Affinity-purified antigen-specific products produced by T cells share epitopes recognized by heterologous antisera raised against several different antigen-specific products from T cells

CELLULAR

IMMUNOLOGY

82, 232-245 (1983)

Affinity-Purified Antigen-Specific Products Produced by T Cells Share Epitopes Recognized by Heterologous Antisera Raised against Several Different Antigen-Specific Products from T Cells R. E. CONE,*” R. W. ROSENSTEIN,?C. A. JANEWAY,? G. M. IVERSON,~ J. H. MURRAY,+ H. CANTOR,~ M. FRESNO,$ J. A. MATTINGLY,” M. CRAMER,# U. KRA~INKEL,’ H. WIGZELL,** H. BINZ,~~ H. FRISCHNECHT,~-~W. F’TAK,$.# AND R. K. GERSHON~ *Departments of Pathology and Surgery, Yale University School of Medicine, New Haven, Connecticut 06510

*Institute for Genetics, University of Cologne, D-5000 Cologne 41, Federal Republic of Germany

THoward Hughes Medical Institute, Department of Pathology, Yale University School of Medicine, New Haven, Connecticut 06510

**Department of Immunology, Uppsala, Sweden

$Immunochemistry Division, New England Nuclear, Boston, Massachusetts 02118

ttlnstitute for Immunology and Virology, University of Zurich, POB CH 8028, Zurich, Switzerland

#Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115

t?Department of Experimental and Clinical Immunology, Copernicus School of Medicine, 31-121 Krakow, ul. Czysta 18, Poland

“Department of Medical Microbiology, Ohio State University, Columbus, Ohio 43210

Received July 13. 1983; accepted July 26, 1983 Heterologous antisera to murine or rat T-cell antigen-binding molecules (T-ABM) were raised in rabbits or sheep. The T-ABM used for immunization were purified by affinity for antigen and did not bear known immunoglobulin isotypes. T-ABM and anti-T-ABM were raised in three separate laboratories. Antisera to T-ABM were exchanged and tested for binding to TABM in three separate laboratories. Thus antisera to at least three distinct T-ABM were tested directly for binding to T-ABM or by adsorption of biological activity. Rabbit antisera to murine trinitrophenol (TNP)-specific T-ABM or rat AgB-specific T-ABM bound both murine or rat TABM, indicating evolutionary conservation of T-ABM. Similar results were found with sheep antisera to murine T-ABM. In addition, all heterologous anti-T-ABM antisera used bound murine ?“-ABM specific for TNP, 4-hydroxy-3-nitrophenyl acetate (NP), SRBC, or T-cell membrane proteins with similar structure. Thus, there is a commonality of antigenic determinants between various T-ABM and T-cell membrane homologues which may be T-cell surface receptors for foreign antigen.

INTRODUCTION Heteroantisera directed at shared (isotypic) epitopes of immunoglobulins have been used to identify constant region determinants on immunoglobulins and B-lymphocyte i To whom correspondence should be addressed. 232 0008-8749183 $3.00 Copyright 0 1983 by Academic Press. Inc. All rights of reproduction in any form revved

T-CELL

RECEPTOR

ISOTYPES

233

cell surface antigen recognition structures (l-3). These anti-isotype reagents have played a key role in distinguishing immunoglobulin classes with different functions. Thus, antisera specific for heavy isotypes identified the role of minor immunoglobulin species such as IgE* in allergic reactions (4) and IgM and IgD as B-cell antigen receptors (1, 3, 5, 6 ). Characterization of molecules that T lymphocytes use to distinguish antigens has proved to be a more difficult task than it was for B cells (7). In particular, the failure of anti-immunoglobulin constant region isotype reagents to identify T-cell receptors indicates that T-lymphocyte antigen recognition structures are not conventional immunoglobulin molecules (7-9). However, analysis of the combining site of T-cellderived antigen recognition structures suggests that such molecules react with antisera against a variable portion of immunoglobulin heavy chains (VH) and that the genes that code for these determinants on T cells are genetically linked with those encoding similar determinants on B cells and their products (7, 10-12). The mapping of genes determining possible T-cell receptor allotypes close to Igh-C (13) and the demonstration that polymorphic determinants linked to &h-V are used by some T-cell subsets to communicate with each other (14) are also consistent with this notion. Less information is available regarding the non-antigen-binding portion of T-cellderived antigen recognition structures. If such molecules bear isotypic determinants, heteroantisera to T-cell antigen recognition structure isotypes would be of great value for the isolation and discrimination of these structures and the identification of Tcell membrane recognition units for antigen. Several laboratories have reported the preparation of heteroantisera to T-cell-derived antigen recognition structures which can be used as described above ( 15- 18). In this report, we demonstrate that several heteroantisera prepared against different antigen-specific T-cell-derived products of rats or mice react with a large portion of murine T cells and with different antigenspecific products derived from rat or murine T cells. The results provide further support for the suggestion (15) that common antigenic determinants exist on T-cellderived antigen recognition structures. MATERIALS T-Cell-Derived

