to control (buffer) in that particular combination of capture and detection MCA. Compatibility or incompatibility between two MCAs was predicted from the previous mapping studies on native hCG and its subunits (Schwarz et al., 1986; Berger et al., 1990). A result was classified as being positive (compatible combination) when the cpm of 1251-MCA bound were at least 30% (A%) higher in the presence of antigen than in its absence; as can be seen at the bottom of the table, positive responses were on average between 384 and 696% higher, while negative results (incompatible combinations) had A% values often below control values and on average between 1.9 and 3.4. Compare also each detection MCA in a compatible and an incompatible combination! Epitope
specificity
Capture MCA 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
(Yl ff3 a2 u2 a4 a4 Pl Pl PI Pl Pl Pl 02 P2 P2 P2 P2 P2 P4 P5 P5 al34 al34 o/34 @4 op4 al al al al Cul Crl a2 a3 Pl Pl P2 P2 02 P3 ;:
All positive ones (mean f SE) All negative ones (mean f SE) a Not done.
of Detection MCA
Predicted compatibility
Native hCG
AsialohCG
DeglyhCG
58 129 1537 86 83 350 559 704 65 216 31 502 845 978 603 527 166 89 1607 1355 697 1064 1038 48 817 117 14 7 -6 4 -11 19 9 2 - 15 13 17 4 - 13 - 12 15 -3
150 238 812 75 82 280 533 696 56 N.D. a N.D. N.D. 614 642 377 302 115 N.D. N.D. N.D. N.D. 858 710 30 643 87 4 13 - 17 -4 3 1 3 17 N.D. N.D. 11 -22 18 4 13 N.D.
998 486 1772 94 260 177 659 876 129 466 132 2679 823 934 534 329 122 128 1714 1164 498 813 685 134 691 98 7 4 -6 - I2 - 10 -6 10 21 - 15 16 18 12 -4 - 18 7 7
(hCG)
+ + + + + + + + + + + + + + + + + + + + + + + + + + _ _ _ _ _ _ _ _ _ _
549 *96 2.8+ 2.9
384 k66 3.4+ 3.4
696 +120 1.9* 3.1
39
-EPlTOPE COMPATIBILITY
TABLE
a2 a3
+
+
a4
-
-
a5
+
+
apI
-
-
+
-
+
up2
-
-
+
-
+
ap3
-
-
+
-
+
-
-
cl!34
+
+
+
+
+
+
+
pl
++++++--+
+ +
-
+
82
+
+
+
+
+
+
+
+
+
+
p3
+
+
+
+
+
+
+
+
+
+
-
p4
-
+
+
+
+
+
+
+
+
+
-
-
ps
++++++++++--a4
a5
PI
p2
p3
al
a2
a3
ap1 apz
ap3 aa4
p4
Fig. 3. The epitope compatibility table was obtained from the reductional analysis of all the results collected from full-scale chessboard two-site MCA binding studies on hCG and its subunits (Schwarz et al., 1985, 1986; Berger et al., 1988, 1990). Incompatible (adjacent or overlapping) epitopes are desig(distinct and distantly located) nated by a ‘-‘, compatible epitopes by a ‘+‘. Identical compatibility patterns were obtained by testing asialo-hCG and degly-hCG, as shown in this study (Table 2), indicating that the number as well as the topographical relationships of the epitopes of hCG are shared by these glycosylation variants of hCG. An epitope map can be constructed from this table by applying two rules outlined previously (Schwarz et al., 1986): (i) the radius of an epitope is practically identical with thf radius of an antibody’s antigen binding domain, i.e. = 10 A (Novotny et al., 1983), and (ii), the minimal spatial requirement for sterical compatibility of two MCAs is that their respective epitopes are at least = 20 A apart.
MCAs, but also attraction, e.g. for the P-MCAs (Table 1). The final question was whether these 14 epitopes are distributed in the same manner as on hCG. We reasoned that, if modifications on or removal of the CHOs would drastically alter the overall conformation of hCG, such would be accompanied by dislocation of epitopes rather than disappearance, so that two epitopes that are adjacent (incompatible) on one form become ‘smeared’ apart and thus compatible on the other form, and vice versa. Either possibility should be detectable by a change in the patterns obtained by the pairwise MCA binding strategy which has, indeed, revealed conformational changes on other globular proteins, e.g. albumin (Lapresle, 1988). As shown in Table 2, every pair of MCAs previ-
ously found on hCG to be compatible or incompatible was so also on asialo-hCG and degly-hCG suggesting that the relative distances of these 14 epitopes from each other are not changed, at least not to an extent that can be detected by the here utilized all-or-none response-type approach. The same was true for hLH (results not shown). It thus appears that both variants (including hLH) occur in an overall conformation that, while not being completely identical, is rather similar to that of native hCG. This similarity still allows for discrete differences in conformation between hCG and degly-hCG as had been proposed from differences in UV absorbance spectra (Manjunath and Sairam, 1982; Merz, 19881, in fluorescence spectra (Manjunath and Sairam, 19821, in circular dichroism spectra (Keutmann et al., 1983; Kalyan and Bahl, 19851, from differences in subunit-subunit association competence (Kalyan and Bahl, 1985) or subunit-subunit dissociation or peptide bond cleavage susceptibility (Merz, 1988), or from differences in antibody affinity (Keutmann et al., 1983, 1985). Yet, the extent of those changes, while being apparently big enough to account for quantitative differences in MCA binding affinity, as seen also in this study (Table 11, is without consequence for the antigenic properties in qualitative terms, i.e. number and topography of epitopes (Table 2, Fig. 3). Definite answer of this issue will await the availability of crystal structure data of hCG in comparison to degly-hCG (Lustbader et al., 1989). In conclusion, while the CHOs markedly influence the metabolic fate as well as receptor selectivity and expression of agonist potency and efficacy, they appear to be, by comparison, of relatively small relevance for the antigenic surface characteristics of hCG. Epitopes on a globular protein appear to be quite robust structures. Acknowledgements
This work was supported in part by the Austrian Academy of Sciences. We thank Ms. Regine Gerth, Renate Goerz, Irene Gaggl and Barbara Mayer for their skillful technical assistance and the National Pituitary Agency in Baltimore, MD for the generous gift of various glycoprotein hormone preparations.
