Cytotoxicity in graft-versus-host reaction. I. Role of donor and host spleen cells.

Under in vitro conditions spleen cells from nonirradiated F(1) hybrids, in which a (graft-vs.-host) (GVH) reaction had been induced with lymphoid cells of parental origin, lysed nonspecifically target cells, i.e., cells syngeneic or allogeneic to the parental genotypes. Furthermore, tumor cells exposed in vitro to spleen cells of F(1) hybrid mice undergoing GVH reaction had markedly decreased ability to grow in syngeneic recipients. Experiments involving inhibition of cytotoxicity with alloantisera indicated that this nonspecific effect was due to host cells. By contrast, spleen cells of lethally irradiated F(1) hybrids undergoing GVH reaction lysed specifically the target cells of the genotype against which the parental (donor) cells had been sensitized; this finding further supports the contribution of host cells to the nonspecific cytotoxic effects in GVH reaction. From these results it was deduced that the cytotoxic effects during GVH reaction involve at least two processes: (a) sensitization of the donor cells to the antigens of the recipient resulting in the activation of their potential to lyse specifically the recipient's cells, and (b) activation of the host's cells into a state of nonspecific cytotoxicity, as a consequence of the immunologically specific attack of the donor cells.

Rejection of tumor cells in vitro.

The emergence of lymphoblast-like cells, capable of rapidly destroying tumor cells, was observed in primary cultures of an antigenic sarcoma transplantable in strain 13 guinea pigs. It is likely that these cytotoxic cells represent the progeny of lymphocytes sensitive to tutmor antigens that had infiltrated the tumor tissue.

Detection of tumor antibodies in patients with gastrointestinal carcinomas by a solid-phase radioimmunoassay.

Metastatic tumors from livers of 5 patients with gastrointestinal carcinomas and from the liver of 1 patient with malignant breast carcinoma were extracted with 3 M KCl; similar extracts were prepared from normal human colon and liver and from human fetal gut. The extracts were depleted of serum globulins by passage through reverse immunoadsorbent columns consisting of rabbit antibodies to the F(ab)2 fragment of human IgG and were then coupled to CNBr-activated paper disks. These "antigen" disks were used in a radioimmunoassay, with the aid of 125I-labeled rabbit antihuman F(ab')2 antibodies for the assay of circulating tumor antibodies produced by cancer patients. Statistical evaluation of the results with plasma samples from 47 patients with colorectal carcinomas and from 7 patients with other gastrointestinal disorders (polyps, villous papilloma, diverticulitis, and Crohn's disease) indicated that a significant number of patients had antibodies to cross-reactive tumor antigen(s). The cross-reactive tumor antigen(s) involved in the reaction was not detected in extracts of the gastrointestinal tract from 12-week human fetuses and did not cross-react with carcinoembryonic antigen.

A microradioimmunoassay for antibodies to tumor-associated antigens.

A versatile microradioimmunoassay for the detection of antibodies to tumor-associated and other tissue antigens was described. The method involved: a) the preparation of solid-phase antigen with cultured (already adhered) or noncultured cells (sedimented by centrifugation) fixed to Micro-Test plates with neutral buffered formaldehyde or absolute methanol; b) the incubation of the antigen with test or control sera; and c) the incubation of the antigen with radioiodinated antiglobulin antibody. The nonspecific background of radioactivity was reduced to an acceptable level by the fixed cells being precoated in the wells with 0.5% bovine serum albumin in phosphate-buffered saline which was also used for the dilution of sera and labeled antiglobulin antibody. Tumor cells in primary cultures gave a high background, as compared to long-term cultures, which was due to the presence of immunoglobulins (most likely tumor-specific antibody). The specific antibody response to a syngeneic mouse tumor was demonstrated by this technique.

Immunochemical characterization of a high molecular weight basic allergen (HMBA) of rye grass (Lolium perenne) pollen.

A high molecular weight basic allergen (HMBA) was isolated from the mixture of non-dialysable components of the aqueous extract of defatted rye grass pollen by a combination of gel filtration and isoelectrofocusing, HMBA, a glycoprotein of mol. wt 56,800 (17% carbohydrate) contained all naturally occurring amino acids. A hyperimmune rabbit anti-HMBA serum gave only a single precipitin band with the crude extract of the rye grass pollen in crossed immunoelectrophoresis. Thus, it was concluded that HMBA was a unique and highly purified antigen. The allergenicity of HMBA was revealed by its ability to elicit immediate skin reactions in grass allergic patients. Moreover, all patients' sera tested had IgE antibodies to HMBA detectable by direct RAST with HMBA allergosorbent discs. These observations indicated that HMBA was a major allergenic constituent of rye grass pollen. Treatment of HMBA by 6 M guanidine HCl led to a significant reduction in its ability to combine with human IgE antibodies. The treatment also resulted in the complete loss of allergenicity (i.e. inability to elicit PCA reactions with a murine reaginic antiserum to HMBA) and antigenicity (inability to form precipitins with rabbit anti-HMBA); hence, it would appear that the allergenic and antigenic determinants of HMBA are 'conformational'.

