Monocyte Subsets Responsible for Immunoglobulin G-Dependent Effector Functions in Vivo

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Immunity Article

Monocyte Subsets Responsible for Immunoglobulin G-Dependent Effector Functions In Vivo

Markus Biburger,1,3 Susanne Aschermann,1,3 Inessa Schwab,1 Anja Lux,1 Heike Albert,1 Heike Danzer,1 Melissa Woigk,1 Diana Dudziak,2 and Falk Nimmerjahn1,* 1Chair of Genetics, Department of Biology, University of Erlangen-Nuremberg, Erwin-Rommel-Str. 3, 91058 Erlangen, Germany 2Laboratory of DC Biology, Department of Dermatology, Nikolaus-Fiebiger-Center for Molecular Medicine, University of Erlangen-Nuremberg, Glu¨ ckstr. 6, 91054 Erlangen, Germany 3These authors contributed equally to this work *Correspondence: [email protected] DOI 10.1016/j.immuni.2011.11.009

SUMMARY mice deficient in the common FcRg chain, essential for cell surface expression and signaling of all activating FcgRs, or Immunoglobulin G (IgG) antibodies confer protection mice deficient in individual activating FcgRs are resistant to the against pathogenic microorganisms, serve as thera- development of disease pathology in active and passive models peutics in tumor therapy, and are involved in destruc- of autoimmune disease and for the activity of therapeutic anti- tion of healthy tissues during autoimmune diseases. bodies in models of IgG-dependent target cell depletion (Hazen- Understanding the molecular pathways and effector bos et al., 1996; Ioan-Facsinay et al., 2002; Kaneko et al., 2006; cell types involved in antibody-mediated effector Meyer et al., 1998; Nimmerjahn et al., 2005; Nimmerjahn and Ravetch, 2005, 2008; Takai, 2002; Takai et al., 1994; Uchida functions is a prerequisite to modulate these activi- et al., 2004). The family of FcgRs consists of three activating ties. In this study we used two independent model (FcgRI, FcgRIII, FcgRIV) and one inhibitory (FcgRIIB) member systems to identify innate immune effector cells in mice and a set of orthologous proteins in nonhuman primates required for IgG activity in vivo. We first defined the and humans (Hirano et al., 2007; Nimmerjahn and Ravetch, precise repertoire of receptors for the IgG Fc frag- 2006). Coexpression of activating and inhibitory FcgRs sets ment (FcgR) on innate immune effector cells in the a threshold for innate immune effector cell activation and regu- blood and on tissue-resident macrophage pop- lates the strength of the initiated responses such as antibody- ulations. Despite expression of relevant activating dependent cellular cytotoxicity (ADCC), phagocytosis of immune FcgRs on various phagocyte populations, our data complexes, and release of proinflammatory mediators. Several indicate that the majority of these cell types are studies have investigated which activating FcgRs are involved dispensable for IgG activity in vivo. In contrast, in mediating the activity of the different IgG subclasses. The picture emerging from this set of data is that the activity of IgG-dependent effector functions were selectively hi lo IgG1 is solely dependent on activating FcgRIII, whereas the impaired in animals lacking the CX3CR1 Ly6C int effector functions triggered by the most potent IgG subclasses CD11c monocyte subset, which expressed the IgG2a or c and IgG2b are largely dependent on FcgRIV or combi- full set of FcgRs required for IgG activity. nations of FcgRI and FcgRIV or FcgRIII and FcgRIV (Baudino et al., 2008; Fossati-Jimack et al., 2000; Giorgini et al., 2008; Hamaguchi et al., 2006; Hazenbos et al., 1996; Meyer et al., INTRODUCTION 1998; Nimmerjahn and Ravetch, 2005; Syed et al., 2009). Depending on the model system, a contribution of the high- Antibodies of the immunoglobulin G (IgG) subclass are of major affinity FcgRI to IgG2a or c activity is observed (Bevaart et al., importance for the control of bacterial and viral infections, the 2006; Ioan-Facsinay et al., 2002; Otten et al., 2008). Confirming killing of tumor cells during antibody therapy, and the destruction the importance of the low-affinity FcgRs for human IgG activity of healthy tissues during autoimmune diseases such as systemic in vivo, several studies with patient cohorts receiving anti- lupus erythematosus (SLE), autoimmune haemolytic anemia tumor antibodies found that patients with high-affinity alleles of (AIHA), autoimmune thrombocytopenia (ITP), and rheumatoid FcgRIIA and FcgRIIIA respond better to therapy (Cartron et al., arthritis (Hogarth, 2002; Nimmerjahn and Ravetch, 2008; Takai, 2002; Weng et al., 2004; Weng and Levy, 2003). Despite this 2002). Despite the capacity of IgG to activate both the classical wealth of information about the role of individual FcgRs for IgG complement pathway and effector cells carrying receptors activity, it remains unclear which effector cells are responsible specific for the IgG Fc fragment (Fcg receptors, FcgRs), studies for these effector functions because most of the potential candi- of many laboratories suggest that FcgRs are responsible date cell populations, such as monocytes, macrophages, and for mediating IgG effector functions in vivo (Nimmerjahn and neutrophils, express an overlapping repertoire of FcgRs (Nim- Ravetch, 2008; Ravetch and Clynes, 1998; Takai, 2002). Thus, merjahn and Ravetch, 2008). Complicating the situation, there

932 Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc. Immunity Ly6Clo Monocytes and IgG Effector Functions

are several monocyte and macrophage subpopulations in the Geissmann et al., 2010; Ziegler-Heitbrock et al., 2010). As shown blood and peripheral tissues such as lung, liver, and spleen in Figure 1A, FcgRIV expression was restricted to the Ly6Clo with an unknown FcgR expression pattern that might participate monocyte subset in the blood, spleen, and bone marrow, identi- in these effector functions (Geissmann et al., 2010; Gordon and fying this activating FcgR as a marker for this monocyte popula- Taylor, 2005; Nimmerjahn and Ravetch, 2008). Thus, we set out tion. A recent study has identified this monocyte subset to be to define the FcgR expression pattern on these different cell pop- expanded in the autoimmune BXSB mouse strain, suggesting ulations and to identify which of them could be involved in IgG- a role for this cell type in autoimmune inflammation (Santiago- mediated platelet phagocytosis as a model system to study Raber et al., 2009, 2011). In general, Ly6Clo monocytes ex- the activity of an autoantibody and IgG-dependent B cell deple- pressed the broadest range of FcgRs including all activating tion, as a model for therapeutic IgG applications in vivo. This FcgRs and the inhibitory FcgRIIB. As shown before, the only analysis revealed that only select monocyte and macrophage other cell type expressing FcgRIV were neutrophils, although populations expressed the relevant set of receptors previously at a lower amount. In contrast to this rather selective expression identified to be important for IgG subclass activity (Baudino pattern of FcgRIV, FcgRIII and FcgRIIB were broadly expressed, et al., 2008; Giorgini et al., 2008; Hamaguchi et al., 2006; Hirano although the amount of expression differed between cell types et al., 2007; Nimmerjahn et al., 2005; Nimmerjahn and Ravetch, (Figure 1A). On tissue macrophages, FcgRIV was detectable 2005; Syed et al., 2009). By using a variety of genetically ablated on red pulp macrophages in the spleen and on Kupffer cells in mouse strains and target cell depletion approaches, we have the liver (Figure 1B). Together with previous data demonstrating hi lo shown that the CX3CR1 Ly6C monocyte subset (also referred FcgRIV expression on alveolar macrophages, these results to as nonclassical or resident subset) expresses the relevant set suggest that this activating FcgR is expressed on tissue-resident of activating and inhibitory FcgRs and is critically involved in macrophages and the Ly6Clo monocyte subset (Mancardi et al., IgG-dependent effector functions in vivo. 2008; Syed et al., 2009).

