Proc. Natl. Acad. Sci. USA
Vol. 73, No. 8, pp. 2862-2866, August 1976
Immunology
Determinants of antigenic molecules responsible for genetically controlled regulation of immune responses (immunogenetics/immune suppression/histocompatibility complex)
MICHAL SCHWARTZ*, CARL WALTENBAUGHt, MARTIN DORFO, RACHEL CESLA*, MICHAEL SELA*, AND BARUj BENACERRAFt t Department of Pathology, Harvard Medical School, Boston, Massachusetts and * the Department of Chemical Immunology, The Weizmann Institute of Science, Rehovot, Israel
Contributed by Baruj Benacerraf, May 18, 1976
ABSTRACT The ability of mice bearing the H-2s haplotype to develop helper responses to the random copolymer of GluMa while they developed suppressor responses to the terpolymer of GluAlaTyr suggested the crucial role of tyrosine in these peptides. On the basis of various considerations, it was postulated that many of the tyrosine residues in GluAlaTyr would be localized at the NH-terminal end of the molecule. To verify this hypothesis, a block terpolymer composed of a short sequence of homopolymer tyrosine covalently bound to the random copolymer of Glua was synthesized. The present studies, using t is block terpolymer, demonstrated that the chemical determinants stimulating helper and suppressor responses are distinct and can be present simultaneously in the same molecule. Thus, addition of COOH-terminal tyrosine residues to the Glu,Ala polypeptide converted this immunogenic molecule to an immunosuppressive molecule in mice bearing the H-2s haplotype. The mechanism by which these short sequences of tyrosine influence H-2-linked immune responses remains to be determined.
to assume that poly(Glu60,Ala30,Tyr'0) is composed of molecules in which most tyrosine residues are clustered at the aminotermini of the chains. This raised the possibility that a stretch of tyrosine residues is capable of blocking the antibody response to the poly(Glu6O,Ala30) portion of the molecules in mice bearing the critical H-2s haplotype. The process controlled by H-linked Ir genes is believed to be concerned with the specific stimulation of helper thymusderived lymphocytes (T cells) and/or with the process whereby these helper T cells exert their regulatory activity on antibody responses (1, 6, 7). The inability of genetic nonresponder animals to form specific antibody responses to the relevant antigen may indeed be overcome by immunization with the antigen coupled to or complexed with an immunogenic protein carrier such as methylated bovine serum albumin (MeBSA) capable of stimulating helper T cells in these animals (9, 10). Furthermore, our recent studies have demonstrated the stimulation of specific suppressor T cells in mice bearing nonresponder H-2 haplotypes for the copolymer GAT immunized with this antigen (10, 11). These suppressor T cells inhibit specifically the responses of these mice to GAT-MeBSA in vio and in vitro, to which, as stated earlier, they normally respond. On the basis of these various considerations, the inability of the mice bearing the H-2s haplotype to develop antibody responses to GAT, while they respond to GA, suggests that the stretch of polytyrosine postulated to be present at the NH2-terminal end of the GAT molecule is capable of suppressing the antibody response to GA determinants in H-2s (SJL) mice. To verify this hypothesis we have prepared a block copolymer composed of a short homopolymer of L-tyrosine covalently bound to a random copolymer of L-glutamic acid and L-alanine, and have investigated the antibody responses to this new
