- Email: [email protected]
Production of monoclonal antibodies to thyroglobulin by in vitro immunization with a free synthetic peptide
0161~5890,‘87$3.00 + 0.00 PergamonJournals Ltd
Molecular fm~unology, Vol. 24, No. 10, pp. 1081-1086,1987 Printedin Great Britain
PRODUCTION OF MONOCLONAL ANTIBODIES TO THYROGLOBULIN BY IN VITRO IMMUNIZATION WITH A FREE SYNTHETIC PEPTIDE MARK
DE BOER,* FERRY A.
OSSENDORP,~ BERT J. M.
AL,*
Jo HILGERS,~ JAN
J. M. DE VIJLDER~and JOSEPH M. TAGER* *Laboratory of Biochemistry, Biotechnology Centre, University of Amsterdam, P.O. Box 201.51,1000HD Amsterdam, The Netherlands; tDivision of Tumor Biology, The Netherlands Cancer Institute (Antoni van Leeuwenhoekhuis), Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands; and IDepartment of Experimental Endocrinology, University of Amsterdam, Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands (First received 28 July 1986; accepted in revisedform23 Apt
1987)
Abstract-A synthetic peptide corresponding to amino acids l-19 of thyroglobulin was used to test the possibility of generating protein-reactive monoclonal antibodies by immunization in vitro with a synthetic peptide as antigen. Splenocytes from non-immunized Balb/c mice were cultured in serum-free medium for 3 days in the presence of thymocyte-conditioned medium and the synthetic peptide prior to fusion with SP2/0 murine myeloma cells. The synthetic peptide was used in its free form, i.e. not coupled to a protein carrier. Hybridomas secreting monoclonal antibodies reactive with the synthetic peptide were obtained after immunization in vitro with as little as 10ng/ml of the synthetic peptide. Between 50 and 70% of the primary clones obtained in different experiments produced monoclonal antibodies also reactive with the intact protein. Six stable hybridomas were isolated; all produced antibodies of the IgM class. We conclude that immunization in vitro with a free synthetic peptide is an efficient method for the generation of
monoclonal antibodies reactive with the intact protein.
INTRODUCTION
Advances in protein chemistry and recombinant DNA technology are making available an increasing number of proteins with known amino acid sequences. It has been suggested that analysis of the amino acid sequence for hydrophilicity allows one to predict potential antigenic sites on the molecule (Hopp and Woods, 1981). Chemically synthesized peptides have proved to be useful as immunogens for the generation of monoclonal antibodies reactive with the native protein (Antoni et al., 1985; Niman et al., 1983). The use of synthetic peptides offers the possibility of producing antibodies with preselected submolecular binding specificities without extensive selection after fusion (Schmitz et al., 1983). For immunization, synthetic peptides are generally coupled to protein carriers such as keyhole limpet haem~yanin (KLH) or bovine serum albumin (BSA), since free peptides are not believed to be very immunogenic. However, the use of carrier proteins does not seem to be essential, provided that the peptide contains an epitope for T-cell recognition. Abbreviations: BSA, bovine serum albumin; DMEM, Dulbecco’s modified Eagle’s medium; ELISA, enzymelinked immunosorbent assay; KLH, keyhole limpet haemocyanin; PBS, phosphate-buffered saline; SDS: sodium dodecyl sulphate; TCM, thymocyte conditioned medium; NRS, normal rabbit serum; TMB, 3,3’,5,5’tetramethyl-~nzidine.
There are several reports on the generation of monoclonal antibodies with free synthetic peptides (Antoni et al., 1985; Schmitz et al., 1983; Young et al., 1983). In order to obtain a simple and efficient procedure for the production of monoclonal antibodies against synthetic peptides, we have used a serum-free immunization procedure in vitro (Ossendorp et al., 1986). The use of immunization in vitro has several advantages: (a) it requires a very short immunization period (3-5 days); (b) little antigen is needed (O.Ol-1Opg); and (c) the technique offers the possibility of producing monoclonal antibodies against antigens that are only weakly immunogenic in viva (Pardue et al., 1983) or even against “self” determinants (Miner et al., 1981; Borrebaeck and Moller, 1986; de Boer et al., 1986). A 19 residue synthetic peptide, corresponding to the sequence 1-19 of thyroglobulin (Rawitch et al., 1983; Mercken et al., 1985), was used as antigen. This peptide contains a tyrosine residue which is involved in the synthesis of thyroxine (T4) in thyroglobulin (Dunn et al., 1981; Marriq et al., 1984). The hormonogenic segment derived from thyroglobulins of a number of different species are identical (Rawitch et al.? 1984), so that the amino acid sequence of the peptide used is highly conserved. Here, we show that using a synthetic peptide for immunization in vitro, we were able to generate monoclonal antibodies against a highly conserved part of thyroglobulin.
1081
MARK
1082
DE &IER
MATERIALS AND METHODS
Muterids
The peptide corr~~nding to amino acids l-19 of thyroglobulin (Nab-Asn-Ile-Phe-Glu-Tyr-Glu-ValAsp-Ala-Glu-Pro-Leu-Arg-Pro-Cys-Glu-Leu-GluArg-COOH) used in this study was synthesized by Bachem Fine Chemicals (Torrance, CA). Dulbecco’s modified Eagle’s medium (DMEM), penicillin and streptom}~cin were obtained from Gibco Biocult (Glasgow, U.K.). bovine thyroglobuli~, porcine thyroglobulin and 3,3’, 5,5’-tetramethylbenzidine (TMB) from Sigma (St Louis, MD), polyethylene glycol (mol. wt 1500) from Merck (Darmstadt, F.R.G.), and foetal calf serum from Sera-lab (Sussex, U.K.)
