monoclonal antibodies recognize conformational alterations leading to decreased bioactivity of human...

6
Molecular and Cellular Endocrinology, 38 (1984) 75-80 Elsevier Scientific Publishers Ireland, Ltd. 75 MCE 01227 Monoclonal antibodies recognize conformational alterations leading to decreased bioactivity of human growth hormone Ingileif Jbnsdbttir I**, Marguerite Luthman 2, Hans-Peter T. Ekre 3 and Peter Perlmann Department of Immunology. University of Stockholm, S- 106 91 Stockholm, ’ Department of Endocrinology, The Karolinska Hospital, S - IO4 01 Stockholm, and ’ Research Department, Biochemistry, KabiVitrum AB, S-I I2 87 Stockholm (Sweden) (Received 4 June 1984; accepted 30 July 1984) Keywords: human growth hormone; conformation; monoclonal antibodies: radioreceptor assay; ELBA. Summary By means of monoclonal antibodies to five different antigenic determinants on human growth hormone (hGH) and polyclonal antisera from mice and rabbits, the immunoactivity of native hGH was compared with that of reduced and S-carboxymethylated hGH, which has an unstable conformation. Native and reduced and S-carboxymethylated hGH were also tested in a radioreceptor assay which reflects the bioactivity of the hormone. The IM-9 cell line was used as a receptor source in this assay. All five monoclonal antibodies were superior in discriminating between the native active form of hGH and its reduced and S-carboxymethylated form, which has a markedly reduced receptor binding activity. Human pituitary growth hormone (hGH) is a protein of molecular weight approx. 22000, con- taining a single polypeptide chain, cross-linked by two disulphide bridges (Wallis, 1978). It has been suggested that the intact three-dimensional struc- ture of the hormone is important for its growth- promoting activity (Aubert et al., 1974). Since in vivo assays for the growth-promoting activity of hGH measure the net effect, the bioactivity can * To whom correspondence and reprint requests should be addressed. Permanent address: Department of Immunology, The National University Hospital, Landspitalinn, 101 Reykjavik (Iceland). Abbreuiations: ELISA, enzyme-linked immunoadsorbent assay; hCS, human chorionic somatomammotropin; hGH, human growth hormone; Mab, monoclonal antibody; RCM-hGH, re- duced and S-carboxymethylated hGH; RIA, radioimmunoas- say; RRA, radioreceptor assay. also be measured by radioreceptor assay (RRA) based on the specific binding of radiolabelled hormone to its receptors. The hybridoma tech- nique (Kohler and Milstein, 1975) for production of monoclonal antibodies (Mabs) provides power- ful tools for investigating the structure-function relationship of protein molecules. We have pro- duced a series of Mabs to hGH that react with distinct epitopes, which are either unique for hGH or cross-react with the structurally related hormone human chorionic somatomammotropin (hCS) (Jonsdottir et al., 1983). Several Mabs to hGH have been reported to recognize conformational antigenic determinants (Retegui et al., 1982; Ivanyi, 1983; Jonsdottir et al., 1983). In this study we have compared the ability of five monoclonal antibodies with that of polyclonal antibodies to recognize conformational alterations or instability of the hGH molecule. The five Mabs

Upload: peter

Post on 31-Dec-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

Molecular and Cellular Endocrinology, 38 (1984) 75-80

Elsevier Scientific Publishers Ireland, Ltd.

75

MCE 01227

Monoclonal antibodies recognize conformational alterations leading to decreased bioactivity of human growth hormone

Ingileif Jbnsdbttir I**, Marguerite Luthman 2, Hans-Peter T. Ekre 3 and Peter Perlmann ’

’ Department of Immunology. University of Stockholm, S- 106 91 Stockholm, ’ Department of Endocrinology, The Karolinska Hospital,

S - IO4 01 Stockholm, and ’ Research Department, Biochemistry, KabiVitrum AB, S-I I2 87 Stockholm (Sweden)

(Received 4 June 1984; accepted 30 July 1984)

Keywords: human growth hormone; conformation; monoclonal antibodies: radioreceptor assay; ELBA.

