Suppression of the immune response by nasal immunization

Intranasal immunization results in both a mucosal and a systemic immune response in humans. Intranasal tetanus toxoid immunization in humans causes an increased serum IgA1 antibody response to tetanus toxoid following a subsequent intramuscular immunization. We hypothesized that intranasal priming with a novel protein antigen, keyhole limpet hemocyanin (KLH), would similarly result in an up-regulated systemic IgA response after a subsequent systemic immunization. To test this hypothesis, five healthy adults received a primary series of intranasal KLH immunizations followed 3 months later by a subcutaneous KLH immunizations Eight healthy adults received only a subcutaneous KLH immunization and served as controls. The nasal immunization resulted in a brisk and sustained serum IgM, IgA, and IgG antibody response and a mucosal IgA response. The subcutaneous immunization alone resulted in a serum antibody response and the development of delayed type hypersensitivity by skin testing. When the nasally primed subjects received a subsequent subcutaneous immunization there was a decline in the serum concentration of IgA and IgG antibodies to KLH. In addition, the nasally primed subjects failed to develop delayed type hypersensitivity to KLH following subcutaneous immunization. These data suggest that the nasal mucosa can induce a mucosal and systemic response; however, it may also suppress a subsequent immune response to systemic immunization.

Molecular characterization of protein Sir, a streptococcal cell surface protein that binds both immunoglobulin A and immunoglobulin G

Cell surface proteins that bind to the Fc part of immunoglobulin (Ig) A and/or IgG are expressed by many strains of the group A Streptococcus, an important human pathogen. Two extensively characterized proteins in this group of molecules are protein Arp that preferentially binds IgA and protein H that binds IgG. In addition, recent work has shown that many group A streptococcal strains express a novel type of Fc-binding protein, designated protein Sir, that binds both IgA and IgG. Protein Sir22, the molecule expressed by a strain of serotype M22, has now been purified and characterized after expression of the cloned gene in Escherichia coli. Dot-blot analysis with a large number of purified monoclonal Igs showed that protein Sir22 reacted with 19 out of 20 IgA proteins and with 19 out of 24 IgG proteins. The affinity constants for the reactions between protein Sir22 and Ig were determined to be 7.0 x 10(8) M-1 for serum IgA, 2.4 x 10(8 M-1 for secretory IgA, and 7.8 x 10(8) M-1 for IgG. Inhibition experiments showed that the bindings of IgA and IgG to protein Sir22 were mutually exclusive, indicating shared or contiguous binding sites. Analysis of the sequence of the sir22 gene indicated a gene product with 365 amino acid residues, including a 41-residue signal peptide. The processed form of the protein, 324 residues, has a calculated M(r) of 37,168. Deletion analysis of the sir22 gene showed that a 156-residue NH2-terminal fragment of protein Sir22 retained the ability to bind both IgA and IgG. The overall organization of protein Sir22 is similar to that of the IgA-binding protein Arp and the IgG-binding protein H. All three of these proteins are members of the M protein family and have a central repeat region of the C type.

Oral tolerance in humans. T cell but not B cell tolerance after antigen feeding

The purpose of this study was to investigate whether oral tolerance, defined as Ag-specific immunologic unresponsiveness after Ag feeding, could be induced in humans after prolonged Ag ingestion. Eight adult volunteers ingested a total dose of 0.5 g of keyhole limpet hemocyanin (KLH) followed by subcutaneous immunization with KLH. Eight controls received only the subcutaneous immunization. In the group fed KLH, there was a significant reduction in KLH-specific T cell proliferation (p = 0.04) and delayed skin test responses (p = 0.07) to KLH. KLH ingestion alone did not induce significant levels of Abs in either serum or secretions. However, after the subsequent subcutaneous immunization, the number of circulating IgG and IgM anti-KLH-producing cells, the titers of serum IgG, IgA, and IgM anti-KLH Abs, and the titers of IgA anti-KLH Abs in saliva and intestinal secretions were significantly greater in the KLH-fed group than in the nonfed group. We conclude that KLH feeding induced systemic T cell tolerance, but B cell priming, at both systemic and mucosal sites. These studies support the concept of using Ag feeding as a treatment for certain immune-mediated diseases.