The serum-free medium used for the immunization was that previously described by Yssel et al. (1984), except that 50,uM 2-mercaptoethanol was added. Murine myeloma cells and hybrid cells were cultured in DMEM supplemented with streptomycin (l~~g~rnl~, penicillin (100 U/ml) and IO or 20% heat-inactivated foetal calf serum. Thymocyte-conditioned medium (TCM) was prepared as follows. Thymus glands were removed from IO- to 14-day-old Balb/c mice and passed through a sterile 50-mesh stainless steel screen. Thymus cells (5 x IO6cells/ml) were cultured in serum-free medium containing 50 PM 2mercaptoethanol. The cells were cultured for 48 hr at 37°C in 7.5% CC&. The TCM was then collected by centrifugation (7 min, 8OOg) to remove the cells and was stored at -70°C. in t$tro
immunization
in vitro
The immunization in vitro was performed as described by Ossendorp ef al. (1986). Briefly, spleen cells (I x IO’ cells/ml) from non-immunized Balb/c mice were cultured in serum-free medium supplemented with 50% (viv) TCM, 50 PM 2-mercaptoethanol and sterile filtered antigen at the desired concn (0.01-10 kg/ml). After 3 days of culture, the cells were sedimented (7 min, 800~) and used for fusion with murine myeloma cells. Immunization
in vivo
Balb/c mice were immuni~ i.p. with 20 pg of the synthetic peptide in an emulsion consisting of equal vols of the peptide in phosphate-buffered saline (PBS) and Freund’s complete adjuvant. On days 7 and 14 the mice received 2Opg of the peptide in Freund’s incomplete adjuvant and after a rest period of 3-6 weeks they received one final injection of 20 pg of the peptide in PBS. Three days later the mice were killed and the splenocytes were used for fusion. Cell fusion
and cloning
Splenocytes and SP2/0 murine myeloma cells (Shulman et al., 1978) were fused with 38% poly-
et al.
ethylene glycol (mol. wt 1500) in DMEM at a ratio of 5: 1, according to the procedure described by ffilkens et ul. (1981). The fused cells were then distributed between the wells of 96well tissue culture plates (Falcon) containing a feeder layer of nonimmune splenocytes (2 x 10’ cells/well). After l&l4 days the supernatants of the hybridomas were screened for specific antibody production (see below). Cells producing specific antibodies were cloned 3 times by limiting dilution under the same feeder conditions as those used after fusion, Screening
procedure
Specific antibody production by the hybrid cells was detected with an ELISA method. The wells of 96well polyvinyl microtitre plates (Falcon) were coated with 0.2 FCC& of synthetic peptide in 50 ~1 PBS, pH 7.4, by drying overnight at 37”‘C. The peptide bound well to plates of this type but poorly to some others. In addition to the synthetic peptide, thyroglobulins from several species were used for screening: human, porcine, bovine, and murine thyroglobul~ns. In these cases, 0.2pg of thyroglobulin in 50 ~1 PBS, pH 7.4, was added to each well and the wells were dried overnight at 37°C. Unbound synthetic peptide or thyroglobulin was removed by washing 5 times with PBSTween (0.1% v/v Tween20). Non-specific sites were blocked by incubation of the plates with I % (v/v) normal rabbit serum (NRS) in PBS for 30 min at 37°C. The wells were washed 5 times with PBS-Tween, 50~1 of hybridoma supernatant was added to each well, and the plates were incubated for 45 min at 37°C. After five washes with PBS-Tween, the wells were incubated for 45 min at 37°C with 50,ul of a I:1000 dilution of rabbit anti-(mouse total Ig) immunoglobulins conjugated to horseradish peroxidase (Dako-Immunoglobulins A.S., Copenhagen, Denmark) in PBS-Tween containing 1% NRS. Unbound peroxidase activity was removed by washing with PBSTween (5 times). Bound peroxidase was revealed by addition of an assay mixture prepared by diluting 100 ~1 of a solution of 6 mgjml TMB in dimethyisulphoxide to 10 ml with 0.1 M sodium phosphate buffer (pH 6.0) and adding 0.03% (v/v) H,O1. The reaction was stopped after 30 min of incubation in the dark by the addition of 50 1.11of 2 M H,SO,. The absorbance was determined s~trophotomet~~Ily at 4.50 nm with a Flow Titertek Multiscan. RESULTS
First we studied the effect of the peptide concn during immuni~tio~ in txlro. Non-immunized splenocytes were cultured in the presence of thym~yteconditioned medium and different peptide concns. After 3 days of culture, the remaining living cells were fused with SP2/0 murine myeloma cells. Table I shows that a somewhat higher number of positive wells was obtained when either 0.1 or 1.O p g/ml of the
1083
Production of monoclonal antibodies to thyroglobulin Table 1. Effect of antigen cone” on immunization in I&RI for the production of specific monoclonal antibodies Concn of peptide in culture medium (fig/ml)
No. of wells containing hybrids” Expt. 1
0 0.01 0.1 1.0 10.0
No. of wells containing anti-(Tg peptide) activity
160 167 143 178 136
Expt. 2.
No. of wells containing anti(Tg peptide) and anti-(thyroglobulin) activity
Expt. 1
Expt. 2
Expt. 1
Expt. 2
0 5 6 7 5
0 I 10 9 6
0 1 4 4 2
0 4 I 6 4
108 131 171 150 144
OOut of a total of 192 wells. Immunization in vifro with different concns of the synthetic peptide was performed by culturing 20 x lo6 splenocytes from non-immunized Balb/c mice as described in Materials and Methods. The results of two different experiments are shown.
Table 2. Effect of the presence of antigen and thymocyte-conditioned medium on immunization in vitro for the production of specific monoclonal antibodies
Additions to splenocyte culture medium None Thymocyte-conditioned medium Tg peptide Thymocyte-conditioned medium + Tg peptide
No. of wells containing hybrids (total 144)
No. of wells containing anti-(Tg peptide) activity
No. of wells containing anti-(Tg peptide) and antiactivity
89
1
0
120 104
2 1
1 1
133
7
3
Immunization with or without the synthetic peptide and with or without thymocyte-conditioned medium was performed by culturing 15 x IO6splenocytes from a non-immunized Balb/c mouse as described in Materials and Methods.
