Abstract
Polyclonal anti-D has been used to prevent RhD-negative mothers from becoming immunized against RhD positive fetal erythrocytes, and this mechanism has been referred as Ab or IgG-mediated immune suppression (AMIS). Although anti-D has been highly successful, the inhibitory mechanisms remain poorly understood. Two major theories behind AMIS involve the binding of IgG to activating or inhibitory FcγR, which can induce either erythrocyte clearance or immune inhibition, respectively. In this work, we explored the absolute role of activating and inhibitory FcγR in the AMIS mechanism using the HOD mouse model of RBC immunization. HOD mice contain a RBC-specific recombinant protein composed of hen egg lysozyme (HEL), OVA and human transmembrane Duffy Ag, and erythrocytes from HOD mice can stimulate an immune response to HEL. To assess the contribution of activating and inhibitory FcγR to AMIS, C57BL/6 versus FcRγ-chain−/− or FcγRIIB−/− mice were used as recipients of HOD-RBC alone or together with anti-HEL Abs (i.e., AMIS) and the resulting immune response to HEL evaluated. We show that anti-HEL polyclonal Abs induce the same degree of AMIS effect in mice lacking these IgG binding receptors as compared with wild-type mice. In agreement with this, F(ab′)2 fragments of the AMIS Ab also significantly reduced the Ab response to the HOD cells. In conclusion, successful inhibition of in vivo Ab responses to HOD-RBC by polyclonal IgG can occur independently of activating or inhibitory FcγR involvement. These results may have implications for the understanding of RhD prophylaxis.
Introduction
During pregnancy, mothers can be immunized by Rh-incompatible fetal RBCs. The subsequent transplacental passage of these Abs can result in fetal RBC destruction, inducing hemolytic disease of the fetus and newborn (HDFN) (1). The administration of polyclonal anti-D Abs to RhD-negative mothers during gestation and immediately after delivery prevents anti-D immunization (2) and is the most successful clinical application of Ab or IgG-mediated immune suppression (AMIS) (3, 4).
The mechanism behind the feedback-inhibition mediated by anti-D is still largely unknown. Studies in humans using monoclonal Abs to mimic the effects of polyclonal anti-D have had, at best, mixed results (5, 6), and several mAbs unexpectedly cause immune enhancement (7). Our inability to properly mimic polyclonal anti-D, both for the prevention of HDFN and as well as the treatment of immune thrombocytopenia (ITP), has actually led to the development of a manufacturing process to generate recombinant polyclonal mixtures of anti-D (8) for preventing HDFN and treating ITP (9). This recombinant polyclonal Ab worked well in the majority of patients with ITP (10), but there have been no reports related to HDFN. Clearly, we require a better understanding of how polyclonal anti-D actually works to prevent HDFN. Although mice expressing an RhD Ag on erythrocytes to study AMIS would be best (11), currently, only solubilized RhD protein (12) and RhD-based peptide immunization (13) have been used to generate immune responses in mice. Although useful for other avenues of research, these models are not sufficient to study AMIS mechanisms.
Several mechanisms have been proposed to explain the inhibition by IgG, and FcRs are thought to play a central role in two of the most widely accepted theories. In the first hypothesis, IgG-coated RBCs are rapidly eliminated by phagocytosis mediated by the binding of the sensitized RBC to activating FcγR on macrophages in the spleen before the RBC are able to induce an immune response (14). The second hypothesis involves the inactivation of the Ag-specific B cells mediated by the inhibitory FcγRIIB. In this hypothesis, the sensitized RBCs simultaneously engage the inhibitory FcγRIIB and the BCR, delivering a negative signal when the ITIM in the cytoplasmatic tail of the FcγRIIB is brought in close proximity to the ITAM of the BCR (15).
