Search

Previous reports have suggested that malignant cells frequently generate a humoral immune response that is ineffective in tumor destruction. Despite coating tumors with IgM and IgG that activate the C system via the classical pathway, normal membrane regulators of C (e.g., membrane cofactor protein and CD59) prevent cytotoxicity. Moreover, C3 deposition on tumors does not result in cytotoxic recognition by phagocytes or NK cells bearing C3 receptors capable of mediating destruction of C3-opsonized bacteria or yeast. The current investigation showed that freshly excised mammary tumors bore IgM, IgG, and C3 detectable by flow cytometry. Normal sera contained natural IgM and IgG Abs reactive with breast tumor cell lines, and IgG Ab titers were increased in patients with breast cancer. Breast tumor cell lines incubated in normal serum from AB+ individuals activated the classical, but not the alternative, pathway of C and became coated with C3. Despite exhibiting membrane-bound C3, serum-opsonized breast tumor cell lines were not killed by CR3 (CD11b/CD18)-bearing NK cells. Priming of NK cell CR3 with small soluble yeast beta-glucan polysaccharides enabled CR3-dependent killing of these same C3-bearing tumor cell lines. Tests of mammary carcinoma cells from freshly excised tumors demonstrated that they also bore sufficient amounts of opsonic C3 for cytotoxic recognition by NK cells bearing polysaccharide-primed CR3, whereas they were largely resistant to NK cells bearing unprimed CR3. This study demonstrates the potential utility of using naturally occurring opsonic C3 on tumor cells for specific immunotherapeutic targeting by NK cells and phagocytes bearing polysaccharide-primed CR3.

SUMMARY Previous studies have shown that an optimal humoral response to a primary protein antigen requires C3 and CR2 (CD21). Sera from non-immunized donors contain natural IgM and IgG antibodies to the primary antigen keyhole limpet haemocyanin (KLH), and these have been previously shown to form immune complexes (IC) that activate the classical pathway of C, fixing iC3b/C3dg onto the KLH antigen. Such KLH IC bind to CR2 on KLH-non-specific B lymphocytes, resulting in antigen processing and MHC class II-dependent presentation to KLH-specific helper T cells. KLH IC also induce B lymphocytes to express the CD80 co-stimulatory molecule via simultaneous CR2 ligation with C3 and FcγRII (CD32) stimulation by IgG natural antibody. The current study demonstrated that KLH IC ligation to either CR2 or FcγRII resulted in activation of a second co-stimulatory molecule, LFA-1 (CD11a, CD18). The possibility of polyclonal B cell stimulation by the presentation of KLH-iC3b/C3dg by antigen-non-specific B cells was excluded by demonstration that in vitro cultivation of peripheral blood mononuclear cells (PBMC) with KLH-iC3b/C3dg elicited only anti-KLH, and did not stimulate synthesis of antibodies to hepatitis C virus (HCV) or tetanus toxoid (TT). Of greatest significance, a specific anti-KLH response was only detectable in cultures stimulated with KLH-iC3b/C3dg and not in cultures stimulated with KLH alone or KLH-IgG. Thus, iC3b/C3dg that was bound to a primary protein antigen enhanced recognition and specific immunoglobulin synthesis by antigen-specific B cells, even though the antigen was taken up and processed via CR2 by both antigen-specific and non-specific B cells.

Zymosan, the cell wall from Saccharomyces cerevisiae, was reported to be a macrophage activator through its beta-glucan over 30 yr ago. Nevertheless, the identity of the beta-glucan receptor has been controversial. This study showed that the alpha M beta 2-integrin, CR3 (Mac-1, CD11b/CD18) served as the beta-glucan receptor through one or more lectin sites located outside of the CD11b I-domain that contains the binding sites for iC3b, ICAM-1, and fibrinogen. Sugar specificity, analyzed with FITC-labeled soluble polysaccharides and flow cytometry, showed CR3-specific staining with several pure beta-glucans but not with alpha-mannan. However, a 10-kDa soluble zymosan polysaccharide (SZP) with high affinity (6.7 x 10(-8) M) for CR3 consisted largely of mannose and approximately 5% glucose. Binding of either SZP-FITC or beta-glucan-FITC to CR3 was blocked not only by pure beta-glucans from yeast, mushroom, seaweed, or barley, but also by N-acetyl-D-glucosamine (NADG), alpha- or beta-methylmannoside, and alpha- or beta-methyl-glucoside. SZP-FITC and beta-glucan-FITC stained all leukocyte types similarly to anti-CR3-FITC, and polysaccharide-FITC staining was inhibited > or = 95% by unlabeled anti-CR3. SZP-FITC staining of cells expressing recombinant chimeras between CR3 and CR4 (p150,95, CD11c/CD18) suggested that both the divalent cation-binding region of CD11b and the region C-terminal to it may regulate binding of polysaccharides to CR3. Unlabeled SZP or beta-glucan also blocked CR3 staining by 11 mAb to C-terminal domain epitopes of CD11b but had no effect on staining by mAb directed to the I-domain. In conclusion, CR3 serves as the leukocyte beta-glucan receptor through a cation-independent lectin site located C-terminal to the I-domain of CD11b. Its sugar specificity is broader than originally appreciated, allowing it to react with certain polysaccharides containing mannose or NADG, as well as glucose.

