THE JOURNAL OF Bro~,ocrcx~ CHEMISTRY Vol. 283, No. 8, Issue of March 25, pp 1832.1837, 1918

Printed m U.S.A.

Anti-glycosyl Antibodies TWO SETS OF ISOANTIBODIES WITH SPECIFICITY FOR DIFFERENT CARBOHYDRATE MOIETIES OF THE SAME GLYCOSYL ANTIGEN*

(Received for publication, July 22, 1977)

JOHN H. PAZUR, KEVIN L. DREHER, AND L. SCOTT FORSBERG

From the Department of Biochemistry and Biophysics, The Pennsylvania State University, University Park, Pennsylvania 16802

Two sets of anti-glycosyl antibodies have been isolated by affinity chromatography methods from the antisera of rab- bits immunized with a vaccine of nonviable cells of Strepto- coccus faecalis, strain N. Both types of antibodies are directed against a diheteroglycan of glucose and galactose present in the cell wall of this organism. The members of one set, anti-galactose antibodies, combine with the termi- nal galactose residues of the glycan and the members of the other set, anti-lactose antibodies, combine with terminal lactose residues of the same glycan. Each set of antibodies is composed of multiprotein components. The electrofocus- ing method had been used to isolate the individual antibody proteins in homogeneous states as shown by both electro- phoresis and ultracentrifugation techniques. Since the com- ponents of each set combine with the same structural unit of the antigen, they have been designated as isoantibodies. The sedimentation constants, electrophoretic properties, carbohydrate constituents, and amino acid compositions of the two sets of antibodies are recorded.

Anti-glycosyl antibodies are defined as those antibodies which are induced in the host by a glycosyl antigen and which combine with a specific carbohydrate moiety of that antigen. In a preliminary note (1) we have described the isolation of two sets of such antibodies from antisera of rabbits immunized with a vaccine of nonviable cells of Streptococcus fuecalis, strain N, which contains an antigenic diheteroglycan of glu- cose and galactose in its cell wall. The antibodies of one set are anti-galactose (anti-Gal) antibodies and these combine with the terminal galactose residues of the antigen while the antibodies of the other set are anti-lactose (anti-Lac) antibod- ies and these combine with terminal lactose residues of the same antigen. The isolation procedure involved adsorption of both types of antibodies on a lactosyl-Sepharose column and sequential elution with gala&se and lactose solutions. Also in the earlier studies (1, 2) it was shown by gel electrophoretic and electrofocusing methods that the anti-Gal antibodies from a particular rabbit consisted of six distinct proteins and the

* This work was taken in part from the MS thesis of K. L. D. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

anti-Lac antibodies consisted of 12 different proteins. Since the individual proteins of each set combine with the same type of structural moiety, i.e. terminal gala&se or terminal lac- tose units of the antigen, the term isoantibody is suggested for the individual members of each set. This report describes details of the procedures for isolating the two sets of anti- glycosyl antibodies and records the sedimentation constants, electrophoretic properties, carbohydrate constituents and amino acid compositions of the two types of anti-glycosyl antibodies. In addition results of electrofocusing experiments are presented showing that the isoantibodies can be resolved into individual components. By the latter technique antibodies directed against a natural antigen have been prepared in a homogeneous state (2). These antibodies should be especially useful for studies on antibody structure and homology (3-51, on the biosynthetic pathways (6) and regulatory mechanisms for antibody synthesis (7, 8).

EXPERIMENTAL PROCEDURES

Preparation of Antigen, Vaccine, and Antiserum -The antigenic glycan of glucose and galactose employed in this study was extracted from the cell wall of Streptococcus fuecalis, strain N, with 10% trichloroacetic acid and purified to homogeneity by fractional precip- itation with ethyl alcohol (9) and Bio-Gel filtration (10). The molec- ular structure of the antigen has been determined by use of a combined analytical scheme based on methylation, periodate oxida- tion, acetolysis, and enzymic hydrolysis (10, 11). Hapten inhibition studies have shown that galactose and lactose are inhibitors of the precipitin reaction and, therefore, terminal galactose and terminal lactose units are immunodeterminant groups of this antigen (9).

A vaccine of nonviable cells of S. faecalis, strain N, was prepared from freshly grown cells collected from a 500-ml culture. A formal- dehyde treatment was used to obtain a suspension of nonviable cells (10). After removal of the formaldehyde by washing the cells with 0.1 M phosphate buffer of pH 7.2 in 0.9% saline, the cells were suspended in 80 ml of sterile saline. The suspension gave an absorbance value of 1.5 at 600 nm, and this suspension has been used to immunize rabbits by intravenous injection. Details of the immunization procedure and regime have been described elsewhere (10). Antisera were prepared from blood samples drawn weekly from the rabbits by standard immunological methods. The antisera were analyzed for protein constituents by microzone electrophoresis. To date unpooled antisera from six different rabbits have been prepared by the above methods and subjected to affinity chromatography as described in the next section.

