Human lymphocyte, monocyte, and neutrophil antibody-dependent cell-mediated cytotoxicity toward human erythrocytes

Complete nucleotide sequence, genome organization, and biological properties of human immunodeficiency virus type 1 in vivo: evidence for limited defectiveness and complementation

Previous studies of the genetic and biologic characteristics of human immunodeficiency virus type 1 (HIV-1) have by necessity used tissue culture-derived virus. We recently reported the molecular cloning of four full-length HIV-1 genomes directly from uncultured human brain tissue (Y. Li, J. C. Kappes, J. A. Conway, R. W. Price, G. M. Shaw, and B. H. Hahn, J. Virol. 65:3973-3985, 1991). In this report, we describe the biologic properties of these four clones and the complete nucleotide sequences and genome organization of two of them. Clones HIV-1YU-2 and HIV-1YU-10 were 9,174 and 9,176 nucleotides in length, differed by 0.26% in nucleotide sequence, and except for a frameshift mutation in the pol gene in HIV-1YU-10, contained open reading frames corresponding to 5'-gag-pol-vif-vpr-tat-rev-vpu-env-nef-3' flanked by long terminal repeats. HIV-1YU-2 was fully replication competent, while HIV-1YU-10 and two other clones, HIV-1YU-21 and HIV-1YU-32, were defective. All three defective clones, however, when transfected into Cos-1 cells in any pairwise combination, yielded virions that were replication competent and transmissible by cell-free passage. The cellular host range of HIV-1YU-2 was strictly limited to primary T lymphocytes and monocyte-macrophages, a property conferred by its external envelope glycoprotein. Phylogenetic analyses of HIV-1YU-2 gene sequences revealed this virus to be a member of the North American/European HIV-1 subgroup, with specific similarity to other monocyte-tropic viruses in its V3 envelope amino acid sequence. These results indicate that HIV-1 infection of brain is characterized by the persistence of mixtures of fully competent, minimally defective, and more substantially altered viral forms and that complementation among them is readily attainable. In addition, the limited degree of genotypic heterogeneity observed among HIV-1YU and other brain-derived viruses and their preferential tropism for monocyte-macrophages suggest that viral replication within the central nervous system may differ from that within the peripheral lymphoid compartment in significant and clinically important ways. The availability of genetically and biologically well characterized HIV-1 clones from uncultured human tissue should facilitate future studies of virus-cell interactions relevant to viral pathogenesis and drug and vaccine development.

Characterization of human alveolar macrophage Fc gamma receptor III: a transmembrane glycoprotein that is shed under in vitro culture conditions

Three classes of Fc gamma receptors (FcR) have been identified on blood leukocytes: FcRI, FcRII, and FcRIII. Two forms of FcRIII have recently been characterized; a phosphatidylinositol linked form is found on neutrophils, whereas a transmembrane form of the molecule is found on a subset of peripheral blood lymphocytes. Peripheral blood monocytes express low levels of FcRIII on their surface, whereas FcRIII is readily expressed by tissue macrophages. The purpose of this investigation was to characterize the form of FcRIII expressed by normal human alveolar macrophages (AM) obtained from normal subjects by bronchoalveolar lavage. We found FcRIII expressed by AM has a molecular mass of 50 to 60 kD on sodium dodecyl sulfate polyacrylamide gel electrophoresis and migrates as a single band with a molecular mass of 35 kD after digestion with endoglycosidase F. Macrophage FcRIII was resistant to cleavage by phosphatidylinositol-specific phospholipase C. These results demonstrate that FcRIII expressed by AM is a transmembrane glycoprotein similar to the molecule found on peripheral blood lymphocytes. Scatchard binding analysis using 125I-labeled mAb 3G8 showed that AM express similar numbers of FcRIII as found on neutrophils (73,300 +/- 16,300 versus 69,300 +/- 8,500 receptor sites/cell, respectively; P = 0.73), whereas fewer binding sites were found on FcRIII-positive peripheral blood lymphocytes (35,300 +/- 13,900; P = 0.04). Of note, we found expression of FcRIII by AM was selectively and dramatically reduced during short term in vitro incubation at 37 degrees C. Receptor shedding as a result of proteolytic cleavage is probably responsible for the reduced expression that occurs during short-term in vitro culture.