Summary

By means of monoclonal antibodies to five different antigenic determinants on human growth hormone (hGH) and polyclonal antisera from mice and rabbits, the immunoactivity of native hGH was compared with that of reduced and S-carboxymethylated hGH, which has an unstable conformation. Native and

reduced and S-carboxymethylated hGH were also tested in a radioreceptor assay which reflects the bioactivity of the hormone. The IM-9 cell line was used as a receptor source in this assay. All five monoclonal antibodies were superior in discriminating between the native active form of hGH and its reduced and S-carboxymethylated form, which has a markedly reduced receptor binding activity.

Human pituitary growth hormone (hGH) is a protein of molecular weight approx. 22000, con- taining a single polypeptide chain, cross-linked by two disulphide bridges (Wallis, 1978). It has been suggested that the intact three-dimensional struc- ture of the hormone is important for its growth- promoting activity (Aubert et al., 1974). Since in vivo assays for the growth-promoting activity of hGH measure the net effect, the bioactivity can

* To whom correspondence and reprint requests should be addressed. Permanent address: Department of Immunology,

The National University Hospital, Landspitalinn, 101 Reykjavik (Iceland).

Abbreuiations: ELISA, enzyme-linked immunoadsorbent assay; hCS, human chorionic somatomammotropin; hGH, human growth hormone; Mab, monoclonal antibody; RCM-hGH, re-

duced and S-carboxymethylated hGH; RIA, radioimmunoas-

say; RRA, radioreceptor assay.

also be measured by radioreceptor assay (RRA) based on the specific binding of radiolabelled hormone to its receptors. The hybridoma tech-

nique (Kohler and Milstein, 1975) for production of monoclonal antibodies (Mabs) provides power-

ful tools for investigating the structure-function relationship of protein molecules. We have pro- duced a series of Mabs to hGH that react with distinct epitopes, which are either unique for hGH or cross-react with the structurally related hormone human chorionic somatomammotropin (hCS) (Jonsdottir et al., 1983). Several Mabs to hGH have been reported to recognize conformational antigenic determinants (Retegui et al., 1982; Ivanyi, 1983; Jonsdottir et al., 1983).

In this study we have compared the ability of five monoclonal antibodies with that of polyclonal antibodies to recognize conformational alterations or instability of the hGH molecule. The five Mabs

76

used here recognize five non-overlapping epitopes in the hGH molecule, revealed by simultaneous

binding of pairs of these Mabs to the hormone (to be published). The immunological activity of re- duced and S-carboxymethylated hGH (RCM-

hGH), which has an unstable conformation (Bew-

ley et al., 1969), was compared with that of native hGH in a competitive ELISA. In order to relate the immunological activity of hGH to its bioactiv-

ity, the receptor binding activity of native and RCM-hGH was measured in a RRA with the IM-9 cell line, which has receptors which are highly specific for primate growth hormone (Lesniak et al., 1974).

Materials and methods

Hormones Pituitary hGH was Crescormon’ from Kabi-

Vitrum AB (Stockholm, Sweden), RCM-hGH was kindly provided by Magnus Hagerman

(KabiVitrum, Stockholm). It was prepared by dis- solving approx. 50 mg hGH in 0.85 M Tris-HCl buffer, pH 8.6, with 6 M guanidine-HCl, 0.3%

EDTA and dithioerythritol in a 5-fold molar ex- cess in a nitrogen atmosphere. After 4 h at room temperature a total of 9 mg iodoacetic acid was added during 45 min. Amino acid analysis of the product showed that all cysteine residues had been converted to S-carboxymethyl cysteine.

Antibodies Mabs to hGH were produced by fusing spleen

cells from mice immunized with hGH with the myeloma cell line Sp 2/O. The production and characte~zation of the Mabs has been described by Jitnsdottir et al. (1983). They were all of the IgG,, k isotype. Mouse antisera to hGH were

pooled sera from mice immunized in the same way as for Mab production (Jonsdottir et al., 1983). Rabbit antisera to hGH were individual or pooled hyperimmune sera of rabbits immunized with Crescormon, kindly provided by Dr. Bengt Granstrand (KabiVit~m, St~kholm).