Antigen-Binding

AND

METHODS

Molecules (T-ABM)2 and Heterologous Anti-T-ABM

The experiments described represent a cross-testing of T-cell-derived antigen-specific molecules or heterologous antisera to such molecules generated in six different laboratories. The details of how the reagents were generated and the techniques used in their characterization can be found in the original manuscripts referenced in Table 1 and briefly below. T-ABM I. Le-anti-DA. This Lewis rat T-cell receptor recognizes the MHC antigens on DA rat cells. The molecules are from Lewis rat serum isolated by reaction with and elution from an anti-idiotype serum coupled to Sepharose and have an apparent molecular ’ Abbreviations used: SDS-PAGE, polyacrylamide gel electrophoresis in sodium dodecyl sulfate; ABM, antigen-binding molecules; TsF, antigen-specific (T-cell derived) suppressor factor; TNP, trinitrophenol; SRBC, sheep red blood cells; MHC, major histocompatibility complex.

234

CONE ET AL.

weight of approximately 70,000 (19). These proteins show no reaction to anti-rat immunoglobulin isotype reagents. ZZ. TNP-TsF. TNP-TsF is a TNP-specific, T-cell-derived suppressor factor from CBA mice. The molecules were isolated by adsorption to and specific elution from TNP coupled to Sepharose (20) and have an apparent molecular weight of 70,000. The hapten-affinity-purified material did not contain (a) immunoglobulin heavy or light chain isotype determinants, (b) murine albumin, (c) GP-70, (d) Iak or H-2k, (e) Lyt-1, or (f) Lyt-2 (15, 20). ZZZ. Clonal Ly-2 TsF: This polypeptide is obtained from an antigen-specific Tsuppressor clone with specific antigen binding and suppressive activity (MW 70,000. The molecule was purified to virtual homogeneity by sequential separation of Sephacryl S-200 and DEAE-cellulose columns followed by isoelectric focusing (21). IV. NP-specific T-cell receptors. 4-Hydroxy-3-nitrophenylacetyl @@)-specific Tcell receptor material produced by murine hapten-sensitized splenic T lymphocytes is isolated on hapten-coupled nylon disks (22, 23) and is 150,000 MW (bivalent) in terms of hapten binding. V. Serum anti-SRBC. Sheep erythrocyte (SRBC)-specific antigen-binding material is found in the serum of BALB/c mice 4 days after immunization with SRBC, but contains no immunoglobulin isotypes (24). These molecules were isolated by adsorption to an antiserum, anti-(anti-SRBC) made by immunization of sheep with SRBC to which anti-SRBC serum had been adsorbed. After removal of anti-immunoglobulin antibodies on mouse immunoglobulin affinity columns, the resulting antiserum was used to isolate the relevant immunizing antigen, i.e., the serum anti-SRBC (24).

Antisera to T-ABM I. Anti-(Le-anti-DA) made in rabbits (16). This serum had no demonstrable activity against anything found in nude mouse serum, nude rat serum, polyvalent rat immunoglobulin, rat serum albumin, mouse serum albumin, or fetal calf serum. Leanti-DA was not GP-70 because it is not glycosylated. ZZ. Anti-TNP-TsF made in rabbits (15). This antiserum was prepared by immunization of rabbits with TNP-specific TsF purified by hapten-affinity chromatography. Anti-TNP-TsF will not bind to bovine or mouse serum albumin, IgG, or B cells, but is reactive with TsF and normal T cells. Anti-TNP-TsF will not bind to T-cell membranes if adsorbed with affinity-purified TNP-TsF. ZZZ. Sheep anti-murine (SZWC-speciJic)TsF (17). C57BL/6 mice fed SRBC were a source of suppressor T cells producing 65,000-75,000 MW SRBC-specific suppressor factors. The factors were adsorbed to SRBC and the coated SRBC were used to immunize sheep. Affinity chromatography was performed using Sepharose 4B (Pharmacia) as the support matrix. Proteins were covalently attached to Sepharose by the cyanogen bromide technique (25).

Radioiodination of Proteins and Cells Protein iodinations and radiolabeling of cell membranes were performed by lactoperoxidase-catalyzed radioiodination as described (26). The final Hz02 concentration for surface labeling was 0.3 mM. Lactoperoxidase was obtained from Sigma. For labeling of cells, 2-5 X lo7 CBA/J thymocytes, spleen cells, or splenic T cells were

T-CELL

RECEPTOR

235

ISOTYPES

labeled with 2 mCi 12’1 (New England Nuclear, Boston, Mass.), as described (26). 1251-Labeled surface proteins were obtained by lysis of labeled cells with 0.05% (v/v) Triton X- 100 (1 ml detergent/lo7 cells) in Tris-EDTA buffer. Aliquots of dialyzed, centrifuged lysates were mixed with test and control antisera for l-2 hr at 4°C and immune complexes are insolubilized by precipitation with sheep anti-rabbit immunoglobulin or adsorption to Staphylococcus aureus, Cowan strain I or to antibodies coupled to Sepharose 4B beads, as described (15). Splenic T cells were purified by adherence of B cells to goat anti-mouse Ig plates, as described (27). Polyacrylamide

Gel Electrophoresis

in Sodium Dodecyl Sulfate (SDS-PAGE)

