Targeting to Fcgamma receptors, but not CR3 (CD11b/CD18), increases clearance of Bordetella pertussis.

In the absence of opsonizing antibodies, Bordetella pertussis, the causative agent of pertussis, readily binds to phagocytes via complement receptor 3 (CR3). After opsonization with antibodies, binding is mediated by IgG receptors (FcgammaR). The effect of targeting B. pertussis to either FcgammaR or CR3 was studied. The fate of unopsonized B. pertussis, IgG-opsonized B. pertussis, and B. pertussis opsonized with bispecific antibodies (BsAbs) directed to CR3 or FcgammaRII/-III was compared. IgG antibodies mediated binding and phagocytosis of B. pertussis via FcgammaR by polymorphonuclear leukocytes (PMNL) in vitro. Opsonization of B. pertussis with BsAbs directed against either CR3 or FcgammaRII/-III facilitated PMNL phagocytosis; however, in vivo studies with BsAb revealed that FcgammaR-mediated uptake facilitates B. pertussis clearance, in contrast to uptake via CR3. Targeting of B. pertussis to FcgammaRII/-III in mice deficient in FcgammaRII or FcgammaRIII indicated that the protective effect is attributable to FcgammaRIII. Competition between uptake via CR3 or FcgammaR may determine the outcome of natural infection.

Activation of complement receptor 3 on human monocytes by cross-linking of very-late antigen-5 is mediated via protein tyrosine kinases.

Bordetella pertussis interacts with very-late antigen-5 (VLA-5) receptors on the human monocyte resulting in cross-linking of these receptors followed by activation of complement receptor 3 (CR3) and firm adhesion of B. pertussis to these monocytes. In the present study we investigated whether protein tyrosine kinases are involved in the activation of CR3 on monocytes, which was assessed by the binding of C3bi-coated erythrocytes (EC3bi). Pre-incubation of monocytes with tyrphostin-A47, a specific protein tyrosine kinase inhibitor, before adherence of the cells to an anti-VLA-5 monoclonal antibody-coated surface, or addition of tyrphostin-A47 within 10 min of the adherence to such surface, reduced the binding of EC3bi to monocytes significantly. Pre-incubation of monocytes with tyrphostin-A47 reduced the binding of B. pertussis to such monocytes as well. Inhibitors of protein kinase A and/or C had no effect on EC3bi binding to monocytes. Cross-linking of VLA-5 on monocytes resulted in tyrosine phosphorylation of several proteins. Together, these results indicate that protein tyrosine kinases are involved in the VLA-5-induced activation of CR3 on human monocytes.

Evidence for an intracellular niche for Bordetella pertussis in broncho-alveolar lavage cells of mice.

Bordetella pertussis can attach, invade and survive intracellularly in human macrophages in vitro. To study the significance of this bacterial feature in vivo, we analyzed the presence of viable bacteria in broncho-alveolar lavage (BAL) cells of mice infected with B. pertussis. We found B. pertussis to be present in a viable state in BAL fluid cells until at least 19 days after infection, suggesting B. pertussis to be able to survive in those cells. This intracellular niche may play an important role in the pathogenesis of pertussis. Pertussis toxin and the RGD sequence of the virulence factor filamentous hemagglutinin (FHA) both play a role in the attachment of B. pertussis to human and mouse macrophages in vitro and we hypothesized these virulence factors to be required for invasion and subsequent intracellular survival of B. pertussis in macrophages in vivo. A B. pertussis double mutant, in which the FHA RGD motif was changed to RAD and the ptx genes were deleted, was also found in a viable state in BAL fluid cells, albeit at lower levels than the wild-type strain. In our model, uptake of B. pertussis by alveolar phagocytes in vivo is thus, at least in part, determined by the bacterial virulence factors FHA and pertussis toxin.

Targeting of influenza epitopes to murine CR1/CR2 using single-chain antibodies.

Single-chain variable fragment (scFv) antibodies are genetically engineered molecules comprising the variable regions responsible for specific binding. scFv that recognize certain surface molecules on professional antigen presenting cells could therefore be suitable for targeting Ag to these cells. We have produced an scFv that recognizes murine complement receptors 1 and 2 (CR1/CR2) and genetically fused it with different numbers of influenza hemagglutinin peptides which contain both B and T cell epitopes. The CR1/CR2 specific hybridoma 7G6 was used for RT-PCR to obtain the variable regions, which were then combined to create an scFv fragment. The influenza hemagglutinin intersubunit peptide HA317-41 (IP) was engineered to the N terminus of the scFv in one, two or three copies. The so obtained IP(1-3)7G6scFv still bound the complement receptors; the peptides in the construct were recognized by the peptide specific monoclonal IP2-11-1 on Western blots and ELISAs. The CR1/CR2 positive B lymphomas A20 and 2PK3 presented the peptide to an I-Ed restricted IP specific T cell hybridoma more efficiently when incubated with the IP(1)7G6 constructs as compared to the free peptide. The results suggest that scFv could work as targeting devices in subunit vaccines.

