Hemophilia A is an inherited bleeding disorder which, due to the lack of clotting FVIII, leads to frequent bleeding episodes. Until now, no cure for this disease exists but prophylactic treatment with plasma-derived or recombinant FVIII products has dramatically improved the life of hemophilia A patients. One of the biggest complications during treatment is the development of antibodies against the exogenous FVIII, which occurs in up to 30% of treated patients.1 FVIII-neutralizing antibodies render the therapy ineffective and lead to increased bleeding episodes, higher morbidity and mortality in these patients.2,3
Enormous efforts have been made in the recent years to develop new recombinant FVIII products which have an extended half-life, thereby reducing FVIII administration frequency and improving patients quality of life.4 However, any modification to the FVIII sequence poses the risk of causing unwanted immunogenicity. Thus, extensive in silico, in vitro, and in vivo analyses are required to evaluate the immunogenicity of each of these new products before they go into clinics.
The recombinant B-domain-deleted FVIII product Efmoroctocog alfa (marketed as Elocta® and Eloctate®) contains a dimeric constant Fc region of human IgG1. This Fc fusion has been reported to prolong the therapeutic half-life through binding to the neonatal Fc receptor (FcRn).5,6 Although the way the Fc domain prolongs FVIII's half-life is known, its effect on immunogenicity is still a matter of debate. While tolerogenic properties of the Fc portion of IgG have been described over the past decades,7 it is well documented that various therapeutic monoclonal antibodies induced neutralizing antibodies.8 Moreover, Fc fusion constructs have been utilized in vaccine development strategies to enhance the immune response towards the immunizing antigen.9–11 Neo-epitopes generated at the hinge region, modified uptake and processing as well as a different bio-distribution of the fusion protein could create a setting which triggers an unwanted immune response.12
The Fc domain not only interacts with the half-life-increasing FcRn, but can also engage in an interaction with various members of the Fcγ receptor (FcγR) family. These receptors are expressed on different immune cells and bind the Fc domain of IgG. In humans, the FcγR family is composed of four activating receptors (FcγRI, FcγRIIa, FcγRIIc and FcγRIIIa) and one inhibitory receptor (FcγRIIb), triggering or inhibiting immune functions, respectively.13 FcγRI is the only FcγR which binds monomeric IgG with high affinity. The other FcγRs are low-affinity IgG receptors, which interact with antibodies present as immune complexes.14
Dendritic cells (DC's), which are the most efficient antigen-presenting cells, co-express the activating receptor FcγRIIa as well as the inhibitory receptor FcγRIIb. The balance between these 2 opposing receptors determines whether DC's are activated, mature and provide signals to activate T cells and subsequently trigger an immune response.15 Recombinant FVIII (rFVIII) itself failed to induce cell maturation in the in vitro model of monocyte-derived DC's (moDC's).16 However, since the Fc domain of rFVIII-Fc is of the IgG1 subclass, it is able to interact with FcγRIIa and FcγRIIb, possibly resulting in an activating or inhibitory signal for DC's. Thus, this study aimed to compare the effect of rFVIII-Fc and rFVIII on moDC maturation by analyzing morphological signs of cell activation, measuring the expression of co-stimulatory molecules as well as the production of pro-inflammatory cytokines. Furthermore, binding studies of rFVIII-Fc and rFVIII with FCGR2A-transfected human embryonic kidney (HEK) 293 cells were carried out. The involvement of FcγRIIa in moDC activation was assessed by measuring the receptor phosphorylation and performing inhibition experiments with an FcγRIIa-specific blocking antibody.
rFVIII-Fc but not rFVIII induces moDC activation and pro-inflammatory cytokine production
DC's are the sentinels of the immune system, constantly monitoring their environment for foreign structures such as pathogens or other non-self proteins. Upon recognition and uptake of foreign structures, immature DC's are converted into mature, potent antigen-presenting cells. This maturation process is accompanied by morphological changes, such as cell clustering, the up-regulation of the expression of molecules involved in antigen presentation (HLA-DR) and co-stimulatory molecules (CD40, CD80, CD86), in addition to the production of various cytokines.
It has been reported that recombinant naïve FVIII itself does not induce DC activation and maturation in vitro.16 In order to see if this is also true for the FVIII-Fc fusion protein, moDC's generated from healthy donors were incubated for 23 hours with 5 nM rFVIII-Fc, 5 nM rFVIII or PBS and analyzed by light microscopy for morphological alterations related to DC activation and maturation. Interestingly, substantially more cell clustering was observed following rFVIII-Fc treatment compared to rFVIII or PBS treatment (Fig. 1). Clusters of DC's have been observed following treatment with various maturation inducing agents,17–19 indicating that rFVIII-Fc induces cell activation.
In order to confirm if the observed morphological changes following rFVIII-Fc treatment are also reflected by the expression of relevant cellular activation markers, the level of CCR7, CD40, CD80, CD86, CD274, and HLA-DR were determined by flow cytometry. Cells treated with PBS served as a control and were used for normalization. As shown in Figure 2, surface expression of the maturation markers HLA-DR, CD40, CD80, and CD86 increased with rFVIII-Fc concentrations above 0.5 nM for most donors, with CD40 and CD80 being significantly upregulated compared to PBS-treated cells. On the contrary, no significant alteration in the expression of these maturation markers was observed for moDC's treated with rFVIII. The expression of CCR7, a C-C chemokine receptor usually up-regulated upon DC maturation declined and the expression of the inhibitory receptor ligand CD274 significantly increased upon rFVIII-Fc treatment. rFVIII had no effect on the expression of CCR7 and CD274. The overall viability of the cells treated with the different concentrations of rFVIII-Fc and rFVIII was not altered (SDC, Fig. 1, http://links.lww.com/HS/A58).
