Interference in immunoenzymometric assays caused by IgM anti-mouse IgG antibodies

Serum samples from a female patient gave falsely elevated results in nine of ten immunoenzymometric assays (IEMAs) that used mouse monoclonal antibodies specific for human chorionic gonadotropin (n = 8), thyrotropin, or creatine kinase-MB isozyme. In contrast, normal results were obtained in five radioimmunoassays that used either mouse monoclonal antibodies or antisera specific for human chorionic gonadotropin (n = 3) or thyrotropin (n = 2). Incubation of her serum with IgG from different species or with F(ab)'2 from mouse IgG prior to IEMA showed that the interference was markedly inhibited by mouse IgG, indicating an antibody specific for the Fc portion of mouse IgG. The interfering activity was bound to and eluted from a column containing Protein A-Sepharose CL-4B. Fractionation of the eluted protein over another column containing Sephacryl S300 showed the activity was enriched in the first protein peak, which contained predominantly IgM. A model is proposed to explain how IgM anti-mouse IgG antibody selectively interferes in IEMAs that use mouse monoclonal antibodies.     

Immunological methods for quantifying free and total serum IgE levels in allergy patients receiving omalizumab (Xolair) therapy

Omalizumab (humanized-IgG1 anti-human IgE Fc, Xolair) complexes circulating IgE, blocking IgE binding to high affinity epsilon Fc receptors (FcepsilonR1) on mast cells and basophils. Free (non-Omalizumab bound) IgE levels in serum are a measure of effective Omalizumab dosing. The goal of this study was to quantify free (non-Omalizumab-complexed) and total serum IgE levels in asthma patients on Xolair. The concentration of (non-Omalizumab bound) free IgE in human serum was measured using a solid phase immunoenzymetric assay (IEMA) in which IgE was captured from serum with monoclonal anti-human IgE (clone HP6061) and detected with labeled-FcepsilonR1alpha. In a companion total human serum IEMA, IgE was captured from serum with the same anti-human IgE (clone HP6061) and all bound IgE was detected with labeled monoclonal anti-human IgE Fc (clone HP6029). Free and total IgE levels were quantified in pre- and 1 and 3 months post Omalizumab therapy sera from 12 allergic asthma patients. In the absence of Omalizumab, working ranges of the free and total IgE IEMAs were comparable (10-1000 kIU/l), with excellent precision, reproducibility and parallelism. Pre-Omalizumab total and free IgE levels by IEMA were highly correlated (r2=0.99, Y=0.9X+0.32, p<0.001), as were total serum IgE levels by IEMA and ImmunoCAP-250 (r2=0.98, Y=1.1X-0.05, p<0.001, n=33). In vitro reduction of free IgE (>90%) occurred at [Omalizumab:IgE] molar ratios of 2-20. Total IgE levels in 12 asthmatics increased from pre-therapy levels (52-658 kIU/l) by 1.5-5.5-fold at 1 month and 1.7-8.6 fold at 3 months of uninterrupted Omalizumab treatment. Free IgE levels fell by 49%-97% at 1 month and 45%-98% by 3 months of Omalizumab treatment. Free and total IgE levels by IEMA aid in monitoring patients receiving Omalizumab therapy.     

Identification of epitopes recognized by a panel of six anti-human IgG2 monoclonal antibodies.

