Antibody-dependent Cellmediated Cytotoxicity (ADCC) Robert F Graziano, Medarex, Inc., Annandale, New Jersey, USA Paul M Guyre, Dartmouth Medical School, Hanover, New Hampshire, USA Antibody-dependent cell-mediated cytotoxicity (ADCC) is the nonphagocytic killing of an antibody-coated target cell by a cytotoxic effector cell. ADCC is triggered through interaction of target-bound antibodies with Fc receptors, molecules on the effector cell surface that recognize the Fc region of immunoglobulins. Effector cells that mediate ADCC include natural killer cells, monocytes, macrophages, neutrophils, eosinophils and dendritic cells. Advanced article Article Contents. Introduction. Fc Receptors are Trigger Molecules for ADCC. Cellular Effectors of ADCC. Mechanisms of Cytotoxicity. Inhibition of Cytotoxicity. Regulation of ADCC. Role of ADCC in Host Immunity. Summary doi: 10.1038/npg.els.0000498 Introduction Antibody-dependent cell-mediated cytotoxicity (ADCC) was initially described by Moller in 1967 as the ability of natural killer (NK) cells at the time identified only as lymphocytes to destroy antibody-coated target cells (Moller, 1967). A nonphagocytic mechanism, whereby most leucocytes can kill target cells in the absence of complement and without major histocompatibility complex (MHC) restriction thus became apparent. The specificity of ADCC is conferred by target-bound antibody and can be induced by antibody concentrations well below those needed to activate complement-mediated lysis. Fc receptors (FcR) on the plasma membrane of the effector cell are necessary but not sufficient for ADCC activity. Once antibody has bound to the target cell, its interaction with either activating or inhibitory FcR on the cytotoxic effector cell appears to dictate the precise signalling pathways that are stimulated. When target-bound antibodies cross-link activating FcR, a cytotoxic pathway is initiated that results in the release of mediators such as perforin, tumour necrosis factor (TNFa), and reactive oxygen intermediates (ROI) which induce target cell death. Conversely, ligation of inhibitory FcR prevents effector cell activation, resulting in survival of the target cell. Fc Receptors are Trigger Molecules for ADCC As noted above, FcR on the surface of cytotoxic effector cells are required for ADCC (Fanger et al., 1989). Separate genes that are members of the immunoglobulin (Ig) supergene family encode different FcR: FcgR specifically bind IgG, FcaR bind IgA, FceR bind IgE and Fcm/aR bind both IgM and IgA (van de Winkel and Capel, 1996; van Egmond et al., 2001a; Dombrowicz et al., 2000; Kinet and Launay, 2000). It is clear that members of the following FcR classes are all capable, when crosslinked, of activating ADCC: FcgRI (CD64), FcgRII (CD32), FcgRIII (CD16), FcaR (CD89) and FceRI. It has also been shown that engagement of at least one IgG-binding FcR, FcgRIIb, leads to inhibition of ADCC activation (Clynes et al., 2000). The balance between activating and inhibitory FcR thus appears to be a crucial determinant of the magnitude of ADCC that is induced when the effector cell engages the antibody-coated target. See also: Fc receptors; Antibody classes; Antibody function; Hypersensitivity: antibodymediated cytotoxic (Type II); Antibodies Cellular Effectors of ADCC Natural killer cells NK cells are perhaps the most reliable effectors of ADCC when tested in cell culture cytotoxicity assays (Perussia, 2000). When freshly isolated from circulating blood, NK cells express relatively high levels of the type IIIa IgG receptor (FcgRIIIa or CD16a) and very efficiently mediate ADCC of tumour cells that are coated with antitumour IgG. NK cells appear not to express other classes of Fc receptors and, therefore, are not activated for ADCC by IgA, IgE or IgM. FcgRIIIa is present on both NK cells and macrophages as a transmembrane receptor that is associated with the FcR g-chain, an activating molecule involved in signal transduction. This association with the g-chain was found to be essential both for surface expression of FcgRIIIa and for signal transduction following engagement of this receptor by IgG. In contrast, FcgRIIIb ENCYCLOPEDIA OF LIFE SCIENCES & 2006, John Wiley & Sons, Ltd. www.els.net 1