Normal immune responses to primary protein Ags (those not seen previously by the immune system) have been shown to require C3 and the C3 receptor CR2 (CD21). This investigation tested the hypothesis that natural Abs to the primary protein Ag keyhole limpet hemocyanin (KLH) exist in normal serum and will form immune complexes (IC) that activate C and generate bound iC3b/C3dg, thereby promoting B cell CR2-dependent Ag processing. Both IgM and IgG anti-KLH were detectable in sera from 11 normal donors. IC generated with fresh serum bore iC3b and bound to CR1 (CD35), CR2, and CR3 (CD11b/CD18). Further treatment of IC with serum containing rCR1 formed IC-bearing C3dg that bound only to CR2. When ICs were mixed with B lymphoblastoid cell clones, CR2-dependent processing of the KLH occurred that was dependent on bound C3dg and CR2. Processing occurred regardless of whether the B cells bore KLH-specific surface Ig, although the efficiency of processing was greater with KLH-specific B cells. Both KLH-specific and nonspecific B cell clones presented KLH to KLH-specific T cells. The binding of KLH IC by normal B lymphocytes induced expression of the B7/BB1 Ag (CD80), the required co-stimulatory ligand for T cell CD28. Blocking experiments indicated that although bound C3 and CR2 were required to mediate IC binding to B cells, induction of CD80 expression required the secondary ligation of IC-associated IgG to B cell FcRII (CD32). These data support the hypothesis that responses to primary protein Ags involve IgG natural Abs and C3 that mediate Ag processing and presentation via B cell CR2 and FcRII.

Reperfusion of ischemic tissue is associated with tissue injury greater than that resulting from ischemia alone. C activation has been hypothesized to mediate the so-called ischemia/reperfusion injury through both membrane attack and C5a-dependent recruitment of neutrophils to sites of C3 fixation on the endothelium via C3 receptors. Adherence of neutrophils is preconditional to expression of their deleterious effects, which are central to the pathophysiology of ischemia/reperfusion injury. This study was designed to evaluate the effect of inhibition of C activation on ischemia/reperfusion injury using a soluble and truncated recombinant human CR1 (sCR1) molecule, a "tail-less" form of the membrane C3b/C4b receptor (CD35) that functions as a regulator of C activation. Capillary perfusion and leukocyte adherence to venular endothelium were measured after reperfusion in a mouse cremaster muscle model that allowed microscopic video observation of microcirculatory changes. Infusion i.v. with sCR1 before a 4-h period of ischemia and during a 3-h subsequent period of reperfusion prevented the increase in leukocyte adherence to venular endothelium seen in controls, and enhanced the number of reperfusing capillaries by 55%. Trypan blue staining showed an increase in muscle cell viability from 11 to 50% in mice receiving sCR1 as compared to controls. Tests of blood samples from mice infused with sCR1 demonstrated nearly complete inhibition of the mouse alternative pathway of C activation, but no detectable loss of the mouse classical pathway of C activation. It was concluded that C activation in this model of skeletal muscle injury is likely to be due to the alternative pathway, and that inhibition of C activation during reperfusion inhibits leukocyte adherence to blood vessel walls and protects the capillary microcirculation.

B lymphocytes and transformed B lymphoblastoid cell lines express CR2 (CD21, C3d/EBV-receptor) that is specific for C3 fragments generated by cleavage of C3b or spontaneously hydrolyzed native C3 (C3i) by the serum enzyme factor I and its cofactor, factor H. It had been shown previously that the Raji B cell line could be cultivated in serum-free medium supplemented with only transferrin and either OKB7 anti-CR2 mAb, C3d, or C3d-derived peptides containing the CR2 binding site. Because these agents appeared to function through ligation of CR2, it was unclear how native C3 could also serve as a growth factor, because C3 does not bind to CR2. It appeared possible that Raji cells might be able to use endogenous factors H and I to generate a CR2 ligand from C3, because previous studies had shown that Raji cells synthesized factor H and probably also synthesized factor I. PCR analysis was used to demonstrate factor I mRNA in Raji cells. Secretion of Raji cell factor I protein was confirmed by a sensitive mAb ELISA. Several B cell lines were examined for C3-dependent growth. Raji cells required both C3 (or OKB7) and transferrin for growth, whereas Wil-2 cells grew with transferrin alone and C3 enhanced the growth-promoting activity of transferrin. Two other B cell lines (Daudi and U698M), the T cell line 8402, and the U937 monocytoid cell line could not be sustained with transferrin plus C3. The C3-dependent growth of Raji cells was inhibited almost completely by either OX-23 anti-factor H or 052.11.3 anti-factor I mAb that also blocked the activity of serum-derived factor H or I, respectively. By contrast, there was no inhibition of growth by either OX-24 anti-factor H or OX-21 anti-factor I mAb that did not block factors H and I activity. After the spontaneous hydrolysis of native C3 to C3i, it is hypothesized that Raji cells convert C3i to iC3i with endogenous factors H and I, and then this iC3i serves as a growth factor by binding to membrane CR2.