The Fc dependance of IgG-mediated suppression has been studied in mice immunized with native and haptened SRBC (16, 17). Using the SRBC model, IgG can suppress the immune response in mice lacking functional FcγRI, III and IV (FcRγ-chain deficient mice), FcγRIIB, neonatal FcR (β2-microglobulin–deficient mice), as well as in double-knockout mice lacking both the FcRγ-chain and FcγRIIB. In addition, anti-hapten F(ab′)2 fragments of monoclonal Abs were able to suppress the Ab response to haptened SRBC (16). These findings argue against a dominant role of FcγRs in AMIS mechanisms when the SRBC model is used.
Although the transfusion of xenogenic SRBC has been widely used to study IgG-mediated immune suppression mechanisms by us (12, 13, 18–21) and others (16, 22–25), validation against an allogenic mouse model is essential as SRBC do not meaningfully enter the mouse circulation and also contain inflammatory ligands and the actual Ags on the SRBC have not been definitively established (3, 26). Contrary to the transfusion of allogeneic mouse RBC that circulate with a normal life span (27, 28), SRBC are cleared from circulation within a few hours (29, 30).
Several different transgenic mouse models of erythrocyte alloimmunization expressing well-defined Ags on the RBCs have been more recently developed. These models include human blood group Ags (glycophorin A, Fyb, or Kell) (31–33) and also model Ags that facilitate the analysis of immune mechanisms (mHEL or HOD) (34, 35). HOD transgenic mice contain an RBC-specific recombinant protein composed of the hen egg lysozyme (HEL) in tandem sequence with chicken OVA and the human transmembrane Duffy Ag (34). These models allow the use of donors and recipients of the same species with well-defined antigenic differences, and because the transgene is present on donor RBCs but absent in the recipients for each of these systems, they more closely resemble the RhD phenomenon as compared with studies with SRBC. The development of these new RBC immunization platforms has allowed us to assess the role of FcγRs in IgG-mediated inhibition in a fully allogenic mouse model of RBC immunization.
In this work, we evaluate the absolute role of activating and inhibitory FcγRs in the AMIS mechanism using the HOD mouse model of erythrocyte immunization and polyclonal Abs against the immunogenic regions of the HOD Ag. We demonstrate in this paper that succesful reduction of the Ab response to the HEL portion on HOD-RBCs mediated by polyclonal IgG can occur via an Fc and FcγR-independent pathway. These results have potential implications for the understanding of RhD prophylaxis.
Materials and Methods
Mice
C57BL/6 mice were purchased from Charles River Laboratories (Montreal, QC, Canada). FcγRIIB−/− (B6129S background; stock number 002848) and B6129S/F2 mice (stock number 101045) were obtained from The Jackson Laboratory (Bar Harbor, ME). FcRγ-chain (FcRγ−/−) (model 583) and FcγRIIB−/− on a C57BL/6 background (model 580) were purchased from Taconic Farms (Germantown, NY). HOD mice on the FVB background consisted of transgenic animals expressing a fusion protein composed of HEL, OVA, and Duffy b (34). All animal studies were approved by the St. Michael’s Hospital animal care committee.
Antibodies
The polyclonal anti-HEL Ab was produced and purified in our laboratory as described previously (36). In brief, C57BL/6 mice were immunized with three doses of 200 μg HEL (Sigma-Aldrich, St Louis, MO) in Freund’s adjuvant (complete Freund adjuvant for the first dose and incomplete for the others) (Chondrex, Redmond, WA), and IgG was purified using protein G affinity chromatography. F(ab′)2 fragments were prepared by digesting purified polyclonal anti-HEL IgG with immobilized pepsin using the Pierce F(ab′)2 Preparation Kit (Thermo Scientific, Rockford, IL). The digested material was passed over a protein A column, which was prebound with rabbit IgG specific for mouse IgG, Fc (Jackson ImmunoResearch Laboratories, Indianapolis, IN) to remove any intact IgG from the F(ab′)2 preparation. The flow-through fraction was tested for contaminating undigested IgG by ELISA using HEL protein and detecting intact IgG with an alkaline phosphatase F(ab′)2 fragment goat anti-mouse IgG, Fc fragment–specific reagent (Jackson ImmunoResearch Laboratories). All F(ab′)2 preparations were dialyzed against PBS (pH 7.22), filtered, and stored at −20°C. The ability of F(ab′)2 fragments to bind HOD erythrocytes was assessed by flow cytometry (37).
