Mitogenic and Binding Properties of Monoclonal Antibodies to the Prolactin Receptor in Nba Rat Lymphoma Cells SELECTIVE ENHANCEMENT BY ANTI-MOUSE

Three monoclonal antibodies (mAbs) (T6, U5, and U6) against prolactin (PRL) receptors in rat liver were studied in the rat lymphoma lactogen-dependent (Nb2-11C) and autonomous (Nb2-SP) cell lines. The mAbs had strong affinity for lactogen receptors (Ka = 12-14 nM-1), similar to that of human growth hormone (hGH) which is a lactogenic hormone. T6 and hGH competed for the same binding site, while U5 and U6 interacted with another epitope. The 125I-hGH-receptor complex could be immunoprecipitated by either U5 or U6, but not by T6. Affinity labeling and immunoblotting revealed that hGH and U6 bind to a protein of 63-65 kDa. T6, U5, and U6 were mitogenic in Nb2-11C cells but their respective potencies were 185-, 70-, and 4700-fold lower than that of hGH. Anti-mouse IgG enhanced the mitogenic effect of all three mAbs and almost completely abolished the differences between them, although their mitogenic activity was still 60-120-fold lower than hGH. Des-13-hGH, a competitive antagonist of hGH which hardly effected the binding of 125I-U5, inhibited the U5-stimulated proliferation of Nb2-11C cells in a noncompetitive manner, indicating that simultaneous binding of both ligands fixed the receptor in a nonactive conformation. A Fab fragment of T6 was not mitogenic, and inhibited the hGH-induced mitogenesis in a competitive manner, but its mitogenicity could be restored by anti-mouse IgG. We suggest that the dimerization or oligomerization of the lactogen receptor in Nb2-11C cells is an obligatory step in the transduction of the mitogenic signal. It may be induced by binding of the mAb to a site, which can be either identical or may even be distinct from that which binds the lactogenic hormone.

A Fab fragment of T6 was not mitogenic, and inhibited the hGH-induced mitogenesis in a competitive manner, but its mitogenicity could be restored by anti-mouse IgG. We suggest that the dimerization or oligomerization of the lactogen receptor in Nb2-1 1C cells is an obligatory step in the transduction of the mitogenic signal. It may be induced by binding of the mAb to a site, which can be either identical or may even be distinct from that which binds the lactogenic hormone.
Prolactin (PRL),' similar to many polypeptide hormones, initiates its biological action by binding to its specific receptor * The costs of publication of this article were defrayed in part by the payment of page charges. This  on the membranes of target cells. Nb, lymphoma cells, which are absolutely dependent on lactogenic hormones for their proliferation (1,2), have proved to be a most suitable and sensitive in vitro cell model for studying the mechanism of mitogenic action of lactogenic hormones, such as PRL or human growth hormone (hGH) (3). Recently, two cell lines derived from the original Nb, strain, one which is lactogendependent (Nb*-11C) and another which is lactogen-independent and autonomous (Nb2-SP), have been established (4,5). Lactogen receptors in both cells lines are equally potent in binding either oPRL or hGH, as judged by the identical affinity for both hormones in the dependent and autonomous cell lines (6). The number of receptors/cell in the autonomous cell line is, however, -2-fold higher (6,7). The molecular mass of the receptor has yet to be fully established and, in different laboratories, varies between 60 and 88 kDa (6,8,9). Binding of the hormone to the receptor is an obligatory step for subsequent proliferation. The minimal length of exposure required for commitment to progression through the cell cycle was 4 h (10). Although some initial events that occur subsequent to the exposure of the Nb2-11C cells to lactogenic hormones have been described recently (ll-16), the mechanism by which the receptor-mediated hormonal effect is transduced is largely unknown.
Polyclonal antibodies raised against PRL receptors in rabbit mammary gland had biological activity and stimulated DNA, casein, and lactose synthesis in normal rabbit (17) and tumor mammary gland explants (18). It was shown, however, that their monovalent fragments were devoid of any PRLlike activities (19). Similarly one out of three monoclonal antibodies raised against these receptors also exhibited PRLlike activity in explants of rabbit mammary gland (20,21). Shiu et al. (3) have shown that polyclonal antibodies against rabbit mammary gland PRL receptors were also capable of stimulating Nb, cells proliferation, while their Fab fragments were devoid mitogenic properties and even inhibited the PRLsimulated proliferation. Due to heterogeneity of these antibodies, calculation of these mitogenic responses on a molecular basis was impossible. To overcome this difficulty, we used recently prepared, specific new mAbs, recognizing various domains of PRL receptors from rat liver. These mAbs showed a cross-reactivity with PRL receptors from other rat tissues, including receptors from Nb, cells (22). In the present study, we have employed three of these antibodies, directed against at least two different epitopes and quantitated their binding properties, biological activities and enhancement by rabbit anti-mouse IgG as compared to the effects of hGH.

