Cellular Signalling Vol. 4, No. 5, pp. 553-558, 1992. Printed in Great Britain.

0898-6558/92 $5.00 + 0.00 © 1992 Pergamon Press Ltd

MODULATION OF H U M A N GROWTH HORMONE BINDING TO SOMATOGENIC A N D LACTOGENIC RECEPTORS BY MONOCLONAL ANTIBODIES TO H U M A N GROWTH HORMONE TAMAR AMIT, RONNIE J. BARKEY, MOUSSA B. H. YOUDIM and Z ~ v HOCl-mERG* Department of Pharmacology, Bruce Rappaport Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, Israel

(Received 10 February 1992; and accepted 9 April 1992) Abstract--The relationship between the structure of human growth hormone (hGH) and the hormone-receptor interaction was investigated by studying the effects of specific monoelonal antibodies (MAbs) to hGH on the binding of [~25I]hGH to rabbit liver and mouse liver microsomes. Receptor binding assays were carded out using a constant dose (1 ng) of [125I]hGHand varying concentrations of MAbs. The assay was carded out in the presence of either excess ovine prolactin for the measurement of somatogenic (SOM) binding sites, or excess bovine growth hormone for the determination of lactogenic (LAC) binding sites. Anti-hGH MAbs were found to have a whole spectrum of effects on hGH binding, including inhibitory, non-effect and enhancing activities. Enhancement of the binding of [~25I]hGHto both SOM and LAC receptors was observed in liver membranes of rabbit or mouse. The observed amplified signal of [mI]hGH binding to various receptors in the presence of MAb no. 8 may be due to conformational changes which occur following MAb binding to hGH. On the other hand, most of the other MAbs caused inhibition of [mI]hGH binding. A negative correlation exists between the cross-reaction of various MAbs with the N-terminus truncated forms of hGH (Metl4-hGH or MetSLeu-hGH) and their respective KD/ICsovalues enabled the evaluation of the crucial role of the N-terminus region in hGH binding to both LAC and SOM receptors. MAb nos 1 and 19, which are directed towards acid residues 95-134 and the C-terminus, inhibited SOM binding more potently than LAC binding. Thus, it seems that these midmolecule and C-terminus regions are also important in hGH binding, and that they play a role in the partial overlap of SOM and LAC binding.

Key words: Growth hormone, growth hormone receptors, monoclonal antibodies, rabbit, mouse.

INTRODUCTION

and specific mutations of the h G H molecule [25]. The second involves the use of monoclonal antibodies (MAbs) directed against the h G H molecule or the growth hormone receptors [69]. Strasburger et al. [10] have mapped the antigenic epitopes o f h G H by using a series of 12 MAbs to h G H , and described h G H binding to N b 2 rat lymphoma cells and IM-9 human lymphoma cells as model systems to evaluate lactogenic and somatogenic binding, respectively [8]. With a view to studying a more physiological model, based on non-cancerous tissue, the present study evaluated the effects of those antih G H MAbs on the binding of h G H to the lactogenic and somatogenic receptors in liver membranes of pregnant mice and rabbits.

IN A VARIETY of species~ human growth hormone ( h G H ) binds to both lactogenic and somatogenic binding sites [1]. Two general strategies have been used to elucidate the importance of specific domains of the h G H molecules in binding and activation of the somatogenic and lactogenic receptors. The first includes chemical modification o f specific amino acid residues, enzymatic fragmentation

*Author to whom correspondence should be addressed at: Technion Faculty of Medicine, POB 9649, Haifa 31096, Israel. Abbreviations: bGH--bovine growth hormone; hGH--human growth hormone; MAb--monocional antibody; oPRL--ovine prolactin. 553

T. AMIT et aL

554

MATERIALS AND METHODS Materials

Recombinant authentic hGH was a gift from BioTechnology, General Ltd. (Rehovot, Israel). Ovine PRL (oPRL; NIADDK-18) and bovine GH (bGH; USDA-B-1) were obtained from the National Hormone and Pituitary Program, NIDKK, NIH (Bethesda, MD). Chloramine-T and BSA (fraction V) were purchased from Sigma Chemical Co. (St Louis, MO); Sephadex G-100 was from Pharmacia (Uppsala, Sweden); Na[~25I] was from the Nuclear Research Center - - Negev (Beersheva, Israel). The preparations and characterization of Met~4-hGH and Met LeuS-hGH have been previously reported [12]. Common European domestic rabbits and IC-R mice were used. MAbs to hGH were provided by Prof. F. Kohen, Department of Hormone Research of the Weizmann Institute of Science, Rehovot, Israel. lodination

hGH was radioiodinated with Na[~25I] by the chloramine-T method [13] and chromatographed on a Sephadex G-100 column (45 × 1.5 on). The specific activity of [1251]hGH ranged from 70 to 80 Ci/g. Binding studies