Transfusion and AMIS induction
Blood from HOD mice was collected by the sephanous vein into PBS containing 1% EDTA as an anticoagulant. HOD-RBC were washed three times with PBS, and the concentration was adjusted to 108 cells/ml. Mice were transfused (tail vein) with 107 HOD-RBC or 107 HOD-RBC presensitized with 0.5 μg polyclonal anti-HEL IgG or the same molar concentration (0.37 μg) of F(ab′)2 fragments of anti-HEL IgG or PBS 1 h before injection without washing away the unbound Ab. Mice were bled for serum on day 0, 7, 14, 21, and 28.
Detection of the HEL-specific Ab response
IgM and IgG subclasses specific for HEL were detected by ELISA as described previously (36). Briefly, ELISA plates were coated with 10 μg/ml HEL overnight at 4°C and blocked with 2% BSA (Sigma-Aldrich) for 2 h at room temperature. The plates were then washed, and serum samples from mice were added (diluted 1:100) for 1.5 h at 22°C. The plates were then washed and incubated with alkaline phosphatase–labeled anti-mouse IgM (μ) (catalog number M31508; Invitrogen, Camarillo, CA) or anti-mouse IgG, Fcγ fragment specific (Jackson ImmunoResearch Laboratories). Plates were developed using 1 mg/ml p-nitrophenyl phosphate (Sigma-Aldrich) and read by an ELISA reader at 405 nm after 15–30 min.
Statistical analysis
Data were expressed as the mean ± SEM and analyzed by the Kruskal–Wallis nonparametric test with Dunn’s multiple comparison test.
Results
AMIS mediated by HEL-specific polyclonal IgG in standard C57BL/6 mice
To evaluate AMIS, C57BL/6 mice were transfused with 107 HOD-RBC in the presence versus absence of 0.5 μg of a mouse polyclonal IgG specific for HEL. The development of IgM and IgG Abs specific for HEL sequences was evaluated at day 7, 14, 21, and 28. Consistent with our previous paper (36), the transfusion of mice with HOD-RBC alone induced both IgM and IgG Abs to the HEL protein (Fig. 1). In contrast, the transfusion of HOD-RBC in the presence of a polyclonal Ab specific for HEL significantly suppressed the development of IgM and IgG anti-HEL (p < 0.0001). The same level of AMIS was confirmed when Ab end-point titration of sera was evaluated instead of a single point dilution (Supplemental Fig. 1). Thus, polyclonal IgG Ab inhibits immunization to the HEL protein on HOD-RBC in C57BL/6 mice.
Induction of AMIS with anti-HEL Abs in C57BL/6 mice. Mice were transfused with 107 HOD RBC alone (HOD) or HOD-RBC in the presence of 0.5 μg IgG anti-HEL (AMIS) or left untreated (Nil). Mice were bled for serum at 1 h (day 0) (to assess the contribution of the prophylactic anti-HEL IgG in the ELISA) and on days 7, 14, 21, and 28. IgM (A) and IgG (B) Abs with specificity for HEL were evaluated by ELISA. Data represent the mean ± SEM of at least three different experiments. HOD n = 23 mice, AMIS n = 16 mice, and Nil n = 12 mice. Statistical analysis of the data from ELISA was done using the Kruskal–Wallis nonparametric test with Dunn’s multiple comparison test. The p values represent significant differences of HOD versus AMIS group (*p < 0.05, **p < 0.01, ****p < 0.0001).