40
References Alonso-Whipple, C., Couet, M.L., Doss, R., Koziarz, J., Ogunro, E.A. and Crowley, W.F. (1988) Endocrinology 123, 1854-1860. Ashwell, G. and Harford, J. (1982) Annu. Rev. Biochem. 51, 531-554. Baenziger, J.U. and Green, E.D. (1988) Biochim. Biophys. Acta 947, 287-306. Benkirane, M.M., Bon, D., Costagliola, S., Paolucci, F., Darbouret, B., Prince, P. and Carayon, P. (1987) Endocrinology 121, 1171-1177. Berger, P., Kofler, R. and Wick, G. (1984) Am. J. Reprod. Immunol. 5, 157-161. Berger, P., Panmoung, W., Khashabi, D., Mayregger, B. and Wick, G. (1988) Endocrinology 123, 2351-2359. Berger, P., Klieber, R., Panmoung, W., Madersbacher, S., Wolf, H. and Wick, G. (1990) J. Endocrinol. 125, 301-309. Bidart, J.-M., Troalen, F., Bohuon, C.J., Hennen, G. and Belle& D.H. (1987) J. Biol. Chem. 262, 15483-15489. Bidart, J.-M., Troalen, F., Ghillani, P., Rouas, N., Razafindratsita, A., Bohuon, C. and Bellet, D. (1990) Science 248, 736-739. Blithe, D.I. (1990) Endocrinology 126, 2788-2799. Canfield, R.E., Morgan, F.J., Kammerman, S., Bell, J.J. and Agosto, G.M. (1971) Recent Prog. Horm. Res. 27, 121-164. Carayon, P., Amr, S., Nisula, B.C. and Lissitzky, S. (1981) Endocrinology 108, 1891-1898. Cassidy, J.T., Jourdian, G.W. and Roseman, S. (1966) Enzymology 8, 680-685. Chen, H.-C., Shimohigashi, Y., Dufau, M.L. and Catt, K.J. (1982) J. Biol. Chem. 257, 14446-14452. Dahl, K.D., Bicsak, T.A. and Hsueh, A.J.W. (1988) Science 239, 72-74. DeLean, A., Munson, P.J. and Rodbard, D. (1972) Am. J. Physiol. 235, 660-672. Feizi, T. (1985) Nature 314, 53-57. Frydman, R., Fernandez, H., Troalen, L.F., Ghillani, P., Rainhorn, J.D. and Bellet, D. (1989) Fertil. Steril. 52, 734-738. Hussa, R.O. (1981) Ligand Rev. 3 (Suppl. 21, 6-43. Kalyan, N.K. and Bahl, O.P. (1983) J. Biol. Chem. 258, 67-74. Keutmann, H.T., McIllroy, P.J., Bergert, E.R. and Ryan, R.J. (1983) Biochemistry 22, 3067-3072. Keutmann, H.T., Johnson, L. and Ryan, R.J. (1985) FEBS Lett. 185, 333-339. Kofler, R., Kalchschmid, E., Berger, P. and Wick, G. (19811 Immunobiology 160, 196-207. Kofler, R., Berger, P. and Wick, G. (1982) Am. J. Reprod. Immunol. 2, 212-216. Lapresle, C. (1988) Anal. Biochem. 174, 308-312.
Liu, L., Southers, J.L., Banks, S.M., Blithe, D.L., Wehmann, R.E., Brown, J.H., Chen, H.-C. and Nisula, B.C. (1989) Endocrinology 124, 175-180. Lustbader, J.W., Birken, S., Pileggi, N.F., Kolks, M.A.G., Pollak, S., Cuff, M.E., Yang, W., Hendrickson, W.A. and Canfield, R.E. (1989) Biochemistry 28, 9239-9243. Manjunath, P. and Sairam, M.R. (1982) J. Biol. Chem. 257, 7109-7115. Mann, K., Berger, P., Schwarz, S. and Spoettl, G. (1989) Acta Endocrinol. 120 (Suppl. l), 235 (abstract). Matzuk, M.M. and Boime, I. (1989) Biol. Reprod. 40, 48-53. Merz, W.E. (1988) Biochem. Biophys. Res. Commun. 156, 1271-1278. Merz, W.E., Hilgenfeldt, U., Doerner, M. and Brossmer, R. (1974) Hoppe-Seyler’s Z. Physiol. Chem. 355, 1035-1045. Moyle, W.R., Bahl, O.P. and Maerz, L. (1975) J. Biol. Chem. 250, 9163-9169. Moyle, W.R., Anderson, D.M., Macdonald, G.J. and Armstrong, E.G. (1988) J. Receptor Res. 8, 419-436. Novotny, J., Bruccoleri, R.E., Newell, J., Murphy, D., Haber, E. and Karplus, M. (1983) J. Biol. Chem. 258,14433-14437. Pierce, J.G. and Parsons, T.G. (1981) Annu. Rev. Biochem. 50, 465-495. Ranta, T., Chen, H.-C., Shimohigashi, Y., Baukal, A.J., Knecht, M. and Catt, K.J. (1985) Endocrinology 116,59-64. Rebois, R.V. and Fishman, P.H. (1984) J. Biol. Chem. 259, 8087-8090. Rebois, R.V. and Liss, M.T. (1987) J. Biol. Chem. 262, 38913896. Ryan, R.J., Keutmann, H.T., Charlesworth, M.C., McCormick, D.J., Milius, R.P., Calvo, F.O. and Vutyavanich, T. (1987) Recent Prog. Horm. Res. 43, 383-429. Sairam, M.R. (1989) FASEB J. 3, 1915-1926. Sairam, M.R. and Bhargavi, G.N. (1985) Science 229, 65-67. Sairam, M.R., Linggen, J. and Bhargavi, G.N. (1988) Biosci. Rep. 8, 271-278. Schwarz, S., Berger, P. and Wick, G. (1985a) Clin. Chem. 31, 1322-1328. Schwarz, S., Berger, P., Kofler, R. and Wick, G. (1985b) in Handbook of Monoclonal Antibodies (Ferrone, S. and Dierich, M., eds.), pp. 195-220, Noyes Publications, Park Ridge, NJ. Schwarz, S., Berger, P. and Wick, G. (1986) Endocrinology 118, 189-197. Schwarz, S., Berger, P., Nelboeck, E. Panmoung, W., Klieber, R. and Wick, G. (1988) J. Receptor Res. 8, 437-453. Singh, I.V., Sairam, M.R., Bhargavi, G.N. and Akhras, R.G. (1989) J. Biol. Chem. 264, 3089-3095.