Membrane fractions were prepared from livers from late pregnant mice and rabbits as previously described [14]. Binding studies were carried out at 4°C for 20h using 0.01M phosphate buffer, containing 1% (w/v) BSA and 50 #M MgCI2, pH 7.6, in a final volume of 0.3 ml. Protein concentration was determined by the method of Lowry et al. [15]. The concentration of microsomal protein was chosen, in preliminary experiments, to ensure that the specific binding would be within the linear part of

the curve. [125I]hGH (1 ng) was preincubated with buffer or serial dilutions of MAbs for 20 h at 4°C prior to addition of the hepatic membranes (0.5 mg protein/tube), incubated in the presence of excess oPRL (1/tg) or bGH (!/~g). Nonspecific binding was determined in the presence of 1 ttg unlabelled hGH. The reaction was stopped by the addition of 2 ml icecold 10mM Tds-HC! (pH7.4). The membranes were sedimented by centrifugation at 3000g for 30min at 4°C and the bound radioactivity was measured in a y-counter. Binding assays were repeated at least twice, in duplicate or in triplicate, and representative data are shown. The variation between replicates was always less than 3%.

RESULTS Inhibiting M A b

The inhibitory capacity of nine MAbs on [m2SI]hGH-specific binding to somatogenic sites (with excess oPRL) or to iactogenic sites (with excess b G H ) in pregnant rabbit and mouse liver membranes is represented in Fig. 1. For those MAbs whose inhibitory capacity was greater than 30% at 4.5#g/ml (which was almost the highest M A b level assayed), the inhibitory concentration 0c50) was estimated (Table 1). Within species, the somatogenic inhibition correlated highly significantly with the lactogenic inhibition (r = 0.99; P < 0.001). The ratio o f lactogenic:somatogenic ic~0 was used to evaluate the relative potency of a M A b towards these sites. Thus, a higher ratio was inversely related to lactogenic potency, but directly

TABLE 1. Tr~ ICs0 OF VAmOUS MAbs FOR INHIBITIONOF [J2~I]hGH BINDINGTO $OMATOGENIC(SOM) AND LACTOaENIC(LAC) BINDING SITES IN RABBIT OR MOUSE LIVER MEMBRANES IC~ value (nM) Rabbit liver

Mouse liver

Code of MAb no.*

KD (nM)*

SOM

LAC

LAC/SOM

SOM

LAC

LAC/SOM

3 4 1 19 9

10.00 1.40 0.10 0.50 5.00

4.74 6.00 3.33 3.40 14.00

4.54 6.59 4.17 5.70 16.67

0.96 1.10 1.25 1.68 1.19

5.36 6.01 2.87 2.80 16.70

4.33 4.69 5.67 3.47 14.10

0.81 0.78 1.98 1.24 0.84

* Code no. and KD of MAb binding to hGH from Strasburger et al. [10].

