T cell cytokine polarity as a determinant of immunoglobulin A (IgA) glycosylation and the severity of experimental IgA nephropathy.

Immunoglobulin A (IgA) glycosylation, recognized as an important pathogenic factor in IgA nephropathy (IgAN), is apparently controlled by the polarity of T helper (Th) cytokine responses. To examine the role of cytokine polarity in IgAN, inbred mice were immunized by intraperitoneal priming with inactivated Sendai virus (SeV) emulsified in either complete Freund's adjuvant (CFA) or incomplete Freund's adjuvant (IFA), which promote Th1- or Th2-immune response, respectively, and then boosted identically twice orally with aqueous suspensions of inactivated virus. Next, some mice were challenged intranasally with infectious SeV. Mice primed with CFA or IFA had equal reductions in nasal viral titre relative to non-immune controls, and equally increased serum levels of SeV-specific IgA antibody. Mice primed with CFA showed higher SeV-specific IgG than those with IFA. Splenocytes from mice primed with IFA produced copious amounts of interleukin (IL)-4 and IL-5, but little interferon-gamma and IL-2; those primed with CFA had reciprocal cytokine recall responses. Total serum IgA and especially SeV-specific IgA from mice primed with IFA showed a selective defect in sialylation and galactosylation. Although the frequency and intensity of glomerular deposits and haematuria did not differ, glomerulonephritis in mice primed with IFA and challenged with infectious virus was more severe than in those given CFA, as judged by serum creatinine level. We conclude that the polarity of T cell cytokines controls the pattern of IgA glycosylation and exerts direct or indirect effects on functional glomerular responses to immune complex deposition.

The role of nasal tolerance in a model of IgA nephropathy induced in mice by Sendai virus.

Mucosal antigenic exposure is implicated in pathogenesis of IgA nephropathy. Although IgG and/or IgM codeposits may promote disease, protracted mucosal antigenic exposure reduces IgG and IgM antibody, a process termed mucosal tolerance. We immunized mice intranasally with infectious or inactivated Sendai virus for 6 or 14 weeks. Anti-virus IgG remained high in mice given infectious virus for 14 weeks, but decreased after 6 weeks in mice given inactivated virus; IgA antibody remained high in both groups. Upon viral challenge, glomerular IgG and complement deposits and the frequency of hematuria, all equal after 6 weeks of immunization, were lower in mice immunized with inactivated virus for 14 weeks but remained high in mice given infectious virus; glomerular IgA increased over time in both immunized groups. Viremia in a non-tolerized immune host can promote glomerulonephritis with IgG and complement codeposits and glomerular dysfunction. These preliminary experiments may guide future, more mechanistic, investigation.

A mucosal IgA-mediated excretory immune system in vivo.

The capacity of mucosal IgA Abs to serve as an excretory immune system in vivo was investigated. Mice expressing a transgenic TCR were immunized intragastrically with the cognate Ag to elicit a vigorous mucosal IgA Ab response. Soon after i.v. challenge, Ag was detected within the epithelial cells of the small intestinal crypts and to a lesser degree within the epithelial cells higher up the villi, paralleling the gradient in expression of the polymeric Ig receptor and the transport of its ligand, oligomeric IgA. Uptake of Ag into the epithelial cells occurred only from the basolateral aspect and only when Ag complexed to IgA Ab could be present in the lamina propria. The results support the concept that local IgA Abs can excrete Ags from the body by transporting them directly through mucosal epithelial cells, using the same mechanism that transports free IgA into the mucosal secretions.

The polymeric immunoglobulin receptor translocates pneumococci across human nasopharyngeal epithelial cells.

