0013-7227/90/1273-1002$02.00/0 Endocrinology Copyright © 1990 by The Endocrine Society
Vol. 127, No. 3 Printed in U.S.A.
Relationship between the Antigenic Topography and the Structure of Human Growth Hormone* M. M. MAZZA, M. G. GOBET, M. J. BISCOGLIO, V. MIHAJLOVICH, J. C. GUILLEMOT, N. VITA, P. FERRARA, AND L. A. RETEGUI Instituto de Quimica y Fisicoquimica Biologicas (UBA-CONICET), Facultad de Farmacia y Bioquimica (M.M.M., M.G.G., M.J.B., V.M., L.A.R.), 1113 Buenos Aires, Argentina; and Unite Biochimie des Proteines, Sanofi Elf Biorecherches (J.C.G., N.V., P.F.), Labege 31328 Cedex, France
ABSTRACT. Monoclonal antibodies (MAb) to human GH (hGH) were used to correlate the antigenic topography of the hormone with its structure. Competition experiments performed in a solid phase RIA system allowed us to measure the reactivity toward the MAb of the following hGH derivatives: hGH 20K (a natural variant lacking residues 32-46), hGH selectively modified in His or Met residues, hGH with the C and/or N-terminal disulfide bond reduced and carbamidomethylated, and hGH cleaved between residues 142-143. Results indicated that fragment 32-46 participates in the structure of epitopes EB1/EB3 and that the C-terminal bridge is located in epitope 10D6, whereas opening of both disulfide bridges alters the entire hGH antigenic surface. His-151 and Met-170 were placed in epitopes NA71 and AC8, respectively, whereas His-18 and Met-14 would be involved in the hGH antigenic domain formed by overlapping epitopes 3C11,10C1, and HG3. MAb AE5, AE12, and AC3 define
a flexible hGH region related to sequence 134-150; the respective epitopes show high conformational mobility induced by modifications in other regions of the molecule. Binding of the different hGH derivatives to lactogenic receptors from female rat liver gave some insights on the localization of the hormone-binding site. Epitopes EB1/EB3 and 10D6 were discarded because there was not a direct correlation between their drastic immunological alterations and the binding properties of the respective hGH derivatives. In the same way, epitopes AE5, AE12, and AC3 were excluded from the hGH-binding domain because a disruption in those sites did not affect the hGH interaction with receptors. We conclude that the hGH structure defined by epitopes 3C11, 10C1, and HG3 is probably related to the binding properties of the hormone. (Endocrinology 127: 1002-1008, 1990)
M
APPING of antigenic determinants (or epitopes) by biochemical techniques may be performed using different experimental approaches. In suitable cases, the study of direct binding of a monoclonal antibody (MAb) to evolutionary or chemical variants of the antigen allows identification of the main amino acid residues implicated in the structure of an epitope (1). Other procedures involve proteolysis of antigen-antibody complexes (2) or chemical modification of an antigen in its free and antibody-bound states (3). The use of synthetic as well as natural fragments of the antigen is also a useful tool to elucidate the protein sequence recognized by a MAb (4). We have recently described the spatial arrangement of 20 epitopes on the surface of the human GH (hGH) molecule (5). Since the epitopes studied covered the whole hGH surface, it was considered feasible to detect,
by immunological methods, the occurrence of minor changes in modified forms of the hormone. RIA competition experiments were used to study the reactivity toward a set of MAb of the following derivatives: hGH 20K (a natural variant lacking residues 32-46), hGH selectively modified in its His and Met residues, hGH lacking the C- and/or N-terminal disulfide bridges, and hGH cleaved between residues 142-143. Results allowed us to ascribe certain hGH regions to a given epitope and, thus, to correlate the antigenic topography of the hormone with its structure. Binding of the different hGH derivatives to female rat liver receptors gave some insights on the localization of the hormone-binding site. Materials and Methods Hormones and peptides
Received February 13, 1990. Address all correspondence and requests for reprints to: Dr. L. A. Retegui, Instituto de Quimica y Fisicoquimica Biologicas, Facultad de Farmacia y Bioquimica, Junin 956, 1113 Buenos Aires, Argentina. * This work was supported by grants from the Consejo Nacional de Investigaciones Cientificas y Tecnicas (CONICET) and Fundacion Antorchas.