DC activation and maturation is usually accompanied by the production of cytokines. Therefore, the amount of the pro-inflammatory cytokines IL-6 and IL-8 was analyzed in the supernatant of rFVIII-Fc- or rFVIII-treated moDC's 23 hours post stimulation via the cytometric bead array. As shown in Figure 3, the levels of both cytokines were strongly increased when the cells were treated with rFVIII-Fc. At 5, 10, and 20 nM rFVIII-Fc a statistically significant difference between rFVIII-Fc-treated cells and PBS-treated cells was observed for IL-6. Interestingly, for 3 out of 5 donors very high levels of IL-8 were detected upon rFVIII-Fc treatment. Cells treated with rFVIII showed no statistically significant difference in IL-6 and IL-8 levels compared to PBS-treated cells. On average, the IL-6 and IL-8 levels for rFVIII-Fc-treated cells were more than 10 times higher than for rFVIII-treated cells. Taken together, the increased secretion of pro-inflammatory cytokines following rFVIII-Fc treatment is in line with the previous finding that moDC's are activated by rFVIII-Fc.
Since foreign proteins exert their immunogenicity only within a certain micro-environment, lipopolysaccharide (LPS) from Pseudomonas aeruginosa was added to the moDC's following a 3-hour incubation period with rFVIII-Fc or rFVIII. As shown in Figure 4, rFVIII-Fc led to a small but significant additional increase in the expression of the maturation markers CD80, CD86, and CD274 when the cells were stimulated with LPS. In conclusion, the LPS-induced maturation signal was amplified by rFVIII-Fc, but not rFVIII.
rFVIII-Fc fusion construct is required for moDC activation
To determine whether a combination of non-covalently bound Fc- and rFVIII-proteins induce a similar moDC activation as seen for rFVIII-Fc, mixtures of 5 or 10 nM human IgG1 Fc and 5 nM rFVIII were added simultaneously to the cells (Fig. 5). Analysis of different activation markers showed no statistically significant difference between cells incubated with rFVIII in the presence or absence of IgG1 Fc. Compared to rFVIII alone, a slight tendency towards increased IL-6 and IL-8 levels was detected, when rFVIII was applied together with 5 nM IgG1 Fc to the cells. However, the increase in IL-6 and IL-8 was not statistically significant and much lower compared to rFVIII-Fc-treated cells. No effect was observed when rFVIII was added together with 10 nM IgG1 Fc. As seen before, incubation of the cells with rFVIII-Fc showed a significantly higher expression of CD40, CD80, CD86, CD274, and HLA-DR and higher levels of IL-6 and IL–8 compared to cells treated with rFVIII. These findings indicate that the strong and potent activation of the moDC's is induced by the covalently-linked FVIII-Fc fusion construct, but not by similar concentrations of a mixture of IgG1 Fc and rFVIII.
Fc fusion proteins can bind to both activating and inhibitory FcγR. Since Fc engagement with activating receptors of the FcγR family has been reported to result in immune cell activation, it is likely that the observed activation of moDC's by rFVIII-Fc is mediated by an interaction with an activating FcγR. The presence of the different FcγR on moDC's was assessed by flow cytometry and is shown in Figure 5C. While expression of FcγRI and FcγRIII was absent or low, moderate to high levels of FcγRII (FcγRIIa and FcγRIIb) were detected on moDC's. Thus, FcγRII is a potential candidate for triggering cell activation following rFVIII-Fc engagement. Moreover, a notably stronger activation following rFVIII-Fc treatment observed in 2 donors (highlighted in Fig. 5 in dark and light red) correlated with higher expression levels of FcγRII. The expression of the macrophage mannose receptor (CD206), which has been implicated in FVIII uptake,20,21 was not increased in these 2 donors. Taken together, these findings support the hypothesis that FcγRII is involved in the rFVIII-Fc-mediated activation of moDC's.
rFVIII-Fc binds to FcγRII on FCGR2A-expressing HEK293 cells but binding to moDC's does not depend on FcγRII
FVIII binding to moDC's has been reported to be mediated by various cell surface receptors, such as the low density lipoprotein receptor-related protein (LRP1), CD206, Siglec-5, galectin-1 and galectin-3.22 Since moDC's express the low-affinity FcγRII, differences in binding between rFVIII-Fc and rFVIII could occur. In order to compare the binding of rFVIII-Fc and rFVIII, moDC's were incubated with increasing FVIII concentrations at 4°C. The binding of rFVIII-Fc and rFVIII was detected with a biotin-labeled FVIII-specific nanobody, followed by an incubation with a APC-streptavidin. A concentration-dependent binding of rFVIII-Fc to moDC's was observed, although 100 nM of rFVIII-Fc were not sufficient to reach saturation (Fig. 6 (A)). Interestingly, the binding characteristics of rFVIII appeared to be similar to those of rFVIII-Fc.
Competition experiments in which the binding of 10 nM rFVIII-Fc was inhibited by the addition of increasing concentrations of rFVIII are shown in Figure 6 (B). rFVIII-Fc binding was detected with an Fc-specific antibody. rFVIII-Fc binding was efficiently inhibited with increasing concentrations of rFVIII. A decrease in rFVIII-Fc binding was detected at concentrations of more than 10 nM. Taken together, these experiments indicate that rFVIII-Fc and rFVIII binding is mediated by the same receptor and that FcγRII plays only a minor role in the binding process. Moreover, binding of rFVIII-Fc was only marginally inhibited with increasing concentrations of human monomeric IgG1 or recombinant IgG1 Fc (Fig. 6 (C)).