The polymeric immunoglobulin receptor (pIgR) plays a crucial role in mucosal immunity against microbial infection by transporting polymeric immunoglobulins (pIg) across the mucosal epithelium. We report here that the human pIgR (hpIgR) can bind to a major pneumococcal adhesin, CbpA. Expression of hpIgR in human nasopharyngeal cells and MDCK cells greatly enhanced pneumococcal adherence and invasion. The hpIgR-mediated bacterial adherence and invasion were abolished by either insertional knockout of cbpA or antibodies against either hpIgR or CbpA. In contrast, rabbit pIgR (rpIgR) did not bind to CbpA and its expression in MDCK cells did not enhance pneumococcal adherence and invasion. These results suggest that pneumococci are a novel example of a pathogen co-opting the pIg transcytosis machinery to promote translocation across a mucosal barrier.

Replication and budding of simian immunodeficiency virus in polarized epithelial cells.

Simian immunodeficiency virus (SIV) infection of primates provides an important model for infection of humans by HIV. Since mucosal epithelium is likely to be an important portal of entry, we decided to study aspects of the interaction of SIV with epithelial cells. SIV was shown to produce virus efficiently in polarized epithelial cells (Vero C1008) transfected with SIVmac239 proviral DNA. The virus titer in the epithelial cell culture fluid reached 10(3) TCID50/ml at day 3 posttransfection. Initially after transfected epithelial cells were plated on a permeable membrane, virus budded at both the apical and the basolateral domains. However, after the cells formed a tight monolayer, 95-100% of the virus particles budded basolaterally, as assessed by release of p27 antigen into the fluid above and below the monolayer. This finding was confirmed by electron microscopy, which showed that the mature virus budded basolaterally in polarized cells. After introduction of the CD4 gene into Vero cells by a retrovirus vector, polarizable cells were able to be infected by cell-free SIVmac239 virus. The virus titer reached 10(4) TCID50/ml in culture fluid and virions also budded basolaterally, the same as the virus from transfected cells. Two viruses (SIVmac1A11 and SIVmac251) that contain truncated TMgp28 instead of TMgp41 also budded basolaterally. Furthermore, we found that HIV-1 with full-length or truncated TMgp41 also budded basolaterally.

Current concepts in mucosal immunity. IV. How epithelial transport of IgA antibodies relates to host defense.

The humoral arm of the mucosal immune system is principally composed of locally synthesized polymeric IgA, whose Fc portion is adapted for binding to the polymeric immunoglobulin receptor that is expressed on the basolateral surface of mucosal epithelial cells, including enterocytes. This receptor mediates the endocytosis and transcytosis of polymeric IgA, which allows IgA to function in host defense at three anatomic levels in relation to mucosal epithelium: IgA antibodies in the lamina propria can bind antigens and excrete them through the epithelium into the lumen; antiviral IgA antibodies in transit through epithelial cells can inhibit virus production by an intracellular action; and IgA antibodies secreted into the lumen can prevent antigens and microbes from adhering to and penetrating the epithelium. The ways in which IgA antibodies function in mucous membranes provide challenging investigative opportunities for cell physiologists and cell biologists.

Interaction of antigens and antibodies at mucosal surfaces.

Infections often involve the mucosal surfaces of the body, which form a boundary with the outside world. This review focuses on immunoglobulin A (IgA) antibodies because IgA is the principal mucosal antibody class. IgA is synthesized by local plasma cells and has a specific polymeric immunoglobulin receptor-mediated transport mechanism for entry into the secretions. By serving as an external barrier capable of inhibiting attachment of microbes to the luminal surface of the mucosal epithelial lining, IgA antibodies form the first line of immune defense. In addition to this traditional mode of extracellular antibody function, recent evidence suggests that IgA antibodies can also function in a nontraditional fashion by neutralizing viruses intracellularly, if a virus is infecting an epithelial cell through which specific IgA antibody is passing on its way to the secretions. IgA antibodies are also envisaged as providing an internal mucosal barrier beneath the mucosal lining. Antigens intercepted by IgA antibodies here can potentially be ferried through the epithelium and thereby excreted. In addition to the polymeric immunoglobulin receptor on mucosal epithelial cells, IgA antibodies can bind to receptors on a variety of leukocytes and have been shown, in some experimental systems, to be capable of activating the alternative complement pathway, making IgA antibodies potential participants in inflammatory reactions. Although the relationship of IgA antibodies to inflammation is not entirely clear, the bias presented is that IgA is basically noninflammatory, perhaps even anti-inflammatory. According to this view, the major role of the Fc portion of IgA antibodies is to transport IgA across mucosal epithelial cells and not, as in the case of the other classes of antibody, to activate secondary phenomena of the kind that contribute to inflammation. Because of IgA's key role as an initial barrier to infection, much current research in mucosal immunology is directed toward developing new vectors and adjuvants that can provide improved approaches to mucosal vaccines. Finally, because of advances in monoclonal antibody technology, topical application of antibodies to mucosal surfaces has significant potential for preventing and treating infections.