peptide was present during the immunization period than when other concns were used. No correlation was found between the antigen concn and the number of wells with both anti-(Tg peptide) and anti(thyroglobulin) activity. About 60% of the hybridomas secreted monoclonal antibodies also reactive with the intact protein. In order to test further whether the anti-(Tg peptide) monoclonal antibodies obtained after immunization in vitro with the synthetic peptide were the result of a specific immune response, splenocytes from non-immunized mice were cultured with or without the synthetic peptide and with or without thymocyte-conditioned medium for a 3 day period. As shown in Table 2, a significant number of positive wells was obtained only when both the synthetic peptide and thymocyte-conditioned medium were present during the culture period. After a 3 day culture period without the peptide and/or thymocyteconditioned medium only a few positive wells were found. The exact specificities of these antibodies is not known, but about 50% of them gave a positive reaction when tested on BSA in the ELISA system, so that they probably represent a background reaction. These results indicate that the generation of antibodies specific for the Tg peptide is dependent both on the antigen and on thymus cell-derived factors which are present in the thymocyteconditioned medium. A number of the primary wells positive for anti-(Tg
peptide activity obtained from these immunization experiments in vitro were subcloned and six stable clones were obtained; all produced antibodies of the IgM class. The binding characteristics in an ELISA of three monoclonal antibodies (C12, Al 1 and D4) obtained after immunization in vitro with the synthetic peptide and one (3D12; see Ossendorp et al., 1986) obtained after immunization in vitro with intact human thyroglobulin are shown in Fig. 1. The monoclonal antibody 3D12 did not react with the synthetic peptide. The monoclonal antibody D4 reacted only with the peptide and the monoclonal antibodies Al 1 and Cl2 (Fig. 1) and F4, B8 and G4 (not shown) showed reaction with both the synthetic peptide and thyroglobulin. Monoclonal antibodies Al 1, C12, F4, B8 and G4 reacted not only with human thyroglobulin but also with thyroglobulin from other species (not shown). None of the monoclonal antibodies reacted when wells were coated with the medium component BSA. An important question is whether these anti-(Tg peptide) monoclonal antibodies also react with native thyroglobulin. We were unable to quantify the affinity of the monoclonal antibodies for native thyroglobulin by immunoprecipitation techniques. We therefore tested the ability of the monoclonal antibodies to bind to thyroglobulin in frozen sections of human thyroid tissue in which thyroglobulin is likely to retain its native conformation. One of the five antibodies, F4, which reacted with thyroglobulin in
1084
MARK
DE BOER et al.
dibtii
of antibody
Fig. 1. The reactivity of three anti-(peptide) monoclonal antibodies (D4, Cl2 and Al 1) and one anti-(thyro~obulin) monoclonal antibody (3D12), tested in a ELISA system. (A) D4; (3) C12; (C) Al I; (D) 3D12. Starting concns in the dilution curves are: D4, 2.2 x lO-2~M; C12, 2.8 x IO-*pM; All, 1.6 x 10m2pM and 3D12,3.1 x 10s2 PM. Microtitre plates were coated with synthetic peptide (m---H), human thyroglobulin (a---0) or BSA (A---&. The ELISA was carried out as described in Materials and Methods.
an ELBA, gave a strong positive reaction; this was restricted to the colloid in the follicles, which contains densely packed thyroglobulin (not shown). Monoclonal antibody G4 gave a positive reaction within the follicular cells but not in the colloid; presumably G4 reacts with breakdown products of thyroglobulin formed after uptake of thyroglobulin by the follicular cells and hydrolysis of the protein to release thyroid hormones. The other antibodies gave no reaction at all. All the primary clones tested so far secreted antibodies of the IgM class, which suggests that the antibodies obtained after immunization in vitro with the synthetic peptide are the result of a primary immune response to the antigen in question (see de Boer et al., 1987). To test whether it was possible to obtain anti-(Tg peptide) monoclonal antibodies after immunization
in vivo with the synthetic peptide in its free form, mice were repeatedly injected with the synthetic peptide. As a control, splenocytes from a non-immunized mouse were used for fusion. Table 3 shows that in two different fusion experiments only five wells positive for the Tg peptide were found and three of them also showed a reaction when tested with BSA. Furthermore, comparable results were obtained after fusion of lymphocytes from the control mouse. We therefore conclude that this 19 amino acid synthetic peptide is not immunogenic in vivo, at least not when used in its free form. DIsCUSSION
In recent years there has been growing interest in the use of synthetic peptides for the production of monoclonal antibodies with preselected binding
Table 3. Immunization in eioo with a 19 amino acid synthetic peptide of thyroglobulin Lymphocytes used for fusion obtained from mouse injected with: Tg peptide in Freunds complete adjuvant Mouse 1 Mouse 2 PBS + Freund’s complete adjuvant ‘Also reactive with BSA.
No. of wells with antibody production (total 288)
No. of wells containing anti-(Tg peptide) activity
No. of wells containing antiactivity
No. of wells containing anti-(Tg peptide) and antiactivity
70 102
1 1
2
I
1” 2”
62
0
2
f”
Production
of monoclonal
specificities. For immunization, these peptides are generally linked to a protein carrier, so that extensive selective screening with peptide-carrier conjugates and carrier alone is needed to avoid cross-reactivity with the carrier protein. Our results show that the serum-free in vitro immunization procedure for the generation of hybridomas producing specific monoclonal antibodies (Ossendorp et al., 1986) can be used for the generation of protein-reactive monoclonal antibodies with a free synthetic peptide as antigen. With immunization in vitro it is theoretically possible to produce monoclonal antibodies to antigens that are only weakly immunogenic in vivo (Pardue et al., 1983). This is of interest in the case of synthetic peptides in general since they are believed not to be very immunogenic, particularly when used in their free form [see Sutcliffe et al. (1983) for a review]. Furthermore, with immunization in vitro it is possible to produce monoclonal antibodies to autologous antigens (Miner et al., 1981; Borrebaeck and Mbller, 1986). It should be pointed out that the synthetic peptide we used is to be highly conserved (Rawitch et al., 1984). Small peptides are thought to exist in different conformational states and the frequency of anti(peptide) antibodies which react with the intact protein is theoretically low (Sutcliffe et al., 1983). It has been shown that the generation of protein-reactive antibodies after immunization with a synthetic peptide occurs more often than expected from such theoretical considerations (Niman et al., 1983). Like others, we found a high percentage of proteinreactive antibodies; about 60% of the primary clones secreted antibodies which were also reactive with thyroglobulin. However, the occurrence of anti(peptide) antibodies that also react with the native protein is primarily determined by the part of the protein from which the sequence of the synthetic peptide was derived. If this part is located inside the folded protein it is unlikely to generate proteinreactive antibodies. However, it is known that in many proteins, N-terminal and C-terminal peptides are able to elicit protein-reactive antibodies; the synthetic peptide we used is derived from the N-terminus. The binding characteristics of the monoclonal antibodies obtained suggest that even on this short fragment of thyroglobulin more than one antigenic determinant exists. The fact that some of the antibodies did not react with thyroglobulin might be explained by the presence of a determinant on the 19 amino acid synthetic peptide which is absent or masked in the intact iodinated protein. When synthetic peptides are coupled to a carrier protein to enhance the immunogenicity, an epitope for T-cell recognition is usually present. However, there are several reports on the generation of monoclonal antibodies when free synthetic peptides are used as antigens for immunization in vivo (Antoni et al., 1985; Schmitz et al., 1983; Young et al., 1983). An
antibodies
to thyroglobulin
1085
important question is whether the anti-(peptide) antibodies obtained after immunization in vitro were generated by a T-cell dependent or T-cell independent mechanism. As shown in Table 2, a significant number of positive wells was only obtained when both the synthetic peptide and thymocyte-conditioned medium were present during the 3 day culture period. The thymocyte-conditioned medium contains thymus cell-derived factors. The dependency on thymocyte conditioned medium thus suggests that the monoclonal antibodies were generated by a T-cell dependent mechanism. The fact that all the anti-(Tg peptide) antibodies are of the IgM class is consistent with the protocol used for immunization in vitro. We have shown previously that antibodies obtained after using this immunization protocol in vitro are the result of a primary immune response to the antigen in question rather than stimulation by T-cell independent mechanisms, and that the antibodies are predominantly of the IgM class (de Boer et al., 1987). Since immunization in vitro yielded only antibodies of the IgM class we tried to produce IgG secreting anti-(Tg peptide) hybridomas by immunization in vivo with the synthetic peptide. However, we found that the synthetic peptide was not immunogenic in vivo, at least not when used in its free form. This might be explained by the nature of the synthetic peptide we used: it is known that the amino terminus of thyroglobulin is highly conserved (Rawitch et al., 1984). Thus, the lack of an immune response to this synthetic peptide in vivo might be due to tolerance mechanisms for self-antigens. Apparently these mechanisms are absent during immunization in vitro. Like others (Borrebaeck et al., 1986) we find it possible to generate monoclonal antibodies to autologous antigens using immunization in vitro (de Boer et al., 1986). Another explanation for the lack of an immune response to the free synthetic peptide in vivo might be the absence of a T-cell epitope on the synthetic peptide, which is needed for a T-cell dependent immune response in vivo. It is known that the factors present in the thymocyte-conditioned medium can substitute for the presence of T-cells during immunization in vitro (Borrebeack et al., 1986). This would explain why it was possible to obtain anti-(Tg peptide) monoclonal antibodies using the synthetic peptide in its free form for immunization in vitro. These two possibilities are at present being investigated.