Zsolation of Anti-Gal and Anti-Lac Antibodies -Affinity chroma- tography (12) on lactosyl-Sepharose was employed for the fractiona- tion of the antiS. fuecalis serum. The lactosyl-Sepharose (13) was prepared from 12 g of cyanogen bromide-activated Sepharose 4B

1832

by guest on August 29, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Anti-glycosyl Antibodies 1833

(Pharmacia) and 50 mg of p-aminophenyl-/3-n-lactoside (Vega-Fox Biochemicals). The above quantities of materials were shaken in 20 ml of 0.1 M NaHC03, pH 9.0, for 24 h at 4”. The reaction product was washed with 3 volumes of the bicarbonate buffer. Remaining un- reacted groups on the Sepharose were blocked by treatment of the suspension with 100 ml of 1 M ethanolamine of pH 8.0 for 2 h at room temperature. The product was poured into a glass column (2.5 x 30 cm) and washed thoroughly with saline/O. 1 M phosphate buffer, pH 7.2. One milliliter of the antiserum was introduced on the column and washing was continued with the buffer. Unadsorbed protein was eluted from the column with approximately 200 ml of salinelphos- phate buffer. The absorbance of the eluate was monitored at 280 nm with an ISCO UV’ analyzer. When the base-line had returned to the original value, the column was washed with 50 ml of 0.1 M galactose solution followed by 50 ml of 0.1 M lactose solution. The fractions corresponding to the 280-nm absorbing material which had been eluted by the galactose and those which were eluted by lactose were combined separately. An equal volume of saturated ammonium sulfate solution was added to the combined samples and precipita- tion was allowed to proceed for 24 h at 4”. The precipitates which formed were collected by centrifugation and dissolved in 1 ml of saline/O.1 M phosphate buffer, pH 7.2. Comparable antibody samples from five different separations on 2.5 ml of 13th week antiserum from the same rabbit were combined and the resulting solutions were dialyzed exhaustively against the saline/phosphate buffer, pH 7.2. The protein content of the samples was measured by UV absorbance at 280 nm.

Anti-Gal and anti-Lac antibodies were also isolated from the antiserum by chromatography on two affinity columns, a galactosyl- Sepharose and a lactosyl-Sepharose. In this procedure the antiserum was first introduced onto a galactosyl-Sepharose column and, after the serum proteins were eluted with the buffer, the column was washed with 50 ml of 0.5 M galactose solution and then with 50 ml of 0.5 M lactose solution. The fractions obtained by elution with the galactose contained protein constituents, but those with lactose did not. The initial fractions containing the serum proteins were com- bined and concentrated by ultrafiltration. The residue was dissolved in 2 ml of saline/O.1 M phosphate buffer and the solution was introduced onto a 1actosylSepharose column. After the serum pro- tein had been washed through the column, 50 ml of 0.5 M lactose was used to elute the adsorbed antibodies. The fractions containing the antibodies were combined and concentrated as described above to obtain preparations of anti-Gal and anti-Lac antibodies.

Ultracentrifugation -The sedimentation rates for the anti-Gal antibodies and anti-Lac antibodies were determined by centrifuga- tion of 0.2 ml of 0.5% solution of the antibody preparations and a mixture of equal parts of both preparations in a sucrose gradient (5 to 40%) in a Spinco SW 65 rotor and model L-65 centrifuge at 65,000 rpm for 14 h. Samples of glucose oxidase and the antibody prepara- tions were centrifuged under identical conditions. From these data, the sedimentation constants and molecular weights of the anti-Gal and anti-Lac antibodies were calculated utilizing an empirical relationship (14). The sedimentation constant for the anti-Gal anti- bodies was also determined from data obtained by centrifugation of a 0.3-ml sample of the 0.5% solution of anti-Gal antibodies in 0.1 M

phosphate buffer in a model E ultracentrifuge at 60,000 rpm. Schlierin optics were employed and photographs were obtained every 8 min. From these data and a specific volume of 0.74 ml/g, which is typical for immunoglobulins, the sedimentation constant and the molecular weight of the anti-Gal antibodies were calculated (15).