Allergenic cross-reactivity of cytochromes c of Kentucky bluegrass and perennial ryegrass pollens.

The allergenic cross-reactivity of cytochromes c isolated from Kentucky bluegrass (KBG) and ryegrass (RG) pollens was demonstrated by the finding that both cytochromes elicited PCA reactions of comparable titers in rats sensitized with murine reaginic antiserum to either KBG or RG cytochrome c. However, allergenic differences were revealed at limiting doses of the cytochromes c; under these conditions RG cytochrome c elicited PCA reactions only with the homologous reaginic serum, whereas KBG cytochrome c elicited PCA reactions with both murine reaginic sera. Moreover, in experiments involving inhibition of PCA, RG cytochrome c failed to neutralize some of the IgE antibodies of the antiserum to KBG cytochrome c, whereas KBG cytochrome c neutralized all IgE antibodies of both murine reaginic antisera. From these results it may be concluded that: (1) the two cytochromes c share common allergenic determinants, and (2) the KBG cytochrome c possesses additional allergenic determinant(s) not present on RG cytochrome c. The radioallergosorbent test, utilizing KBG cytochrome c discs and a pool of sera of individuals allergic to KBG, was inhibited to the extent of 87 or 41% by the addition of 10 micrograms of KBG or RG cytochrome c, respectively. By contrast both cytochromes c at 10 micrograms were equally effective in the inhibition (92%) of the binding of the IgE antibodies in this serum pool to RG cytochrome c allergosorbent discs. These experiments confirmed that the two cytochromes share common allergenic determinants.

Partial characterization of an antigenic site of high molecular weight basic antigen, a ryegrass pollen allergen, using a monoclonal antibody.

We have previously reported on the production of murine monoclonal antibodies (Mabs) to the retentate fraction of the aqueous extract of Kentucky bluegrass pollen (KBG-R) [Kisil et al. (1980) Fedn Proc. 39, 3479]. In the present study, the ability of one of these Mabs (Mab 12) to recognize various antigenic components present in KBG-R and the corresponding fraction of ryegrass (rye-R) was evaluated by the Western and immunoblot methods. Thus, KBG-R and rye-R were resolved electrophoretically into a large number of components. Remarkably, the concurrent immunoblot analysis with Mab 12 detected only a single antigenic component in each of the retentate fractions. The position of the antigenic component observed on these immunoblots was identical to that obtained with the rye allergen high mol. wt basic antigen (mol. wt 56,800). To characterize the antigenic site recognized by the Mab, the size of HMBA was reduced by cleavage with CNBr, the resulting fragments separated by high-performance liquid chromatography on a reverse-phase column and their antigenic activity analyzed by immunoblot. Two peptides, CB-1 (mol. wt = 17,400) and CB-2 (mol. wt = 22,000), retained the capacity to react with Mab 12 and also IgE antibodies present in a pool of sera from grass allergic individuals.

Mapping of antibody binding epitopes of a recombinant Poa p IX allergen.

Antibody binding epitopes of a recombinant Poa p IX allergen were delineated using recombinant DNA and solid-phase peptide synthesis procedures. The full-length cDNA clone KBG60 and its four overlapping recombinant fragments, KBG60.1, KBG60.2, KBG8.3 and KBG10 which spanned the entire molecule were synthesized in E. coli with aid of the plasmid expression vector, pWR590.1. The antigenic and allergenic sites of these recombinant proteins were analyzed by ELISA using human IgE and murine IgG antibodies. It was thus demonstrated that although the epitopes were found on all the fragments tested, the majority of these were located on a C-terminal fragment, rKBG8.3. Furthermore, synthetic peptides were also employed to identify the epitopes of rKBG60 protein. The use of antisera raised against native KBG pollen extract and the recombinant KBG8.3 protein to scan a total of 56 overlapping deca-penta peptides, covering the entire rKBG60 protein, revealed that 10 positive peptides involved in the antibody-binding site(s). Taken together, the results of these studies indicate that rKBG60 protein possesses at least 10 antibody binding epitopes.

