Anti‐ BDCA 2 monoclonal antibody inhibits plasmacytoid dendritic cell activation through Fc‐dependent and Fc‐independent mechanisms

Research Article11 March 2015Open Access Source Data Anti-BDCA2 monoclonal antibody inhibits plasmacytoid dendritic cell activation through Fc-dependent and Fc-independent mechanisms Alex Pellerin Alex Pellerin Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Karel Otero Karel Otero Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Julie M Czerkowicz Julie M Czerkowicz Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Hannah M Kerns Hannah M Kerns Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Renée I Shapiro Renée I Shapiro Biologics Drug Discovery, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Ann M Ranger Ann M Ranger Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Kevin L Otipoby Kevin L Otipoby Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Frederick R Taylor Frederick R Taylor Biologics Drug Discovery, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Thomas O Cameron Thomas O Cameron Biologics Drug Discovery, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Joanne L Viney Joanne L Viney Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Dania Rabah Corresponding Author Dania Rabah Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Alex Pellerin Alex Pellerin Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Karel Otero Karel Otero Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Julie M Czerkowicz Julie M Czerkowicz Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Hannah M Kerns Hannah M Kerns Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Renée I Shapiro Renée I Shapiro Biologics Drug Discovery, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Ann M Ranger Ann M Ranger Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Kevin L Otipoby Kevin L Otipoby Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Frederick R Taylor Frederick R Taylor Biologics Drug Discovery, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Thomas O Cameron Thomas O Cameron Biologics Drug Discovery, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Joanne L Viney Joanne L Viney Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Dania Rabah Corresponding Author Dania Rabah Immunology Research, Biogen Idec, Cambridge, MA, USA Search for more papers by this author Author Information Alex Pellerin1, Karel Otero1, Julie M Czerkowicz1, Hannah M Kerns1, Renée I Shapiro2, Ann M Ranger1, Kevin L Otipoby1, Frederick R Taylor2, Thomas O Cameron2, Joanne L Viney1 and Dania Rabah 1 1Immunology Research, Biogen Idec, Cambridge, MA, USA 2Biologics Drug Discovery, Biogen Idec, Cambridge, MA, USA *Corresponding author. Tel: +617 679 6255; Fax: +617 679 3208; E-mail: [email protected] EMBO Mol Med (2015)7:464-476https://doi.org/10.15252/emmm.201404719 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions Figures & Info Abstract Type I interferons (IFN-I) are implicated in the pathogenesis of systemic lupus erythematosus (SLE). In SLE, immune complexes bind to the CD32a (FcγRIIa) receptor on the surface of plasmacytoid dendritic cells (pDCs) and stimulate the secretion of IFN-I from pDCs. BDCA2 is a pDC-specific receptor that, when engaged, inhibits the production of IFN-I in human pDCs. BDCA2 engagement, therefore, represents an attractive therapeutic target for inhibiting pDC-derived IFN-I and may be an effective therapy for the treatment of SLE. In this study, we show that 24F4A, a humanized monoclonal antibody (mAb) against BDCA2, engages BDCA2 and leads to its internalization and the consequent inhibition of TLR-induced IFN-I by pDCs in vitro using blood from both healthy and SLE donors. These effects were confirmed in vivo using a single injection of 24F4A in cynomolgus monkeys. 