RESULTS
Binding Experiments-Binding of radiolabeled '251-hGH or mAbs '251-T6, '251-U5, and '251-U6 to homogenates prepared from Nb2-SP cells was studied in the presence of increasing concentrations of the nonlabeled mAbs or hGH (Fig. 1). As can be seen, the binding of hGH or T6 on one hand and the binding of U5 and U6 on the other, showed mutual competition. U5 and U6 competed for only -50% of bound '251-hGH, and hGH inhibited only 50% of the binding of either lz51-U5 or '251-U6. Partial (-50%) competition was obtained by U6 and T6, while U5 and T6 had almost no effect on the binding of either ligand. Des-13-hGH, a truncated form of hGH which competes for binding with hGH, but has an 85-fold lower affinity toward the receptor (13), exhibited only a slight ability to inhibit '251-U5 binding. Only -20% inhibition was observed at concentration of des-13-hGH, that was 104-fold higher than the concentration of U5 required to achieve the same effect. The association constants for hGH, T6, U6, and U5 calculated from homologous binding experiments (Fig. 1, A-D), using Scatchard plots, were respectively 16, 14, 13, and 12 nM-'. Linear curves indicated the existence of single receptor populations (not shown).
Similar results were also obtained using intact Nb2-SP cells (see Table I   To evaluate whether initial binding of mAb U6 or hGH affects the subsequent binding of '251-hGH or lz51-U6, Nb2-SP homogenates were preincubated from 0 to 12 h with either U6 (20 nM) or hGH (45 nM). Then, every 2 h the labeled ligands were added and the reaction was terminated 16 h later. In both cases, preincubation had absolutely no effect on the subsequent binding of the ligand (not shown). These competition experiments indicate that three forms of lactogen binding sites may exist on the cell surface and in the cell homogenates, (a) sites that recognize hGH, T6, U5, and U6 but are able to bind only one of them at a time, (b) sites that bind hGH or T6 but not U5 or U6, (c) sites that bind U5 or U6 but not hGH or T6. The present results do not exclude the possibility that (b) and (c) are in fact identical, but these sites are able to bind simultaneously U5 or U6 and hGH or T6 and therefore U5 or U6 do not compete or compete only partially with either hGH-or TG-specific sites.
Down-regulation Experiments-The results of the downregulation experiments in Nbz-SP cells are summarized in Table I. Exposure to hGH resulted in down-regulation of '70-80% of receptors that recognize all four ligands. Exposure to T6 was less effective, despite the fact that the receptor binds both ligands with similar affinity. Only 31-37% of receptors that recognize hGH and/or T6 were down-regulated.
However, in another experiment (not shown), in which the T6 concentration was elevated to 27 nM for 3 h, 75% downregulation occurred. Similarly T6 was also capable of downregulating lactogen receptors in the Nb2-11C cells. A 3-h exposure to 2.7 and 27 nM T6 resulted, respectively, in 43 and 86% down-regulation.
Since the U5-and UG-receptor complexes do not dissociate completely in pH 2.3-4.5, while at pH 11, when dissociation of the ligand does occur, the cells are killed, and the receptor is degraded, we were unable to assay whether these mAbs were also able to down-regulate lactogen receptors. It is, however, evident that the proportion of the down-regulated receptors in which '251-hGH, '251-T6, or '251-U6 can be displaced by either hGH, T6, U6, or U5 (see Table I) was roughly equal. A similar proportion of downregulated receptors also was observed using lZ51-U5 as a tracer, after exposure to hGH. The exposure to a low concentration of T6 barely had any effect. This finding suggests the possibility that hGH and T6 also down-regulate the receptor form that binds simultaneously U6 or U5 and either hGH or T6. The most reasonable interpretation therefore is that the receptor forms (5) and (c), mentioned in the previous paragraph, are indeed identical.
Effect of T6, U5, and U6 on the Proliferation of Nb2-1lC Cells-As shown in Fig. 2, all three mAbs exhibited mitogenic activity in Nb, cells. Surprisingly, despite the fact that the affinity of the mAbs and hGH for the receptor is similar, the molar concentrations of T6, U5, and U6 required for halfmaximal effect were much higher than that of hGH. Nonspecific mouse IgG had no mitogenic effect (not shown).
Quantitative analysis of these results was performed by plotting the reciprocals of the number of doublings versus the reciprocal of hGH or mAb concentration. Previous experiments indicated that this plot gives a linear regression with correlation coefficient >0.98 (4). The intersection on the abscissa is equal to -l/concentration of the hormone required for half-maximal proliferation. Using this method, we replotted the results presented in Fig. 2 (not shown). Similarly to hGH, T6, U5, and U6 yielded straight lines and the respective concentrations required for half-maximal proliferation were 0.0027, 0.503, 0.187, and 12.7 nM.