CR3 (CD11b/CD18) functions both as an iC3b-receptor and as an adhesion molecule for cellular ligands such as ICAM-1. Although CR3 has been well characterized on phagocytic cells, much less is known about CR3 on lymphocytes. In this study, the expression of CR3 was examined on resting and stimulated B, T, and natural killer (NK) cells by three-color flow cytometry. Biotinylated anti-CR3 mAb and streptavidin-FITC were used in combination with anti-CD3 mAb conjugated with peridinin chlorophyll-alpha protein (PerCP) and phycoerythrin-labeled mAbs to CD4, CD8, CD19, or CD56. Among resting lymphocytes, CR3 was expressed on nearly all NK cells (CD56+CD3-), 1% of CD4+CD3+ helper T cells, 7% of CD8+CD3+ cytotoxic T cells, and 20% of B cells (CD19+). Among the 5% of T cells (CD3+) expressing CR3, the majority was CD56+. Incubation of PBMC for 30 min with PMA induced a three- to fivefold increase in CR3 expression on NK cells and a twofold increase on T cells but did not change the expression of CR3 on B cells. This effect of PMA was not blocked by the presence of cycloheximide, suggesting the presence of cytoplasmic (granule) stores of CR3 in these lymphoid cells resembling those previously reported in neutrophils and monocytes. When PBMC were incubated with rIFN-alpha, rIL-2, beta-glucan, or high concentrations of LPS, expression of CR3 on NK cells increased significantly, but > or = 4 hr of stimulation was required. Other cytokines (rIFN-gamma, rIL-1, rIL-4, rIL-6, TNF-alpha) and rC5a had no significant effect on CR3 expression. Among NK cells, both the CD56bright and the CD56dim cells expressed CR3, and the expression of CR3 on both of these NK cell subsets was increased in a similar manner by PMA. However, rIL-2 stimulated a greater increase in CR3 expression on CD56bright cells than on CD56dim cells. These studies suggest that CR3 expressed by NK cells or cytotoxic T cells resembles phagocyte CR3 in that cellular activation stimulates increased surface expression of CR3 derived from cytoplasmic reserves of the receptor.

C receptor CR3 (iC3b-receptor, CD11b/CD18) plays an essential role in several phagocytic and adhesive neutrophil functions. Recent evidence suggests that stimulus-induced phosphorylation of the CR3 beta-chain, CD18, may mediate certain neutrophil functions by transiently converting the molecule to an activated state. Staurosporine, a protein kinase C inhibitor that blocks PMA-induced CD18 phosphorylation, was used to study the functional relevance of this event. Neutrophils adhered to glass were assayed for binding and phagocytosis of iC3b-opsonized sheep E (EC3bi) in the presence or absence of PMA and/or staurosporine. Binding of EC3bi was markedly increased, not only by PMA, but also by staurosporine and by a combination of both agents (three- to sevenfold). The enhancement of rosetting by staurosporine was likely caused by increased surface expression of CR3 via exocytosis of specific granular contents. In contrast, staurosporine alone did not stimulate phagocytosis of EC3bi and markedly inhibited PMA-induced phagocytosis. Staurosporine also inhibited phagocytosis of yeast beta glucan particles, a CR3 ligand that, in contrast to EC3bi, is bound and ingested without additional prior treatment with PMA. beta glucan phagocytosis was associated with a low level of CD18 phosphorylation. Staurosporine did not block phagocytosis in general, because this agent had relatively little effect on FcR-mediated phagocytosis. These data demonstrate that phagocytosis mediated by CR3 requires activation of CR3 via a staurosporine-sensitive pathway. Increased binding of EC3bi, a function of increased surface expression of CR3, does not require activation of CR3 by such a pathway, confirming previous evidence for the independence of these two phenomena. A direct role for CD18 phosphorylation in the activation of CR3 for phagocytosis is consistent with these data.

Mac-1/CR3 functions as both an adhesion molecule mediating the diapedesis of leukocytes across the endothelium and a receptor for the iC3b fragment of complement responsible for phagocytic/degranulation responses to microorganisms. Mac-1/CR3 has many functional characteristics shared with other integrins, including bidirectional signaling via conformational changes that originate in either the cytoplasmic domain or extracellular region. Another key to its functions is its ability to form membrane complexes with glycosylphosphatidylinositol (GPI)-anchored receptors such as Fc gammaRIIIB (CD16b) or uPAR (CD87), providing a transmembrane signaling mechanism for these outer membrane bound receptors that allows them to mediate cytoskeleton-dependent adhesion or phagocytosis and degranulation. Many functions appear to depend upon a membrane-proximal lectin site responsible for recognition of either microbial surface polysaccharides or GPI-linked signaling partners. Because of the importance of Mac-1/CR3 in promoting neutrophil inflammatory responses, therapeutic strategies to antagonize its functions have shown promise in treating both autoimmune diseases and ischemia/reperfusion injury. Conversely, soluble beta-glucan polysaccharides that bind to its lectin site prime the Mac-1/CR3 of circulating phagocytes and natural killer (NK) cells, permitting cytotoxic degranulation in response to iC3b-opsonized tumor cells that otherwise escape from this mechanism of cell-mediated cytotoxicity.