33
0 1991 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/91/$03.50 MOLCEL 02568
Number and topography of epitopes of human chorionic gonadotropin (hCG) are sh ared by desialylated and deglycosylated hCG S. Schwarz ‘, H. Krude I, R. Klieber 2, S. Dirnhofer 2, C. Lottersberger W.E. Merz 3, G. Wick 1 and P. Berger 2
*,
’ I~titute of General and ~~er~rne~tal Pat~oIo~~ Faculty of Medicine, ~‘m’versi~ of In~b~~k, A-6020 Innsb~c~ Austria, II, ’ Immunoendocrinolo~ Research Unit, Austrian Academy of Sciences, A-6020 Innsb~ck, Austria, ’ Institute of 3i~~e~t~ University of ~eideIberg,O-6900 Heidelberg I, F.R.G. (Received 25 March 1991; accepted 10 May 1991)
Key words: Monoclonal antibody; Epitope mapping; Microheterogeneity;
Carbohydrate moiety; Glycoprotein hormone
Summary A previously established map of the surface epitopes of human chorionic gonadotropin (hCG) served as template for the present study in which we investigated the antigenic surfaces of two glycosylation variants of hCG, i.e. desialylated hCG (asialo-hCG) and deglycosylated hCG (degly-hCG). This map allocates five epitopes to the a subunit, five to the p subunit and four +? epitopes to structures formed only by the a/P heterodimer holo-hCG (Schwarz et al. (1986) Endocrinology 118, 189-197; Berger et al. (1990) J. Endocrinol. 125, 301-309). Here it is described that both variants complied with this template: each of the 14 distinct monoclonal antibodies with which the epitopes of hCG were defined reacted with radiolabeled asialo-hCG and degly-hCG as well and generally bound degly-hCG with greater affinity than hCG. Moreover, every combination of capture and radiolabeled detection antibody that was either compatible or incompatible on unlabeled hCG was so also on unlabeled asialo-hCG and degly-hCG. It thus appears that alterations of the carbohydrate structure of hCG can be associated with a change in affinity between some antibodies and their respective epitopes but not with a loss of an epitope or with a change in the topographical relationships of the 14 epitopes.
Introduction Human chorionic gonadotropin (hCG) belongs to the glycoprotein hormone family which com-
Address for correspondence: S. Schwarz, Institute of Genera1 and Experimental Pathology, Faculty of Medicine, University of Innsbruck, A-6020 Innsbruck, Austria. Tel. 43 (512) 507-2266, Fax 43 (512) 507-2507.
prises in addition follicle-stimulating hormone (FSH), th~oid-stimulating hormone (TSH) and luteinizing hormone (III). They all resemble each other in molecular mass (M, 2: 301, overall conformation (elipsoid globular shape) and quaternary structure: they are non-covalently linked heterodimers consisting of an (Y subunit that is singular within a given species, and a p subunit that is unique for each hormone and that, by combination with the (Ysubunit, confers receptor
34
binding ability, receptor specificity and biological activity (for reviews, see Hussa, 1981; Pierce and Parsons, 1981; Ryan et al., 1987). The carbohydrate moieties (CHO) of hCG serve diverse functions. Intracellularly, they influence the folding, assembly competence, sorting and stability of the subunits, while extracelluIarly, the CHOs affect the circulatory half life and, more importantly, the biological activity of the secreted a/P hormone, i.e. the degree of hCG receptor activation (involving G, protein coupling, CAMP generation and steroidogenesis). Not only is the structural integrity of the CHOs important (Baenziger and Green, 1988; Sairam, 1989) but also their correct linkage to the protein backbone: while the metabolic fate of hCG through Ca2+-dependent binding to the asialogalactose receptor of hepatocytes (Ashwell and Harford, 1982) is determined by the sialic residues on all CHOs, activation of hCG receptors on target cells requires only the cYAsn-52linked CHO (Matzuk and Boime, 1989). Previously, we had probed the number and the topographical arrangement of the surface epitopes of hCG (Schwarz et al., 1985, 1986) and of the free subunits of hCG (Berger et al., 1990) by using a large panel of monocIona1 antibodies (MCA) prepared in our laboratory (Kofler et al., 1981, 1982; Berger et al., 1984,1988, 1990). Fourteen surface epitopes could thus be distinguished which are designated according to their localization as al to (~5, 81 to /35, and apl to a/34 (cw/?or c-epitopes are present only on the complete hormone; some may arise from two incomplete epitopes as shown by Bidart et al., 1990). Knowledge of the antigenic structure of hCG permitted us, first, to construct immunoassays of predictable specificity for holo-hCG or either of its subunits (Schwarz et al., 1985; Berger et al., 1990), and secondly, to study the relationship of these antibody epitopes with receptor-interaction domains (Schwarz et al., 1988). Here we asked whether or not desialylated hCG (asialo-hCG) and deglycosylated (degIy)hCG differ from native hCG (i) as to the number of epitopes, (ii) as to the affinities with which these epitopes are recognized by the MCAs, and (iii) as to the topographical relationships of these epitopes.