The binding of rFVIII and rFVIII-Fc appeared to be similar and this raised the question whether rFVIII-Fc is capable of binding to FcγRIIa. Therefore, binding studies with HEK293 cells transiently transfected with an FCGR2A-expression vector were carried out. As shown in Figure 7 (A), rFVIII-Fc efficiently bound to FcγRIIa-expressing cells. For rFVIII, the binding was low and comparable to non-transfected cells (Fig. 7 (B)). Competition with an anti-FcγRIIa antibody showed an efficient inhibition of rFVIII-Fc binding, while much higher concentrations of the isotype control were required to achieve inhibition of binding (Fig. 7 (C)). These findings support the hypothesis that rFVIII-Fc is able to bind to FcγRIIa.
rFVIII-Fc increases FcγRIIa phosphorylation
moDC's coexpress the activating (FcγRIIa) and inhibitory (FcγRIIb) isoforms of the FcγRII.15 Due to their highly homologous extracellular domains, discrimination between the 2 receptors was not possible. However, since rFVIII-Fc caused moDC activation, it can be concluded that FcγRIIa is targeted by rFVIII-Fc. Activation of FcγRIIa is mediated by phosphorylation of the immunoreceptor tyrosine-based activation motif (ITAM) in the cytoplasmic tail.23 This initiates a cascade of signaling events which results in DC maturation and enhanced expression of pro-inflammatory cytokines.15 To evaluate if FcγRIIa is targeted by rFVIII-Fc, moDC's were incubated with 5 nM rFVIII-Fc and the phosphorylation of various immune receptors, including FcγRIIa, was analyzed via a phospho-immunoreceptor antibody array. Incubation with rFVIII or PBS served as a control. As shown in Figure 8, increased phosphorylation of FcγRIIa was induced by incubating the cells with rFVIII-Fc, but not with rFVIII or PBS. These findings support the hypothesis of moDC activation by rFVIII-Fc via FcγRIIa.
FcγRIIa blockade reduces rFVIII-Fc-mediated moDC activation
In order to confirm that FcγRIIa in fact provokes activation of moDC's via FcγRIIa, cells were pre-incubated for 1 hour with an FcγRIIa-specific blocking antibody before adding rFVIII-Fc. Cells treated with the same concentration of an isotype-matched antibody served as a control. As shown in Figure 9 (A), incubation with an FcγRIIa-specific blocking antibody prior to cultivation with rFVIII-Fc reduced the cell clustering. Moreover, the FcγRIIa-specific blocking antibody significantly reduced the expression of CD40, CD80, CD86, CD274, and HLA-DR on moDC's and decreased the level of IL-6 and IL-8 in the supernatant (Fig. 9 (B) and (C)). The FcγRIIa-specific blocking antibody resulted in similarly reduced surface marker expression and cytokine levels as observed when the cells were treated with rFVIII. In contrast, pre-incubation of the isotype control antibody with rFVIII-Fc did not affect the rFVIII-Fc-induced cell clustering nor did it alter the expression of the surface markers or cytokine levels. Taken together, these findings strongly support the hypothesis that the observed activation of moDC's by rFVIII-Fc is mediated via FcγRIIa signalling.
Hemophilia A patients are dependent on the substitution with exogenous FVIII to achieve hemostasis. One of the most serious complications during FVIII replacement therapy is the development of anti-FVIII antibodies, which interfere with the infused FVIII and make replacement therapy ineffective. Thus, a major concern when bioengineering FVIII is to increase the immunogenicity of FVIII. Fusion of the human IgG1 Fc domain to FVIII has been shown to prolong the half-life of the clotting factor. However, the immunological consequences that arise from the Fc fusion are not completely understood. Clinical data show that previously treated patients with severe hemophilia did not develop neutralizing antibodies to rFVIII-Fc.24 However, these patients are less likely to develop inhibitors. Unfortunately, data from clinical trials conducted with previously untreated patients, which would shed more light on the immunogenicity of rFVIII-Fc, are not yet available.
Extrapolating immunological data from other Fc-containing therapeutic proteins appear to be difficult. Among other factors, it depends on the protein that carries the Fc part as well as the applied in vitro or in vivo models. While some publications support the idea that Fc fusion proteins can promote tolerance,7 it should be noted that a number of human monoclonal antibodies induce anti-drug antibodies in a substantial number of patients.8 Moreover, Fc fusion vaccines have been shown to induce a much stronger antigen-specific immune response compared to vaccines with the native antigen.9,25,26 Thus, careful evaluation of immunological properties of a new Fc fusion construct is crucial.
DC activation and maturation leads to various changes, such as an reduction in antigen uptake, enhanced presentation of antigen-derived peptides by MHC class II, altered expression of chemokine receptors and adhesion molecules involved in migration27 as well as the formation of cell clusters.17 These functional alterations are tightly linked to the reorganization of the cytoskeleton28 and are believed to be required for the DC's to efficiently present antigens to T cells.29 The in vitro moDC model applied in this study revealed several hallmarks of cell activation following rFVIII-Fc exposure as described above. Cell clustering, up-regulation of co-stimulatory molecules such as CD40 and CD80, as well as increased levels of the pro-inflammatory cytokines IL-6 and IL-8 were exclusively observed after treatment of cells with rFVIII-Fc, but not rFVIII. In contrast to rFVIII-Fc, the mixture of rFVIII and purified IgG1 Fc did not trigger moDC activation.