Monoclonal antibodies to hGH

MAb-1

555

MAb-4

MAb-10

lOO 50

E

103

10

104

11

4

MAb-2

.E

0b

MAb-7

lOO

103

104

MAIm12

,,.,=,.... /

0"~=-0~0,,

•

5O e-,

~Q

'l-

I

103

e-

o

I

104 i(

103

MAb-3 100

50

104

0b

MAb-9 ,oo

1~

10~

MAb-19 lOO_

•

103

104 0(

103

104 012

103

104

MAb (ng/ml) FIG. 1. Inhibitory effects of various anti-hGH MAbs on [12SI]hGH binding to somatogenic or lactogenic receptors in mouse or rabbit liver membranes. [125I]hGH (1 rig/0.1 ml) was preincubated in the absence or presence of increasing concentrations of different MAbs for 20 h at 4°C. Incubation was carried out as described in Materials and Methods in the presence of excess oPRL (1 #g added to saturate the lactogenic sites) to verify the inhibition of the somatogenic receptors in the mouse (©) or in the rabbit (I--1);or in the presence of excess bGH (1 #g, which saturates somatogenic sites), to verify the inhibition of lactogenic receptors in the mouse (O) or in the rabbit (1). Results are expressed as a percentage of hGH-specific binding. Maximum [12SI]hGH specific binding, which served here as the 100% control, was 24.6+0.9%/mg and 40.16+ 1.2%/mg protein to the rabbit somatogenic and lactogenic liver receptors, respectively, and 18.6+1.1%/mg and 26.4+ 1.5%/mg protein to mouse somatogenic and lactogenic liver receptors, respectively.

related to somatogenic inhibitory potency. M A b no. 1, which is directed against the C-terminal epitope of h G H [10], had ratios of 1.25 and 1.98 for rabbit and mouse liver membranes, respectively. Thus, a more potent inhibition of somatogenic rather than lactogenie binding was observed with M A b no. 1 and expressed especially in the mouse liver membranes. A control, non-relevant MAb, which is not directed against h G H molecule, did not affect hGH-specific binding at a wide

range of concentrations up to 8 ng/ml (data not shown).

Enhancing MAb When M A b no. 8 was preincubated with h G H , it enhanced hGH-specific binding to both somatogenic and lactogenic binding sites in liver membranes o f rabbit (Fig. 2a) or mouse (Fig. 2b). However, this increase in h G H binding was not associated with any changes in

556

T. AMIT et al.

~

100. 4oo-(

•

300 -

~

g oo-

10

-

ao

(a)

%9o

}

1000

~_ "10 e-

•

10~

10'

~

°d.01

0.1

1

10

2OO

.8

(b)

-I-

e 1 ~ - \\

"6 ¢-.

e,-

0 I~

10 m

2

1

da

(b)

150-

100- o

o

\.3 I

lo 3

104

IvlAb (ng/ml)

FIG. 2. The" enhancing effect of MAb no. 8 on [J:5I]hGH binding to the somatogenic (I-q); ( 0 ) or lactogenic (11); (@) receptors in liver membranes from rabbit (a) or mouse (b). The [125I]hGH specific binding is expressed as a percentage of the maximal binding in the absence of MAb.

the level of non-specific binding. The enhancement was only pronounced when MAb no. 8 was preincubated with hGH, before adding the liver membranes; MAb no. 8 did not enhance h G H binding when co-incubated with h G H or when preincubated with the membranes (data not shown).

Truncated hGH fragments To evaluate the role of the N-terminal amino acids, we have utilized analogue fragments of the h G H molecule; Met, LeuS-hGH, which is devoid of the seven N-terminal amino adds, and Metm4-hGH, which is devoid of the 13 N-terminal amino acids. The inhibitory activity of the MAbs, expressed by the index ratio [8] o f the respective MAbs-hGH dissociation constant (KD) relative to its ICs0, was correlated against their respective cross-reaction with

0.1

0.001

0.01

0.1

I

10

~I,C5o

FIG. 3. Correlation between KD/ICsovalues and the cross-reactivity of the various MAbs with Met]4-hGH. The Kn/ZCsovalues of different MAbs for inhibition of ['2SI]hGH binding to somatogenic (@) or lactogenic (©) receptors in liver membranes from rabbit (a) or mouse (b). The KD/ICs0 value for each MAb is plotted against the cross-reactivity with Met~4-hGH (%). The correlation coctIicients (r) were (a): -0.93 (@) and -0.94 (©), P < 0.05; and (b): -0.92 (@), P < 0.05. Met]4-hGH (Fig. 3). Significant negative correlation with the logarithmic value of cross-reaction with Metl4-hGH was found for the somatogenic binding in both rabbit (r = - 0 . 9 3 ; P < 0.05) and mouse (r = - 0 . 9 2 ; P < 0.05) and for the rabbit lactogenic binding as well ( r - - - 0 . 9 4 ; P < 0.05). Similarly, significant negative correlation was found with the crossreaction to Met, LeuS-hGH for the somatogenic binding in rabbit (r = - 0 . 9 3 ; P < 0.05) and in mouse ( r = - 0 . 9 1 ; P < 0 . 0 5 ) , and for the rabbit lactogenic binding ( r = - 0 . 9 4 ; P< 0.05). DISCUSSION Anti-hGH MAbs, mapped for their antigenic epitope on the h G H molecule [10], provide an