beta-Glucans were identified 36 years ago as a biologic response modifier that stimulated tumor rejection. In vitro studies have shown that beta-glucans bind to a lectin domain within complement receptor type 3 (CR3; known also as Mac-1, CD11b/CD18, or alphaMbeta2-integrin, that functions as an adhesion molecule and a receptor for factor I-cleaved C3b, i.e., iC3b) resulting in the priming of this iC3b receptor for cytotoxicity of iC3b-opsonized target cells. This investigation explored mechanisms of tumor therapy with soluble beta-glucan in mice. Normal mouse sera were shown to contain low levels of Abs reactive with syngeneic or allogeneic tumor lines that activated complement, depositing C3 onto tumors. Implanted tumors became coated with IgM, IgG, and C3, and the absent C3 deposition on tumors in SCID mice was reconstituted with IgM or IgG isolated from normal sera. Therapy of mice with glucan- or mannan-rich soluble polysaccharides exhibiting high affinity for CR3 caused a 57-90% reduction in tumor weight. In young mice with lower levels of tumor-reactive Abs, the effectiveness of beta-glucan was enhanced by administration of a tumor-specific mAb, and in SCID mice, an absent response to beta-glucan was reconstituted with normal IgM or IgG. The requirement for C3 on tumors and CR3 on leukocytes was highlighted by therapy failures in C3- or CR3-deficient mice. Thus, the tumoricidal function of CR3-binding polysaccharides such as beta-glucan in vivo is defined by natural and elicited Abs that direct iC3b deposition onto neoplastic cells, making them targets for circulating leukocytes bearing polysaccharide-primed CR3. Therapy fails when tumors lack iC3b, but can be restored by tumor-specific Abs that deposit iC3b onto the tumors.

CR3 (Mac-1; alphaMbeta2 integrin) functions as both a receptor for the opsonic iC3b fragment of C3 triggering phagocytosis or cytotoxicity and an adhesion molecule mediating leukocyte diapedesis. Recent reports have suggested that a CR3 lectin site may be required for both cytotoxic responses and adhesion. Cytotoxic responses require dual recognition of iC3b via the I domain of CD11b and specific microbial surface polysaccharides (e.g., beta-glucan) via a separate lectin site. Likewise, adhesion requires a lectin-dependent membrane complex between CR3 and CD87. To characterize the lectin site further, a recombinant baculovirus (rBv) system was developed that allowed high level expression of rCD11b on membranes and in the cytoplasm of Sf21 insect cells. Six rBv were generated that contained truncated cDNA encoding various CD11b domains. Immunoblotting of rBv-infected Sf21 cells showed that some native epitopes were expressed by five of six rCD11b fragments. Lectin activity of rCD11b proteins was evaluated by both flow cytometry with beta-glucan-FITC and radioactive binding assays with [125I]beta-glucan. Sf21 cells expressing rCD11b that included the C-terminal region, with or without the I-domain, exhibited lectin activity that was inhibited by unlabeled beta-glucan or anti-CR3 mAbs. The smallest rCD11b fragment exhibiting lectin activity included the C-terminus and part of the divalent cation binding region. The beta-glucan binding affinities of the three C-terminal region-containing rCD11bs expressed on Sf21 cell membranes were not significantly different from each other and were similar to that of neutrophil CR3. These data suggest that the lectin site may be located entirely within CD11b, although lectin site-dependent signaling through CD18 probably occurs with the heterodimer.

Mouse leukocyte CR3 (Mac-1, alphaMbeta2 integrin) was shown to function as a receptor for beta-glucans in the same way as human CR3. Soluble zymosan polysaccharide (SZP) or pure beta-glucans labeled with FITC or 125I bound in a saturable and reversible manner to neutrophils, macrophages, and NK cells. This lectin activity was blocked by anti-CD11b mAb M1/70 or 5C6 and did not occur with leukocytes from CR3-/- (CD11b-deficient) mice. SZP preparations containing primarily mannose or glucose bound to CR3, and the binding of 125I-labeled beta-glucan to CR3 was competitively inhibited by beta-glucans from barley or seaweed, but not by yeast alpha-mannan. Also, as with human CR3, the lectin site of mouse CR3 was inhibited by alpha- or beta-methylglucoside (but not D-glucose), alpha- or beta-methylmannoside, and N-acetyl-D-glucosamine. Phagocytosis of zymosan and serum-opsonized zymosan was partially inhibited by anti-CR3 and was reduced to40% of normal with leukocytes from CR3-/- mice. As with neutrophils from patients with CD18 deficiency, neutrophils from CR3-/- mice exhibited no phagocytosis of particulate beta-glucan. SZP or beta-glucans primed CR3 of neutrophils, macrophages, and NK cells for cytotoxicity of iC3b-opsonized tumor cells that otherwise did not trigger killing. beta-Glucan priming for cytotoxicity was inhibited by anti-CR3 and did not occur with leukocytes from CR3-/- mice. The primed state of macrophage and NK cell CR3 remained detectable for 18 to 24 h after pulsing with beta-glucans. The similarity of mouse and human CR3 in response to beta-glucans highlights the utility of mouse tumor models for development of therapeutic beta-glucans.