Acknowledgements-The authors are grateful to John Hilkens, Peter Bruning and Ronald Oude Elferink and for helpful comments, Professor R. D. Hess (UniversitSt Hanover) for his help in making the thyroglobulin synthetic peptide available to us and Wendy van Noppen for help in the preparation of the manuscript. This work was supported by a grant from the Netherlands Cancer Foundation (Queen Wilhelmina Fund for Cancer Research, The Netherlands).
1086
MARK
DE BOER
REFERENCES
Antoni G., Mariani M., Presentini R., Lafata M., Neri P., Bracci L. and Cianfriglia M. (1985) Production and characterization of monoclonal antibodies to anti-human chorionic somatomammotropin by immunization with two free synthetic peptides. Molec. Immun. 22,1237-1241. Boer M. de, Ossendorp F. A., Bruning P. F. and Tager J. M. (1986) Absence of self-tolerance during in vitro immunization for the production of monoclonal antibodies. Sixth International Congress of Immunology, July 1986, Toronto, Canada, No. 3.15.13, p. 284. Boer M. de, Ossendorp F. A., Bruning P. F. and Tager J. M. (1987) Direct evidence for a primary immune response of murine B-lymphocytes after in oitro immunization at dissociated splenocytes. Hybridoma 6, 253-258. Borrebaeck C. A. K. and Mijller S. A. (1986) In vitro immunization. Effect of growth and differentiation factors on antigen-specific B cell activation and production of monoclonal antibodies to autologous antigens and weak immunogens. J. Immun. 136, 371&3715. Dunn J. T., Dunn A. D., Heppner D. G., Jr and Kim P. S. (1981) A discrete thyroxine-rich iodopeptide of 20,000 daltons from rabbit thyroglobulin. J. biol. Chem. 256, 942-947.
Hilkens J., Tager J. M., Buijs F., Brouwer-Kelder B., Thienen G. M. van, Tegelaers F. P. W. and Hilgers J. (198 1) Monoclonal antibodies against human acid a-glucosidase. Biochim. biophys. Acta 678, 7-l 1. Hopp T. P. and Woods K. R. (1981) Prediction of protein antigenic determinants from amino acid sequences. Proc. natn. Acad. Sci. U.S.A. 78, 38243833.
Marriq C., Lejeune P. J., Malthiery Y., Rolland M. and Lissitzkv S. (1984) Precursor-product relationship between
ef d.
Niman H. L., Houghten R. A., Walker L. E., Reisfeld R. A., Wilson I. A., Hogle J. M. and Lemer R. A. (1983) Generation of protein-reactive antibodies by short peptides is an event of high frequency: implications for the structural basis of immune recognition. Proc. natn. Acad. Sci. U.S.A. 80, 49494953.
Ossendorp F. A., Boer M. de, Al E. J. M., Hilgers J., Bruning P. F. and Tager J. M. (1986) Production of murine monoclonal antibodies against human thyroglobulin using an in vitro immunization procedure in serum-free medium. J. Zmmun. Meth. 91, 257-264. Pardue R. L., Brady R. C., Perry G. W. and Dedman J. R. (1983) Production of monoclonal antibodies against calmodulin by in vitro immunization of spleen cells. J. Cell Biol, 96, 1149-l 154. Rawitch A. B., Chernoff S. B., Litwer M. R., Rouse J. B. and Hamilton J. W. (1983) Thyroglobulin structurefunction. J. biol. Chem. 258, 2079-2082. Rawitch A. B.. Litwer M. R.. Greee J.. Turner C. D.. Rouse J. B. and Hamilton J. W.‘(l98iyThe isolation of indentical thyroxine containing amino acid sequences from bovine ovine and porcine thyroglobulins. Biochem. biophys. Res. Commun. 118, 423-429.
Schmitz H. E., Atassi H. and Atassi M. 2. (1983) Production of monoclonal antibodies with preselected submolecular binding specificities to protein antigenic sites: antibodies to sperm whale myoglobin sites. Molec. Immun. 20, 719-726. Shulman M. H., Wilde C. D. and Kijhler G. (1978) A better cell line for making hybridomas secreting specific antibodies. Nature, Land. 276, 269-270. Sutcliffe J. G., Shinnick T. M., Green N. and Lemer R. A. (1983) Antibodies that react with predetermined sites on proteins. Science 219, 66&666. Young C. R., Schmitz H. E. and Atassi M. Z. (1983) Antibodies with specificities to preselected protein regions evoked by free synthetic peptides representing protein antigenic sites or other surface locations: demonstration with myoglobin. Molec. Immun. 20, 567-570. Yssel H., Vries J. E. De, Koken M., Blitterswijk W. Van and Spits H. (1984) Serum-free medium for generation and propagation of functional human cytotox; and helper T cell clones. J. Immun. Meth. 72, 219-227.