Cytokine modulation of adhesion molecules in the regulation of immunologic cytotoxicity of epidermal targets

Immunologic cytotoxicity is an important endpoint of the immune response to tumors, viral infected cells, grafted tissues, and exogenous microorganisms, and is also an important mechanism of disease, especially in autoimmunity. There are multiple mechanisms of immunologic cytotoxicity, but each has three major stages: leukocyte/target attachment, specific recognition, and target lysis following effector activation. Adhesion molecules present on leukocytes and potential targets appear to be involved in all three stages of cytotoxicity. A major factor in all types of cellular cytotoxicity is the interaction of LFA-1 on leukocytes and CAM-1 on targets. Modulation of ICAM-1 levels on target by the cytokines TFN-g, IL-1, and TNF-a is a major point of control of the susceptibility of targets to cytotoxicity by many different cytotoxic mechanisms. It also appears that modulation of the avidity of LFA/ICAM-1 binding is another important control point in modulating immunologic cytotoxicity. Cytokines also have important effects on immunologic cytotoxicity in ways other than adhesion molecule induction: effector priming to better respond to specific recognition signals, effector mobilization into tissue, and expansion of cytotoxic effector populations. ICAM-1 on the surface of epidermal keratinocytes and melanocytes is likely to greatly influence cytotoxic damage of these cells in diseases as photosensitive lupus erythematosus, lichen planus, erythema multiforme, and vitiligo. It has been found that the epidermal staining pattern for ICAM-1 in each of these diseases in distinctive and different in each disease. It is proposed that disease-specific induction of ICAM-1 by factors such as UVR and herpes-virus is an important determinant in triggering these skin diseases and in determining the pattern of disease.

Inhibition of antibody-dependent lymphocyte cytotoxicity by homologous restriction factor incorporated into target cell membranes

The 65 kD homologous restriction factor (HRF) was isolated from normal human erythrocytes (E) by immunoadsorption using rabbit anti-human HRF. The protein was radiolabeled and incorporated into the membrane of sheep erythrocytes (Es). Es bearing HRF exhibited a markedly reduced susceptibility to reactive lysis by C5b-9. Es-HRF with 1,000-3,000 HRF molecules per cell and sensitized with rabbit IgG anti-Es also were less susceptible to lysis by human large granular lymphocytes (LGL) than untreated Es sensitized with IgG antibody. Similarly, human E of a patient with paroxysmal nocturnal hemoglobinuria (PNH), lacking HRF and sensitized with IgG antibody underwent lysis by human LGL. Lysis was abrogated by incorporation of isolated human HRF. Incorporation of human decay-accelerating factor (DAF) into sensitized Es had no effect on antibody-dependent, cell-mediated cytotoxicity. Furthermore, lysis of Es by the isolated cytolytic C9-related protein (C9RP) of human cytotoxic lymphocytes could be inhibited by cell bound human HRF. These results suggest that HRF inhibits channel formation not only by C5b-9, but also by cytotoxic lymphocytes.

Immune thrombocytopenia: effects of maternal gamma globulin infusion on maternal and fetal serum, platelet, and monocyte IgG

A 24-year-old woman with lupus-like serologic abnormalities had immune thrombocytopenia that resolved after splenectomy, but increased quantities of platelet surface IgG persisted. Three years later, during the 36th week of her first pregnancy, gamma globulin (400 mg/kg daily for 5 days) was administered intravenously to decrease the risk and/or severity of immune thrombocytopenia in her infant. The infusion produced marked but transient elevations of maternal concentrations of serum IgG and quantities of monocyte surface IgG, but no significant changes in Fc receptor-mediated rosetting of peripheral blood monocytes with antibody-sensitized platelets occurred. Modest increases in quantities of platelets and plasma platelet-specific IgG were demonstrated. The infant, delivered by cesarean section 2 days after the end of the infusion, had a normal platelet count; cord blood had a normal concentration of serum IgG, but an elevated quantity of platelet surface IgG (by comparison with values for normal adults). Infant values of plasma platelet-specific IgG, monocyte surface IgG, and monocyte/platelet rosettes also were within the range of normal for adults. Anticytomegalovirus antibody was present in large amounts in the gamma globulin infused, first appeared in maternal serum after therapy, and was detected in cord serum. The significance of these observations to the management of immune neonatal thrombocytopenia is discussed.