Monoclonalantibodiesto hGH effective tool for research of the biological significance of various domains of the hGH molecule. In this study, the use of rabbit and mouse liver membranes, which contain both lactogenic and somatogenic receptors, enabled us to evaluate both receptor regions in the same non-cancerous tissue. The use of hGH enabled us to study both lactogenic and somatogenic receptors with the same ligand [16]. Incubation was carried out in the presence of excess oPRL to saturate lactogenic sites and verify the effect of MAbs on somatogenic binding; or in the presence of bGH to saturate somatogenic sites and verify the effect of MAbs on lactogenic binding. In previous reports these regions were assessed for their role in affecting hGH binding or induced proliferation of tumour cell lines. Indeed, anti-hGH MAbs were found to have a whole spectrum of effects on hGH binding, including inhibitory, non-effect and enhancing activities. Previously this panel of MAbs was tested for their ability to inhibit hGH binding either to lactogenic receptors in Nb 2 rat lymphoma cells or somatogenic receptors in IM-9 human lymphocytes [8]. In assessing the inhibitory activity of the MAbs on hGH binding, two factors have to be considered: the MAb affinity towards hGH (KD) and the inhibition capacity (ICs0) of hGH binding to the respective receptors, lactogenic or somatogenic. Thus the inhibitory power of a given MAb on hGH binding was represented by the previously introduced ratio of Ko/Icso [8]. The negative correlation which exists between the cross-reaction of various MAbs with the N-terminus truncated forms of hGH (Met~4-hGH or Meta-Leu-hGH) and their respective KD/[Cso values [8], enabled the evaluation of the crucial role of the N-terminus region in hGH binding to the somatogenic and lactogenic receptors. This is consistent with previous data on the importance of this region in binding to the lactogenic receptors in Nb 2 and somatogenic receptors in IM-9 cell lines [8], as well as the studies of bGH after replacement of 23 N-terminal amino acids [4, 5, 12]. Yet this interpretation is now being refined: MAbs nos 2 and 7 do not cross-react with the truncated CELLS 4:5-6

557

forms, and, hence, seem to bind to the N-terminus. These MAbs do not interfere with hGH binding to rabbit liver, and MAb no. 2 does not inhibit hGH binding to mouse liver. They do, however, interfere with hGH binding and proliferation of Nb 2 cells [8]. MAbs nos 1 and 19 identify the sequence 95-134 and the C-terminus, respectively. They inhibit somatogenie binding more potently than lactogenic binding. It seems, therefore, that these midmolecule and C-terminus regions are also important in the hGH binding, and that they play a role in the partial overlap of somatogenic and lactogenic binding. These results are in keeping with those of Cunningham et al. [17] who demonstrated the hGH binding site to the extracellular portion of the somatogenic liver receptor to be the loop between mid C-terminus residues 54-74 in the fourth helix and, to a lesser extent, the N-terminal region of the first helix. MAb no. 8 possessed a unique capacity to enhance hGH binding to liver membranes. This MAb does not cross-react with hPL, has a 50% cross-reaction with Met-LeuS-hGH, 14-25% cross-reaction with Met~4-hGH and negligible cross-reaction with a C-terminus truncated hGH fragment [10]. The specificity of the enhancing effect was carefully controlled. It was evident only after preincubation of the MAb with hGH, and was not apparent after preincubation with the membrane preparation nor when added in the binding assay itself. This may be related to the relatively low affinity of the MAb to hGH, which is one order lower than receptor affinity. The enhancing activity was more pronounced in the rabbit, compared with the mouse, and was more marked for somatogenic than for lactogenic binding. These results are another example of inconsistency with a previous report on mild inhibition by MAb no. 8 of Nb 2 proliferation and [~25I]hGH binding to Nb2 and IM-9 cell lines [8]. Enhancement of hormonal activity by MAbs, or by polyclonal antibodies, has previously been reported for insulin and for epidermal growth factor [18, 19]. Similar observations were made for MAbs complexed with hGH and

