Production of Antibodies of Identical Idiotype but Diverse Immunoglobulin

Transcription

1 Proc. Nat. Acad. Sci. USA Vol. 72, No. 5, pp , May 1975 Production of Antibodies of Identical Idiotype but Diverse Immunoglobulin Classes by Cells Derived from a Single Stimulated B Cell (V gene sharing/simultaneous class production/monoclonal antibody) PATRICIA J. GEARHART, NOLAN H. SIGAL, AND NORMAN R. KLINMAN Department of Pathology, School of Medicine, University of Pennsylvania, Philadelphia, Pa Communicated by Herman N. Eisen, February 14, 1976 ABSTRACT The availability of anti-phosphocholine antibody of the TEPC 15 idiotype from the clonal progeny of a single precursor cell, stimulated in vitro, permitted the demonstration of monoclonal antibodies with as many as three immunoglobulin classes with identical variable regions. This demonstration was dependent on sensitive radioimmunoassays which showed a one to one relationship between the total anti-phosphocholine antibody produced by a clone and the sum of anti-phosphocholine antibody of the different classes as well as the amount of antibody of the TEPC 15 idiotype. The class distribution was confirmed by isoelectric focusing identification of IgM, IgA, and IgGI antibodies of the TEPC 15 idiotype produced by single clones which showed characteristic pl values for each immunoglobulin class. Thus, within the generative phase of a single antibody-producing cell clone, various heavy chain constant regions can be linked to the same variable region, and single precursor cells have the capacity to generate progeny expressing at least three distinct immunoglobulin classes. The heterogeneity of immunological responses to most antigens represents the composite antibody product of many monospecific precursor cells responsive to those antigens (1). In addition to heterogeneity of the variable domains and their respective antigen-binding sites, heterogeneity also exists in the constant region of the heavy chain, since antibodies of similar specificity are produced by individual animals in several immunoglobulin classes (2). Considerable evidence is available in support of the hypothesis that the variable and constant regions are under separate genetic control and are joined to form a heavy or light chain (3-5). Inherent in this "two gene-one polypeptide chain" hypothesis is the concept that any constant region gene may be associated with any variable region gene (6). The most compelling evidence for this is derived from the study of sera from myeloma patients, which contain monoclonal IgM and monoclonal IgG2 proteins identical in the heavy chain variable region by partial sequencing and idiotype (5). The sharing of idiotypic markers by IgM and IgG immunoglobulin has been reported in other systems (7, 8), and recently several investigators have demonstrated idiotypic cross reactions between a mouse myeloma protein of the IgA class with specificity for phosphocholine (PC) and anti-pc antibody of the IgM class from Balb/c mice (9, 10). Whether immunoglobulins containing the same variable region on different heavy chains are produced by progeny of the same precursor cell or different precursor Abbreviations: PC, phosphocholine (frequently termed phosphorylcholine); Dnp, 2,4-dinitrophenyl; B cell, bone-marrowderived antibody-forming cell precursor; PPC-TGG-Hy, 3-(pazophenylphosphocholine) N-acetyl- L- tyrosylglycylglycine-he- - mocyanin; PC-BSA-BAC, phosphocholine-bovine serum albumin-bromoacetyl cellulose cells, however, is still not certain. Definitive evidence would require the analysis of the variable region on different immunoglobulin classes produced by a single cell or a clone of cells derived from a single stimulated precursor cell. Studies by several investigators have suggested that one cell or its clonal progeny can produce both IgM and IgG antibody. Pernis et al. (11) have shown that