Phylogeny of Immunoglobulin Structure and Function. XV. Idiotypic Analysis of Shark Antibodies

DEVELOPMENTAL AND COMPARATIVE IMMUNOLOGY, Vo. 6, pp. 463-472, 1982. 0145-305X/82/030463-10$03.00/0 Printed in the USA. Copyright (c) 1982 Pergamon Press Ltd. All rights reserved.

PHYLOGENY OF IMMUNOGLOBULIN STRUCTURE AND FUNCTION. XV. IDIOTYPIC ANALYSIS OF SHARK ANTIBODIES!

L. William Clem and Gerrie A. Leslie Department of Microbiology, University of Mississippi Medical Center, Jackson, MS and Department of Microbiology and Immunology, University of Oregon Health Sciences Center, Portland, OR

ABSTRACT. Nurse shark antibodies to the streptococcal A-variant carbohydrate were specifically purified from the sera of four individual animals and used as immunogens in guinea pigs. The resultant guinea pig antisera contained antioodies with apparent idiotypic specificities for the homologous shark proteins. The shark idiotypic sites were located on the Fab fragments and appeared to require the participation of H and L chains for full expression. Tests for cross reactivity employing the four guinea pig anti-idiotypic sera and antibodies from 13 immunized sharks were positive in only two cases (heterologous inhibition). These findings indicate that the idiotypic library (and by inference the antibody combining site repertoire) of nurse sharks to the streptococcal A-variant antigen is proDably quite extensive. INTRODUCTION The immune response of higher vertebrates is characterized by an ability to produce antibodies specific for a vast array of different antigenic determinants, probably on the order of lo 6-lo7. Similarly, these animals can produce large numbers (probably up to 100) in some cases of different combining sites for any given antigenic determinant (1). This extensive and heterogeneous antibody combining site repertoire results, for the most part, from an almost unlimited number of different functionally effective combinations of antibody H and L chain variable regions (2), The available data suggest that lower vertebrates, such as sharks and fish, may also have the ability to produce antibodies to a large array of different antigenic determinants (3). This implies that the mechanism(s) for generating antibody combining site diversity arose early in phylogenesis.

1.

This work was supported in part by NSF grants PCM75-16479 and 79-04954 and a grant from the Medical Research Foundation of Oregon, It was initiated while the senior author was at the University of Florida College of Medicine, Gainesville, FL. 463

464

SHARK IDIOTYPES

Vol. 6, No. 3

However there is little data to indicate whether or not antibodies from lower vertebrates display heterogeneity in the combining sites to a single antigen. ! priori it is impossible to say whether the shark antibody combining site repertoire for a given antigenic determinant is more or less extensive than that for the same determinant in higher species. Much of our current knowledge of the structural heterogeneity of antibodies from higher vertebrates has resulted from amino acid sequence studies. However, the use of antibodies to idiotypes as probes for variable region differences and similarities in antibody molecules has proven to be an effective way of assessing the heterogeneity within and between antibody populations (4). Therefore, taking advantage of the ability of nurse sharks to produce large amounts of antibodies of restricted heterogeneity to the group A-variant streptococcal carbohydrate (5,6), studies were undertaken to determine if these shark proteins contained idiotypic determinants. The strategy employed involved immunizing guinea pigs with specifically purified antibodies from individual sharks, adsorbing the xenogeneic antisera to render them specific for putative idiotypic sites, localizing these sites on the shark antibody molecules and determining if antibodies from other sharks also contained these idiotypic markers. The results obtained with the presumed idiotypes from four nurse sharks are reported here.

MATERIALS AND METHODS Preparation of shark antisera and antibodies Antisera from individual nurse sharks immunized with group A-variant streptococcal vaccine were obtained as described previously (5,6); in the cases of individual animals not discussed in previous publications, antisera were obtained 50-70 days after one course of immunization (three intravenous injections of vaccine spaced two days apart). All antisera used contained between 2-10 mg antibody/ml as determined by the quantitative precipitin reaction with the group A-variant streptococcal carbohydrate. Sera were stored at -20°C. Nurse shark antibodies to the streptococcal group A-variant carbohydrate were purified from affinity columns as described previously (7). Nurse shark antibodies to the DNP moiety were the same as used in another study (8). All shark antibodies used here were of the 19S type (5) as judged by immunoelectrophoretic, analytical ultracentrifugal and SDS polyacrylamide gel criteria. They were also at least 90% active as judged by their ability to readsorb to the appropriate affinity matrix. Purified proteins were stored at -20°C in phosphate buffered saline, pH7.4 at 5-10 mg/ml. Preparation of anti-idiotypic sera Groups of three guinea pigs were immunized with specifically purified antibodies obtained from single bleedings from individual nurse sharks 50-70 days after initial immunization. Each guinea pig received 500 ~g of antigen in complete Freund's adjuvant distributed between the hind footpads and subcutaneously on the back, followed one month.later by 100 ~g of antigen in complete adjuvant in multiple sites on the back. Sera from each guinea pig were obtained at 7 and 10 days after the secondary immunization and stored at -20°C. As discussed in more detail under RESULTS, some of these guinea pig