Oral tolerance in humans. T cell but not B cell tolerance after antigen feeding

The purpose of this study was to investigate whether oral tolerance, defined as Ag-specific immunologic unresponsiveness after Ag feeding, could be induced in humans after prolonged Ag ingestion. Eight adult volunteers ingested a total dose of 0.5 g of keyhole limpet hemocyanin (KLH) followed by subcutaneous immunization with KLH. Eight controls received only the subcutaneous immunization. In the group fed KLH, there was a significant reduction in KLH-specific T cell proliferation (p = 0.04) and delayed skin test responses (p = 0.07) to KLH. KLH ingestion alone did not induce significant levels of Abs in either serum or secretions. However, after the subsequent subcutaneous immunization, the number of circulating IgG and IgM anti-KLH-producing cells, the titers of serum IgG, IgA, and IgM anti-KLH Abs, and the titers of IgA anti-KLH Abs in saliva and intestinal secretions were significantly greater in the KLH-fed group than in the nonfed group. We conclude that KLH feeding induced systemic T cell tolerance, but B cell priming, at both systemic and mucosal sites. These studies support the concept of using Ag feeding as a treatment for certain immune-mediated diseases.

Intraperitoneal immunization of human subjects with tetanus toxoid induces specific antibody-secreting cells in the peritoneal cavity and in the circulation, but fails to elicit a secretory IgA response

Five patients on continuous ambulatory peritoneal dialysis (CAPD) were immunized intraperitoneally with tetanus toxoid (TT) through an indwelling catheter. Four control patients on CAPD received the same dose of TT intramuscularly. Before immunization, virtually no anti-TT antibody-secreting cells (AbSC) were detected by the enzyme-linked immunospot (ELISPOT) assay in peripheral blood or peritoneal fluid from patients of either group. One to 2 weeks after immunization, high frequencies of TT-specific AbSC were detected in the circulation and peritoneal cavity. More than 80% of those cells were of the IgG isotype, with IgA accounting for most of the remainder. Patients receiving TT by the i.p. route showed significantly higher frequencies of specific IgG and IgA AbSC in the peritoneal cavity than patients immunized intramuscularly. Frequencies of AbSC in peripheral blood did not significantly differ between the two groups. Immunization with TT by both routes resulted in a significant increase of IgG anti-TT antibodies in serum, saliva and peritoneal fluid. A significant IgA antibody response was seen only in serum and peritoneal effluents. Therefore, i.p. immunization of human subjects with TT elicited both a localized response in the peritoneal cavity as well as a systemic response in serum, but did not induce a salivary IgA response.

Antisperm antibodies (ASAs) in infertile males: subclass distribution of IgA antibodies and the effect of an IgA1 protease on sperm-bound antibodies

PROBLEM

(1) To determine the IgA subclass distribution of antibodies in the serum and on the sperm of infertile male patients. (2) To determine the effect of an IgA1 protease on the binding of IgA antisperm antibodies (ASA).

METHOD

Fifteen infertile males with ASA in serum (10) or on sperm (5) were recruited for this study. Duration of infertility was at least one year. Monoclonal antibodies to human IgA1 and IgA2 were conjugated to immunobeads. The distribution of IgA1 and IgA2 subclass ASA was determined for positive serum and sperm samples. The effects of an IgA1 protease (isolated from Neisseria meningitidis strain HF13) on sperm-bound antibodies was evaluated by immunobead binding.

RESULTS

In serum, the IgA1 subclass ASA was predominate (91%) when compared with IgA2 (9%) subclass. Direct sperm-bound antibodies displayed a distribution more characteristic of the secretory immune system with IgA1 accounting for 63% and IgA2 accounting for 37% of the total IgA ASA. Enzyme treatment dramatically reduced the amount of serum IgA antibodies bound to sperm (P < 0.05). Similarly, a significant reduction in direct sperm-bound antibodies was observed after enzymatic treatment with no loss in sperm motility.

