Messenger RNA for an antigen-specific binding molecule from an ...

Proc. Nati. Acad. Sci. USA Vol. 81, pp. 1524-1528, March 1984 Immunology

Messenger RNA for an antigen-specific binding molecule from an antigen-specific T-cell hybrid (translation/ receptors/genes/hybridoma) KENNETH D. BEAMAN*, NANCY H. RUDDLEt, ALFRED L. M. BOTHWELL*t,

AND

ROBERT E. CONE*§¶

Departments of *Pathology and §Surgery, tSchool of Epidemiology and Public Health; and tThe Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, CT 06510

Communicated by Dorothy M. Horstmann, November 14, 1983

The T-cell hybridoma 51H7D specifically ABSTRACT binds the hapten azobenzenearsonate (ABA). An antiserum (anti-PClF) specific for antigenic determinants shared by antigen-binding molecules of T cells (TABM) was used to precipitate polyribosomes containing mRNA for TABM from this Tcell bybridoma. The resultant mRNA was translated in vitro. The translated product (TrP51H7D) bound specifically ABA and was bound by both anti-PCIF and anti-T-cell suppressor factor. The parent lymphoma BW5147 yielded a similar translated product with the same antiserum used to isolate specific mRNA containing polysomes. This product (TrPBW) was bound by the antiserum used but did not bind ABA. The specific translated protein from both cells, had a pI of -5.0, and apparent molecular weight of 71,000 (reduced) or 145,000 (nonreduced). Both protein products, when treated with guanidine to break down all noncovalent bonds, revealed an elemental peptide of Mr 23,500. The cDNA made from the isolated mRNA had 600-900 bases. mRNA of this size is expected for a protein of Mr 25,000. Our data indicate that a TABM specific for ABA is composed of peptides of Mr 23,500.

describe in this report the isolation of mRNA for a haptenspecific T-cell antigen-binding molecule.

MATERIALS AND METHODS Cell Lines. The murine T-cell hybrid 51H7D, which specifically binds azobenzenearsonate (ABA), was prepared and characterized as described (11). Murine T-lymphoma BW5147 TG was originally obtained from R. Goldsby and maintained in our laboratory for several years. All T-cell lines lacked immunoglobulin by immunofluorescence and did not secrete immunoglobulin. Cells were grown in 100 ml of Dulbecco's modified Eagle's medium supplemented with 10% heat-inactivated fetal bovine serum to which 1 mCi of [3H]uridine was added (1 Ci = 37 GBq) 24 hr before harvesting. Medium and serum were obtained from GIBCO. Both cells produced cell-surface components that were bound by the antiserum used in this study as shown by immunofluorescence. Hybrid 51H7D formed rosettes with ABA sheep erythrocytes and suppressed an in vitro plaqueforming response to this antigen, whereas BW5147 neither formed rosettes nor suppressed the response (12). Heterologous Antisera. Rabbit antisera to murine antigenspecific molecules produced by T cells were prepared as described (4, 13). Briefly, rabbits were immunized with trinitrophenol-specific (T-cell derived) antigen-binding proteins purified by hapten-affinity chromatography. The resultant antisera, anti-PClF and anti-TsF, were specific for T lymphocytes and T-cell-derived antigen-binding molecules (4,

Some subsets of T lymphocytes recognize antigen specifically and have the ability to regulate the response of the immune system to antigen; other T-cell subsets have cytolytic activity directed by antigen-specific molecules (1). Unlike the antigen-binding molecules of B cells (immunoglobulin), identification of T-cell antigen-binding molecules (TABM) has proven to be more difficult. Recently, a number of significant advances have been made in identifying TABM. Antigen-specific T-cell hybridomas provide an abundant source of homogeneous cell populations for the isolation of TABM (2, 3). Heterologous antisera against affinity-purified antigen-specific T-cell products have been used to isolate several TABM (4, 5). This commonality of epitopes between TABM indicates the utility of these heterologous antisera in the isolation of such proteins (6). Other investigators have prepared monoclonal antibodies against cloned T-cell lines (7-9) and have reported the production of a monoclonal antibody against a genetically restricted T-cell antigen receptor (8). A specific heteroantiserum used in these studies, antiPC1F, binds TABM from a variety of sources (4). This antiserum was prepared against an antigen-binding protein factor (PCIF) released by murine Ly 1+, 2- T cells obtained from mice that had been sensitized by skin painting with picrylchloride, an analog of the trinitrophenol hapten (10). Anti-PClF binds to PCIF, TABM from a variety of sources (6), and to the antigen-binding component of a trinitrophenol-specific T-cell suppressor factor (4). Using both heterologous antiserum and hapten-specific T-cell hybrids, we