The immune response to thymus-dependent antigens is under the control of genes in the major histocompatibility complex in mice (1), rats (2, 3), guinea pigs (4) and rhesus monkeys (5). Synthetic polypeptide antigens have been very useful in the study of the activity of these genes, which have been termed H-linked Ir genes, and which differ for distinct antigens (6). The responses of mouse strains bearing different H-2 haplotypes to the two related random linear copolymers of L-amino acids, poly(Glu6O,Ala4O) and poly(Glu6O,AlaW,Tyr'0) (abbreviated to GA and GAT, respectively), are examples of antigens under such distinct genetic control, although the antibodies produced against the two copolymers are very crossreactive. Mice bearing H-2abdkg haplotypes respond, whereas mice bearing H-2p q n haplotypes do not respond to both antigens. However, mice bearing the H-2s haplotype, such as the SJL or the congenic resistant BlO.S strains, respond to GA but not to GAT. Thus, the presence of 10% tyrosine in the latter copolymer affects critically the immunogenicity of this antigen for H-2s mice (7). The amino acid copolymers used in these studies are prepared by polymerization of the appropriate N-carboxyamino acid anhydrides, in which the various reactive functions are reversibly blocked. The polymerization reaction starts from an initiation point and is propagated via the free amino group. The distribution of the different amino acid residues along the polymeric chains reflects the rate of polymerization of the respective N-carboxyamino acid anhydrides (8). Since the Ncarboxyanhydride of tyrosine polymerizes more slowly than those of y-benzylglutamate and alanine, it seemed reasonable
copolymerpoly(Glu65, Ala3l)-poly(Tyr4) [abbreviated to (GA)T] in mouse strains bearing different H-2 haplotypes. The antibody responses to (GA)-T were compared to those elicited in the same strains by GA and GAT. In addition, we have compared the ability of (GA)-T to suppress the primary plaque-forming cell (PFC) responses of H-2s SJL mice to GAT-MeBSA, (G,A)-T-MeBSA, and GA. The results demonstrate that the newly prepared block copolymer (GA)-T resembles very closely, both in its immunogenic and suppressive properties, the random copolymer poly(Glu60,Ala30,Tyr10). It is, therefore, reasonable to conclude that the inability of mice bearing the H-2s haplotype to produce antibodies against GA in response to GAT or (GA)-T is due to the unique ability of oligo(L-tyrosine) sequences to stimulate suppressor cells capable of suppressing the response to other determinants in this molecule normally immunogenic for these mice. This is, therefore, an example of the demonstration on the same molecules of different determinants capable of stimulating helper and suppressor responses under the control of H-2-linked genes.
Abbreviations: GA, poly(Glu6O,Ala40); GAT, poly(Glu6O,Ala30, Tyrl');(GA)-T, poly(Gluo5, Ala3l)-poly(Tyr4); T cells, thymus-derived lymphocytes; MeBSA, methylated bovine serum albumin; PFC, plaque-forming cells.
2862
Proc. Nati. Acad. Sci. USA 73 (1976)
Immunology: Schwartz et al. (a)
2863
1.4
(b)
1.2
6.0 0
E C
5.0
_ 1.0
4.0F 3.0 F
0.8
0.6
2.0
0.4
It
1.0k
c
9;
_0.2
II 20 40 60 80 100 120 1 140 80 0 FRACTION NUMBER FIG. 1. (a) Chromatographic separation of oligotyrosine on a Sephadex LH-20 column. (b) Chromatographic separation of (GA)-T on a Sephadex G-50 column. The materials under the first peaks were collected and lyophilized as indicated in Materials and Methods.
0
20
40
60
MATERIALS AND METHODS Animals. Mice were obtained from the Experimental Unit of the Weizmann Institute of Science for experiments carried out in Rehovot. The mice studied at Harvard Medical School were either purchased from the Jackson Laboratories, Bar Harbor, Maine, or bred in the animal facility of the Department of Pathology. Antigens. These polymers were used: poly(Glu6O, Ala3°,Tyr'0), abbreviated GAT, with an average molecular weight of 47,000, and poly(Glu50,Ala50), abbreviated GA, with an average molecular weight of 45,000, which were purchased from Miles Laboratories, Inc., Miles Research Division, Elkhart, Md. and Pilot Chemicals, Watertown, Mass., respectively. The synthesis of the block copolymer poly(Glu,Ala)-poly(Tyr) [abbreviated to (GA)-T] is described below. The block copolymer was prepared by making use of the amino terminus of an oligotyrosine. N-Carboxytyrosine anhydride was dissolved in dioxane, with butylamine (12) as initiator for the polymerization, and kept for 3 days under stirring. The product, oligotyrosine, precipitated out while the excess of the N-carboxytyrosine anhydride remained in the solution. The precipitate was dissolved in dimethylformamide and the solution was passed on a Sephadex LH-20 column, using dimethyl-
