In defence of indirect costs.

In all nations that do science, researchers need support not only for the direct costs of experimentation, but also for the direct costs of their institutions. The US is about to see a vast change in the system for such costs that will affect its ability to do science and, quite possibly, limit the opportunities for researchers from other countries who are studying or working in US laboratories.

Mac-1 (CD11b/CD18) is an oligodeoxynucleotide-binding protein.

We have studied the interactions of phosphodiester and phosphorothioate oligodeoxynucleotides with Mac-1 (CD11b/CD18; alpha M beta 2), a heparin-binding integrin found predominantly on the surface of polymorphonuclear leukocytes (PMNs), macrophages and natural killer cells. Binding of a homopolymer of thymidine occurred on both the alpha M and beta 2 subunits. Soluble fibrinogen, a natural ligand for Mac-1, was an excellent competitor of the binding of a phosphorothioate oligodeoxynucleotide to both TNF-alpha-activated and nonactivated PMNs. Upregulation of cell-surface Mac-1 expression increased cell-surface binding of oligodeoxynucleotides. Binding was inhibited by anti-Mac-1 monoclonal antibodies, and the increase in cell-surface binding was correlated with a three- to fourfold increase in internalization by PMNs. An oligodeoxynucleotide inhibited beta 2-dependent migration through Matrigel, but the production of reactive oxygen species in PMNs adherent to fibrinogen dramatically increased. Thus, our data demonstrate that Mac-1 is a cell-surface receptor for oligodeoxynucleotides that can mediate their internalization and that this binding may have important functional consequences.

Tyrosine phosphorylation is required for Fc receptor-mediated phagocytosis in mouse macrophages.

Although Fc receptor-mediated phagocytosis is accompanied by a variety of transmembrane signaling events, not all signaling events are required for particle ingestion. For example, Fc receptor-mediated phagocytosis in mouse inflammatory macrophages (Di Virgilio, F., B. C. Meyer, S. Greenberg, and S. C. Silverstein. 1988. J. Cell Biol. 106:657; Greenberg, S., J. El Khoury, F. Di Virgilio, and S. C. Silverstein. 1991. J. Cell Biol. 113:757) and neutrophils (Della Bianca, V., M. Grzeskowiak, and F. Rossi. 1990. J. Immunol. 144:1411) occurs in the absence of cytosolic calcium transients. We sought to identify transmembrane signaling events that are essential for phagocytosis. Here we show that tyrosine phosphorylation is an early event after Fc receptor ligation in mouse inflammatory macrophages, and that the formation of tyrosine phosphoproteins coincides temporally with the appearance of F-actin beneath phagocytic cups. The distribution of tyrosine phosphoproteins that accumulated beneath phagocytic cups was punctate and corresponded to areas of high ligand density on the surface of the antibody-coated red blood cells, which provided the phagocytic stimulus. A tyrosine kinase inhibitor, genistein, but not several inhibitors of protein kinase C, blocked the appearance of tyrosine phosphoproteins as assessed by immunofluorescence, the focal accumulation of F-actin beneath immunoglobulin G-opsonized particles, and the ingestion of these particles as well. We suggest that tyrosine phosphorylation is a critical signaling event that underlies Fc receptor-mediated phagocytosis in mouse macrophages, and is necessary for the engulfment per se.

Receptors for C3b and C3bi promote phagocytosis but not the release of toxic oxygen from human phagocytes.

We have measured the release of H2O2 from granulocytes, monocytes, and macrophages during spreading on ligand-coated culture surfaces. While IgG-coated surfaces stimulate vigorous release of H2O2, neither C3b- nor C3bi-coated surfaces promoted appreciable release of H2O2 despite full ligation of C3b and C3bi receptors. We also measured release of H2O2 from cultured monocytes spreading on surfaces coated with both fibronectin and C3. Under such circumstances, the C3 receptors elicit a strong phagocytic response, but no H2O2 release was recorded. We conclude that the C3b and C3bi receptors of monocytes and granulocytes do not signal the generation of toxic oxygen intermediates from these cells.

Tumor-promoting phorbol esters stimulate C3b and C3b' receptor-mediated phagocytosis in cultured human monocytes.