In vitro effects of gammaglobulin (IgG) on human monocyte Fc receptor function. I. Effect on monocyte membrane-associated IgG and Fc receptor-dependent binding of antibody-coated platelets

Intravenous immunoglobulin (IV IgG) has been reported to be clinically beneficial for the treatment of immune thrombocytopenic purpura (ITP). The mechanism of effect still remains unknown. We examined the in vitro effects of two commercially available IV IgG preparations. Exposure of normal monocytes to IgG in vitro produced a significant increase in monocyte-bound IgG. Prior treatment of the commercial IgG preparations by filtration through a 0.2-micron Millipore filter or ultracentrifugation caused a dramatic decrease in IgG bound to the monocyte surface, indicating that IgG aggregates were responsible for this effect. Exposure of monocytes to IgG levels as high as 150 mg (Sandoglobulin) and 400 mg (Gamimune) did not result in a statistically significant inhibition of monocyte-platelet interaction as examined by a morphologic rosetting assay. Thus, despite the ability of IV IgG preparations to cause substantial increments in monocyte surface IgG, impairment of Fc receptor-mediated monocyte binding of antibody-coated platelets was not observed.

Antibody-dependent cell-mediated cytotoxicity and phagocytosis of autologous red blood cells in alphamethyldopa-induced haemolysis

7 alphamethyldopa (AMD)-treated patients with positive direct antiglobulin test (DAT) were investigated. Peripheral blood mononuclear phagocytes of 2 patients suffering from haemolysis caused lysis and phagocytosis of autologous DAT-positive red blood cells (RBC). Eluates from RBC of both patients contained antibodies of IgG1 subclass and supported antibody-dependent cell-mediated cytotoxicity (ADCC) and antibody-dependent phagocytosis (ADPh) of the patients' RBC after remission. Both cytotoxic activities were proportional to the serum concentration and to the number of attacking cells. The 5 patients without overt haemolysis did not show in vitro lysis or phagocytosis of autologous RBC. These results suggest that ADCC as well as ADPh participate in the destruction of RBC in AMD-induced haemolysis in vivo.

Antibody-dependent cellular cytotoxicity and skin disease

Antibody dependent cellular cytotoxicity (ADCC) is a recently described mechanism of immunologic lysis in which cellular targets sensitized by specific antibodies are efficiently and selectively lysed by Fc receptor (FcR) bearing nonspecific effectors. Immunoglobulins of various classes (IgG, IgM, IgA, IgE) and various cellular effectors (large granular lymphocytes, monocyte/macrophages, T lymphocytes, neutrophils, and eosinophils) can induce ADCC in vitro, and the importance of ADCC in vivo is being tested experimentally in resistance to viral, bacterial, and parasitic infection, in tumor surveillance, in allograft rejection, and in inflammatory diseases. There is much indirect evidence that ADCC may be the mechanism of damage of different cellular targets in skin diseases, but the best direct evidence concerns immunologic keratinocyte damage, especially in cutaneous lupus erythematosus (LE). We have shown that keratinocytes of several species are highly susceptible to lymphocyte and monocyte-mediated ADCC, but not to neutrophil or eosinophil ADCC in vitro using two different cytotoxicity assays. In contrast, complement was a relatively ineffective mediator of lysis of metabolically intact keratinocyte targets. Patients with certain cutaneous lupus syndromes have serum antibodies capable of inducing monocyte and lymphocyte ADCC of targets coated with extractable nuclear antigens. We have shown that these antigens apparently move to the cell membrane of keratinocytes in vitro following ultraviolet irradiation. In an animal model, we have shown that antibodies to SSA/Ro bind to human keratinocytes in vivo, especially after ultraviolet irradiation. This antigen/antibody system is highly associated with 3 different photosensitive LE syndromes. The experimental linkage of UV radiation to autoantibody binding to keratinocytes and the demonstration of mononuclear cell-mediated ADCC causing keratinocyte lysis support our hypothesis that the keratinocyte damage and mononuclear cell infiltrate seen histologically in cutaneous LE are part of an ADCC process.

Human monocyte interaction with antibody-coated platelets. I. General characteristics

The interaction of human monocytes with antibody-coated and normal platelets was studied using an assay system that employed rate zonal centrifugation to separate monocyte-associated (bound) platelets from free platelets. At a monocyte:platelet ratio of 1:10, monocytes bound 21.2 +/- 6.6% of antibody-coated platelets and less than 2.2 +/- 1.3% of control platelets. Monocyte binding of antibody-coated platelets was rapid and inhibited by monomeric IgG and staphylococcal protein A. Specific binding was positively related to the number of monocytes present and the amount of IgG displayed on the platelet surface. Transmission and scanning electron micrographs illustrate the membrane binding of antibody-coated platelets to monocytes and suggest that phagocytosis of platelet targets occurs as well. Thus, these observations indicate that monocytes can participate in Fc-receptor binding of anti-PlA1-sensitized platelets. This novel assay may be useful in the analysis of monocyte-macrophage Fc-receptor interaction with platelets sensitized with IgG from both iso- and autoimmune disorders as well as in the characterization of treatment modalities which may alter monocyte Fc-receptor recognition and binding.