Molecular fm~unology, Vol. 24, No. 10, pp. 1081-1086,1987 Printedin Great Britain
PRODUCTION OF MONOCLONAL ANTIBODIES TO THYROGLOBULIN BY IN VITRO IMMUNIZATION WITH A FREE SYNTHETIC PEPTIDE MARK
DE BOER,* FERRY A.
OSSENDORP,~ BERT J. M.
AL,*
Jo HILGERS,~ JAN
J. M. DE VIJLDER~and JOSEPH M. TAGER* *Laboratory of Biochemistry, Biotechnology Centre, University of Amsterdam, P.O. Box 201.51,1000HD Amsterdam, The Netherlands; tDivision of Tumor Biology, The Netherlands Cancer Institute (Antoni van Leeuwenhoekhuis), Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands; and IDepartment of Experimental Endocrinology, University of Amsterdam, Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands (First received 28 July 1986; accepted in revisedform23 Apt
1987)
Abstract-A synthetic peptide corresponding to amino acids l-19 of thyroglobulin was used to test the possibility of generating protein-reactive monoclonal antibodies by immunization in vitro with a synthetic peptide as antigen. Splenocytes from non-immunized Balb/c mice were cultured in serum-free medium for 3 days in the presence of thymocyte-conditioned medium and the synthetic peptide prior to fusion with SP2/0 murine myeloma cells. The synthetic peptide was used in its free form, i.e. not coupled to a protein carrier. Hybridomas secreting monoclonal antibodies reactive with the synthetic peptide were obtained after immunization in vitro with as little as 10ng/ml of the synthetic peptide. Between 50 and 70% of the primary clones obtained in different experiments produced monoclonal antibodies also reactive with the intact protein. Six stable hybridomas were isolated; all produced antibodies of the IgM class. We conclude that immunization in vitro with a free synthetic peptide is an efficient method for the generation of
monoclonal antibodies reactive with the intact protein.
INTRODUCTION
Advances in protein chemistry and recombinant DNA technology are making available an increasing number of proteins with known amino acid sequences. It has been suggested that analysis of the amino acid sequence for hydrophilicity allows one to predict potential antigenic sites on the molecule (Hopp and Woods, 1981). Chemically synthesized peptides have proved to be useful as immunogens for the generation of monoclonal antibodies reactive with the native protein (Antoni et al., 1985; Niman et al., 1983). The use of synthetic peptides offers the possibility of producing antibodies with preselected submolecular binding specificities without extensive selection after fusion (Schmitz et al., 1983). For immunization, synthetic peptides are generally coupled to protein carriers such as keyhole limpet haem~yanin (KLH) or bovine serum albumin (BSA), since free peptides are not believed to be very immunogenic. However, the use of carrier proteins does not seem to be essential, provided that the peptide contains an epitope for T-cell recognition. Abbreviations: BSA, bovine serum albumin; DMEM, Dulbecco’s modified Eagle’s medium; ELISA, enzymelinked immunosorbent assay; KLH, keyhole limpet haemocyanin; PBS, phosphate-buffered saline; SDS: sodium dodecyl sulphate; TCM, thymocyte conditioned medium; NRS, normal rabbit serum; TMB, 3,3’,5,5’tetramethyl-~nzidine.
There are several reports on the generation of monoclonal antibodies with free synthetic peptides (Antoni et al., 1985; Schmitz et al., 1983; Young et al., 1983). In order to obtain a simple and efficient procedure for the production of monoclonal antibodies against synthetic peptides, we have used a serum-free immunization procedure in vitro (Ossendorp et al., 1986). The use of immunization in vitro has several advantages: (a) it requires a very short immunization period (3-5 days); (b) little antigen is needed (O.Ol-1Opg); and (c) the technique offers the possibility of producing monoclonal antibodies against antigens that are only weakly immunogenic in viva (Pardue et al., 1983) or even against “self” determinants (Miner et al., 1981; Borrebaeck and Moller, 1986; de Boer et al., 1986). A 19 residue synthetic peptide, corresponding to the sequence 1-19 of thyroglobulin (Rawitch et al., 1983; Mercken et al., 1985), was used as antigen. This peptide contains a tyrosine residue which is involved in the synthesis of thyroxine (T4) in thyroglobulin (Dunn et al., 1981; Marriq et al., 1984). The hormonogenic segment derived from thyroglobulins of a number of different species are identical (Rawitch et al.? 1984), so that the amino acid sequence of the peptide used is highly conserved. Here, we show that using a synthetic peptide for immunization in vitro, we were able to generate monoclonal antibodies against a highly conserved part of thyroglobulin.
1081
MARK
1082
DE &IER
MATERIALS AND METHODS
Muterids
The peptide corr~~nding to amino acids l-19 of thyroglobulin (Nab-Asn-Ile-Phe-Glu-Tyr-Glu-ValAsp-Ala-Glu-Pro-Leu-Arg-Pro-Cys-Glu-Leu-GluArg-COOH) used in this study was synthesized by Bachem Fine Chemicals (Torrance, CA). Dulbecco’s modified Eagle’s medium (DMEM), penicillin and streptom}~cin were obtained from Gibco Biocult (Glasgow, U.K.). bovine thyroglobuli~, porcine thyroglobulin and 3,3’, 5,5’-tetramethylbenzidine (TMB) from Sigma (St Louis, MD), polyethylene glycol (mol. wt 1500) from Merck (Darmstadt, F.R.G.), and foetal calf serum from Sera-lab (Sussex, U.K.)