Modulation of the humoral immune response by antibody-mediated antigen targeting to complement receptors and Fc receptors.

During an ongoing immune response, immune complexes, composed of Ag, complement factors, and Igs, are formed that can interact with complement receptors (CRs) and IgG Fc receptors (Fc gamma R). The role of CR1/2 and Fc gamma R in the regulation of the immune response was investigated using OVA that was chemically conjugated to whole IgG of the rat anti-mouse CR1/2 mAb 7G6. FACS analysis using the murine B cell lymphoma IIA1.6 confirmed that the 7G6-OVA conjugate recognized CR1/2. Incubating IIA1.6 cells with 7G6-OVA triggered tyrosine phosphorylation and Ag presentation to OVA-specific T cells in vitro. Immunizing mice with 7G6-OVA at a minimal dose of 1 microgram i.p. per mouse markedly enhanced the anti-OVA Ig response, which was primarily of the IgG1 isotype subclass. The 7G6-OVA did not enhance the anti-OVA response in CR1/2-deficient mice. OVA coupled to an isotype control Ab induced a considerably lower anti-OVA response compared with that induced by OVA alone, suggesting inhibition by interaction between the Fc part of the Ab and the inhibitory Fc gamma RIIb on B cells. This findings was supported by the observation that IIA1.6 cells which were incubated with 7G6-OVA lost the ability to present Ag upon transfection with Fc gamma RIIb. In sum, 7G6-conjugated OVA, resembling a natural immune complex, induces an enhanced anti-OVA immune response that involves at least CR1/2-mediated stimulation and that may be partially suppressed by Fc gamma RIIb.

FcgammaRIII (CD16)-deficient mice show IgG isotype-dependent protection to experimental autoimmune hemolytic anemia.

In autoimmune hemolytic anemia (AIHA), there is accumulating evidence for an involvement of FcgammaR expressed by phagocytic effector cells, but demonstration of a causal relationship between individual FcgammaRs and IgG isotypes for disease development is lacking. Although the relevance of IgG isotypes to human AIHA is limited, we could show a clear IgG isotype dependency in murine AIHA using pathogenic IgG1 (105-2H) and IgG2a (34-3C) autoreactive anti-red blood cell antibodies in mice defective for FcgammaRIII, and comparing the clinical outcome to those in wild-type mice. FcgammaRIII-deficient mice were completely resistent to the pathogenic effects of 105-2H monoclonal antibody, as shown by a lack of IgG1-mediated erythrophagocytosis in vitro and in vivo. In addition, the IgG2a response by 34-3C induced a less severe but persistent AIHA in FcgammaRIII knock-out mice, as documented by a decrease in hematocrit. Blocking studies indicated that the residual anemic phenotype induced by 34-3C in the absence of FcgammaRIII reflects an activation of FcgammaRI that is normally coexpressed with FcgammaRIII on macrophages. Together these results show that the pathogenesis of AIHA through IgG1-dependent erythrophagocytosis is exclusively mediated by FcgammaRIII and further suggest that FcgammaRI, in addition to FcgammaRIII, contributes to this autoimmune disease when other IgG isotypes such as IgG2a are involved.

Murine IgG1 complexes trigger immune effector functions predominantly via Fc gamma RIII (CD16).

Previously, we have demonstrated that phagocytosis of IgG1-coated particles by macrophages in vitro is impaired by deletion of Fc gamma RIII in mice, suggesting that IgG1 may interact preferentially with Fc gamma RIII. In the present study, the biologic relevance of this observation was addressed by triggering various effector functions of the immune system in Fc gamma RIII(-/-) mice, using panels of mAbs of different IgG subclasses. Both binding and phagocytosis of IgG1-coated sheep or human erythrocytes by Fc gamma RIII(-/-) macrophages in vitro were strongly impaired, indicating that the impaired ingestion of complexed IgG1 by Fc gamma RIII(-/-) macrophages is due to a defect in binding. An in vivo consequence of the defective phagocytosis was observed by resistance of Fc gamma RIII-deficient mice to experimental autoimmune hemolytic anemia, as shown by a lack of IgG1-mediated erythrophagocytosis in vivo by liver macrophages. Furthermore, trapping of soluble IgG1-containing immune complexes by follicular dendritic cells in mesenteric lymph nodes from Fc gamma RIII(-/-) mice was abolished. Whole blood from Fc gamma RIII(-/-) mice was unable to induce lysis of tumor cells in the presence of IgG1 antitumor Abs. Finally, IgG1 mAbs proved unable to mount a passive cutaneous anaphylaxis in Fc gamma RIII(-/-) mice. Together, these results demonstrate that IgG1 complexes, either in particulate or in soluble form, trigger in vitro and in vivo immune effector functions in mice predominantly via Fc gamma RIII.