Agar Diffusion-Samples of 10 ~1 of the 0.5% solution of the two antibody preparations were tested with the 0.1 and 0.5% solutions of the antigenic diheteroglycan for precipitin formation by the method of double diffusion in agar. A similar diffusion procedure was used to determine the immunoglobulin types of the two types of antibodies utilizing goat anti-IgA, anti-IgG and anti-IgM sera (Cappell Labo- ratories). Agar diffusion tests were also performed with the anti-Gal and anti-Lac antibodies in polyacrylamide gels on which the anti- bodies were resolved into individual components by electrophoresis. The unstained gels were embedded in agarose solution and after the agarose had solidified, a trough was cut along side of the gel. A 0.15% solution of the diheteroglycan was placed in the trough and diffusion was allowed to proceed at room temperature for 48 h.

Hapten Inhibition -Qualitative and quantitative precipitin tests ~-

’ The abbreviations used are: UV, ultraviolet; dansyl, 5-dimeth- ylaminonaphthalene-1-sulfonyl.

were performed by procedures previously described (9). For the antibody preparations containing 0.5% protein, optimum concentra- tion of antigen was 5 pg with both antibody types. This quantity of antigen was used in hapten inhibition tests with a variety of carbohydrates and carbohydrate derivatives. For the inhibition tests with the anti-Gal antibodies the concentration of inhibitors ranged from 0 to 200 mM while for the anti-Lac antibodies the concentration ranged from 0 to 5 mM.

Polyacrylamide Gel Electrophoresis -Samples of 50 fig of the anti- Gal and the anti-Lac antibody preparations were subjected to elec- trophoresis in 5% polyacrylamide gels (Buchler Instruments) at a current of 2.5 mA/gel for 6 h. A mixture of the two antibody preparations was also subjected to electrophoresis under the above conditions. One set of finished gels was stained with Coomassie blue G 250 and the stained gel was photographed by an indirect lighting method. A second set of unstained gels was used in agar diffusion tests as outlined in a preceding section. The location of the precipitin bands in the agar plate was compared with the location of the protein components in the stained gel. A third and fourth set of gels were stained with glycoprotein staining reagents, periodate-Schiff (16) and periodate-dansylhydrazine (17).

Carbohydrate Constituents -Samples of 2 mg of anti-Gal and anti- Lac antibody preparations were dissolved in 0.2 ml of 0.2 N HCl. The vials were sealed and heated in a boiling water bath for 2 h. The hydrolysates and reference carbohydrates (fucose, mannose, galac- tose, glucosamine, galactosamine, and neuraminic acid) were sub- jected to paper chromatography in a solvent system of n-butyl alcohobpyridinelwater (6:4:3) by volume). One chromatogram was stained with silver nitrate reagent (18) and another was sprayed with copper sulfate reagent followed by molybdic acid reagent (19). The R, values of the hydrolytic products and of reference compounds were determined.

A sample of 2 ml of the anti-Lac antibodies (10 mg of protein) was taken to dryness by lyophilization. The dried sample and a standard sample containing 0.1 mg of each standard (fucose, mannose, and galactose) were subjected separately to methylation analysis (20, 21) under identical reaction conditions. The final methyl alditol acetates from these samples were dissolved in a small volume of chloroform and samples of 1 ~1 were subjected to analysis on OV-225 in a Varian aerograph series 1400 gas chromatograph at temperature of 170”. The components which were separated were analyzed individually in a DuPont 21-490 mass spectrometer, and the nature and abun- dance of the m/e fragments were determined.

Amino Acid Analysis -Samples of 0.5 ml of the anti-Gal and anti- Lac antibody solutions (1 mg of antibody protein in each sample) were dialyzed against distilled water and then taken to dryness by lyophilization in glass vials. One-half milliliter of 6 N HCl was added and the tubes were sealed and heated at 110” for 24 h. The hydrochlo- ric acid was removed by drying in a vacuum desiccator containing NaOH pellets and the residue was dissolved in 1 ml of water. Samples of 0.1 ml of these solutions were used for amino acid analysis employing a Beckman model 120 amino acid analyzer.

Gradient El&ion -A sample of the anti-Lac preparation (10 mg) was used in a gradient elution experiment. The antibodies in the sample were adsorbed on the lactosyl-Sepharose column and the column was washed with a lactose gradient (0 to 0.5 M lactose). Fractions of 1 ml were collected from the column and all fractions were dialyzed against saline/phosphate buffer to remove the lactose. Samples of 10 ~1 of the fractions were subjected to polyacrylamide gel electrophoresis as described in a preceding section. Qualitative precipitin tests were performed with all fractions and the antigenic glycan.