1251-labeled surface proteins were dissolved or eluted from S. Immunoprecipitated, aureus or Sepharose beads by boiling in SDS-PAGE sample buffer containing 2% 2mercaptoethanol and were resolved by SDS-PAGE, as described (15). ‘311-Labeled, reduced and alkylated MOPC 104E, bovine serum albumin, ovalbumin, or chymotrypsin were coelectrophoresed with immunoprecipitated proteins and served as molecular weight markers. RESULTS The data given below represent the results of several exchanges of antisera made against T-cell-derived antigen-binding molecules and/or T-ABM. We wished to determine the binding activity of three different heterologous antisera raised against Tcell ABM to different T-ABM. The products and antisera raised against them are summarized in Table 1. Analysis by SDS-PAGE

or Afinity-PuriJied

T-Cell-Derived Antigen-Binding

Molecules

Two of the polypeptides used in these studies have been purified by adsorption to anti-&e-anti-DA) idiotype affinity Sepharose beads (rat Le-anti-DA T-ABM) or TNPBGG Sepharose beads (murine TNP-TsF) and elution with glycine-HCl and MgC12 or TNP-ethylaminocaproic acid, respectively. These polypeptides were labeled with 12’1 and the radiolabeled proteins were resolved by SDS-PAGE. As shown in Fig. 1A, both 1251-labeled, reduced polypeptides were resolved into peaks with apparent molecular weights of 70,000. When nonreduced polypeptides were analyzed (Fig. TABLE T-Cell-Derived Antigen-Binding Antigen-binding molecule

1

Molecules and Heterologous Antisera to T-ABM Reference

Antiserum

Used in This Study Reference

Rat L&anti-DA

Binz and Wigzell (19)

Rabbit anti-(Lx-anti-DA)

Binz and Wigzell (16)

Murine

TNP-TsF

Rosenstein et al. (20)

Rabbit anti-(TNP-TsF)

Cone el al. ( 15)

Murine

serum anti-SRBC

Iverson

et al. (24)

Sheep anti-mu&e

Mattingly

Clonal Ly-2 SRBC TsF

Fresno

etal.

NP-specific T-cell

Kmwinkel et al. (22) Cramer and Krawinkel (23)

receptors

(2 I)

SRBC-TSF

etal

(I 7)

236

CONE ET AL. BSA

Relative

WA

Chymo

Mobility

RG. 1. SDS-PAGE protile of affinity-purified ‘2sI-labeled Le-anti-DA (0) and TNP-TsF (0) before. immunoprecipitation with anti-@-anti-DA) or anti-TNP-TsF in both reduced and unreduced (B) states. g, OVA, L, BSA, and chymo refer to mobilities of ‘3’I-labeled MOPC 104E c chains, ovalbumin, MOPC 104E light chains, bovine serum albumin, or chymotrypsin internal gel standards for each gel. Acrylamide concentration is 10%.

lB), some aggregated material not entering the gel was observed, but the major peak in the gel had an apparent molecular weight of 70,000. Thus, 1251-labeled murine TNP-TsF was indistinguishable in size from iz51-labeled rat Le-anti-DA and behaved in SDS gels as a monomeric molecular species. It should be stressed here that these molecules were not bound by anti-murine or anti-rat albumin ( 16, 20). Nor did antisera to these polypeptides contain anti-albumin activity (15, 20) as measured by the radioimmune assay or ELISA assays. Moreover, no mouse albumin was detected in murine TsF as determined by peptide mapping.

Reaction of Various Sera with Ajinity-PuriJied, Idiotypic Rat T-Cell Protein The binding of various antisera to Lewis rat T-cell proteins bearing the idiotypic determinants present on Le-anti-DA antibody was tested by immunoprecipitation or adsorption of these proteins to antisera or antibody coupled to Sepharose, respectively. Biosynthetically labeled, idiotype-bearing T-cell proteins were purified from culture fluids of Lewis T cells by culture of Lewis T cells with 3H-labeled amino acids and adsorption of culture medium to and elution from anti-@&anti-DA) serum immunoadsorbent. The 70,000 MW polypeptides purified in this manner were then used in binding assays to affinity columns conjugated with various anti-T-ABM sera. As shown in Table 2, these molecules were bound specifically by rabbit anti-(Le-antiDA protein), anti-TNP-TsF, and sheep anti-murine-SRBC-TsF. In similar experiments, affinity-purified Le-anti-DA protein was radiolabeled with lz51 (see Fig. 1) and the ‘251-labeled protein was immunoprecipitated with various antisera. The radioiodinated molecules were bound specifically and to a similar extent (Table 2) by rabbit anti-(Le-anti-DA protein) and rabbit anti-TNP-TsF. Moreover, the SDS-PAGE profiles of the immunoprecipitated molecules are identical (Fig. 2). The precipitated proteins show a major peak with an apparent molecular weight of 70,000 and other peaks with apparent molecular weights of 45,000 and 25,000. The change in profile of Le-anti-DA from that seen in Fig. 1 can most likely be attributed

T-CELL

RECEPTOR

237

ISOTYPES

TABLE 2 Binding of Affinity-Purified I-e-anti-DA Protein by Heterologous Anti-T-ABM