Induction of AMIS with anti-HEL Abs in C57BL/6 mice. Mice were transfused with 107 HOD RBC alone (HOD) or HOD-RBC in the presence of 0.5 μg IgG anti-HEL (AMIS) or left untreated (Nil). Mice were bled for serum at 1 h (day 0) (to assess the contribution of the prophylactic anti-HEL IgG in the ELISA) and on days 7, 14, 21, and 28. IgM (A) and IgG (B) Abs with specificity for HEL were evaluated by ELISA. Data represent the mean ± SEM of at least three different experiments. HOD n = 23 mice, AMIS n = 16 mice, and Nil n = 12 mice. Statistical analysis of the data from ELISA was done using the Kruskal–Wallis nonparametric test with Dunn’s multiple comparison test. The p values represent significant differences of HOD versus AMIS group (*p < 0.05, **p < 0.01, ****p < 0.0001).
AMIS in FcRγ-chain–deficient mice
To evaluate the contribution of activating FcγRs to the induction of an AMIS effect, FcRγ-chain−/− mice were used as recipients of HOD-RBC, and the resulting immune response to the HEL Ag was evaluated. Mice lacking the FcRγ-chain, which is a necessary signaling constituent of FcγRI, FcγRIII, and FcγRIV, do not express any activating FcγR function (38). All activating mouse FcγRs bind to IgG2a and IgG2b, whereas FcγRIII additionally binds to IgG1 and FcγRI binds to IgG3 (38). The polyclonal anti-HEL used in this study was produced in C57BL/6 mice, which are known to express IgG2c instead of IgG2a (39). The binding of IgG2c to FcγRI has been previously demonstrated (40), but to the best of our knowledge, interactions with other FcγRs are not yet known.
Mice injected with HOD-RBC alone made a significant IgM and IgG response as expected (Fig. 2), whereas the IgM and IgG immune responses in the AMIS group was clearly inhibited. It was noteworthy that the IgG Ab response in FcRγ-chain−/− mice (Fig. 2B) was lower than in wild-type C57BL/6 mice (Fig. 1B), potentially contributing to the lack of statistical significance between AMIS and HOD groups at the IgG level in FcRγ-chain−/− mice. FcRγ-chain−/− mice have been previously reported to generate lower IgG Ab response compare with wild-type mice (41). Thus, polyclonal IgG Ab specific for the HOD Ag on erythrocytes inhibits immunization in the absence of activating FcγR function.
Induction of AMIS with anti-HEL Abs in FcRγ-chain−/− mice. Mice were untransfused (Nil), transfused with 107 HOD-RBC alone (HOD) or HOD-RBC in the presence of 0.5 μg IgG anti-HEL (AMIS). Mice were bled for serum at 1 h (to assess the contribution of the prophylactic anti-HEL IgG in the ELISA) and on days 7 and 14. IgM (A) and IgG (B) Abs with specificity for HEL were evaluated by ELISA. Data represent the mean ± SEM of at least three different experiments. HOD cells n = 16 mice, AMIS n = 9 mice, and Nil n = 8 mice. Statistical analysis of data from the ELISA was done using the Kruskal–Wallis nonparametric test with Dunn’s multiple comparison test (*p < 0.05, **p < 0.01, ***p < 0.001).
Induction of AMIS with anti-HEL Abs in FcRγ-chain−/− mice. Mice were untransfused (Nil), transfused with 107 HOD-RBC alone (HOD) or HOD-RBC in the presence of 0.5 μg IgG anti-HEL (AMIS). Mice were bled for serum at 1 h (to assess the contribution of the prophylactic anti-HEL IgG in the ELISA) and on days 7 and 14. IgM (A) and IgG (B) Abs with specificity for HEL were evaluated by ELISA. Data represent the mean ± SEM of at least three different experiments. HOD cells n = 16 mice, AMIS n = 9 mice, and Nil n = 8 mice. Statistical analysis of data from the ELISA was done using the Kruskal–Wallis nonparametric test with Dunn’s multiple comparison test (*p < 0.05, **p < 0.01, ***p < 0.001).