Determinants of ryegrass pollen cytochrome c recognized by human IgE and murine monoclonal antibodies.

In attempts to produce fragments of an allergenic molecule which would retain allergenic and/or antigenic determinant(s), the cytochrome c of ryegrass (RG) pollen, which had been shown to be an allergenic constituent of this pollen, was digested with trypsin and chymotrypsin and the resulting fragments were separated by high performance liquid chromatography. Several of these fragments were shown, with the aid of the radioallergosorbent test and solid phase radioimmunoassays, to bind IgE antibodies present in a pool of six sera from grass-sensitive patients and three murine monoclonal antibodies, designated as Mab 41, Mab 42 and Mab 43, which had been originally produced against the crossreacting cytochrome c of Kentucky bluegrass (KBG). In summary, (i) fragments C-67 and C-74 reacted with all antibodies, (ii) fragments T-45, T-46 and C-69 bound to human IgE antibodies as well as to Mab 41 and Mab 42, but not to Mab 43, (iii) fragment T-44 reacted only with Mab 41 and Mab 42, and (iv) fragment C-83 bound only Mab 42 and Mab 43. On the basis of these results, it is concluded that (i) immunochemically active fragments of the RG cytochrome c can be readily produced by enzymatic degradation, (ii) there is significant crossreaction between the antigenic determinants of RG and KBG cytochromes c, (iii) whereas all fragments possessed at least two of the original antigenic determinants, fragments C-83 and T-44 were devoid of allergenic determinants, (iv) the antigenic determinants recognized by Mab 41 and Mab 42 were different from those reacting with human IgE antibodies and Mab 43, (v) each of the three monoclonal antibodies recognized a distinct antigenic determinant, (vi) fragments C-67 and C-74 possessed all determinants recognized by the human IgE and mouse antibodies used.

Immunomodulation in rats by transplantable anterior pituitary tumors.

The effect on the immune response of the MtT/W5 (W5), MtT/W10 (W10) tumors of Wistar-Furth rats and of MtT/F4 (F4) tumor transplantable in Fischer 344 rats was examined. Antibody response to sheep red blood cells and contact sensitivity reactions to dinitrochlorobenzene (DNCB) were used as immune parameters. In intact rats the W5 tumor suppressed the antibody response as much as did hypophysectomy (Hypox). The antibody response of Hypox rats was marginally improved by this tumor. Contact sensitivity was not suppressed in intact animals and the poor reactivity of Hypox rats was reconstituted to normal levels by W5. Treatment with bromocriptine (BRC) had no influence on tumor growth or on the immune reactivity of tumor-bearing hosts. The W10 tumor suppressed antibody formation, but not contact sensitivity in intact animals; the antibody and the DNCB response of Hypox animals was reconstituted partially, by this tumor. BRC treatment of tumor-bearing animals mimicked the effect of Hypox in this system. The F4 tumor suppressed antibody formation in intact rats and did not reconstitute the reactivity of Hypox rats. The DNCB response was not influenced by this tumor in intact animals and it was partially reconstituted in Hypox rats. Again, the reactivity of BRC-treated animals was similar to that of Hypox tumor-bearing animals. These results indicate that pituitary tumors may alter the immune reactivity of their hosts significantly.

Radioimmunoassay for human growth hormone using monoclonal antibodies.

Radioimmunoassay (RIA's) for human growth hormone (hGH) employing monoclonal antibodies (Mab) were compared with a conventional clinical RIA using polyclonal antibodies (PAb). Several MAb displayed sensitivities (defined as the concentration of hormone required for 50% inhibition of binding of 125I-hGH) for hGH in the RIA which were equivalent to those of PAb. Those MAb with the greatest sensitivity for hGH displayed an equal affinity for human placental lactogen (hPL), whereas those with lower sensitivity for hGH, cross-reacted minimally with hPL. The concentrations of hGH in human serum samples were determined with MAb and the values showed a high degree of correlation with those obtained by PAb. It is therefore concluded that MAb are useful for the development of clinical RIA's for polypeptide hormones.

The influence of antibody affinity on the radioallergosorbent test (RAST) and in vitro histamine release. Studies with hapten-specific monoclonal IgE antibodies.

Four murine monoclonal IgE antibodies specific for the hapten, 4-hydroxy-3-nitrophenylacetyl (NP), had been previously found to be heteroclitic in nature in that they bound the crossreacting hapten, 4-hydroxy-3-iodo-5-nitrophenylacetyl (NIP), with greater affinity than NP. The influence of antibody affinity on the results of two commonly used assays for IgE, namely the radioallergosorbent test (RAST) and histamine release from rat peritoneal mast cells, was studied using these antibodies. In general, in agreement with previous reports, it was found that affinity influences both RAST and histamine release; however, the affinity constants deduced from equilibrium dialysis measurements for the reactions with monovalent haptens were not directly related to the activities of the antibodies as reflected in assays using multivalent hapten protein conjugates.