When phagocyte CR3 binds to iC3b on bacteria or yeast, phagocytosis and degranulation are triggered because of simultaneous recognition of iC3b via a CD11b I-domain binding site and specific microbial polysaccharides via a lectin site located COOH-terminal to the I-domain. By contrast, when phagocyte or natural killer (NK) cell CR3 adheres to iC3b on erythrocytes or tumor cells that lack CR3-binding membrane polysaccharides, neither lysis nor cytotoxicity are stimulated. This investigation showed that soluble CR3-specific polysaccharides such as beta-glucan induced a primed state of CR3 that could trigger killing of iC3b-target cells that were otherwise resistant to cytotoxicity. Anti-CR3 added before sugars prevented priming, whereas anti-CR3 added after sugars blocked primed CR3 attachment to iC3b-targets. Polysaccharide priming required tyrosine kinase(s) and a magnesium-dependent conformational change of the I-domain that exposed the CBRM1/5 activation epitope. Unlike LPS or cytokines, polysaccharides did not up-regulate neutrophil CR3 expression nor expose the mAb 24 reporter epitope representing the high affinity ICAM-1-binding state. The current data apparently explain the mechanism of tumoricidal beta-glucans used for immunotherapy. These polysaccharides function through binding to phagocyte or NK cell CR3, priming the receptor for cytotoxicity of neoplastic tissues that are frequently targeted with iC3b and sparing normal tissues that lack iC3b.

This review will focus on the complement receptor type 1 (CR1, CD35) expressed by erythrocytes and will cover its structure, molecular biology, and function as a membrane inhibitor of complement activation. The CR1 present on phagocytic cells and lymphocytes has similar functions in regulation of complement activation and serves as a receptor responsible for triggering cellular activation events such as phagocytosis or Ig synthesis. These latter receptor functions of CR1 are not covered in this review, and the reader is referred to several past reviews about leukocyte CR1 for this information (Ross and Medof 1985; Wright and Griffin 1985; Fearon and Ahearn 1989; Ross et al. 1989). Erythrocyte CR1 has three functions in regulation of complement activation that are covered in this review: (a) inhibition of the C3 and C5 convertases of the classical and alternative pathways of complement activation; (b) factor I cofactor activity for cleavage of C3b and iC3b; (c) adsorption of soluble immune complexes, thereby inhibiting complement-mediated inflammation. Table 1 summarizes the attributes of CR1 on all cell types.

Treatment of neutrophils with phorbol myristate acetate (PMA) increases surface expression of CR3 (iC3b-receptor; CD11b/CD18). This up-regulation was examined in whole-mount preparations of adherent neutrophils by stereo high voltage immunoelectron microscopy. In the absence of PMA, immunogold-labeled CR3 was uniformly distributed in the plasma membrane. After 5 to 15-min incubations with PMA, when the highest rates of specific granule exocytosis occurred, the average density of CR3 in the membrane doubled. During this time, dense foci of CR3 were observed in addition to the uniform distribution of CR3. These dense concentrations of CR3 colocalized with secreted lactoferrin (LF), a specific granule marker, above assemblies of cytoplasmic vesicles that were morphologically similar to specific granules and contained LF. After longer incubations in PMA, LF secretion ceased, LF staining became rare, and the high density areas of CR3 were no longer present. These data demonstrate that incipient CR3 appear on the cell surface in high concentration at sites of specific granule exocytosis and then diffuse out into the plasmalemma. PMA also induced shedding of CR3 from the cell surface at the cell margins on structures which also contained LF. Shed CR3 was immunoprecipitated from incubation supernatants and shown to be of identical subunit composition to surface CR3. Although others have shown that the mobile, incipient surface CR3 do not mediate neutrophil adherence to endothelium or homotypic aggregation, the current experiments demonstrated that such CR3 do mediate iC3b-dependent adhesion. The rapid appearance and subsequent dissipation of high concentrations of CR3 on the neutrophil surface caused by specific granule fusion may be essential for neutrophil function at sites of complement deposition.

This paper describes the assay system for two different types of lymphocyte complement receptors, the immune adherence receptor (C3b receptor) and the C3d receptor.

Whether homotypic neutrophil aggregation depends on the quantitative increase of gp165/95 molecules (Mac 1, CR3) recruited to the cell surface during activation was studied using mAb of the CD11b group that recognize distinct epitopes encoded by the alpha-subunit of this glycoprotein. After the addition of antibody MN41, neutrophils did not aggregate in response to a chemoattractant, FMLP. Blockade of preexisting surface gp165/95 by mAb MN41, followed by removal of the excess antibody from the mixture, was used to show that the molecules of gp165/95 newly expressed in response to stimulation by a chemoattractant were incapable of effectively mediating the induced cell-cell interactions of aggregation. Flow cytometry studies confirmed that binding of unlabeled antibody MN41 did not block further increases in surface expression of gp165/95 after stimulation with FMLP. These data suggest that molecules of gp165/95 exhibit two functionally distinct forms, one, present on the surface of freshly isolated neutrophils, that becomes competent to mediate the aggregation response upon activation by a stimulus and a second form that can be translocated to the cell surface by the stimulus but is greatly diminished if not lacking in the ability to participate in that aggregation event.

A sensitive and specific fluorescence assay for membrane complement (C) receptors (CR1 and CR2) was developed with purified C3b and C3d fragments coupled to fluorescent microspheres (0.9 mu diameter). C3-microspheres (C3-ms) bound to cells with low numbers of receptors that were undetectable by other assay techniques. Inhibition studies with anti-CR1 and anti-CR2 demonstrated that C3b-ms and C3d-ms bound exclusively to CR1 and CR2, respectively. Preparation of the C3-ms required only small amounts of partially purified C3 and no immunoglobulin or other C components. Once formed, the C3-ms were stable for up to 4 mo at 4 degrees C.