CONCLUSIONS

(1) Although IgA1 direct sperm-bound antibodies were dominant, when compared with serum there was a higher proportion of IgA2 subclass, which suggests a local production of IgA. (2) Specific IgA1 protease treatment is capable of reducing the amount of immunobead-detectable IgA on sperm. Hamster oocyte sperm-penetration assays are ongoing to determine if this treatment might improve sperm penetration rates.

Lectin receptors on IgA isotypes

It has been shown previously that secretory IgA interacts with the mannose-specific lectin of Escherichia coli. The purpose of the study described here was to evaluate whether the N-linked oligosaccharide chains of the human IgA isotypes IgA1 and IgA2 differ in lectin receptor activity. A range of plant lectins specific for N-linked oligosaccharide chains were tested for their ability to precipitate IgA1 and IgA2 myeloma proteins, secretory IgA and free secretory component. IgA2 myeloma proteins reacted more strongly than IgA1 with the mannose-specific lectin ConA, whereas IgA1 myeloma proteins reacted more strongly than IgA2 with two galactose-specific lectins, Ricinus communis agglutinin I and Abrus precatorius agglutinin. This suggests that IgA2 possesses a larger proportion of short truncated complex type oligosaccharide chains and/or oligomannose type chains than IgA1. Further, IgA2 reacted more strongly than IgA1 myeloma proteins with Lens culinaris (lentil) lectin, and Pisum sativum (pea) lectin, suggesting that IgA2 exposes more of short, complex type chains fucosylated on the core than IgA1. The differences demonstrated in receptor activity between IgA1 and IgA2 may be important in their interaction with the microbial flora, as well with endogenous lectins, such as phagocyte receptors.

Secretory immunity in the female reproductive tract

PROBLEM

Antibodies and antibody-producing cells display a different and characteristic distribution in body fluids and tissues.

METHOD

We have investigated the tissues of the female reproductive tract to determine whether the distribution of immunoglobulin-producing cells and the contents of cervical secretions were similar to those found in tissues of the secretory immune system.

RESULTS

Immunohistochemical examinations of female genital tissues revealed the presence of plasma cells that secrete IgA (and in lower numbers IgM and IgG) especially in the subepithelial layers of the uterine endo- and ectocervix, fallopian tubes, and vagina. Both IgA1- and IgA2-producing plasma cells were found in approximately equal proportions. The presence of J-chain in the IgA-secreting cells suggests the synthesis of polymeric IgA (pIgA). Epithelial cells lining the fallopian tube and endocervix were positive for secretory component (SC), which is required for the transepithelial transport of pIgA into external secretions. Cervical mucus was collected and the molecular forms of IgA were separated using column chromatography. Approximately 80% of IgA in cervical mucus was polymeric compared with 55% in the vaginal fluid.

CONCLUSIONS

These data indicate that all effector components of the mucosal immune system are present in the female reproductive tract. The immunization routes that lead to a secretory IgA (S-IgA) response need to be further explored.

Early expression of human J chain and mu chain gene in the fetal liver

The expression of J chain and mu chain genes has been investigated during human fetal ontogeny by the polymerase chain reaction performed on liver sections. With this technique, we have been able to detect expression of J chain and mu chain in a single 6-10-microns-thick formalin-fixed paraffin-embedded section of fetal liver tissue. J chain expression reached a maximum at 16 weeks of gestation but was clearly detectable in the liver at the 6th week. Although not detectable at the 6th week, mu chain mRNA became readily detectable at 7 weeks of gestation. These results indicate that J chain expression precedes that of mu chain by at least one week. Therefore, our results imply that J chain is the first immunoglobulin-related polypeptide expressed during the embryogenesis and differentiation of B cells in the fetal liver.