(CD16b) found on human neutrophils is expressed as a phosphatidyl inositol glycan-linked receptor that does not associate with the g-chain and does not activate tumour cell killing. Only FcgRIIIa, and not FcgRIIIb, is found in mice. Monocytes, macrophages and dendritic cells Monocytes and macrophages express all three classes of IgG receptors, FcgRI (CD64), FcgRII (CD32) and FcgRIII (CD16), as well as the receptor for IgA (CD89) (van de Winkel and Capel, 1996; van Egmond et al., 2001a). Each has been shown to be capable of triggering phagocytosis and ADCC by monocytes and macrophages. Although only low levels of FcgRIIIa are expressed on freshly isolated monocytes, this receptor can be markedly increased when monocytes mature into macrophages or are cultured with activating cytokines. Cytokine activation has also been shown to induce expression of FceRI. It has been noted that activated macrophages have high FcgR IIIa/IIb ratios, favouring cell activation, whereas quiescent effectors have low IIIa/IIb ratios, providing a high threshold for cell activation (Ravetch and Lanier, 2000). Human blood dendritic cells also express IgG FcR and are potent effectors in antibody-dependent cellular cytotoxicity (Schmitz et al., 2002). Neutrophils and eosinophils Freshly isolated neutrophils do not lyse tumour targets through FcgR, but can be activated by granulocyte macrophage colony-stimulating factor (GM-CSF) to mediate killing through FcgRIIa, and by interferon g (IFN-g) to mediate killing through FcgRI and FcgRIIa. The type of FcgRIII expressed by human neutrophils (FcgRIIIb) does not mediate cytotoxicity of tumour targets, even following cell activation by cytokines. On the other hand, FcgRIIIa on murine polymorphonuclear neutrophil (PMN) does induce ADCC of tumour cells. Cytolysis of red blood cells occurs via FcgRIIa, FcgRIIIa and FcgRIIIb even when resting human neutrophils are used as effectors. These observations suggest that erythroid and tumour targets are killed through different pathways, and that triggering of FcgRIa or FcgRIIa on neutrophils activates lytic mechanisms distinct from those initiated via FcgRIIIb (Fanger et al., 1989). Of particular interest, neutrophils can efficiently kill human IgA-coated tumour cells and ADCC is highest when both IgA and IgG receptors are coordinately triggered (van Egmond et al., 2001b). Furthermore, freshly isolated neutrophils carry out FcaR-dependent lysis without cytokine preactivation. These findings suggest different pathways of cell activation via FcgR versus FcaR. Although freshly isolated eosinophils express FcgRII as well as low levels of FcgRIII and FceRI, they are not potent ADCC effectors unless activated with GM-CSF, interleukin 3 (IL-3) or IL-5. After cytokine activation, killing of tumour or erythrocyte targets is mediated through FcgRII but not FcgRIII. Thus, like neutrophils, eosinophils require cytokine activation to perform ADCC through FcgRII. Also, like neutrophils, freshly isolated eosinophils are able to mediate ADCC via FcaRI. Eosinophils have also been shown to mediate ADCC of parasites via FceR following activation by cytokines (Dombrowicz et al., 2000). Mechanisms of Cytotoxicity It is now well established that antibodies act as a bridge between FcR on the effector cell and the target antigen on the cell that is to be killed. Crosslinking of receptors is required to activate the cytotoxic event. Even for highaffinity receptors, binding of monomeric Ig is insufficient to transmit an activation signal. The nature of the biological response and the subsequent cytotoxicity initiated by FcR crosslinking depends on the integration of a variety of signals that are dependent on the class and subclass of both the antibody and the FcR involved, on activation of other signalling molecules, on the effector cell type, and on the nature of the target cell. Cell surface molecules other than FcR also may play a role by promoting adhesion between effector and target cells and/or by providing synergistic signals. Crosslinking of activating FcR initiates a signalling cascade that culminates in release of cytoplasmic granules and the synthesis of surface activation molecules including CD69 and CD25. It often begins with the stimulation of immunoreceptor tyrosine-based activation motif (ITAM) sequences (Billadeau and Leibson, 2002; Ravetch and Lanier, 2000). ITAMs have been identified within the cytoplasmic domain of the so-called alpha or immunoglobulinbinding chains of some FcR (e.g. FcgRIIa), and also within associated signalling molecules (e.g. g-chain in the case of FcgRIa, FcgRIIIa, FcaRI and FceRI). Granule release is dependent on activation of the mitogen-activated protein kinase (MAPK) family member ERK (extracellular signalrelated kinase), and includes intermediate interactions involving the p56lck and many other protein tyrosine kinases (reviewed by Perussia, 2000). Many studies have focused on dissecting the mechanisms of target cell destruction mediated by cytolytic T cells (CTL) and NK effector cells. Although the manner in which T cells and NK cells recognize their targets is different, these cells share similar cytotoxic mechanisms. For one such mechanism it is proposed that, upon recognition of target, the contents of intracellular granules within the effector cell are released by a calcium-dependent exocytotic process. Perforin, or cytolysin, is one component that is released from granules. Perforin inserts and polymerizes within the target cell membrane, in a manner that is analogous to the membrane insertion of the terminal components of the complement cascade, resulting in the formation of a pore in the target cell. Calcium is required 2