AMIS in FcγRIIB-deficient mice
FcγRIIB is the only classical FcγR expressed on B cells, and it has been shown to negatively regulate B cell activation (38, 42–44). FcγRIIB is a low-affinity receptor that binds mouse IgG1, IgG2a, and IgG2b (38). To the best of our knowledge, its affinity for IgG2c is unknown. To evaluate the role of FcγRIIB in AMIS, we first transfused FcγRIIB−/− mice on the B6129S background (versus control B6129S mice) with polyclonal IgG anti-HEL and HOD-RBC. Anti-HEL polyclonal Ab induced the same degree of the AMIS effect in mice lacking the the inhibitory FcγRIIB as compared with wild-type B6129S mice (Fig. 3).
Induction of AMIS with anti-HEL Abs in FcγRIIB-deficient mice on the B6129S background. Mice were untransfused (Nil) or transfused with 107 HOD-RBC alone (HOD) or HOD-RBC in the presence of 0.5 μg IgG anti-HEL (AMIS). Mice were bled for serum at 1 h (to assess the contribution of the prophylactic anti-HEL IgG in the ELISA) and on days 7 and 14. IgM (A and C) and IgG (B and D) Abs with specificity for HEL were evaluated by ELISA in B6129S FcγRIIB-deficient mice (C and D) versus control B6129S/F2 mice (A and B). Data represent the mean ± SEM of at least three different experiments. FcγRIIB−/− experiments: HOD n = 14 mice, AMIS n = 10 mice, and Nil n = 14 mice. B6129S/F2 experiments: HOD n = 9 mice, AMIS n = 9 mice, and Nil n = 10 mice. Statistical analysis of data from ELISA was done using the Kruskal–Wallis nonparametric test with Dunn’s multiple comparison test (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).
Induction of AMIS with anti-HEL Abs in FcγRIIB-deficient mice on the B6129S background. Mice were untransfused (Nil) or transfused with 107 HOD-RBC alone (HOD) or HOD-RBC in the presence of 0.5 μg IgG anti-HEL (AMIS). Mice were bled for serum at 1 h (to assess the contribution of the prophylactic anti-HEL IgG in the ELISA) and on days 7 and 14. IgM (A and C) and IgG (B and D) Abs with specificity for HEL were evaluated by ELISA in B6129S FcγRIIB-deficient mice (C and D) versus control B6129S/F2 mice (A and B). Data represent the mean ± SEM of at least three different experiments. FcγRIIB−/− experiments: HOD n = 14 mice, AMIS n = 10 mice, and Nil n = 14 mice. B6129S/F2 experiments: HOD n = 9 mice, AMIS n = 9 mice, and Nil n = 10 mice. Statistical analysis of data from ELISA was done using the Kruskal–Wallis nonparametric test with Dunn’s multiple comparison test (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).
We additionally evaluated the suppressive effect of polyclonal IgG anti-HEL in FcγRIIB−/− mice on the C57BL/6 background (Fig. 4) because the amelioration of ITP on FcγRIIB−/− mice on the mixed B6129S background has given different results as compared with this knockout on C57BL/6 background mice (45–47). The background of the FcγRIIB−/− mice appeared to have no effect on the AMIS mediated by polyclonal anti-HEL Abs. The response to the HOD-RBC was also efficiently suppressed in this strain of FcγRIIB−/− mice.