Surface expression of secretory component and HLA class II DR antigen on glandular epithelial cells from human endometrium and two endometrial adenocarcinoma cell lines

The expression of secretory component (SC) by human glandular endometrial cells cultured in vitro was significantly increased by estradiol in the medium. Interferon-gamma and interleukin-4 stimulated the expression of SC only in the presence of estrogen. Tumor necrosis factor-alpha plus estrogen also caused a significant increase in the number of cells expressing SC. HLA class II antigen DR was detected on few glandular epithelial cells of human endometrium cultured in control medium, whereas interferon-gamma and interleukin-4, but not tumor necrosis factor-alpha, caused significant increases in the expression of DR. Estrogen in the culture medium did not significantly affect DR expression. The human endometrial adenocarcinoma cell lines, HEC and RL-95, expressed SC in approximately 50 and 20% of the cells. Also, approximately 20% of the RL-95 cells stained for DR antigen. Interferon-gamma did not influence the degree of expression of either surface marker of the two cell lines. Cells of both lines bound polymeric IgA and IgM but showed little to no binding of monomeric IgA, IgG, or an IgM previously shown not to bind SC.

Immune response of the female rat genital tract after oral and local immunization with keyhole limpet hemocyanin conjugated to the cholera toxin B subunit

The immune response of the female rat genital tract was evaluated with Lewis rats given primary and secondary immunizations with keyhole limpet hemocyanin (KLH) alone or coupled to the cholera toxin (CT) B subunit (CTB) by the oral or intravaginal-uterine route or a combination of routes. CT (2 to 5 micrograms) was administered as an adjuvant with the KLH-CTB conjugate. While a significant mucosal immunoglobulin A (IgA) response was induced by KLH, there were no significant differences among the immunized groups in the levels of IgA antibodies in salivary gland, gut, vaginal, and uterine secretions, with the exception that rats immunized only orally with the KLH-CTB conjugate lacked a detectable vaginal response. Levels of IgA antibodies to CT, however, were significantly increased in genital tract secretions of rats immunized locally versus orally with the KLH-CTB conjugate. Antibody activity of the IgG isotype against both KLH and CT was significantly elevated in genital tract secretions of rats immunized with KLH-CTB by the oral or intravaginal-uterine route and given genital tract boosters, in comparison with the results for the other groups. IgM antibody titers were generally negligible in the different secretions. An enzyme-linked spot-forming assay revealed IgA and IgG antibody-secreting cells in salivary gland and uterine tissues. A highly significant correlation between the numbers of antibody-secreting cells and antibody titers existed for uterine IgG but not IgA responses to KLH among the different groups of rats. In conclusion, a vigorous local immune response was induced after immunization of the female rat reproductive tract alone or in combination with peroral challenge with the KLH-CTB conjugate.

Cytokines and periodontal disease: immunopathological role of interleukins for B cell responses in chronic inflamed gingival tissues

Localized and chronically-inflamed gingival tissues of adult periodontitis (AP) are generally characterized as a hyper-responsiveness of B lineage cells where increased numbers of plasma cells occur. It was previously shown that high numbers of IgG subclass antibody-secreting cells (e.g., IgG1 > IgG2 > IgG3 > or = IgG4) with significant numbers of IgA subclass antibody-producing cells were seen in enzymatically dissociated gingival mononuclear cells (GMC) from inflamed periodontal tissues. An interesting finding was that the frequency of IgA2 plasma cells was elevated in the severe stage of AP when compared with the moderate stage. IgM plasma cells were essentially not found in these tissues. To understand the cytokine involvement in these increased B cell responses in inflamed gingiva, GMC isolated from inflamed tissues of AP patients were examined for cytokine production, specifically for IL-2, IL-4, IL-5, and IL-6 at both the protein and mRNA levels, since these cytokines have been shown to be essential interleukins for the regulation of the B cell response. Freshly-isolated GMC and peripheral blood mononuclear cells (PBMC) from AP patients were initially examined for IL-6 production because of its essential role for the terminal differentiation of B cells to become Ig-producing plasma cells. High levels of IL-6 were produced by GMC but not by PBMC unless cells were stimulated with T cell mitogen. A similar findings was also obtained when levels of IL-6 specific mRNA were examined in GMC and PBMC.(ABSTRACT TRUNCATED AT 250 WORDS)