Materials
and methods
Preparation of hCG, desialylated hCG and deglycosylated hCG and iodinated derivatives
Purified hCG was prepared from crude hCG (Schering, Berlin, F.R.G.) by a modification of the procedure of Canfield et al. (19’71), as described in detail elsewhere (Merz et al., 1974). From this preparation ~12,000 IU/mg protein), desialylated hCG was obtained by digestion (1618 h, 37°C) with neuraminidase (EC 3.2.1.18; from Clostridium perfringens) immobilized on agarose beads (type X-A, Sigma Chemical Co., Deisenhofen, F.R.G.) in a buffer containing 50 mM sodium phosphate and 50 mM citric acid pH 5.6 according to Cassidy et al. (1966). The thus obtained asialo-hCG was desalted on a Biogel P-30 column (Bio-Rad Labs., Munich, F.R.G.) equilibrated with 50 mM NH,HCO, and Iyophilized. Deglycosylated hCG was prepared (Merz, 1988) by treating purified hCG in a closed polyterafluoroethylene (Teflon) apparatus with anhydrous hydrogen fluoride (HF) for 1 h at 0°C. After the HF had been removed under vacuum, degly-hCG was recovered in 0.5 M NH,HCO,, and after an overnight incubation at 37°C subjected to gel filtration (Biogel P-60 column 1.5 X 66 cm eluted with SO mM NH~HCO~). Purity and preservation of receptor binding activity of these three preparations were assessed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with silver or immunostaining and by radioreceptor displacement assays, all done in comparison with CR-123 as reference standard (kindly provided by the National Pituitary Agency, Baltimore, MD, U.S.A.). lZSI-labeled derivatives of hCG and the variants were prepared as described (Schwarz et al., 1986, 1988). One-site binding protocol MCA titration assay. Each of the 14 different
MCAs (stock: l-10 mg protein/ml) was serially diluted with 0.05 M phosphate buffer containing 1% (v/v) bovine serum albumin (PBS/BSA). Aliquots of 0.1 ml were incubated in parallel with 25,000 cpm of ““I-hCG, ‘251-asialo-hCG or “‘Idegly-hCG, respectively (0.1 nM, in a total volume of 0.3 ml PBS/BSA), for 12-16 h at 22°C. Unbound iz51-antigen was separated by the dou-
35
ble antibody-immunosorbent procedure (for details see Berger et al., 1990) and bound ‘251-antigen was measured. Production, characterization, purification (by ammonium sulfate and/or highperformance liquid chromatography (HPLC)) and radiolabeling of these MCAs has been described previously (Kofler et al., 1981, 1982; Berger et al., 1984, 1988, 1990; Schwarz et al., 1985, 1986, 1988). MCA displacement assay. Aliquots of 0.1 ml of MCA, appropriately diluted so as to bind < 50% of ‘251-hCG added (as determined in the foregoing MCA titration assay), were incubated with 25,000 cpm “‘1-hCG in the absence or presence of increasing concentrations of unlabeled hCG (in 0.2 ml PBS/BSA) for 12-16 h at 4°C. Conditions for separation of bound and free ‘251hCG were as above. The inhibitory concentration 50% (IC,,) of a given MCA was determined using the computer program ALLFIT (DeLean et al., 1972). Two-site sandwich protocol
Polystyrol tubes were coated with = 10 pg capture MCA (= 300 nM IgG), and after washing, incubated for 2 h at 37°C with 1 nM hCG or asialo-hCG or degly-hCG, respectively (1 nM was found optimal in preliminary dose-response experiments). Control tubes received only PBS/ BSA. After washing, = 0.5 nM (= 200,000 cpm) of ‘2”I-labeled detection MCA (from which nonspecific binding material was removed by ultracentrifugation) was added. After an incubation of 4 h at 37°C and washing, radioactivity bound to tubes was measured. This approach provides an all-or-none type of response: a positive reaction signals sterical compatibility between two epitopes, a negative result (mutual exclusion of binding) signals identity or steric proximity, as outlined in detail previously (Schwarz et al., 1985, 1986). Results
At low dilution (1: loo), a given MCA could bind 100% of each of the added 1251-antigens, i.e. native, asialo- and degly-hCG. This binding was titratable as shown by diluting the MCA further up to 1: l,OOO,OOO. As can be seen in Fig. 1, the titer (dilution of MCA that would yield 50% of
111000 1110,0001/100,0001/1000,000
dilutionof antibody Fig. 1. Binding of ‘“I-hCG (Xl, “‘1-asialo-hCG (squares) and ‘251-degly-hCG (+I as a function of dilution of MCA (dilution being indicated on the abscissa). As an example, INN-hCG-2, a MCA specific for the pl-epitope, is shown. Note the rank order in titer: degly-hCG > asialo-hCG > hCG. Cpm bound (B) are expressed as a percentage of total (T) radioactivity added. Each point is the mean of duplicates of a typical experiment repeated once. Curves were generated by the computer program ALLFIT (DeLean et al., 1972).