Human IgG2 contains several subclass specific amino acid residues or deletions in the CH1 and CH2 domains and also in the hinge region. These substituted residues are the structural correlates for IgG2 specific epitopes. Since human IgG2 has different biological properties from other subclasses, some IgG2 epitopes may be located in regions correlating with sites determining the biological functions. Previously, we produced three anti-IgG2 monoclonal antibodies (mAbs) with highly specific and interesting reactivities using improved immunization protocols. However, it has been almost impossible to identify epitopes conventionally, because human IgG2 is so resistant to proteolysis that various proteolytic fragments could not be isolated. In this study, we identified the epitopes recognized by anti-IgG2 mAbs by SDS-PAGE, Western blotting, amino acid sequence analysis and peptide/mAb binding ELISA, thus overcoming the need for fragment isolation. A panel of six anti-human IgG2 mAbs, including the current WHO/IUIS specificity standards (HP6002, HP6008, HP6014) and our own (HG2-6A, HG2-30F, HG2-56F), reacted with distinct epitopes. The residues essential to expression of the epitopes recognized by the mAbs were: Pro234, Val235 and Val309 for HG2-56F, HG2-30F and HP6008, respectively. HP6014 reacted with the epitope expressed by Thr214 and its neighboring residues. HG2-6A was reactive with the hinge region, and HP6002 was assumed to be directed against discontinuous epitopes requiring intact Fc for expression. Through these studies, the pepsin and papain cleavage sites of human IgG2 were also clarified.     

Presence of IgG4 on the membrane of human basophils. Histamine release is induced by monoclonal antibodies directed against the Fab but not the Fc region of the IgG4 molecule

We have studied the possible role of human IgG4 as an anaphylactic antibody. For that purpose, we have determined the induction of histamine release (HR) from human basophils by anti-IgE and anti-IgG4 monoclonal antibodies (MoAbs) recognizing different epitopes located at the Fc and Fab regions of the IgG4 molecule. The results show that anti-IgG4 (Fab) MoAb was able to induce HR in 93% of donors tested, with no differences between atopics and non-atopics. That HR is calcium dependent and is accompanied by the synthesis and release of leukotriene C4. In contrast, no HR could be induced by anti-IgG4(Fc) MoAbs in any individual, even in the presence of D2O or after a second challenge with a polyclonal goat anti-mouse IgG antibody. The results obtained suggest the presence of IgG4 on the basophil membrane and that the epitope recognized by the anti-IgG4 (Fc) MoAbs is probably hidden in cell-bound IgG4. This was demonstrated by immunofluorescence techniques: IgG4 bound to the basophil membrane could be detected with anti-IgG4(Fab) but not with anti-IgG4(Fc) MoAbs. In addition, we found that nine donors were unresponsive to an anti-IgE stimulus, while they released histamine efficiently after challenge with anti-IgG4(Fab), suggesting the existence of different receptors for both immunoglobulins.     

A quantitative ELISA for antigen-specific IgG subclasses using equivalence dilutions of anti-kappa and anti-subclass specific secondary reagents Application to the study of the murine immune response against the capsular polysaccharide of Neisseria meningitidis serogroup B

We have developed an enzyme-linked immunosorbent assay (ELISA) to measure murine antigen-specific IgG antibodies of defined subclass using precalibrated equivalence dilutions of anti-κ (in the standard) and each anti-IgG subclass-specific polyclonal secondary antibody (in the test sample). The calibration of secondary reagents could be carried out easily with a set of monoclonal antibodies (MoAbs) specific for all IgG subclasses. These MoAbs do not require purification or standardization. In addition the MoAbs can be of different antigenic specificity. Once the equivalence dilutions have been determined, they can be applied in a quantitative ELISA using the same antigen in the standard and sample, and using only one IgG subclass standard for the determination of all the IgG subclasses. The method is easy to standardize for many antigenic systems. It is particularly useful when the only standard available is one standardized MoAb of the appropriate specificity, and it could be adapted to use with standard polyclonal antibodies having a known content of total antigen-specific IgG bearing κ chains but unknown IgG subclass composition. The use of this method to quantitate IgG specific for the capsular polysaccharide of Neisseria meningitidis serogroup B (CpsB) gave highly reproducible measures with an interbatch CV of 5–6% similar for all IgG subclasses and low detection limits ranging from 0.3 ng/well for IgG3 to 0.8 ng/well for IgG2a. The IgG subclass response observed after immunization with live meningococci was mainly IgG2a (74%) and IgG2b (18%). Hyperimmunization modified this IgG distribution to one of mainly IgG3 (62%) and IgG1 (28%) which was maintained in the response to a single immunization 4 weeks later, possibly indicating the generation of resting B cells during continuous stimulation.     