RESULTS Involvement of Tissue-Resident Macrophages in ITP In chronic human ITP, removal of the spleen can stabilize platelet FcgR Expression Pattern on Monocyte and Macrophage counts in about 60%–70% of patients, which has been used as Cell Populations indirect evidence that splenic-resident macrophages might be Given the important role of selective FcgRs for IgG-mediated involved in autoantibody-mediated platelet phagocytosis (Bib- effector functions, we reasoned that identification of the FcgR urger et al., 2010; Cines and McMillan, 2005; Mizutani et al., expression repertoire on innate immune effector cells might 1992). Moreover, liver-resident macrophages (Kupffer cells) are provide an insight into responsible effector cell populations. suggested to be involved in removal of opsonized platelets and Most importantly, the precise expression pattern of mouse acti- do express the relevant activating FcgRIV (Figure 1B; Coetzee vating FcgRIV, which is essential for IgG2a- and IgG2b-medi- et al., 2000; Heyns et al., 1980; Schmidt and Rasmussen, ated effector functions in a wide variety of in vivo model systems, 1985; Stratton et al., 1989). To analyze whether autoantibody has not been determined in detail so far (Baudino et al., 2008; deposits in the spleen and liver, we labeled the anti-platelet anti- Giorgini et al., 2008; Hamaguchi et al., 2006; Nimmerjahn body 6A6-IgG2a and an isotype control antibody with quantum et al., 2005, 2010; Nimmerjahn and Ravetch, 2005, 2006; Otten dots that can be identified by fluorescent microscopy in tissue et al., 2008; Syed et al., 2009). An initial characterization of the sections after antibody injection. Indeed, the autoantibody but expression pattern of FcgRIV demonstrated that blood mononu- not the isotype control could be identified in the red pulp of the clear cells, neutrophils, and ex-vivo-generated bone-marrow- spleen and in the liver (Figures 2A and 2C). To investigate derived monocytes and thioglycolate-elicited macrophages did whether the spleen is required for IgG-mediated platelet deple- express FcgRIV (Hirano et al., 2007; Nimmerjahn et al., 2005). tion, we splenectomized mice and investigated the capacity of It remained unclear, however, on which tissue-resident macro- two 6A6 subclass switch variants to deplete platelets in these phages and monocyte subsets in the blood, bone marrow, and animals. Surprisingly, neither of the two IgG subclasses was spleen this receptor was expressed and how the expression dif- impaired in its activity in vivo, suggesting that splenic-resident fered between these cells. Based on the expression of cell phagocytic cell populations were not required for platelet surface markers including the chemokine receptors CX3CR1, depletion in this acute model system in vivo (Figure 2B). Further CCR2, and accompanying markers CD11c and Ly6C (Gr-1), analysis of liver tissue sections showed that the deposited auto- monocytes can be distinguished into two major subsets (Geiss- antibodies were not localized close to F4/80-positive Kupffer mann et al., 2010; Gordon and Taylor, 2005; Ziegler-Heitbrock cells but rather in proximity to endothelial cells (Figures 2C hi lo hi et al., 2010). The Ly6C CX3CR1 CD11c CCR2 subset (also and 2D). Because endothelial cells in the liver selectively ex- called inflammatory or classical monocytes) has been shown pressed the inhibitory FcgRIIB (Figure S1 available online), they to be recruited to tissues during inflammation (Geissmann may not contribute to platelet removal according to previous lo hi et al., 2003; Serbina and Pamer, 2006). The Ly6C CX3CR1 studies obtained in FcgRIIB-deficient mice, showing that the CD11cintCCR2lo subset (also called resident or nonclassical inhibitory Fc receptor is a negative regulator of platelet depletion monocytes) has been initially hypothesized to be important for rather than contributing to platelet phagocytosis (Nimmerjahn maintaining the repertoire of tissue macrophages during the and Ravetch, 2005). Consistently, depletion of liver-resident steady state, but recent data tend to indicate that they are local- macrophages by clodronate liposomes did not impair auto- ized mainly at the endothelium where these cells may have antibody deposition (Figure 2D). The importance of hematopoi- a gatekeeper function (Auffray et al., 2007; Cros et al., 2010; etic cells in IgG-mediated platelet phagocytosis was further

Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc. 933 Immunity Ly6Clo Monocytes and IgG Effector Functions

A FcγRI FcγRIII FcγRIV FcγRIIB

Neutrophils Blood Eosinophils Ly6Chi monocytes Ly6Clo monocytes Lymphocytes 0 1 2 0 7.5 15 0 10 20 0 5 10 oemarrow Bone Neutrophils Eosinophils Ly6Chi monocytes Ly6Clo monocytes Lymphocytes 0 1 2 0 7.5 15 0 10 20 0 5 10

Neutrophils Spleen Eosinophils Ly6Chi monocytes Ly6Clo monocytes Lymphocytes 0 1 2 0 7.5 15 0 10 20 0 5 10

ΔMFI [x10³ AU]

B F4/80 9E9 (FcγRIV) Overlay Spleen Liver

Figure 1. FcgR Expression Pattern on Innate Immune Effector Cells (A) The relative expression of all activating FcgRs (FcgRI, FcgRIII, FcgRIV) and the inhibitory FcgRIIB was determined on a variety of innate immune effector cells including neutrophils, eosinophils, the Ly6Clo and the Ly6Chi monocyte subsets, and lymphocytes in peripheral blood, bone marrow, and spleen. Shown is the increase in the median fluorescence intensity (DMFI) caused by the respective fluorochrome-labeled FcgR-specific antibodies. (B) Identification of FcgRIV expression on tissue-resident macrophages in the spleen and the liver. Macrophages were identified by staining with F4/80, and FcgRIV expression was identified with the monoclonal antibody 9E9. The scale bars represent 100 mm. Shown are representative pictures from three independent experiments. confirmed by using bone marrow chimeras. As shown in Fig- deficient animals. Despite the absence of CD11b, IgG-mediated ure 2E, platelet phagocytosis was restored if common FcRg platelet depletion was not impaired (Figure 2F). Similar results / chain-deficient mice (Fcer1g ), which are resistant to IgG were obtained in CX3CR1-deficient mice as shown in Figure 2G. subclass-dependent platelet depletion, were reconstituted with bone marrow of wild-type mice. Conversely, no platelet deple- Mast Cells, Basophils, Eosinophils, and Neutrophils Are tion was observed in wild-type mice reconstituted with bone Not Required for IgG-Dependent Platelet Depletion marrow of g chain-deficient animals. To investigate whether Mast cells are essential for the initiation of many IgG-dependent recruitment of neutrophils to peripheral tissues was important proinflammatory reactions, and the absence of mast cells abro- for platelet removal, we induced platelet depletion in CD11b- gates the induction of joint inflammation in a model of serum

934 Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc. Immunity Ly6Clo Monocytes and IgG Effector Functions

A B Iso-IgG2a-Qdot CD31 Ly6G 6A6-IgG2a-Qdot CD31 Ly6G 100 C57BL/6 C57BL/6 splenectomized 80

0 Residual platelet count [%] 6A6-IgG1 6A6-IgG2a

C D PBS Clodronate (150 l) Iso-IgG2a-Qdot CD31 F4/80 6A6-IgG2a-Qdot F4/80 20 min 6A6 20 min

6A6-IgG2a-Qdot CD31 F4/80

6A6-IgG2a-Qdot CD31 120 min 6A6 120 min

E F G Cd11b-/- Cx3cr1GFP/GFP 100 125 125

80 100 100

60 75 75

40 50 * 50 * 20 25 25 Residual plateletResidual count [%] Relative plateletRelative count [%] 0 plateletRelative count [%] 0 0 -/- -/- -/- 0 hr 4 hr 0 hr 4 hr C57BL/6 C57BL/6 Fcer1g Fcer1g Fcer1g + C57BL/6 +

Figure 2. Involvement of Spleen- and Liver-Resident Phagocytes in Autoantibody Activity (A) Deposition of the quantum dot (Qdot)-labeled platelet-specific antibody 6A6-IgG2a or an isotype control antibody in the spleen 20 min after intravenous injection. Endothelial cells were identified by staining with a CD31-specific antibody and neutrophils were identified by staining with a Ly6G-specific antibody. The inset shows a selected area of CD31 and 6A6-IgG2a-Qdot staining. (B) Shown is the platelet-depleting activity of the 6A6-IgG1 and 6A6-IgG2a platelet-specific antibodies in C57BL/6 mice either left untreated (n = 5) or splenectomized (n = 3). (C) Deposition of the quantum dot (Qdot)-labeled platelet-specific antibody 6A6-IgG2a or an isotype control antibody in the liver 20 min after intravenous injection. As before, CD31 was used as an endothelial marker and F4/80 as a marker for liver-resident macrophages. (D) Deposition of quantum dot (Qdot)-labeled 6A6-IgG2a in the liver of PBS or clodronate liposome-pretreated mice at the indicated time points after intravenous injection. Tissue-resident macrophages were identified by staining with an F4/80-specific antibody. See also Figure S1.

Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc. 935 Immunity Ly6Clo Monocytes and IgG Effector Functions

transfer arthritis and impairs the Arthus reaction (Lee et al., 2002; by the treatment (Figures 5A–5C). This selective reduction of Nigrovic et al., 2007). To investigate whether mast cells or mast Ly6Clo monocytes translated into impaired autoantibody activity cell-derived cytokines are involved in IgG-dependent platelet (Figure 5D). Similar results were obtained in Cd11c-DTR mice, depletion, we induced ITP in mast cell-deficient Kit/ mice or which express the simian diphtheria toxin receptor under the in mice lacking the receptors for the proinflammatory cytokines control of the Cd11c promoter (Jung et al., 2002). Besides den- IL-1 or TNF-a. Compared to wild-type animals, no difference in dritic cells, CD11c is selectively expressed on the Ly6Clo but not the activity of platelet depletion was observed, suggesting that on the Ly6Chi monocyte subset or on tissue-resident macro- mast cells were not involved in anti-platelet IgG activity (Fig- phages of the liver and spleen (Hashimoto et al., 2011). Indeed, ure 3A). Besides Ly6Clo monocytes, neutrophils do express injection of diphtheria toxin resulted in a reduction of Ly6Clo FcgRIV and may have the capacity to phagocytose opsonized blood monocytes but did not deplete liver-resident Kupffer cells material. To specifically deplete neutrophils, we used the or splenic red pulp macrophages as shown before (Figures Ly6G-specific antibody 1A8. As shown in Figures 3B and 3C, S3A–S3C; Jung et al., 2002; Probst et al., 2005). Consistent injection of this antibody diminished neutrophil numbers by with the results of the clodronate experiment, this translated more than 90%, without affecting other cell populations. Despite into reduced autoantibody activity (Figure 5E). Further support- the absence of neutrophils, however, IgG-dependent platelet ing these data, mice deficient in FcgRI and FcgRIV were resistant depletion was not affected compared to mock-treated animals to IgG2a-dependent platelet depletion (Figure 5F). As shown in (Figure 3D). Similar results were obtained in mice deficient for Figure 1A, this combination of activating FcgRs is selectively the transcription factor GATA1, which lack eosinophils as well expressed on Ly6Clo monocytes in the blood. Moreover, when as after depletion of basophils with an FcεRI-specific antibody plotting the cell numbers of neutrophils and Ly6Chi and Ly6Clo (Figures 3E–3G). Thus, this set of experiments suggests that monocytes of all our previous experiments against the platelet- neither mast cells, basophils, eosinophils, nor neutrophils are depleting activity of the autoantibody, a clear correlation involved in platelet removal. between IgG activity and cell number could be seen only for Ly6Clo monocytes (Figure S3D). Involvement of Monocyte Subpopulations in IgG Activity Previous studies have shown that both monocyte subsets can In Vivo phagocytose particulate matter (Nahrendorf et al., 2007; Panizzi To identify cells that could bind the autoantibody, we injected et al., 2010; Tacke et al., 2006). To study the endocytic and into mice fluorescently labeled normal or mutated anti-platelet phagocytic activity of both monocyte subsets in the blood during antibodies deficient in FcgR binding because of absent IgG the steady state, we first injected liposomes containing the fluo- heavy chain glycosylation (N297A). Analysis of autoantibody rescent dye PKH26, which can be detected by flow cytometric binding to different cell populations expressing FcgRs 20 min analysis upon uptake into cells. As shown in Figures 6A–6C, after injection showed binding to both Ly6Chi and Ly6Clo mono- Ly6Clo monocytes in the blood had the highest capacity to cyte populations in the blood with only minor binding to other acquire the labeled liposomes by fluid phase endocytosis. In innate immune effector cells including neutrophils (Figures 4A contrast, the same monocyte population was less active in lipo- and 4B). Depletion of both monocyte subsets (and all tissue- some uptake in the bone marrow (Figure S4). To address the resident macrophages) resulted in a block of autoantibody capacity to phagocytose particulate matter, mice were injected activity indicating that either or both of these monocyte subsets with fluorescently labeled beads. Under these conditions, both are crucial for IgG activity (Figures S2A–S2C). To distinguish Ly6Chi and Ly6Clo monocyte subsets were able to phagocytose between the contribution of the Ly6Clo and the Ly6Chi monocyte the beads although the latter showed a trend toward a higher subset for IgG activity, we specifically depleted the latter sub- phagocytic activity during the first 2 hr, consistent with previous population with the CCR2 chemokine receptor-specific antibody reports (Figure 6D; Tacke et al., 2006). MC-21. As shown in Figures 4C and 4D, injection of this antibody resulted in an efficient and highly specific removal of Ly6Chi Role of Ly6Clo Monocytes for IgG-Dependent B Cell monocytes without a major impact on the Ly6Clo monocyte pop- Depletion ulation, neutrophils, and eosinophils. Despite the absence of To establish a broader role of the Ly6Clo monocytes for IgG these cells, the efficiency of IgG-dependent platelet depletion re- activity, we used an IgG2a antibody recognizing mouse CD20 mained unchanged, arguing against a major involvement of to deplete B cells. As shown for other IgG2a antibodies, B cell these cells (Figures 4E and 4F). To achieve a selective depletion depletion by this antibody occurred in an Fcg-dependent manner of the FcgRIV-positive Ly6Clo monocyte subset, we titrated the (Figures 7A, 7B, and S5; Hamaguchi et al., 2006). Consistent with amount of clodronate liposomes down to a single injection of our results in the platelet depletion model, B cell depletion 10 ml 2 hr before autoantibody injection. At this dose, the number in the blood was not affected by the absence of splenic-resident of Ly6Clo monocytes was reduced to about 15%–20% of its orig- monocytes and macrophages, neutrophils, or the Ly6Chi mono- inal number, whereas neither tissue-resident macrophages in cyte subset (Figures 7C–7E). In contrast, selective depletion the liver nor Ly6Chi monocytes and neutrophils were affected of Ly6Clo monocytes by injection of low-dose clodronate

The scale bars represent 50 mm. (E) Platelet-depleting activity of the 6A6-IgG2a antibody in C57BL/6 (n = 5), Fcer1g/ (n = 5), Fcer1g/ mice transplanted with C57BL/6 bone marrow (n = 3) and C57BL/6 mice transplanted with bone marrow of Fcer1g/ (n = 3) animals. / (F and G) Shown are the relative platelet counts 4 hr after injection of the 6A6-IgG2a antibody in CD11b-deﬁcient (Cd11b ) (n = 4) and CX3CR1-deﬁcient (Cx3cr1GFP/GFP) (n = 4) animals. *p < 0.05.