Induction of AMIS with anti-HEL Abs in FcγRIIB-deficient mice on the C57BL/6 background. Mice were untransfused (Nil) or transfused with 107 HOD-RBC alone (HOD) or HOD-RBC in the presence of 0.5 μg IgG anti-HEL (AMIS). Mice were bled for serum at 1 h (to assess the contribution of the prophylactic anti-HEL IgG in the ELISA) and on days 7 and 14. IgM (A) and IgG (B) Abs with specificity for HEL were evaluated by ELISA. Data represent the mean ± SEM of three different experiments. HOD n = 5 mice, AMIS n = 6 mice, and Nil n = 5 mice. Statistical analysis of data from ELISA was done using the Kruskal–Wallis nonparametric test with Dunn’s multiple comparison test (*p < 0.05).
Induction of AMIS with anti-HEL Abs in FcγRIIB-deficient mice on the C57BL/6 background. Mice were untransfused (Nil) or transfused with 107 HOD-RBC alone (HOD) or HOD-RBC in the presence of 0.5 μg IgG anti-HEL (AMIS). Mice were bled for serum at 1 h (to assess the contribution of the prophylactic anti-HEL IgG in the ELISA) and on days 7 and 14. IgM (A) and IgG (B) Abs with specificity for HEL were evaluated by ELISA. Data represent the mean ± SEM of three different experiments. HOD n = 5 mice, AMIS n = 6 mice, and Nil n = 5 mice. Statistical analysis of data from ELISA was done using the Kruskal–Wallis nonparametric test with Dunn’s multiple comparison test (*p < 0.05).
AMIS mediated by F(ab′)2 fragments of polyclonal IgG
Because FcRγ-chain deficient mice express FcγRIIB and FcγRIIB−/− mice reciprocally express activating FcγRs, we evaluated AMIS mediated by F(ab′)2 fragments of polyclonal anti-HEL as a condition where no FcγR can a priori be involved. The F(ab′)2 fragments produced retained their ability to bind HOD erythrocytes as efficiently as the intact Ab, as assessed by flow cytometry (Fig. 5A). Contaminating intact IgG was not detected in the F(ab′)2 preparation, and based on the sensitivity of the ELISA, contamination with more than 8 ng/ml intact IgG can be excluded (a dose that does not induce suppression). The F(ab′)2 fragments of polyclonal IgG retained their ability to suppress 73 and 61% of the IgM and IgG Ab response, respectively (Fig. 5B, 5C). Thus, the majority of Ab mediated inhibition can still occur in the absence of the Fc portion of the IgG molecule.
Induction of AMIS with F(ab′)2 fragments of polyclonal IgG anti-HEL in C57BL/6 mice. Mice were untransfused (Nil) or transfused with 107 HOD-RBC alone (HOD) or HOD-RBC in the presence of 0.37 μg F(ab′)2 fragments of IgG anti-HEL (AMIS F(ab′)2), or in the presence of 0.5 μg IgG anti-HEL (AMIS IgG). The reactivity of the F(ab′)2 fragments of IgG with the HOD cells was assessed by flow cytometry (solid line) in comparison with the intact Ab (dotted line) using a PE-conjugated anti-mouse Fab specific as secondary Ab (A). Unstained HOD RBC (solid black histogram) and HOD RBC incubated with secondary Ab only (solid gray histogram) were used as negative controls. Mice were bled for serum at 1 h (to assess the contribution of the prophylactic anti-HEL IgG in the ELISA) and on days 7 and 14. IgM (B) and IgG (C) Abs were evaluated by ELISA. Data represent the mean ± SEM of at least three different experiments. Nil n = 9 mice, HOD n = 17 mice, AMIS F(ab′)2 n = 9 mice, and AMIS IgG n = 9 mice. Statistical analysis of data from ELISA was done using the Kruskal–Wallis nonparametric test with Dunn’s multiple comparison test (*p < 0.05, ***p < 0.001, ****p < 0.0001).