Electrofocusing -Electrofocusing (22) of the antibody preparations was performed in an LKB instrument in a sucrose-stabilized pH- gradient (Ampholine 6 to 8). Samples of approximately 10 mg of the antibody preparations were introduced into a llO-ml gradient and electrofocusing was performed at 800 V for 65 h. Fractions of 1 ml were collected and monitored for UV absorbance at 280 nm. Dupli- cate samples of 100 ~1 of the peak fractions were used for polyacryl- amide gel electrophoresis as described in a preceding section. One set of gels was stained for protein components and the other was tested for antibody activity by the agar diffusion technique described above.

RESULTS

Microzone electrophoretic patterns of antisera from rabbits immunized with nonviable cells of Streptococcus faecalis,

by guest on August 29, 2018

http://ww

w.jbc.org/

Dow

nloaded from

1834 Anti-glycosyl Antibodies

strain N, contained high levels of immunoglobulins in the blood samples collected after the 10th week of immunization. Agar diffusion tests revealed that these sera contained anti- bodies directed against a diheteroglycan of glucose and galac- tose isolated from the cell wall of the organism. Hapten inhibition data showed that both lactose and galactose were inhibitors of the reaction between the diheteroglycan and the antiserum. This result was interpreted as evidence for the presence of two populations of antibodies, one which combined with the gala&se moiety and the other which combined with the lactose moiety of the antigen. Fig. 1 shows affinity chromatography elution patterns for preimmune serum and serum collected from a single rabbit after the 13th week of immunization. It will be noted in the figure that the preim- mune serum yielded only a UV-absorbing peak of serum proteins, but the immune serum yielded the serum protein peak and two other UV-absorbing peaks, one eluting with the galactose and the other eluting with lactose. The latter materials were recovered from the eluates as described under “Experimental Procedures.” From 12.5 ml of antiserum, 40 mg of anti-Gal antibodies and 60 mg of anti-Lac antibodies were obtained. The antibody preparations were stored in saline/O.1 M phosphate buffer, pH 7.2, at a concentration of 0.5% protein in the frozen state. All analyses were performed on the antibodies isolated from the serum of a single rabbit. Exami- nation of antisera from five other rabbits by the affinity chromatography technique revealed that all samples also contained two sets of antiglycosyl isoantibodies. However, the structural and immunological characterization of these sam- ples has not yet been undertaken.

That both preparations react with the diheteroglycan to yield antigen-antibody precipitins is shown in the agar diffu- sion patterns in Fig. 2. Other agar diffusion tests showed that both types of antibodies reacted with goat anti-IgG serum, but not with goat anti-IgA or anti-IgM sera. Thus, both the anti- Gal and anti-Lac antibodies are of the IgG immunoglobulin type.

Data on the sedimentation patterns obtained by centrifuga- tion on sucrose density gradients and in the analytical ultra- centrifuge have been published (1, 2). Both antibody prepara- tions yielded single symmetrical peaks on ultracentrifugation, characteristic of material of uniform molecular size. Utilizing these data and appropriate formulae (14, 15), sedimentation constants for both preparations were calculated to be 6.9 S and the molecular weights to be 1.5 x 10”.

A variety of carbohydrates and derivatives were tested as potential inhibitors of the precipitin reaction. Some of these

FRACTION NUMBER

FIG. 1. Elution patterns from lactosyl-Sepharose of pre-immune (too frame) and immune serum (bottom frame) from a rabbit , im&kized’ with nonviable cells of S. faecalis. The first arrow indicates the beginning of elution with 0.1 M galactose and the second arrow, with 0.1 M lactose.

data are plotted in Fig. 3. It will be noted in the figure that only lactose and lactosyl derivatives functioned as inhibitors for the anti-Lac antibodies, but galactose, lactose, and glyco- sides of these compounds functioned as inhibitors for the anti- Gal antibodies. It should also be noted that a IO-fold difference in the concentration of inhibitors for 50% inhibition of the anti-Gal and the anti-Lac antibodies existed.

Gel electrophoretic patterns of the anti-Gal and the anti-Lac antibodies were published earlier (l), showing that the anti- Gal antibodies consisted of six distinct proteins and the anti- Lac antibodies consisted of 12 different proteins. Gel electro- phoretic patterns of an equal mixture of the two types of antibodies yielded a composite pattern containing the individ- ual bands of each set. Unstained gels of anti-Gal and anti-Lac antibodies were tested by the agar diffusion method to deter- mine if the individual components were antibody proteins. In these tests it was found that all of the protein components in both antibody preparations yielded a precipitin band with the diheteroglycan. Other polyacrylamide gels of the anti-Gal and anti-Lac antibodies were stained with glycoprotein staining reagents, periodate-Schiff (16) and periodate-dansyl hydrazine (17). Positive tests were obtained with all components of the anti-Gal and the anti-Lac antibodies, showing that all of the components were glycoproteins.