Subclass specificities of rabbit antibodies to human antidextran

Rabbits were immunized with purified antidextran containing IgG2. Half of the animals were treated with commercial human IgG intravenously to induce tolerance. The antisera were absorbed and/or eluted from IgG-Sepharose columns and the antibodies tested with a panel of at least eight human myeloma proteins of the four IgG subclasses, both kappa and lambda types. All animals produced mainly anti-IgG2 or IgG4, sometimes intechanged between these maxima or between kappa and lambda types over a series of bleedings. Absorption with IgG2 and/or IgG4 myeloma proteins resulted in sera or antibody preparations specific for the IgG1 or IgG3 subclasses, and the amount of those antibodies also varied between bleedings from the same animal. It may be concluded that anti-IgG2 and anti-IgG4 syntheses are related in a complementary manner or are equally likely to be produced in a response to IgG2; while IgG1 and IgG3 are also interrelated, but with an inverse relationship to IgG2 and IgG4. Alternatively anti-IgG2 and anti-IgG4 antibodies could be highly cross-reactive, and so could anti-IgG1 and anti-IgG3 antibodies.     

Anti-IgG antibodies in rheumatic diseases cross-react with Streptococcus mutans SR antigen.

We have previously shown that SR protein, a S. mutans major cell wall protein, as well as the recombinant protein SR (rSR) share common epitopes with human IgG. Since this antigenic mimicry could play a role in the induction of anti IgG, we have examined, in k-ELISA, the presence of antibodies reacting with S. mutans SR proteins and S. mutans whole cells in sera from 36 patients with rheumatic diseases. The majority of the 36 sera showed a high reactivity with rSR when compared with control sera. Eight highly positive sera were further purified on rSR and human IgG sorbents and tested against both rSR and IgG in ELISA and Western blotting. The affinity-purified antibodies reacted strongly with rSR, IgG and IgG Fab fragments but failed to react with IgG Fc fragment. In Western blotting the addition of unlabelled IgG abolished the reactivity of affinity-purified biotinylated antibodies with all antigens, confirming the existence of a common epitope shared by rSR and human IgG heavy chain. We show the existence in rheumatic diseases of high titres of anti-human IgG antibodies cross-reactive with S. mutans SR proteins. Those antibodies are principally IgG and react with the Fd part of the Fab fragment. We can hypothesize from the above data that this antigenic mimicry existing between S. mutans SR-related antigens and human IgG could play a role in the synthesis of at least a part of the anti-IgG antibodies present in rheumatic diseases sera.     

A triple antibody assay for the quantitation of plasma IgG subclass antibodies to acetylcholine receptors in patients with myasthenia gravis

Monoclonal anti-human IgG subclass antibodies have been used in an immunoprecipitation assay for the determination of anti-acetylcholine receptor IgG subclasses in plasma from patients with myasthenia gravis. Solubilized acetylcholine receptors labelled with 125I-alpha-bungarotoxin were incubated with patient plasma. Monoclonal mouse antibodies to human IgG subclasses 1-4 were added to the incubation and finally precipitated with anti-mouse IgG antibody. A maximal IgG subclass precipitation of 62-76% was determined with 125I-labelled myeloma IgG subclasses 1-4 added to normal human plasma. The anti-IgG subclass antibodies were added in excess which ensured that the precipitation of IgG2, IgG3 or IgG4 were unchanged, and that of IgG1 was only reduced by 17%, when the plasma IgG concentration was increased by a factor of two. The anti-IgG subclass antibodies were highly specific for their complementary subclasses. Determination of the IgG subclass of the anti-acetylcholine receptor antibodies from 8 patients with myasthenia gravis showed that IgG1 and IgG3 antibodies are predominant. This may support the hypothesis that complement mediated lysis of the neuromuscular end-plate plays a pathogenetic role in myasthenia gravis.     