Human leukocyte complement receptor type three (CR3) was shown to be lectin-like and to resemble bovine serum conglutinin (K) in that it bound to both iC3b and unopsonized yeast (Saccharomyces cerevisiae), and was inhibited by EDTA or N-acetyl-D-glucosamine (NADG). CR3 and K also bound to zymosan (Z), a yeast cell wall extract that contains primarily polysaccharide and no detectable protein. However, structural differences and the absence of K on bovine phagocytes indicated that CR3 was not the human homologue of bovine K. Phagocytic and respiratory responses to unopsonized Z were CR3 dependent because they were inhibited by monoclonal antibodies specific for the alpha-chain of CR3 and did not occur with phagocytes from patients with a genetic deficiency of CR3. The binding of CR3 to Z did not require opsonization of the Z with neutrophil-secreted C3, as Z binding and responses were not inhibited by Fab anti-C3. In addition, CR3-dependent binding of yeast occurred with neutrophils from which protein secretion was blocked by fixation with paraformaldehyde. Rabbit erythrocytes (RaE) also bound weakly to neutrophil CR3 and triggered ingestion. Anti-CR3 not only blocked the binding and ingestion of RaE but also blocked selectively the ingestion of RaEC3b without affecting the strong binding mediated by CR1. Even though sheep E and sheep EC3b were not ingested by neutrophils, a weak binding of CR3 to sheep E was suggested by the finding of 20 to 40% inhibition of sheep EAIgG ingestion by anti-CR3. Such inhibition was only observed in buffers that allowed activity of the CR3 binding site and not in buffers containing either EDTA or NADG. An apparently contradictory finding was that the weak CR3-dependent binding of Z triggered neutrophil ingestion and a superoxide burst, whereas the avid CR3-dependent binding of sheep EC3bi did not induce significant ingestion or a respiratory burst. Blocking studies with monoclonal antibodies specific for different epitopes of the alpha-chain of CR3 suggested that this might result from the presence of two distinct binding sites in CR3: one site for fixed iC3b that did not trigger functions, and a second function-triggering site for Z that did not bind to fixed iC3b.

Erythrocytes, bone marrow-derived lymphocytes, monocytes, and granulocytes were shown to have a receptor activity for C4. Theis C4 receptor activity was studied in relation to the previously identified C3b and C3d receptors. By assay for inhibition of rosette formation by fluid-phase complement (C), only two different lymphocyte C receptors were demonstrated. The immune adherence receptor, the only one of the two shared in common with erythrocytes, was specific for C4 or the C3c region of C3b, but was unreactive with C3d. The other lymphocyte receptor, the C3d receptor, was specific for C3d fragments, but would also react to a lesser extent with the C3d region of uncleaved C3b. ThC3d receptor did not react with either C3c or C4. This specificity of the C3d receptor allowed certain cells which contained only C3d receptors to form rosettes with EAC1-3b and EAC1-3d, but not with EAC14. However, because C3d receptors bound EAC1-3d or C3d fragments more firmly than they did EAC1-3b or C3b fragments, many other types of cells containing only C3d receptors, formed rosettes with EAC1-3d but not with EAC1-3b. Erythrocytes and those lymphocytes which contained only immune adherence receptors, formed rosettes with EAC14 and EAC1-3D but not with EAC1-3d. A double-label assay was devised for the simultaneous detection of both types of C receptors on individual lymphocytes. This assay involved fluorescence labeling of one of the two C receptors with soluble C fragments in combination with the usual rosette method for labeling the other type of C receptor. With this double-label assay, it was observed that the two different lymphocyte C receptors capped independently and thus were located on different molecules which could each move through the fluid membrane matrix independently of the other.

Six different immune complex test systems for the detection of IgG Fc receptors were applied to the study of various human lymphocyte populations. The extent of binding varied widely according to the system and the cell type employed. Two systems bound preferentially to a high proportion of B lymphocytes from peripheral blood or tonsils, one of which bound with only a very few T cells. In contrast, four other test systems which bound well with the Fc receptors on T lymphocytes gave weaker reactions with Fc receptors on B cells. The reactivity of Fc receptors on null or third population lymphocytes was similar to that of the Fc-positive T cells. Pronase digestion experiments showed a graded selective loss of reactivity with the different Fc reagents. No one system was optimal for all of the lymphocyte populations, although aggregated IgG exhibited the broadest spectrum of reactivity. A pronounced effect of temperature was evident on the binding reactions, and native IgG showed strong binding at 4 degrees C, particularly to the Fc receptors on T cells.