Competitive ELISA Competitive ELISA was used as described by

Jonsdottir et al. (1983) to measure immunological activity. The antibodies (Mabs or polyclonal,

titrated to give an OD,,, of 1.0 in 60 min) were incubated with dilutions of the competing antigen in microplates (Dynatech M-129B, Dynatech, Switzerland) which were precoated with 1 ,ng hGH per ml in 0.1 M sodium carbonate buffer (pH 9.6)

at 4’C overnight, and saturated with 1% bovine serum albumin. Bound antibodies were detected

by incubation with alkaline phosphatase con-

jugated second antibody to mouse or rabbit im- munoglobulins (Sigma, St. Louis, MO). The plates were developed by incubation with p- nitrophenylphosphate (Sigma), and reading the absorbance at 405 nm in a Multiskan automated spectrophotometer (Flow Laboratories, Irvine, U.K.). The results were expressed as percentage of antibody bound at a given antigen concentration, 100% being antibody alone.

Radioreceptor assay The radioreceptor assay was a modification

(Luthman et al., in preparation) of that described by Rosenfeld and Hintz (1980). Duplicate samples of IM-9 cells at a final concentration of 3 x lO’/ml, [‘251]hGH (0.2 ng/ml) and unlabelled hormone at a final concentration of O-10 pg/ml

were incubated in 12 x 75 mm polypropylene tubes for 120 min at 30°C under gentle shaking in a waterbath. Triplicate 200 ~1 aliquots were re- moved, layered over 200 ~1 ice-cold buffer in plastic microtubes and centrifuged. The super- natants were discarded and the pellets counted in a Packard Autogamma 800C scintillation counter (Packard Instrument Co. Inc., Downers Grove, IO, USA). Non-specific binding was defined as the radioactivity bound in the presence of 10 pg un- labelled hGH per ml, and was subtracted from all data points. The data are presented as percentage of maximal specific binding of [‘*‘I]hGH to the IM-9 cells.

Results

We have compared the immunolo~cal activity of native hGH with that of RCM-hGH in an indirect competitive ELISA. Fig. 1 shows the re- sults of experiments where individual or pooled polyclonal antisera from rabbits (R) and mice (M) were used to measure the immunological activities of these two forms of the hormone. The results are

11

Ra hGH 1 RahGH 2

-2 -1 0 1 2 -2 -1 0 1 2

log,, pglml competing antigen Fig. 1. Competitive ELISA. Abscissa: log,, concentration of competing antigen (pg/ml) (0. human growth hormone; I, reduced and

S-carboxymethylated human growth hormone). Ordinate: % binding of antibody. 100% is OD,, of antibody alone.

expressed as percentage of antibody bound at a given concentration of competing hormone, 100% being antibody only. The polyclonal antibodies reveal no differences in the reactivity of these two forms of hGH, except for the small difference in activity to R anti-hGH,. This is in accordance

with previous reports, indicating a slightly de- creased immunological activity of RCM-hGH in a conventional radioimmunoassay (RIA) (Aubert et al., 1974).

When the 5 different Mabs were used to com- pare the native and RCM-hGH in ELBA, the activity of the RCM-hGH was markedly reduced

(Fig. 2). The RCM-hGH activity was about 100 times lower, as measured with Mab 1, which reacts with an epitope common to hGH and hCS, or with Mab 28, which is hGH-specific. Mab 10, which is cross-reactive with hCS, and the hGH-specific Mabs 3 and 5 (Jonsdottir et al., 1983) discriminate even more efficiently between these two forms of

hGH. We also measured immunological activity with a mixture of 12 Mabs (of equal titres) di-

rected to these same 5 antigenic determinants of hGH (to be published). As shown in Fig. 2, there was a clear difference between the activity of the

native and modified hormone, although this was less pronounced than when individual Mabs were

used. When we compared the receptor binding activ-

ity of hGH to the IM-9 cell line before and after reduction and S-carboxymethylation, the RCM- hGH showed only a slight activity compared to that of native hGH. The results are presented in Fig. 3, as percentage of maximum specific binding of radiolabelled hGH to the IM-9 cells in the presence of various concentrations of native or RCM-hGH. Total binding of [‘2sI]hGH (0.2 ng/ml) and non-specific binding were 9.5% and 3.5% of total counts, respectively, in this experi- ment.