Induction of AMIS with F(ab′)2 fragments of polyclonal IgG anti-HEL in C57BL/6 mice. Mice were untransfused (Nil) or transfused with 107 HOD-RBC alone (HOD) or HOD-RBC in the presence of 0.37 μg F(ab′)2 fragments of IgG anti-HEL (AMIS F(ab′)2), or in the presence of 0.5 μg IgG anti-HEL (AMIS IgG). The reactivity of the F(ab′)2 fragments of IgG with the HOD cells was assessed by flow cytometry (solid line) in comparison with the intact Ab (dotted line) using a PE-conjugated anti-mouse Fab specific as secondary Ab (A). Unstained HOD RBC (solid black histogram) and HOD RBC incubated with secondary Ab only (solid gray histogram) were used as negative controls. Mice were bled for serum at 1 h (to assess the contribution of the prophylactic anti-HEL IgG in the ELISA) and on days 7 and 14. IgM (B) and IgG (C) Abs were evaluated by ELISA. Data represent the mean ± SEM of at least three different experiments. Nil n = 9 mice, HOD n = 17 mice, AMIS F(ab′)2 n = 9 mice, and AMIS IgG n = 9 mice. Statistical analysis of data from ELISA was done using the Kruskal–Wallis nonparametric test with Dunn’s multiple comparison test (*p < 0.05, ***p < 0.001, ****p < 0.0001).
Discussion
The Fc dependence of anti–erythrocyte-mediated AMIS effects has been suggested from the finding that the inhibition can be nonepitope specific, where Abs to one Ag can inhibit the response to another Ag on the same erythrocyte (14, 17, 48–51), as well as from several studies demostrating that F(ab′)2 fragments could be poor suppressors (17, 48, 52, 53). A study by Karlsson and colleagues (16) using haptenated SRBC showed that anti-hapten mAbs suppress the Ab response in mice lacking the FcγRIIB as well as activating FcγRs or the neonatal Fc receptor (FcRn). More data about the role of FcγRs in the suppressive effect of IgG are required in models that mimic the noninflammatory immunization mediated by allogenic erythrocytes.
In the current study, to our knowledge, we have for the first time assessed the involvement of the activating and inhibitory FcγR for IgG in AMIS mechanisms using non-inflammatory (27) mouse erythrocytes expressing a unique, well-characterized immunological Ag (34). A polyclonal Ab against the HEL protein (analogous to the use of anti-D in humans) was used as the suppressing Ab and only the resulting Ab response specific for the HEL epitopes was measured as other portions of the HOD molecule do not stimulate an Ab response (54). Polyclonal IgG anti-HEL completely inhibited the Ab response to the HEL portion of the HOD erythrocytes (and both IgM and IgG responses were fully suppressed under AMIS conditions) in mice lacking functional activating FcγR activity or the presence of the inhibitory FcγRIIB. F(ab′)2 fragments of the polyclonal IgG, which cannot interact with both activating and inhibitory FcγRs, significantly reduced the Ab response to the immunogenic HEL portion of the molecule. These results support the previous findings using the SRBC model (16) and demonstrate that successful reduction of in vivo Ab responses can occur independent of FcγR function when a fully allogenic mouse model of erythrocyte immunization and polyclonal IgG are used.
Nonetheless, it is noteworthy that the F(ab′)2 fragments of polyclonal anti-HEL were less efficient than the whole Ab at inducing an AMIS effect, although significant suppression was still observed. These results could be interpreted in several ways: 1) F(ab′)2 fragments are more rapidly eliminated as compared with intact IgG and this may allow a partial immune response; 2) a small yet significant portion of AMIS effect is mediated by some combination of activating FcγR with FcγRIIB; 3) the Fc portion of the IgG could be mediating other non-FcγR mechanisms that contribute to inducing complete AMIS; 4) the AMIS effect is complete but that F(ab′)2 fragments induce immune enhancement, which counteracts the AMIS effect; and 5) AMIS is mediated by more than one mechanism where the Fc region of IgG plays a role in one of these mechanisms.