for the insertion and assembly of perforin in the target cell membrane. A critical role for perforin in host immune defence has been confirmed in perforin-deficient mice. Other components of granules released by exocytosis have also been implicated in the mechanism of CTL and NK cellmediated cytotoxicity. One such granule component, granzyme B, may be involved in deoxyribonucleic acid (DNA) fragmentation observed in target cells after engaging a cytotoxic effector cell. However, studies using granzyme B- deficient mice suggest a less important role for granzyme B in cytotoxicity. Another mechanism of target cell destruction is the engagement of Fas by Fas ligand (FasL), molecules expressed, respectively, on the surface of some target and effector cells. This interaction initiates a cascade of signals in the target cell resulting in apoptosis of the target. Crosslinking of FcgRIIIa on NK cells induces the expression of Fas L, enabling them to kill Fas+ targets. Therefore, FasL/Fas interaction may be an important mechanism of ADCC mediated by NK cells. Myeloid cells are able to destroy antibody-opsonized targets by phagocytosis as well as by extracellular lysis. Targets ingested phagocytically are destroyed in intracellular phagolysosomes of the effector cell. The mechanism(s) by which myeloid cells mediate extracellular lysis of antibody-opsonized target cells has been controversial and may be different from ADCC mediated by NK cells. In one in vitro system, the mechanism of ADCC mediated by myeloid cells has been shown to be distinct from that mediated by NK cells based on divalent cation requirements (Graziano et al., 1989). ADCC mediated by NK cells via FcgRIII was Ca 2+ dependent and Mg 2+ independent. In contrast, ADCC mediated by monocytes or by IFN-g-activated PMN via FcgRIa or FcgRIIa, or by peritoneal macrophages via FcgRIa, FcgRIIa or FcgRIIIa was Mg 2+ dependent and Ca 2+ independent. Thus, even though macrophages and NK cells both mediate ADCC via FcgRIIIa, the mechanism by which these two types of effector cells mediate their killing is distinct. Monocytes, macrophages and neutrophils release ROI when FcR are crosslinked, and ROI have been implicated as the mediators of target cell death. However, other mechanisms must also exist since neutrophils isolated from patients with chronic granulomatous disease CGD, whose cells lack the ability to form ROI, were able to mediate ADCC as well as neutrophils isolated from normal donors (Roberts et al., 1993). Finally, it has been shown that when antibody binds to an apoptotic trigger molecule on the target cell, FcR crosslinking of the target-bound immunoglobulin can induce apoptotic cell death, apparently without a requirement for active participation of the effector cell such as release of toxic mediators by the effector cell. In conclusion multiple pathways may be involved in how an effector cell mediates ADCC of an antibody-opsonized target cell. These include perforin, cytolytic enzymes, reactive oxygen intermediates and apoptotic signalling both with and independent of Fas/FasL interactions. Several mechanisms may combine to mediate cytotoxicity depending on a variety of factors. The relative importance of each in vivo and ex vivo will continue to be dissected using knockout mice and/or effector cells from individuals deficient in specific effector mechanisms. Inhibition of Cytotoxicity Fine tuning between activating and inhibitory pathways is now also understood to be a critical determinant of ADCC activity (Billadeau and Leibson, 2002; Ravetch and Lanier, 2000). FcgRIa (CD64), FcgRIIa (CD32) and FcgRIIIa (CD16a); FcaR (CD89) and FceRI are all capable of mediating ADCC when engaged by their respective ligands and crosslinked. Conversely, an inhibitory signal is generated upon crosslinking of FcgRIIb which contains cytoplasmic immunoreceptor tyrosine-based inhibitory motif (ITIM) sequences, aborting cellular activation through