On paper chromatographic analyses of dilute acid hydroly- sates of the anti-Gal and the anti-Lac antibodies two principal monosaccharide constituents with RF values of 0.28 and 0.11 were obtained from both preparations. These compounds in the hydrolysate of the anti-Gal preparation were isolated by preparative paper chromatography. The compound with RF value of 0.28 yielded a positive test with galactose oxidase (23, 24) and with the hexosamine reagent (2.5) while the compound with RF value of 0.11 yielded a positive test with the thiobar-

FIG. 2. Agar diffusion patterns for the anti-Gal and anti-Lac antibodies and the diheteroglycan from S. faecalis. Center wells contain the antibodies and theperipheral wells contain 0.1 and 0.5% solutions of the diheteroglycan.

ANTI-GAL ANTI-LAC

/NH,+-P-LAC

0 40 80 120 160 200 0 I 2 3 4 5 CONCENTRATION OF INHIBITOR (mM)

FIG. 3. Hapten inhibition of the anti-Gal and anti-Lac antibodies. MEL, melibiose.

by guest on August 29, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Anti-glycosyl Antibodies 1835

bituric acid reagent (26). These tests and the RF values establish the former to be galactosamine and the latter to be neuraminic acid. In the native antibodies the amino sugars are most likely in the N-acetylated form.

In the dilute acid hydrolysates, neutral sugars with RF values of fucose (0.50), mannose (0.411, and gala&se (0.33) were also faintly visible. The presence of these monosaccha- rides in the antibody preparation was established by methyl- ation analysis. The identities of the various components in the methylation mixture were established by retention times on gas chromatography on OV-225 at 170” and characteristic mle fragments on mass spectrometry. Thus the carbohydrate de- rivatives identified from the anti-Lac preparation were: 1,5-di- O-acetyl-2,3,4-tri-O-methyl-fucitol, retention time of 0.55 and m/e fragments of 117 (loo), 131 (401, 161 (201, and 175 (10); 1,3,4,5-tetra-O-acetyl-2-O-methyl-fucitol, retention time of 1.43 and m/e fragment of 117 (100); 1,5-di-O-acetyl-2,3,4,6- tetra-O-methyl-mannitol, retention time of 0.97 and m/e frag- ments of 45 (70), 117 (1001, 161 (50), and 205 (20); 1,5-di0 acetyl-2,3,4,6-tetra-O-methyl-galactitol, retention time of 1.17 and m/e fragments of 45 (80), 117 (loo), 161 (60), and 205 (30); 1,2,5-tria-acetyl-3,4,6-tri-O-methyl-mannitol, retention time of 1.80; 1,4,5-tri-O-acetyl-2,3,6-tri-O-methyl-mannitol, reten- tion time of 2.02 and m/e fragments of 45 (401, 117 (1001, and 233 (20) and 1,2,3,5-tetra-O-acetyl-4,6-di-O-methyl-mamitol, retention time of 2.72. The numbers in parentheses represent relative abundance of the fragments with the most abundant fragment assigned an arbitrary number of 100. These values are in agreement with published values (20) and values determined in this laboratory for standard compounds. Meth- ylated amino sugar derivatives were also detected on temper- ature program analysis of the methylation mixture but the exact nature of the derivatives has not yet been determined. Also the exact molecular structures of the oligosaccharide chains of the antibodies have not yet been determined. Work is in progress on this aspect of the structures of the anti-Gal and anti-Lac antibodies to determine whether these antibodies possess oligosaccharide chains with the basic structure of oligosaccharides from other types of immunoglobulins (27,281.

TABLE I

Number of amino acid residues per molecular weight of 150,000 for anti-Gal and anti-Lac antibodies

Residue Anti-Gal Anti-Lac

Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine l/z Cystine” Valine MethionineO Isoleucine Leucine Tyrosine Phenylalanine Histidine Lysine Arginine TrvptophanO

109 111 165 170 160 159 133 132 105 100 114 116 82 78

122 128

48 47 86 84 44 41 41 42 15 16 72 73 37 37

a Not yet determined for the anti-Gal and anti-Lac antibodies and in the calculations for this table, the values for the myeloma immunoglobulin (Ref. 30) have been employed for these amino acids.