78

0.

0.

01

ot

t

T Mab3

Mab 28

,

I

m Mab mixture

-2 -1 0 1 2 -2 -1 0 1 2

log,, pg/ml competing antigen

Fig, 2. Competitive ELISA. Abscissa: log,, concentration of competing antigen (pg/ml) (0, human growth hormone; n , reduced and S-carboxymethylated human growth hormone). Ordinate: % binding of monoclonal antibody. 100% is OD,, of monoclonal antibody alone.

Discussion

It has previously been reported that the confor- mation of hGH is only slightly altered after reduc- tion and S-carboxymethylation, whereas its con- formational stability is markedly reduced, as shown

by the increased rate of proteolysis by trypsin (Bewley et al., 1969). However, the immunological activity of RCM-hGH as measured by polyclonal antisera is only slightly decreased (Aubert et al., 1974). Our results confirm these findings (Fig. 1). RCM-hGH has little growth-promoting activity in

19

_ 0 1 2 3 4

log,, ng/ml unlabelled hormone Fig. 3. Radioreceptor assay. Abscissa: log,, concentration of

unlabelled hormone (q/ml) (0, human growth hormone; n ,

reduced and S-carboxymethylated human growth hormone).

Ordinate: % [ ‘251]hGH bound to IM-9 cells, mean + SD. 100%

is maximum specific binding of [‘*‘I]hGH in the absence of

unlabelled hormone.

vivo, which has been explained by rapid enzymatic

degradation (Bewley et al., 1969). It has also been suggested that the reduction and S-carboxymethy- lation of the disulphide bond between Cys,, and

CYS,,, in the hGH molecule is responsible for the

loss of biological activity (G&f et al., 1976). We have shown here that the receptor binding activity of RCM-hGH to IM-9 cells as measured in vitro is

markedly reduced compared to that of native hGH (Fig. 3). This has also been seen when liver mem- branes from pregnant rabbits were used as the receptor source (Aubert et al., 1974). This clearly demonstrates that the immunological activity of the hormone as measured by polyclonal antisera may differ markedly from its bioactivity as mea- sured by RRA. However, Mabs to 5 different antigenic determinants of the hormone efficiently discriminate between the native active hGH and the inactive RCM-hGH (Figs. 2 and 3). We as- sume that the reduced binding of RCM-hGH to the Mabs is due to conformational alterations whereas it is unlikely that the introduction of the carboxymethyl groups contributes directly to al- tered binding of 5 distinct epitopes of the hGH molecule. In agreement with our results, Mabs have recently been reported to discriminate be- tween native active and inactive forms of inter-

feron (Pestka et al., 1983). The reduced tmmuno- logical activity of RCM-hGH to the pool of 12 Mabs is not surprising since its activity to each of these Mabs individually is weak. However, poly- clonal antisera contain a heterogeneous population of antibodies directed to the whole antigenic surface, where a large number of antibodies will

react with each epitope, recognizing it with differ- ent degrees of affinities. This can possibly explain

why polyclonal antisera do not recognize minor conformational alterations.

Pure hGH preparations contain several mass and charge variants of hGH, which differ in de-

gree of growth promotion, immunological activity and receptor binding (for review see Chawla et al.,

1983). Mabs have been successfully applied in studies of the immunological nature of mass (Wal- lis et al., 1982) and charge (Jonsdottir et al., 1984) variants of hGH as well as in those of proteolyti- tally modified forms (Aston and Ivanyi, 1983; Wallis and Daniels, 1983).