13). Isolation of Polysomal RNA. T-cell hybrids were harvested at a density of 1 to 5 x 105 per ml of culture in 100-ml volumes. The pelleted cells were suspended in 3 ml of lysis buffer/25 mM Tris'HCl, pH 7.5/25 mM NaCl/5 mM MgCl2/2% Triton X-100/5% sucrose (vol/vol) and lysed in a Dounce homogenizer (14). Both the lysis and subsequent precipitation buffers contained 10 mM vanadyl ribonucleoside complex (Bethesda Research Laboratories, Gaithersburg, MD). The lysate was centrifuged at 27,200 x g for 10 min. An equal volume of polysomal buffer (lysis buffer containing 100 mM MgCl2) was added to the resultant supernatant. This solution was layered over an equal volume of 200 mM sucrose/25 mM Tris-HCl/25 mM NaCl/100 mM MgCl2, pH 7.5, and centrifuged at 27,200 x g for 15 min. The pelleted polysomes were then suspended in suspension buffer (10 mM Tris*HCl/100 mM NaCl/5 mM EDTA, pH 7.4) and immunoprecipitated with 20 1.l of anti-PClF containing 25 units of heparin per ml and 100 units of human placental RNase inhibitor (Bethesda Research Laboratories) (15). Abbreviations: ABA, azobenzenearsonate; TABM, T-cell antigenbinding molecules. ITo whom reprint requests should be addressed at: Department of Pathology, Yale University School of Medicine, New Haven, CT

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

06510.

1524

Immunology: Beaman et aL The precipitated polysomes were extracted with phenol and precipitated with ethanol. The pelleted RNA was suspended in a binding buffer and passed over an oligo(dT) column (Bethesda Research Laboratories). The bound polyadenylylated RNA was eluted and ethanol precipitated. In Vitro Translation. The final ethanol precipitate was suspended in 100 /l of water and translated in a wheat germ system (Bethesda Research Laboratories) that contained human placental RNase inhibitor and [35S]methionine. The translated products were either precipitated with 10% trichloroacetic acid or dialyzed against 50 mM Tris-HCl/10 mM EDTA buffer, pH 8.0, and stored frozen at -70'C for later analysis. NaDodSO4/PAGE and Two-Dimensional Electrophoresis. The translated product was placed in NaDodSO4 sample buffer and reduced with 5% 2-mercaptoethanol and alkylated with iodoacetamide as described (4). The samples were then run on 9% polyacrylamide slab gels. 125I-labeled reduced MOPC 104E (IgM), bovine serum albumin, and ovalbumin were used as molecular weight standards. Two-dimensional gels were run as described (16). The samples were placed in IEF sample buffer, reduced, alkylated, and run on a pre-equilibrated pH 3-10 gradient tube gel at 250 V for 16-20 hr. The gels were removed and the final pH gradients were measured. These gels were placed on 9% NaDodSO4/polyacrylamide slab gels and run in the second dimension.