Monocytes were isolated in high yield (approximately 80%) and purity (greater than 90%) by Percoll gradient centrifugation and incubated in Teflon culture vessels. Using this culture method, we routinely recovered 80% of the cells originally placed into culture. Studies of the C3b and C3b' receptors of these monocytes showed that the function of both receptors could be dramatically altered by treating the cells with tumor-promoting phorbol esters. Both C3b and C3b' receptors of human monocytes efficiently mediate attachment of erythrocytes coated with the corresponding ligands, but do not promote their ingestion. However, monocytes treated with phorbol myristate acetate (PMA) or phorbol didecanoate ingest C3b- and C3b'-coated erythrocytes. Phorbol esters that are inactive as tumor promoters do not stimulate C3 receptor-mediated phagocytosis. The ability of monocytes to respond to PMA by activation of C3 receptors is developmentally regulated. Freshly isolated monocytes do not take up C3b- or C3b'-coated erythrocytes in response to PMA, but after 3 d of culture they show strong PMA-stimulated uptake. The stimulatory effect of PMA on monocyte C3b and C3b' receptor function occurs within minutes, is stable for hours, is cycloheximide insensitive, and can be inhibited with colchicine. Several lines of evidence indicates that phagocytosis of C3b or C3b'-coated erythrocytes is specifically mediated by the monocytes' C3b and C3b' receptors. First, erythrocytes attached to monocytes with concanavalin A are not ingested when the monocytes are treated with PMA. Second, monocytes plated on IgG-bearing substrates lose Fc receptor activity on their nonadherent surfaces but retain the capacity to ingest C3b- or C3b'-coated erythrocytes after PMA treatment. Third, PMA-treated monocytes plated on C3b-coated surfaces lose C3b receptor activity on their nonadherent surfaces but retain the capacity to ingest C3b'-coated erythrocytes. Conversely, PMA-treated monocytes plated on C3b'-coated surfaces show reduced C3b' receptors activity on their nonadherent surfaces but retain the capacity to ingest C3b-coated erythrocytes.

2-Deoxyglucose selectively inhibits Fc and complement receptor-mediated phagocytosis in mouse peritoneal macrophages II. Dissociation of the inhibitory effects of 2-deoxyglucose on phagocytosis and ATP generation.

Macrophages incubated in 2-deoxy-D-glucose (2-dG)-containing medium showed a marked decrease in cellular ATP content, and were unable to ingest IgG- and complement-coated erythrocytes via the corresponding membrane receptors for these ligands. However, the inhibitory effects of 2-dG on Fc- and C3 receptor-mediated phagocytosis were not a consequence of lowered macrophage ATP levels since addition of glucose or mannose to the culture medium restored the capacity of the macrophages to ingest IgG- and C3-coated particles without increasing ATP levels. These results indicate that Fc- and C3 receptor-mediated phagocytosis (opsonin dependent) differs qualitatively from the ingestion of latex and zymosan particles (opsonin independent); they suggest that the same regulatory molecules govern the responses of phagocytic cells to signals initiated by both the Fc and C3 receptors. The possibility that these molecules are regulated by glycosylation is discussed.

Characterization of the macrophage receptro for complement and demonstration of its functional independence from the receptor for the Fc portion of immunoglobulin G.

The complement receptor of the macrophage membrane recognizes particle-bound C3b but does not recognize particle-bound C3d. C3-b-coated sheep erythrocytes were bound to macrophages via their C3b receptors, and the preparations were then incubated with either latex particles or opsonized pneumococci (test particles). Macrophages ingested the test particles, but erythrocytes were not ingested; they remained bound to C3b receptors of the macrophage plasma membrane. Thus, a signal initiating ingestion via one type of receptor is not transmitted to all receptors which have the potential to mediate phagocytosis.

Studies of the macrophage complement receptor. Alteration of receptor function upon macrophage activation.