SHARK IDIOTYPES

Vol. 6, No. 3

465

antisera contained antibodies to shark isotypic and presumed idiotypic determinants. They were rendered specific for idiotypic determinants by multiple (at least 4) passages through affinity columns prepared by CNBr coupling pooled normal nurse shark (from at least 10 animals) 19S immunoglobulin to Sepharose 4B. Each antiserum judged to contain idiotypic antibodies by immunodiffusion analysis was also passed over an additional affinity column prepared with 19S immunoglobulin isolated from the normal serum of the shark that had produced the antibody in question. Quantification of shark idiotypes The radioimmunoassay employed for quantifying the shark idiotypic reactions involved mixing various amounts of radioiodinated shark antibody (9) with 30 ~1 guinea pig antiserum at 37°C for 1 hour followed by 25°C for 3 hrs. These reactions were conducted in phosphate buffered saline, pH7.4, containing 10-2~ EDTA and 1% normal nurse shark 19S Ig. The immune complexes were then precipitated (at equivalence) with rabbit antiserum against guinea pig immunoglobulin at 37°C for 1 hour and 2.5°C for 12 hrs. The immune precipitates were washed twice with cold phosphate buffered saline and counted for radioactivity; corrections for tube binding (usually representing <10% of the total) were included in the calculations. The percent antigen precipitated was calculated by dividing the counts in the precipitate by the counts added (XlOO). Inhibition of the idiotypic reactions with homologous or heterologous molecules, fragments or polypeptide chains was accomplished by mixing an excess (see RESULTS) of unlabeled inhibitor with the radiolabeled homologous antigen prior to adding the guinea pig antiserum. The data were calculated and expressed as above. RESULTS ~unodiffusion

analysis

Each guinea pig antiserum was tested by immunodiffusion against the homologous antigen, pooled normal shark immunoglobulin and a battery of anti-streptococcal carbohydrate antibodies from other sharks (heterologous antibodies). Each of the antisera exhibited a precipitin band with each of the shark antibodies (or immunoglobulin) studied and some formed immune precipitates with a component preeent in the homologous, but not in the heterologous, shark antibody preparations (Figure 1). Also depicted in this figure is an illustration of the finding that adsorption of such positive guinea pig sera with normal shark immunoglobulin removed the visible reactions with the heterologous proteins. Additional immunodiffusion experiments (data not shown) indicated that such adsorbed (presumably idiotypically positive) guinea pig antisera did not form immune precipitates with any of the other 9 heterologous shark proteins examined. It should also be mentioned that the antigen in whole shark serum demonstrable by immunodiffusion analysis with these guinea pig antisera was restricted to those homologous shark sera containing significant levels of antibody to the group A-variant streptococcal antigen (see 6). The quantitative features of these studies can be summarized as follows: Sera from 3/3 guinea pigs immunized with antibody from shark 225, 2/3 with antibodies from sharks 19 and 1903 and 1/3 with antibodies from shark Frances were judged by immunodiffusion criteria to contain antiidiotypic antibodies for the respective shark proteins. Since each of the adsorbed guinea pig antisera for each shark gave reactions of identity, the ~ositive sera were

466

SHARK IDIOTYPES

Vol. 6, No. 3

2

2

•

1

5• FIG. 1 Immunodiffusion analysis of the reactions between guinea pig antiserum to nurse shark 225 antibody to the A-variant streptococcal carbohydrate and various nurse shark proteins. Well contents: well 1, nurse shark 19S Ig; well 2, antibody from nurse shark 225; wells 3,4 and 5, antibodies from nurse sharks 19, Frances and 1904 respectively; well A, unadsorbed guinea pig antiserum; well B, guinea pig antiserum adsorbed with normal nurse shark 195 immunoglobulin.