tracer binding) of this particular PI-MCA was = 25 or = 20 times higher for ‘Z51-degly-hCG than for ‘251-hCG or ‘251-asialo-hCG, respectively. All 14 MCAs were tested in the same way and found to be able to bind ‘*‘I-asialo- and ‘251-degly-hCG as well. In addition, each of the 14 MCAs was analyzed by a displacement assay using 1251-hCG in the presence of unlabeled hCG
0
-10
-9 -8 log [hCG], Ii4
-7
-6
Fig. 2. Binding of ‘%hCG to five different P-epitope-specific MCAs (Bl, /32, p3, /34, @S) as a function of increasing concentrations of unlabeled hCG. Cpm bound (BI are expressed as a percentage of cpm bound in the absence of hCG (B,). Each point is the mean of triplicates of a typical experiment repeated once. Curves were generated by the computer program ALLFIT (DeLean et at., 1972).
36 TABLE
I
APPARENT DEGLY-hCG
AFFINITIES
Epitope specificity and designation of MCA (Yl cu2 a3 a4 ru5 a&2 a/33 QP4 Pl P2 P3 P4 P5
INN-hFSH-73 INN-hFSH-8 INN-hFSH-15 INN-hFSH-132 INN-hFSH 158 INN-hCG-10 INN-hCG-40 INN-hCG-45 INN-hCG-26 INN-hCG-2 INN-hCG-22 INN-bLH-1 INN-hCG-24 INN-hCG-58
(K,)
OF
MONOCLONAL
ANTIBODIES
(MCA)
FOR
NATIVE
K, (liter.mol~‘) Native hCG a
hCG,
ASIALO-hCG
Rank order Asialo-hCG
b
Degly-hCG
AND
’
b
( + SEM) 6.9 + 0.58 x 3.3 + 0.29 x n.d. 2.6 k 0.27 x 1.0+0.09x10s 1.3f0.11X10s 1.5~0.15x10’ 6.4 + 0.54 x 2.4+0.25x 2.5f0.26x107 1.1f0.09X10’ 8.6 f 0.70 x 1.4~0.13x10s 3.1 _to.31 x
10h 10” 10s
10h lox
10’ 10s
1.1 x 10h 1.5 x 109 0.2x 0.3 x 1.2x 1.4x 3.5 x 2.0 x 4.7 x 1.8x 10.0x 2.1 x 3.5 x
10s 10” 10s 10’ 10h 10s 10’ 10” 10’ 10s lox
1.6x lo6 3.3 x 10’ 1.0x 2.5 x 2.4 x 2.7 x 53.0x 6.1 x 57.0 x 3.2 x 72.0 x 7.8 x 39.0 x
10s 10s 10s 10’ 10h 10s 10’ 10’ 10’ 10s 10s
n>d>a n=d>a d>n>a n>d>a d>n>a d>n>a d>n>a d>n>a d>n>a d>a>n d>a>n d>a>n d>a>n d>a>n
aKaoc,;, = 1/Ics,y~,,~
h Kacvarlantj = Ka(hCC).[titer(,,,,,,,) /fiter,t,d ’ Native hCG, n; asialo-hCG,
a; degly-hCG,
d.
(Fig. 2). The reciprocal of the herefrom determined IC,,, was taken as an estimate of the affinity (K,) of a MCA for native hCG. The apparent K, of a MCA for the variants was calculated by multiplying KachCGj with the relative titer, i.e. titer(variant/ titer(,,,,. These values are summarized in Table 1. Three different patterns of affinities with respect to native (n), asialo(a) and degly-hCG (d) emerged: (i) n 2 d 2 a, shared by the crl-, a2- and a4-MCAs; (ii) d > n 2 a, shared by the (r3- and a5-MCAs and all a@-MCAs; and (iii) d > a > n, shared by all /3MCAs (as exemplified in Fig. 1). Since all 14 epitopes, previously detected on hCG, were thus found to be present also on asialo-hCG and degly-hCG (Table 11, we asked whether these variants would share also the topographical relationships of the epitopes with hCG. Therefore, each of the 14 MCAs was tested also in a two-site binding protocol in one or the other configuration (i.e. as capture or as detection MCA). Since the epitope map of hCG was already known from previously performed complete 14 x 14 chessboard matrix experiments (Schwarz et al., 1985, 1986; Berger et al., 19901,
the present question required retesting of only the 42 most critical and informative MCA combinations (omitting all self-by-self and all reciprocally identical ones). As shown in Table 2, every of the 26 pairs of MCAs predicted to be positive (compatible), turned out to be so on the variants. The same was true for the 16 pairs predicted to be negative (incompatible). These data are summarized as an epitope compatibility table in Fig. 3. Discussion Glycoprotein hormones are of similar general architecture (Pierce and Parsons, 1981). Different laboratories arrived at similar numbers and topographical distributions of surface epitopes when investigating hCG (Schwarz et al., 1985, 1986; Bidart et al., 1987; Berger et al., 1990), hLH (Alonso-Whipple et al., 19881, hTSH (Benkirane et al., 1987) or hFSH (Berger et al., 1988). The carbohydrate moieties (CHO), representing = 30% of the total molecular mass of hCG, constitute hydrophilic, hence surface exposed portions of the molecule. In order to investigate the possi-
37
bility whether CHOs influence the epitopes that also are on the surface of hCG, two glycosylation variants of hCG were mapped, i.e. asialo-hCG and degly-hCG. The former retains most of its potency at the hCG receptor (Moyle et al., 1975) whereas the latter is a hCG receptor-bindable ligand but, in terms of CAMP response, a competitive hCG antagonist (Moyle et al., 1975; Chen et al., 1982; Sairam and Bhargavi, 1985; Baenziger and Green, 1988; Matzuk and Boime, 1989; Sairam, 1989). In addition, while native hCG displays negligible crossreactivity, asialo-hCG considerably crossreacts (as an antagonist) with the TSH receptor (Carayon et al., 1981) and deglyhCG (as a partial agonist) with the FSH receptor (Ranta et al., 1985). The principal finding of this study was that asialo-hCG and degly-hCG share with native hCG the same 14 epitopes previously identified on hCG. The very same MCAs could recognize also every band of different pl obtained by isoelectrofocussing of a microheterogenous preparation of hCG (Mann et al., 1989). Asialo-hCG and deglyhCG (and most, if