Although rFVIII-Fc treatment predominantly induced signals typical of cell activation and maturation, reduced levels of the cytokine receptor CCR7 and increased levels of the inhibitory receptor ligand CD274 were detected as well. CCR7 is implicated in migratory activity of DC's and a reduction in this receptor might have functional consequences with respect to the homing of DC's to the lymphoid organs. While the increased expression of CD40, CD80, CD86 and HLA-DR following rFVIII-Fc treatment provides a strong stimulatory signal for T-cell activation, an increased expression of CD274 (also called programmed death-ligand 1 or PD-L1) on DC's transmits an inhibitory signal to T cells via PD-1.30 Engagement of CD274 with PD-1 on T cells can block T cell proliferation, cytokine production and inhibit T cell survival.31 However, this inhibitory signal can be overturned by CD28 co-stimulation or IL-2.32,33 Moreover, up-regulation of CD274 has also been observed following stimulation with pro-inflammatory cytokines in various cell types,34,35 and appears to be a regulatory feedback mechanism to limit the intensity of DC activation. In line with this, an up-regulation of CD274 following LPS stimulation after 23 hours was observed (SDC, Fig. 2, http://links.lww.com/HS/A58), while all other co-stimulatory molecules as well as HLA-DR were upregulated. Thus, the up-regulation of CD274 following rFVIII-Fc treatment probably reflects a common physiological regulatory feedback mechanism that follows a strong stimulation event rather than indicating an overall inhibitory effect of rFVIII-Fc.
Although a dose-dependent increase in the expression of the surface markers was detected following rFVIII-Fc cultivation, it appears that rFVIII-Fc did not activate all moDC's (SDC, Figure 3, http://links.lww.com/HS/A58). Especially for the expression of CD86 and HLA-DR, two populations appeared in the histogram following rFVIII-Fc stimulation. When 5 nM or more of rFVIII-Fc were applied to the cells, more than 40% of the living cells were activated, that is, high in CD86 and HLA-DR expression. The high expression of CD86 correlated with the high expression of HLA-DR, showing that these are the same cells. Activation of all cells might be achieved by using higher concentrations of rFVIII-Fc.
The observation that rFVIII-Fc causes the activation of moDC's is in part contradictory to previous results from 2 other studies. Krishnamoorthy et al investigated the immune response of rFVIII-Fc in FVIII-deficient mice.36 The immune response following rFVIII-Fc treatment appeared to be strongly concentration-dependent in this model. At 50 and 100 IU/kg rFVIII-Fc fewer total anti-FVIII antibodies were detected compared to a treatment with recombinant FVIII lacking the Fc domain. Moreover, rFVIII-Fc appeared to induce the expression of molecules involved in immune suppression or tolerance induction, including an increased expression of CD274 on murine DC, which is in agreement with our results. By contrast, treatment with 250 IU/kg rFVIII-Fc resulted in increased levels of non-neutralizing and neutralizing anti-FVIII antibodies and significantly boosted T cell proliferation and IFN-γ production. It is important to note, that the source of study material was different in the study by Krishnamoorthy: Human embryonic kidney cell-derived rFVIII-Fc was compared to rFVIII products produced in Chinese hamster ovary cells. Both expression systems differ in the nature of their post-translational modifications, which might have an impact on their potential to induce an immune response. Moreover, the applied hemophilia A mice model is intrinsically limited with respect to evaluating Fc fusion constructs, as Fc receptors differ between humans and mice. Thus, certain immune responses triggered by the human Fc domain might not be detected in the murine model. In the other rFVIII-Fc-related study by Kis-Toth using human monocyte-derived macrophages,37 rFVIII-Fc induced phenotypical alteration, which the authors interpreted as regulatory macrophage polarization. However, it is worth noting that these phenotypic alterations were achieved following treatment with a very high concentration of rFVIII-Fc (200 nM or 400 U/ml), which is more than 100 times higher than the naturally occurring concentration.
The findings presented here give evidence that the observed activating effect of rFVIII-Fc is mediated by the interaction with FcγRIIa, which is expressed together with FcγRIIb on moDC's.15 Blockade of FcγRIIa almost completely abolished the effect of rFVIII-Fc. Moreover, a rapid increase of FcγRIIa phosphorylation was detected upon incubation with rFVIII-Fc, which supports the hypothesis that rFVIII-Fc is capable of activating FcγRIIa signaling. In line with this, two donors with significantly increased FcγRII expression levels showed stronger activation signals upon rFVIII-Fc incubation.
The ability of rFVIII-Fc to bind to FcγRIIa was shown with FCGR2A-transfected HEK293 cells. Here rFVIII-Fc bound efficiently to the transfected cells, while the binding of rFVIII was low and comparable between FCGR2A-transfected and empty vector-transfected HEK293 cells. Interestingly, on moDC's the binding pattern did not differ between rFVIII-Fc and rFVIII, indicating that FcγRIIa does not play an important role in the overall rFVIII-Fc binding. This is not surprising, since FcγRIIa has a low affinity to monomeric IgG's and binds only multimeric or aggregated IgG efficiently. The applied rFVIII-Fc and rFVIII preparations contained only a very low percentage of aggregates and the ratio of FVIII activity to FVIII antigen was comparably high for both recombinant FVIII products (SDC, Fig. 4, http://links.lww.com/HS/A58). Therefore, main FVIII binding is most likely mediated by FVIII-binding receptors. This is supported by the finding that rFVIII-Fc binding is efficiently inhibited with rFVIII and only high concentrations of human IgG1 or Fc are capable of reducing rFVIII-Fc binding (Fig. 6).