hGH was prepared in our laboratory following the method of Mills et al. (6). The hormone preparation met all the criteria of purity and homogeneity described previously (7). [125I]hGH was obtained as indicated by Roth (8). Specific radioactivities of 90-120 fiCi/ng were usually achieved. The natural variant hGH 20K, which has a 15-amino acid deletion (32-46), as well as synthetic peptides corresponding to
1002
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 15:59 For personal use only. No other uses without permission. . All rights reserved.
hGH ANTIGENIC STRUCTURE the hGH sequences 1-43 and 32-46 were kindly provided by Dr. U. J. Lewis (Whittier Institute, La Jolla, CA). The structural and biological properties of hGH 20K have been extensively characterized (for a review, see Ref. 9). An enzymatically modified form of hGH, termed hGH-(142143) was obtained by some of the authors (Guillemot, J. C, N. Vita, and. P. Ferrara, manuscript in preparation). Chemical studies showed that hGH-(142-143) is identical to hGH except for a cleavage between Thr-142 and Tyr-143. Bovine GH (bGH) was obtained according to the method of Dellacha and Sonenberg (10). MAb MAb anti-hGH QA68, NA39, NA71, NA27, and antihuman placental lactogen (hPL), termed EBl, EB2, and EB3, were generously supplied by Dr. J. Ivanyi (Welcome Research Laboratory, London, United Kingdom); their properties were described previously (11, 12). MAb anti-hGH F l l , AE5, AE12, AC3, 3C11, HG3,10C1,10D6, AC8, 04C11, AE6, and AC6 have been prepared and characterized in our laboratory (5, 7, 13). Chemical modifications of hGH Reduced and tetra-S-carbamidomethylated hGH was prepared in the presence of 8 M urea, as described by Callaway Necessary et al. (14), followed by dialysis against solutions containing decreasing concentrations of urea to prevent an incorrect protein folding. The amino acid analysis indicated a value of 3.8 mol carboxymethylcysteine/mol protein, corresponding to the reduction and alkylation of both C-terminal (Cys-184-Cys-189) and N-terminal (Cys-53-Cys-165) disulfide bonds. This derivative is hereafter called hGH-RCAMc+N. Partially reduced and carbamidomethylated hGH was prepared according to the procedure of Graf et al. (15), using a 10 molar ratio of dithiothreitol to protein. The carboxymethylcysteine content estimated by amino acid analysis was 2.3 mol/ mol protein, indicating that predominantly the C-terminal disulfide bond had been reduced and alkylated (15). This derivative of hGH is hereafter called hGH-RCAMc. Both hGH-RCAMc+N and hGH-RCAMc were submitted to HPLC in a BIO-SIL TSK 250 column (Bio-Rad Laboratories, Richmond, CA) to eliminate aggregated material. Fractions containing the monomeric derivatives were pooled and stored at -20 C until used. Selective modification of His residues of hGH with ethoxyformic anhydride was carried out as described previously (16, 17). Reaction of hGH with chloramine-T (Cl-T) under mild conditions allowed the modification of Met residues to the corresponding sulfoxides (18). The hormone was incubated with different concentrations of Cl-T (Cl-T/Met molar ratios, 0.5, 5, and 20), and after digestion with trypsin the corresponding peptides were purified by HPLC. Identification and quantification of the modified Met residues were performed as previously reported (18). Solid phase RIA MAb reactivity toward each hGH derivative was measured using a solid phase RIA, as described previously (13). Briefly,
1003
polyvinyl microplates (Dynatech Laboratories, Inc., Alexandria, VA) coated with a MAb (ascitic fluid diluted 1:500 in 5 mM glycine buffer, pH 9.2) were incubated with [125I]hGH (final concentration, 0.6 ng/ml) and serial dilutions of hGH or its derivatives. Both the tracer and the competitor were diluted in PBS with 10 mg/ml BSA. The bound radioactivity was counted after incubation for 4 h at 37 C and overnight at room temperature and washed with PBS containing 0.125 ml Tween-20/ liter. The reactivity of either hGH 20K or hGH-(142-143) was tested against native hGH as a control, whereas in the case of chemically modified hGH derivatives, an internal control of the reaction (processed simultaneously, but omitting the reactive) was used. Binding to lactogenic rat liver receptors Microsomes from female rat liver were prepared as indicated by