Data on the amino acid composition of the antibodies have been obtained on the two sets of isoantibodies. The values recorded in Table I are average values of triplicate analyses on two different antibody preparations. Also correction factors of 1.13 and 1.25 have been employed in the calculations of the threonine and serine values (29). The number of residues of each amino acid per mole of antibody has been calculated for a molecular weight of 1.5 x lo”. Since cysteine, tryptophan, and methionine were not determined in our analyses, values for these amino acids were taken from Table I of Ref. 30 and assumed to be a reasonable estimate for these amino acids.

A partial separation of the anti-Lac isoantibodies was achieved by adsorption on lactosyl-Sepharose and elution with a lactose gradient of 0 to 0.5 M. The gel patterns obtained for several fractions from the column are reproduced in Fig. 4. The gradient elution technique resulted in a resolution of the components into several subsets. A resolution into individual components was achieved by electrofocusing. The electrofocus- ing pattern for the anti-Lac preparation and gel electropho- retie patterns for the individual components are shown in Fig. 5. The components are numbered with Roman numerals with

anti.lac 1 2 3 4

FIG. 4. Gel electrophoretic patterns for the anti-Lac antibody preparations and Fractions 1, 2, 3, and 4 obtained by elution of the antibodies from a 1actosylSepharose column with a lactose gradient.

I I I I I , I I t I , IO 20 30 40 50 69 70 80 90 loo 110

FRACTION NUMBER

FIG. 5. Electrofocusing and gel electrophoretic patterns for the anti-Lac isoantibodies and the individual components. Roman nu- merals, isoantibody components; M, original mixture of anti-Lac isoantibodies; R, reconstituted sample from the individual anti-Lac isoantibodies.

by guest on August 29, 2018

http://ww

w.jbc.org/

Dow

nloaded from

1836 Anti-glycosyl Antibodies

I being the component with the slowest electrophoretic mobil- ity. Similar results for the anti-Gal preparation have been published (2).

DISCUSSION

Two sets of antibodies directed at two different immunode- terminant groups of the same glycan have been isolated from antisera of a rabbit immunized with a vaccine of non-viable cells of Streptococcus faecalis, strain N, containing an anti- genie glycan of glucose and gala&se in its cell wall. The isolation was achieved by affinity chromatography on lactosyl- Sepharose and sequential elution with galactose and lactose solutions. One set, anti-Gal antibodies, is directed against terminal galactose units of the antigen and the other, anti- Lac antibodies, is directed against terminal lactose units of the same antigen. Gel electrophoresis and electrofocusing experiments revealed that both sets of antibodies consisted of several protein components with each component exhibiting antibody activity. Since the members of each set combine with the same structural unit of the antigen, i.e. terminal galactose or terminal lactose units, such antibodies are appropriately termed isoantibodies.

The isoantibodies of the anti-Gal set were separated into individual components by electrofocusing, and six distinct protein species were obtained. The isoantibodies of the anti- Lac set were also separated into individual components and 12 isoantibodies were obtained from this set. The analytical data reported in this paper have been obtained on the antibodies from a single rabbit. However, the antisera from five addi- tional rabbits have been analyzed and in every case two sets of isoantibodies were obtained. Isoantibodies are produced not only against streptococcal glycans, but recent reports show that other antigens such as pneumococcal glycans (31, 321, staphylococcal ribonuclease (33) and synthetic carbohydrate- protein conjugates (34) elicit immune responses, resulting in the synthesis of sets of isoantibodies directed at different structural units of the same antigen. This phenomenon is apparently a general property of the immune response.

To ascertain the chemical nature and the physical proper- ties of the two types of antibodies, experiments have been performed on the mixture of isoantibodies of each set which have been isolated form the serum of a single rabbit. Thus ultracentrifugation by the density gradient method and in the analytical ultracentrifuge yielded sharp symmetrical sedi- mentation patterns for both sets of antibodies, indicating that the antibodies of each set are of identical molecular size. Further, a mixture of the two types of antibodies yielded a symmetrical pattern showing the molecular weights of the two sets of antibodies were also identical. The molecular weight calculated for both types of antibodies from these data was 1.5 x lo”.