pooled for each shark and used in the studies reported below. It should also be mentioned that 0/3 guinea pigs immunized with antibodies from sharks 226 and 269 responded with the production of detectable anti-idiotypic antibodies. Crossreactivity of shark idiotypes Since none of the adsorbed guinea pig anti-idiotypic sera formed immune precipitates with any of the heterologous shark antibodies, it was decided to employ a more sensitive radioimmunoassay to detect possible cross reactive idiotypes. The assay was based upon inhibiting, with an excess of unlabeled protein, the homologous binding reaction between labeled shark antibody and the guinea pig antiserum. The assays were constructed such that a ten fold excess of unlabeled homologous inhibitor would totally inhibit binding of the homologous radiolabeled protein; the amounts of radiolabeled protein used were 2 ~g for sharks 1904 and 225 and 2.5 ~g for shark 19 and represented the minimal amounts of material that was maximally precipitated by the homologous antiserum. The addition of a 50 fold excess of unlabeled heterologous inhibitor should give significant inhibition if as little as 5% of the molecules contain the idiotypic determinant in question. It should be pointed out that significant amounts of the radioactivity in each of the shark antibody preparations were specifically precipitated by the homologous guinea pig antiserum, i.e. 48%, 49%, 24% and 8% for sharks 225, 1904, 19 and Frances respectively. In no case was a significant amount of radioactivity bound by any of the four heterologous antisera. The results of the inhibition experiments employing homologous and heterologous shark 19S antibodies are shown in Table 1. It can be seen that, with two exceptions, none of the heterologous shark proteins gave significient inhibition of the homologous reaction. No inhibition of the

467

SHARK IDIOTYPES

Vol. 6, No. 3

TABLE I

Inhibition of the Homologous Idiotypic-Anti-Idiotypic Reactions of Nurse Shark Antibodies to the Group A-variant Streptococcal Carbohydrate with Antibodies from other Sharks. % Shark 225 Ab

Precipitated

% Shark 19 Ab Precipitated

None

48.:t2t

24+2

Shark Ig

50.:!::2

23.±1

225 {60 day)

<1

20+2

225 {120 day)

<1

21+2

51±2

19 (75 day)

50.:!::2

<1

52+2

19 (108 day)

49±2

<1

51±2

1904 (61 day)

51.:!::3

22+2

<1

1904 (221 day)

50~:.2

Inhibitor *

226

47±1

% Shark 1904 Ab

Precipitated

<1

21+2

13

22+1

15

22+2

269

20.±_2

267

21+1

52+2

261

49±2

22+2

48+2

259

47±1

22+2

22+2

97

24.±_1

46±2

9

23±1

49±1

Frances

21+1

52+2

69 (Ab to DNP)

23+1

29 (Ab to DNP)

25+3

51+2

"' All

inhibitors were antibodies {from the indicated shark and days postimmunization) to the group A-variant streptococcal carbohydrate except where otherwise noted and were added at a 50 fold excess over the radiolabeled antigen.

t

The values indicated represent the means of triplicate determinations + the standard deviations. The values underlined represent the significant cross reactions.

Vol. 6, No. 3

SHARK IDIOTYPES

468

rather weak homologous reaction with nurse shark Frances was observed (data not shown). The exceptions noted were in the cases where the shark 225 homologous system was inhibited (~17%) by antibodies from shark 15 and the shark 1904 system was inhibited (~55%) by antibodies from shark 259. Since kinetic experiments comparing the inhibition by different amounts of these two heterologous proteins with the homologous inhibitors indicated quite different affinities (Figure 2), it is suggested that these heterologous reactions represented cross-reactive idiotypes.

NO INHIBITOR

50 ,/

20

40

60

80

100

120

lltJ UNLABELED INHIBITOR FIG. 2 Inhibition of the idiotypic-anti-idiotypic reaction of nurse shark 1904 e, shark 1904) and heterologous with various amounts of homologous (• (o----o, shark 259) antibodies to the group A-variant streptococcal carbohydrate. Localization of idiotypic determinants Since idiotypic determinants on antibody molecules are generally considered to represent antigenic determinants associated with the variable regions of the H and/or L chains, attempts at localizing the putative shark idiotypes to various parts of the shark antibody molecules were undertaken. Preliminary immunodiffusion experiments involving the use of Fab fragments and reductive 7S subunits from antibodies from sharks 225, 19 and 1904 indicated that none of these formed immune precipitates with the homologous adsorbed guinea pig antiserum; each however appeared to inhibit the homologous precipitation reaction (see 10). On the basis of these observations, studies involving the inhibition of the homologous radioimmune assay (as discussed in the preceeding section) were undertaken. Both Fab fragments and reductive subunits from shark 225 were capable of totally inhibiting the homologous reactions (Table II); similar results were obtained with sharks 19 and 1904. Although no inhibition of the antigen binding reaction was

Vol. 6, No. 3

SHARK IDIOTYPES

469

TABLE II Inhibition of the Homologous Idiotypic-Anti-Id iotypic Reaction of Nurse Shark Number 225 with Fragments and Polypeptide Chains.