Bonifacino et al. (19). Binding of the hGH derivatives was measured by overnight incubation at 4 C [126I]hGH (100,000 cpm; ~1 ng) in the presence of the competitors with 400-600 ng microsomal proteins in a total volume of 0.5 ml. Incubation medium consisted of 25 mM Tris-HCl (pH 7.4), 10 mM CaCl2, and 0.1% (wt/vol) BSA. The reaction was stopped by dilution with 4 ml cold incubation medium, followed by centrifugation at 1,400 x g for 25 min at 4 C. Supernatants were discarded, and pellets were counted for radioactivity. hGH-specific binding varied from 12-15% of the added radioactivity, and nonspecific binding, determined in the presence of an excess (2 ng) of unlabeled hormone, varied from 2-3%. Results The reactivity of each MAb toward the hGH control and that toward the natural or chemically modified derivative were measured simultaneously; the results originated paired competition curves. The data were expressed as relative reactivity (RR), i.e. the concentration of hGH necessary to produce 50% competition divided by the concentration of derivative needed to achieve the same effect. Therefore, a RR value of about 1 indicates that the epitope recognized by the MAb is structurally similar in the control and the derivative, and a RR value less than 1 suggests that the epitope tested was altered in the modified hormone. Reactivity of hGH 20K Table 1 collects the RR values shown by various hGH derivatives studied with a set of 17 MAb. The results obtained with hGH 20K indicated that MAb 10D6, AC8, 04C11, and NA39 had similar affinities for hGH or hGH 20K (RR = 1), whereas MAb QA68, F l l , 3C11, NA71, NA27, HG3, and 10C1 bound the natural variant with lower affinity. The RR values obtained with MAb EBl and EB3 were very low, suggesting that the corresponding epitopes were significantly altered in hGH 20K.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 15:59 For personal use only. No other uses without permission. . All rights reserved.
hGH ANTIGENIC STRUCTURE
1004
TABLE 1. Reactivity of various MAb anti-hGH with hGH derivatives
hGH 20K
Endo • 1990 Voll27»No3 h6H
1C! - K3
RR° MAb
c
hGH 20K" hGH-(142-143) hGH-RCAMcd hGH-RCAMc+Ne
Vol. 127, No. 3 Printed in U.S.A.
Relationship between the Antigenic Topography and the Structure of Human Growth Hormone* M. M. MAZZA, M. G. GOBET, M. J. BISCOGLIO, V. MIHAJLOVICH, J. C. GUILLEMOT, N. VITA, P. FERRARA, AND L. A. RETEGUI Instituto de Quimica y Fisicoquimica Biologicas (UBA-CONICET), Facultad de Farmacia y Bioquimica (M.M.M., M.G.G., M.J.B., V.M., L.A.R.), 1113 Buenos Aires, Argentina; and Unite Biochimie des Proteines, Sanofi Elf Biorecherches (J.C.G., N.V., P.F.), Labege 31328 Cedex, France
ABSTRACT. Monoclonal antibodies (MAb) to human GH (hGH) were used to correlate the antigenic topography of the hormone with its structure. Competition experiments performed in a solid phase RIA system allowed us to measure the reactivity toward the MAb of the following hGH derivatives: hGH 20K (a natural variant lacking residues 32-46), hGH selectively modified in His or Met residues, hGH with the C and/or N-terminal disulfide bond reduced and carbamidomethylated, and hGH cleaved between residues 142-143. Results indicated that fragment 32-46 participates in the structure of epitopes EB1/EB3 and that the C-terminal bridge is located in epitope 10D6, whereas opening of both disulfide bridges alters the entire hGH antigenic surface. His-151 and Met-170 were placed in epitopes NA71 and AC8, respectively, whereas His-18 and Met-14 would be involved in the hGH antigenic domain formed by overlapping epitopes 3C11,10C1, and HG3. MAb AE5, AE12, and AC3 define
a flexible hGH region related to sequence 134-150; the respective epitopes show high conformational mobility induced by modifications in other regions of the molecule. Binding of the different hGH derivatives to lactogenic receptors from female rat liver gave some insights on the localization of the hormone-binding site. Epitopes EB1/EB3 and 10D6 were discarded because there was not a direct correlation between their drastic immunological alterations and the binding properties of the respective hGH derivatives. In the same way, epitopes AE5, AE12, and AC3 were excluded from the hGH-binding domain because a disruption in those sites did not affect the hGH interaction with receptors. We conclude that the hGH structure defined by epitopes 3C11, 10C1, and HG3 is probably related to the binding properties of the hormone. (Endocrinology 127: 1002-1008, 1990)