The serum-free medium used for the immunization was that previously described by Yssel et al. (1984), except that 50,uM 2-mercaptoethanol was added. Murine myeloma cells and hybrid cells were cultured in DMEM supplemented with streptomycin (l~~g~rnl~, penicillin (100 U/ml) and IO or 20% heat-inactivated foetal calf serum. Thymocyte-conditioned medium (TCM) was prepared as follows. Thymus glands were removed from IO- to 14-day-old Balb/c mice and passed through a sterile 50-mesh stainless steel screen. Thymus cells (5 x IO6cells/ml) were cultured in serum-free medium containing 50 PM 2mercaptoethanol. The cells were cultured for 48 hr at 37°C in 7.5% CC&. The TCM was then collected by centrifugation (7 min, 8OOg) to remove the cells and was stored at -70°C. in t$tro
immunization
in vitro
The immunization in vitro was performed as described by Ossendorp ef al. (1986). Briefly, spleen cells (I x IO’ cells/ml) from non-immunized Balb/c mice were cultured in serum-free medium supplemented with 50% (viv) TCM, 50 PM 2-mercaptoethanol and sterile filtered antigen at the desired concn (0.01-10 kg/ml). After 3 days of culture, the cells were sedimented (7 min, 800~) and used for fusion with murine myeloma cells. Immunization
in vivo
Balb/c mice were immuni~ i.p. with 20 pg of the synthetic peptide in an emulsion consisting of equal vols of the peptide in phosphate-buffered saline (PBS) and Freund’s complete adjuvant. On days 7 and 14 the mice received 2Opg of the peptide in Freund’s incomplete adjuvant and after a rest period of 3-6 weeks they received one final injection of 20 pg of the peptide in PBS. Three days later the mice were killed and the splenocytes were used for fusion. Cell fusion
and cloning
Splenocytes and SP2/0 murine myeloma cells (Shulman et al., 1978) were fused with 38% poly-
et al.
ethylene glycol (mol. wt 1500) in DMEM at a ratio of 5: 1, according to the procedure described by ffilkens et ul. (1981). The fused cells were then distributed between the wells of 96well tissue culture plates (Falcon) containing a feeder layer of nonimmune splenocytes (2 x 10’ cells/well). After l&l4 days the supernatants of the hybridomas were screened for specific antibody production (see below). Cells producing specific antibodies were cloned 3 times by limiting dilution under the same feeder conditions as those used after fusion, Screening
procedure
Specific antibody production by the hybrid cells was detected with an ELISA method. The wells of 96well polyvinyl microtitre plates (Falcon) were coated with 0.2 FCC& of synthetic peptide in 50 ~1 PBS, pH 7.4, by drying overnight at 37”‘C. The peptide bound well to plates of this type but poorly to some others. In addition to the synthetic peptide, thyroglobulins from several species were used for screening: human, porcine, bovine, and murine thyroglobul~ns. In these cases, 0.2pg of thyroglobulin in 50 ~1 PBS, pH 7.4, was added to each well and the wells were dried overnight at 37°C. Unbound synthetic peptide or thyroglobulin was removed by washing 5 times with PBSTween (0.1% v/v Tween20). Non-specific sites were blocked by incubation of the plates with I % (v/v) normal rabbit serum (NRS) in PBS for 30 min at 37°C. The wells were washed 5 times with PBS-Tween, 50~1 of hybridoma supernatant was added to each well, and the plates were incubated for 45 min at 37°C. After five washes with PBS-Tween, the wells were incubated for 45 min at 37°C with 50,ul of a I:1000 dilution of rabbit anti-(mouse total Ig) immunoglobulins conjugated to horseradish peroxidase (Dako-Immunoglobulins A.S., Copenhagen, Denmark) in PBS-Tween containing 1% NRS. Unbound peroxidase activity was removed by washing with PBSTween (5 times). Bound peroxidase was revealed by addition of an assay mixture prepared by diluting 100 ~1 of a solution of 6 mgjml TMB in dimethyisulphoxide to 10 ml with 0.1 M sodium phosphate buffer (pH 6.0) and adding 0.03% (v/v) H,O1. The reaction was stopped after 30 min of incubation in the dark by the addition of 50 1.11of 2 M H,SO,. The absorbance was determined s~trophotomet~~Ily at 4.50 nm with a Flow Titertek Multiscan. RESULTS
First we studied the effect of the peptide concn during immuni~tio~ in txlro. Non-immunized splenocytes were cultured in the presence of thym~yteconditioned medium and different peptide concns. After 3 days of culture, the remaining living cells were fused with SP2/0 murine myeloma cells. Table I shows that a somewhat higher number of positive wells was obtained when either 0.1 or 1.O p g/ml of the
1083
Production of monoclonal antibodies to thyroglobulin Table 1. Effect of antigen cone” on immunization in I&RI for the production of specific monoclonal antibodies Concn of peptide in culture medium (fig/ml)
No. of wells containing hybrids” Expt. 1
0 0.01 0.1 1.0 10.0
No. of wells containing anti-(Tg peptide) activity
160 167 143 178 136
Expt. 2.
No. of wells containing anti(Tg peptide) and anti-(thyroglobulin) activity
Expt. 1
Expt. 2
Expt. 1
Expt. 2
0 5 6 7 5
0 I 10 9 6
0 1 4 4 2
0 4 I 6 4
108 131 171 150 144
OOut of a total of 192 wells. Immunization in vifro with different concns of the synthetic peptide was performed by culturing 20 x lo6 splenocytes from non-immunized Balb/c mice as described in Materials and Methods. The results of two different experiments are shown.
Table 2. Effect of the presence of antigen and thymocyte-conditioned medium on immunization in vitro for the production of specific monoclonal antibodies
Additions to splenocyte culture medium None Thymocyte-conditioned medium Tg peptide Thymocyte-conditioned medium + Tg peptide
No. of wells containing hybrids (total 144)
No. of wells containing anti-(Tg peptide) activity
No. of wells containing anti-(Tg peptide) and antiactivity
89
1
0
120 104
2 1
1 1
133
7
3
Immunization with or without the synthetic peptide and with or without thymocyte-conditioned medium was performed by culturing 15 x IO6splenocytes from a non-immunized Balb/c mouse as described in Materials and Methods.
peptide was present during the immunization period than when other concns were used. No correlation was found between the antigen concn and the number of wells with both anti-(Tg peptide) and anti(thyroglobulin) activity. About 60% of the hybridomas secreted monoclonal antibodies also reactive with the intact protein. In order to test further whether the anti-(Tg peptide) monoclonal antibodies obtained after immunization in vitro with the synthetic peptide were the result of a specific immune response, splenocytes from non-immunized mice were cultured with or without the synthetic peptide and with or without thymocyte-conditioned medium for a 3 day period. As shown in Table 2, a significant number of positive wells was obtained only when both the synthetic peptide and thymocyte-conditioned medium were present during the culture period. After a 3 day culture period without the peptide and/or thymocyteconditioned medium only a few positive wells were found. The exact specificities of these antibodies is not known, but about 50% of them gave a positive reaction when tested on BSA in the ELISA system, so that they probably represent a background reaction. These results indicate that the generation of antibodies specific for the Tg peptide is dependent both on the antigen and on thymus cell-derived factors which are present in the thymocyteconditioned medium. A number of the primary wells positive for anti-(Tg
peptide activity obtained from these immunization experiments in vitro were subcloned and six stable clones were obtained; all produced antibodies of the IgM class. The binding characteristics in an ELISA of three monoclonal antibodies (C12, Al 1 and D4) obtained after immunization in vitro with the synthetic peptide and one (3D12; see Ossendorp et al., 1986) obtained after immunization in vitro with intact human thyroglobulin are shown in Fig. 1. The monoclonal antibody 3D12 did not react with the synthetic peptide. The monoclonal antibody D4 reacted only with the peptide and the monoclonal antibodies Al 1 and Cl2 (Fig. 1) and F4, B8 and G4 (not shown) showed reaction with both the synthetic peptide and thyroglobulin. Monoclonal antibodies Al 1, C12, F4, B8 and G4 reacted not only with human thyroglobulin but also with thyroglobulin from other species (not shown). None of the monoclonal antibodies reacted when wells were coated with the medium component BSA. An important question is whether these anti-(Tg peptide) monoclonal antibodies also react with native thyroglobulin. We were unable to quantify the affinity of the monoclonal antibodies for native thyroglobulin by immunoprecipitation techniques. We therefore tested the ability of the monoclonal antibodies to bind to thyroglobulin in frozen sections of human thyroid tissue in which thyroglobulin is likely to retain its native conformation. One of the five antibodies, F4, which reacted with thyroglobulin in
1084
MARK
DE BOER et al.