% Shark 225 Ab Precipitated

Inhibitor

49::!:3

None 225 Ab

<1

225-7S Subunits

<1

225-Fab Fragments

<1

*

225-L Chains 225-H Chains

27::!:1

Ig-L Chains

51+2

Ig-H Chains

51+3

Recombinants <1

225 H -225 L 225 H -Ig L

Ig H - 225 L Ig H - Ig L

*The

values indicated represent the means of triplicate determinations the standard deviations. The inhibitors were added to the reaction mixtures at a SO fold molar excess over the radiolabeled antigen.

±

observed with mildly reduced and alkylated L chains, some inhibition was seen With H chains. Since mildly reduced shark H chains routinely contain a small amount (~5%) of L chains, the possibility existed that the H chain inhibition was due to unseparated H-L chain pairs. This interpretation was substantiated by the observation (with sharks 225 and 1904) that ten fold less material failed to significantly inhibit the homologous reaction, Furthermore, recombination of the appropriate shark antibody H and L chains to 7S subunits (11) appeared to totally restore the inhibitory activity (Table II). Hence it seems likely that full expression of the shark idiotypic determinants requires participation of both the H and L chains. DISCUSSION This study was undertaken to assess the degree of combining site

470

SHARK IDIOTYPES

Vol. 6, No. 3

heterogeneity of shark antibodies to the streptococcal A-variant carbohydrate. Previous studies had indicated considerable restriction in electrophoretic heterogeneity of antibodies from some sharks whereas considerable heterogenity was seen with antibodies from other sharks. Similarly, the electrophoretic mobilities of those antibodies exhibiting restricted heterogeneity appeared to differ considerably from one animal to another (6). While these observations suggest considerable diversity, of some unknown nature, between shark antibodies to a defined antigen, they do not really address the question of combining site heterogenity. To this end alternative methods for probing shark antibody site structural features were sought. Idiotypes are antigenic determinants situated near or in the antibody combining site and are generated by the V-regions of H and/or L chains. It therefore seemed reasonable to search for idiotypes on shark antibodies. The results presented here support the idea that some of the guinea pig antisera against shark antibodies did in fact contain anti-idiotypic antibodies. The findings that the adsorbed guinea pig antisera did not react with normal shark immunoglobulin or exhibit extensive crossreactions with antibodies from other sharks is suggestive of an anti-idiotypic specificity. Similarly the apparent monovalent nature (inhibitory but not precipitating) of the homologous Fab fragment reaction is a feature expected of an idiotypic reaction. However, in the absence of ligand inhibition (an experiment not possible with the streptococcal A-variant carbohydrate due to the lack of a suitable hapten, see 12), a major concern in establishing the idiotypic nature of an antigenic determinant using xenogeneic antisera is whether or not the determinant being detected might be of an "hidden" isogeneic or allogeneic nature in very low concentrations in normal serum. It would seem that these possibilities with the shark antibodies are made unlikely, but unfortunately not eliminated, by the results obtained with H and L chains. It is difficult to see how isogeneic or allogeneic determinants would require the "proper" H-L chain combination. Hence it seems more likely that the shark determinants were idiotypic (and by inference involved the antibody combining site region). The data presented here suggest that the idiotypic library of nurse sharks to the streptococcal antigen may be relatively large. Documentation of at least four primary (or homologous) idiotypes and two cross reacting idiotypes has been obtained. Similarly, the finding that significant amounts of the radiolabeled antibodies from each of the four idiotype positive sharks did not react with the homologous guinea pig antiserum suggests the presence of additional idiotypes. Similar reasoning may also be applied to the antibodies from sharks for which no idiotypic markers were defined. This indication of a rather large combining site repertoire of nurse sharks for a particular antigen is in contrast with the results obtained by others with carp (bony fish) wherein extensive idiotypic cross reactions were seen with antibodies to the DNP moiety from different animals (13). In this context it must be pointed out that while idiotypic analysis can be used to establish the presence of a large combining site repertoire in a species, it is of questionable value in establishing a limited repertoire. For example, the finding of extensive sharing of idiotypes between individuals of a species can be explained, at least in some cases, by a preferential expression of one idiotype over others, e.g. the phosphorylcholine and arsonate responses in mice (4,14). Thus the interpretation of the carp results as indicating a very limited repertoire of combining sites must be taken with caution. On the other hand, the results presented here with sharks do not suffer from this potential pitfall in that very little crossreactivity or shared idiotypy was seen.