Sera

Binding of rat Lewis anti-DA protein isotope label [3H]Leucine (cpm in eluate/ cpm in filtrate X10-*)’

Antibody system Rabbit anti-@e-anti-DA) sepbarose Rabbit anti-TNP-TSF sepharose Sheep anti-SRBC-TsF sepharose Normal rabbit serum sepharose Rabbit anti-&e-anti-DA) serum + S. aureus Rabbit anti-TNP-TsF serum + S. aureus Normal rabbit serum + S. aureus

lz51 (cpm bound X10-4)b nt’ nt nt nt 3.7 2.8 0.56

160118 150135 140/60 25/180 nt nt nt

a A total of 2 X 10’ cpm was added to each immunoadsorbent. Anti-rat albumin and anti-rat polyvalent immunoglobulin immunoadsorbents gave results which were indistinguishable from normal serum controls. Results are representative of five experiments with the same result. Elution was with glycine-HCl buffer, pH 2.4. b A total of 6 X lo4 cpm was used for immunoprecipitation. ’ nt, not tested.

to proteolytic degradation occurring at the immunoprecipitation step, and is characteristic of all the T-ABM we have studied by this technique (I 5). Binding of Anti- T-ABM Sera to Murine

TNP- TsF

The results presented above suggest that murine T-ABM and rat T-ABM may have determinants that cross-react when recognized by antibodies prepared in rabbits and sheep. To test this possibility further, TNP-TsF (see Table 1) was bound to TNP20

OL 0

P

OVA

L

P

OVA

L

1

I

1

I

1

k

I

I

0.25

0.5

I

0.75 Relative

1.00

1

I

025

0.5

1

0.75

I.0

Mobility

FIG. 2. (A) SDS-PAGE profiles of affinity-purified 125-labeled Le-anti-DA after immunoprecipitation with anti-&e-anti-DA) (0) or anti-TNP-TsF (0) and sheep anti-rabbit serum and reduction of precipitate. See legend to Fig. 1 for standards. (B) SDA-PAGE profiles of affinity-purified lz51 TNP-TsF after immunoprecipitation with anti-TNP-TsF (0) or anti-&e-anti-DA) and sheep anti-rabbit serum (0) and reduction, alkylation of precipitate. See legend to Fig. I for standards.

238

CONE ET AL.

Sepharose, isolated by hapten elution, and radioiodinated with ‘251. Nonlabeled TNPTsF was also used to prepare microtiter plates for radioimmunoassay or ELISA tests. The latter procedures detected specific binding of anti-&e-anti-DA), anti-SRBC-TsF, and anti-TNP-TsF to TNP-TsF (data not shown). When TNP-TsF was immunoprecipitated with either anti-&e-anti-DA) or antiTNP-TsF, again identical amounts (40-60% of i2’I-labeled protein) were bound and identical SDS-PAGE profiles were obtained (Fig. 2). In this instance, as seen in Fig. 2 with Le-anti-DA, some degradation of the nonimmunoprecipitated affinity-purified material (Fig. 1) was seen after immunoprecipitation. Precipitation of these proteins by rabbit anti-mouse albumin was no different than that using normal serum controls (approx l-2% of *251-labeled protein were bound). Binding of Various Anti-T-ABM ceptor Material

Heteroantisera

to Isolated, NP-Specific,

T-Cell Re-

Hapten-specific T-cell receptor material can be isolated on hapten-coupled nylon discs from sensitized splenic T lymphocytes of an as yet unknown functional activity (Table 1). This material is serologically, idiotypically, and genetically very well characterized (22, 23) and represents an intrinsic T-cell product (23). Murine NP-specific T-cell receptors of C57BL/6 origin were bound to immunoabsorbents prepared from three directed against three seemingly unrelated T-ABM (Table 3). The results show that all three antisera contain antibodies which are reactive with isolated NP-specific 20

66K I

I

G

45K I

25K I

.

z 5 “9

IO ; u”

20

68K

45K

25K

I

L

I

Spleen

0’ 0

0.2

0.4

Relative

L 0.6

0.0

, 1.0

Mobility

FIG. 3. SDS-PAGE profiles of reduced, ‘251-labeled T-cell membrane proteins immunoprecipitated with anti-&e-anti-DA) (0) or anti-TNP-TsF (0) and sheep anti-rabbit serum. No ‘231-labeled polypeptide peaks were resolved in precipitates with NRS and sheep anti-rabbit serum (X). See legend to Fig. 1 for standards.