We have examined the roles of Fc receptors and complement receptors in mediating the interaction of sensitized sheep erythrocytes (E) with activated and with nonactivated mouse peritoneal macrophages. Both activated and nonactivated macrophages ingest IgG-coated erythrocytes [E(IgG)]; activated cells intest 1.5-2 times as man E(IgG) as do nonactivated macrophages. Thus, there is a quantitative difference in Fc receptor-mediated ingestion between activated and nonactivated macrophages. There is, however, a qualitative difference in function of complement receptors of activated and nonactivated macrophages. Nonactivated macrophages avidly bind complement-coated E [E(IgM)Ia1, but do not ingest them to a significant degree. Activated macrophages, on the other hand, bind and ingest E(IgM)C. The possibility of Fc receptor participation in mediating ingestion of E(IgM)C by activated macrophages was eliminated by blocking Fc receptors with an antimacrophage IgG fraction. Activated macrophages treated with antimacrophage IgG did not ingest E(igG) but did ingest both E(IgM)C AND E(IgM)C. Nonactivated macrophages treated with antimacrophage IgG did not interact at all with E(IgG). These cells bound, but did not ingest, E(IgM)C and E(IgM)C. Complement receptor-mediated ingestion is a marker for macrophage activation and may be physiologically important in the elimination of complement-coated particles.

Immunological analysis of plasminogen activators from normal and transformed hamster cells. Evidence that the plasminogen activators produced by SV40 virus-transformed hamster embryo cells and normal hamster lung cells are antigenically identical.

Rabbits were immunized against the plasminogen activator released by SV4- virus-transformed hamster embryo cells. The resulting antiplasminogen activator immunoglobulin (APA-IgG) inhibited the enzymatic activity of the plasminogen activator produced by SV40-transformed hamster cells, and the plasmin-catalyzed release of these cells from the tissue culture dish. APA-IgG was not cytotoxic for these cells even in the presence of complement and did not inhibit their release of plasminogen activator. APA-IgG formed a single precipitin line in immunodiffusion plates using highly purified plasminogen activator as antigen. APA-IgG inhibited the plasminogen activator produced by newborn hamster lung cells and by an established diploid line (DON) of hamster lung cells, but did not inhibit plasminogen activators produced by normal or transformed hamster kidney cells or by cells of other species (mouse and human). We derive three major conclusions from these data: (a) There are several immunologically distinguishable forms (isozymes) of plasminogen activators in normal hamster tissues. (b) The plasminogen activators produced by normal hamster lung cells and by SV40 virus-transformed hamster embryo cells share antigenic determinants and are presumably the same isozyme. (c) The plasminogen activators produced by different hamster tumor cells do not share antigenic determinants and are presumably different isozymes.

Activated human monocytes inhibit the intracellular multiplication of Legionnaires' disease bacteria.

We have examined the interaction between virulent egg yolk-grown L. pneumophila, Philadelphia 1 strain, and in vitro-activated human monocytes, under antibiotic-free conditions. Freshly explanted human monocytes activated by incubation with concanavalin A (Con A) and human lymphocytes inhibited the intracellular multiplication of L. pneumophila. Both Con A and lymphocytes were required for activation. Con A was consistently maximally effective at greater than or equal to 4 mug/ml. Monocytes activated by incubation with cell-free filtered supernatant from Con A-sensitized mononuclear cell cultures also inhibited the intracellular multiplication of L. pneumophil a. The most potent supernatant was obtained from mononuclear cell cultures incubated with greater than or equal to 15 mug/ml Con A for 48 h. The degree of monocyte inhibition of L. pneumophila multiplication was proportional to the length of time monocytes were preincubated with supernatant (48 {greater than} 24 {greater than} 12 h) and to the concentration of supernatant added (40 percent {greater than} 20 percent {greater than} 10 percent {greater than} 5 percent). Monocytes treated with supernatant daily were more inhibitory than monocytes treated initially only. With time in culture, monocytes progressively lost a limited degree of spontaneous inhibitory capacity and also lost their capacity to respond to supernatant with inhibition of L. pneumophila multiplication. Supernatant-activated monocytes inhibited L. pneumophila multiplication in two ways. They phagocytosed fewer bacteria, and they slowed the rate of intracellular multiplication of bacteria that were internalized. As was the case with nonactivated monocytes, antibody had no effect on the rate of intracellular multiplication in supernatant-activated monocytes. Neither supernatant-activated nor nonactivated monocytes killed L. pneumophila in the absence of antibody. Both killed a limited proportion of these bacteria in the presence of antibody and complement. We have previously reported that anti-L, pneumophila antibody and complement neither promote effective killing of L. pneumophila by human polymorphonuclear leukocytes and monocytes nor inhibit the rate of L. pneumophila multiplication in monocytes. These findings and our present report that activated monocytes do inhibit L. pneumophila multiplication indicate that cell-mediated immunity plays a major role in host defense against Legionnaires' disease.