formamide as a chromatography solvent. The oligopeptide under the first peak was precipitated with absolute ether (Fig. la). Paper electrophoresis at pH 1.9 served as indication for purity. Neutral equivalent, determined by anhydrous titration, was 1430, which means an oligopeptide of five residues. N-Carboxy-'y-benzyl-L-glutamate anhydride and N-carboxy-L-alanine anhydride were dissolved separately in dioxane, then mixed together, added to a solution of oligotyrosine in dimethylformamide under stirring, and kept for 3 days. The product was precipitated with water and dried over phosphorus pentoxide; the protecting groups were removed with HBr in acetic acid. The reaction product was dialyzed using Spectrapor membrane tubing, with molecular weight cutoff 6000-8000 (Spectrum Medical Industries, Los Angeles), lyophilized, and chromatographed on a Sephadex G-50 fine column with 0.01 ammonium bicarbonate as the running buffer (Fig. 1). The material under the first peak was collected and lyophilized. Amino acid analysis of the polymer showed the following amino acid residue ratios: Glu:Ala:Tyr, 65:31:4. The average molecular weight of the final block copolymer was calculated to be 14,000, taking into account the amino acid composition and the fact that the initiator oligotyrosine was on the average a pentamer. Upon sedimentation the polymer gave, in 0.01 M sodium
Table 1. Comparison of immune responses to GA and (G,A)-T (arithmetic mean ± SE) in strains of mice bearing different
H-2 haplotypes Strains C3H.SW
B1O.HTG SWR/J SJL/J C3H.SW
B1O.HTG BlO.F SWR/J B1O.S SJL/J
H-2
No. of animals
b g q s
5 4 5 15
b g
5 4 4 5 5 13
n q s s
(G,A)-T
No. of animals
GA
P value
[% binding of (G,A)-T] 26 ± 3 26 ±6 5±1 11 ± 2 (% binding of GA-tyramine) 26 ± 3 25±6 1±4 5±1 6 ±2 18 ±2
5 5 1 16
26 ± 5 33 ±5 0 22 ± 3
5 5 4 5 4 11
26 ±4 38±6 4±1 7±1 16 ±3 29 ±2
NS >0.3 0.1 NS NS
Vol. 73, No. 8, pp. 2862-2866, August 1976
Immunology
Determinants of antigenic molecules responsible for genetically controlled regulation of immune responses (immunogenetics/immune suppression/histocompatibility complex)
MICHAL SCHWARTZ*, CARL WALTENBAUGHt, MARTIN DORFO, RACHEL CESLA*, MICHAEL SELA*, AND BARUj BENACERRAFt t Department of Pathology, Harvard Medical School, Boston, Massachusetts and * the Department of Chemical Immunology, The Weizmann Institute of Science, Rehovot, Israel
Contributed by Baruj Benacerraf, May 18, 1976
ABSTRACT The ability of mice bearing the H-2s haplotype to develop helper responses to the random copolymer of GluMa while they developed suppressor responses to the terpolymer of GluAlaTyr suggested the crucial role of tyrosine in these peptides. On the basis of various considerations, it was postulated that many of the tyrosine residues in GluAlaTyr would be localized at the NH-terminal end of the molecule. To verify this hypothesis, a block terpolymer composed of a short sequence of homopolymer tyrosine covalently bound to the random copolymer of Glua was synthesized. The present studies, using t is block terpolymer, demonstrated that the chemical determinants stimulating helper and suppressor responses are distinct and can be present simultaneously in the same molecule. Thus, addition of COOH-terminal tyrosine residues to the Glu,Ala polypeptide converted this immunogenic molecule to an immunosuppressive molecule in mice bearing the H-2s haplotype. The mechanism by which these short sequences of tyrosine influence H-2-linked immune responses remains to be determined.