the same rabbit allotype is associated with a cell's surface IgM and cytoplasmic IgG; however, the genetic regulation of allotype in the variable region may be distinct from that which controls the hypervariable region (12). Nossal et al. (13) have demonstrated that single antibody-forming cells can release IgM and IgG; furthermore, Press and Klinman have shown that monoclonal anti-2,4-dinitrophenyl (Dnp) antibody from splenic foci generated in vitro can contain both IgM and IgG1 (14). These observations, while indicating that one cell or a clone of cells can produce antibody of two classes with similar antigen specificity, do not necessarily establish the identity of the variable region of the two immunoglobulin classes. Since such a demonstration would provide crucial proof for the stated postulate, we have examined antibody produced by isolated clones, derived from single stimulated precursors to antibody-forming cells specific for PC and derived from bone marrow (B cells) for specificity, class, idiotype, and isoelectric spectrum. Evidence is presented that single clones can produce antibody with apparently identical variable regions but different heavy chain constant regions; therefore, regardless of the class, the idiotype of the variable region is conserved at the clonal level. MATERIALS AND METHODS Fragment Culture. The splenic focus technique of antigenically stimulating individual B cells in fragment cultures has been described previously (15). Limiting doses of spleen cells from unimmunized donor Balb/c mice were injected intravenously into lethally irradiated hemocyanin-primed syngeneic recipients. Sixteen hours later, the recipient's spleens were diced and the fragments were stimulated in vitro with 3-(p-azophenylphosphocholine) -N-acetyl-L-tyrosylglycylglycine-hemocyanin (PPC-TGG-Hy) (16). The culture supernatants collected 8-13 days after stimulation were analyzed by radioimmunoassay for the presence of anti-pc antibody as well-as the immunoglobulin class and idiotype of that antibody. Radioimmunoassays for Fab, Immunoglobulin Class, and Idiotype. Specific antibody bound to PC-bovine serum albumin-bromoacetyl cellulose (PC-BSA-BAC) immunoadsorbant (17) was quantified with rabbit 1i5I-labeled antibody to mouse

2 1708 Immunology: Gearhart et al. Proc. Nat. Acad. Sci. USA 72 (1976) 0 2 a: CD *17 i 8,,-X,12/ '16-oo 0~~~~ ~~ ~ ~ ~ ~ ~ ~ ~~ NANOGRAMS I DIOTYPE 16 is I 20 Linear relationship of multi-heavy chain class mono- FIG. 1. clonal antibody for Fab and TEPC 15 idiotype content. Samples from culture fluids or TEPC 15 ascites were tested for Fab, idiotype, IgM, IgGl, and IgA antibody as described in Materials and Methods. The standard curves for the Fab antibody and idiotype assays were derived using known amounts of purified, reduced and alkylated TIPC 15 protein. Crosses represent two dilutions of a 45% (NH4)2S04 preparation of TEPC 15 ascites. Numbered closed circles denote monoclonal anti-pc antibody containing the following heavy chain classes: 1l-1, Syl; 2-pA, y1; 3-iu, a; 4-u, yl, a; S-A, -yl, a; 6-mu, yl, a; 7-Kyl, a, 8-Aj, yl, a; 9P-g, a; l0a, Lyl, a; 11-,, a, 12-1a,yl 13-1, a; l4-, -yl, a; i-lu, a; l6-u, a; 17--, yl, a. Fab (15). Heavy chain classes were quantified by the same procedure, only employing "2'I-labeled goat antibody to mouse IgG1, IgM, and IgA as the detecting reagent (14, 17). The TEPC 15 plasma cell tumor was obtained from Michael Potter, National Cancer Institute, National Institutes of Health, Bethesda, Md., and maintained by serial passage in Balb/c mice. Protein was purified from ascites fluid by adsorption to and elution from PC-glycyltyrosine-Sepharose 4B (Pharmacia Fine Chemicals, Inc., Piscataway, N.J.) as described previously (17). The idiotypic determinant of the TEPC 15 protein, which has binding specificity for PC (18), is found on the majority of antibody to PC in the