We also found that U6, partially inhibited US-stimulated mitogenesis (not shown), suggesting that U5 and U6 interact at a very similar site on the lactogen receptor. To investigate whether the mitogenic effect of U5 is mediated through a binding site that recognizes hGH, the effect of des-IS-hGH, a competitive antagonist of hGH and oPRL (13,24), was tested.
As shown in Fig. 3A, a strong dose-dependent of inhibition U5-stimulated proliferation was obtained. Replotting of the results in a double-reciprocal manner (Fig. 3B) revealed that unlike hGH-or oPRL-stimulated proliferation of Nb2-1lC cells (see Fig. 1 in Ref. 13) the inhibition was noncompetitive. The Ki value (mean + S.E.) calculated from two inhibition plots was 830 + 180 nM.
Previously, we have found that 12-0-tetradecanoylphorbol 13-acetate enhanced the hGH-or oPRL-stimulated mitogenesis of Nb2-11C cells (4). A similar effect, namely a -2O-25% increase in doubling rate, also was observed using concentrations of U5, U6, or T6 required for half-maximal stimulation of the proliferation (data not shown). These findings suggest that the post-receptor effects enhanced by 12-O-tetradeca-

Cells by Rabbit Anti-mouse
IgG-As shown in Fig. 4, the addition of anti-mouse IgG increased the mAb-stimulated proliferation in a dose-dependent manner. At 500 pM mAb, the maximal enhancement was obtained in 1:25,000 dilution, at 100 pM at 1:125,000 dilution, and at 20 pM between 1:125,000 and 1:625,000 dilutions. At higher IgG concentrations the mitogenic effect was gradually attenuated. The effect was, however, inversely correlated with the ability of a particular mAb to stimulate proliferation in the absence of antimouse IgG: in the case of U5 the enhancing effect was relatively small, with T6 the effect was stronger, and with U6 there was dramatic increase. In fact, addition of anti-mouse IgG almost completely abolished the differences between T6, U5, and U6. It should be mentioned that, in the absence of mAbs, the anti-mouse IgG had no stimulatory effect on cell proliferation and did not influence the hGH-stimulated mitogenesis (not shown).
Binding Properties and Biological Activity of Fab (T6) Fragment-The ability of Fab fragment to compete with lz51-hGH for binding to Nb2-SP cells was reduced -30-fold as compared to the intact T6 (Fig. 5A). It has no mitogenic activity in Nb,-11C cells even at the concentration of 400 nM (not shown) but inhibited the hGH-stimulated mitogenesis in a competitive manner (Fig. 5B). The K, value (mean f S.E.) calculated from two inhibition plots, using different concentrations of Fab (130-390 nM), was 107 f 11 nM. The mitogenic activity of the Fab could however, be restored in a dose-dependent pattern by simultaneous addition of rabbit anti-mouse IgG (Fig. 6). Interestingly, the Fab fragment did not inhibit the U5-stimulated proliferation of the Nb2-11C cells (not shown).

Immunoprecipitation
Experiments-Immunoprecipitation of Triton X-lOO-solubilized receptor-'251-hGH complexes was 2.0 . carried out to determine whether U5 or U6 and hGH can bind simultaneously to lactogen receptors. As shown in Fig. 7, a dose-dependent precipitation was achieved with both U5 and U6, evidently indicating that simultaneous binding of hGH and these Abs occurs. T6 did not precipitate the complex, further indicating that this mAb and hGH compete for the same binding site. No precipitation occurred in the presence of nonspecific mouse IgG.
Determination of the Molecular Weight of hGH and U6 Binding Sites-Cross-linking experiments of lz51-hGH to intact cells, followed by SDS-PAGE and autoradiography, revealed an appearance of a single band of M, 87,000 (Fig. 8). This band appears within as early as 7 min of exposure to '251-hGH. Its intensity increases through 30 min and then gradually decreases, most likely due to down-regulation of the occupied receptor and degradation of the hormone-receptor complex. Additional minor bands with a lower M, values probably represent partially proteolyzed forms that appear after 15 min. Assuming that one molecule of hGH interacts with one binding site, the M, of the binding unit was estimated to be 65,000. This value is very similar to our previously published results (6) and those of others (8,9).
Western blot analysis of the membrane fraction solubilized in 0.5% CHAPS (Fig. 9A) or 1% Triton X-100 (Fig. 9B) revealed one major 63-kDa protein band that was specifically recognized by '251-U6. Similar dose-response interactions with increasing amounts of membrane protein were observed in both cases. Aggregates and a few minor bands with lower h4, values, probably representing proteolytic fragments of the receptor, also interacted with '251-U6. that U5 and T6 did not compete with each other indicate that these mAbs can bind simultaneously to the receptor existing in both forms and that U5 and U6 bind in a similar but not identical mode.