not all, other glycosylation variants) are antigenically quite similar and thus targets for anti-hCG antibodies. This has several practical consequences: first, advantageous ones as to detection of hCG producing tumors by immunoassays. Secondly, disadvantageous ones as it becomes clear that immunoassays generally overestimate the actual bioactive concentration in a given sample, in contrast to bioassays and ‘bioimmunoassays’ (Moyle et al., 1988). The use of agonist-specific diagnostic assays becomes compelling in view of the recent detection of natural FSH ‘antihormones’, likely to be equivalents of degly-FSH (Dahl et al., 1988). Furthermore, one has to recognize that if antibodies occur in vivo, they are likely to react with all glycoforms of hCG. Such antibodies may be passively administered (treatment of ectopic pregnancy: Frydman et al., 1989) or may arise upon application of asialo-hCG and degly-hCG (Liu et al., 1989) or hCG-toxin conjugates (‘hormonotoxins’) (Singh et al., 1989). If one of the 14 epitopes is formed by a particular CHO, the absence of the CHOs should lead to abolished binding of the respective MCA. Since this was not the case, all the 14 epitopes of hCG
appear to be shaped predominantly by the protein backbone, as has been suggested also by others (Bidart et al., 1987). This explains on the single epitope level why polyclonal antisera, produced against native hCG, recognized degly-hCG with similar affinity as hCG (Chen et al., 1982; Manjunath and Sairam, 1982; Rebois and Fishman, 1984) and vice versa (Rebois and Liss, 1987). Thus, while degly-hCG is antigenically not different from hCG, it may still be immunogenically different. Rebois et al. (1984) and Sairam (1989) have discovered such degly-specific antibodies (of P-specificity) within an hCG-absorbed antiserum raised against degly-hCG suggesting that this protein backbone epitope may have been concealed by the CHOs on hCG. Yet per se, the CHOs seem to be of little immunogenicity in mice in contrast to CHO antigens in other contexts, e.g. blood group or tumor-associated antigens (Feizi, 1985). A second finding of this study was that each of the MCAs recognized its epitope with affinities that were often quite different depending on whether the epitope was measured on native hCG or on the variants. The prevailing pattern, however, was that degly-hCG was bound by 11 of 14 MCAs with greater affinity than native hCG. Similar was observed by Moyle et al. (1988) and Sairam et al. (1988) with a single MCA. We do as yet not know the relationship of those 14 epitopes to the primary structure of hCG and thus to the known attachment sites of the CHOs. Degly-hCG is devoid of all the complex-type N-linked sugar antennae except for the basal Gal-NAc residues (Chen et al., 1982; Manjunath and Sairam, 1982; Keutmann et al., 1983, 1985; Kalyan et al., 1985; Blithe, 1990) but still contains the O-linked CHO. These, however, reside on the C-terminal extension (Baenziger and Green, 1988) while all ‘our’ five p epitopes localize in the core of the p subunit (Kofler et al., 1981). The here observed behavior of degly-hCG relative to hCG suggests that the CHOs compromise as a whole the binding energy of a given MCA sterically (by partially shielding an epitope), i.e. by their size, length and hydration. The behavior of asialo-hCG (lacking only the terminal sialic acid residues) relative to hCG suggests that there may be also electrostatic repulsion for certain MCAs, e.g. the (Y-and (Y@-
38 TABLE
2
TWO-SITE
MONOCLONAL
ANTIBODY
BINDING
TO hCG, ASIALO-hCG
AND DEGLY-hCG
Results in each row represent ‘251-MCA binding to either hCG, asialo-hCG or degly-hCG whereby the cpm bound were expressed as A% to control (buffer) in that particular combination of capture and detection MCA. Compatibility or incompatibility between two MCAs was predicted from the previous mapping studies on native hCG and its subunits (Schwarz et al., 1986; Berger et al., 1990). A result was classified as being positive (compatible combination) when the cpm of 1251-MCA bound were at least 30% (A%) higher in the presence of antigen than in its absence; as can be seen at the bottom of the table, positive responses were on average between 384 and 696% higher, while negative results (incompatible combinations) had A% values often below control values and on average between 1.9 and 3.4. Compare also each detection MCA in a compatible and an incompatible combination! Epitope
specificity
Capture MCA 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
(Yl ff3 a2 u2 a4 a4 Pl Pl PI Pl Pl Pl 02 P2 P2 P2 P2 P2 P4 P5 P5 al34 al34 o/34 @4 op4 al al al al Cul Crl a2 a3 Pl Pl P2 P2 02 P3 ;:
All positive ones (mean f SE) All negative ones (mean f SE) a Not done.
of Detection MCA
Predicted compatibility
Native hCG
AsialohCG
DeglyhCG
58 129 1537 86 83 350 559 704 65 216 31 502 845 978 603 527 166 89 1607 1355 697 1064 1038 48 817 117 14 7 -6 4 -11 19 9 2 - 15 13 17 4 - 13 - 12 15 -3
150 238 812 75 82 280 533 696 56 N.D. a N.D. N.D. 614 642 377 302 115 N.D. N.D. N.D. N.D. 858 710 30 643 87 4 13 - 17 -4 3 1 3 17 N.D. N.D. 11 -22 18 4 13 N.D.