T-CELL

RECEPTOR

239

ISOTYPES

TABLE 3 Binding of Anti-T-ABM

Sera to NP-Specific T-Cell Receptors’ of C57BL/6 Origin Isolated NP-specific T-cell receptor material b

Immunoabsorbent

A’

B

C

-

D

E

0

5

35 0

23 6

9%Absorption d Normal rabbit serum Anti-&e-anti-DA) Anti-TNP-TsF Anti-SRBC-TsF Anti-mouse Ig

3

6

0

24

45

4

3

87 5

a Binding of NP-specific material to insolubilized antisera is monitored by titrating the materials before and after absorption in the haptenated bacteriophage inactivation assay. * NP-specific receptor material is obtained from spleen cells. In a first absorption step, the B-cell receptor material is removed by absorption to a polyspecific rabbit or goat anti-mouse Ig immunoabsorbent. The remaining antGIg- fraction of NP-specific molecules represents T-cell receptor material (17). In a second absorption step this material is absorbed on the respective immunoabsorbents. ’ A-E stands for individual T-cell receptor preparations. Apart from the normal rabbit serum and the anti-mouse Ig controls, absorptions with anti-idiotypic reagents-conventional or monoclonal-and/or anti-VH antisera were performed in parallel in every experiment. d For the normal rabbit serum control the first absorption step with anti-mouse Ig served as a reference. The other data given are relative to this normal rabbit serum control.

T-cell receptors. The degree of cross-reactivity seems to depend on both the antiserum and the individual T-cell receptor preparation. Isolation of a Molecule in Hyperimmune Mouse Anti-sheep Red Blood Cell Serum (Serum Anti-SRBC) That Reacts with Heteroantisera to T-ABM Using antiserum (anti-(serum anti-SRBC) (Table 1) coupled to Sepharose that defines the antigen combining site of SRBC-specific Lyt-2+ T cells, antigen-binding material was purified from the serum of mice hyperimmunized with SRBC (serum anti-SRBC) (Table 1). After clearing all Ig+ material in the isolate by passage through a goat anti-mouse Ig column, the non-Ig, antigen-specific material in hyperimmune serum was isolated and radiolabeled (‘251). On SDS-PAGE, serum anti-SRBC closely resembles the molecules previously described in this paper (data not shown). Isolated, unlabeled material was tested for binding to anti-(Le-anti-DA), anti-TNP-TsF, and anti-SRBC-TsF (Table 4). All three of these sera reacted with serum anti-SRBC (Table 4), while anti-Ig sera did not bind this protein (data not shown). These results demonstrate the potential use of the various heteroantisera to characterize ABM in the serum that do not express conventional Ig markers. Reaction of Biosynthetically Labeled, Antigen-SpeciJic Suppressor Factor from a Cloned Suppressor T-Cell Line with Heteroantisera to T-ABM Fresno et al. (21) have recently described the purification of a biologically active suppressor T-cell factor from a T-cell line (clonal Lyt-2 TsF, Table 1). This material is specific for sheep glycophorin and resembles molecules l-4 in Table 1 in all pa-

240

CONE ET AL. TABLE 4 Binding of Various Antisera to Serum Anti-SRBC Serum 1. 2. 3. 4. 5. 6.

cpm bound”

Normal rabbit serum Normal sheep serum Anti-TNP-TsF Anti-SRBC-TsF Anti-&e-anti-DA) Anti-&rum anti-SRBC)

50 100 1800 1900 1900

1800

’ Serum anti-SRBC (1 &well) was put on plastic (Cook microtiter wells), and a 1: 10 dilution of test materials was then added. The number of cpm of ‘%labeled anti-rabbit Ig (groups 1, 3, 5) or anti-sheep Ig (groups 2, 4, 6) that bound to the antigen-antibody complexes in the wells was then determined.

rameters studied. This material was labeled biosynthetically (35S]methionine) and purified by size and charge. It was then tested for reactivity with anti-TNP-TsF. The results (Table 5) show that all of the antigen-binding material from the cloned cells bound to anti-TNP-TsF. Thus, anti-T-ABM specificity in this serum was unambiguously confirmed.

Isolation of T-Cell Membrane Proteins Using Heteroantisera to T-ABM The results presented above demonstrated that murine or rat T-ABM express similar and/or cross-reactive epitopes recognized by heteroantisera produced by rabbits or sheep. Moreover, rat &e-anti-DA) and murine (TNP-TsF) T-ABM are structurally similar. The commonality of epitopes expressed by T-ABM specific for four distinct antigens that are bound by heterologous anti-T-ABM sera suggests strongly that TABM bear isotypic determinants. Conceivably, T-cell membrane antigen recognition structures will also express such shared isotypes and react with these sera. To test this possibility, various T-cell preparations were surface radioiodinated and lysed with nonionic detergent and radiolabeled proteins were immunoprecipitated with various anti-T-ABM sera. As shown in Fig. 3, SDS-PAGE profiles of T-cell membrane proteins bound by rabbit antimurine TNP-TsF or anti-&e-anti-DA) were similar. As noted previously ( 15), resolution by SDS-PAGE of such immunoprecipitated material under reducing conditions tends to give three peaks of 70,000,45,000, and 25,000 molecular TABLE 5 Binding of Biosynthetically Labeled” Affinity-Purified Clonal Ly-2 TsF to Anti-TNP-TsF Conjugated to Sepharoseb Material on immunoabsorbent

cpm in eluate’/ cpm in liltrate (X IO-*)

Rabbit gamma globulin Rabbit anti-TNP-TsF

3196 99/l

D[3%]Methionine. b A total of 10’ cpm was added. cGlycine-HCl buffer, pH 2.4.