Serum IgG and IgM levels to bacterial antigens in necrotizing ulcerative gingivitis

Previous studies of the systemic antibody response in necrotizing ulcerative gingivitis (NUG) have elicited varying results. The purpose of this study was to determine the humoral response to site-specific isolates of microbiota associated with NUG. Sera from 21 active NUG subjects and 21 age-sex-race matched controls were assessed for IgG and IgM antibodies to 4 clinical isolates of Prevotella intermedia and 3 clinical isolates of Treponema species. P. intermedia and Treponema strains were isolated from active and inactive sites of NUG patients and gingivitis sites of controls. P. intermedia was also isolated from noninflamed sites of the controls. IgG and IgM serum levels to these 7 bacteria were measured by ELISA. Compared to control subjects, the NUG sera exhibited significantly lower IgG and IgM levels to all 4 isolates of P. intermedia (P < or = 0.001). It was also noted that sera from NUG subjects had elevated IgM levels to all 3 spirochete isolates but significantly higher only to the spirochete isolated from a gingivitis site of a control subject (P < or = 0.005). The data suggest that failure to mount a substantial antibody response to P. intermedia may be associated with onset of disease activity in NUG. However, the assumed lack of "biological significance" between differences in antibody responses measured indicates the relationship is weak or nonexistent. It also appears that antibody response to spirochetes is not associated with onset of NUG.

Interactions of cell-surface galactosyltransferase with immunoglobulins

Detection of the activity of beta-1,4-galactosyltransferase (beta-1,4-GT) in suspensions of viable mouse hepatocytes, the human hepatoma cell line Hep G2, the human colonic adenocarcinoma cell line HT-29, the monocyte-like cell line U937, and human splenic B and T lymphocytes suggested the presence of beta-1,4-GT, in an enzymatically active form, on plasma membranes. The presence of beta-1,4-GT on cell surfaces was also indicated from the effect of trypsinization of live cells, which significantly reduced cell surface beta-1,4-GT activity, but did not affect the activity associated with cytoplasmic membranes. Furthermore, the presence of beta-1,4-GT on the cell surface was demonstrated by indirect immunofluorescence staining of cells with anti-beta-1,4-GT antibody. The detection of radioactivity in immunoglobulins (Ig) and their component chains after incubation with suspensions of intact cells in the presence of Mn2+ and UDP-[3H]-galactose, indicated that Ig molecules were galactosylated. In the absence of UDP-[3H]-galactose, beta-1,4-GT on cell surfaces, or immobilized on Sepharose-4B, formed stable complexes with galactose acceptors, including Ig. The efficiency of binding decreased in the order: J chain > alpha chain > mu chain > polymeric IgA2 > monomeric/polymeric IgA1 > IgM > IgG. Thus, beta-1,4-GT could act as a cell-surface receptor for Ig through a cation-dependent, lectin-like association of the beta-1,4-GT with the carbohydrate moieties of the Ig. This was confirmed by indirect surface immunofluorescence and radiolabeled ligand binding assays. The binding was inhibitable by EDTA, alpha-lactalbumin (in the presence of glucose), GlcNAc, or uridine 3',5'dialdehyde. At 37 degrees C, the apparent affinity constants and association rate constants of interaction between cell surface beta-1,4-GT on glutaraldehyde-fixed HT-29 and U937 cells and alpha 2 chain or monomeric IgA1 were in the range from 7.1 x 10(7) to 4.6 x 10(8) M-1 and from 1 x 10(5) to 3 x 10(6) M-1 s-1, respectively. The dissociation rate constants and half time of dissociation calculated from these data were in the range from 2.1 x 10(-2) to 5.0 x 10(-4) s-1 and from 33 to 1380 s, respectively. The number of alpha 2 or IgA1 molecules bound per HT-29 and U937 cell were in the range from 1.9 x 10(5) to 1.3 x 10(6). The binding of IgA by the cell surface beta-1,4-GT was not associated with internalization or the catabolic degradation of the ligand.