558

T. A~,tll"et al.

b G H [20-22]. In vivo experiments with antih G H , anti-hGH MAbs and polyclonal antibodies enhanced growth. However, there has not been a documentation of correlation between enhancement of receptor binding and acceleration of growth with such MAbs. We have not studied the growth effect o f M A b no. 8. Moreover, antibodies in the serum of GH-deficient children on h G H therapy, mostly of a polyclonal nature, rarely interfere with growth [23]. One can only speculate on the mechanism of binding enhancement at the present time. This may involve the longer availability of complexed M A b - h G H to the receptor, the bivalent nature of M A b molecules, conformational changes of the h G H by the MAb, binding to a distant epitope from the binding epitope, or the restriction of a particular receptor subtype by the MAb. Acknowledgements--We are grateful to PRov. FORTUNA COHEN of the Department of Hormone Research, Weizmann Institute of Science, Rehovot, Israel, for providing the anti-hGH MAbs and to BioTechnology General Ltd., Rehovot, Israel, for providing the recombinant human growth hormone. We thank MRs JovcE CARP for typing the manuscript.

REFERENCES 1. Issakson O. G. P., Eden S. and Jansson J. O. (1985) A. Rev. Physiol. 447, 483-525. 2. Li C. H. and Bewley T. A. (1976) Proc. nam. Acad. Sci. U.S.A. 73, 1476-1477. 3. Li C. H. and Blake J. (1979) Proc. nam. Acad. Sci. U.S.A. 76, 6124-6127. 4. Gertler A., Shamay A., Cohen N., Ashkenazi A., Friesen H. G., Levanon A., Gorecki M., Aviv H., Hadar D. and Vogel L. 0986) Endocrinology 118, 720-726.

5. Binder L., Vogel T., Hadar D., Elberg G. and Gertler A. (1989) Molee. Endocr. 3, 923-930. 6. Retogui L. A., de Meyts P., Pena C. and Masson P. L. (1982) Endocrinology 111, 668-675. 7. Thomas H., Green I. C., Wallis M. and Aston R. (1987) Biochem. J. 243, 365-372. 8. Strasburger C. J., Binder L., Elbcrg G., Cohen-Chapnik N., Kohen F. and Gertler A. (1989) Molec. cell. Endocr. 67, 55-62. 9. Barnard R., Bundesen P. G., Rylatt B. and Waters M. J. (1985) Biochem. J. 231, 459-468. 10. Strasburger C. J., Kostyo J., Vogel T., Barnard G. J. and Kohen F. (1989) Endocrinology 124, 1548-1588. 11. Posner B. I., Kelly P. A., Shiu R. P. C. and Friesen H. G. (1984) Endocrinology 95, 521-528. 12. Ashkenazi A., Vogel T., Barash I., Hadany D., Levanon A., Gorecki M. and Gertler A. (1987) Endocrinology 121, 414-419. 13. Greenwood F. C., Hunter W. M. and Glover J. S. (1963) Biochem. J. 89, 114-123. 14. Amit T., Barkey R. J., Gavish M. and Youdim M. B. H. (1986) Endocrinology 118, 835-843. 15. Lowry O. H., Rosebrough N. J., Farr A. L. and Randall R. J. (1951) J. biol. Chem. 193, 265-275. 16. Bick T., Youdim M. B. H. and Hochherg Z. (1989) Endocrinology 125, 1711-1717. 17. Cunningham B. C. and Wells J. A. (1989) Science 244, 1081-1085. 18. Sehechter Y., Chang K. J., Jacobs S. and Cuatrecasas P. (1979) Proc. natn. Acad. Sci. U.S.A. 76, 2720-2724. 19. Schechter Y., Hernaze L., Schlessinger J. and Cuatrecasas P. (1979) Nature 278, 835-836. 20. Holder A. T., Astort R., Preece M. A. and Ivanyi J. (1985) J. Endocr. 107, Rg-RI2. 21. Aston R., Holder A. T., Preece M. A. and Ivanyi J. (1986) J. Endocr. 110, 381-388. 22. Aston R., Holder A. T., Ivanyi J. and Bomford R. (1987) Molec. Immun. 24, 143-150. 23. Ammann A. J. (1986) In Immunological Aspects of Growth Hormone (R. D. G. Milner and H. Flodh, Eds), pp. 33-41. Medical Education Services, Oxford.