Review article: Epithelial disposition of antigen.

Exogenous antigens can impinge upon the luminal surface of mucous membranes and to a limited degree, can even penetrate into the lamina propria. As a result of infection, microbial antigens can be present in mucosal epithelium or lamina propria. At all three levels (lumen, epithelium, lamina propria), locally produced antibodies, principally immunoglobulin A (IgA), can combine with antigens. As a result, antigens can be prevented from attaching to the epithelium, viruses can be neutralized inside epithelial cells and immune complexes in the lamina propria can be excreted into the lumen. As the Fc portion of IgA is specialized to facilitate epithelial transport of IgA rather than to activate inflammatory mediator systems (as is the case with other classes of immunoglobulin), the predominance of IgA among mucosal antibodies serves to limit the degree of local inflammation following antigen-antibody reactions.

Hepatic pathology resulting from mouse hepatitis virus S infection in severe combined immunodeficiency mice.

Mouse hepatitis virus (MHV) is a pervasive pathogen that causes morbidity and mortality in mouse colonies worldwide. Although it is not a major cause of mortality in immunocompetent mice, infections from MHV strains of lower virulence can be fatal to athymic nude mice. The histopathologic features and alterations of serum biochemical parameters resulting from infection with a low-virulence MHV strain in severe combined immunodeficiency (scid) mice has not been well described. Thus we recently studied the disease caused by MHV-S in scid mice after intranasal inoculation. Mouse hepatitis virus infection in scid mice, which have severe defects of B and T cells, may be highly lethal, resulting in immediate mortality. However, our results indicate that scid mice survived for an average of 12 to 14 days after infection with doses of MHV up to 10(7) PFU/mouse. The virus caused a significant increase in serum enzyme activities and bilirubin concentration associated with histologically demonstrable hepatocellular injury at postinoculation days 3, 4, and 8. Furthermore, virus was detected in mouse liver homogenates and nasal and bronchial lavage specimens. These results provide valuable information regarding the histopathologic and biochemical consequences of MHV-S infection in scid mice.

Comparative studies of transcytosis and assembly of secretory IgA in Madin-Darby canine kidney cells expressing human polymeric Ig receptor.

Epithelial transport of polymeric IgA (pIgA) from its site of synthesis to the mucosal lumen is mediated by the polymeric Ig receptor (pIgR). During transcytosis, a disulfide bond forms between pIgR and pIgA, resulting in secretion of a covalently linked complex. To dissect further the intracellular processing and functions of pIgR, we have expressed the entire coding sequence of human pIgR cDNA in Madin-Darby canine kidney (MDCK) cells. Cloned transfected cells express human pIgR, as detected by immunofluorescence and by quantification of the cleaved extracellular domain of pIgR in culture supernatants. The function of transfected pIgR was confirmed by measuring vectorial transcytosis of 125I-labeled pIgA and its disulfide bonding to pIgR. Species specificity of transcytosis was determined by comparing transport of human, rat, and mouse pIgA in MDCK cells expressing either human or rabbit pIgR. pIgA from all three species was transported by both human and rabbit pIgR, with rat pIgA being transported to the greatest extent in each case. However, disulfide bonding was observed only with human pIgR, and was found to occur mainly inside the cell. Our results suggest that conformational differences between human and rabbit pIgR may account for differences in disulfide bonding to pIgA, and show that efficient transcytosis of pIgA is correlated better with noncovalent than covalent binding to pIgR.