Clinical investigations have revealed discrepan- cies between plasma hGH RIA and RRA values in

acromegalic patients (Gorden et al., 1976). From studies of short-statured children the concept of ‘bioinactive hGH’ with a subnormal RRA/RIA

ratio has been introduced (Kowarski et al., 1978). Our results show that Mabs are superior to poly-

clonal antisera in detecting alterations or insta-

bility in the hGH molecule leading to decreased bioactivity. Thus, Mabs should be powerful tools to study the variants of hGH present in plasma of patients with growth disorder. The plasma hGH activity to various Mabs can be related to the location of the epitopes that they recognize in the hGH molecule, with respect to the receptor bind- ing site (Retegui et al., 1982; Cadman et al., 1982; Jonsdottir et al., in preparation). Such studies should help to explain growth disorder at a molec- ular level. In addition, we conclude that Mabs which discriminate between native active and inac- tive forms of hGH should be especially useful for purification and in vitro assay of the hormone.

Acknowledgements

We thank Mr. Rahul Ghosh for excellent technical assistance and Dr. Linda Fryklund, Re- search and Development, KabiVitrum AB, Stock-

80

holm, for reading the manuscript. This work was supported by a grant from KabiVitrum AB.

References

Aston, R. and Ivanyi, J. (1983) EMBO J. 2, 493-497.

Aubert, M.L., Bewley, T.A., Grumbach, M.M. and Kaplan,

S.L. (1974) In: Advances in Human Growth Hormone

Research, No. (NIH) 74-612, Ed.: S. Raiti (DHEW Pub].,

Washington, DC) pp. 434-462.

Bewley, T.A., Brovetto-Cruz, J. and Li, C.H. (1969) Biochem-

istry 8, 4701-4708.

Cadman, H.F., Wallis, M. and ivanyi, J. (1982) FEBS Lett.

137, 149-152.

Chawla, R.K., Parks, J.S. and Rudman, D. (1983) Annu. Rev.

Med. 34, 519-547.

Gorden, P., Lesniak, M.A., Eastman, R., Hendricks, CM. and

Roth. J. (1976) J. Clin. Endocrinol. Metab. 43, 364-373.

Grlf, L., Borvendeg, J., Barat, E., Hermann, 1. and Patthy, A.

(1976) FEBS Lett. 66, 233-237.

Ivanyi, J. (1983) In: Protein Conformation as an Immunologi-

cal Signal, Eds.: F. Celada, V.N. Schumacher and E.E.

Sercarz (Plenum Publishing Corporation, New York) pp.

191-200.

Jonsdbttir, I., Ekre, H.-P.T. and Perlmann, P. (1983) Mol. Immunol. 20, 871-876.

Jbnsdottir. I., Ekre. H.-P.T., Skoog, B. and Perlmann, P. (1984)

FEBS Lett. 167, 15-18.

Kowarski, A.A., Schneider, J., Ban Galim, E., Weldon, V.V.

and Daughaday (1978) J. Clin. Endocrinol. Metab. 47.

461-464.

Lesniak, M.A., Gorden, P., Roth, J. and Gavin 111. J.R. (1974)

J. Biol. Chem. 249. 1661-1667.

Pestka, S., Kelder, B.. Langer, J.A. and Staehelin, T. (1983)

Arch. B&hem. Biophys. 224, 111-116.

Retegui, L.A., Milne, R.W., Cambasio, CL. and Masson, P.L.

(1982) Mol. Immunol. 19, 865-875.

Retegui, L.A., De Meyts, P., Peha, C. and Masson, P.L. (1982)

Endocrinology 111, 668-676.

Rosenfield, R.G. and Hintz, R.L. (1980) J. Clin. Endocrinol.

Metab. 50, 62-69.

Wailis, M. (1978) In: Chemistry and Biochemistry of Amino

Acids, Peptides and Proteins. Vol. 5, Ed.: B. Weinstein

(Marcel Dekker, New York) pp. 213-230.

Wallis, M. and Daniels, M. (1983) FEBS Lett. 159, 241-245.

Wallis, M., Ivanyi, J. and Surowy, T.K. (1982) Mol. Cell.

Endocrinol. 28, 363-372.