24F4A also inhibited pDC activation by SLE-associated immune complexes (IC). In addition to the inhibitory effect of 24F4A through engagement of BDCA2, the Fc region of 24F4A was critical for potent inhibition of IC-induced IFN-I production through internalization of CD32a. This study highlights the novel therapeutic potential of an effector-competent anti-BDCA2 mAb that demonstrates a dual mechanism to dampen pDC responses for enhanced clinical efficacy in SLE. Synopsis This study describes a novel therapeutic approach to dampen pDC responses whereby an anti-BDCA2 mAb, 24F4A, internalizes BDCA2 and inhibits TLR7 and TLR9-induced type I IFN while simultaneously down-modulating CD32a through the Fc region, from the surface of pDCs. pDC derived type I IFN is thought to play an important role in the pathogenicity of systemic lupus erythematous (SLE). Blood Dendritic Cell Antigen (BDCA2) is a receptor that is specifically expressed on human and non-human primates pDCs that inhibits TLR7 and TLR9-mediated type I IFN. In vitro experiments indicate that BDCA2 is rapidly internalized and internalization correlates with type I IFN inhibition. 24F4A administration leads to BDCA2 internalization in vivo and dampens TLR9-mediated type I IFN without depleting pDCs in cynomolgus monkeys. 24F4A inhibits type I IFN from immune complex-stimulated pDCs through BDCA2 ligation as well as Fc-dependent down-modulation of CD32a. Introduction Systemic lupus erythematosus (SLE) is a chronic autoimmune inflammatory disease that can affect multiple organs. SLE is characterized by the presence of pathogenic autoantibodies against self-nucleoproteins and DNA (Tan, 1989). Although SLE is a multifactorial disease, increasing evidence indicates that type I interferon (IFN-I) plays a pivotal role in the disease pathogenesis (Crow, 2010); (Elkon & Wiedeman, 2012). IFN-I constitutes a family of cytokines (IFN-α, IFN-β IFN-ε, IFN-κ, IFN-ω, IFN-δ and IFN-τ) that are important for clearance of viral infections but whose uncontrolled production contributes to autoimmune and inflammatory conditions (Theofilopoulos et al, 2005). While most cells can produce IFN-I in response to nucleic acids, plasmacytoid dendritic cells (pDCs) are considered the major producers of IFN-I (Liu, 2005). PDCs are a specialized population of bone marrow-derived cells that can produce as much as a 1,000-fold more IFN-I than other cells in response to ligands that engage the endosomal Toll-like receptor (TLR)7 and TLR9 (Siegal et al, 1999). PDCs accumulate in skin lesions of SLE patients as well as in other target organs and have been suggested to be a major source of IFN-I in SLE (Blomberg et al, 2001; Farkas et al, 2001; Tucci et al, 2008; Tomasini et al, 2010; Ghoreishi et al, 2012). Direct evidence of the critical role of pDCs in SLE pathogenesis has been recently obtained from animal models of SLE in which pDCs were depleted or inactivated (Rowland et al, 2014). Autoreactive immune complexes (IC) are potent inducers of IFN-I in SLE (Lovgren et al, 2006). IC bind to low-affinity FcγRIIa (CD32a) Fc receptors on the surface of pDCs and are internalized. In the endosome, single-stranded RNA or DNA containing unmethylated CpG sequences present in these complexes stimulates the nucleic acid sensors TLR7 and TLR9, respectively, leading to the production of IFN-I (Means et al, 2005; Lovgren et al, 2006). Persistent IC-mediated stimulation of TLR7 and TLR9 in pDCs is suspected to be one of the primary mechanisms whereby pDCs release IFN-I and contribute to SLE disease progression (Ronnblom & Alm, 2001; Swiecki & Colonna, 2010; Ganguly et al, 2013). Blood dendritic cell antigen 2 (BDCA2) is a C-type lectin exclusively expressed on the surface of human pDCs (Dzionek et al, 2000). BDCA2 consists of a single extracellular carbohydrate recognition domain, a transmembrane region, and a short cytoplasmic tail that does not harbor any signaling motifs. BDCA2 transmits intracellular signals through an associated transmembrane adaptor, the FcεRIγ, and induces a B-cell receptor (BCR)-like signaling cascade. Antibody-mediated ligation of BDCA2 leads to recruitment of SYK to the phosphorylated immunoreceptor tyrosine-based activation motif (ITAM) of FcεRIγ. SYK activation leads to the activation of BTK and PLCγ2 leading to calcium mobilization. BDCA2 receptor engagement has been shown to inhibit TLR7- or TLR9-induced production of IFN-I and other pDC-derived pro-inflammatory mediators (Dzionek et al, 2001; Fanning et al, 2006; Cao et al, 2007; Rock et al, 2007). In addition to inhibiting IFN-I production by pDCs, ligation of BDCA2 with an antibody leads to rapid internalization of BDCA2 by clathrin-mediated endocytosis (Dzionek et al, 2000, 2001; Jaehn et al, 2008). In this study, we describe a novel antibody, 24F4A, which binds BDCA2 and leads to its internalization and the consequent inhibition of TLR9-induced IFN-I by pDCs both in vitro and in vivo. In addition, we describe two distinct mechanisms by which 24F4A can inhibit SLE-IC-mediated pDC activation. The first mechanism is mediated by the induction of signaling from BDCA2, while the second mechanism is mediated by the depletion of Fc receptor (FcR) from the surface of pDCs. Results Novel anti-BDCA2 monoclonal antibody (mAb) inhibits IFN-I production by pDCs in healthy human donors and SLE patients BDCA2 ligation with a monoclonal antibody against BDCA2 (clone AC144) has been shown to suppress the ability of human pDCs to produce IFN-I in response to TLR7 and TLR9 ligands (Dzionek et al, 2001; Blomberg et al, 2003; Cao et al, 2007). We generated mouse monoclonal antibodies (mAbs) against human BDCA2 and investigated their ability to inhibit TLR9-induced IFNα by pDCs. Of the 92 hybridomas isolated, only 10 anti-BDCA2 mAbs, belonging to 4 primary-sequence defined families, demonstrated binding to both human and cynomolgus BDCA2 and inhibited TLR9-induced IFNα from human PBMC (Supplementary Table S1 and Supplementary Fig S1). 24F4A demonstrated high potency, comparable to AC144, and was therefore chosen for humanization (Supplementary Table S1 and Supplementary Fig S1). Next, we investigated the potency of 24F4A in whole-blood assays. To this end, whole blood from healthy human donors (n = 12) was stimulated with the TLR9 ligand CpG-A in the presence of increasing concentrations of 24F4A. BDCA2 ligation by 24F4A led to a dose-dependent inhibition of TLR9-induced IFNα production (Fig 1A) with an average IC50 of 0.06 μg/ml (Fig 1B). 24F4A displayed similar potency in purified peripheral blood mononuclear cells (PBMC) isolated from healthy human donors and SLE patients (Fig 1B). To ensure that 24F4A specifically inhibited pDC-derived IFN-I, PBMC from healthy donors were stimulated with the TLR3 ligand poly(I:C). PDCs do not express TLR3 and therefore do not contribute to IFN-I production induced by poly(I:C) in PBMC culture (Kadowaki et al, 2001; Matsumoto et al, 2003). As expected, 24F4A inhibited TLR9-induced IFNα but did not impact TLR3-induced IFNα production by PBMC (Supplementary Fig S2). Figure 1. Anti-BDCA2 mAb inhibits IFNα production by pDCs and induces internalization of BDCA2 on the surface of pDCs A, B. Whole blood or PBMC were treated with increasing concentrations of 24F4A and stimulated with CpG-A for 16 h at 37°C. Anti-BDCA2-mediated IFNα inhibition in whole-blood assays (A). IFNα levels were detected using human IFNα ELISA. Graph depicts average of duplicate wells of one representative donor (n = 10). IC50 of 24F4A-mediated IFNα inhibition (B) in whole blood (circles) (n = 10) and PBMC (n = 18) from healthy human donors (squares) or SLE patients (triangles) (n = 11). Horizontal bar represents the mean IC50 for each sample type. Error bars represent SD of IC50 between donors. C. Anti-BDCA2-mediated internalization in whole-blood assays. Whole blood was treated with increasing concentrations of 24F4A for 16 h. Mean fluorescence intensity (MFI) of BDCA2 was determined with a non-cross-blocking anti-BDCA2 mAb (2D6). Shown is a representative plot of 10 experiments conducted. D. The IC50 of 24F4A-mediated IFNα inhibition was compared to the EC50 of 24F4A-induced BDCA2 internalization. Download figure Download PowerPoint Anti-BDCA2-induced BDCA2 internalization is correlated with the inhibition of IFNα production in human pDCs Antibody-mediated BDCA2 ligation was previously reported to induce rapid receptor internalization (Dzionek et al, 2000; Jahn et al, 2010). To determine whether 24F4A could induce BDCA2 internalization, whole blood from healthy human donors was incubated overnight