Interactions of galactosyltransferase with serum and secretory immunoglobulins and their component chains

Assay of the activity of beta-1,4-galactosyltransferase (beta-1,4-GT) revealed that in addition to serum, milk, colostrum, amniotic and cerebrospinal fluids and malignant effusions, this enzyme is present also in tears and saliva. Molecular-sieve chromatography of human colostral whey and serum and subsequent assay of beta-1,4-GT activity have shown that beta-1,4-GT was present as a free enzyme (55 kDa) and associated with components of larger molar mass. The elution pattern did not change when the chromatography was carried out in a buffer devoid of, or enriched with, Mn2+, a cofactor of beta-1,4-GT activity. However, the activity associated with the large molar mass components was absent when the chromatography was carried out in the presence of a chelating agent (EDTA). Analyses of the eluted material by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate (SDS-PAGE), and by immunodiffusion indicated that the major colostral component in beta-1,4-GT activity-containing fractions was secretory IgA (S-IgA); in addition, the beta-1,4-GT activity was detected in fractions that contained lactoferrin and alpha-lactalbumin. Interactions of beta-1,4-GT with S-IgA and lactoferrin in colostrum were also demonstrated by the detection of radioactivity in precipitin lines obtained by immunoelectrophoresis and autoradiography of the colostral whey after it had been incubated with UDP-[3H]-galactose. Furthermore, radioactively labeled S-IgA and alpha-chain were detected when colostral whey incubated with UDP-[3H]-galactose was analyzed by SDS-PAGE under non-reducing and reducing conditions, respectively. In serum, the beta-1,4-GT-binding components identified in fractions after molecular-sieve chromatography were IgG, IgA, IgM and transferrin. The binding of beta-1,4-GT to immunoglobulins (Ig) was also demonstrated by assaying the beta-1,4-GT activity associated with Sepharose-4B-immobilized Ig of various isotypes and molecular forms, which were incubated with colostral beta-1,4-GT in the presence of Mn2+. Beta-1,4-GT measured by enzyme activity was bound to these Ig in order: polymeric IgA2 > monomeric IgA1 = polymeric IgA1 = secretory IgA = pentameric IgM > IgG. Immobilized component chains, namely alpha, mu and J chains, bound beta-1,4-GT more effectively than native Ig. Incubation of the IgA1 myeloma protein with crude human colostral galactosyltransferase in the presence of UDP[3H]-galactose and Mn2+ resulted in galactosylation of both N- and O-linked carbohydrate side chains.(ABSTRACT TRUNCATED AT 400 WORDS)

Murine model for evaluation of protective immunity to influenza virus

We have characterized a murine model as an inexpensive, readily available and sensitive animal model for the evaluation of protective immune responses induced by various routes of administration of influenza A vaccine preparations. Using a non-mouse-adapted human influenza virus to infect unanaesthetized animals intranasally, we established that the optimum dose for infection of Balb/c mice was 10(4) plaque forming units of virus and that the optimum sampling time for measurement of virus yields in the organs of the respiratory tract was 72 h after challenge. We found that the infection was initiated in the nose and progressed by descending into the trachea and lungs over a period of days. Evaluation of protection against infection clearly showed that the tissues of the mouse respiratory tract were completely protected after administration of whole killed virus intranasally and partly protected when virus was administered subcutaneously. The protection correlated with the level of virus-specific IgA antibodies in saliva.