Immunoprecipitation and Antigen-Binding Assays. The ability of antiserum (4, 13) to bind the translated products was determined by the addition of the translated product (25,000-40,000 trichloroacetic acid-precipitable cpm of 35S) to a solution containing the antiserum to be tested. This mixture was incubated 1 hr at room temperature with intermittent shaking and was immunoprecipitated with a 10-fold excess of sheep anti-rabbit antiserum overnight at 4°C. The resultant pellet was washed 3 times and assayed. To determine the ability of the translated product to bind antigen, ABA-coupled sheep erythrocytes were made as described (17). A volume of translated product appropriate to produce 30,000-50,000 trichloroacetic acid-precipitable cpm was added to a phosphate-buffered saline solution containing -50 ,Al of pelleted ABA-sheep erythrocytes or normal sheep erythrocytes. This mixture was incubated 1 hr at room temperature and washed 4 times in phosphate-buffered saline (pH 7.4). To more clearly define whether the translated product bound to ABA hapten or to the ABA hapten-sheep erythrocyte complex, ovalbumin or ABA-ovalbumin prepared as described (17) was preincubated at various concentrations with the translated product and allowed to incubate at room temperature for 15 min. This mixture was then added to normal or ABA-sheep erythrocytes and treated as described above. cDNA Synthesis. cDNA was prepared as described to produce full-length cDNAs (18). Briefly, the complementary DNA was synthesized using the sodium pyrophosphate procedure to which radiolabeled CTP was added and the reaction was incubated for 1 hr at 37°C. The reaction was terminated by phenol extraction and the products were purified by Sephadex G-100 chromatography. The resultant cDNA was ethanol precipitated and aliquots were analyzed on an alkaline agarose gel. RESULTS To obtain poly(A) RNA for an antigen-specific T-cell product, the T-cell hybrid specific for ABA (51H7D) was used as the source of polyribosomes that were precipitated by rabbit anti-PClF, specific for TABM. Cellular RNA was labeled by incorporation of [3H]uridine and 0.23% of the total uridinelabeled poly(A) RNA was immunoprecipitated by anti-PClF as opposed to 0.0019% by normal rabbit antiserum. Immuno-

Proc. Natl. Acad. Sci. USA 81 (1984)

1525

precipitable poly(A) RNA from the parent tumor (BW5147) 0.56% of the total. The poly(A) RNA obtained as described above was translated into protein to determine whether the peptides obtained could be bound by antiserum specific for TABM. A high percentage of translated products (50-61%) from both partially purified mRNAs (TrP51H7D from 51H7D, TrPBW from BW5147) were bound by anti-PClF and anti-TsF (Table 1). When the total mRNA from both BW5147 and 51H7D was translated, 0.5% of 5S-labeled TrP51H7D cpm). To determine the molecular weight of the translated proteins, these proteins were resolved, with or without reduction, by NaDodSO4/polyacrylamide gels. As shown in Fig. 2, the major reduced product(s) had an apparent molecular weight of 71,000, while the nonreduced product had an apparent molecular weight of 145,000 (see Fig. 4). Similar profiles were obtained when TrP5lH7D- and TrPBW-translated products bound by anti-PClF were resolved (Fig. 2). When TrP51H7D was first allowed to bind to ABA-sheep erythrocytes and run on a 9% polyacrylamide gel, the pattern was the same as shown in Fig. 2. To totally reduce the product, 5% 2-mercaptoethanol was added for 1 hr before running NaDodSO4/polyacrylamide gels. Lesser concentrations and incubation times resulted in only partial reduction as observed by NaDodSO4/PAGE. Repeated freeze-thaw cycles or prolonged dialysis caused additional breakdown of the product and bands of Mr 45,000 and 28,000 became more apparent in the gel patterns. The was

Table 1. Immunoprecipitation and antigen binding of translated products Translated products bound, cpm TrPBW TrP51H7D Antiserum Anti-PC1F 12,372 ± 1782 13,318 ± 2268 Anti-TsF 14,487 ± 1848 12,156 ± 2629 Normal rabbit antiserum 1,072 ± 103 2,911 ± 540 Antigen 406 ± 98 ABA-sheep erythrocytes 20,065 ± 3210 291 ± 106 Sheep erythrocytes 2,392 ± 241 the translated of For the immunoprecipitation products with antiserum, a total of 25,197 (TrP51H7D) and 26,268 trichloroacetic acidprecipitable counts were added to each assay. For the antigen-binding assay, 34,652 (TrP51H7D) and 37,343 (TrPBW) trichloroacetic acid-precipitable counts were added. All data are expressed as the mean of 3-5 experiments (±SD).

Proc. NatL Acad Sci. USA 81

Immunology: Beaman et al.