with increasing concentrations of 24F4A. Surface expression of BDCA2 was detected on pDCs by flow cytometry using an antibody directed to a non-overlapping epitope of BDCA2 (antibody clone 2D6; Supplementary Fig S3). Treatment with 24F4A led to a dose-dependent decrease in BDCA2 surface expression on pDCs with an average EC50 of 0.017 μg/ml (Fig 1C). The EC50 of 24F4A-mediated BDCA2 internalization correlated with the IC50 of IFNα inhibition (n = 10 healthy donors) with an R2 value of 0.68 (Fig 1D). The correlation between BDCA2 internalization and IFN-I inhibition was further supported by experiments using another anti-BDCA2 mAb from our panel, murine 6G6. While 6G6 bound BDCA2 with high affinity (Supplementary Table S1) and achieved full receptor occupancy (Supplementary Fig S4C) it only led to modest BDCA2 internalization and similarly, modest inhibition of TLR9-induced IFNα in PBMC culture (Supplementary Fig S4A and B). Together, these data suggest that BDCA2 internalization and inhibition of TLR9-induced IFN-I production may be mechanistically linked. Anti-BDCA2 mAb induces rapid and persistent intracellular localization of BDCA2 and sustained inhibition of IFN-I Studies of 24F4A-mediated BDCA2 internalization kinetics showed that BDCA2 surface expression rapidly decreased following treatment with 1 μg/ml 24F4A, reaching background levels within 1 h of treatment. Treatment with tenfold lower concentration of 24F4A delayed surface BDCA2 downmodulation by 2 h, while a 100-fold lower concentration of 24F4A was able to cause significant receptor internalization only after overnight culture (Fig 2A). These data indicate that BDCA2 is rapidly internalized upon ligation with 24F4A with dose-dependent kinetics. Figure 2. 24F4A induces rapid internalization of BDCA2 resulting in trafficking to LAMP1+ compartments and sustained IFNα inhibition A. Whole blood was treated with the indicated concentrations of 24F4A or the IgG1 isotype control antibody for 0, 1, 3, 6, 9, and 24 h. FACS analysis was performed to determine BDCA2 levels on pDCs. Fluorescence minus one (FMO) represents background staining of BDCA2. Shown is a representative plot of three independent experiments. B. Isolated pDCs were treated with 10 μg/ml 24F4A-AF647 at 4°C or at 37°C and analyzed at the indicated time points. Subcellular localization of BDCA2 (red) compared to LAMP1 (green) was assessed using confocal microscopy. Phalloidin was used to delineate the cell membrane. Yellow represents co-localization of BDCA2 and LAMP1. Shown is a representative image of four experiments conducted. C, D. Whole blood was either treated with 24F4A for 1 h at 37°C (pre-incubation) or left untreated. PBMC were isolated from each condition. PBMC from the untreated whole blood were subsequently treated with 10 μg/ml 24F4A, and cells from all conditions were stimulated with CpG-A for 16 h. Plot represents mean of duplicate wells. Flow cytometry was used to measure BDCA2 levels (MFI). Shown is a representative donor of eight donors tested. E. Whole blood was pre-treated for 0, 1, 3, 6, and 9 h with 10 μg/ml of 24F4 or the isotype control and then stimulated with CpG-A for additional 16 h. IFNα levels were measured by ELISA. Shown is a representative plot of two independent experiments. Source data are available online for this figure. Source Data for Figure 2 [emmm201404719-sup-0014-Sourcedata-Fig2.pdf] Download figure Download PowerPoint Next, to visualize the internalization of BDCA2 and subcellular localization after ligation with 24F4A, purified pDCs were incubated with AlexaFluor647-labeled 24F4A and analyzed by confocal microscopy. As shown in Fig 2B, labeled 24F4A bound to BDCA2 was localized on the cell surface of pDCs at 4°C. After a 10 min incubation at 37°C, labeled 24F4A was detected inside pDCs presumably complexed with BDCA2. After a 1 h incubation at 37°C the 24F4A/BDCA2 complex was localized in the LAMP1+ endolysosomal compartments where it persisted for up to 14 h post-treatment (Fig 2B). While it has been shown that anti-BDCA2 mAb can lead to BDCA2 internalization and accumulation in intracellular