Oral immunization with influenza virus in biodegradable microspheres

Polymeric microspheres were evaluated as an oral antigen delivery system for immunization with influenza virus. The immune responses obtained were compared after either oral or systemic immunization of BALB/c mice using purified, formalin-inactivated influenza virus type A/H3N2, either encapsulated in biodegradable and biocompatible microspheres or free in solution. The immunogenicity of formalin-treated influenza vaccine was preserved during the microencapsulation process, and the microencapsulated antigen induced protective immune responses after systemic immunization that were equal to or higher than those induced by conventional vaccine. When administered orally to primed animals, microencapsulated antigen induced levels of anti-influenza antibodies in saliva that were higher than and in serum that were comparable to those obtained by systemic immunization. Furthermore, oral booster immunization provided virtually complete protection of animals challenged with live virus.

Cytokine-induced immunoglobulin production in primary IgA nephropathy

Increased IgA synthesis probably plays a role in the pathogenesis of IgA nephropathy (IgAN). We investigated whether an increased sensitivity to the effect of various growth factor combinations leads to increased immunoglobulin synthesis by peripheral blood mononuclear cells (PBMC) from IgAN patients, in comparison to healthy controls. Although none of the growth factors studied (pokeweed mitogen [PWM], interleukin [IL]-2, IL-6, transforming growth factor-beta [TGF-beta], and combinations) led to greater IgA synthesis in IgAN patients than in controls, the IgA subclass ratio was shifted in favor of IgA1. In controls, but not in IgAN patients, IL-2 enhanced the production of IgA and IgA1 compared with media alone. This possibly reflects previous in vivo activation by IL-2 in IgAN patients. The suppressive effect of TGF-beta on immunoglobulin synthesis was modestly greater in IgAN patients than in controls. Increased production of IL-2 and perhaps other cytokines by T cells in vivo may be responsible for the elevated IgA immune response in these patients.

Transforming growth factor-beta inhibits the production of IgG, IgM, and IgA in human lymphocyte cultures

Transforming growth factor beta (TGF beta) has potent immunoregulatory effects acting on both T and B cells. It strongly inhibits secretion of IgG and IgM in human and murine B cell cultures, but has been shown to have an enhancing effect on IgA production in the mouse. We have studied the effect of TGF beta on the production of IgA in human lymphocyte cultures. The addition of TGF beta to pokeweed-stimulated peripheral blood lymphocytes resulted in a suppression of IgA production of both subclasses, similar in magnitude to the suppression of IgG and IgM production. Membrane IgA expression was not increased by culturing tonsillar lymphocytes with TGF beta. In conclusion, we find no evidence for a selective enhancing effect of TGF beta on IgA synthesis in humans, in contrast to the findings reported in mice.

IgA-containing immune complexes after challenge with food antigens in patients with IgA nephropathy

The possibility that patients with IgA nephropathy (IgAN) might have abnormal IgA immune responses to immunogens commonly encountered at mucosal surfaces, resulting in the formation of circulating immune complexes (CIC), was examined. Since it is generally held that such increased IgA responses are characterized by detectable aberrancies in handling of IgA-containing CIC, IgAN patients and controls were given a large volume of bovine milk (after dietary deprivation of bovine antigens) and immune complex levels were measured over a period of 12 h. An assay based on binding of CIC containing C3 to solid-phase anti-C3 and subsequent development with isotype-specific antibody revealed no differences in responses of patients and controls with respect to IgG- and IgM-containing CIC. Although IgAN patients tended to have higher levels of IgA-containing CIC, there were no differences in response patterns when IgA CIC levels after ingestion of the milk stimulus were related to baseline levels. Polymorphonuclear leucocytes (PMNC), which bear surface receptors for IgA, were isolated from some subjects at the same times as the samples for CIC levels and examined by two-colour immunofluorescence for the coincident presence of IgA and milk antigens. In contrast to the data obtained in the CIC assays, these experiments revealed the simultaneous presence of IgA and two of three milk proteins in PMNC of IgAN patients but not controls. Follow-up experiments designed to assess more quantitatively the coincidental presence of IgA and milk antigens indicated no significant differences between patients and controls. However, milk proteins seemed to be more commonly associated with IgA in PMNC of IgAN patients, suggesting the presence of non-complement-fixing IgA/antigen CIC after mucosal challenge of some IgAN patients.