Vol. 6, No. 3

SHARK IDIOTYPES

471

The findings presented here, indicating a wide idiotypic reportoire for shark antibodies to the streptococcal A-variant antigen, are consistent with the results obtained with shark antibodies to other antigens. For example, the observation that nurse shark antibodies to the DNP moiety (8) and to the S pneumococcal polysaccharide (11) exhibited heterogeneity of both ligand ~inding and isoelectric focusing patterns certainly suggest that in nurse sharks (either as individuals or as a species) every antibody site to a particular ligand is not the same as every other antibody site to that same ligand. Hence with the recent documentation of variable and constant regions in shark antibody L chains (7) it seems that the multi-V-region, discontinuous germ line gene arrangement followed by somatic cell rearrangement characteristic of the antibody capability of higher animals (16) may be ubiquitous among vertebrates. This should come as no surprise to those who believe that the immune response has survival value, A "primitive" reportoire of potential combining sites, while perhaps appealing on a developmental basis, would certainly not have the flexibility that may be required to confer survival value; It might be appropriate in closing to suggest that while considerable evolution (i.e. phylogenetic changes) of H chain C-regions may have occurred in the different vertebrates, perhaps few, if any, changes have occurred in the "business" end (combining site) of the molecule. REFERENCES 1.

NISONOFF, A., HOPPER, J, and SPRING, S, Academic Press, N.Y., 1975.

2.

SEIDMAN, J., LEDER, A., NAU, M., NORMAN, B. and LEDER, P. diversity. Science 202, 11, 1978.

3.

CARTON, Y. La response immunitaire chez les agnathes et les poissons. Structure des immunoglobulins. Ann. Biol, 12, 139, 1973.

4.

RODKEY, J.S. Autoregulation of immune responses via idiotype network interactions. Microbiol. Revs. 44, 631, 1980.

5.

CLEM, L.W. and LESLIE, G.A. Production of 19S IgM antibodies with restricted heterogeneity from sharks. Proc. Nat. Acad. Sci., U.S.A. 68, 139, 1971.

6.

CLEM, L.W., McLEAN, W.E. and SHANKEY, V. Qualitative and quantitative aspects of the antibody library of sharks. Adv. Exp. Med. Biol. 64, 231, 1975.

7.

CLEM, L.W. ~ LESLIE, G.A. Phylogeny of immunoglobulin structure and function. XIV. Peptide map and amino acid composition studies of shark antibody light chains. (Submitted)

8.

SHANKEY, v.~nd CLEM, L.W. Phylogeny of immunoglobulin structure and function. IX. Intramolecular heterogeneity of shark 19S IgM antibodies to the dinitrophenyl hapten, J. Immunol, 125, 2690, 1980.

9.

The Antibody Molecule, Antibody

SMALL, P.A., KLAPPER, D. and CLEM, L.W. Half-lives, body distribution and lack of interconversion of serum 19S and 7S IgM of sharks. J. Immunol, 105, 29, 1970.

472

SHARK IDIOTYPES

Vol. 6, No. 3

10.

LESLIE, G.A. and CLEM, L.W. Phylogenetic perspectives of idiotype regulation. In The Biologic Significance Immune Regulation (L. Ruben and E. Gershwin, eds.) Marcel Dekker Publ., New York, (in press).

11.

SHANKEY, v. and CLEM, L.W. Phylogeny of immunoglobulin structure and function. VIII. Intermolecular heterogeneity of shark 19S IgM antibodies to p~coccal polysaccharide. Mol. Immunol. 17, 365, 1980.

12.

SHANKEY, v. Heterogeneity of IgM ligand binding: an inter -or intramolecular phenomenon? Doctoral Dissertation, University of Florida, Gainesville, FL, 1977.

13.

MACHULLA, H., RICHTER, R. and AMBROSIUS, H. Study of antibody heterogeneity of carp. The idiotypic specificity of anti-DNP antibodies. Immunol. Letters 1, 329, 1980.

14.

EICHMANN, K. Expression and function of idiotypes on lymphocytes. Adv. Immunol. 26, 195, 1978.

15.

MAX, E., SEIDMAN, J. AND LEDER, P. Sequences of five potential recombination sites encoded close to an immunoglobulin K constant region gene. Proc. Nat. Acad. Sci., U.S.A. 76, 3450, 1979. Received : May 1981 Accepted : August 1981