Human and murine natural killer cell activity may require lipoxygenation of arachidonic acid

Natural killing (spontaneous cell-mediated cytotoxicity) by either human blood cells or mouse spleen cells is reversibly inhibited by nordihydroquaiaretic acid or BW755C, agents that antagonize the lipoxygenation of arachidonic acid. Cultured mouse natural killer cells are similarly inhibited. These studies suggest that natural killing may be dependent on the lipoxygenation of arachidonic acid.

The role of RNP, Sm, and SS-A/Ro-specific antisera from patients with lupus erythematosus in inducing antibody-dependent cellular cytotoxicity (ADCC) of targets coated with nonhistone nuclear antigens

To better understand potential antibody-dependent mechanisms of tissue damage in lupus erythematosus (LE), an examination of whether antibodies to nonhistone nuclear antigens in LE patients' sera can induce ADCC of cellular targets coated with the corresponding antigens was undertaken. With high titer anti-RNP sera, significant ADCC was seen with monocyte (P less than 0.01), T-lymphocyte (P less than 0.001), and low-density lymphocyte (P less than 0.001) effectors. Using monocyte effectors, significant ADCC was seen with anti-RNP (P less than 0.01), anti-Sm (P less than 0.01), and anti-SSA/Ro (P less than 0.01), with the most profound lysis being with the anti-SSA/Ro sera. Neutrophils were ineffective in any nuclear antigen-antibody system tested. The effective mononuclear cell-mediated ADCC seen with anti-RNP, anti-Sm, and anti-SSA antisera may be related to the mononuclear cell-associated tissue change seen in cutaneous lupus lesions.

Prediction of the outcome of rhesus haemolytic disease of the newborn: additional information using an ADCC assay

An in vitro test system is described which measures the ability of anti-D sera to lyse Rh(D)-positive red cells. This test was applied to anti-D sera from 11 cases of HDN selected in Glasgow and tested 'blind' in Edinburgh. Evidence is presented to support the view that the ADCC (antibody-dependent cell-mediated cytotoxicity ) assay can correctly identify those cases where there is a satisfactory clinical outcome despite a high level of anti-D suggesting otherwise. Amniocentesis might therefore be avoided in this group.

Simultaneous detection of two mechanisms of immune destruction of penicillin-treated human red blood cells

Two separate processes of putative red-cell destruction in penicillin-induced immune hemolysis were measured simultaneously by a rapid (3 hour) assay utilizing 51Cr-labelled red blood cells (RBC). Antibody-dependent, cell-mediated cytotoxicity (ADCC) was estimated by release of 51Cr; and antibody-dependent phagocytosis (ADPh) by quantitation of 51Cr uptake into mononuclear phagocytes as well as by counts of engulfed RBC. Attacking cells were obtained by Ficoll-Hypaque separation of peripheral blood from normal donors. Phagocytosis as well as lysis were proportional to anti-penicillin antiserum concentration, to incubation time, and to the concentration of the attacking cells. Enrichment of mononuclear phagocytes in the attacking cell population by albumin gradient separation led to an increase in phagocytosis as well as in cytotoxicity. Depletion of mononuclear phagocytes resulted in a decline in both processes. Dilution of antiserum abolished ADCC but affected ADPh only slightly. Iodoacetate as well as colchicine inhibited both activities. These results indicate that both processes may be operative in the immune destruction of RBC in vivo.

Antibody-dependent cell-mediated cytotoxicity in human cancer: characterization of patient leukocyte activity and treatment effects

Antibody-dependent cell-mediated cytotoxicity (ADCC), medicated by peripheral blood Hypaque-Ficoll separated mononuclear cells, was studied in humans using chicken erythrocytes (CRBC) incubated in a 1:1200 dilution of rabbit anti-CRBC and human B erythrocytes (HRBC) incubation in a 1:20 dilution of isoantibody. At the optimal target effector ratio of 3:1, ADCC to both CRBC and HRBC was significantly higher than normal in 27 lung cancer, 18 malignant melanoma, and seven colon cancer patients, but not in 20 breast cancer patients. Chemotherapy (single-agent or combination) in 12 patients did not effect ADCC in vitro but significantly suppressed ADCC to both targets after only four or five days of therapy in vivo (ADCC to CRBC, 47.4 to 24.1% lysis: ADCC to HRBC, 48.1 to 16.3% lysis). Immunotherapy with intravenous (IV) corynebacterium parvum or IV methanol extraction residue of BCG (MER) boosted ADCC to both targets within four to seven days of the first dose. It was found that ADCC to HRBC but not to CRBC was completely absent in three cases of active hairy cell leukemia but was present in two cases in remission. The ADCC to HRBC showed an age-dependent increase in both the 51 normal subject and the cancer patients. This was not observed for ADCC to CRBC. The ADCC to CRBC was mediated mainly by an Fc-receptor-positive, nonadherent, small lymphocyte, and ADCC to HRBC was mediated entirely by an adherent monocyte. The ADCC did not correlate significantly with the H3 thymidine incorporation of peripheral blood mononuclear cells, cultured without stimulation for either one or seven days. It also did not correlate with the number of residual granulocytes in the mononuclear cell suspensions. Measurement of ADCC is a useful method of characterizing host defense in malignant disease and its modification by therapy.