Previous investigations of p150,95 (CD11c), the third member of the CD18 membrane glycoprotein family that includes CR3 (Mac-1 or CD11b) and LFA-1 (CD11a), had demonstrated that solubilized p150,95 bound to iC3b-agarose in a manner similar to isolated CR3. The current study showed that membrane surface p150,95 also expressed iC3b-receptor activity and was probably the same as the neutrophil receptor for iC3b- or C3dg-coated erythrocytes (EC3bi or EC3dg) that had been previously designated CR4. Normal neutrophil and macrophage CR4-dependent EC3bi rosettes were inhibited by monoclonal anti-p150,95, and cells from a patient with CD18 deficiency did not form CR4-dependent EC3bi rosettes. With neutrophils that bore large amounts of CR1 and CR3 and little p150,95, EC3bi were found primarily via CR1 and CR3, and demonstration of p150,95-dependent rosettes required large amounts of fixed iC3b, low-ionic strength buffer, and antibody blockade of CR1 and CR3. By contrast, culture-derived macrophages expressed eight times more p150,95 than did monocytes and EC3bi were bound to both p150,95 and CR3 when EC3bi bore small amounts of fixed iC3b and assays were carried out in isotonic buffer. Comparison of the amounts of CR1, CR3, and CR4 in various tissues by immunoperoxidase staining revealed that CR4 was the most abundant C3 receptor molecule on tissue macrophages, and suggested that CR4 might be involved in clearance of C3-opsonized particles or immune complexes.

Human monocytes and monocyte-derived macrophages were examined for their ability to bind and ingest C3-coated sheep erythrocytes (E). Greater than 90% of monocytes and macrophages formed rosettes with EC3b and EC3bi prepared with 15,000-20,000 molecules of C3 per E. Binding of EC3b to the monocyte or macrophage surface was inhibited by rabbit anti-C3b receptor (CR1), but was not inhibited by two different monoclonal anti-C3bi receptor (CR3) antibodies. EC3bi rosette formation was inhibited by monoclonal anti-CR3, but not by anti-CR1. Monocytes and macrophages did not form rosettes with similarly prepared EC3d,g or EC3d. Macrophages cultured for 7 days, but not freshly isolated monocytes, phagocytosed both EC3b and EC3bi. This ability was a consequence of macrophage maturation, as no external stimuli were present during in vitro culture. Experiments directed to determine if EC3b was converted to EC3bi before ingestion suggested that macrophage CR1 and CR3 mediated phagocytosis independently. No evidence was obtained that during the phagocytosis assay, macrophage factor I converted EC3b to EC3bi. The number of E bound or ingested by monocytes and macrophages was dependent on the number of molecules of C3b or iC3b bound per E. Monocytes and macrophages did not require the presence of either Ca++ or Mg++ for rosette formation with EC3b, whereas both divalent cations were required for optimum rosette formation with EC3bi. The presence of divalent cations was required for macrophage phagocytosis of EC3b and EC3bi. For ingestion of EC3b, Mg++ alone was sufficient, and the addition of Ca++ did not increase the number of EC3b ingested. For ingestion of EC3bi, both Ca++ and Mg++ were required for optimal phagocytosis, and their effect was concentration dependent and additive.

Human B lymphocyte antigens analogous to the murine Ia determinants were found on myeloblasts and promyelocytes but not on more mature granulocytes. This was apparent by fluorescent staining with both human alloantisera and rabbit antisera to the isolated Ia-like proteins. The cells of patients with chronic myelocytic leukemia showed this difference especially clearly. Separation of the myeloblasts and promyelocytes by multistep density gradient fractionation produced a marked enrichment of the positive cells. The remaining cells from higher density fractions were more-mature neutrophils that were essentially negative. In acute myeloid leukemia, in which myeloid cells early in differentiation predominate, the vast majority of cells were strongly positive. Similar results were obtained with normal bone marrow cells. Here also, only the early forms of the myeloid series separated by gradient centrifugation had Ia antigens. Evidence was also obtained for the presence of Ia determinants on cells with the appearance of early erythroid precursors. Support for the presence of the Ia determinants on granulocyte-macrophage committed stem cells was provided by the inhibition of granulocyte colony formation in agar cultures following preincubation of normal bone marrow with antiserum and complement. Cross absorptions with purified preparations of immature cells provided evidence for the close similarity of the antigenic determinants on both myeloblasts and B cells. A 28,000-37,000-dalton bimolecular complex obtained from myeloblast membranes contained the Ia determinants and was similar to that obtained from peripheral blood B cell membranes.

In the course of mammary tumor development, a population of nylon nonadherent cells with CR appears in the spleens of tumor-bearing mice although none are ever detected in normal mice. These cells apparently arise in response to immunologic stimulation. In a series of studies we have further characterized subsets of T cells (CR+ and CR-) with regard to their responses to mitogens in the lymphocyte transformation assay. Nylon column nonadherent cells from the spleens of tumor-bearing mice were rosetted in a complement receptor assay using EAC rosetting, and CR+ cells were separated from CR- by centrifugation in a discontinuous Ficoll gradient. CR+ T cells responded strongly to PHA and Con A and in addition responded to LPS, an activity not usually associated with conventional T cells. In contrast, CR- T cells from tumor-burdened mice responded to PHA but failed to respond to Con A or LPS.

Mouse leukocytes were studied for membrane receptors for the third C component by rosette formation with C coated erythrocytes (EAC). Methods were devised for the preparation of EAC complexes containing either mouse C3b or mouse C3d. EAC 1-3dmo were prepared from EA treated with whole mouse serum while EAC 1-3bmo were produced from EAC 142hu treated with whole mouse serum containing sodium suramin. The specificity of the EAC complexes for mouse leukocytes was confirmed by inhibition experiments using fluid phase human C3d. Low concentrations of fluid phase human C3d inhibited EAC1-3dmo rosettes but failed to inhibit EAC 1-3bmo rosettes. Eight-fold higher concentrations of fluid phase C3d caused partial inhibition of EAC1-3bmo rosette formation with lymphocytes, but not with other types of murine leukocytes. Thus mouse leukocytes apparently contain the same two types of C receptors as do human and guinea pig leukocytes. Mouse CR1 is specific for a non-C3d region of C3b, (possibly analogous to human C3c) whereas mouse CR2 is specific for both C3d and the C3d region of C3b.