Balb/c strain (9). Anti-idiotypic serum was raised in A/He mice by multiple injections of TEPC 15 protein (19). The serum was used in a solid phase radioimmunoassay in which the culture fluid containing antibody sharing the TEPC 15 idiotype was detected by inhibition of 12"L-labeled TEPC protein binding to the anti-idiotype antibody immobilized on plastic tubes (17). Radioimmunoassays for Fab, idiotype, and a heavy chain class were standardized using purified, reduced and alkylated TEPC 15 protein. The quantity of antibody of the IgM and IgG1 class was determined by comparing the amount of antibody detected by anti-fab from foci containing solely IgM or IgG1 and the corresponding counts detected by anti-p or anti--yl. The specificity of assays for heavy chain class and idiotype has been previously shown (17). Isoelectric Focusing. Isoelectric focusing analyses of culture supernatants were performed by a micro-adaptation using a TABLE 1. Quantitation of monoclonal anti-pc antibody by Class and idiotype Nanograms antibody in sample* Heavy chain class Clone Idiotype Is 71 a PC17A PC17C PC17D PC17G PC17H * Twenty-microliter aliquots from culture fluids collected 8-10 days after in vitro stimulation were quantified for idiotype and heavy chain class as described in Materials and Methods. discontinuous sucrose density gradient containing 5% Ampholine (LKB Instruments, Rockville, Md.) as described by Press and Klinman (20). Ampholine mixtures containing 50% ph 4-6 and 50% ph 5-7 provided a linear gradient from ph 4 to ph 6.5 and were used to focus culture fluids containing solely IgM, and IgM plus IgA. IgA was focused in an Ampho". line mixture containing 66% ph 5-8, 17% ph 4-6, and 17% ph 7-9, which yielded a linear gradient from ph 5 to ph 7. An Ampholine mixture of 50% ph 6-8 and 50% ph 7-9 gave a linear gradient from ph 6 to ph 8 and was used to focus the IgG1 antibody as well as fluids containing IgG1 and IgA. Two-tenths milliliter of culture fluid was distributed among 0.25 ml volumes of 20%, 30%, 40%, and 50% solutions of sucrose containing the appropriate Ampholine; or 0.1 jul of TEPC 15 ascites fluid was added to the 30% layer. The solutions were then successively layered into a 1 ml disposable syringe closed at one end with dialysis tubing held in place by a rubber gasket. The tubes were focused for 6-8 hr at 40 at 160 V, after which approximately 100 drops were individually collected into wells of a collection tray containing 0.3 ml of 0.15 M NaCl- After the ph was read, one-half of each sample was assayed for the presence of the idiotype by inhibition in the anti-idiotype assay, and the other half was tested for anti-pc antibody by binding to PC-BSA-BAC. Bound antibody was detected with "15I-labeled anti-p, and again relabeled with iodinated anti-yl, and then with iodinated anti-a. Radioactivity from a previous labeling was subtracted from each new labeling, and increments exceeding the background adsorption- Were denoted as heavy chain positive. RESULTS Simultaneous Expression of Several Immunoglobulin Classes by Clones. Fig. l presents a comparison of the amount of anti-pc antibody on a weight basis detected by the antiidiotype assay versus that detected by analysis of Fab determinants. In this figure, all of the data points were obtained from clones producing antibody of more than one immunoglobulin class, and in every case, the experimental values fall on a theoretical line constructed for a one to one correlation. Thus, all of the anti-pc antibodies in these mixed class clones are of the TEPC 15 idiotype. Table 1 demonstrates that in these fragment cultures readily detectable amounts of anti-pc antibody of each of the immunoglobulin classes are being synthesized-and that the sum of the antibody detected for each imtnunoglobulin class approximates and does not exceed the values obtained by analysis of the idiotype.