Impaired IgG-dependent anaphylaxis and Arthus reaction in Fc gamma RIII (CD16) deficient mice.

The family of receptors for IgG (Fc gamma R) plays an essential role in antibody-mediated effector functions of the immune system. However, the specific contribution of each of the Fc gamma R classes to in vivo immune reactions is still unclear. Here, we demonstrate that mice deficient for the ligand-binding alpha chain of Fc gamma RIII lack NK cell-mediated antibody-dependent cytotoxicity and phagocytosis of IgG1-coated particles by macrophages. Strikingly, these mice lack IgG-mediated mast cell degranulation, are resistant to IgG-dependent passive cutaneous anaphylaxis, and exhibit an impaired Arthus reaction. These results indicate a prominent role for Fc gamma RIII in inflammatory and anaphylactic responses, making this receptor a potential target in immunotherapy.

Binding of FimD on Bordetella pertussis to very late antigen-5 on monocytes activates complement receptor type 3 via protein tyrosine kinases.

Nonopsonized Bordetella pertussis bind to human monocytes by means of the virulence factors filamentous hemagglutinin (FHA), pertactin, and the minor fimbrial subunit FimD. Receptors on monocytes that mediate binding of B. pertussis to these cells include complement receptor type 3 (CR3), which binds to FHA of B. pertussis, and very late antigen-5 (VLA-5), which binds to an, as yet, unknown ligand on these bacteria. In the present study, the possibility that FimD acts as a ligand for VLA-5 was investigated. Soluble fibronectin, which is the natural ligand for VLA-5, or mAbs against VLA-5 inhibited binding to monocytes of B. pertussis strains that express FimD but not of mutant strains that lack FimD. Beads that were coated with the fusion protein maltose-binding protein-FimD bound to adherent monocytes, and this binding was inhibited by soluble fibronectin or mAb against the alpha- or beta-chain of VLA-5, while soluble collagen or mAb against VLA-4, VLA-6, CR3, or HLA class II had no effect. Down-modulation of VLA-5 on the apical surface of monocytes by plating the cells onto surfaces precoated with anti-VLA-5 mAb also inhibited binding of beads coated with maltose-binding protein-FimD to monocytes, while precoating of the surfaces with mAb against VLA-6 or CR3 had no effect. These results indicate that VLA-5 on monocytes serves as a receptor for FimD on B. pertussis. Binding of C3bi-coated erythrocytes to monocytes, which is a measure of the binding activity of CR3, was enhanced when monocytes were adhered onto plates precoated with purified fimbriae of B. pertussis, while precoating with fimbriae lacking FimD had no effect. Precoating of the plates with FimD-containing fimbriae also enhanced binding of B. pertussis, which express FHA, but not of strains that lack FHA, to monocytes. The enhanced binding of C3bi-coated erythrocytes and B. pertussis to monocytes could be markedly inhibited by tyrphostin-47, a protein tyrosine kinase inhibitor. These results demonstrate that interaction of FimD of B. pertussis with VLA-5 on monocytes activates CR3, which requires protein tyrosine kinases and results in enhanced binding of B. pertussis to the latter receptor via FHA.

Bordetella pertussis fimbriae bind to human monocytes via the minor fimbrial subunit FimD.

Nonopsonized Bordetella pertussis can bind to and become ingested by human monocytes. Previous studies demonstrated that mutant B. pertussis strains that lack fimbriae express a reduced adherence to monocytes. The present study was undertaken to investigate the involvement of the minor fimbrial subunit FimD in the adherence of B. pertussis to human monocytes using purified fimbriae, FimD, and strains with mutations in fimbrial genes. Flow cytometry demonstrated that purified B. pertussis fimbriae avidly bind to monocytes in a dose-dependent manner but bind less to neutrophils and hardly to lymphocytes; FimD-lacking fimbriae did not bind to monocytes. Purified fimbriae or FimD inhibited the binding of wild-type B. pertussis to monocytes in a dose-dependent manner and similarly inhibited the binding of a mutant defective in the major fimbrial subunits. It did not affect the binding of strains defective in FimD. These results prove that B. pertussis bind to human monocytes via the fimbrial minor subunit FimD.

Virulence factors determine attachment and ingestion of nonopsonized and opsonized Bordetella pertussis by human monocytes.