Several receptors which are involved in FVIII binding, uptake and immune-recognition by DC's have been identified 22,38 and could, together with FcγRIIa, be responsible for the observed cell activation. Potential interaction partners that are present on moDC's are LRP1, CD206, Siglec-5,39 galectin-1, and galectin-3.40 LRP1 binds FVIII with moderate affinity and was shown to be involved in the clearance of FVIII in mice.41 However, in a human in vitro model LRP1 does not seem to play a role in FVIII uptake and activation of T cells.21,42 The impact of CD206 on FVIII uptake and immunogenicity in moDC has been a matter of debate: While in one study saturating and blocking CD206 reduced the uptake of FVIII in moDC's and decreased T-cell activation,20 no inhibitory effect on FVIII binding and uptake was detected in another study.21 Similar to the latter results, we did not observe a reduction of rFVIII-Fc or rFVIII binding if the cells were pre-incubated with different concentrations of mannan (SDC, Fig. 5, http://links.lww.com/HS/A58). Amongst the FVIII-binding lectins, galectin-1 and -3 appear to have the highest affinity towards FVIII. 40 However, whether these lectins are involved in FVIII uptake and the extent to which they contribute to FVIII immunogenicity remains to be established.
moDC activation relies on the presence of the rFVIII-Fc fusion construct. Neither rFVIII or IgG1 Fc alone, nor rFVIII together with IgG1 Fc (applied as 2 separate proteins) have a comparable activating effect on moDC's. Moreover, the addition of human IgG1 Fc together with rFVIII-Fc is also not capable of blocking the rFVIII-Fc-induced cell activation (SDC, Fig. 6, http://links.lww.com/HS/A58). Thus, we speculate that close proximity of rFVIII and Fc, as it is achieved in the fusion construct, is required to cluster FVIII receptors together with FcγRIIa. This initiates cellular signaling cascades that lead to moDC activation and maturation. It has been assumed that ligand binding to members of the FCɣR family leads to the activation of Src-family kinases, which in turn phosphorylate the ITAM motif present in the cytosolic domain of the receptor. The phosphorylation creates a docking site for the kinase Syk, which then phosphorylates down-stream substrates, leading to the initiation of various signaling pathways and cell activation.13,43 Interestingly, Src-and Syk-initiated signaling cascades following immune-receptor stimulation fed into the Rho/WASP pathway, which plays an important role in modulating the structure of the actin cytoskeleton.44 It is tempting to speculate that the observed cell-clustering process is rFVIII-Fc-driven and functions via the Src- and Syk-initiated signaling pathways.
MoDC's express both activating and inhibitory FcγRII's, which possess opposing functions due to the presence of the cytosolic immunoreceptor tyrosine-based activating or inhibitory motif. While ligation of FcγRIIa on DC's was shown to induce cell maturation, increase stimulation of T cells and elevate the expression of pro-inflammatory cytokines, binding to FcγRIIb rather inhibits the immune response.15 Based on the data presented here, rFVIII-Fc seems to specifically target FcγRIIa and not or at least to a lesser extent FcγRIIb. Given the structural similarity of the extracellular domain of the two FcγRII with 89% sequence identity, the question remains: Why does rFVIII-Fc signal via FcγRIIa? Affinity and preference towards these highly related receptors are not well understood. However, one factor that greatly influences the binding to Fc receptors, and thus the biological activity of the Fc domain, is its glycosylation. For example, terminal sialylation of the Fc domain has been shown to be important for the anti-inflammatory activity of IgG's,45,46 probably by promoting an anti-inflammatory milieu through the interaction with sialic acid recognizing receptors.47 Interestingly, Kannicht et al showed that the Fc domain of rFVIII-Fc contains a remarkably low level of sialylation (less than 1%).48 In contrast, the sialylation degree of the Fc domain of plasma-derived human IgG's is around 13%.49 Thus, it can be speculated, that the very low sialic acid content of the fusion construct (and the abolishment of binding to sialic acid-specific receptors) in combination with the binding to FVIII-specific receptors and a co-ligation with FcγRIIa supports signaling pathways that drive DC activation.
The in vitro model of human moDC's applied in this study has been extensively used in basic research to study DC biology and was implemented successfully in numerous clinical trials.50 While the ready availability of large quantities of cells is a big advantage of this model, it should be kept in mind that moDC's differ in terms of function from in vivo-occurring human DC subsets. In general, moDC's share many features of DC's following stimulation: They are capable of capturing, processing and presenting antigens to T cells and are able to produce important chemokines and cytokines.51–53 However, recent evidence suggests that moDC's are rather inflammatory monocytes with DC-like functions.54,55 Thus, translation to human DC's in vivo should be made with caution. Analysis of naturally in vivo-occurring DC subsets, such as plasmacytoid or conventional DC's would be of great value to ascertain whether rFVIII-Fc behaves similarly to the moDC's.
Here we report for the first time that rFVIII-Fc leads to activation of moDC's at therapeutically relevant doses in vitro. Mechanistically, this effect appears to be mediated by FcγRIIa. Although two other studies report tolerogenic properties of rFVIII-Fc, the presented interim study with previously untreated patients 56 did not show lower rates of inhibitor development in this patient group compared to conventional rFVIII products, indicating that rFVIII-Fc is not tolerogenic per se. Moreover, studies with hemophilia mice, which do not express the genetic equivalent of the human FcγRIIa,13 might underestimate the unwanted immunological consequences caused by the Fc fusion construct. Thus, careful investigations in appropriate in vivo and in vitro models are required to better understand the immunogenicity of Fc fusion proteins.