M
APPING of antigenic determinants (or epitopes) by biochemical techniques may be performed using different experimental approaches. In suitable cases, the study of direct binding of a monoclonal antibody (MAb) to evolutionary or chemical variants of the antigen allows identification of the main amino acid residues implicated in the structure of an epitope (1). Other procedures involve proteolysis of antigen-antibody complexes (2) or chemical modification of an antigen in its free and antibody-bound states (3). The use of synthetic as well as natural fragments of the antigen is also a useful tool to elucidate the protein sequence recognized by a MAb (4). We have recently described the spatial arrangement of 20 epitopes on the surface of the human GH (hGH) molecule (5). Since the epitopes studied covered the whole hGH surface, it was considered feasible to detect,
by immunological methods, the occurrence of minor changes in modified forms of the hormone. RIA competition experiments were used to study the reactivity toward a set of MAb of the following derivatives: hGH 20K (a natural variant lacking residues 32-46), hGH selectively modified in its His and Met residues, hGH lacking the C- and/or N-terminal disulfide bridges, and hGH cleaved between residues 142-143. Results allowed us to ascribe certain hGH regions to a given epitope and, thus, to correlate the antigenic topography of the hormone with its structure. Binding of the different hGH derivatives to female rat liver receptors gave some insights on the localization of the hormone-binding site. Materials and Methods Hormones and peptides
Received February 13, 1990. Address all correspondence and requests for reprints to: Dr. L. A. Retegui, Instituto de Quimica y Fisicoquimica Biologicas, Facultad de Farmacia y Bioquimica, Junin 956, 1113 Buenos Aires, Argentina. * This work was supported by grants from the Consejo Nacional de Investigaciones Cientificas y Tecnicas (CONICET) and Fundacion Antorchas.
hGH was prepared in our laboratory following the method of Mills et al. (6). The hormone preparation met all the criteria of purity and homogeneity described previously (7). [125I]hGH was obtained as indicated by Roth (8). Specific radioactivities of 90-120 fiCi/ng were usually achieved. The natural variant hGH 20K, which has a 15-amino acid deletion (32-46), as well as synthetic peptides corresponding to
1002
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 15:59 For personal use only. No other uses without permission. . All rights reserved.
hGH ANTIGENIC STRUCTURE the hGH sequences 1-43 and 32-46 were kindly provided by Dr. U. J. Lewis (Whittier Institute, La Jolla, CA). The structural and biological properties of hGH 20K have been extensively characterized (for a review, see Ref. 9). An enzymatically modified form of hGH, termed hGH-(142143) was obtained by some of the authors (Guillemot, J. C, N. Vita, and. P. Ferrara, manuscript in preparation). Chemical studies showed that hGH-(142-143) is identical to hGH except for a cleavage between Thr-142 and Tyr-143. Bovine GH (bGH) was obtained according to the method of Dellacha and Sonenberg (10). MAb MAb anti-hGH QA68, NA39, NA71, NA27, and antihuman placental lactogen (hPL), termed EBl, EB2, and EB3, were generously supplied by Dr. J. Ivanyi (Welcome Research Laboratory, London, United Kingdom); their properties were described previously (11, 12). MAb anti-hGH F l l , AE5, AE12, AC3, 3C11, HG3,10C1,10D6, AC8, 04C11, AE6, and AC6 have been prepared and characterized in our laboratory (5, 7, 13). Chemical modifications of hGH Reduced and tetra-S-carbamidomethylated hGH was prepared in the presence of 8 M urea, as described by Callaway Necessary et al. (14), followed by dialysis against solutions containing decreasing concentrations of urea to prevent an incorrect protein folding. The amino acid analysis indicated a value of 3.8 mol carboxymethylcysteine/mol protein, corresponding to the reduction and alkylation of both C-terminal (Cys-184-Cys-189) and N-terminal (Cys-53-Cys-165) disulfide bonds. This derivative is hereafter called hGH-RCAMc+N. Partially reduced and carbamidomethylated