In the present study, the role of virulence factors in and the effect of opsonization on the interactions between Bordetella pertussis and human monocytes were investigated. The methods used facilitated the distinction between attachment and ingestion of bacteria by monocytes. Nonopsonized virulent B. pertussis cells attached to monocytes. Nonopsonized B. pertussis mutant strains deficient in filamentous hemagglutinin, fimbriae, or pertactin exhibited a reduced adherence to monocytes compared with that of their respective parental strains. Nonopsonized avirulent B. pertussis cells did not attach to monocytes. These results led to the conclusion that fimbriae and pertactin are involved in the adherence of nonopsonized virulent B. pertussis cells to monocytes and confirm the role of filamentous hemagglutinin in this process. In the absence of opsonins, about 40% of the monocyte-associated virulent B. pertussis cells were ingested. When B. pertussis cells were preopsonized with inactivated normal serum, about 50% of the monocyte-associated virulent B. pertussis cells were phagocytosed and about 80% of the monocyte-associated avirulent B. pertussis cells were ingested. These results indicate that virulence factors inhibit opsonin-mediated ingestion of B. pertussis by monocytes.

Antimicrobial functions of mononuclear phagocytes.

One of the major functions of mononuclear phagocytes, i.e., monocytes and macrophages, is the phagocytosis and killing of microorganisms. To obtain more insight into the pathogenesis of infectious diseases and to develop new therapies against these diseases, a better understanding of the antimicrobial mechanisms employed by mononuclear phagocytes is essential. The present review gives a short description of the mononuclear phagocyte system and summarizes various methods that are used to study the antimicrobial mechanisms of mononuclear phagocytes.

Very late antigen-5 and complement receptor type 3 cooperatively mediate the interaction between Bordetella pertussis and human monocytes.

Nonopsonized Bordetella pertussis, the causative agent of whooping cough, can attach to and become ingested by human monocytes. It has been reported that complement receptor type 3 (CR3) on human monocyte-derived macrophages binds filamentous hemagglutinin expressed on B. pertussis. In the present study, the role of very late antigen-5 (VLA-5) in the attachment of B. pertussis to adherent human monocytes was investigated. It was found that soluble fibronectin and soluble mAb against VLA-5 markedly inhibited the attachment of B. pertussis to monocytes. When VLA-5 on monocytes was cross-linked by plating these cells onto surfaces precoated with fibronectin or mAb against VLA-5, the binding of both B. pertussis and C3bi-coated sheep erythrocytes to these cells was significantly enhanced, whereas the binding of a B. pertussis mutant strain deficient in filamentous hemagglutinin was not affected. The enhanced attachment of B. pertussis to monocytes plated onto fibronectin-coated surfaces was markedly inhibited by soluble mAb against CR3. Neutrophils, which express similar levels of CR3 and about 10-fold lower levels of VLA-5 as compared with monocytes, did not bind B. pertussis. Together, these results indicate that VLA-5 is involved in the attachment of B. pertussis to monocytes and that cross-linking of VLA-5 enhances the attachment of B. pertussis to monocytes by augmenting the binding activity of CR3. We propose that the attachment of B. pertussis to monocytes occurs in two steps: binding and cross-linking of VLA-5 by B. pertussis enhances the binding activity of CR3, which in turn facilitates the subsequent binding of these bacteria to the latter receptor.

T-cell reactivity to 38 kD insulin-secretory-granule protein in patients with recent-onset type 1 diabetes.

Type 1 diabetes seems to be an autoimmune disease in which T cells have a substantial role. A possible target antigen was suggested by the proliferation of CD4 T cells from a newly diagnosed patient in response to a 38 kD polypeptide of the insulin-secretory-granule membrane. To see whether this reactivity is widespread at disease onset, we have generated T-cell lines in vitro from peripheral blood mononuclear cells of nineteen children of caucasoid origin with newly diagnosed type 1 diabetes and sixteen healthy controls matched for age and HLA antigens. The procedure involved two cycles of incubation with a rat beta-cell tumour subcellular fraction enriched in secretory granules and plasma membrane components, followed by a proliferation assay. Fourteen (74% [95% confidence interval 49-91%]) of the patients' cell lines showed a positive proliferative response on subsequent exposure to the islet-cell antigen preparation compared with only two (13% [2-38%]) of the controls (p = 3 x 10(-4); difference 61% [44-87%]). Two subjects who had high titres of islet-cell autoantibodies (ICA) without clinical diabetes produced responsive T-cell lines. Reactivity towards the 38 kD fraction of insulin-secretory-granule membranes was found only in patients (eight of ten responders tested; 95% CI 44-98%) and one ICA-positive non-diabetic subject. Detection of an ongoing autoimmune T-cell response might be useful diagnostically and could lead to prevention of diabetes through specific immunotherapy.