Agar diffusion tests (Fig. 2) showed that the anti-Gal and anti-Lac preparations reacted with the diheteroglycan. Hap- ten inhibition data (Fig. 3) verify that the anti-Gal antibodies combine with terminal gala&se units and the anti-Lac anti- bodies combine with terminal lactose units of the same anti- gen. Agar diffusion tests were also used to show that the individual components of each set possessed antibody activity and all are of the IgG class of immunoglobulins. Equine anti- lactose antibodies against the same diheteroglycan have been prepared in another laboratory (351, but these antibodies were shown to be of the IgM class. The immunization regime and the species difference may be responsible for the difference in immunoglobulin types. Also, anti-galactose antibodies di-

rected against terminal gala&se units of other antigens (36, 37) have been observed and such antibodies would be useful for comparative studies.

Some differences in the ability of the anti-Lac isoantibodies to bind 1actosylSepharose do exist. On elution of the antibod- ies from lactosyl-Sepharose with a lactose gradient, the anti- Lac antibodies eluted in subsets of two to five components. Apparently the binding constants for the lactosyl moiety and the individual isoantibodies vary sufficiently to allow for separation of the components. With the availability of the individual pure components, this property of the isoantibodies can be more fully explored.

The carbohydrate content and composition of the two sets of antibodies is very similar. In dilute acid hydrolysates of both preparations galactosamine and neuraminic acid, which in the native antibodies are likely acetylated, were present in highest concentration. The neutral sugars, fucose, mannose and galactose were detected as their alditol acetate derivatives on methylation analysis. The structure of the oligosaccharide side chains in the antibodies has not yet been determined. On the basis of the derivatives thus far identified, it appears that the oligosaccharides in the antibody proteins differ somewhat in structure from the oligosaccharides in myeloma immuno- globulins (27, 28). The above finding is not surprising in view of a recent report (38) on the variations in carbohydrate composition of IgG antibodies isolated from rabbits and chick- ens immunized with fluorescyly-globulin conjugates. The amino acid composition of the two sets of isoantibodies is very similar (Table I). The structural differences in the two sets of antibodies and among the individual members of each set most probably resided in the hypervariable regions of the antibody molecules and the electrophoretic differences may be due to different numbers of amide groups in the individual antibodies.

Our findings show that two different structural units of an antigenic diheteroglycan elicit immune responses leading to the synthesis of different types of proteins. Whether the phenomenon is related to the genetic constitution of the animals (39-411, to the manner of attachment of the antigen at the receptor sites or to some yet unknown factor needs to be ascertained. Some evidence for a role for the receptor sites has been presented in recent studies on the nature of the receptor sites of membrane glycoproteins for various types of lectins (42, 43). The overall result with the S. faecalis diheteroglycan is the initiation of two series of biochemical reactions, one leading to synthesis of anti-gala&se isoantibodies and the other to the synthesis of anti-lactose isoantibodies. Since it has been possible to separate the isoantibodies into individual components, these materials should be useful for studies on questions relating to antibody homology, antibody diversity, biosynthetic pathways and regulatory mechanisms for the synthesis of antibodies.

Acknowledgments-We express appreciation to the follow- ing individuals of this department for their assistance: A. T. Phillips with the ultracentrifugation, B. Lewis with the gel electrophoresis, T. Kirby with the amino acid analysis, and D. Sharkey with the electrofocusing.

REFERENCES

1. Pazur, J. H., Miller, K. B., Dreher, K. L., and Forsberg, L. S. (1976)Biochem.Biophys. Res. Commun. 70, 545-550

2. Pazur, J. H., and Dreher, K. L. (1977) Biochem. Biophys. Res. Commun. T&818-824

by guest on August 29, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Anti-glycosyl Antibodies 1837

3. 4.

5.

6. 7. 8. 9.

10.

11.

12.

13. 14.

15. 16.

17.

18.

19.

20.

21.

22. 23.

Krause, R. M. (1970) Adu. Immunol. 12, l-56 Edelman, G. M., and Gall, W. E. (1969) Annu. Reu. Biochem.

38, 415-466 Hood, L., Eichmann, K., Lackland, H., Krause, R. M., and

Ohms, J. J. (1970) Nature 228, 1040-1044 Capra, J. D., and Kindt, T. J. (1975)Zmmunogenetics 1, 417-427 Eichmann, K. (1975) Immunogenetics 2, 491-506 Hood, L. (1976) Fed. Proc. 35, 2158-2167 Pazur, J. H., Anderson, J. S., and Karakawa, W. W. (1971) J.

Biol. Chem. 246, 1793-1798 Pazur, J. H., Cepure, A., Kane, J. A., and Hellerquist, C. G.

(1973) J. Biol. Chem. 248, 279-284 Pazur, J. H., and Forsberg, L. S. (1978) Carbohydr. Res., 60,

167-178 Cuatrecasas, P., Wilcheck, M., and Antinsen, C. B. (1968)Proc.