1526 lI)

145,000

50

+,I

71,000

23,500

30 40 Gel slice no.

50

20

u

0

40

+

L

(1984)

II

10 30

= x

c-

20 T

0

ma-oc 0

m

20

10 x

0

E0. 10

lo-,

10-2

Inhibitor,

mg

u

FIG. 1. Competitive inhibition of TrP51H7D binding to ABAsheep erythrocytes by ABA-ovalbumin. *, ABA-ovalbumin; o, ovalbumin.

reduced product after 3-5 cycles of freeze-thawing totally broke down to these lower molecular weight species. However, the nonreduced product remained unchanged. To determine if the Mr 71,000 protein was composed of smaller peptide subunits, we incubated the translated product TrP51H7D in 5 or 7 M guanidine. This procedure totally disrupts all noncovalent bonds and inhibits proteolytic digestion. The treated nonreduced polypeptides were incubated at 40°C for 24 hr and dialyzed into 8 M urea/Tris HCl/EDTA buffer, pH 8.0, and then with NaDodSO4/PAGE sample buffer (16). The proteins were then run on 9% polyacrylamide tube gels and then cut and assayed (Fig. 3, Lower). As b

C

20

F

10 F I.

_

o

10

-I .

20

60

FIG. 3. Three separate treatments of the TABM mRNA translated protein TrP51H7D were run on a 9% NaDodSO4/PAGE tube gel that was sliced and assayed. (Upper) Nonreduced protein of Mr 145,000. (Middle) Protein reduced with 5% 2-mercaptoethanol, Mr 71,000. (Lower) Digestion of the reduced protein in 7 M guanidine. The resultant peptide is M, 23,500.

shown in Fig. 4, the reduced, guanidine-treated product was Mr 23,500 while the nontreated products were Mr 71,000 reduced and Mr 145,000 nonreduced. Similar results have been observed using TABM isolated from T-cell membranes (unpublished data). To determine isoelectric points of the translated proteins, the reduced polypeptides were resolved by two-dimensional electrophoresis. As shown in Fig. 4, TrP51H7D was resolved as Mr 71,000 polypeptides with a pl range of 4.8 to 5.1. Similar proteins were obtained for two-dimensional gel analysis of BW5147. Finally, the mRNA from the hybrid 51H7D was transcribed into cDNA by reverse transcriptase. The cDNA was run on a 1.4% alkaline agarose gel (Fig. 5). The mRNA tran-

d

01:..

._-

ova

pH 8

o-

7

6

L oWV

FIG. 2. NaDodSO4/PAGE of the 35S-labeled reduced protein translated from purified poly(A) RNA. The 9% gels were run and the location of iodinated standards are indicated. Lanes: a and b, translated products from BW5147, TrPBW, and the hybrid 51H7D, TrP51H7D; c and d, translated products bound by anti-PCIF (lane c, TrPBW; lane d, TrP51H7D). When TrP51H7D was allowed to bind to ABA-sheep erythrocytes, the banding patterns of the NaDodSO4/ PAGE remained the same. A, ,u heavy chain; L, light chain; ova, ovalbumin.

1I

FIG. 4. Two-dimensional gel electrophoresis of the "S-labeled reduced translated product from 51H7D, TrP51H7D. The pH ranges are indicated on the horizontal axis. The molecular weight standards relative positions are on the vertical axis. Proteins were loaded from the alkaline end of the gel. A, ,u heavy chain; L, light chain; ova, ovalbumin.

Immunology:

Proc. Natd Acad ScL USA 81 (1984)

Beaman et aL a

b

c

1000 -

450-

9

_

200-

FIG. 5. cDNA was made from TABM mRNA of the T-cell hybrid S1H7D and run on a 1.4% alkaline agarose gel. The bands in lanes a and c are standards of 450 bases. The position of other marker standards are indicated on the Fig. The band in the center is the cDNA of the TABM of the hybrid.

scribed is a homogeneous population of 600-900 bases. This size range of cDNA is consistent with a mRNA that could code for a Mr 23,000 peptide. This further substantiates the protein data that indicate that these TABM are composed of multimeric peptides.