IgA and mucosal defense.

The traditional role of IgA antibodies in mucosal defense has been considered as providing an immune barrier to keep exogenous substances, including microbial pathogens, from penetrating the mucosa. In this way infections can be prevented. More recently, studies in vitro and in vivo are providing evidence to suggest that IgA may have additional roles in mucosal defense. For example, during their passage through the lining epithelial cells of mucous membranes en route to the secretions, IgA antibodies may have an opportunity to neutralize intracellular pathogens like viruses. Also, IgA antibodies in the mucosal lamina propria have opportunities to complex with antigens and excrete them through the adjacent mucosal epithelium, again by the same route to the secretions that is taken by free IgA. These latter functions could aid in recovery from infection.

Disulfide bond formation between dimeric immunoglobulin A and the polymeric immunoglobulin receptor during hepatic transcytosis.

The polymeric immunoglobulin receptor on rat hepatocytes binds dimeric IgA on the sinusoidal surface and mediates its transport to the canaliculus, where the complex of dimeric IgA and secretory component, the cleaved extracellular domain of polymeric immunoglobulin receptor, is secreted into bile. This process is unique in that disulfide bonds are formed between dimeric IgA and polymeric immunoglobulin receptor during transcytosis, permanently preventing their dissociation. Here we present three lines of evidence that disulfide bonding between dimeric IgA and polymeric immunoglobulin receptor occurs predominantly in a late transcytotic compartment and that hepatic transcytosis can proceed in the absence of disulfide bond formation. First, throughout the course of transcytosis the percentage of intracellular dimeric IgA disulfide bonded to polymeric immunoglobulin receptor is less than half that in bile, suggesting that disulfide bond formation is a late event in transcytosis. Second, dimeric IgA that recycles from early endocytotic compartments into the circulation is mostly noncovalently bound to secretory component. Finally, the rate of transcytosis of dimeric IgA and its appearance in bile are not affected when disulfide bond formation with polymeric immunoglobulin receptor is inhibited by blocking of free thiol groups on dimeric IgA with iodoacetamide. These results are consistent with other findings in the literature and indicate that the main physiological role of disulfide bond formation between dimeric IgA and polymeric immunoglobulin receptor is not to facilitate transcytosis but, rather, to stabilize the dimeric IgA-secretory component complex after its release into external secretions such as bile and intestinal secretions.

Experimental IgA nephropathy secondary to hepatocellular injury induced by dietary deficiencies and heavy alcohol intake.

BACKGROUND: In humans, alcoholic liver disease is frequently associated with IgA mesangial deposits, microscopic hematuria and a small amount of proteinuria, identifying a secondary form of IgA nephropathy. Alcoholic liver disease is almost always associated with nutritional deficiencies. EXPERIMENTAL DESIGN: In order to examine the relationship between alcohol intake and/or inadequate diet and IgA nephropathy, groups of 4 week-old-male Lewis rats were maintained on a lipotrope-deficient (LD) diet (N = 20), intragastric infusions of a commercial whiskey (1.5 ml/100 gm body weight) three times a week, and regular chow (N = 23) or both intragastric whiskey infusion and an LD diet (N = 17). A fourth control group (N = 19) was given no whiskey and normal chow. RESULTS: All rats given the LD diet had marked steatosis and elevated "liver" enzymes. Changes were more severe, and with early bridging fibrosis and nodule formation in those also given whiskey, associated with increased hepatic content of mRNA encoding transforming growth factor-beta. A moderate steatosis without alteration in serum enzymes or transforming growth factor-beta expression was found in rats given whiskey (all p < 0.0001) compared with controls. IgA accumulated in hepatic sinusoids instead of in canaliculi and bile ducts, suggesting impaired transport of IgA and IgA immune complexes from blood to bile, in rats given an LD diet and/or whiskey infusion. A moderate increase in mesangial matrix was observed only in rats given both whiskey and an LD diet. Bright granular IgA and mild granular C3 mesangial deposits and electron-dense deposits were evident in 63 to 70% of experimental rats (all p < 0.001) versus only trace deposits in 5 to 11% of controls. Moderate IgG codeposits were present in 34 to 55% of rats given the LD diet and/or whiskey (all p < 0.02), versus trace deposits in 10% of controls. Significant hematuria and proteinuria were observed in rats given the LD diet and/or whiskey (p < 0.0001) versus controls. Intestinal permeability measured by xylose absorption was significantly increased relative to controls only in rats given both whiskey and the LD diet (p < 0.001). Serum IgA specific for selected alimentary antigens was increased relative to controls in 75 to 100% of the experimental rats. CONCLUSIONS: The combination of LD diet and alcohol intake, which mimics the human alcoholic condition, promotes hepatic and renal changes, leading to hepatocellular injury and a secondary form of IgA nephropathy.