936 Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc. Immunity Ly6Clo Monocytes and IgG Effector Functions

NK Ly6Chi Neu Eo A B low SSC CD11bhi 100 PBS

50 1A8

25 CD62L Gr-1 Residual platelet count [%] 0 - / -/- 6 -/- - -/- lo / a - t CD11b Ly6C SSC g 1 -/ i b K 1r1 1 Il C57BL Fcer1 Tnfrsf Tnfrsf

C D PBS 1A8 0 hr 4 hr 0.8 100 /µl] 0.6 4 75 /µl] 3

0.4 50

0.2 25 * * Cell count [x10 Platelet count [x10 0.0 0 PBS 1A8 lo hi hils C C p Ly6 Ly6 monocytes monocytesEosino Neutrophils

E F G Gata1-/- anti- FcεRI 150 0.25 150 /µl]

/µl] 125 4 125 0.20 4

100 100 0.15 75 75 0.10 50 50

0.05 25 Basophil count [%] 25 Platelet count [x10 Platelet count [x10 * * * 0 0 0 0 hr 4 hr PBS MAR-1 0 hr 4 hr

Figure 3. Involvement of Mast Cells, Neutrophils, Basophils, and Eosinophils in Autoantibody-Dependent Platelet Depletion (A) Shown is the residual platelet count 4 hr after injection of the 6A6-IgG2a antibody in the indicated mouse strains. (B) Flow cytometric analysis of cell populations in the peripheral blood of PBS-treated mice or animals injected with the neutrophil-specific antibody 1A8. Natural killer (NK) cells and Ly6Clo and Ly6Chi monocytes were identified by low side scatter and by staining with CD62L, CD11b, and NK1.1 (not shown) (left). Neutrophils (Neu) and eosinophils (Eo) were identified by high expression of CD11b, Gr-1, and high side scatter (SSC) (right). (C) Numbers of Ly6Clo and Ly6Chi monocytes, neutrophils, and eosinophils in the peripheral blood of PBS- or 1A8-treated animals 36 hr after neutrophil depletion (n = 3). (D) Shown is the platelet count 4 hr after injection of the 6A6-IgG2a antibody in mice pretreated with PBS (n = 3) or 1A8 to deplete neutrophils (n = 3). (E) Gata1/ mice lacking eosinophils were injected with 6A6-IgG2a and platelet counts were determined before and 4 hr after antibody injection. (F) Shown are the basophil counts in the peripheral blood of mice 24 hr after injection of the FcεRI-specific MAR-1 antibody. (G) Platelet counts in the peripheral blood of mice that were depleted of basophils before (0 hr) and 4 hr after induction of ITP by injection of 6A6-IgG2a. *p < 0.05. liposomes or deletion of FcgRI and FcgRIV selectively coex- of the two monocyte subsets and neutrophils to kill B cells via pressed on this monocyte subset efficiently interfered with B antibody-dependent cellular cytotoxicity (ADCC), we isolated cell depletion (Figures 7F and 7G). To directly assess the capacity these different cells from the peripheral blood by cell sorting

Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc. 937 Immunity Ly6Clo Monocytes and IgG Effector Functions

0 hr 4 hr 0 hr 4 hr

Figure 4. Involvement of Ly6Chi Monocytes in Autoantibody-Mediated Platelet Depletion (A and B) Binding of the fluorescently labeled platelet-specific antibodies 6A6-IgG2a (A) and 6A6-IgG1 (B) to the indicated cell populations in the blood 20 min after injection. Aglycosylated antibody variants deficient in FcgR binding (N297A) served as controls. One out of three representative experiments is shown. (C) Flow cytometric analysis of cell populations in the peripheral blood after injection of an isotype control antibody or the MC21 antibody recognizing CCR2 on Ly6Chi monocytes. Natural killer (NK) cells, Ly6Clo monocytes, and Ly6Chi monocytes were identified by low side scatter and by staining with CD62L, NK1.1, and CD11b (left). Neutrophils (Neu) and eosinophils (Eo) were identified by high expression of CD11b, high side scatter (SSC), and the presence (Neu) or absence (Eo) of Gr-1 and FcgRIV (right). (D) Quantification of neutrophils, eosinophils, and Ly6Clo and Ly6Chi monocyte subsets after injection of MC21 or an isotype-matched control antibody. (E and F) Shown is the platelet count before and 4 hr after injection of the 6A6-IgG1 (E) and 6A6-IgG2a (F) antibodies in mice pretreated with MC21 (n = 3) or the isotype control (n = 3). See also Figure S2. and performed ADCC assays with sorted B cells in vitro. As de- DISCUSSION picted in Figure 7H, Ly6Clo monocytes had the greatest capacity to kill B cells, suggesting that this monocyte subset may be of It is generally accepted that cellular FcgRs are crucial for general importance for IgG-dependent effector functions in vivo. IgG-dependent effector functions including phagocytosis,

938 Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc. Immunity Ly6Clo Monocytes and IgG Effector Functions

lo AB CFigure 5. Involvement of Ly6C Monocytes in Autoantibody-Mediated Platelet Deple- μl) tion (A) Quantiﬁcation of cell numbers of Ly6Chi and Ly6Clo monocytes in mice that were either left untreated or injected with PBS or with 10 ml of clodronate liposomes to deplete the Ly6Clo monocyte subset. (B) Shown are liver sections stained with an F4/80 antibody speciﬁc for liver-resident macrophages 2 hr after injection of 10 ml clodronate liposomes or PBS. (C) Blood neutrophil counts in mice left either untreated or injected with PBS or 10 ml of clodronate liposomes 2 hr after injection. (D) Platelet depletion activity of the 6A6-IgG2a DE Fantibody in mice either left untreated or pretreated with PBS or clodronate liposomes 2 hr before antibody injection. (E) Platelet-depleting activity of the 6A6-IgG2a antibody in the indicated mouse strains 24 hr after pretreatment with diphtheria toxin (+DT) or PBS (DT). (F) Residual platelet count in C57BL/6 and Fcgr1/Fcgr4/ mice 4 hr after injection of the 6A6-IgG2a anti-platelet antibody. *p < 0.05. See also Figure S3.

amount on neutrophils and on tissue-resident macrophages. Based on sequence similarity, FcgRIV may be the orthologous antibody-dependent cell-mediated cytotoxicity (ADCC), and the receptor to human FcgRIIIA, because the expression pattern on release of a variety of cytokines, chemokines, and cytotoxic Ly6Clo monocytes is consistent with the corresponding nonclas- substances (Nimmerjahn and Ravetch, 2008; Takai, 2002). sical or resident monocyte subset in humans (Gordon and Because the members of this receptor family are widely ex- Taylor, 2005; Nimmerjahn et al., 2005; Ziegler-Heitbrock et al., pressed on cells of the innate immune system and especially 2010). The importance of FcgRIIIA for human ITP is shown by on phagocytic lineages such as monocytes, macrophages, the efficient inhibition of platelet depletion in patients and in and neutrophils, the identification of cell populations partici- transgenic mice after infusion of blocking antibodies for this re- pating in these IgG-dependent activities is demanding. Under ceptor (Clarkson et al., 1986; Samuelsson et al., 2001). With conditions of acute or chronic inflammation, different cell types respect to B cell depletion, several studies indicate that patients become recruited to sites of inflammation, making it difficult to carrying FcgRIIIA alleles with increased affinity for human IgG1 delineate which of the individual cell types are important for anti- show better therapeutic outcomes (Cartron et al., 2002; Weng body-dependent effects. Therefore, we chose platelet- and B et al., 2004; Weng and Levy, 2003). cell-depletion model systems with a fast kinetic excluding the Our data, acquired by a variety of pharmacological, sur- need for inflammation- and cytokine-dependent recruitment of gical, and genetic approaches, suggest that the FcgRIV+ cells to sites of IgG activity. This short assay time enabled us CD11cintLy6Clo monocyte subset in the blood may be important to use a variety of transient target cell-depletion strategies that for IgG-dependent platelet removal and B cell depletion in vivo. are difficult to use in models that require more time for the However, previous studies suggest that tissue-resident phago- readout and the consecutive recruitment of different cell types. cyte populations of the spleen and liver might be involved in A similar kinetic of antibody-mediated platelet or B cell depletion autoantibody-mediated platelet depletion, and removal of the can be observed in healthy human volunteers injected with spleen is an effective treatment in humans with a therapeutic serum from patients with ITP (Sprague et al., 1952) or in tumor success rate of 60%–70% (Bussel, 2006; Cines and McMillan, patients treated with B cell-depleting antibodies such as Rituxi- 2005; Heyns et al., 1980; Mizutani et al., 1992; Schmidt and Ras- mab. Starting from our previous studies that identified FcgRIV as mussen, 1985; Stratton et al., 1989). There are several possible one of the important activating Fcg receptors for IgG2a activity in explanations for these contradictory results. The most apparent the two model systems, we now performed a detailed analysis of one is that the model system used in this study is an acute and the cellular distribution of all Fcg receptors (Hamaguchi et al., only transient model with a fast recovery of platelet numbers. 2006; Nimmerjahn et al., 2005, 2007; Nimmerjahn and Ravetch, In ITP patients, autoantibodies are produced constantly by 2005). These experiments showed that FcgRIV is selectively ex- plasma cells, which might result in the recruitment of additional pressed on the Ly6CloCD11cint monocyte subset, at a lower cell types of the mononuclear phagocyte system. Moreover,

Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc. 939 Immunity Ly6Clo Monocytes and IgG Effector Functions

A B Figure 6. Endocytic and Phagocytic Activity of 0 min 20 min 700 0 min Monocyte Subsets 20 min Ly6Chi Ly6Clo Ly6Chi Ly6Clo 600 (A) Shown is the uptake of ﬂuorescently labeled (PKH26) lo hi 500 liposomes by Ly6C and Ly6C monocyte subsets of the 400 peripheral blood before (0 min) and 20 min after intrave- MFI 300 nous injection. 200 (B) Quantiﬁcation of PKH26-labeled liposome uptake by the indicated cell types in the peripheral blood by analysis

PKH26 100 0 of their median ﬂuorescence intensity (MFI) prior and FcγRIV 20 min after injection. (C) Kinetics of the uptake of PKH26-labeled liposomes by lo hi

NK cells the indicated cell types in the peripheral blood. See also monocytes Ly6C monocytes Ly6C Neutrophils Eosinophils Figure S4. Lymphocytes (D) Phagocytosis of ﬂuorescently (ﬂash-red) labeled beads by Ly6Chi and Ly6Clo monocytes at the indicated time C Ly6Clo monocytes D lo Ly6C monocytes points after intravenous injection. Ly6Chi monocytes Ly6Chi monocytes

7500 8

6 5000 the in vivo data, sorted Ly6Clo monocytes were hi 4 superior to neutrophils or Ly6C monocytes in 2500 cells [%] antibody-dependent killing of B cells in vitro.

PKH26 median 2 fluorescence [AU] Flash Red-positive Taken together, our experiments demon-

0 0 strate that despite widespread expression of 0 30 60 90 120 0 30 60 90 120 FcgRs on innate immune effector cells, a rather Time after injection [min] Time after injection [min] select subpopulation of monocytes may be crucial for antibody-mediated platelet phagocytosis and B cell depletion in mice in vivo. In the removal of the spleen might stabilize platelet counts not only light of the recent finding that an FcgRIV-positive monocyte by removing resident phagocytic cell populations but also by subset is selectively expanded in lupus-prone mice, it seems impairing the generation of autoantibodies by plasma blasts or possible that this cell type is centrally involved in mediating plasma cells. Moreover, the accumulation of radioactivity in the IgG-dependent tissue pathology (Santiago-Raber et al., 2009, spleen and liver observed upon injection of radiolabelled plate- 2011). Despite functional overlaps between mouse and human lets is an indirect readout and does not prove that these sites FcgRs and the respective mouse and human monocyte subsets, are dominant players in IgG-dependent platelet removal. We there are also important functional differences, necessitating also observed an accumulation of autoantibody in spleen and further studies before definitive conclusions can be drawn for liver, although it was not required for removing the majority of the human system (Gordon and Taylor, 2005; Ziegler-Heitbrock platelets from the circulation and occurred in the absence of et al., 2010). organ-resident macrophages. Further evidence arguing against an involvement of liver-resident Kupffer cells for IgG-mediated EXPERIMENTAL PROCEDURES platelet depletion is their lack of FcgRIIB expression, which has been identified as a negative regulator of IgG (autoantibody Mice and therapeutic IgG) activity in vivo (Kumar et al., 2006; Nimmer- Female mice at 8–15 weeks of age on the C57BL/6 background obtained from Janvier (Le Genest-Saint-Isle, France) were used in all experiments. Kit- jahn and Ravetch, 2005). Formally excluding a role of the liver or deficient (Kit / ), GATA1-deficient (Gata1 / ), CD11b-deficient (Cd11b / ), liver-resident cells is challenging, because in contrast to the interleukin-1-receptor-deficient (Il1r1/), and tumor necrosis factor receptor spleen it is impossible to remove this organ without affecting a- and b-deficient (Tnfrsf1a/ Tnfrsf1b/) mice were obtained from the Jack- viability. Moreover, it is possible that cell types such as neutro- son Laboratories. FcRg-chain-deficient mice (Fcer1/) were provided by phils or classical monocytes might participate in an indirect J. Ravetch. Cx3cr1GFP/GFP and Cd11c-DTR transgenic mice were kindly cooperative manner in these IgG-mediated effector functions. provided by D. Littman through Jackson Laboratories. All mice were main- tained under specific-pathogen-free conditions and according to the guide- With respect to IgG-dependent B cell depletion, our results are lines of the National Institutes of Health and the legal requirements in Germany in line with previous reports suggesting that myeloid cells might and the USA. be important for ADCC activity in mice. For example, ADCC reactions mediated by IgG2a antibodies are found to be depen- In Vivo Experiments dent on activating FcgRIV and are enhanced in mice lacking ITP was induced by intravenous injection of 4 mg of 6A6 IgG subclass FcgRIIB (Hamaguchi et al., 2006; Nimmerjahn et al., 2010; Nim- switch variants as described (Nimmerjahn et al., 2005; Nimmerjahn and merjahn and Ravetch, 2005). By using an Fcgr1/Fcgr4/ Ravetch, 2005). Platelet counts were determined by one of two methods: mouse, we have shown that both of these receptors are essential (1) blood was diluted in PBS and 5% BSA (1:2 to maximum 1:10 dilution) and analyzed in an Advia 120 hematology system (Siemens), or (2) blood for IgG activity. This FcgR combination is selectively expressed lo hi was diluted 1:100, stained with anti-CD61-FITC (dilution 1:50–1:100), and on Ly6C monocytes but not on the Ly6C subset, not on neu- analyzed in a final dilution of 1:1,000 or 1:2,000 by flow cytometry. In the trophils, natural killer (NK) cells, or other innate immune effector latter method, platelets were identified by their light scatter characteristics cells in the blood, spleen, and bone marrow of mice. In line with (determining forward scatter and side scatter on a logarithmic scale)

940 Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc. Immunity Ly6Clo Monocytes and IgG Effector Functions

A B C C57BL/6 Fcer1g-/- Splenectomized 150 150 150

100 100 100

50 50 50 Peripheral B cells [%] Peripheral B cells [%] 0 Peripheral B cells [%] 0 0 0624 0624 0624 Time [hr] Time [hr] Time [hr]

D E F 1A8 MC21 10 µl Clodronate 150 150 150

100 100 100

50 50 50 Peripheral B cells [%] Peripheral cells B [%]

0 0 Peripheral cells B [%] 0 0624 0624 0624 Time [hr] Time [hr] Time [hr]