the heterogeneity of bovine IgG2—IV. Structural differences between IgG2a molecules of the A1 and A2 allotypes

The major product of digestion of bovine IgG2a(A1) with immobilized pepsin is F(ab')₂ while similar treatment of IgG2a(A2) yields Fab, Fc and other products. It is postulated that structural differences in the hinge region, including the absence of the most N-terminal disulfide bridge in IgG2a(A2) and a displacement of the primary pepsin cleavage site toward the Fd, can explain this effect. The immunodominant A1 allotope(s) appears to be located in the CH3 domain of IgG2a(A1) while the A2 allotopes are located elsewhere and are apparently affected by digestion. The allotypic bias of rabbit anti-IgG2a is also present in anti-IgG2a raised in goats. However, the A2 allotypic determinants of bovine IgG2a are recognized by goat precipitins although precipitins of this specificity are not detectable in rabbits immunized with IgG2a(A2). Rabbit anti-A2 antibodies are detectable using the ELISA in rabbits immunized with IgG2a(A2).     

ELISA quantitation of IgG subclass antibodies to dietary antigens

The IgG subclasses of human antibodies against 2 dietary antigens, ovalbumin (OA) and beta-lactoglobulin (BLG), were studied by ELISA using monoclonal anti-human IgG subclass antibodies. Under the assay conditions used, the anti-IgG subclass antibodies were subclass specific. Quantitative estimates of the subclass antibodies were obtained by reference to a 'capture' assay using F(ab')2 anti-light chain antibody as ligand and IgG myelomas as standards. The validity of these estimates was supported by antibody quantitation using the Farr assay. In healthy adults with serum anti-OA or anti-BLG antibodies, anti-OA antibodies were found mainly in the IgG1 (9/11) and IgG4 (6/11) subclasses, whereas 5 sera showed high levels of IgG2 antibodies. In contrast, the IgG subclass distribution of anti-BLG antibodies was predominantly IgG4 (10/10).     

A Three-Layer Immunoradiometric Assay for Determination of IgG Subclass Antibodies in Human Sera ("IgG Subclass RAST")

We report the development of a three-layer immunoradiometric assay (TIRA) for measurement of IgG antibodies of all four subclasses in human sera. The first layer consists of diluted human serum, the second layer is monoclonal mouse antibodies to human IgG subclasses., and the third layer is ¹²⁵I-labelled rabbit anti-mouse IgG. Monoclonal anti-IgG1. anti-IgC3 and anti-IgG4 reacted only with their complementary IgG subclass, whereas the anti-lgG2 showed slight cross-reactivity to immunoglobins of other subclasses and classes and to light chain proteins. The observed cross-reactivity was found to be without importance, when the TIRA was applied to measurement of IgG subclass antibodies. Equipotency was established by use of appropriate dilutions of the monoclonal antibodies, and the assay was calibrated by use of human reference serum. The TIRA therefore permits reliable inter-individual and intra-individual comparisons of the IgG antibody response in all four subclasses. Von-sped fie binding obtained with pooled normal human serum was below 0.33'#. Inter-assay coefficient of variation was between 18 and 27%, The TIRA was applied to measurement of IgG subclass antibodies to timothy grass pollen in sera from grass pollen allergies undergoing immunotherapy.     

Obtention of monoclonal antibodies against human IgG4 using two different immunization strategies: development of a biotin-based ELISA for IgG4 quantitation.