dibtii
of antibody
Fig. 1. The reactivity of three anti-(peptide) monoclonal antibodies (D4, Cl2 and Al 1) and one anti-(thyro~obulin) monoclonal antibody (3D12), tested in a ELISA system. (A) D4; (3) C12; (C) Al I; (D) 3D12. Starting concns in the dilution curves are: D4, 2.2 x lO-2~M; C12, 2.8 x IO-*pM; All, 1.6 x 10m2pM and 3D12,3.1 x 10s2 PM. Microtitre plates were coated with synthetic peptide (m---H), human thyroglobulin (a---0) or BSA (A---&. The ELISA was carried out as described in Materials and Methods.
an ELBA, gave a strong positive reaction; this was restricted to the colloid in the follicles, which contains densely packed thyroglobulin (not shown). Monoclonal antibody G4 gave a positive reaction within the follicular cells but not in the colloid; presumably G4 reacts with breakdown products of thyroglobulin formed after uptake of thyroglobulin by the follicular cells and hydrolysis of the protein to release thyroid hormones. The other antibodies gave no reaction at all. All the primary clones tested so far secreted antibodies of the IgM class, which suggests that the antibodies obtained after immunization in vitro with the synthetic peptide are the result of a primary immune response to the antigen in question (see de Boer et al., 1987). To test whether it was possible to obtain anti-(Tg peptide) monoclonal antibodies after immunization
in vivo with the synthetic peptide in its free form, mice were repeatedly injected with the synthetic peptide. As a control, splenocytes from a non-immunized mouse were used for fusion. Table 3 shows that in two different fusion experiments only five wells positive for the Tg peptide were found and three of them also showed a reaction when tested with BSA. Furthermore, comparable results were obtained after fusion of lymphocytes from the control mouse. We therefore conclude that this 19 amino acid synthetic peptide is not immunogenic in vivo, at least not when used in its free form. DIsCUSSION
In recent years there has been growing interest in the use of synthetic peptides for the production of monoclonal antibodies with preselected binding
Table 3. Immunization in eioo with a 19 amino acid synthetic peptide of thyroglobulin Lymphocytes used for fusion obtained from mouse injected with: Tg peptide in Freunds complete adjuvant Mouse 1 Mouse 2 PBS + Freund’s complete adjuvant ‘Also reactive with BSA.
No. of wells with antibody production (total 288)
No. of wells containing anti-(Tg peptide) activity
No. of wells containing antiactivity
No. of wells containing anti-(Tg peptide) and antiactivity
70 102
1 1
2
I
1” 2”
62
0
2
f”
Production
of monoclonal
specificities. For immunization, these peptides are generally linked to a protein carrier, so that extensive selective screening with peptide-carrier conjugates and carrier alone is needed to avoid cross-reactivity with the carrier protein. Our results show that the serum-free in vitro immunization procedure for the generation of hybridomas producing specific monoclonal antibodies (Ossendorp et al., 1986) can be used for the generation of protein-reactive monoclonal antibodies with a free synthetic peptide as antigen. With immunization in vitro it is theoretically possible to produce monoclonal antibodies to antigens that are only weakly immunogenic in vivo (Pardue et al., 1983). This is of interest in the case of synthetic peptides in general since they are believed not to be very immunogenic, particularly when used in their free form [see Sutcliffe et al. (1983) for a review]. Furthermore, with immunization in vitro it is possible to produce monoclonal antibodies to autologous antigens (Miner et al., 1981; Borrebaeck and Mbller, 1986). It should be pointed out that the synthetic peptide we used is to be highly conserved (Rawitch et al., 1984). Small peptides are thought to exist in different conformational states and the frequency of anti(peptide) antibodies which react with the intact protein is theoretically low (Sutcliffe et al., 1983). It has been shown that the generation of protein-reactive antibodies after immunization with a synthetic peptide occurs more often than expected from such theoretical considerations (Niman et al., 1983). Like others, we found a high percentage of proteinreactive antibodies; about 60% of the primary clones secreted antibodies which were also reactive with thyroglobulin. However, the occurrence of anti(peptide) antibodies that also react with the native protein is primarily determined by the part of the protein from which the sequence of the synthetic peptide was derived. If this part is located inside the folded protein it is unlikely to generate proteinreactive antibodies. However, it is known that in many proteins, N-terminal and C-terminal peptides are able to elicit protein-reactive antibodies; the synthetic peptide we used is derived from the N-terminus. The binding characteristics of the monoclonal antibodies obtained suggest that even on this short fragment of thyroglobulin more than one antigenic determinant exists. The fact that some of the antibodies did not react with thyroglobulin might be explained by the presence of a determinant on the 19 amino acid synthetic peptide which is absent or masked in the intact iodinated protein. When synthetic peptides are coupled to a carrier protein to enhance the immunogenicity, an epitope for T-cell recognition is usually present. However, there are several reports on the generation of monoclonal antibodies when free synthetic peptides are used as antigens for immunization in vivo (Antoni et al., 1985; Schmitz et al., 1983; Young et al., 1983). An
antibodies
to thyroglobulin
1085
important question is whether the anti-(peptide) antibodies obtained after immunization in vitro were generated by a T-cell dependent or T-cell independent mechanism. As shown in Table 2, a significant number of positive wells was only obtained when both the synthetic peptide and thymocyte-conditioned medium were present during the 3 day culture period. The thymocyte-conditioned medium contains thymus cell-derived factors. The dependency on thymocyte conditioned medium thus suggests that the monoclonal antibodies were generated by a T-cell dependent mechanism. The fact that all the anti-(Tg peptide) antibodies are of the IgM class is consistent with the protocol used for immunization in vitro. We have shown previously that antibodies obtained after using this immunization protocol in vitro are the result of a primary immune response to the antigen in question rather than stimulation by T-cell independent mechanisms, and that the antibodies are predominantly of the IgM class (de Boer et al., 1987). Since immunization in vitro yielded only antibodies of the IgM class we tried to produce IgG secreting anti-(Tg peptide) hybridomas by immunization in vivo with the synthetic peptide. However, we found that the synthetic peptide was not immunogenic in vivo, at least not when used in its free form. This might be explained by the nature of the synthetic peptide we used: it is known that the amino terminus of thyroglobulin is highly conserved (Rawitch et al., 1984). Thus, the lack of an immune response to this synthetic peptide in vivo might be due to tolerance mechanisms for self-antigens. Apparently these mechanisms are absent during immunization in vitro. Like others (Borrebaeck et al., 1986) we find it possible to generate monoclonal antibodies to autologous antigens using immunization in vitro (de Boer et al., 1986). Another explanation for the lack of an immune response to the free synthetic peptide in vivo might be the absence of a T-cell epitope on the synthetic peptide, which is needed for a T-cell dependent immune response in vivo. It is known that the factors present in the thymocyte-conditioned medium can substitute for the presence of T-cells during immunization in vitro (Borrebeack et al., 1986). This would explain why it was possible to obtain anti-(Tg peptide) monoclonal antibodies using the synthetic peptide in its free form for immunization in vitro. These two possibilities are at present being investigated.