DISCUSSION The T-cell hybrid 51H7D binds ABA-sheep erythrocytes and suppresses an in vitro response to this antigen (11, 12). In this report, we describe the immunological purification of mRNA from this hybrid, which when translated produces a protein capable of binding antigen specifically. This product (TrP51H7D) is recognized by the heterologous antisera, antiTsF and anti-PClF. These antisera were prepared against products that bind to and modulate the immune response to trinitrophenol and other antigen-specific molecules synthesized by T cells (4, 5, 6, 13). The translated protein of this mRNA (TrP51H7D), which bound to the ABA hapten, did not bind to other haptens such as trinitrophenol. Other investigators, using T-cell hybrids specific for ABA, reported the isolation of a protein that bound ABA and was found in the cytoplasm of a variety of cells, including the parent hybridoma BW5147 (19). However, our results are not to be confused with such an observation because we isolated a mRNA from BW5147, which produced a product (TrPBW) that was bound by anti-PClF and anti-TsF, had the same apparent molecular weight as TrP51H7D, but did not bind to ABA. TrP51H7D isolated by us had an apparent molecular weight of 71,000 and a pI of 4.8-5.1, in contrast to the cytoplasmic molecule (19), which had an apparent molecular weight of 62,000 and a pI of 6.9. The size of these in vitro translated TABM was the same as that reported for PClF (10) and that reported by other investigators for other TABM (6, 20). Some investigators have reported the presence of TABM and other T-cell products of Mr -46,000 and Mr -24,000 (7, 8, 9). We have noted that TrP51H7D and PC1F frequently break down to molecules of Mr 46,000 and Mr 24,000 after

1527

multiple freeze-thaw cycles or with prolonged storage at ±40C. We report here molecular biological and biochemical evidence indicating that these antigen-binding molecules are composed of Mr 23,500 peptides that may combine in a variety of multimeric forms. When the mRNA was translated, the product (TrP51H7D) had an apparent molecular weight of 145,000 (nonreduced) and 71,000 (reduced). TrP51H7D was reduced with 5% 2mercaptoethanol and digested in 5-8 M guanidine in Tris/ EDTA buffer, pH 8.0, to break all noncovalent bonds and reveal the elemental peptides. The reduced digestion product, when dialyzed against a Tris/EDTA buffer, pH 8.0, reverts to the size of the original material Mr 71,000 with an additional band of Mr 47,000 appearing and a band of Mr 23,500 remaining (unpublished results). This digestion pattern was identical to that of PCIF (10). In addition, 0.1 M HCl produced similar results; however, prolonged digestion in acid broke the peptide bonds and completely digested the protein. Initially, the breakdown of TrP51H7D and PCIF was thought to be attributed to protease digestion. However, the digestion in 7 M guanidine coupled with the data from cDNA indicate to us that TABM are actually composed of Mr 23,500 peptides, which exist in a variety of multimeric forms. In our hands, the Mr 47,000 dimer resulted when prolonged incubation occurred with the antibody. This form is rarely seen with TrP51H7D, which is the only TABM not directly harvested by antibody affinity chromatography. However, when TrP51H7D is eluted from an ABA-coupled Sepharose affinity column, the Mr 47,000 species does become evident. Finally, our data suggest that in general a family of TABM are Mr 71,000 proteins, which may form disulfide-linked dimers of Mr 145,000. However, based on the guanidine digest and the size of the cDNA from mRNA for a TABM, the fundamental size of these proteins appears to be approximately Mr 25,000. Thus, the Mr 45,000 or Mr 70,000 TABM may be dimers or trimers of the Mr 25,000 peptides. Much of the data from our laboratory and others support the likelihood that TABM may be multimers of the fundamental peptide (6, 8, 10). Our data cannot predict the exact size of the actual "biologically active" species or whether additional peptide chains may be part of the multimeric complex. These additional peptide chains may provide for the biological activity or self-recognition necessary for a "complete" T-cell receptor molecule. The TABM peptide reported here may be the functional analog of X or K chains responsible for the antigen specificity of a variety of antigen-specific T-cell products. We thank Ms. Katherine Mahoney and Ms. Donna Guralski for excellent technical assistance, and Mrs. Marjorene Ainley for patience and expertise in the preparation of the manuscript. This paper is dedicated to the memory of Richard K. Gershon, friend and colleague. This work was supported by U.S. Public Health Service Grants A116942 and PQlCA29606, National Cancer Institute Grant NCI CA16885, and National Science Foundation Grant PCM7824905. K.D.B. was supported by a National Institutes of Health training grant from the Department of Pathology (Immunology), Yale University, NIH A107019. N.H.R. was the recipient of an American Cancer Society Faculty Research Award FRA-196.