Epithelial transcytosis of monomeric IgA and IgG cross-linked through antigen to polymeric IgA. A role for monomeric antibodies in the mucosal immune system.

We recently demonstrated that the polymeric IgR (pIgR) mediates epithelial transcytosis of immune complexes (IC) containing dimeric IgA (dIgA). In vivo, this "excretory" pathway could allow direct elimination of IgA IC at the mucosal sites where they are likely to form. In contrast, IC containing only monomeric IgA (mIgA) were not transported, consistent with the specificity of pIgR for polymeric IgA. However, the potential exists in vivo that monomeric Ig like mIgA or IgG could become associated through binding to multivalent Ag with IC containing dIgA, and that such mixed IC could act as ligands for pIgR. In the present work, using Madin-Darby canine kidney epithelial cells that express pIgR, we showed that 125I-labeled anti-DNP mIgA or IgG in the same IC with unlabeled dIgA antibody and DNP-BSA Ag was vectorially transported from the basolateral to the apical surface and then released. However, 125I-mIgA IC or 125I-IgG IC (without dIgA) and 125I-mIgA or 125I-IgG in the presence of dIgA antibody (but without Ag) were not transported, demonstrating the necessity for monomeric Ig to be in an IC with dIgA to be transported. Transcytosis of mixed mIgA/dIgA or IgG/dIgA IC was mediated by pIgR because no transport was observed in untransfected, wild-type Madin-Darby canine kidney cells lacking pIgR. The data demonstrate that mIgA and IgG can participate along with dIgA in the "excretory" pathway for local elimination of IgA IC, thus providing a means by which monomeric antibodies have the potential to participate in the mucosal immune system.

A three-tiered view of the role of IgA in mucosal defense.

Mucosal IgA has generally been viewed as an immune barrier to prevent the adherence and absorption of antigens. Recent studies employing polarized epithelial monolayers have suggested two additional functions for mucosal IgA. One is to neutralize intracellular microbial pathogens, such as viruses, directly within epithelial cells. The second is to bind antigens in the mucosal lamina propria and excrete them through the adjacent epithelium into the lumen, thereby ridding the body of locally formed immune complexes and decreasing their access to the systemic circulation.

Unstable inter-H chain disulfide bonding and non-covalently associated J chain in rat dimeric IgA.