Several monoclonal antibodies (MAbs) against human IgG4 have been produced in two fusion experiments. In fusion I, Balb/c mice were conventionally immunized and most of the antibodies obtained recognized IgG class-specific epitope(s). Therefore, a second fusion experiment was performed using IgG4-immunized mice that had been made tolerant to epitopes shared by the four IgG subclasses, and most of the hybridomas derived from this second fusion secreted antibodies specific for IgG4. Five MAbs which were found to be specific for both IgG4 isoallotypes (4a and 4b) were selected from the two fusion experiments. These antibodies define at least two non-overlapping epitopes located on the Fab and Fc regions of IgG4 molecule, respectively. Two MAbs were used to develop a two-site ELISA for IgG4 quantitation, in which one MAb (anti-Fab) was immobilized in the solid phase and the other one (anti-Fc) was biotin-labeled. The assay had a detection limit below 1 ng/ml and a practical working range between 4 and 200 ng/ml. The use of these MAbs should improve our understanding of the role of IgG4 in different clinical conditions.     

Studies of human anti-IgM anti-IgG cryoglobulins. I. Patterns of reactivity with autologous and isologous human IgG and its subunits.

Monoclonal human anti-IgG preparations purified from mixed IgM-IgG cryoglobulins were tested for their antigenic specificity by haemagglutination-inhibition assay. A panel of fourteen IgG preparations of the four gamma chain subclasses were prepared from myeloma sera and used as inhibitors of haemagglutination. Each of six IgM anti-globulins demonstrated different reactivity profiles with these IgG preparations. In addition, the fraction of the serum IgG which had bound to and cryoprecipitated with the IgM preparations, termed 'antigen-IgG', was purified and assayed for subclass content. The gamma chain subclasses found in the 'antigen-IgG' fractions showed that each IgM cryoprecipitated an IgG from serum which had different quantities of the subclasses present. These 'autologous' reactivity patterns were in instances different from the specificities expected from the results obtained with the myeloma proteins. When all antigen-IgG pools were tested with each IgM, some antiglobulins showed stronger reactivity with isologous than with their own, antigen-IgG pools. The IgM anti-IgG preparations were also compared in reactivity with IgG and its subunits in order to localize the antigenic determinant(s) with which these autoantibodies react. Heavy chains showed far greater reactivity than Fc fragment for 5/6 IgM preparations. Light chains, F(ab')2, pFc' and Fab were non-reactive. A relationship between the length of papain digestion and Fc reactivity was demonstrated. Based on the data, possible locations for the antigenic determinant(s) were considered.     

Determination of human IgG subclass by EIA]

The human IgG subclasses have specific characteristics in response to respective antigens, and determination of their characteristics is important in analysis of infectious, autoimmune and other diseases. We developed quantitative EIA (ELISA) assays for Japanese human IgG subclasses with excellent reproducibilities using monoclonal antibodies. The assays enable determinations between 0.4 ng/ml and 2,000 ng/ml. The serum IgG subclass levels (mean+SD) in healthy adults are as follows: IgG1, 5.93 + 2.07 mg/ml (59.94%); IgG2, 3.36 + 1.84 mg/ml (32.97%); IgG3, 0.30 + 0.19 mg/ml (3.07%); and IgG4, 0.22 + 0.15 mg/ml (2.34%). Excellent correlation was obtained between total IgG level measured by EIA using anti-whole human serum and the sum of the four IgG subclasses (IgG1+IgG2+IgG3+IgG4). Samples from every patient with M protein of IgG1 and IgG2 types shown by M band on immunoelectrophoresis also had the highest level in four IgG subclasses.     

Rapid immunoassays for the measurement of immunoglobulin G subclass concentration in immunoglobulin preparations and human serum.

Enzyme immunoassays (EIA) capable of determining total IgG1, IgG2, IgG3 and IgG4 subclass concentrations in human serum preparations have been developed. Subclass-specific monoclonal antibodies (mAbs) are bound to polyacrylamide bead-conjugated anti-mouse immunoglobulin antibodies. Bound immunoglobulins are detected with a peroxidase-conjugated anti-IgG antibody or a biotin-conjugated anti-IgG antibody followed by peroxidase streptavidin. The standard curves were found to be linear in the regions 16.0-2.0 micrograms/ml for IgG1, 4.0-0.5 micrograms/ml for IgG2, 0.4-0.06 micrograms/ml for IgG3 and 0.25-0.05 micrograms/ml for IgG4. Coefficient of variation (CV) values range from 0.32-7.32% for IgG1, 0.66-4.85% for IgG2, 1.62-6.85% for IgG3 and 0.05-6.47% for IgG4 standard curves. The inter-assay variability for the control human serum samples was 9.6% for IgG1, 6.7% for IgG2, 9.5% for IgG3 and 6.8% for IgG4.     