Materials and methods
Two batches of the 2 different recombinant B-domain deleted FVIII product, rFVIII-Fc (Elocta®, 2000 Units, Sobi) and rFVIII (Nuwiq®, 2000 Units, Octapharma), were dissolved in sterile water according to the manufacturers recommendations, aliquoted and stored at -80°C until further usage. FVIII activity and antigen concentration was determined by chromogenic FVIII assay (Chromogenix) and FVIII antigen ELISA (ASSERACHROM VIII:Ag, Stago), respectively. Plasmatic human IgG1 was obtained from Athens Research & Technology, recombinant human IgG1 Fc (Thr106-Lys330, 293E expressed) from Biolegend. Fluorophore-coupled monoclonal antibodies against human HLA-DR (Clone L234), CD40 (Clone 5C3), CD86 (Clone BU63), CCR7 (Clone G043H7), CD274 (Clone 29E.2A3), FcγRI (Clone 10.1), FcγRII (Clone FUN-2), FcγRIII (Clone 3G8), CD206 (Clone 15–2), APC-streptavidin were purchased from Biolegend and against human CD80 (Clone L397.4) from BD. APC-coupled F(ab’)2 fragment specific for human IgG Fc-part was purchased from Jackson Immuno Research. Tag-free expression vector containing the sequence of human FCGR2A (NM_001136219) as well as an empty control vector (pCMV6-XL5) were purchased from OriGene.
Monocyte purification, moDC stimulation, and flow cytometric analysis
Monocytes from healthy donors, obtained from buffy coats from the DRK-Blutspendedienst Ost, were enriched via Ficoll (GE Healthcare) gradient and subsequently CD14+ monocytes were purified by magnetic cell sorting (CD14+ monocyte isolation Kit, Miltenyi Biotec). The purity of the cells was measured by flow cytometry and was always more than 85% of the living cells. To obtain immature moDC's, CD14+ monocytes were cultivated for 5 to 6 days in RPMI medium supplemented with 10% fetal calf serum (FCS, Biochrome), 100 U/ml penicillin, 100 μg/ml streptomycin, 100 U/ml granulocyte-macrophage colony-stimulating factor (GMCSF, R&D) and 200 U/ml interleukin-4 (IL-4, R&D). In order to analyze the effect of the FVIII products on cell maturation, 2.5x105 moDC's per well of 48-well plate were incubated for 23 hours with indicated concentrations of FVIII. Cells treated with the same volume of phosphate buffered saline (PBS) served as a control. Morphological alterations of the cells following rFVIII-Fc, rFVIII or PBS treatment were documented with the Axio Observer Z1 microscope (Zeiss) using a 100-fold magnification. Since the micro-environment can affect the immunogenicity of a foreign protein, the impact of the FVIII products in the presence of LPS was investigated. Therefore, moDC's were pre-incubated for 3 hours with different concentrations of FVIII and then stimulated with 1 μg/ml LPS from P aeruginosa serotype 10 (Sigma) for 20 hours. Expression of surface markers involved in T cell stimulation (CCR7, CD40, CD80, CD86, CD274, and HLA-DR) was determined subsequently by flow cytometry. In order to detect fluorescence signals obtained from unspecific binding of the antibodies to dead cells, cells were treated with Zombie Aqua dye (Biolegend). Samples were analyzed with the flow cytometer FACSVerse (BD). Final examination of the data was carried out using FlowJo software (Tree Star, Inc.). To normalize for variability between the different donors, expression of surface receptors was expressed as the ratio of the median fluorescence intensity (MFI) by dividing the MFI of FVIII-incubated cells by the MFI of cells incubated with the same volume of PBS.
In parallel to the quantification of surface markers, the cytokine profile secreted by the moDC's was analyzed 23 hours post stimulation following supernatant collection via the cytometric bead array CBA Flex (BD) detecting simultaneously interleukin-6 (IL-6) and interleukin-8 (IL-8) according to the manufacturer's recommendation. Samples were analyzed with the flow cytometer FACSVerse (BD). Final analysis and calculation of the cytokine concentration were carried out using FCAP Array software (BD).
Binding experiments were carried out with moDC's or transiently-transfected HEK293 cells. For expression of human FcγRIIa, 2x106 HEK293 cells were seeded in a 10 cm dish in Roti-CELL DMEM High Glucose (Roth) supplemented with 5% FCS, 100 U/ml penicillin and 100 μg/ml streptomycin. On the following day, cells were transfected using Lipofectamine 2000 Transfection Reagent according to the manufacturer's recommendation with 10 μg of the expression vector for FcγRIIa or empty vector. On the next day, successful transfection was verified by flow cytometry and cells were subjected to binding studies.
For rFVIII binding studies, 2x105 moDC's or transfected HEK293 cells per test were washed with FACS-buffer (PBS containing 1% FCS) and incubated with various concentrations of rFVIII-Fc or rFVIII for 30 minutes on ice. Following two wash steps with FACS buffer, cells were incubated with the CaptureSelect Biotin Anti-FVIII Conjugate (ThermoFisher Scientific) for 30 minutes on ice. Unbound detection reagent was removed by washing once with FACS buffer. For detecting bound rFVIII, APC streptavidin was applied to the cells for 30 minutes on ice, followed by two wash steps with FACS buffer. MFI of APC was quantified via flow cytometry. For binding inhibition experiments, the cells were pre-treated for 10 minutes on ice with various concentrations of human IgG1 or recombinant human IgG1 Fc, followed by the addition of 10 nM rFVIII-Fc for further 30 minutes on ice in the first step. Detection of bound rFVIII-Fc was carried out as described above.
For competition binding experiments, 2×105 moDC's per test were washed with FACS-buffer and pre-treated with various concentrations of rFVIII for 10 minutes on ice, followed by the addition of 10 nM rFVIII-Fc and an incubation for 30 minutes on ice. After two wash steps with FACS buffer, rFVIII-Fc-treated cells were incubated with Fc-specific APC AffiniPure F(ab’)2 fragment for 30 minutes on ice. Unbound streptavidin was removed by washing with FACS buffer. Binding of rFVIII-Fc was determined by measuring the MFI of APC via flow cytometry. Final examination of the data was carried out using FlowJo software.