Neutrophil-mediated cytotoxicity to schistosomula of Schistosoma mansoni in vitro: studies on the kinetics of complement and/or antibody-dependent adherence and killing

The capacity of rat peritoneal neutrophils to adhere to and kill schistosomula of Schistosoma mansoni in vitro has been investigated. Neutrophils adhere readily to schistosomula in the presence of antibody plus complement (C) and C alone (fresh normal rat serum), but not with heat-inactivated normal rat serum. However, schitosomular killing is only achieved with neutrophils and fIRS or fNRS. In the presence of hiIRS the cells detach after 6 h without producing a significant level of parasite death. The system involving neutrophils plus fIRS is the most efficient in terms of serum dilution and the rate of schistosomular killing. The complement-dependent antibody involved in this system belongs to the class IgG and occurs in rat serum at peak titres, 6-8 wk after a primary schistosome infection. Neutrophil adherence in the presence of fNRS depends upon the generation of C3b molecules at the parasite surface via the alternative pathway of C activation. Studies on the antibody alone system indicate that the lack of significant schistosomular killing might result from the absence of factors which stimulate cell migration, since if a chemokinetic agent is introduced into the assay a 30% increase in mortality is recorded. The possible participation of neutrophils in the destruction of a primary and/or challenge infection in vivo is discussed.

Antibody-dependent, cell-mediated cytotoxicity against human red blood cells: correlation of effector cell type with enzymatic alteration of the target cell surface

Target cell factors, which contribute to the determination of the effector cell type in an antibody-dependent, cell-mediated cytotoxicity system were studied. Human red blood cells (RBC) were treated with papain and investigated by transmission and scanning electron microscopy (TEM and SEM). Both untreated (native) and treated 0 Rh+ RBC were labeled with 51Cr, sensitized with anti-D immunoglobulin and incubated with unfractionated autologous peripheral blood mononuclear cells. With native RBC, immune lysis was proportional to the number of phagocytes: enrichment of effectors in phagocytes increased lysis, while depletion of phagocytes from effectors decreased lysis. Following papain treatment of target RBC, lysis by unfractionated mononuclear effectors was markedly augmented; since this effect was not diminished by decrease of phagocytes to less than 2%, the augmented lysis was not due to the number of phagocytic effectors. TEM and SEM of enzyme-treated RBC showed spherocytosis with varying degrees of crenation and blurring and irregularities of the cellular membranes. The results suggest that papain-induced alterations in the target RBC rendered them susceptible to lysis by interaction between anti-D antibody and peripheral blood lymphocytes.

Oxidative mechanisms of monocyte-mediated cytotoxicity

Human monocytes stimulated with phorbol myristate acetate were able to rapidly destroy autologous erythrocyte targets. Monocyte-mediated cytotoxicity was related to phorbol myristate acetate concentration and monocyte number. Purified preparations of lymphocytes were incapable of mediating erythrocyte lysis in this system. The ability of phorbol myristate acetate-stimulated monocytes to lyse erythrocyte targets was markedly impaired by catalase or superoxide dismutase but not by heat-inactivated enzymes or albumin. Despite a simultaneous requirement for superoxide anion and hydrogen peroxide in the cytotoxic event, a variety of hydroxyl radical and singlet oxygen scavengers did not effect cytolysis. However, tryptophan significantly inhibited cytotoxicity. The myeloperoxidase inhibitor cyanide enhanced erythrocyte destruction, whereas azide reduced it modestly. The inability of cyanide to reduce cytotoxicity coupled with the protective effect of superoxide dismutase suggests that cytotoxicity is independent of the classic myeloperoxidase system. We conclude that monocytes, stimulated with phorbol myristate acetate, generate superoxide anion and hydrogen peroxide, which together play an integral role in this cytotoxic mechanism.