compartments (Jaehn et al, 2008), it is unknown whether BDCA2 is recycled to the cell surface after 24F4A-mediated internalization. To address this question, whole blood was pre-incubated in the absence or presence of 10 μg/ml of 24F4A for 1 h at 37°C to obtain maximal internalization. PBMC were subsequently isolated, thereby removing unbound 24F4A that could engage re-expressed BDCA2. After 16 h at 37°C, BDCA2 levels remained low on pDCs that were pre-incubated with 24F4A but cultured in the absence of 24F4A and were comparable to the BDCA2 levels on pDCs in control cultures that were continuously exposed to 10 μg/ml of 24F4A (Fig 2C). Furthermore, when stimulated with CpG-A, IFNα production was inhibited in cells pre-incubated with 24F4A to levels indistinguishable from that seen in PBMC continuously exposed to 24F4A (Fig 2D). These data demonstrate that BDCA2 is not rapidly recycled to the cell surface of pDCs after 24F4A-mediated BDCA2 internalization. Furthermore, the data show that 1 h pre-incubation with 24F4A is sufficient to inhibit TLR9-induced IFNα production. We extended these studies to ascertain whether internalized BDCA2 was still capable of inhibiting IFNα over longer pre-incubation periods with 24F4A. To this end, whole blood from healthy human donors was pre-incubated with 24F4A at 37°C for various periods up to 9 h, followed by stimulation with CpG-A. As shown in Fig 2E, pre-incubation with 24F4A for up to 9 h led to maximal inhibition of TLR9-induced IFNα production. Our data suggest that the persistent localization of the 24F4A/BDCA2 complex in LAMP1+ endolysosomal compartment in the absence of recycling could be important for its ability to mediate IFN-I inhibition. Anti-BDCA2 mAb leads to BDCA2 internalization and IFN-I inhibition in vivo in cynomolgus monkeys We next tested the pharmacokinetic properties and biological activity of 24F4A in vivo. Since rodents do not express BDCA2 (Dzionek et al, 2000), we performed these studies in the cynomolgus monkey. Cynomolgus monkey BDCA2 protein shares 90.6% homology with human BDCA2. 24F4A binds similarly to cynomolgus and human BDCA2 with an average EC50 of 0.63 μg/ml (n = 5) and 0.7 μg/ml (n = 8), respectively (Supplementary Fig S5). Nine cynomolgus monkeys were divided into three groups that received a single intravenous (IV) injection of vehicle (sodium citrate buffer), 1 mg/kg 24F4A, or 10 mg/kg 24F4A. Animals were bled at various time points before and after 24F4A administration. First, we addressed whether administration of 24F4A leads to BDCA2 internalization in vivo, using flow cytometry. Because the 2D6 anti-BDCA2 clone does not cross-react with cynomolgus BDCA2 (Supplementary Table S1), a two-step approach was used to detect internalization of BDCA2 on cynomolgus monkey pDCs. Unoccupied surface BDCA2 was detected on pDCs in whole blood using fluorescently labeled 24F4A (direct method), while surface BDCA2 bound to 24F4A was detected using a fluorescently labeled anti-human IgG1 (indirect method). The lack of unoccupied BDCA2 (direct method) coupled with loss of detectable 24F4A (indirect method) indicated BDCA2 internalization. Results from a representative animal from both the vehicle-treated group and the 1 mg/kg 24F4A-treated group are shown in Fig 3A and B. Prior to vehicle and 24F4A administration, the baseline surface expression of BDCA2 was assessed for each cynomolgus monkey using the direct method (Fig 3A-i and B-i, dotted red line). In addition, maximal binding of BDCA2 to 24F4A was established prior to 24F4A administration by “spiking” whole blood with saturating amounts of 24F4A in vitro and measuring bound 24F4A by the indirect method (Fig 3A-ii and B-ii, solid red line). Within 6 h of 24F4A administration at 1 mg/kg, BDCA2 expression on the surface of pDCs decreased to almost undetectable levels (Fig 3B-iii, dotted red line) but not in the vehicle-treated group (Fig 3A-iii, dotted red line). In addition, the levels of bound 24F4A (Fig 3B-iv, solid black line) were indistinguishable from the vehicle-treated group (Fig 3A-iv, solid black line). The lack of available BDCA2 receptor together with the lack of detectable 24F4A on the surface of pDCs indicated internalization of BDCA2. Over 95% of surface BDCA2 was internalized in all animals within 6 h of IV treatment (1 and 10 mg/kg) (Fig 3C). Internalization of BDCA2 correlated with circulating levels of 24F4A, establishing a pharmacokinetic/pharmacodynamic (PK/PD) relationship in vivo. When 24F4A serum concentrations decreased to a range of 0.1–0.03 μg/ml, the level of BDCA2 recovered to > 70% of the baseline level (Fig 3D-i–iii), establishing an EC50 of 0.133 μg/ml (Fig 3D-iv). Figure 3. 