A comparison was made between the binding sites of two receptors that are believed to be closely associated on human B lymphocytes: complement receptor type two (CR2) that is specific for C3d fragments, and the receptor (EBVR) for Epstein Barr virus (EBV). Isolated fluid-phase CR2 bound to C3d on erythrocytes (EC3d) and inhibited both B cell-EC3d rosettes and the agglutination of EC3d by anti-C3d, it failed to inhibit either the binding or superinfection of B cells by EBV. By contrast, isolated fluid-phase EBVR inhibited EBV B cell binding activity and superinfection but had no CR2 activity. In addition, radiolabeled CR2 bound to EC3d and anti-CR2-Sepharose, whereas radiolabeled EBVR did not. Purified fluid-phase C3d fragments inhibited EC3d rosette formation with CR2+/EBVR+ cells but did not inhibit EBV binding. However, EBV binding to B cells did inhibit EC3d rosette formation. Clones of human/mouse somatic cell hybrids made from CR2+/EBVR+ human B lymphoblastoid cell and CR2-/EBVR- mouse myeloma cell parents expressed either EBVR or CR2 but only rarely expressed both EBVR and CR2. This suggested that the genes for EBVR and CR2 were located on two different human chromosomes. Thus it was concluded that CR2 is probably not the binding site for EBV.

Monoclonal antibody (mAb) anti-Leu M5 reacts with a two-chain molecule composed of a 150-kDa alpha subunit noncovalently associated with a 95-kDa beta subunit and probably is specific for an epitope on the 150-kDa alpha chain. This p150/95 antigen is the third member of a family of polypeptides sharing a common 95-kDa beta chain, which includes the lymphocyte function-associated antigen LFA-1 (p177/95) and complement receptor CR3 (Mo1/MAC-1/OKM1; p165/95) antigens. Sequential immunoprecipitation with anti-p95 beta chain mAb specifically removed the antigens detected by anti-LFA-1, anti-CR3 and anti-Leu M5 mAb. Certain patients with recurrent bacterial infections are genetically deficient in expression of the LFA-1 and Mo1 antigens, and have impaired granulocyte function. Granulocytes from a patient with this disease also failed to react with anti-Leu M5. Stimulation of normal granulocytes with f-Met-Leu-Phe, C5a-desArg, or calcium ionophore resulted in increased expression of Mo1 and Leu M5 antigens on the cell surface, but did not significantly increase expression of LFA-1 antigen. In functional assays, anti-Leu M5 did not inhibit T cell-mediated or natural killer cell-mediated cytotoxicity. In addition, anti-Leu M5 neither inhibited the binding of complement-coated particles to CR1 or CR3 nor did it affect the binding of EC3dg to neutrophils (CR4). These studies clearly indicate that the p150/95 antigen recognized by the anti-Leu M5 antibody is a structurally distinct member of the LFA-1/CR3 family.

In the present study it was shown that normal peripheral lymphocytes have two different complement receptors: one for C3b (the immune adherence receptor) and one for C3b subsequent to its cleavage by C3b inactivator. The two receptors are not cross-reactive and were shown by tests with various antisera to be antigenically distinct. Both the immune adherence receptor and the receptor for C3b inactivator-cleaved C3b were found on normal peripheral lymphocytes and on cultured lymphoblastoid cells. In 15 out of 18 chronic lymphatic leukemia patients, the immune adherence receptor was either partially or completely missing from the peripheral lymphocytes, while the lymphocyte receptor for C3b inactivator-cleaved C3b was retained. Normal erythrocytes, on the other hand, were found to have only the immune adherence receptor. Granulocytes from normal peripheral blood appeared to have only a receptor for C3b and did not have a receptor for C3b inactivator-cleaved C3b.

Sensitization of sheep E with nurse shark (N) serum resulted in the formation of EAN if the serum was heat-inactivated or reacted in 0.04 M EDTA at 30°C, of EANC′1N if the serum was freeze-inactivated or reacted in 0.04 M EDTA at 0°C. Guinea pig (g.p.) complement or sup. I readily lysed EANC′1N but not EAN due to a complete C′1 incompatibility, which could also be demonstrated in vivo: in the absence of C′1N, the then non-complement-fixing antibody was not only tolerated by guinea pigs, but it protected them against lethal Forssman shock. Shark C′1 lysed EARa and its intermediate complexes with g.p.-C′ only to the extent of its contamination with antibody; a noteworthy exception was the complex EARaC′1–C′8 g.p. A large variety of Forssman positive and negative erythrocytes were agglutinated and sensitized by the shark's natural antibodies. Young animals had much less antibody but a fully reactive complement system. A powerful, nontoxic inactivator of mammalian C′4 could be isolated and separated from antibody, C′1, and most other serum proteins by gel filtration of low ionic strength precipitates from normal shark serum.