hGH was prepared according to the procedure of Graf et al. (15), using a 10 molar ratio of dithiothreitol to protein. The carboxymethylcysteine content estimated by amino acid analysis was 2.3 mol/ mol protein, indicating that predominantly the C-terminal disulfide bond had been reduced and alkylated (15). This derivative of hGH is hereafter called hGH-RCAMc. Both hGH-RCAMc+N and hGH-RCAMc were submitted to HPLC in a BIO-SIL TSK 250 column (Bio-Rad Laboratories, Richmond, CA) to eliminate aggregated material. Fractions containing the monomeric derivatives were pooled and stored at -20 C until used. Selective modification of His residues of hGH with ethoxyformic anhydride was carried out as described previously (16, 17). Reaction of hGH with chloramine-T (Cl-T) under mild conditions allowed the modification of Met residues to the corresponding sulfoxides (18). The hormone was incubated with different concentrations of Cl-T (Cl-T/Met molar ratios, 0.5, 5, and 20), and after digestion with trypsin the corresponding peptides were purified by HPLC. Identification and quantification of the modified Met residues were performed as previously reported (18). Solid phase RIA MAb reactivity toward each hGH derivative was measured using a solid phase RIA, as described previously (13). Briefly,
1003
polyvinyl microplates (Dynatech Laboratories, Inc., Alexandria, VA) coated with a MAb (ascitic fluid diluted 1:500 in 5 mM glycine buffer, pH 9.2) were incubated with [125I]hGH (final concentration, 0.6 ng/ml) and serial dilutions of hGH or its derivatives. Both the tracer and the competitor were diluted in PBS with 10 mg/ml BSA. The bound radioactivity was counted after incubation for 4 h at 37 C and overnight at room temperature and washed with PBS containing 0.125 ml Tween-20/ liter. The reactivity of either hGH 20K or hGH-(142-143) was tested against native hGH as a control, whereas in the case of chemically modified hGH derivatives, an internal control of the reaction (processed simultaneously, but omitting the reactive) was used. Binding to lactogenic rat liver receptors Microsomes from female rat liver were prepared as indicated by Bonifacino et al. (19). Binding of the hGH derivatives was measured by overnight incubation at 4 C [126I]hGH (100,000 cpm; ~1 ng) in the presence of the competitors with 400-600 ng microsomal proteins in a total volume of 0.5 ml. Incubation medium consisted of 25 mM Tris-HCl (pH 7.4), 10 mM CaCl2, and 0.1% (wt/vol) BSA. The reaction was stopped by dilution with 4 ml cold incubation medium, followed by centrifugation at 1,400 x g for 25 min at 4 C. Supernatants were discarded, and pellets were counted for radioactivity. hGH-specific binding varied from 12-15% of the added radioactivity, and nonspecific binding, determined in the presence of an excess (2 ng) of unlabeled hormone, varied from 2-3%. Results The reactivity of each MAb toward the hGH control and that toward the natural or chemically modified derivative were measured simultaneously; the results originated paired competition curves. The data were expressed as relative reactivity (RR), i.e. the concentration of hGH necessary to produce 50% competition divided by the concentration of derivative needed to achieve the same effect. Therefore, a RR value of about 1 indicates that the epitope recognized by the MAb is structurally similar in the control and the derivative, and a RR value less than 1 suggests that the epitope tested was altered in the modified hormone. Reactivity of hGH 20K Table 1 collects the RR values shown by various hGH derivatives studied with a set of 17 MAb. The results obtained with hGH 20K indicated that MAb 10D6, AC8, 04C11, and NA39 had similar affinities for hGH or hGH 20K (RR = 1), whereas MAb QA68, F l l , 3C11, NA71, NA27, HG3, and 10C1 bound the natural variant with lower affinity. The RR values obtained with MAb EBl and EB3 were very low, suggesting that the corresponding epitopes were significantly altered in hGH 20K.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 15:59 For personal use only. No other uses without permission. . All rights reserved.
hGH ANTIGENIC STRUCTURE
1004
TABLE 1. Reactivity of various MAb anti-hGH with hGH derivatives
hGH 20K
Endo • 1990 Voll27»No3 h6H
1C! - K3
RR° MAb
c
hGH 20K" hGH-(142-143) hGH-RCAMcd hGH-RCAMc+Ne