Specific cleavage sites on human IgG subclasses by cruzipain, the major cysteine proteinase from Trypanosoma cruzi

Cruzipain, the major cysteine proteinase of Trypanosoma cruzi, might have other biological roles than its metabolic functions. In this report, we have explored the interaction of cruzipain with molecules of the immune system. The enzyme was used to digest all human IgG subclasses at different pH values and lengths of time. At pH 7.3, all subclasses were readily split at the hinge region. Immunoblot and amino acid sequence analysis showed fragments of IgG1 and IgG3 to be compatible with Fab and Fc, whereas IgG2 and IgG4 rendered Fab2 and Fc. In all cases the fragments produced might impair the binding capacities and the effector functions of specific IgG. At these cleavage sites cruzipain displays cathepsin L and/or cathepsin B activities and shows a clear preference for Pro at the P'2 position and polar residues at P1. Despite the activity of cruzipain within the hinge, the enzyme also cleaved all heavy chains between the CH2 and CH3 domains; producing Fc'-like-fragments of 14 kDa. These fragments are potential candidates to block or saturate Fc receptors on immunocompetent cells. At mild acidic pH cruzipain produced further degradation of the Fc of all subclasses, the Fd of IgG4 and partially the Fd of IgG1, with the consistent loss of any antibody activity. The L chains apparently were not affected. Thus, cruzipain should be able to modulate, depending on the subclass selected and the pH of the environment, the production and the length of different biologically active/inactive IgG fragments.     

Activity of Two Types of Fc Receptors, FcγRI and FcγRII, in Human Monocyte Cytotoxicity to Sensitized Erythrocytes

We investigated the cytotoxicity of human monocytes mediated by two types of receptors for the Fc portion of IgG, Fc gamma RI and Fc gamma RII. Erythrocytes sensitized with human IgG (EA-human IgG) were used to assay Fc gamma RI function, and erythrocytes sensitized with mouse IgG1 (EA-mouse IgG1) were used to assay Fc gamma RII. Both types of Fc gamma R were observed to mediate antibody-dependent cell-mediated cytotoxicity (ADCC), which was further characterized by using different monoclonal anti-Fc gamma R antibodies (MoAb) and monomeric IgG. Lysis of EA-human IgG was inhibited by both monomeric human IgG and mouse IgG2a in a dose-dependent way, and also by anti-Fc gamma RI MoAb 10.1. Cytolysis of EA-mouse IgG1 was inhibited by monomeric mouse IgG1 and by two anti-Fc gamma RII MoAb, IV.3 and CIKM5. Antibodies of the mouse IgG2b isotype affected neither type of ADCC. The effectiveness of cytotoxicity mediated by either of the Fc gamma R was studied by means of targets sensitized with a calibrated number of IgG molecules. At least 20 times more IgG molecules per target cell were necessary to obtain half-maximal cytotoxicity mediated by Fc gamma RII than for Fc gamma RI-mediated cytolysis. Furthermore, the previously described polymorphism of Fc gamma RII was also reflected in Fc gamma RII-dependent cytotoxicity. These studies demonstrate that Fc gamma RII can mediate ADCC, although a higher degree of target cell sensitization is required than for Fc gamma RI-mediated ADCC.     

