Dennis W. Metzger

A monoclonal antibody (SJ-9A4) to P24 present on common alls, neuroblastomas and platelets - I. Characterization and development of a unique radioimmunometric assay.

We report the development and characterization of SJ-9A4, a monoclonal antibody (MoAb) produced against common acute lymphoblastic leukemia (C-ALL) cell lines. SJ-9A4 reacted with C-ALL, B-cell chronic lymphocytic leukemia (B-CLL), platelets and C-ALL neuroblastoma (NB) and the K562 cell lines. It had no significant reactivity with erythrocytes, granulocytes, circulating T or B lymphocytes, monocytes, granulocytic cell lines or a Ewings sarcoma cell line. SJ-9A4 was shown to recognize the same region as two other MoAb to the p24 antigen, BA-2 and DU-ALL-1, as demonstrated by their ability to inhibit the binding of labeled SJ-9A4 to NALM-1 and NB cells. Other MoAb: J5, PI 153/3 and monoclonal anti-HLA-DR antibodies gave no inhibition. A solid phase indirect radioimmunometric assay (IRA) was developed which enabled the detection of P24 from C-ALL cells, utilizing its ability to bind the Ricinus communis agglutinin (RCA1) or wheat germ agglutinin (WGA) and SJ-9A4 simultaneously. When BA-2 and DU-ALL-1 were used in place of SJ-9A4, similar IRA results were obtained. Using the RCA1/SJ-9A4-IRA, P24 from as few as 1.6 X 10(4) cells of a C-ALL cell line could be detected; however, similar extracts of NB cell lines were negative despite high levels of SJ-9A4 binding to intact cells. The presence of P24 in NB extracts was demonstrated by (1) preincubation of NB extracts with SJ-9A4 which blocked MoAb binding to P24 and (2) immunoadsorption of P24 from solubilized membranes of 35S-methionine (met) labeled NB cells. Treatment of NB cells with neuraminidase did not result in IRA binding when either RCA1 or WGA were used as the solid phase lectin indicating that the differences in lectin affinity are not due to over sialation of NB membrane glycoproteins. These findings demonstrate a difference in the glycosylation of P24 from C-ALL and NB cells.

Characterization of a monoclonal antibody reactive with rabbit T lymphocytes and neutrophils

An IgG1 monoclonal antibody, termed ACM-1, has been shown to react with rabbit T cells, but not Ig+ cells or macrophages. This antibody appears to recognize the same epitope as the previously described 9AE10 antibody and, together with 9AE10, has been used to obtain highly pure and fully functional T- and B-cell populations. However, the relevant epitope does not appear to be homologous to rodent Thy-1 since quantitative absorptions failed to show reactivity with rabbit brain. Furthermore, attempts to obtain in vivo T-cell depletion resulted in larger decreases in white blood cells than would be expected for simple T-cell removal. In vitro assays on enriched neutrophil preparations revealed that 80-95% of these cells were reactive with ACM-1 and 9AE10. Thus, it appears that in the rabbit, T cells and neutrophils share a major epitope.