G H Fcgr1-/- Fcgr4-/- Ly6Clo monocytes Ly6Chi monocytes Neutrophils 150 100

100 75

50 25

0 Peripheral B cells [%] 0 B cell depletion [%] 0624 -25 1:2 5:1 10:1 1:2 5:1 10:1 1:2 5:1 10:1 Time [hr] E/T E/T E/T

CD20-IgG2a Isotype

Figure 7. Role of Monocyte Subpopulations in IgG-Dependent B Cell Depletion (A–G) Shown is the percentage of peripheral blood B cells 6 and 24 hr after injection of 100 mg of an CD20-specific IgG2a antibody in C57BL/6 mice (A), Fcer1g/ mice (B), splenectomized mice (C), mice depleted of neutrophils (D), Ly6Chi monocytes (E), Ly6Clo monocytes (10 ml clodronate) (F), or in mice deficient for the activating FcgRs I and IV (Fcgr1/ Fcgr4 /) (G). In all experiments four to five mice per group were used. See also Figure S5. (H) Shown is the capacity of Ly6Clo monocytes, Ly6Chi monocytes, and neutrophils to mediate antibody-dependent killing of B cells in vitro at the indicated effector to target cell ratios in the presence of the CD20-specific IgG2a or an isotype control antibody. Shown is one out of three independent experiments. together with positive staining for CD61, and their relative numbers per Depletion of Innate Immune Effector Cells blood volume were calculated based on the platelet numbers counted in For depletion of neutrophils, mice were injected with 0.5 mg of the a defined period at constant flow rates. B cell depletion was induced by Ly6G-specific antibody (clone 1A8; provided by T. Malek via BioXcell) 30 hr intravenous injection of 100 mg of a CD20-specific antibody generated in prior to ITP induction as described previously (Daley et al., 2008). For depletion the institute by T. Winkler. B cell counts were determined before and 6 or of Ly6Chi monocytes, mice were injected intraperitoneally with 50 mgof 24 hr after antibody injection by FACS analysis. B cells counts before anti- a CCR2-specific antibody (clone MC-21) the day before ITP induction. Baso- body injection were set to 100%. phil depletion was achieved by two daily injections of 5 mg of the FcεR1- specific MAR1 antibody 3 consecutive days prior to 6A6-IgG2a injection. In Situ Analysis of Autoantibody Binding For depletion of phagocytic cell populations, mice were injected with the indi- To analyze localization of anti-platelet antibodies in liver or spleen in situ, cated amounts of liposome-encapsulated dichloromethylene-bisphospho- purified 6A6-IgG2a was labeled with QDot705 nanochrystals with QDot705 nate (clodronate liposomes) 2 hr or 20 hr before ITP induction. Depletion of antibody conjugation kit (Invitrogen). To detect binding of 6A6-IgG subclasses CD11c-positive cells in Cd11c-DTR mice was achieved by a single injection to cell populations by flow cytometry, the antibody was conjugated with of 4 ng/g diphtheria toxin as described (Jung et al., 2002). The efficiency of Alexa Fluor 647 with the Alexa Fluor 647 monoclonal antibody labeling kit cell depletion was verified by flow cytometry and immunofluorescence (Invitrogen). analysis.

Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc. 941 Immunity Ly6Clo Monocytes and IgG Effector Functions

Fluorescence Labeling of Phagocytic Cell Populations by Liposome ACKNOWLEDGMENTS Uptake PBS liposomes were labeled with the PKH26 cell linker mini kit (Sigma-Aldrich This study was supported by the German Research Foundation (SFB 643 to Chemie GmbH, Munich, Germany). Per staining reaction, 1 ml of PBS lipo- F.N. and D.D.), the Emmy-Noether programme (DU548 2-1; D.D.), and the somes were sedimented by centrifugation (200 3 g, 3 min) and supernatant Bavarian Genome Research Network (BayGene to F.N. and D.D.). We are was removed. Liposomes were resuspended in 1 ml of diluent C. This suspen- grateful to N. van Roijen for clodronate liposomes, M. Mack (University Hospital sion was mixed with an equal volume of diluent C containing PKH26 and the Regensburg, Germany) for the CCR2 antibody, T. Winkler (University of reaction was incubated for 5 min at room temperature prior to addition of Erlangen) for the CD20-IgG2a antibody), T. Malek (University of Miami) for 1 ml of fetal calf serum for termination of the reaction. Liposomes were washed the 1A8 hybridoma, and J. Ravetch (Rockefeller University, New York), once in PBS with 1% BSA and several times with PBS. Finally, PKH26-labeled U. Schleicher (Department of Microbiology at the University Hospital of liposomes were resuspended in 600 ml PBS and 150 ml of this suspension was Erlangen), and D. Littman (Skirball Institute, New York) for mice. used for injection. Received: August 16, 2010 Bone Marrow Chimeras Revised: September 27, 2011 For the generation of bone marrow chimeric mice, ﬁve million bone marrow Accepted: November 7, 2011 cells from C57BL/6 or Fcer1g / mice were injected into lethally irradiated Published online: December 8, 2011 (9 Gy) C57BL/6 or Fcer1g/ recipient mice. Two months later, mice were eval- uated for the presence of donor cells by ﬂow cytometric analysis. For subse- REFERENCES quent experiments, only mice with a chimerism greater than 95% were used.