Interaction of the Legionnaires' disease bacterium (Legionella pneumophila) with human phagocytes. I. L. pneumophila resists killing by polymorphonuclear leukocytes, antibody, and complement.

We have previously reported that virulent egg yolk-grown Legionella pneumophila, Philadelphia 1 strain, multiplies intracellularly in human blood monocytes. We now report on the interaction between virulent L. pneumophila and human polymorphonuclear leukocytes (PMN), antibody, and complement, in vitro, under antibiotic-free conditions. L. pneumophila in concentrations ranging from 10(3) to 10(6) colony forming units (CFU)/ml are completely resistant to the bactericidal effects of 0-50 percent fresh normal human serum, even in the presence of high concentrations of rabbit or human anti-L. pneumophila antibody. L. pneumophila bacteria fix the third component of complement (C3) to their surfaces, as measured by fluorescence microscopy using rhodamine- conjugated goat anti-human C3 IgG, only when the bacteria are incubated with both specific anti-L. pneumophila antibody and complement. Similarly, L. pneumophila adhere to PMN, as measured by fluorescence microscopy, only in the presence of both specific antibody and complement. Electron microscopy revealed that these opsonized bacteria are phagocytosed by the PMN. PMN require both antibody and complement to kill L. pneumophila; even then, PMN reduced CFU of L. pneumophila by only 0.5 log under conditions in which they reduce CFU of a serum-resistant encapsulated strain of Escherichia coli by 2.5 logs. Separation of PMN-associated and nonassociated CFU of L. pneumophila revealed that the major proportion of the surviving bacteria are PMN associated. Thus, the ineffective killing of opsonized L. pneumophila is a result of a failure of PMN to kill these bacteria after they become PMN- associated. With or without antibody, PMN do not support the growth of L. pneumophila. These findings suggest that PMN, even in conjunction with the humoral immune system, do not play a decisive role in defense against the Legionnaires' disease bacterium.

Interaction of the legionnaires' disease bacterium (Legionella pneumophila) with human phagocytes. II. Antibody promotes binding of L. pneumophila to monocytes but does not inhibit intracellular multiplication.

In an accompanying paper (13), we reported that human polymorphonuclear leukocytes kill only a limited proportion (0.5 log) of an inoculum of Legionella pneumophila (Philadelphia 1 strain) in the presence of human anti-L. pneumophila antibody and complement. We now report on the effect of anti-L. pneumophila antibody on L. pneumophila-monocyte interaction. The studies were carried out under antibiotic-free conditions. Monocytes bind more than three times as many viable L. pneumophila bacteria in the presence of both antibody and complement than in the presence of complement alone. Monocytes requires both antibody and complement to kill any L. pneumophila: however, even then, monocytes kill only a limited proportion (0.25 log) of an inoculum. The surviving bacteria multiply several logs in the monocytes and multiply as rapidly as when the bacteria enter monocytes in the absence of antibody. These findings suggest that humoral immunity may not be an effective host defense against L. pneumophila. Consequently, a vaccine that resulted only in antibody production against the Legionnaires' disease bacterium may not be efficacious.

Segmental response of the macrophage plasma membrane to a phagocytic stimulus.

A method of attaching mouse RBCs to mouse macrophages is described. Both cell types were coated with rabbit anti-mouse macrophage F(ab')(2), and cross-linkage of cells was effected with sheep F(ab')(2) directed against rabbit F(ab')(2). 98% of macrophages attached an average of 11 RBCs each. Attachment occurred at 37 degrees C and was stable for at least 4 h. Less than 0.1% of macrophages ingested RBCs under these conditions. Latex particles and opsonized pneumococci were ingested as avidly by RBC-coated macrophages as by native macrophages. Ingestion of these particles did not prompt ingestion of attached RBCs. When anti-RBC IgG was added, however, over 90% of macrophages ingested an average of six RBCs each. Thus, ingestion of one particle does not trigger generalized phagocytosis of all particles attached to the cell's plasma membrane, and the phagocytic stimulus is confined to the segment of the cell's plasma membrane immediately adjacent to the particle being ingested.