3 Proc. Nat. Acad. Sci. USA 720 (1975) TABLE 2. Frequency of foci releasing anti-pc antibody containing more than one immunoglobulin class Heavy chain class Number of foci* A only 4 yl only 4 a only 3 A + -yl 7 (0.58)t J + a 4 (0.51)t '4 + a 5 (0.53)t + yl + a 9 (0.01)t * Total fragments analyzed = The fragments analyzed were from hemocyanin-primed recipients injected with small numbers of donor spleen cells, which yielded a low multiplicity of PC precursors when stimulated with 0.5 1AM PPC-TGG-Hy. t Number of foci predicted for the random occurrence of two (or three) simultaneous, independent events. The predicted values were obtained by multiplying the frequency of total foci containing one immunoglobulin class by the frequency of foci containing other classes. It has been previously shown that the frequency of foci producing more than one antibody specificity is close to that predicted by the random distribution of precursors (14). The results of Table 2 indicate that the frequency of clones producing anti-pc antibody of more than one immunoglobulin class is significantly greater than the number predicted by the random occurrence of more than one precursor in the same fragment. Thus, the vast majority of fragment cultures producing anti-pc antibody even of two or more immunoglobulin classes are derived from a single precursor cell. Analysis of Clonal Antibody Separated by Isoelectric Focusing. The isoelectric spectrum of each immunoglobulin class of TABLE 3. Isoelectric focusing analysis of idiotype positive monoclonal antibody to phosphocholine* Nanograms antibody in -a representative aliquot chain Clone class pi range Fabt Idiotypet PC14F16 ' a PC16E34 '4y a PC14G16 ' a PC14F50 a ;& PC14G46 a JA GD3D62 -a 7.4t ,u * Culture fluids were focused.in ph gradients covering ph 6-9 idiotype positive monoclonal antibody was determined by isoelectric focusing analyses of culture supernatants using a micromethod described by Press and Klinman (20). Since the supporting medium in this technique is sucrose, focusing of IgM and IgA as well as IgG in an Ampholine ph gradient can be accomplished. After the ph gradient was analyzed, the eluted samples were divided and one-half was assayed for anti-pc antibody and heavy chain class while the other half was assayed for idiotype content. The focusing patterns presented in Fig. 2 show that antibody containing the TEPC 15 idiotype of the IgM class has a pi range of with a peak at , IgGi antibody bearing the idiotype focuses with a pi of , and IgA antibody of the idiotype focuses in a pi range of with a peak at 6.1. The broadness in the focusing patterns of the monoclonal antibodies may reflect in vitro post-synthetic modification. In vivo modifications have been previously reported (21) and seem to be present in the TEPC 15 myeloma protein found in ascites fluid, which focuses in the same pi range as monoclonal IgA antibody but with a less distinct peak. The most direct demonstration of monoclonal production of several immunoglobulin classes with identical idiotypes is presented in Table 3. Here the immunoglobulins from fragment culture supernatants containing more than one class were separated by isoelectric focusing ph gradients. As seen in Table 3, antibody from a clone that produced any combination of IgM, IgGi, and IgA also had the idiotype region associated with each class. Since the amount of antibody detected by the anti-idiotype assay usually approximated that observed with anti-fab, each peak represents antibody with only the TEPC 15 idiotype. In addition, antibody containing the different heavy chain products focused consistently within the ph range delineated for each class in Fig. 2; thus the pi values served as invariant markers for the TEPC 15 variable region and appropriate heavy chain classes. The focusing analysis is further strengthened by preliminary observations that non-tepc 15 idiotype clonal antibodies focus at several pi values, distinct from those found for the three classes shar- for IgGi and IgA antibody and ph 4-7 for-iga and IgM. Samples from the collected gradients were assayed for immunoglobulin class, Fab, and idiotype as described in Materials and Methods. t In some instances, accurate quantitation was precluded by insufficient material for analysis. $ph determinations higher than this value were beyond the range of Ampholine used. Idiotypic Identity of Monoclonal Immunoglobulins 1709 a 1.0. CLONE GD 3D53 0;Of/ CD LZ_.bX91 IVA z C- 3.0 Z 1.0 CLONE PC 14F16 ~~~~~~~~~CLONE PC TEPC 15 ASCITES i,. I ph FIG. 2. Identification by isoelectric focusing of IgM, IgGI, and IgA proteins containing the TEPC 15 idiotype. Culture fluids containing monoclonal antibody or TEPC 15 ascites fluid were focused and the collected fractions were assayed for ph, idiotype, and heavy chain class. ing the TEPC 15 idiotype.