Development and characterization of ouabain-resistant human fusion partners.

The ouabain-resistant mutant cell lines, HOA-1 and HOA-20 were developed from WI-L2-729-HF2 by cloning with increasing concentration of ouabain. Both parent and mutant cell lines were resistant to base analogues, 6-thioguanine (6-TG) and 8-azaguanine (8-AG) to the level of 20 micrograms/ml in the culture medium. The parent cell line WI-L2-729-HF2 was highly sensitive to ouabain, whereas HOA-1 and HOA-20 were resistant to ouabain to the level of 1 microM and 20 microM, respectively. However, all the cell lines were sensitive to HAT-selective medium which is essential for hybrid selection after fusion. All three lymphoblastoid cell lines were positive for Epstein-Barr virus nuclear antigen (EBNA), secreted TNF-beta (lymphotoxin) without any external stimulation, secreted trace amounts of IgG(kappa), which was also present in their cytoplasm and had IgM(kappa) as surface bound immunoglobulin. They also expressed the CD20, CD71 (transferrin receptor) as surface antigens. In addition to these antigens, HOA-20 also expressed CD38 antigen. The karyotype analysis of these cell lines revealed modal chromosomal numbers ranging from 40 to 47. The HLA-A, -B and -C antigens expressed by WI-L2-729-HF2 and its mutants HOA-1 and HOA-20 were identical. Both the HOA-1 and HOA-20 mutants were found suitable for the generation of hybrids after fusion with EBV-transformed human B-lymphocytes.

A method for the preparation of protein-protein conjugates of predetermined composition.

A novel procedure for the synthesis of well-defined protein-protein conjugates is described using ovalbumin (OA) and IgG as test proteins. This procedure involves the highly selective and rapid reaction of alkyl halide and sulfhydryl groups, which have been grafted, respectively, onto the proteins to be conjugated. Accordingly, iodoacetylated IgG, (ICH2CO)nIgG, was prepared by the reaction of the epsilon-amino groups of IgG with the N-hydroxysuccinimide ester of iodoacetic acid (NHIA), the degree of iodoacetylation (n) being proportional to the concentration of NHIA. OA was reacted with S-acetylmercaptosuccinic anhydride (SAMSA) under conditions yielding, on the average, a monosubstituted derivative. Following removal of the protective S-acetyl group, the resulting -SH derivative of OA was reacted with (ICH2CO)nIgG. The OAx-IgG conjugates so produced were characterized by gel filtration, specific radioactivity (using tritiated OA) and immunodiffusion. It was found that the average number of OA molecules coupled per IgG molecule could be controlled by varying the degree of iodoacetylation of IgG.

The development of a radioallergosorbent test (RAST) for murine IgE antibodies.

A purified monoclonal IgE preparation, isolated from the ascitic fluid of mice bearing a hybridoma secreting IgE with specificity to ovalbumin, was used for the production of goat anti-murine IgE (GAME) antiserum, which was then rendered monospecific for the epsilon chain. Another monoclonal hybridoma IgE preparation with specificity for the 2,4-dinitrophenyl group was isolated from ascitic fluid in a relatively pure state by affinity chromatography and used in the form of an immunosorbent to isolate antibodies from the monospecific goat serum. The GAME antibodies were 125I-labeled and used to develop a radioallergosorbent test (RAST) for the quantitation of murine IgE antibodies specifically adsorbed onto antigen-coupled paper discs. The RAST was specific for antibodies of the IgE class only and was as sensitive as and more accurate than PCA assay. RAST results on sera of mice treated with tolerogenic conjugates indicated a reduction in the affinity and concentration of the IgE antibody populations on suppression of the IgE response. The effect of interference by non-IgE antibody populations on the RAST curves has been discussed.

Human--human hybridomas secreting lipid A reactive monoclonal antibodies.

Five human-human hybridoma cell lines secreting monoclonal antibodies (mAbs) against lipid A (LA) were produced by cell fusion of Epstein-Barr virus (EBV) transformed human peripheral blood lymphocytes and a human lymphoblastoid cell line KR-4. All these mAbs were isotyped as IgM(kappa) and reacted with lipopolysaccharides (LPS) and LA of various gram-negative bacteria. Whereas the binding of only four of the five mAbs to solid-phase LA was blocked by polymyxin-B sulphate, the mitogenic effect of LPS and LA on murine B lymphocytes was inhibited by all five mAbs. These results demonstrate that the human immune system recognizes at least two common epitopes in lipid A of various gram-negative bacteria.