ITAM-containing receptors. The presence or absence of signalling through other immune inhibitory receptors on ADCC effector cells also influences cytotoxic activity. These receptors, with acronyms such as killer inhibitory receptor (KIR) (or CD158), CD94/NKG2 and immunoglobulin-like transcript (ILT) (also called leucocyte inhibitory receptor (LIR) or CD85) associate with SHP-1 and SHP-2 phosphatases to inhibit cell activation. The effect of FcgRIIb on ADCC and tumour growth in vivo was demonstrated in FcgRIIb-deficient mice. Using models of antibody-mediated tumour treatment Clynes et al. (2000) demonstrated that FcgRIIb-deficient mice were able to inhibit tumour growth and protect from tumour metastasis as compared to wild-type mice. This finding confirmed that FcgRIIb acted as an inhibitory receptor. Regulation of ADCC The ability of an effector cell to mediate ADCC depends on a variety of parameters including the type of effector cell, the class of FcR(s) engaged, the number of FcR per cell and the state of activation of the effector cell. Treatment of effector cell populations with cytokines or other physiologic mediators may increase or decrease the expression level of a particular FcR on an effector cell, thereby enhancing or depressing ADCC through that receptor. Cytokine treatment may also change the activation state of an effector cell leading to enhanced or depressed ADCC without altering the number of FcR on the effector cell. The expression of other molecules on the effector cell such as inhibitory receptors or adhesion molecules also impacts the ability of an effector cell to mediate ADCC. It seems likely that the specific tissue microenvironment, which dictates the overall molecular interactions of the FcR-bearing 3
cell, plays a crucial role in the ultimate level of function that is triggered by engagement of FcR. IFN-g is a potent modulator of ADCC. Treatment of monocytes, macrophages or neutrophils with IFN-g increases the expression level of FcgRIa on these cells more than 10-fold. ADCC mediated via FcgRIa is similarly increased, especially when limited antibody is available on the target. Interestingly, even though IFN-g treatment does not increase the expression level of FcgRIIa on neutrophils, it does enable ADCC through FcgRIIa, perhaps by altering the cytotoxic potential of the neutrophil or by changing the FcgRIIa/FcgRIIb ratio. Furthermore, IFN-g induces higher expression of the Fc receptor g-chain which is essential for ADCC triggered through FcgRIa, and may also participate in signal transduction following engagement of FcgRIIa and FcgRIIIa. Enhancement of ADCC by cytokines other than IFN-g has also been demonstrated. GM-CSF and TNF-a enhance ADCC by monocytes cultured in vitro through FcgRI, FcgRII and FcaRI. ADCC mediated by PMN via FcgRII is enhanced by in vitro culture in the presence of G- CSF. Treatment of patients in vivo with granulocyte colony-stimulating factor (G-CSF) or IFN-g leads to upregulation of FcgRI on neutrophils and enhanced ex vivo ADCC via this receptor. Role of ADCC in Host Immunity Early studies showed that tumour-specific antibodies that were effective in promoting ADCC in vitro were also efficacious in curing tumours in vivo using mouse models (Herlyn and Koprowski, 1982). This correlative evidence was confirmed more recently in studies using mice deficient in FcRg. FcRg refers to the g-chain-signalling molecule that was initially described as a subunit of FceRI, but later shown to be required for optimal expression of and signal transduction by FcgRIa and FcgRIIIa. NK cells and macrophages from FcRg / mice are deficient in expression of these activating FcgR and ineffective in mediating ADCC in vitro. The physiologic importance of this deficiency has also been demonstrated in vivo. Functional FcR were required for both passive and active immunity to tumours. Whereas, human IgG1 and murine IgG2a antitumour antibodies dramatically inhibited tumour growth in intact mice, they were without effect in FcRg chain / mice (Clynes et al., 2000). Similarly, when the Fc region of an antitumour antibody was mutagenized to eliminate its binding to FcgRIII, it lost its capacity to inhibit tumour growth in vivo. Evidence implicating the role of ADCC in human host immunity is mostly circumstantial. However, several reports correlating the ability of human effector cells to mediate ADCC, or the presence of ADCC-mediating antibody titres, with the lack of disease progression have been published. For example, in one study, ADCC activity of mononuclear cells isolated from immunocompromised HIV patients correlated with