Human phospho-immunoreceptor array
For determining the influence of FVIII on tyrosine phosphorylation of human immunoreceptors, moDC's were incubated for 10 minutes with 5 nM rFVIII-Fc or rFVIII and then lysed. PBS-treated cells served as a control. Protein concentration was determined via BCA (Pierce) and the Proteome Profiler Human Phospho-Immunoreceptor Array (R&D systems) was carried out according to the manufacturer's recommendations using 500 to 1000 μg cell lysate. Images of the membranes were captured using GeneSys imaging software on a G:BOX Chemi system. Raw volume of the spots was determined using GeneTools analysis software (Syngene) and fold change of raw volume of FVIII-stimulated cells over raw volume of PBS-stimulated cells was calculated using Excel.
FcγRIIa-specific blocking experiment
2.5×105 moDC's per well of 48-well plate were pre-incubated with 1 mg/ml FcγRIIa-specific blocking antibody (clone IV.3, Hölzel Biotech) or with the same concentration of LEAF™ Purified Mouse IgG2b, κ Isotype control antibody (Biolegend). After 1 hour, 5 nM rFVIII-Fc were added and the cells were incubated for another 23 hours. Cells not treated with the blocking antibody or the isotype control antibody served as a control. Analysis of surface marker expression and cytokine levels were carried out as described above.
The authors gratefully acknowledge the expertise of Aaron Juche and the generosity and commitment of the blood donors of the DRK.
1. Ing M, Gupta N, Teyssandier M, et al. Immunogenicity of long-lasting recombinant factor VIII products. Cell Immunol.
2. Kessler C, Oldenburg J, Ettingshausen CE, et al. Spotlight on the human factor: Building a foundation for the future of haemophilia A management: Report from a symposium on human recombinant FVIII at the World Federation of Hemophilia World Congress, Melbourne, Australia on 12 May 2014. Haemophilia.
2015;21 (Suppl 1):1–12.
3. O’Hara J, Hughes D, Camp C, et al. The cost of severe haemophilia in Europe: The CHESS study. Orphanet J Rare Dis.
4. Mannully ST LNR, Pulicherla KK. Perspectives on progressive strategies and recent trends in the production of recombinant human factor VIII. Int J Biol Macromol.
5. Powell JS, Josephson NC, Quon D, et al. Safety and prolonged activity of recombinant factor VIII Fc fusion protein in hemophilia A patients. Blood.
6. Dumont JA, Liu T, Low SC, et al. Prolonged activity of a recombinant factor VIII-Fc fusion protein in hemophilia A mice and dogs. Blood.
7. Blumberg RS, Lillicrap D. Tolerogenic properties of Fc portion of IgG and its relevance to the treatment and management of hemophilia. Blood.
8. Harding FA, Stickler MM, Razo J, et al. The immunogenicity of humanized and fully human antibodies: residual immunogenicity resides in the CDR regions. MAbs.
9. Guyre PM, Graziano RF, Goldstein J, et al. Increased potency of Fc-receptor-targeted antigens. Cancer Immunol Immunother.
10. Konduru K, Bradfute SB, Jacques J, et al. Ebola virus glycoprotein Fc fusion protein confers protection against lethal challenge in vaccinated mice. Vaccine.
11. Pleass RJ. Fc-receptors and immunity to malaria: From models to vaccines. Parasite Immunol.
12. Levin D, Golding B, Strome SE, et al. Fc fusion as a platform technology: Potential for modulating immunogenicity. Trends Biotechnol.
13. Nimmerjahn F, Ravetch JV. Fcgamma receptors as regulators of immune responses. Nat Rev Immunol.
14. Bruhns P, Iannascoli B, England P, et al. Specificity and affinity of human Fcgamma receptors and their polymorphic variants for human IgG subclasses. Blood.
15. Boruchov AM, Heller G, Veri M-C, et al. Activating and inhibitory IgG Fc receptors on human DCs mediate opposing functions. J Clin Invest.
16. Pfistershammer K, Stockl J, Siekmann J, et al. Recombinant factor VIII and factor VIII-von Willebrand factor complex do not present danger signals for human dendritic cells. Thromb Haemost.
17. Delemarre FG, Hoogeveen PG, de Haan-Meulman M, et al. Homotypic cluster formation of dendritic cells, a close correlate of their state of maturation. Defects in the biobreeding diabetes-prone rat. J Leukoc Biol.
18. Termeer C, Johannsen H, Braun T, et al. The role of CD44 during CD40 ligand-induced dendritic cell clustering and maturation. J Leukoc Biol.
19. Lokshin AE, Kalinski P, Sassi RR, et al. Differential regulation of maturation and apoptosis of human monocyte-derived dendritic cells mediated by MHC class II. Int Immunol.
20. Dasgupta S, Navarrete AM, Bayry J, et al. A role for exposed mannosylations in presentation of human therapeutic self-proteins to CD4+ T lymphocytes. Proc Natl Acad Sci U S A.
21. Herczenik E, van Haren SD, Wroblewska A, et al. Uptake of blood coagulation factor VIII by dendritic cells is mediated via its C1 domain. J Allergy Clin Immunol.
22. Hartholt R, van Velzen AS, Peyron I, et al. To serve and protect: The modulatory role of von Willebrand factor on factor VIII immunogenicity. Blood Rev.
23. Isakov N. ITAMs: immunoregulatory scaffolds that link immunoreceptors to their intracellular signaling pathways. Recep Channels.
24. Mahlangu J, Powell JS, Ragni MV, et al. Phase 3 study of recombinant factor VIII Fc fusion protein in severe hemophilia A. Blood.