Acknowledgements-The authors are grateful to John Hilkens, Peter Bruning and Ronald Oude Elferink and for helpful comments, Professor R. D. Hess (UniversitSt Hanover) for his help in making the thyroglobulin synthetic peptide available to us and Wendy van Noppen for help in the preparation of the manuscript. This work was supported by a grant from the Netherlands Cancer Foundation (Queen Wilhelmina Fund for Cancer Research, The Netherlands).
1086
MARK
DE BOER
REFERENCES
Antoni G., Mariani M., Presentini R., Lafata M., Neri P., Bracci L. and Cianfriglia M. (1985) Production and characterization of monoclonal antibodies to anti-human chorionic somatomammotropin by immunization with two free synthetic peptides. Molec. Immun. 22,1237-1241. Boer M. de, Ossendorp F. A., Bruning P. F. and Tager J. M. (1986) Absence of self-tolerance during in vitro immunization for the production of monoclonal antibodies. Sixth International Congress of Immunology, July 1986, Toronto, Canada, No. 3.15.13, p. 284. Boer M. de, Ossendorp F. A., Bruning P. F. and Tager J. M. (1987) Direct evidence for a primary immune response of murine B-lymphocytes after in oitro immunization at dissociated splenocytes. Hybridoma 6, 253-258. Borrebaeck C. A. K. and Mijller S. A. (1986) In vitro immunization. Effect of growth and differentiation factors on antigen-specific B cell activation and production of monoclonal antibodies to autologous antigens and weak immunogens. J. Immun. 136, 371&3715. Dunn J. T., Dunn A. D., Heppner D. G., Jr and Kim P. S. (1981) A discrete thyroxine-rich iodopeptide of 20,000 daltons from rabbit thyroglobulin. J. biol. Chem. 256, 942-947.
Hilkens J., Tager J. M., Buijs F., Brouwer-Kelder B., Thienen G. M. van, Tegelaers F. P. W. and Hilgers J. (198 1) Monoclonal antibodies against human acid a-glucosidase. Biochim. biophys. Acta 678, 7-l 1. Hopp T. P. and Woods K. R. (1981) Prediction of protein antigenic determinants from amino acid sequences. Proc. natn. Acad. Sci. U.S.A. 78, 38243833.
Marriq C., Lejeune P. J., Malthiery Y., Rolland M. and Lissitzkv S. (1984) Precursor-product relationship between
ef d.
Niman H. L., Houghten R. A., Walker L. E., Reisfeld R. A., Wilson I. A., Hogle J. M. and Lemer R. A. (1983) Generation of protein-reactive antibodies by short peptides is an event of high frequency: implications for the structural basis of immune recognition. Proc. natn. Acad. Sci. U.S.A. 80, 49494953.
Ossendorp F. A., Boer M. de, Al E. J. M., Hilgers J., Bruning P. F. and Tager J. M. (1986) Production of murine monoclonal antibodies against human thyroglobulin using an in vitro immunization procedure in serum-free medium. J. Zmmun. Meth. 91, 257-264. Pardue R. L., Brady R. C., Perry G. W. and Dedman J. R. (1983) Production of monoclonal antibodies against calmodulin by in vitro immunization of spleen cells. J. Cell Biol, 96, 1149-l 154. Rawitch A. B., Chernoff S. B., Litwer M. R., Rouse J. B. and Hamilton J. W. (1983) Thyroglobulin structurefunction. J. biol. Chem. 258, 2079-2082. Rawitch A. B.. Litwer M. R.. Greee J.. Turner C. D.. Rouse J. B. and Hamilton J. W.‘(l98iyThe isolation of indentical thyroxine containing amino acid sequences from bovine ovine and porcine thyroglobulins. Biochem. biophys. Res. Commun. 118, 423-429.
Schmitz H. E., Atassi H. and Atassi M. 2. (1983) Production of monoclonal antibodies with preselected submolecular binding specificities to protein antigenic sites: antibodies to sperm whale myoglobin sites. Molec. Immun. 20, 719-726. Shulman M. H., Wilde C. D. and Kijhler G. (1978) A better cell line for making hybridomas secreting specific antibodies. Nature, Land. 276, 269-270. Sutcliffe J. G., Shinnick T. M., Green N. and Lemer R. A. (1983) Antibodies that react with predetermined sites on proteins. Science 219, 66&666. Young C. R., Schmitz H. E. and Atassi M. Z. (1983) Antibodies with specificities to preselected protein regions evoked by free synthetic peptides representing protein antigenic sites or other surface locations: demonstration with myoglobin. Molec. Immun. 20, 567-570. Yssel H., Vries J. E. De, Koken M., Blitterswijk W. Van and Spits H. (1984) Serum-free medium for generation and propagation of functional human cytotox; and helper T cell clones. J. Immun. Meth. 72, 219-227.