Shared and Nonshared Idiotypes on Rabbit Anti‐Allotype Antibodiesa

The sharing of cross-reactive idiotype IdX markers by rabbit,’-6 mouse,’ and anti-allotype antibodies has been described by several laboratories. This report summarizes our studies of idiotypic determinants on rabbit anti-allotype antibodies and discusses the implications of our results with regard to the general nature and biological significance of IdX structures. We4 as well as others’ have previously demonstrated that in every rabbit examined, almost all antibody directed to the rabbit V, allotype, a I , bears a predominant idiotype (IdX-al), in addition to unique (IdI) determinants. The presence of IdX-a1 in unrelated animals, initially observed by radioimmunoassay, has also been visualized using anti-idiotype absorption of antibody isoelectrofocusing (IEF) patterns. In this assay, formation of idiotype-anti-idiotype complexes prior to focusing results in the elimination of IdX-positive bands.” The IEF patterns of four anti-al preparations that had been preincubated with a two-fold excess of either normal rabbit immunoglobulin as a control or with anti-idiotype antibody are shown i n FIGURE I . I t can be seen that, in agreement with our earlier results, most of the heterogeneous anti-al activity in each case was removed by the anti-idiotype absorption. Further experiments showed that the IdX-a1 determinant(s) is within, or close to, the antigen-combining site and expressed primarily on isolated H chains, but not L chains, of anti-al a n t i b ~ d y . ~ I n addition, IdX-al was found in rabbits that had been completely suppressed for V, a subgroup expression and then autoimmunized to induce anti-al antibody. This latter finding demonstrates that a l rabbits contain IdX-al within their genetic repertoire and that IdX-al can be associated with both the major ( a ’ ) and minor ( a ) V, subgroups as well as with various H and L chain allotypes. At this point, one must ask why heterogeneous anti-al antibodies would share an IdX marker. Two explanations can be offered. I . ) IdX-al plays an integral role in regulating allotype expression. The highly conserved gene segment encoding IdX-a 1 would most likely be carried within the J or D region, thereby allowing rearrangement with various V,, genes. 2.) Injection of anti-al, rather than inducing classical anti-idiotype, actually stimulates the production of molecules expressing “a1 -like” determinants in the antigen-combining site, that is, internal images of the a l antigen. Reaction of anti-allotype antibody with internal image epitopes would thus give the appearance of IdX sharing. This concept has been advanced by Jerne et ~ 1 . ~ to explain the sharing of idiotypes by rabbit anti-b6 antibodies. In order to distinguish between these two possibilities, anti-a1 antibody was isolated from a goat and tested for the ability to inhibit rabbit IdX-al binding to anti-idiotype. As seen in FIGURE 2, goat anti-a1 inhibits binding nearly as well as

Heating of immunoglobulins for immunoblot analysis destroys variable region antigenicity.

The antigenicity of rabbit immunoglobulin heavy chain variable region allotypes was examined by immunoblotting after sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The standard method of sample preparation resulted in a total loss of immunoreactivity with anti-allotype antibody due to heating of the samples prior to their electrophoresis. Thus, caution is warranted in interpreting results obtained from analysis of Ig when such proteins are heated before or during separation. In addition, we describe a method that permits not only separation of Ig heavy and light chains in SDS-PAGE but also preserves V region antigenicity.

EPITOPE MIMICRY BY ANTI-IDIOTYPE SEQUENCES IN REVERSE ORIENTATION

Antigen mimicry by anti—idiotypic internal images offers one approach to vaccine development and immune modulation. However, the rational design of internal image vaccines and an understanding of idiotypic interactions will depend upon a precise understanding of the rules which govern antigen recognition and of the structural basis for internal image expression. Previous studies using the reovirus hemagglutinin and a synthetic peptide (GAT) identified determinants in antibody internal image variable (V) regions that appeared to match continuous sequences within the respective antigens (Ollier et al., 1985 ; Bruck et al., 1986).

In vivo activation of quiescent B cells by anti-immunoglobulin. II: Production of lysozyme-specific monoclonal antibodies of desired isotype

We wished to determine whether injection of mice with anti-isotype antibody would be a means to regulate in vivo isotype expression and to obtain hybridomas secreting monoclonal antibodies (mAbs) of desired antigen specificity and isotype. Treatment with a rat mAb (7D2) reactive with both the IgG2a and IgG2b isotypes of mouse Ig resulted in large increases in the serum concentrations of mouse IgG2(a + b). Moreover, injection of antigen-7D2 conjugates had a profound effect on the isotype distribution of hybridomas subsequently obtained from these animals. Thus, while greater than 95% of anti-hen eggwhite lysozyme (HEL) mAbs prepared from mice immunized with HEL alone were of the IgG1 isotype, 12/15 (80%) of the mAbs from mice injected with HEL-7D2 conjugates were of the IgG2a or IgG2b isotype. When tested for effector functions using HEL-coated erythrocytes, the mAbs showed the expected activities, i.e., the IgG2, but not IgG1 anti-HEL mAbs were able to fix complement, bind protein A, and mediate antibody-dependent cytotoxicity and phagocytosis. These results indicate that in vivo immunization with anti-isotype-antigen conjugates can be used to produce hybridomas of predetermined antigen and isotype specificities.