Disulfide bonds are a major force in stabilizing the three-dimensional structure of immunoglobulins. To determine the pattern of interchain disulfide bonding between the four H chains, four L chains and single J chain of rat dimeric IgA (dIgA), we analyzed dIgA from the LO DNP-64 hybridoma by diagonal SDS-PAGE. Bands corresponding to one, two, three and four H chains, one and two L chains and the free J chain were observed under non-reducing conditions, suggesting that the interchain disulfide bonds in rat dIgA are unstable under denaturing conditions. Similar patterns of disulfide bonding were observed in three other hybridoma or myeloma dIgAs from LOU/CN rats. In contrast, when dIgA pretreated with iodoacetamide (IA) was analyzed by the same technique, only bands corresponding to four H chains, one and two L chains and the free J chain were observed, suggesting that blocking free sulfhydryl groups stabilizes the inter-H chain disulfide bonds. Reaction of dimeric LO DNP-64 dIgA with 5,5'-dithiobis-(2-nitrobenzoic acid) or with 14C-IA demonstrated that this dIgA contains an average of 4 moles of free sulfhydryl groups per mole of protein under non-denaturing conditions and 9 moles of free sulfhydryl groups under denaturing conditions. Taken together, the results suggest that interchain disulfide bonds in rat dIgA are unstable, presumably due to the influence of nearby free sulfhydryl groups, and that non-covalent forces are critical for stabilizing the dIgA complex. The results also indicate that J chain is entirely non-covalently associated with the H chains, an apparently unique feature of rat dIgA. A model for interchain disulfide bonding in rat dIgA is proposed.

Intracellular neutralization of virus by immunoglobulin A antibodies.

IgA is thought to neutralize viruses at the epithelial surface of mucous membranes by preventing their attachment. Since IgA, a polymeric immunoglobulin, is transported through the lining of epithelial cells by the polymeric-immunoglobulin receptor and since viruses are obligate intracellular parasites, we hypothesized that IgA antibodies may also interfere with viral replication by binding to newly synthesized viral proteins within infected cells. Polarized monolayers of Madin-Darby canine kidney epithelial cells expressing the polymeric-immunoglobulin receptor were infected on the apical surface with Sendai virus. Anti-Sendai virus IgA monoclonal antibody delivered from the basolateral surface colocalized with viral protein within the cell, as documented by immunofluorescence. More importantly, anti-viral IgA reduced virus titers greater than 1000-fold (P less than 0.0001) in apical supernatants and greater than 10-fold (P less than 0.0001) in cell lysates from monolayers treated with anti-viral IgA compared with those treated with either anti-viral IgG or an irrelevant IgA monoclonal antibody. We believe that the differences in viral titers between cell layers treated with specific IgA, which enters the epithelial cell by binding to the polymeric-immunoglobulin receptor, and those treated with specific IgG, which does not enter the cells, or irrelevant IgA indicate that specific intracellular IgA antibodies can inhibit viral replication. Thus, in addition to the classical role of humoral antibodies in extracellular defense, IgA antibody may be able to neutralize microbial pathogens intracellularly, giving IgA a role in host defense that has traditionally been reserved for cell-mediated immunity.

Glomerular hemodynamics and eicosanoid synthesis in a rat model of IgA nephropathy.

We sought to study glomerular pathophysiology in a model of IgA nephropathy (IgAN). Preliminary experiments with oral immunization indicated that Lewis rats had higher IgA levels and IgA/IgG ratios than Wistar, Fischer or Sprague-Dawley rats. Six groups of Lewis rats were studied: four continuously orally immunized for eight weeks with bovine gamma globulin (BGG) in the drinking water, the other two non-immunized controls. Groups of immunized rats were treated with a thromboxane receptor antagonist (SQ 29,485) and/or a thromboxane synthase inhibitor (UK 38,485). After systemic challenge, microscopic hematuria was present in 84% of immunized rats not given anti-thromboxane drugs versus 18% of non-immunized rats (P less than 0.01). Immunized rats showed predominantly IgA glomerular deposits with lesser IgG and C3, and produced more glomerular thromboxane than controls, but no significant increase in prostaglandin E2. Immunized rats also had reduced GFR and RPF, but not a reduced filtration fraction, compared to controls. Thromboxane synthase inhibitor diminished glomerular thromboxane and increased prostaglandin E2 in immunized rats. Anti-thromboxane therapy reduced hematuria and apparently re-established the RPF but not the GFR in immunized rats, yielding a reduced filtration fraction. We propose that increased thromboxane, in concert with mesangial contraction that is unaffected by anti-thromboxane drugs, contributes to the pathophysiology in this model of IgAN.