24F4A mediates BDCA2 internalization and type I IFN inhibition in vivoCynomolgus monkeys were administered 24F4A (10 or 1 mg/kg) or vehicle (n = 3 for each dose group) intravenously. Cynomolgus monkeys were bled at various time points, and flow cytometry was used to measure BDCA2 expression and receptor occupancy. PDCs were defined as CD20−, CD14−, CD123+, and HLADR+. A–C. Prior to in vivo dosing, baseline surface levels of BDCA2 for both the vehicle (Ai) and 1 mg/kg (Bi) animals (red, dotted line) were established by staining with fluorescently labeled 24F4A (direct method). Maximal binding of 24F4A to BDCA2 was also established pre-dose in the vehicle (Aii) and 1 mg/kg (Bii) animals (red, solid line) by treating whole blood with 10 μg/ml of 24F4A at 4°C and then detecting bound 24F4A with a fluorescently labeled anti-human IgG1 (indirect method). The direct method was used to stain whole blood from both the vehicle (Aiii) and 1 mg/kg 24F4A (Biii) animals 6 h post-dose (red, dotted line). In a separate stain, the indirect method was used to detect bound 24F4A in the vehicle (Aiv) and 1 mg/kg (Biv) treated animals (black, solid line). (C) Percent BDCA2 internalization relative to pre-dose BDCA2 levels 6 h post-dose with vehicle, 10 mg/kg, or 1 mg/kg 24F4A. Graph shows mean ± standard deviation for each group (n = 3). D. PK/PD relationship between 24F4A serum concentrations (red triangle, left axis) and BDCA2 expression on pDCs (black squares, right axis, normalized to pre-dose levels) from the 1 mg/kg group (i–iii). Serum 24F4A was measured by ELISA. (iv) Percent BDCA2 internalization versus serum concentration of 24F4A for all dosed cynomolgus monkeys at all time points tested. E. Whole blood from vehicle- or 1 mg/kg 24F4A-treated monkeys was stimulated with CpG-A, and induction of IFN-I was measured by MxA bioassay at various time points pre- and post-treatment. Horizontal black lines represent the model-based estimates of the geometric mean of IFN-I in pre- and post-dose samples. Duplicate symbols represent independent replicates of the MxA bioassay for that time point. Statistical analysis was performed using a two-way mixed-effects analysis of variance (ANOVA). Download figure Download PowerPoint Next, we determined whether 24F4A, when administered to cynomolgus monkeys in vivo, could inhibit the production of IFN-I after TLR9 stimulation ex vivo. As the reagents for stimulating and detecting cynomolgus IFN-I are not adequate, we utilized an established MxA bioassay (Wadhwa et al, 2013) to indirectly measure IFN-I production from CpG-A-stimulated whole blood. A two-way mixed-effects analysis of variance (ANOVA) was used to estimate and compare the mean log10 IFN-I concentrations for each treatment group before and after administration of 24F4A up to day 31 or prior to loss of the pharmacodynamics effect (BDCA2 internalization). Vehicle injection did not significantly affect the level of IFN-I, while 1 mg/kg 24F4A led to a 46% decrease in IFN-I production (95% CI: 18–65%; P = 0.004) (Fig 3E). IFN-I production was also significantly decreased following 10 mg/kg IV administration (52%, 95% CI: 22–70%; P = 0.003) (Supplementary Fig S6). There was no significant change in the frequency of circulating pDCs after 24F4A administration (Supplementary Fig S6). Together, these results demonstrate for the first time that administration of single dose of anti-BDCA2 mAb in cynomolgus monkeys causes rapid internalization of BDCA2 and a significant reduction in TLR9-induced IFN-I production in ex vivo whole-blood assays. The Fc