1. Cone, R. E. (1981) Prog. Allergy 29, 182-221. 2. Marrack, P., Graham, S. D., Leibson, H. J., Roehm, N., Wegmann, D. & Kappler, J. W. (1982) in Isolation, Characterization and Utilization of T Lymphocyte Clones, eds. Fattra, C. G. & Fitch, F. W., (Academic, New York), pp. 119-151. 3. Ruddle, N. H. (1982) Lymphokines 5, 49-76. 4. Cone, R. E., Rosenstein, R. W., Tite, J., Ptak, W. & Gershon, R. K. (1983) J. Immunol. 130, 2083-2087. 5. Cone, R. E., Murray, J. H., Rosenstein, R. W., Ptak, W., Iverson, G. M. & Gershon, R. K. (1981) in Immunoglobulin Idiotypes, eds. Janeway, C., Sercarz, E. & Wigzell, H., (Academic, New York), pp. 493-511.

1528

Immunology: Beaman et al.

6. Cone, R. E., Rosenstein, R. W., Janeway, C. A., Iverson, G. M., Murray, J. H., Cantor, H., Fresno, M., Mattingly, J. A., Cramer, M., Krawinkel, U., Wigzell, H., Binz, H., Frischnecht, H., Ptak, W. & Gershon, R. K. (1983) Cell Immunol. 182, 232-245. 7. Allison, J. P., McIntyre, B. W. & Block, D. (1982) J. Immunol. 129, 2293-2298. 8. Haskins, K., Kubo, R., White, J., Pigeon, M., Kappler, J. & Marrack, P. (1983) J. Exp. Med. 4, 1149-1164. 9. Reinherz, E. L., Meur, S. C. & Schlossman, S. F. (1983) Immunol. Today 4, 5-8. 10. Ptak, W., Gershon, R. K., Rosenstein, R. W., Murray, J. H. & Cone, R. E. (1983) J. Immunol. 131, 2859-2863. 11. Ruddle, N. H., Beezley, B., Lewis, G. K. & Goodman, J. W. (1980) Mol. Immunol. 17, 925-931. 12. Goodman, J. W., Lewis, G. K., Primi, D., Hornbeck, P. & Ruddle, N. H. (1980) Mol. Immunol. 17, 933-939.

Proc. NaiL Acad Sci. USA 81 (1984) 13. Cone, R. E., Rosenstein, R. W., Murray, J. H., Iverson, G. M., Ptak, W. & Gershon, R. K. (1981) Proc. Natl. Acad. Sci. USA 78, 6411-6415. 14. Palmiter, R. D., Christensen, A. K. & Schimke, R. T. (1970) J. Biol. Chem. 245, 833-841. 15. Lheureux, M., Lheureux, M., Vervoort, T., VanMeirvenne, N. & Steinert, M. (1979) Nucleic Acids Res. 7, 595-607. 16. Takeda, A., Waldron, J. A., Ruddle, N. H. & Cone, R. E. (1983) Exp. Cell Res. 148, 83-93. 17. Isakson, P. C., Honegger, J. L. & Kinksky, S. C. (1979) J. Immunol. Methods 25, 89-96. 18. Bothwell, A. L. M., Paskind, M., Beth, M., Imanishi-Kari, T., Rajewski, K. & Baltimore, D. (1981) Cell 24, 625-637. 19. Clark, A. F. & Capra, D. J. (1982) J. Exp. Med. 155, 611-615. 20. Hunt, J. C., Vasta, G. P. & Marchalonis, J. J. (1983) Cell. Immunol. 75, 22-42.