2-Deoxyglucose selectively inhibits Fc and complement receptor-mediated phagocytosis in mouse peritoneal macrophages. I. Description of the inhibitory effect.

Incubation of normal or thioglycollate-elicited mouse peritoneal macrophages with 2-deoxy-D-glucose (2-dG) inhibits the capacity of these macrophages to phagocytize IgG- or complement-coated particles via their Fc and C3 receptors. 2-dG has no inhibitory effect on the capacity of these macrophages to phagocytize latex or zymosan particles, which are ingested in the absence of specific opsonins, and it does not inhibit binding of IgG- or C3-coated particles to their respective receptors on the macrophage's plasma membrane. 2-dG exerts its inhibitory effect on the macrophage and not on the opsonized particle. The inhibition is independent of particle size, occurs within 15-30 min of addition of this glucose analogue to the medium at 37 degrees C, cannot be overcome by supra-agglutinating amounts of opsonizing antibody, and is completely reversible by substitution of 5.5 mM glucose for 50 mM 2-dG in the medium. Addition of equimolar amounts of glucose or mannose, but not of fructose, galactose, fucose, or glucosamine, to medium containing 50 mM 2-dG results in substantial reversal of the inhibitory effect of 2-dG on Fc and C3 receptor mediated phagocytosis.

Colocalization of F-actin and talin during Fc receptor-mediated phagocytosis in mouse macrophages.

We have studied the distribution of talin in J774 cells and mouse peritoneal macrophages undergoing Fc receptor-mediated phagocytosis. At early stages of phagocytosis, talin accumulates in the cells' cortical cytoplasm adjacent to the forming phagosome and extends into pseudopods that are encircling the particle. Talin colocalizes with F-actin at these sites. After particle ingestion is completed, F-actin and talin are no longer concentrated adjacent to phagosomes. Thus, talin and F-actin undergo dynamic and coordinate changes in their cytoplasmic location during Fc receptor-mediated phagocytosis.

Fibronectin and serum amyloid P component stimulate C3b- and C3bi-mediated phagocytosis in cultured human monocytes.

Fibronectin (FN) and serum amyloid P component (SAP) markedly enhance phagocytosis mediated by the C3b and C3bi receptors of cultured human monocytes but not of granulocytes. (The C3b and C3bi receptors of granulocytes can be activated by treatment of these phagocytes with PMA.) Activation of monocyte C3 receptors by FN is developmentally regulated: Freshly explanted monocytes respond to FN with a small increase in C3 receptor-mediated phagocytosis while monocytes matured in culture exhibit a much greater response. The mechanism of action of FN on C3 receptors of cultured monocytes is unique in two respects. First, while substrate-bound FN or SAP activate monocyte C3 receptors, soluble FN does not. Second, stimulation of the basal surface of monocyte plasma membranes by substrate-bound FN activates C3b and C3bi receptors on the apical surface of the plasma membrane, i.e., at sites remote from the segments of membrane in contact with the FN or SAP.

Modulation of Fc receptors of mononuclear phagocytes by immobilized antigen-antibody complexes. Quantitative analysis of the relationship between ligand number and Fc receptor response.

Macrophages plated on surfaces coated with antigen-IgG complexes lose the capacity to bind and ingest IgG-coated particles via their Fc receptors (FcR). Macrophages plated on surfaces containing a similar number of IgG molecules that are not complexed to antigen show little or no decrease in FcR activity. Using a rat monoclonal antibody (2.4G2 IgG) directed against the trypsin-resistant FcR (FcRII) of mouse macrophages we show that the decrease in receptor activity induced by substrate-adherent immune complexes is caused by the physical removal of 60 and 75% of FcRII from the nonadherent membrane surfaces of resident and thioglycollate broth-induced macrophages, respectively. Macrophages maintained on antigen-IgG-coated surfaces for up to 44 h show no recovery in FcRII activity or number, while macrophages on control surfaces exhibit two and threefold increases, respectively, in these parameters. Macrophages maintained for 72 h on antigen-IgG-coated surfaces show a small recovery in FcRII activity, and in the number of FcRII that is accessible to bind 125I-2.4G2 IgG. FcRII modulation, as measured by the binding of 125I-labeled 2.4G2 IgG, is initiated when the number of IgG molecules bound to the substrate is approximately equal to the total number of FcRII on the plasma membranes of all the macrophages on the substrate. FcRII activity and number decrease linearly as the number of substrate-bound IgG molecules increases exponentially, and are maximally reduced when the number of IgG molecules on the substrate is 20-fold greater than the total number of all FcRII on the surfaces of all the macrophages in the culture. Thus there is a stoichiometric relationship between the number of IgG molecules on the substrate and the extent of FcRII modulation.