T-CELL

RECEPTOR

241

ISOTYPES

weight. That three different anti-T-ABM sera bind the same family of T-cell surface molecules was tested by adsorption of ‘251-labeled murine T-cell surface proteins to anti-TNP-TsF Sepharose and elution of these proteins. The ability of several anti-TABM sera to bind similar T-cell membrane proteins was tested by immunoprecipitation of anti-TNP-TsF Sepharose-isolated membrane proteins by other anti-T-ABM sera. As shown in Table 6, rabbit anti-&e-anti-DA) and sheep anti-mm-me SRBC-TsF all bind T-cell surface proteins isolated with anti-TNP-TsF Sepharose. Moreover, precipitation of ‘251-labeled T-cell surface proteins with anti-TNP-TsF Sepharose or anti&e-anti-DA) serum removes the majority of proteins detected by all three antisera in a sequential precipitation of the detergent lysate (Table 7). Thus, heteroantisera raised against three independently obtained mouse or rat T-cell ABM all react with the same family of molecules found on the membrane of mouse T lymphocytes. In addition to using the various antisera to precipitate solubilized membrane proteins, we also used them to detect the antigens on intact cell membranes by indirect immunofluorescence. The cellular distribution of the antigens detected by the anti-TNPTsF has been reported (15). The pattern of staining murine lymphocytes with anti&e-anti-DA) is similar to anti-TNP-TsF. Eighty percent of thymocytes were weakly stained, 80% of Thy-l+ spleen cells and essentially all Lyt-l-, 2+ T cells were brightly stained, and 50% of Lyt-l+, 2- T cells were stained in dull fluorescence. On the other hand, anti-SRBC-TsF did not stain the cells, suggesting that this reagent bound membrane determinants which are not exposed on the surface on intact cells. DISCUSSION The results we have reported demonstrate a commonality of epitopes between Tcell-derived antigen-binding molecules obtained from rats or mice. The antigen-binding molecules were obtained from serum, the supematants of cultured cells, or an Lyt2+ T-cell clone and were purified by adsorption to antigen or anti-idiotype affinity adsorbents. Despite the various origins of these T-cell-produced materials, we were able to raise antisera in rabbits or sheep to murine or rat T-ABM which cross-react with many T-ABM. These findings are summarized in Table 8. TABLE 6 Rebinding of ‘2sI-I.abeled T-Cell Membrane Protein Initially Isolated with Anti-TNP-TsF or Other Anti-T-ABM Sera” Antiserum on immunoabsorbent used for reisolation

cpm (X10e4)*

Anti-TNP-TsF Rabbit gamma globulin Anti-&e-anti-DA) Normal rabbit serum Anti-SRBC-TsF Normal sheep serum

2.6 0.4 2.1 0.5 2.0 0.5

by the Same

’ Detergent lysates of splenic T cells labeled with “‘1 were absorbed to anti-TNP-TsF Sepharose and bound proteins were eluted with 0.2 M NazC03. After dialysis into 0.05% Triton X-IOO-Tris-HCl buffer, the labeled proteins were tested with the reagents listed above. ’ 4.5 X 10’ cpm of labeled membrane proteins was added.

242

CONE ET AL. TABLE I

Precipitation of 1251-Labeled T-Cell Membrane Proteins Bound by One Anti-T-ABM Bound by Other Anti-T-ABM Sera”

Removes Protein

% Removal by 6rst precipitation of specific cpm bound by the following sera* Immunoabsorbent used in first precipitation Anti-TNP-TsF Anti-&e-anti-DA)

Anti-TNP-TsF

Anti-&e-anti-DA)

89 80

Anti-SRBC-TsF

65 80

80 NT

a Detergent lysates of splenic T cells labeled with i*‘I by lactoperoxidase catalyzed radioiodination were reacted with either anti-TNP-TsF or normal rabbit gamma globulin or anti-(&anti-DA) or normal rabbit serum and then removed by treatment with sheep anti-rabbit Ig serum. Supematants from the lirst precipitation were then precipitated with the above reagents. *Counts bound by control reagents (normal rabbit gamma globulin or normal rabbit serum) were subtracted before the percentage was calculated.

Most of the T-ABM studied display a marked structural similarity in which the antigen-binding polypeptide is approximately 70,000 MW. The rat and murine TABM also appear to be monomeric polypeptides, although sometimes noncovalently associated oligomers have been found. It should be stressed that no murine or rat albumin has been detected in these preparations. Nor do the antisera have antialbumin activity. The NP-specific T-ABM differ from those above in that these molecules appear to be 150,000 MW dimers and the molecule is bivalent in hapten binding (22, 23). It is unclear whether the latter proteins represent another family of T-ABM (e.g., membrane associated receptors) or the differences represent aggregation changes, etc. Many of the T-ABM studied show the same propensity toward breakdown and/ or dissociation to 45,000 and 25,000 MW polypeptides. This apparent pleomorphism of T-ABM indicates that caution should be exercised in comparing the structure of different T-ABM. Despite some differences in the NP-specific T-cell receptors studied, TABLE 8 Summary Antisera used to identify T-cell products0 T-Cell nroduct studied 1. 2. 3. 4. 5. 6. 7.