ELISA measurement of IgG subclass production in culture supernatants using monoclonal antibodies

Specific and sensitive ELISA to quantitate the human IgG subclasses in cell culture supernatants are described. These assays detect a minimum of 5 ng/ml IgG1, 90 ng/ml IgG2, 8 ng/ml IgG3 and 8 ng/ml IgG4 and can generally measure IgG subclasses in lymphocyte cultures containing a minimum of 200 ng/ml of total IgG. The isotype specificity of these ELISA is demonstrated and each individual ELISA shown to react with a number of paraproteins of the relevant subclass independently of their light chain type or their (major Caucasian) allotype. These assays have been used to determine the IgG subclass response of normal human lymphocytes to pokeweed mitogen in vitro.     

Clinical significance of IgG subclasses of anti-Sm and U1 ribonucleoprotein antibodies in patients with systemic lupus erythematosus and mixed connective tissue disease

IgG subclasses of anti-Sm and anti-U1 ribonucleoprotein (U1 RNP) antibodies were determined using a new clone of the anti-IgG2 antibody (HG2-56F). Although the predominance of IgG1 coincided with previous reports, IgG2 anti-Sm and U1 RNP antibodies were detected in numerous patients. IgG3 anti-Sm antibody significantly correlates with joint involvement and a high titer of anti-DNA antibody. On the other hand, IgG4 anti-U1 RNP antibody significantly correlated with esophageal dilation and muscular involvement. These results may suggest that some IgG subclasses are related to a specific clinical feature or manifestation.     

Study of induction of activation of human peripheral blood mononuclear cells with a non-activating form of anti-CD3 MoAb in autoimmune thyroid disease (AITD).

Anti-CD3 (OKT3) MoAb is a mitogenic agent which activates lymphocytes. We have studied the effects of murine anti-human OKT3 MoAb (IgG1) alone or in combination with IL-2, human thyroglobulin (Tg) and thyroperoxidase (TPO) antigens on the proliferation of whole peripheral blood mononuclear cells (PBMC) (including monocytes) or subtypes (T, CD4+, CD8+, B) as measured by tritiated thymidine (3H-TdR) incorporation. B cell differentiation was studied by measuring numbers of IgG-secreting cells and specific anti-TPO/anti-Tg-secreting cells by SPOT ELISA. PBMC or lymphocyte subtypes, obtained from 45 patients with Hashimoto's thyroiditis (HT), 40 Graves' disease (GD) and 51 normal controls were cultured in 96 microtitre plates for 6 days in the presence of OKT3 MoAb at final concentrations 25-250 ng/ml, IL-2 15 U/ml, Tg and TPO (1 micrograms/ml). Then cultures were pulsed with 0.2 microCi 3H-TdR/well and incorporation was measured after 18 h. IgG and anti-TPO/Tg-secreting cells were detected at 7 days. Higher proliferative responses from whole PBMC preparations in response to any of the combinations including OKT3 MoAb were observed in the HT preparations, while the basal values were the lowest. IL-2 alone increased these responses markedly, but equally in all groups. IL-2 in combination with OKT3 had an additive effect on proliferation, with higher responses in HT. Tg and TPO antigens did not change these responses. Most HT preparations responded with their maximum proliferation to the lowest concentration of OKT3 MoAb (25 ng/ml), whereas in GD and control preparations of PBMC these responses were shifted to higher concentrations (250 ng/ml); even with those, proliferation was not so enhanced in controls when compared with HT and GD preparations. In contrast, the proliferative responses of T cells alone and subpopulations of CD8+ suppressor/cytotoxic cells were decreased in HT preparations compared with controls. Monocytes were necessary for proliferation. In the subpopulation of B cells (> 95% pure) and CD4+ helper/inducer cells, differences did not reach significance. In spite of the effect on proliferation, OKT3 MoAb only mildly but significantly increased the numbers of IgG-secreting cells in HT and GD preparations and did not stimulate synthesis of specific antibodies. Our data suggest that the increased proliferative responses of whole PBMC to OKT3 MoAb in HT preparations might be due to insufficient activation of T suppressor/cytotoxic cells.