Macrophages elicited with heat-killed bacillus Calomette-Guérin protect C57BL/6J mice against a syngeneic melanoma.

We have demonstrated that a murine cytotoxic peritoneal cell can be elicited by intraperitoneal immunization with heat-killed Mycobacterium bovis, strain Bacillus Calmette-Guérin (BCG). When these cells are injected together with cells of clone B(5)59 of B16 melanoma in a Winn-type transfer assay into syngeneic C57BL/6J mice, the tumorigenic potential of the melanoma is completely abrogated. Similarly, mice immunized intraperitoneally with dead BCG are protected against intraperitoneal challenge with a number of B16 melanoma cells sufficient to cause tumors in 100% of control mice. However, mice immunized intraperitoneally with dead BCG are not protected against tumor formation when B16 melanoma cells are injected subcutaneously. Co-injection of BCG-elicited peritoneal cells with B16 melanoma cells into nude or sublethally irradiated (650 rad) mice inhibits tumor formation in > 85% of the mice, indicating that additional participation of host bone marrow- or thymus-derived leukocytes is not required to eradicate the tumor implant. The effector cell in the BCG-induced peritoneal exudate is adherent and phagocytic and is a mononuclear phagocyte. Nonadherent lymphoid cells from the same BCG-induced peritoneal exudate and from thioglycollate-broth-elicited granylocytes and macrophages neither prevent nor delay B16 tumor formation.

Effects of immobilized immune complexes on Fc- and complement-receptor function in resident and thioglycollate-elicited mouse peritoneal macrophages.

We have examined the Fc- and complement-receptor function of resident and thioglycollate-elicited mouse peritoneal macrophages plated on surfaces coated with rabbit antibody-antigen complexes and with complement. We derive four major conclusions from these studies. (a) The trypsin-resistant Fc receptors of resident and thioglycollate-elicited macrophages are completely modulated when these cells are plated on rabbit antibody-antigen complexes. Residual Fc receptor activity is a result of the incomplete modulation of trypsin-sensitive IgG2a receptors. (b) The complement receptors of thioglycollate-elicited macrophages, but not of resident macrophages, are modulated when these cells are plated on complement-coated surfaces. The capacity of the two cell types to modulate their complement receptors is correlated with their ability to ingest complement-coated erythrocytes. (c) The complement and Fc receptors of both types of macrophages move independently of one another. (d) Complement masks the Fc segments of IgG in immune complexes thereby rendering them ineffective as ligands for macrophage Fc receptors.

Studies on the mechanism of phagocytosis. II. The interaction of macrophages with anti-immunoglobulin IgG-coated bone marrow-derived lymphocytes.

We have examined the effect of the distribution of anti-immunoglobulin IgG molecules on the surface of bone marrow-derived lymphocytes upon the interaction of these cells with macrophages. Lymphocytes which were diffusely coated with antibodies to surface immunoglogulin were ingested by macrophages. Lymphocytes which had the same number of anti-immunoglobulin IgG molecules redistributed to one pole of the surface bound to the macrophages' Fc receptors but were not ingested. These results confirm our previous hypothesis that ingestion of an immunologically coated particle requires the sequential, circumferential binding of specific receptors on the plasma membrane of a phagocytic cell to immunologic ligands distributed over the entire particle surface. Macrophages which had bound capped lymphocytes by the macrophages' Fc receptors removed the immune complex caps from the lymphocyte surface without destroying the lymphocytes. These lymphocytes remained attached to the macrophage surface. The finding that macrophages can phagocytize immune complexes from the surface of a cell without destroying the cell to which these complexes are attached may be important in understanding the effects of antigens and antibodies on cells participating in a humoral immune response, in identifying the mechanisms by which chronic viral infections are established, and in defining the roles of blocking antibodies in tumor immunity.