L&anti-DA TNP-TsF SRBC-TsF NP-specific T-cell receptors Serum anti-SRBC Clonal TsF (a) T-membrane isolates (b) T-membrane staining

Anti(Le-anti-DA)

AntiTNP-TsF

AntiSRBC-TsF

+ + NT + + NT + +

+ +

+ + +

NT + + + + +

+ + NT f -

Anti-(serumanti-SRBC)

t-1 t-1 NT NT (G (+I (-)

a +, T-cell product bound by antisera; -, T-cell product not bound by antisera; ( ), data not presented in this manuscript; NT, not tested.

T-CELL

RECEPTOR

ISOTYPES

243

it should be stressed that three different heterologous anti-T-ABM reacted with these proteins. The tendency of the 70,000 MW T-ABM to break down into 45,000 and 25,000 MW proteins suggests that this instability may be related to the function of the molecule. Conceivably, this breakdown could be considered to be a separation of three domains, each the approximate size of a light chain. This analysis would predict that the 45,000 MW piece is likely to be composed of two separate domains. Preliminary evidence from studies with clonal Lyt-2 TsF indicates that a 25,000 MW piece binds antigen and anti-(serum anti-SRBC) but not anti-TNP-TsF and is not suppressive while a 45,000 MW piece does not bind antigen, does not react with anti-(serum anti-SRBC), does react with anti-TNP-TsF, and is nonspecifically suppressive at high concentrations (M. Fresno, unpublished observations). The other antisera have not yet been used in these studies. Thus, a 45,000 MW piece appears to be responsible for the factor’s biological activity and carries the shared isotypic determinant, while a 25,000 MW piece gives the molecule antigen specificity (21). It should be stressed here that the binding of clonal Lyt-Zderived, antigen-specific TsF with rabbit anti-TNP-TsF provides strong validation of the anti-T-ABM specificity in these sera. We have no evidence that antisera produced by rabbits or sheep distinguish rat T-cell ABM from murine T-cell ABM. We do have preliminary evidence which demonstrates that sheep anti-SRBC-TsF and rabbit anti-mm-me TsF do bind some different epitopes on the same molecule. Moreover, while rabbit antisera demonstrate cell surface related molecules by immunofluorescence, the sheep antisera do not react with intact cells. Taken together, the results suggest a constant region of T-ABM, and a portion of this constant region is conserved in evolution. This is similar to the observation that rabbit anti-mouse or anti-rat immunoglobulin reagents distinguish poorly between rat and mouse immunoglobulin. In addition, as anti-immunoglobulin isotype reagents bind B cells but not T cells (7), heterologous anti-T-ABM sera bind T cells but not B cells (15, 16). Since certain T-cell membrane molecules share antigenic determinants with soluble antigen-binding proteins derived from T cells, it is likely that the cell surface molecules represent T-cell membrane antigen recognition structures. This notion follows the same logic that was used with anti-immunoglobulin isotype reagents to demonstrate membrane immunoglobulin as the B-lymphocyte antigen recognition structure ( I-3). Some T-cell-derived antigen-binding molecules behave like immunoglobulins in certain ways. Serum-derived T-ABM copurified with immunoglobulins when standard procedures of serum fractionation are used (24). The antigen-binding sites of T- and B-ABM heavy chains are serologically similar (11, 12, 16, 23). Moreover, allotypic determinants associated with T-cell antigen recognition structures have been detected by genes closely linked to immunoglobulin C-region genes (13, 28). Given the similarities between these T-cell antigen recognition structures it seems reasonable to speak of the molecules recognized by these sera as T-cell isotypes. The present studies do not directly speak to the question of whether multiple T-cell isotypes exist. We have shown that the Lyt-1 ‘- and Lyt-2+-derived subfactors of TsF are serologically distinct (29). Human and marmoset T-cell products bearing Ig VH determinants structurally similar to molecules described herein have been reported (34). These observations, taken together with those of Owen and co-workers ( 13,28) suggest that more than one T-cell isotype exists. Moreover, some preliminary results with cloned, Ia-restricted antigen-specific helper T cells indicate that these cells do not react with these sera used in this study (J. Kaye and C. Janeway, Jr., unpublished results). Thus,

244

CONE ET AL.

functionally distinct T-cell subsets may express and/or secrete different T-cell receptor &types and different isotypes may have different functional properties. Moreover, receptor molecules structurally distinct from those reported herein have been reported (30). Thus far, no T-ABM reported herein can be shown to express antigens recognized by anti-MHC sera. Our findings suggest that MHC gene products are not an intrinsic part of such T-cell receptors. These observations must be reconciled with those of others (31) who have found T-cell produced antigen-specific material inseparable from certain MHC gene products. We suggest that MHC-associated antigen-specific T-cell factors may arise through covalent and/or other high afhnity binding of two polypeptides encoded by different genes. Evidence which supports this contention has been presented (32, 33). We would like to stress that the strategy of making antibodies against antigenspecific, T-cell-derived proteins will be useful for isolating and characterizing these molecules from cells or serum. Such antisera could also be used to monitor levels of such molecules in serum or culture supematants and thus directly assessT-cell function in a fashion analogous to the techniques employed to measure the production of immunoglobulin by B cells. These antisera could also be used for immunoprecipitation of in vitro translated products obtained when messenger RNA for T-cell receptors is used. The molecular biology and genetics of antigen-binding structures produced by T cell could be studied. ACKNOWLEDGMENTS This work was supported in part by USPHS Grants AI 16942 and PQl CA 29606 and by National Science Foundation (USA) Grant PCM 292 1779. We thank Ms. Pallas Lo, Faith Bockelman, and Maureen Wescott for excellent technical assistance and Astrid Swanson and Majorette Ainley for manuscript prep aration. We thank Dr. Dennis Spencer, Department of Surgery, Yale University, for his interest and support.