to assume that poly(Glu60,Ala30,Tyr'0) is composed of molecules in which most tyrosine residues are clustered at the aminotermini of the chains. This raised the possibility that a stretch of tyrosine residues is capable of blocking the antibody response to the poly(Glu6O,Ala30) portion of the molecules in mice bearing the critical H-2s haplotype. The process controlled by H-linked Ir genes is believed to be concerned with the specific stimulation of helper thymusderived lymphocytes (T cells) and/or with the process whereby these helper T cells exert their regulatory activity on antibody responses (1, 6, 7). The inability of genetic nonresponder animals to form specific antibody responses to the relevant antigen may indeed be overcome by immunization with the antigen coupled to or complexed with an immunogenic protein carrier such as methylated bovine serum albumin (MeBSA) capable of stimulating helper T cells in these animals (9, 10). Furthermore, our recent studies have demonstrated the stimulation of specific suppressor T cells in mice bearing nonresponder H-2 haplotypes for the copolymer GAT immunized with this antigen (10, 11). These suppressor T cells inhibit specifically the responses of these mice to GAT-MeBSA in vio and in vitro, to which, as stated earlier, they normally respond. On the basis of these various considerations, the inability of the mice bearing the H-2s haplotype to develop antibody responses to GAT, while they respond to GA, suggests that the stretch of polytyrosine postulated to be present at the NH2-terminal end of the GAT molecule is capable of suppressing the antibody response to GA determinants in H-2s (SJL) mice. To verify this hypothesis we have prepared a block copolymer composed of a short homopolymer of L-tyrosine covalently bound to a random copolymer of L-glutamic acid and L-alanine, and have investigated the antibody responses to this new
The immune response to thymus-dependent antigens is under the control of genes in the major histocompatibility complex in mice (1), rats (2, 3), guinea pigs (4) and rhesus monkeys (5). Synthetic polypeptide antigens have been very useful in the study of the activity of these genes, which have been termed H-linked Ir genes, and which differ for distinct antigens (6). The responses of mouse strains bearing different H-2 haplotypes to the two related random linear copolymers of L-amino acids, poly(Glu6O,Ala4O) and poly(Glu6O,AlaW,Tyr'0) (abbreviated to GA and GAT, respectively), are examples of antigens under such distinct genetic control, although the antibodies produced against the two copolymers are very crossreactive. Mice bearing H-2abdkg haplotypes respond, whereas mice bearing H-2p q n haplotypes do not respond to both antigens. However, mice bearing the H-2s haplotype, such as the SJL or the congenic resistant BlO.S strains, respond to GA but not to GAT. Thus, the presence of 10% tyrosine in the latter copolymer affects critically the immunogenicity of this antigen for H-2s mice (7). The amino acid copolymers used in these studies are prepared by polymerization of the appropriate N-carboxyamino acid anhydrides, in which the various reactive functions are reversibly blocked. The polymerization reaction starts from an initiation point and is propagated via the free amino group. The distribution of the different amino acid residues along the polymeric chains reflects the rate of polymerization of the respective N-carboxyamino acid anhydrides (8). Since the Ncarboxyanhydride of tyrosine polymerizes more slowly than those of y-benzylglutamate and alanine, it seemed reasonable
copolymerpoly(Glu65, Ala3l)-poly(Tyr4) [abbreviated to (GA)T] in mouse strains bearing different H-2 haplotypes. The antibody responses to (GA)-T were compared to those elicited in the same strains by GA and GAT. In addition, we have compared the ability of (GA)-T to suppress the primary plaque-forming cell (PFC) responses of H-2s SJL mice to GAT-MeBSA, (G,A)-T-MeBSA, and GA. The results demonstrate that the newly prepared block copolymer (GA)-T resembles very closely, both in its immunogenic and suppressive properties, the random copolymer poly(Glu60,Ala30,Tyr10). It is, therefore, reasonable to conclude that the inability of mice bearing the H-2s haplotype to produce antibodies against GA in response to GAT or (GA)-T is due to the unique ability of oligo(L-tyrosine) sequences to stimulate suppressor cells capable of suppressing the response to other determinants in this molecule normally immunogenic for these mice. This is, therefore, an example of the demonstration on the same molecules of different determinants capable of stimulating helper and suppressor responses under the control of H-2-linked genes.
Abbreviations: GA, poly(Glu6O,Ala40); GAT, poly(Glu6O,Ala30, Tyrl');(GA)-T, poly(Gluo5, Ala3l)-poly(Tyr4); T cells, thymus-derived lymphocytes; MeBSA, methylated bovine serum albumin; PFC, plaque-forming cells.