Direct action of cyclosporin A on human B-lymphocytes with regard to immunoglobulin production.

Highly purified preparations of human B-lymphocytes were cultured with or without cyclosporin A (CyA; 1 microgram/ml) for 8 days with pokeweed mitogen (PWM) in the presence of helper factors or with Epstein-Barr virus (EBV). The amounts of IgG, IgM, IgA and IgE produced in cultures were measured by radioimmunoassays or by reverse hemolytic plaque-forming cell assays. The results demonstrate that whereas CyA had a strong suppressive effect on the production of immunoglobulins (Ig) by PWM-activated B-cells, it had an enhancing effect on EBV-activated B-cells. It is concluded that CyA has a direct effect on human B-lymphocytes and that it may suppress or enhance their activation depending on the stimulant employed to trigger the cells.

Tolerogenic polyethylene glycol derivatives of xenogeneic monoclonal immunoglobulins.

It is becoming apparent that the effectiveness of xenogeneic monoclonal antibodies (XIg), which are increasingly used for diverse therapeutic purposes in man, may be counteracted by their inherent immunogenicity. Since conjugates of proteins with monomethoxypolyethylene glycol (mPEG) have proved to be effective tolerogens in other systems, we have used an experimental model in mice to explore the tolerogenicity of mPEG conjugates of a human monoclonal IgG (HIgG), i.e. a myeloma protein. Administration of these conjugates prior to immunization with heat aggregated HIgG (ha-HIgG) resulted in specific tolerance, as manifested by a marked reduction in the level of antibodies to HIgG, which was related to the degree of conjugation and the dose of conjugate administered. Thus, administration of HIgG(mPEG)20 6 to 43 days prior to immunization with ha-HIgG resulted in an inhibition of anti-HIgG antibody formation of the order of 85-90%, in relation to the titres of mice receiving PBS in lieu of HIgG(mPEG)20; these results hold the promise that mPEG conjugates of XIg may prove therapeutically useful in man in relation to organ transplantation, localization of tumours by immuno-imaging and tumour destruction by immunotoxins.

Down-regulation of secondary in vitro antibody responses by suppressor T cells of mice treated with a tolerogenic conjugate of ovalbumin and monomethoxypolyethylene glycol, OVA(mPEG)13.

In previous studies from this laboratory it was shown that OVA(mPEG)n conjugates induced: (i) tolerance in mice with respect to IgG and IgE antibody responses to dinitrophenylated OVA (DNP-OVA); and (ii) OVA-specific suppressor T (Ts) cells which could down-regulate a primary immune response in vivo. For the present study, we have developed an in vitro culture system for assessing the activity of Ts cells of mice tolerized by an OVA(mPEG)13 conjugate. Spleen cells from mice which had been primed with DNP4-OVA in Al(OH)3 gel were cultured with DNP4-OVA to induce a secondary antibody response in vitro. After 6 days, cells secreting anti-DNP antibodies of the IgG1 class were enumerated by an immunoenzymatic plaque-forming cell assay. Addition to the culture of T cells from mice treated with 3 i.p. injections of 500 micrograms of OVA(mPEG)13 resulted in a 29-61% reduction in the number of IgG1 anti-DNP antibody-forming cells, in comparison with the effect of T cells from mice treated with PBS. It was concluded that this tolerogenic conjugate induced splenic Ts cells which were capable of suppressing secondary in vitro anti-DNP responses.

Suppression of antibody responses in rats to murine anti-CD4 monoclonal antibodies by conjugates with monomethoxypolyethylene glycol.

The effectiveness of therapeutically relevant xenogeneic monoclonal antibodies (MoAb) may be counteracted by their inherent immunogenicity. Since conjugates of diverse proteins with mono-methoxypolyethylene glycol (mPEG) were shown to induce Ag-specific tolerance in mice and rats, we used outbred rats in this study as an experimental model for establishing the tolerogenicity of mPEG conjugates of murine MoAb. The results demonstrate that: (i) murine anti-rat CD4 MoAb (W3/25) were more immunogenic than murine anti-human CD4 MoAb (MAX.16H5) in rats; (ii) W3/25 preferentially induced an anti-idiotypic (anti-id) antibody response; and (iii) antibodies to both common and idiotypic determinants could be suppressed in rats by treatment with W3/25(mPEG)28.