In the absence of a clear FcγR-mediated AMIS mechanism, the evidence in the mouse SRBC literature favors the steric hindrance or epitope masking model, which postulates that IgG could mask antigenic epitopes and thereby prevent B cells from recognizing and responding to the Ag (16, 23, 55). However, we have previously demonstrated using HOD erythrocytes that Abs directed to the Duffy portion of the HOD molecule suppress the Ab response to the HEL Ag without interference in the binding between the anti-Duffy and anti-HEL Abs (36). Significant suppression of the anti-HEL Ab response by anti-Duffy Abs can still occur in the absence of activating FcγR function (unpublished data). This would indicate that AMIS can occur through steric hindrance-independent mechanisms. There are also previous studies in humans demonstrating that anti-Kell Abs can suppress the immune response to the RhD Ag, which also argues against the steric hindrance hypothesis (14). Taken together, the AMIS experiments performed with the HOD-RBC model indicate that the major suppressive effect of Abs to allogenic erythrocytes can be independent of FcγRs or epitope masking, suggesting that other mechanisms might contribute to AMIS effects.
Another possible explanation for IgG-mediated inhibition involves FcγR-independent phagocytosis, where glycans on IgG molecules may play a crucial role. It has been previously shown that an agalactosyl form of IgG with exposed N-acetylglucosamine residues (called G0 IgG, when glycosylated IgG lacks the terminal galactose and sialic acid residues, exposing N-acetylglucosamine residues) on their Fc portion can be internalized by mannose receptors on macrophages (56). This could be a pathway by which Ag–Ab complexes can be internalized by macrophages independently of FcγR, especially considering that mouse IgG Fc has a high percentage of N-acetylglucosamine sugars (57). This hypothesis could also be consistent with the finding that deglycosylated monoclonal IgG was deficient in suppressing the immune response in vitro (58).
Polyclonal anti-HEL Abs can also induce some degree of Ag loss or Ag modulation, where the RBCs lose the incompatible Ag from their surfaces by trogocytosis and then circulate normally (28, 59). Although Ag loss has been reported involving Abs against several different blood group Ags (60–65), there is little understanding of this phenomenon or its contribution to IgG-mediated suppression. Studies in animal models have shown that it requires the binding of multiple Abs to the RBC, and individual monoclonal Abs did not induce Ag loss (59). Ag loss appears to require the function of FcγR (28), and this could be an additional reason why the F(ab′)2 only gave partial AMIS activity. Although we cannot exclude the possibility of having some degree of Ag loss when intact polyclonal anti-HEL is used to induce AMIS, we believe that this is unlikely to be the main mechanism involved because mAbs with specificity for the HOD molecule (which would not be expected to induce Ag loss) also induce an AMIS effect (36).
In summary, to our knowledge, these experiments are the first to evaluate the role of activating and inhibitory FcγRs in AMIS in a fully allogenic model of RBC immunization. We have demonstrated that significant suppression of the Ab response to the immunogenic HEL portion of the Ag on HOD RBCs mediated by polyclonal IgG can occur via an FcγR-independent pathway, although the Fc part of the IgG may be required for IgG to fully suppress the Ab response. Although the mechanism of AMIS remains to be fully resolved, these findings have implications for helping understand how anti-RBC Abs may be able to attenuate the immune response to erythrocytes.
Acknowledgements
We thank Andrew Crow, Joan Legarda, Danielle Marjoram, Melissa Menard, Dr. Xiaojie Yu, Dr. Wang Lin, and the St Michael’s Hospital Research Vivarium stuff.
Footnotes
This work was supported by Health Canada (Grant CBS221511) as part of the Canadian Blood Services/Canadian Institutes for Health Research Partnership Fund (to A.H.L.). L.B. and H.Y. were supported by postdoctoral scholarships from the Canadian Blood Services. The views expressed herein do not necessarily represent the view of the federal government of Canada.
The online version of this article contains supplemental material.
References
Disclosures
The authors have no financial conflicts of interest.