Flow Cytometry Auffray, C., Fogg, D., Garfa, M., Elain, G., Join-Lambert, O., Kayal, S., Sarnacki, S., Cumano, A., Lauvau, G., and Geissmann, F. (2007). Monitoring Flow cytometric analysis was done on a FACS Canto II (BD Biosciences, of blood vessels and tissues by a population of monocytes with patrolling Heidelberg, Germany) with single-cell suspensions of peripheral blood, bone 5 6 behavior. Science 317, 666–670. marrow, and spleen. Unless otherwise indicated, 1 3 10 to 1 3 10 cells were incubated on ice for at least 10 min with Fc block (clone 2.4G2, 0.5 mg Baudino, L., Nimmerjahn, F., Azeredo da Silveira, S., Martinez-Soria, E., Saito, T., per sample) to reduce unspecific binding to Fc receptors. When staining for Carroll, M., Ravetch, J.V., Verbeek, J.S., and Izui, S. (2008). Differential contribu- the individual FcgRs, Fc block was omitted. Analysis was restricted to viable tion of three activating IgG Fc receptors (FcgammaRI, FcgammaRIII, and cells, which were identified by exclusion of nucleic acid binding dye 40,6- FcgammaRIV) to IgG2a- and IgG2b-induced autoimmune hemolytic anemia in Diamidino-2-phenylindol (DAPI). To exclude cell aggregates from analysis, mice. J. Immunol. 180, 1948–1953. we used a gating strategy excluding events with an unproportionally large Bevaart, L., Jansen, M.J., van Vugt, M.J., Verbeek, J.S., van de Winkel, J.G., area of the forward scatter signal relating to its height. Cell numbers per blood and Leusen, J.H. (2006). The high-affinity IgG receptor, FcgammaRI, plays volume were calculated based on the respective numbers counted in a defined a central role in antibody therapy of experimental melanoma. Cancer Res. period at defined flow rates. Data acquisition and analysis was performed with 66, 1261–1264. the FACS Diva software (BD Biosciences). A complete list of antibodies and Biburger, M., Aschermann, S., Lux, A., and Nimmerjahn, F. (2010). The role of a precise description of the gating strategy used to identify monocyte subpop- Fcgamma receptors in murine autoimmune thrombocytopenia. Annu. ulations can be found in the Supplemental Information. Hematol. 89, S25–S30. Immunohistochemistry Bussel, J. (2006). Treatment of immune thrombocytopenic purpura in adults. 6 mm sections of frozen tissue were air-dried overnight followed by fixation in Semin. Hematol. 43, 3–10. acetone. When biotinylated antibodies were used on liver sections, blocking Cartron, G., Dacheux, L., Salles, G., Solal-Celigny, P., Bardos, P., Colombat, of endogenous biotin was performed with an Avidin-Biotin blocking kit P., and Watier, H. (2002). Therapeutic activity of humanized anti-CD20 mono- (Invitrogen) according to manufacturer’s instructions. The sections were clonal antibody and polymorphism in IgG Fc receptor FcgammaRIIIa gene. analyzed with an Axiovert 200M microscope (Carl Zeiss AG, Oberkochen, Blood 99, 754–758. Germany) supplied with an AxioCam MRM camera. A complete list of anti- Cines, D.B., and McMillan, R. (2005). Management of adult idiopathic throm- bodies used for immunohistochemistry can be found in the Supplemental bocytopenic purpura. Annu. Rev. Med. 56, 425–442. Information. Clarkson, S.B., Kimberly, R.P., Valinsky, J.E., Witmer, M.D., Bussel, J.B., In Vitro ADCC Assays Nachman, R.L., and Unkeless, J.C. (1986). Blockade of clearance of immune Naive B cells, neutrophils, and Ly6Clo and Ly6Chi monocytes were isolated complexes by an anti-Fc gamma receptor monoclonal antibody. J. Exp. from C57BL/6 mice by flow cytometric sorting on a FACS Aria III (BD). Cells Med. 164, 474–489. were combined at the indicated effector to target cell ratios in 96-well plates Coetzee, L.M., Pieters, H., van Wyk, V., Cooper, S., Roodt, J., De Reys, S., and with the CD20-specific IgG2a antibody or an isotype-matched control anti- Badenhorst, P.N. (2000). The effect of monoclonal anti-human-platelet antibody and incubated for 16–20 hr at 37 C before determination of total cell bodies on platelet kinetics in a baboon model: IgG subclass dependency. numbers and flow cytometric analysis to quantify the cell types present in Thromb. Haemost. 83, 148–156. the culture. Cros, J., Cagnard, N., Woollard, K., Patey, N., Zhang, S.Y., Senechal, B., Puel, A., Biswas, S.K., Moshous, D., Picard, C., et al. (2010). Human CD14dim Statistical Analysis monocytes patrol and sense nucleic acids and viruses via TLR7 and TLR8 All data are expressed as mean ± standard deviation. Data were analyzed receptors. Immunity 33, 375–386. and plotted with GraphPad Prism software (GraphPad Software Inc., San Diego, CA). For calculation of statistical significance, data were analyzed Daley, J.M., Thomay, A.A., Connolly, M.D., Reichner, J.S., and Albina, J.E. with Student’s t test or, in the case of unequal variances, Student’s t test (2008). Use of Ly6G-specific monoclonal antibody to deplete neutrophils in with Welch modification. p % 0.05 was considered significant. mice. J. Leukoc. Biol. 83, 64–70. Fossati-Jimack, L., Ioan-Facsinay, A., Reininger, L., Chicheportiche, Y., SUPPLEMENTAL INFORMATION Watanabe, N., Saito, T., Hofhuis, F.M., Gessner, J.E., Schiller, C., Schmidt, R.E., et al. (2000). Markedly different pathogenicity of four immunoglobulin G Supplemental Information includes Supplemental Experimental Procedures isotype-switch variants of an antierythrocyte autoantibody is based on their and five figures and can be found with this article online at doi:10.1016/ capacity to interact in vivo with the low-affinity Fcgamma receptor III. J. Exp. j.immuni.2011.11.009. Med. 191, 1293–1302.

942 Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc. Immunity Ly6Clo Monocytes and IgG Effector Functions