4 1710 Immunology: Gearhart et al. DISCUSSION A previous report from this laboratory indicated that 16% of primary Dnp-specific Balb/c B cells and 14% of secondary Drp-specific B cells yielded clonal progeny producing anti- Dnp antibody of both the u and 'y1 heavy chain classes (14). Absolute evidence for two immunoglobulins sharing the same variable heavy region in these clones was lacking, however, since the idiotype of the antibody product was not identified. The experiments reported here extend this analysis and serve as an unambiguous demonstration of the production of antibody with differentheavy chain constant regions butapparently identical variable domains by the clonalprogeny ofa single B cell responding to antigenic stimulation. The experimental design not only insured the single cell origin of the antibody-producing cell population, but also provided sufficient antibody from a single clone to permit establishing its immunoglobulin class as well as the identity of the variable regions, both by isoelectric focusing and anti-idiotype antibody analysis. Although other workers have indicated that the TEPC 15 idiotype can only be found on IgM antibody after antigenic stimulation of Balb/c mice (10, 22, 23), we have previously demonstrated that maximizing carrier recognition allows the production of other immunoglobulin classes both in vivo and in vitro (17). Much evidence is available indicating the single cell origin of antibody-forming cells in fragment cultures. This includes the linear dose-response relationship between the number of cells transferred and the resulting number of positive fragment cultures found for responses to a variety of haptens (15). The antibody produced by such fragment cultures is homogeneous by the criteria of isoelectric focusing spectrum (20) and equilibrium dialysis analysis of hapten binding characteristics before and after polypeptide chain recombination (1, 24). Furthermore, previous analyses have shown that in spite of a potentially heterogeneous response to PC in Balb/c mice, fragments producing antibody of the TEPC 15 idiotype do not produce anti-pc antibody bearing another idiotype (17). The presence of two clones in a single fragment culture has been reported to be a statistically random event (14). However, the frequency of clones producing anti-pc antibody of more than one class is significantly higher than that predicted by the random distribution of more than one precursor in a fragment, and this has also been observed in double-classproducing anti-dnp foci (14). Thus it would seem from this and the linear relationship shown between Fab and idiotype in Fig. 1 that the vast majority of fragment cultures producing anti-pc antibody even of two or more immunoglobulin classes are derived from a single precursor cell. This is most apparent in the frequency of cultures producing all three immunoglobulin classes and demonstrates that the lineage of cells producing IgM, IgG, and IgA can derive from a single immunocompetent cell. The most definitive evidence in support of monoclonal production of several immunoglobulin classes with identical idiotypes has been obtained by isoelectric focusing analysis where it is shown in Fig. 2 and Table 3 that antibody of the TEPC 15 idiotype and a given heavy chain class will always focus in the same pi range. Accordingly, the immunoglobulin populations from clones producing more than one heavy chain class were separated by isoelectric focusing and each peak was independently analyzed for idiotype and class. It is clearly demonstrated from these experiments that regardless of im- munoglobulin class, all of the anti-pc antibody from these Proc. Nat. Acad. Sci. USA 72 (1975) clones could be identified as being of the TEPC 15 idiotype by both anti-idiotype analysis and isoelectric focusing pattern. The validity of the focusing technique is confirmed by the demonstration of similarity between the pi value for TEPC 15 protein, the product of a plasmacytoma, and induced monoclonal antibody to PC. In this regard, the identity in pi as well as heavy chain class and idiotype provides strong evidence that myeloma proteins represent products of normal precursor cells (25). While analyses of myeloma proteins have demonstrated that identical variable regions can be associated with distinct immunoglobulin classes (5, 8, 9), the single cell origin and physiological relevance of these immunoglobulin-producing cells is