region of anti-BDCA2 mAb enhances the inhibition of immune complex-induced IFN-I production by human pDCs Bivalent binding of anti-BDCA2 mAb to BDCA2 is an essential requirement for the agonistic activity of the mAb. Monovalent Fab fragments do not elicit BDCA2 signaling and do not inhibit TLR7 or TLR9-induced IFN-I production by pDCs (Jahn et al, 2010). It is not known, however, whether Fc engagement of 24F4A contributes to the cross-linking of BDCA2, thereby enhancing its ability to inhibit IFN-I. To address this possibility, we generated an Fc-effectorless form of 24F4A (24F4A-ef) with a human IgG4.P/human IgG1 chimeric Fc region that cannot bind Fc receptors (Supplementary Fig S7). PDCs were incubated in the presence of increasing concentrations of 24F4A or 24F4A-ef and stimulated with synthetic TLR7 ligand (R848), synthetic TLR9 ligand (CpG-A), or a disease-relevant ligand, SLE immune complexes (SLE-IC), Sm/RNP. Sm/RNP autoantigens, consisting of U1 RNA bound by Smith antigen (Sm), are frequently targeted by autoantibodies in SLE and are capable of activating TLR7 (Lau et al, 2005; Vollmer et al, 2005; Christensen et al, 2006). 24F4A and 24F4A-ef were equipotent at inhibiting R848 and CpG-A-induced IFNα by pDCs (Fig 4A–D). In contrast to the synthetic TLR7 or TLR9 ligands, 24F4A was 40-fold more potent at inhibiting Sm/RNP IC-induced IFNα with an average IC50 of 0.03 μg/ml, compared to 24F4A-ef, which inhibited with an average IC50 of 1.3 μg/ml (Fig 4E and F). Treatment with effector-competent isotype control antibody at the highest concentration of 10 μg/ml did not affect the induction of IFNα by R848, CpG-A, or SLE-IC (Sm/RNP) (Fig 4A, C and E). The difference in potency of 24F4A versus 24F4A-ef was not due to a difference in signaling downstream of BDCA2 engagement as 24F4A and 24F4A-ef were equally capable of inducing phosphorylation of Syk and PLCγ2 (Fig 4G). Figure 4. The Fc region of 24F4A enhances the inhibition of immune complex-induced IFNα production by human pDCs A–F. Isolated human pDCs were treated with increasing concentrations of 24F4A (red circles), 24F4A-ef (black squares), or 10 μg/ml of the isotype control (black triangle) and stimulated with 5 μM of R848 (A and B), 1 μM of CpG-A (C and D), or immune complexes (E and F) for 16 h at 37°C. IFNα concentrations in cultured supernatants were determined using ELISA. Representative plots for each stimulation (A, C and E) and the average IC50 of 24F4A and 24F4A-ef for each stimulation condition (B, D and F) are shown (n = 3). Horizontal bars represent mean IC50, and error bars represent SD of the mean from three independent experiments. G. Isolated human pDCs were treated with 10 μg/ml of 24F4A (black line), 24F4A-ef (gray shaded histogram), or the isotype control (gray dotted line) for 10 min at 37°C. Flow cytometry was used to evaluate pSYK and pPLCγ2. Shown is a representative plot of two independent experiments. H. Isolated pDCs treated with 10 μg/ml of anti-CD32a (AT10) and stimulated with CpG-A, R848, or immune complexes. Error bars represent SD of percent inhibition of IFNα from three independent experiments. Download figure Download PowerPoint While the uptake of synthetic TLR ligands is independent of Fc receptors (Vollmer et al, 2004), IC need to be internalized by CD32a (FcγRIIa) (Means et al, 2005). Although CD32a was reported to be the only Fc receptor expressed on pDCs (Bave et al, 2003), recent transcript profiling data indicated that CD32b could be expressed at similar levels in pDCs (Guilliams et al, 2014). To address this question, we evaluated the transcript levels of CD32a and CD32b in isolated pDCs using Q-PCR and confirmed the exclusive expression of CD32a in pDCs (Supplementary Fig S8). As expected, CD32a blockade led to complete inhibition of IC-mediated IFNα production by pDCs, but did not affect IFNα production mediated by synthetic TLR7 or TLR9 ligands (CpG-A or R848) (Fig 4H). Taken together, these data show that while the Fc domain of 24F4A did not impact the signaling activity of the mAb and its inhibitory activity on synthetic TLR ligands, it contributed to the potent inhi