This hypothesis is further supported by additional experiments that showed: (a) preincubation of homogenates with either hGH or U6 for 12 h prior to addition of the labeled ligand did not affect the binding properties; (b) hGH and T6 down-regulated to similar degrees the receptors in which these ligands did or did not compete with U6 or U5 (Table I); (c) immunoprecipitation studies indicated that the receptor can simultaneously bind hGH and U5 or U6 (Fig. 7); (d) the M, of the main form that binds hGH in the intact cells (Fig. 8) is almost identical to that which binds U6 in solubilized receptors (Fig. 9); and (e) U6 inhibits partially the U5-stimulated proliferation of Nb,-11C cells. The physiological importance of these two epitopes is unknown, but the fact that the binding patterns to cell homogenates and intact cells were very similar hints at that both are present on the cell surface.
The biological effect of anti-PRL receptor mAbs was monitored by measuring their effect on the proliferation of Nb 11C lymphoma cells. As shown in Fig. 2, all three mAbs were mitogenic. Surprisingly, however, despite the fact that the binding association constants of all three mAbs were almost identical and were similar to that of hGH, the respective molar concentrations U5, T6, and U6 required for half-maximal mitogenic effect (Fig. 2) were 70-, 185-, and 4700-fold higher. Simultaneous addition of mAbs and anti-mouse IgG at optimal concentrations enhanced the mitogenic activity of all three mAbs and greatly diminished the differences between the mAbs themselves as well as between the mAbs and hGH. The respective concentrations required for a similar mitogenic response (as calculated from the data presented in Fig. 4) were 59-fold (U5), 60-fold (T6), and 120-fold (U6) higher than that of hGH.
The present results suggest that, similar to epidermal growth factor (27)(28)(29)(30), insulin-like growth factor-I (31), or platelet-derived growth factor (32) receptors, dimerization or oligomerization of lactogen receptors in Nb2-11C cells is an obligatory step in transduction of the biological signal. Our present finding that the Fab fragment prepared from T6 became an antagonist of hGH, lost its mitogenic activity but regained it in the presence of an anti-mouse IgG strongly supports this hypothesis. Although definite evidence is still missing, it would appear that binding of hGH (or other lactogenic hormones) may also induce a conformational change that results in dimer formation.
Recent reports indicating that anti-hGH antibodies enhanced the somatogen receptor-mediated bioactivity (33) and that the lactogen and somatogen receptors are structurally related (34) support this notion. If this assumption is true, then the fact that the mitogenic activity of all three mAbs is lower than that of hGH may be related to their impaired ability to induce dimerization of the receptor. Two possible explanations may be offered: (a) the conformational change resulting from binding of hGH shifts the monomer/dimer equilibrium to the right, while the binding of the mAb, has the same effect but to a much lesser degree; (b) mAbs bind but lack the ability to induce dimerization. Only random dimerization that leads to further biological response occurs. The present data are insufficient to select which explanation is the most likely, although selective enhancement of mitogenic activity by anti-mouse IgG does not favor the latter explanation.
Binding of des-13-hGH, a weak antagonist and competitive inhibitor of hGH or oPRL in Nbs-11C cells (13), did not prevent simultaneous binding of U5 (Fig. 1) (and most likely also U6) but fixed the receptor in a form that either prevented its dimerization or formed a nonactive dimer. This possibility is in full agreement with the noncompetitive inhibition pattern of the USstimulated Nb2-11C cells proliferation (Fig. 3).
In an attempt to understand the role of receptor dimerization in transduction of epidermal growth factor-stimulated biological signal, an "allosteric oligomerization model" has been proposed (35). According to this model, monomeric epidermal growth factor receptors are in equilibrium with oligomeric receptors. The binding of the ligand stabilizes the oligomeric state and subsequently leads to phosphorylation of the protein kinase domain by mutual phosphorylation between the neighboring cytosolic domains (36).
A similar hypothesis may apply also to platelet-derived growth factor or insulin-like growth factor-I receptors that have tyrosine kinase in the intracellular domain (37,38). Cloning of lactogen (34,39) and somatogen (40) receptors revealed that both are a single transmembrane proteins that show regions of high similarity. However both receptors lack sequences associated with protein tyrosine kinase activity or autophosphorylation sites.
Thus, although our present results strongly indicate that dimerization or oligomerization of receptors is an obligatory step in the mitogenic process of Nbz-11C lymphoma cells, we have, at present, no indication concerning its mechanism. Elucidation of early events occurring after exposure to hormone or antibody and follow up after receptor trafficking in the stimulated cells may be helpful in resolving this question.