2862
Proc. Nati. Acad. Sci. USA 73 (1976)
Immunology: Schwartz et al. (a)
2863
1.4
(b)
1.2
6.0 0
E C
5.0
_ 1.0
4.0F 3.0 F
0.8
0.6
2.0
0.4
It
1.0k
c
9;
_0.2
II 20 40 60 80 100 120 1 140 80 0 FRACTION NUMBER FIG. 1. (a) Chromatographic separation of oligotyrosine on a Sephadex LH-20 column. (b) Chromatographic separation of (GA)-T on a Sephadex G-50 column. The materials under the first peaks were collected and lyophilized as indicated in Materials and Methods.
0
20
40
60
MATERIALS AND METHODS Animals. Mice were obtained from the Experimental Unit of the Weizmann Institute of Science for experiments carried out in Rehovot. The mice studied at Harvard Medical School were either purchased from the Jackson Laboratories, Bar Harbor, Maine, or bred in the animal facility of the Department of Pathology. Antigens. These polymers were used: poly(Glu6O, Ala3°,Tyr'0), abbreviated GAT, with an average molecular weight of 47,000, and poly(Glu50,Ala50), abbreviated GA, with an average molecular weight of 45,000, which were purchased from Miles Laboratories, Inc., Miles Research Division, Elkhart, Md. and Pilot Chemicals, Watertown, Mass., respectively. The synthesis of the block copolymer poly(Glu,Ala)-poly(Tyr) [abbreviated to (GA)-T] is described below. The block copolymer was prepared by making use of the amino terminus of an oligotyrosine. N-Carboxytyrosine anhydride was dissolved in dioxane, with butylamine (12) as initiator for the polymerization, and kept for 3 days under stirring. The product, oligotyrosine, precipitated out while the excess of the N-carboxytyrosine anhydride remained in the solution. The precipitate was dissolved in dimethylformamide and the solution was passed on a Sephadex LH-20 column, using dimethyl-
formamide as a chromatography solvent. The oligopeptide under the first peak was precipitated with absolute ether (Fig. la). Paper electrophoresis at pH 1.9 served as indication for purity. Neutral equivalent, determined by anhydrous titration, was 1430, which means an oligopeptide of five residues. N-Carboxy-'y-benzyl-L-glutamate anhydride and N-carboxy-L-alanine anhydride were dissolved separately in dioxane, then mixed together, added to a solution of oligotyrosine in dimethylformamide under stirring, and kept for 3 days. The product was precipitated with water and dried over phosphorus pentoxide; the protecting groups were removed with HBr in acetic acid. The reaction product was dialyzed using Spectrapor membrane tubing, with molecular weight cutoff 6000-8000 (Spectrum Medical Industries, Los Angeles), lyophilized, and chromatographed on a Sephadex G-50 fine column with 0.01 ammonium bicarbonate as the running buffer (Fig. 1). The material under the first peak was collected and lyophilized. Amino acid analysis of the polymer showed the following amino acid residue ratios: Glu:Ala:Tyr, 65:31:4. The average molecular weight of the final block copolymer was calculated to be 14,000, taking into account the amino acid composition and the fact that the initiator oligotyrosine was on the average a pentamer. Upon sedimentation the polymer gave, in 0.01 M sodium
Table 1. Comparison of immune responses to GA and (G,A)-T (arithmetic mean ± SE) in strains of mice bearing different
H-2 haplotypes Strains C3H.SW
B1O.HTG SWR/J SJL/J C3H.SW
B1O.HTG BlO.F SWR/J B1O.S SJL/J
H-2
No. of animals
b g q s
5 4 5 15
b g
5 4 4 5 5 13
n q s s
(G,A)-T
No. of animals
GA
P value
[% binding of (G,A)-T] 26 ± 3 26 ±6 5±1 11 ± 2 (% binding of GA-tyramine) 26 ± 3 25±6 1±4 5±1 6 ±2 18 ±2
5 5 1 16
26 ± 5 33 ±5 0 22 ± 3
5 5 4 5 4 11
26 ±4 38±6 4±1 7±1 16 ±3 29 ±2
NS >0.3 0.1 NS NS