Geissmann, F., Jung, S., and Littman, D.R. (2003). Blood monocytes consist of myocardium sequentially mobilizes two monocyte subsets with divergent and two principal subsets with distinct migratory properties. Immunity 19, 71–82. complementary functions. J. Exp. Med. 204, 3037–3047. Geissmann, F., Manz, M.G., Jung, S., Sieweke, M.H., Merad, M., and Ley, K. Nigrovic, P.A., Binstadt, B.A., Monach, P.A., Johnsen, A., Gurish, M., Iwakura, (2010). Development of monocytes, macrophages, and dendritic cells. Y., Benoist, C., Mathis, D., and Lee, D.M. (2007). Mast cells contribute to initi- Science 327, 656–661. ation of autoantibody-mediated arthritis via IL-1. Proc. Natl. Acad. Sci. USA Giorgini, A., Brown, H.J., Lock, H.R., Nimmerjahn, F., Ravetch, J.V., Verbeek, 104, 2325–2330. J.S., Sacks, S.H., and Robson, M.G. (2008). Fc gamma RIII and Fc gamma RIV Nimmerjahn, F., and Ravetch, J.V. (2005). Divergent immunoglobulin g are indispensable for acute glomerular inflammation induced by switch variant subclass activity through selective Fc receptor binding. Science 310, 1510– monoclonal antibodies. J. Immunol. 181, 8745–8752. 1512. Gordon, S., and Taylor, P.R. (2005). Monocyte and macrophage heteroge- Nimmerjahn, F., and Ravetch, J.V. (2006). Fcgamma receptors: old friends and neity. Nat. Rev. Immunol. 5, 953–964. new family members. Immunity 24, 19–28. Hamaguchi, Y., Xiu, Y., Komura, K., Nimmerjahn, F., and Tedder, T.F. (2006). Nimmerjahn, F., and Ravetch, J.V. (2008). Fcgamma receptors as regulators of Antibody isotype-specific engagement of Fcgamma receptors regulates immune responses. Nat. Rev. Immunol. 8, 34–47. B lymphocyte depletion during CD20 immunotherapy. J. Exp. Med. 203, Nimmerjahn, F., Bruhns, P., Horiuchi, K., and Ravetch, J.V. (2005). 743–753. FcgammaRIV: a novel FcR with distinct IgG subclass specificity. Immunity Hashimoto, D., Miller, J., and Merad, M. (2011). Dendritic cell and macrophage 23, 41–51. heterogeneity in vivo. Immunity 35, 323–335. Nimmerjahn, F., Anthony, R.M., and Ravetch, J.V. (2007). Agalactosylated IgG antibodies depend on cellular Fc receptors for in vivo activity. Proc. Natl. Acad. Hazenbos, W.L., Gessner, J.E., Hofhuis, F.M., Kuipers, H., Meyer, D., Heijnen, Sci. USA 104, 8433–8437. I.A., Schmidt, R.E., Sandor, M., Capel, P.J., Dae¨ ron, M., et al. (1996). Impaired IgG-dependent anaphylaxis and Arthus reaction in Fc gamma RIII (CD16) defi- Nimmerjahn, F., Lux, A., Albert, H., Woigk, M., Lehmann, C., Dudziak, D., cient mice. Immunity 5, 181–188. Smith, P., and Ravetch, J.V. (2010). FcgRIV deletion reveals its central role for IgG2a and IgG2b activity in vivo. Proc. Natl. Acad. Sci. USA 107, 19396– Heyns, A.D., Lo¨ tter, M.G., Badenhorst, P.N., van Reenen, O.R., Pieters, H., 19401. Minnaar, P.C., and Retief, F.P. (1980). Kinetics, distribution and sites of destruction of 111indium-labelled human platelets. Br. J. Haematol. 44, Otten, M.A., van der Bij, G.J., Verbeek, S.J., Nimmerjahn, F., Ravetch, J.V., 269–280. Beelen, R.H., van de Winkel, J.G., and van Egmond, M. (2008). Experimental antibody therapy of liver metastases reveals functional redundancy between Hirano, M., Davis, R.S., Fine, W.D., Nakamura, S., Shimizu, K., Yagi, H., Kato, Fc gammaRI and Fc gammaRIV. J. Immunol. 181, 6829–6836. K., Stephan, R.P., and Cooper, M.D. (2007). IgEb immune complexes activate macrophages through FcgammaRIV binding. Nat. Immunol. 8, 762–771. Panizzi, P., Swirski, F.K., Figueiredo, J.L., Waterman, P., Sosnovik, D.E., Aikawa, E., Libby, P., Pittet, M., Weissleder, R., and Nahrendorf, M. (2010). Hogarth, P.M. (2002). Fc receptors are major mediators of antibody based Impaired infarct healing in atherosclerotic mice with Ly-6C(hi) monocytosis. inflammation in autoimmunity. Curr. Opin. Immunol. 14, 798–802. J. Am. Coll. Cardiol. 55, 1629–1638. Ioan-Facsinay, A., de Kimpe, S.J., Hellwig, S.M., van Lent, P.L., Hofhuis, F.M., Probst, H.C., Tschannen, K., Odermatt, B., Schwendener, R., Zinkernagel, van Ojik, H.H., Sedlik, C., da Silveira, S.A., Gerber, J., de Jong, Y.F., et al. R.M., and Van Den Broek, M. (2005). Histological analysis of CD11c-DTR/ (2002). FcgammaRI (CD64) contributes substantially to severity of arthritis, GFP mice after in vivo depletion of dendritic cells. Clin. Exp. Immunol. 141, hypersensitivity responses, and protection from bacterial infection. Immunity 398–404. 16, 391–402. Ravetch, J.V., and Clynes, R.A. (1998). Divergent roles for Fc receptors and Jung, S., Unutmaz, D., Wong, P., Sano, G., De los Santos, K., Sparwasser, T., complement in vivo. Annu. Rev. Immunol. 16, 421–432. Wu, S., Vuthoori, S., Ko, K., Zavala, F., et al. (2002). In vivo depletion of CD11c+ Samuelsson, A., Towers, T.L., and Ravetch, J.V. (2001). Anti-inflammatory dendritic cells abrogates priming of CD8+ T cells by exogenous cell-associ- activity of IVIG mediated through the inhibitory Fc receptor. Science 291, ated antigens. Immunity 17, 211–220. 484–486. Kaneko, Y., Nimmerjahn, F., Madaio, M.P., and Ravetch, J.V. (2006). Santiago-Raber, M.L., Amano, H., Amano, E., Baudino, L., Otani, M., Lin, Q., Pathology and protection in nephrotoxic nephritis is determined by selective Nimmerjahn, F., Verbeek, J.S., Ravetch, J.V., Takasaki, Y., et al. (2009). engagement of specific Fc receptors. J. Exp. Med. 203, 789–797. Fcgamma receptor-dependent expansion of a hyperactive monocyte subset Kumar, V., Ali, S.R., Konrad, S., Zwirner, J., Verbeek, J.S., Schmidt, R.E., and in lupus-prone mice. Arthritis Rheum. 60, 2408–2417. Gessner, J.E. (2006). Cell-derived anaphylatoxins as key mediators of anti- Santiago-Raber, M.L., Baudino, L., Alvarez, M., van Rooijen, N., Nimmerjahn, body-dependent type II autoimmunity in mice. J. Clin. Invest. 116, 512–520. F., and Izui, S. (2011). TLR7/9-mediated monocytosis and maturation of Lee, D.M., Friend, D.S., Gurish, M.F., Benoist, C., Mathis, D., and Brenner, Gr-1(hi) inflammatory monocytes towards Gr-1(lo) resting monocytes impli- M.B. (2002). Mast cells: a cellular link between autoantibodies and inflamma- cated in murine lupus. J. Autoimmun. 37, 171–179. tory arthritis. Science 297, 1689–1692. Schmidt, K.G., and Rasmussen, J.W. (1985). Kinetics and distribution in vivo of Mancardi, D.A., Iannascoli, B., Hoos, S., England, P., Dae¨ ron, M., and Bruhns, 111In-labelled autologous platelets in idiopathic thrombocytopenic purpura. P. (2008). FcgammaRIV is a mouse IgE receptor that resembles macrophage Scand. J. Haematol. 34, 47–56. FcepsilonRI in humans and promotes IgE-induced lung inflammation. J. Clin. Serbina, N.V., and Pamer, E.G. (2006). Monocyte emigration from bone Invest. 118, 3738–3750. marrow during bacterial infection requires signals mediated by chemokine Meyer, D., Schiller, C., Westermann, J., Izui, S., Hazenbos, W.L., Verbeek, receptor CCR2. Nat. Immunol. 7, 311–317. J.S., Schmidt, R.E., and Gessner, J.E. (1998). FcgammaRIII (CD16)-deficient Sprague, C.C., Harrington, W.J., Lange, R.D., and Shapleigh, J.B. (1952). mice show IgG isotype-dependent protection to experimental autoimmune Platelet transfusions and the pathogenesis of idiopathic thrombocytopenic hemolytic anemia. Blood 92, 3997–4002. purpura. J. Am. Med. Assoc. 150, 1193–1198. Mizutani, H., Furubayashi, T., Kashiwagi, H., Honda, S., Take, H., Kurata, Y., Stratton, J.R., Ballem, P.J., Gernsheimer, T., Cerqueira, M., and Slichter, S.J. Yonezawa, T., Tarui, S., and Ikehara, S. (1992). Effects of splenectomy on (1989). Platelet destruction in autoimmune thrombocytopenic purpura: immune thrombocytopenic purpura in (NZW x BXSB) F1 mice: analyses of kinetics and clearance of indium-111-labeled autologous platelets. J. Nucl. platelet kinetics and anti-platelet antibody production. Thromb. Haemost. Med. 30, 629–637. 67, 563–566. Syed, S.N., Konrad, S., Wiege, K., Nieswandt, B., Nimmerjahn, F., Schmidt, Nahrendorf, M., Swirski, F.K., Aikawa, E., Stangenberg, L., Wurdinger, T., R.E., and Gessner, J.E. (2009). Both FcgammaRIV and FcgammaRIII are Figueiredo, J.L., Libby, P., Weissleder, R., and Pittet, M.J. (2007). The healing essential receptors mediating type II and type III autoimmune responses via

Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc. 943 Immunity Ly6Clo Monocytes and IgG Effector Functions

FcRgamma-LAT-dependent generation of C5a. Eur. J. Immunol. 39, 3343– B lymphocytes through Fc receptor-dependent mechanisms during anti-CD20 3356. antibody immunotherapy. J. Exp. Med. 199, 1659–1669. Tacke, F., Ginhoux, F., Jakubzick, C., van Rooijen, N., Merad, M., and Weng, W.K., and Levy, R. (2003). Two immunoglobulin G fragment C receptor Randolph, G.J. (2006). Immature monocytes acquire antigens from other cells polymorphisms independently predict response to rituximab in patients with in the bone marrow and present them to T cells after maturing in the periphery. follicular lymphoma. J. Clin. Oncol. 21, 3940–3947. J. Exp. Med. 203, 583–597. Weng, W.K., Czerwinski, D., Timmerman, J., Hsu, F.J., and Levy, R. (2004). Takai, T. (2002). Roles of Fc receptors in autoimmunity. Nat. Rev. Immunol. 2, Clinical outcome of lymphoma patients after idiotype vaccination is correlated 580–592. with humoral immune response and immunoglobulin G Fc receptor genotype. Takai, T., Li, M., Sylvestre, D., Clynes, R., and Ravetch, J.V. (1994). FcR J. Clin. Oncol. 22, 4717–4724. gamma chain deletion results in pleiotrophic effector cell defects. Cell 76, Ziegler-Heitbrock, L., Ancuta, P., Crowe, S., Dalod, M., Grau, V., Hart, D.N., 519–529. Leenen, P.J., Liu, Y.J., MacPherson, G., Randolph, G.J., et al. (2010). Uchida, J., Hamaguchi, Y., Oliver, J.A., Ravetch, J.V., Poe, J.C., Haas, K.M., Nomenclature of monocytes and dendritic cells in blood. Blood 116, and Tedder, T.F. (2004). The innate mononuclear phagocyte network depletes e74–e80.

944 Immunity 35, 932–944, December 23, 2011 ª2011 Elsevier Inc.