998 486 1772 94 260 177 659 876 129 466 132 2679 823 934 534 329 122 128 1714 1164 498 813 685 134 691 98 7 4 -6 - I2 - 10 -6 10 21 - 15 16 18 12 -4 - 18 7 7
(hCG)
+ + + + + + + + + + + + + + + + + + + + + + + + + + _ _ _ _ _ _ _ _ _ _
549 *96 2.8+ 2.9
384 k66 3.4+ 3.4
696 +120 1.9* 3.1
39
-EPlTOPE COMPATIBILITY
TABLE
a2 a3
+
+
a4
-
-
a5
+
+
apI
-
-
+
-
+
up2
-
-
+
-
+
ap3
-
-
+
-
+
-
-
cl!34
+
+
+
+
+
+
+
pl
++++++--+
+ +
-
+
82
+
+
+
+
+
+
+
+
+
+
p3
+
+
+
+
+
+
+
+
+
+
-
p4
-
+
+
+
+
+
+
+
+
+
-
-
ps
++++++++++--a4
a5
PI
p2
p3
al
a2
a3
ap1 apz
ap3 aa4
p4
Fig. 3. The epitope compatibility table was obtained from the reductional analysis of all the results collected from full-scale chessboard two-site MCA binding studies on hCG and its subunits (Schwarz et al., 1985, 1986; Berger et al., 1988, 1990). Incompatible (adjacent or overlapping) epitopes are desig(distinct and distantly located) nated by a ‘-‘, compatible epitopes by a ‘+‘. Identical compatibility patterns were obtained by testing asialo-hCG and degly-hCG, as shown in this study (Table 2), indicating that the number as well as the topographical relationships of the epitopes of hCG are shared by these glycosylation variants of hCG. An epitope map can be constructed from this table by applying two rules outlined previously (Schwarz et al., 1986): (i) the radius of an epitope is practically identical with thf radius of an antibody’s antigen binding domain, i.e. = 10 A (Novotny et al., 1983), and (ii), the minimal spatial requirement for sterical compatibility of two MCAs is that their respective epitopes are at least = 20 A apart.
MCAs, but also attraction, e.g. for the P-MCAs (Table 1). The final question was whether these 14 epitopes are distributed in the same manner as on hCG. We reasoned that, if modifications on or removal of the CHOs would drastically alter the overall conformation of hCG, such would be accompanied by dislocation of epitopes rather than disappearance, so that two epitopes that are adjacent (incompatible) on one form become ‘smeared’ apart and thus compatible on the other form, and vice versa. Either possibility should be detectable by a change in the patterns obtained by the pairwise MCA binding strategy which has, indeed, revealed conformational changes on other globular proteins, e.g. albumin (Lapresle, 1988). As shown in Table 2, every pair of MCAs previ-
ously found on hCG to be compatible or incompatible was so also on asialo-hCG and degly-hCG suggesting that the relative distances of these 14 epitopes from each other are not changed, at least not to an extent that can be detected by the here utilized all-or-none response-type approach. The same was true for hLH (results not shown). It thus appears that both variants (including hLH) occur in an overall conformation that, while not being completely identical, is rather similar to that of native hCG. This similarity still allows for discrete differences in conformation between hCG and degly-hCG as had been proposed from differences in UV absorbance spectra (Manjunath and Sairam, 1982; Merz, 19881, in fluorescence spectra (Manjunath and Sairam, 19821, in circular dichroism spectra (Keutmann et al., 1983; Kalyan and Bahl, 19851, from differences in subunit-subunit association competence (Kalyan and Bahl, 1985) or subunit-subunit dissociation or peptide bond cleavage susceptibility (Merz, 1988), or from differences in antibody affinity (Keutmann et al., 1983, 1985). Yet, the extent of those changes, while being apparently big enough to account for quantitative differences in MCA binding affinity, as seen also in this study (Table 11, is without consequence for the antigenic properties in qualitative terms, i.e. number and topography of epitopes (Table 2, Fig. 3). Definite answer of this issue will await the availability of crystal structure data of hCG in comparison to degly-hCG (Lustbader et al., 1989). In conclusion, while the CHOs markedly influence the metabolic fate as well as receptor selectivity and expression of agonist potency and efficacy, they appear to be, by comparison, of relatively small relevance for the antigenic surface characteristics of hCG. Epitopes on a globular protein appear to be quite robust structures. Acknowledgements
This work was supported in part by the Austrian Academy of Sciences. We thank Ms. Regine Gerth, Renate Goerz, Irene Gaggl and Barbara Mayer for their skillful technical assistance and the National Pituitary Agency in Baltimore, MD for the generous gift of various glycoprotein hormone preparations.