Natl. Acad. Sci. U. S. A. 61, 636-643 Wofsy, L., and Burr, B. (1969) J. Immunol. 103, 380-382 Martin, R. G., and Ames, B. N. (1961)J. Biol. Chem. 236, 1372-

1379 Schachman, H. K. (1957) Methods Enzymol. 4, 32-103 Fairbanks, G., Steck, T. L., and Wallach, D. F. H. (1971)

Biochemistry 10, 2606-2617 Eckhardt, A. E., Hayes, C. E., and Goldstein, I. J. (1976) Anal.

Biochem. 73, 192-197 Mayer, F. C., and Larner, J. (1959) J. Am. Chem. Sot. 81, 188-

193 French, D., Knapp, D. W., and Pazur, J. H. (1950)5. Am. Chem.

Sot. 72, 5150-5152 Bjorndal, H., Helerquist, C. G., Lindberg, B., and Svensson, S.

(1970) Angew. Chem. Znt. Ed. Engl. 9, 610-619 Pazur, J. H., Dropkin, D. J., Dreher, K. L., Forsberg, L. S., and

Lowman. C. S. (1976)Arch. Biochem. BioDhvs. 176. 257-266 Valmet, E.’ (1968) Sci. Tools 15, 8-11 1 ” Avigad, G., Amaral, D., Asensio, C., and Horecker, B. L. (1962)

J. Biol. Chem. 237, 2736-2743

24.

25. 26. 27.

28.

29.

30.

31.

32.

33.

34.

35. 36.

37.

38.

39.

40.

41. 42.

Ostrand-Rosenberg, S. (1976) Immunogenetics 3, 53-64 Ahmann, G. B., and Sage, H. J. (1974) Cell. Immunol. 13, 407-

415 43. Gurd, J. W. (1977) Biochemistry 16, 369-374

Pazur, J. H., Knull, H. R., and Chevalier, G. E. (1977) J. Carbohydr. Nucleosides Nucleotides 4, 129-146

Blix. G. (1948) Acta. Chem. Stand. 2. 467-473 Warren, L. (1959) J. Biol. Chem. 234; 1971-1975 Clamp, J. R., and Putman, F. W. (1964) J. Biol. Chem. 239,

3233-3240 Kornfeld, R., Keller, J., Baenziger, J., and Kornfeld, S. (1971)

J. Biol. Chem. 246. 3259-3268 Junge, J. M., Stein,’ E. A., Neurath, H., and Fischer, E. H.

(1959) J. Biol. Chem. 234, 556-561 Edelman, G. M., Gall, W. E., Waxdal, M. J., and Konigsberg,

W. H. (1968) Biochemistry 7, 1950-1958 Freedman. M. H.. Guver. R. B.. and Terrv. W. D. (1972) J. Biol.

Chem. i47, 7051-7Okl ’ ’ “’ Chen, F. W., Strosbera. A. D.. and Haber. E. (1973) J. Immunol.

110, 98-106 -’ Curd, J. G., Schechter, A. N., and Anfmsen, C. B. (1975) Fed.

hoc. 34, 550 Kieda, C. M. T., Delmotte, F. M., and Monsigny, M. L. P. (1977)

Proc. Natl. Acad. Sci. U. S. A. 74. 168-172 Kim, Y. D., and Karush, F. (1973)Z~munochemistry 10,365-371 Glaudemans, C. P. J., Zissis, E.. and Jollev. M. E. (1974)

Carbohydr. Res. 40, 129-135 ”

Eichmann, K., Uhlenbruck, G., and Baldo, B. A. (1976) Immu- nochemistry 13, l-6

Mitchell, J. E., Conrad, H. E., and Voss, E. W., Jr. (1976) Immunochemistry 13, 659-666

Eichmann, K., Braun, D. G., and Krause, R. M. (1971) J. Erp. Med. 134, 48-65

Freedman, M., Yeger, H., Milandre, J., and Slaughter, M. (1975) Immunochemistq 12, 137-147

by guest on August 29, 2018

http://ww

w.jbc.org/

Dow

nloaded from

J H Pazur, K L Dresher and L S Forsbergcarbohydrate moieties of the same glycosyl antigen.

Anti-glycosyl antibodies. Two sets of isoantibodies with specificity for different

1978, 253:1832-1837.J. Biol. Chem. 

  http://www.jbc.org/content/253/6/1832Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/253/6/1832.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on August 29, 2018

http://ww

w.jbc.org/

Dow

nloaded from