REFERENCES 1. 2. 3. 4. 5. 6. 7.

8. 9. 10. 11. 12. 13. 14.

15. 16. 17. 18.

Vitetta, E. S., and Uhr, J. W., Science 189, 964, 1975. Warner, N. L., Advan. Immunol. 19, 67, 1974. Pemis, B., Immunol. Rev. 37, 210, 1977. Ishizaka, K., and Ishizaka, T., J. Zmmunol. 99, 1187, 1967. Rowe, D. S., Hug, K., Fomi, L., and Pemis, B., J. Exp. Med. 139, 965, 1973. Vitetta, E. S., Neither, V., McWilliams, M., Phillips-Quagliana, T., Lamm, M., and Uhr, J. W., J. Exp. Med. 141, 206, 1975. Cone, R. E., Prog. Allergy 29, 182, 198 1. Cone, R. E., and Rosenstein, R. W., Mol. Immunol. 18, 67, 1981. Warr, G. W., Marton, G., Szenberg, A., and Marchalonis, J. J., Immunochemistry, 15,615, 1978. Burgois, A., Kahn-Perles, B., and Rubin, B., In “Immunoglobulin Idiotypes: UCLA Symposia on Molecular and Cellular Biology Series” (C. A. Janeway, E. E. Sercatz, and H. Wigzell, Eds), pp. 449-458. Academic Press, New York, 1981. Binz, H., and Wigzell, H., Contemp. Top. Immunobiol. 7, 113, 1981. Eichmann, IL, Advan. Immunol. 26, 195, 1978. Gwen, F. L., Riblet, R., and Taylor, B. A., J. Exp. Med. 153, 801, 198 I. Eardley, D. D., Shen, F. W., Cantor, H., and Getshon, R. K., J. Exp. Med. 150, 44, 1979. Cone, R. E., Rosenstein, R. W., Murray, J. H., Iverson, G. M., Ptak, W., and Gershon, R. K., Proc. Natl. Acad. Sci USA 78, 6411, 1981. Binz, H., and Wigzell, H., J. Exp. Med. 154, 1261, 1981. Mattingly, J., Kaplan, M. D., and Janeway, C. A., Jr., J. Exp. Med. 152, 545, 1980. Kontiainen, S., and Feldmann, M., Thymus 1, 59, 1979.

T-CELL

RECEPTOR

ISOTYPES

245

19. Binz, H., and Wigzell, H., Rand. J. Immunol. 5, 559, 1976. 20. Rosenstein, R. W., Cone, R. E., Murray, J. H., Ptak, W., and Gershon, R. K., Proc. Natl. Acad. Sci. USA 18, 5821, 1981. 2 I. Fresno, M., McVay-Boudreau, L., Mabel, G., and Cantor, H., J. Exp. Med. 15, 1260, 198 1. 22. Krawinkel, V., Cramer, M., Kindred, B., and Rajewsky, K., Eur. J. Immunol. 9, 815, 1979. 23. Cramer, M., and Krawinkel, V., In “Regulatory T Lymphocytes” (B. Pemis and H. J. Vogel, Eds.). pp. 39-45. Academic Press, New York, 1980. 24. Iverson, G. M., Eardley, D. D., Janeway, C. A., and Gershon, R. K., Proc. Natl. Acad. Sci. USA 80, 1435, 1983. 25. Axen, R., Porath, J., and Emach, S., Nature (London) 214, 1302, 1967. 26. Emerson, S. G., Reilly, P., and Cone, R. E., J. Immunogenet. 6, 87, 1979. 27. Wysocki, L. J., and Sato, V. C., Proc. Natl. Acad. Sci. USA 75, 2844, 1978. 28. Owen, F. L., Advan. Immunol. in press. 29. Cone, R. E., Rosenstein, R. W., Tite, J., Ptak, W., and Gershon, R. K., J. Immunol. 130, 2083, 1983. 30. Haskins, K., Kubo, R., White, J., Pigeon, M., Kappler, N., and Marrack, P., J. Exp. Med. 157, 1149, 1983. 3 1. Germain, R. N., and Benacerraf, B., Springer Semin. Immunopathol. 3, 93, 1980. 32. Taussig, M. J., and Holliman, A., Nature (London) 277, 308, 1979. 33. Taniguchi, M., Saito, T., Takei, I., and Tokuhisa T., J. Exp. Med. 153, 1672, 1981. 34. Marchalonis, J. J., Vasta, G. R., Hunt, J. C., Wang, A. C., and Minowada, J. Cell Immunol. 77, 161, 1983.