40
References Alonso-Whipple, C., Couet, M.L., Doss, R., Koziarz, J., Ogunro, E.A. and Crowley, W.F. (1988) Endocrinology 123, 1854-1860. Ashwell, G. and Harford, J. (1982) Annu. Rev. Biochem. 51, 531-554. Baenziger, J.U. and Green, E.D. (1988) Biochim. Biophys. Acta 947, 287-306. Benkirane, M.M., Bon, D., Costagliola, S., Paolucci, F., Darbouret, B., Prince, P. and Carayon, P. (1987) Endocrinology 121, 1171-1177. Berger, P., Kofler, R. and Wick, G. (1984) Am. J. Reprod. Immunol. 5, 157-161. Berger, P., Panmoung, W., Khashabi, D., Mayregger, B. and Wick, G. (1988) Endocrinology 123, 2351-2359. Berger, P., Klieber, R., Panmoung, W., Madersbacher, S., Wolf, H. and Wick, G. (1990) J. Endocrinol. 125, 301-309. Bidart, J.-M., Troalen, F., Bohuon, C.J., Hennen, G. and Belle& D.H. (1987) J. Biol. Chem. 262, 15483-15489. Bidart, J.-M., Troalen, F., Ghillani, P., Rouas, N., Razafindratsita, A., Bohuon, C. and Bellet, D. (1990) Science 248, 736-739. Blithe, D.I. (1990) Endocrinology 126, 2788-2799. Canfield, R.E., Morgan, F.J., Kammerman, S., Bell, J.J. and Agosto, G.M. (1971) Recent Prog. Horm. Res. 27, 121-164. Carayon, P., Amr, S., Nisula, B.C. and Lissitzky, S. (1981) Endocrinology 108, 1891-1898. Cassidy, J.T., Jourdian, G.W. and Roseman, S. (1966) Enzymology 8, 680-685. Chen, H.-C., Shimohigashi, Y., Dufau, M.L. and Catt, K.J. (1982) J. Biol. Chem. 257, 14446-14452. Dahl, K.D., Bicsak, T.A. and Hsueh, A.J.W. (1988) Science 239, 72-74. DeLean, A., Munson, P.J. and Rodbard, D. (1972) Am. J. Physiol. 235, 660-672. Feizi, T. (1985) Nature 314, 53-57. Frydman, R., Fernandez, H., Troalen, L.F., Ghillani, P., Rainhorn, J.D. and Bellet, D. (1989) Fertil. Steril. 52, 734-738. Hussa, R.O. (1981) Ligand Rev. 3 (Suppl. 21, 6-43. Kalyan, N.K. and Bahl, O.P. (1983) J. Biol. Chem. 258, 67-74. Keutmann, H.T., McIllroy, P.J., Bergert, E.R. and Ryan, R.J. (1983) Biochemistry 22, 3067-3072. Keutmann, H.T., Johnson, L. and Ryan, R.J. (1985) FEBS Lett. 185, 333-339. Kofler, R., Kalchschmid, E., Berger, P. and Wick, G. (19811 Immunobiology 160, 196-207. Kofler, R., Berger, P. and Wick, G. (1982) Am. J. Reprod. Immunol. 2, 212-216. Lapresle, C. (1988) Anal. Biochem. 174, 308-312.
Liu, L., Southers, J.L., Banks, S.M., Blithe, D.L., Wehmann, R.E., Brown, J.H., Chen, H.-C. and Nisula, B.C. (1989) Endocrinology 124, 175-180. Lustbader, J.W., Birken, S., Pileggi, N.F., Kolks, M.A.G., Pollak, S., Cuff, M.E., Yang, W., Hendrickson, W.A. and Canfield, R.E. (1989) Biochemistry 28, 9239-9243. Manjunath, P. and Sairam, M.R. (1982) J. Biol. Chem. 257, 7109-7115. Mann, K., Berger, P., Schwarz, S. and Spoettl, G. (1989) Acta Endocrinol. 120 (Suppl. l), 235 (abstract). Matzuk, M.M. and Boime, I. (1989) Biol. Reprod. 40, 48-53. Merz, W.E. (1988) Biochem. Biophys. Res. Commun. 156, 1271-1278. Merz, W.E., Hilgenfeldt, U., Doerner, M. and Brossmer, R. (1974) Hoppe-Seyler’s Z. Physiol. Chem. 355, 1035-1045. Moyle, W.R., Bahl, O.P. and Maerz, L. (1975) J. Biol. Chem. 250, 9163-9169. Moyle, W.R., Anderson, D.M., Macdonald, G.J. and Armstrong, E.G. (1988) J. Receptor Res. 8, 419-436. Novotny, J., Bruccoleri, R.E., Newell, J., Murphy, D., Haber, E. and Karplus, M. (1983) J. Biol. Chem. 258,14433-14437. Pierce, J.G. and Parsons, T.G. (1981) Annu. Rev. Biochem. 50, 465-495. Ranta, T., Chen, H.-C., Shimohigashi, Y., Baukal, A.J., Knecht, M. and Catt, K.J. (1985) Endocrinology 116,59-64. Rebois, R.V. and Fishman, P.H. (1984) J. Biol. Chem. 259, 8087-8090. Rebois, R.V. and Liss, M.T. (1987) J. Biol. Chem. 262, 38913896. Ryan, R.J., Keutmann, H.T., Charlesworth, M.C., McCormick, D.J., Milius, R.P., Calvo, F.O. and Vutyavanich, T. (1987) Recent Prog. Horm. Res. 43, 383-429. Sairam, M.R. (1989) FASEB J. 3, 1915-1926. Sairam, M.R. and Bhargavi, G.N. (1985) Science 229, 65-67. Sairam, M.R., Linggen, J. and Bhargavi, G.N. (1988) Biosci. Rep. 8, 271-278. Schwarz, S., Berger, P. and Wick, G. (1985a) Clin. Chem. 31, 1322-1328. Schwarz, S., Berger, P., Kofler, R. and Wick, G. (1985b) in Handbook of Monoclonal Antibodies (Ferrone, S. and Dierich, M., eds.), pp. 195-220, Noyes Publications, Park Ridge, NJ. Schwarz, S., Berger, P. and Wick, G. (1986) Endocrinology 118, 189-197. Schwarz, S., Berger, P., Nelboeck, E. Panmoung, W., Klieber, R. and Wick, G. (1988) J. Receptor Res. 8, 437-453. Singh, I.V., Sairam, M.R., Bhargavi, G.N. and Akhras, R.G. (1989) J. Biol. Chem. 264, 3089-3095.