uncertain. The studies reported here demonstrate the sharing of idiotypic specificity among immunoglobulin classes produced in response to antigenic stimulation of a single precursor cell and thus indicate that the genetic information coding for the variable heavy-chain region can be joined to any of several constant heavy chain genes within the generative phase of a single antibody cell clone. The authors thank Mses. Judy Owen, Geraldine Ball, and Alicia Scott for fine technical assistance. This investigation was supported by U.S. Public Health Service Grant A , CA , Training Grants A (P.G.), and MSTP GM (N.S.), and a Career Development Award 1-K04-A to N.K. 1. Klinman, N. R. (1971) "Purification and analysis of 'monofocal' antibody," J. Immunol. 106, Rockey, J., Klinman, N. & Karush, F. (1964) "Equine antihapten antibody. I. 7S132A- and 10S 71-globulin components of purified anti-3-lactoside antibody," J. Exp. Med. 120, Todd, C. W. (1963) "Allotype in rabbit 19S proteins," Biochem. Biophys. Res. Commun. 11, Frangione, B. & Milstein, C. (1969) "Partial deletion in the heavy chain disease protein ZUC," Nature 224, Wang, A., Wilson, S., Hopper, J., Fudenberg, H. & Nisonoff, A. (1970) "Evidence for control of synthesis of the variable regions of the heavy chains of immunoglobulins G and M by the same gene," Proc. Nat. Acad. Sci. USA 66, Hood, L. & Talmage, D. (1970) "Mechanism of antibody diversity: germ line basis for variability," Science 168, Oudin, J. & Michel, M. (1969) "Idiotypy of rabbit antibodies. II. Comparison of idiotypy of various kinds of antibodies formed in the same rabbits against Salmonella typhi," J. Exp. Med. 130, Penn, G., Kunkel, H. & Grey, H. (1970) "Sharing of individual antigenic determinants between a TG and a -ym protein in the same myeloma serum," Proc. Soc. Exp. Biol. Med. 135, Cosenza, H. & K6hler, H. (1972) "Specific inhibition of plaque formation to phosphorylcholine by antibody against antibody," Science 176) Sher, A. & Cohn, M. (1972) "Inheritance of an idiotype associated with the immune response of inbred mice to phosphorylcholine," Eur. J. Immunol. 2, Pernis, B., Forni, L. & Amante, L. (1971) "Immunoglobulins as cell receptors," Ann. N.Y. Acad. Sci. 190, Kindt, T., Klapper, D. & Waterfield, M. (1973) "An idiotypic cross-reaction between allotype a3 and allotype a negative rabbit antibodies to streptococcal carbohydrate," J. Exp. Med. 137, Nossal, G., Warner, N. & Lewis, H. (1971) "Incidence of cells simultaneously secreting IgM and IgG antibody to sheep erythrocytes," Cell. Immunol. 2, Press, J. & Klinman, N. (1973) "Monoclonal production of both IgM and IgG1 antihapten antibody," J. Exp. Med. 138,

5 Proc. Nat. Acad. Sci. USA 72 (1976) 15. Klinman, N. R. (1972) "The mechanism of antigenic stimulation of primary and secondary clonal precursor cells," J. Exp. Med. 136, Sigal, N., Gearhart, P. & Klinman, N. (1975) "The frequency of phosphorylcholine-specific B cells in conventional and germfree Balb/c mice," J. Immunol., in press. 17. Gearhart, P., Sigal, N. & Klinman, N. (1975) "Heterogeneity of the Balb/c anti-phosphorylcholine antibody response at the precursor cell level," J. Exp. Med. 141, Potter, M. & Lieberman, R. (1970) "Common individual antigenic determinants in five of eight Balb/c IgA myeloma proteins that bind phosphorylcholine," J. Exp. Med. 132, Lieberman, R. & Humphrey, W. (1971) "Association of H-2 types with genetic control of immune responsiveness to IgA allotypes in the mouse," Proc. Nat. Acad. Sci. USA 68, Press, J. & Klinman, N. (1973) "Isoelectric analysis of neo- Idiotypic Identity of Monoclonal Immunoglobulins 1711 natal monofocal antibody," Immunochemistry 10, Williamson, A., Salaman, M. & Kreth, H. (1973) "Microheterogeneity and allomorphism of proteins," Ann. N.Y. Acad. Sci. 209, Lee, W., Cosenza, H. & Kohier, H. (1974) "Clonal restriction of the immune response to phosphorylcholine," Nature 247, Claffin, J., Lieberman, R. & Davie, J. (1974) "Clonal nature of the immune response to phosphorylcholine. II. Idiotypic specificity and binding characteristics of anti-phosphorylcholine antibodies," J. Immunol. 112, Klinman, N. R. (1971) "Regain of homogeneous binding activity after recombination of chains of 'monofocal' antibody," J. Immunol. 106, Granato, D., Braun, D. & Vassalli, P. (1974) "Induction of anti-dnp antibodies: Suppressive effect of circulating anti-idiotypic antibodies to mouse myeloma protein MOPC 315," J. Immunol. 113,