longer survival in these patients (Forthal et al., 1999). In a separate study, the levels of ADCC activity via anti-gp120 antibodies in serum of human immunodeficiency virus 1 (HIV 1) positive individuals correlated with successful host defense as defined by rate of progression to acquired immune deficiency syndrome (AIDS). Furthermore, effector cells isolated from humans that were capable of mediating ADCC in conjunction with antibody to herpes simplex virus (HSV) were able to protect neonatal mice from lethal infection with HSV. Some bacterial pathogens have developed mechanisms to disrupt Ig FcR interactions in order to circumvent ADCC or phagocytosis. These mechanisms provide the pathogens an advantage by interfering with host clearance mechanisms implicating the importance of ADCC and phagocytosis in host immunity. Summary ADCC occurs when antibodies create a bridge between antigenic determinants on target cells and FcR on cytotoxic effector cells. ADCC is activated by the crosslinking of effector cell FcR, which can trigger the destruction of erythrocytes, allogeneic cells, virus-infected cells, tumour cells and parasites. The mechanisms of killing include target cell apoptosis as well as effector cell release of toxic molecules such as perforin, granzyme B, TNFa, reactive oxygen and reactive nitrogen. References Billadeau DD and Leibson PJ (2002) ITAMs versus ITIMs: striking a balance during cell regulation. Journal of Clinical Investigation 109: 161 168. Clynes RA, Towers TL, Presta LG and Ravetch JV (2000) Inhibitory Fc receptors modulate in vivo cytoxicity against tumor targets. Nature Medicine 6: 443 446. Dombrowicz D, Quatannens B, Papin JP, Capron A and Capron M (2000) Expression of a functional Fc epsilon RI on rat eosinophils and macrophages. Journal of Immunology 165: 1266 1271. Fanger MW, Shen L, Graziano RF and Guyre PM (1989) Cytotoxicity mediated by human Fc receptors for IgG. Immunology Today 10:92 99. Forthal DN, Landucci G, Haubrich R et al. (1999) Antibody-dependent cellular cytotoxicity independently predicts survival in severely immunocompromised human immunodeficiency virus-infected patients. Journal of Infectious Diseases 180: 1338 1341. Graziano RF, Erbe DV and Fanger MW (1989) The mechanisms of antibody-dependent killing mediated by lymphoid and myeloid cells are distinct based on different divalent cation requirements. Journal of Immunology 143: 3894 3900. Herlyn D and Koprowski H (1982) IgG2a monoclonal antibodies inhibit human tumor growth through interaction with effector cells. Proceedings of the National Academy of Sciences of the USA 79: 4761 4765. Kinet JP and Launay P (2000) Fc alpha/micror: single member or first born in the family? Nature Immunology 1: 371 372. 4
Moller E (1967) Cytotoxicity by nonimmune allogeneic lymphoid cells. Specific suppression by antibody treatment of the lymphoid cells. Journal of Experimental Medicine 126: 395 405. Perussia B (2000) Signaling for cytotoxicity. Nature Immunology 1: 372 374. Ravetch JV and Lanier LL (2000) Immune inhibitory receptors. Science 290: 84 89. Roberts RL, Ank BJ, Fanger MW, Shen L and Stiehm ER (1993) Role of oxygen intermediates in cytotoxicity: studies in chronic granulomatous disease. Inflammation 17: 77 92. Schmitz M, Zhao S, Schakel K et al. (2002) Native human blood dendritic cells as potent effectors in antibody-dependent cellular cytotoxicity. Blood 100: 1502 1504. van Egmond M, Damen CA, van Spriel AB et al. (2001a) IgA and the IgA Fc receptor. Trends in Immunology 22: 205 211. van Egmond M, van Spriel AB, Vermeulen H et al. (2001b) Enhancement of polymorphonuclear cell-mediated tumor cell killing on simultaneous engagement of fcgammari (CD64) and fcalphari (CD89). Cancer Research 61: 4055 4060. Further Reading Boruchov AM, Heller G, Veri MC et al. (2005) Activating and inhibitory IgG Fc receptors on human Dcs mediate opposing functions. Journal of Clinical Investigation 115: 2914 2923. Colucci F, Di Santo JP and Leibson PJ (2002) Natural killer cell activation in mice and men: different triggers for similar weapons? Nature Immunology 3: 807 813. Lanier LL (2005) NK cell recognition. Annual Review of Immunology 23: 225 274. Perussia B (1998) Fc receptors on natural killer cells. Current Topics in Microbiology and Immunology 230: 63 88. Pleass RJ and Woof JM (2001) Fc receptors and immunity to parasites. Trends in Parasitology 17: 545 551. Takai T (2005) Fc receptors and their role in immune regulation and autoimmunity. Journal of Clinical Immunology 25: 1 18. van de Winkel JG and Capel PJA (1996) Human IgG Fc Receptors. Austin: R.G. Landes Company. 5