25. Zaharatos GJ, Yu J, Pace C, et al. HIV-1 and influenza antigens synthetically linked to IgG2a Fc elicit superior humoral responses compared to unmodified antigens in mice. Vaccine.
26. Soleimanpour S, Farsiani H, Mosavat A, et al. APC targeting enhances immunogenicity of a novel multistage Fc-fusion tuberculosis vaccine in mice. Appl Microbiol Biotechnol.
27. Banchereau J, Briere F, Caux C, et al. Immunobiology of dendritic cells. Annu Rev Immunol.
28. Verdijk P, van Veelen PA, de Ru AH, et al. Morphological changes during dendritic cell maturation correlate with cofilin activation and translocation to the cell membrane. Eur J Immunol.
29. Burns S, Thrasher AJ. Dendritic cells: the bare bones of immunity. Curr Biol.
30. Chen L. Co-inhibitory molecules of the B7-CD28 family in the control of T-cell immunity. Nat Rev Immunol.
31. Riley JL. PD-1 signaling in primary T cells. Immunol Rev.
32. Freeman GJ, Long AJ, Iwai Y, et al. Engagement of the PD-1 immunoinhibitory receptor by a novel B7 family member leads to negative regulation of lymphocyte activation. J Exp Med.
33. Carter L, Fouser LA, Jussif J, et al. PD-1:PD-L inhibitory pathway affects both CD4(+) and CD8(+) T cells and is overcome by IL-2. Eur J Immunol.
34. Loke P’n, Allison JP. PD-L1 and PD-L2 are differentially regulated by Th1 and Th2 cells. Proc Natl Acad Sci U S A V 100.
35. Mazanet MM, Hughes CCW. B7-H1 is expressed by human endothelial cells and suppresses T cell cytokine synthesis. J Immunol.
36. Krishnamoorthy S, Liu T, Drager D, et al. Recombinant factor VIII Fc (rFVIIIFc) fusion protein reduces immunogenicity and induces tolerance in hemophilia A mice. Cell Immunol.
37. Kis-Toth K, Rajani GM, Simpson A, et al. Recombinant factor VIII Fc fusion protein drives regulatory macrophage polarization. Blood Adv.
38. Lai JD, Georgescu MT, Hough C, et al. To clear or to fear: An innate perspective on factor VIII immunity. Cell Immunol.
39. Pegon JN, Kurdi M, Casari C, et al. Factor VIII and von Willebrand factor are ligands for the carbohydrate-receptor Siglec-5. Haematologica.
40. O'Sullivan JM, Jenkins PV, Rawley O, et al. Galectin-1 and Galectin-3 constitute novel-binding partners for factor VIII arterioscler. Thromb Vasc Biol.
41. Bovenschen N, Herz J, Grimbergen JM, et al. Elevated plasma factor VIII in a mouse model of low-density lipoprotein receptor-related protein deficiency. Blood.
42. Dasgupta S, Navarrete AM, André S, et al. Factor VIII bypasses CD91/LRP for endocytosis by dendritic cells leading to T-cell activation. Haematologica.
43. Lowell CA. Src-family and Syk kinases in activating and inhibitory pathways in innate immune cells: signaling cross talk. Cold Spring Harb Perspect Biol.
44. Mócsai A, Ruland J, Tybulewicz VLJ. The SYK tyrosine kinase: a crucial player in diverse biological functions. Nature reviews. Immunology.
45. Kaneko Y, Nimmerjahn F, Ravetch JV. Anti-inflammatory activity of immunoglobulin G resulting from Fc sialylation. Science.
46. Anthony RM, Wermeling F, Karlsson MCI, et al. Identification of a receptor required for the anti-inflammatory activity of IVIG. Proc Natl Acad Sci U S A.
47. Nimmerjahn F, Ravetch JV. Fc-receptors as regulators of immunity. Adv Immunol.
48. Kannicht C, Kröning M, Solecka-Witulska BA, et al. Comparative N-Glycosylation Analysis of the Fc Portions of a Chimeric Human Coagulation Factor VIII and Immunoglobulin G1 //Comparative N-Glycosylation Analysis of the Fc Portions of a Chimeric Human Coagulation Factor VIII and Immunoglobulin G1. Bioengineering (Basel).
49. Flynn GC, Chen X, Liu YD, et al. Naturally occurring glycan forms of human immunoglobulins G1 and G2. Mol Immunol.
50. Ahmed MS, Bae Y-S. Dendritic cell-based therapeutic cancer vaccines: past, present and future. Clin Exp Vaccine Res.
51. Bures EJ, Hui JO, Young Y, et al. Determination of disulfide structure in agouti-related protein (AGRP) by stepwise reduction and alkylation. Biochemistry.
52. Chow KV, Lew AM, Sutherland RM, et al. Monocyte-derived dendritic cells promote Th polarization, whereas conventional dendritic cells promote Th proliferation. J Immunol.
53. Kuhn S, Yang J, Ronchese F. Monocyte-derived dendritic cells are essential for CD8(+) T cell activation and antitumor responses after local immunotherapy. Front Immunol.
54. Sander J, Schmidt SV, Cirovic B, et al. Cellular differentiation of human monocytes is regulated by time-dependent Interleukin-4 signaling and the transcriptional regulator NCOR2. Immunity.
55. Segura E, Touzot M, Bohineust A, et al. Human inflammatory dendritic cells induce Th17 cell differentiation. Immunity.
56. Königs C. Inhibitors in Previously Untreated Patients With Severe Haemophilia A Treated With rFVIIIFc: PIPs A-LONG Study, Interim data: EAHAD 2019. EAHAD 2019; February 6, 2019; Prague, Czech Republic.