Inhibition of human FcγRIII prevents platelet destruction in a dual-humanized murine model of FNAIT

Key Points

• •FcγRIII is a key mediator of anti–HPA-1a–driven platelet phagocytosis in vitro, underscoring its role in platelet destruction in FNAIT.
• •Blockade of FcγRIII abrogates anti–HPA-1a–mediated platelet clearance in vivo, demonstrating its potential as a therapeutic target in FNAIT.

Visual Abstract

Abstract

Fetal and neonatal alloimmune thrombocytopenia (FNAIT) occurs when maternal anti–fetal platelet antibodies, most frequently against human platelet antigen 1a (HPA-1a), cause fetal platelet destruction. The precise pathogenic mechanisms remain incompletely understood. Here, we examined potential mechanisms underlying anti–HPA-1a–mediated platelet clearance, focusing on phagocytosis, complement fixation, platelet activation, and desialylation. Anti–HPA-1a antibodies in pooled plasma and in individual samples of 9 of 10 patients with FNAIT induced Fc gamma receptor (FcγR)–dependent platelet phagocytosis in vitro, indicating this as the major contributor to platelet destruction. In contrast, complement fixation, platelet activation (CD62P expression), and desialylation (Ricinus communis agglutinin I binding) were observed in only a minority of patients. To model anti–HPA-1a–mediated platelet clearance in vivo, we developed a dual-humanized mouse model combining HPA-1a–expressing platelets (from APLDQ transgenic mice [C57BL/6 mice expressing A₃₀P₃₂L₃₃D₃₉Q₄₇₀ on a murine GPIIIa backbone]) with mice engineered to express human FcγRs in place of murine FcγRs. Platelets from APLDQ mice sensitized with anti–HPA-1a sera underwent rapid clearance in these FcγR-humanized recipients. Pretreatment with an FcγRIII-blocking monovalent antibody significantly reduced clearance, confirming FcγRIII’s central role in mediating platelet destruction. These results demonstrate that FcγRIII-dependent phagocytosis is a dominant mechanism in anti–HPA-1a–driven platelet clearance in this model. Although complement activation and platelet desialylation were observed in a minority of samples, their contribution to platelet destruction remains speculative. Together, these findings highlight the complexity of FNAIT pathophysiology yet identify FcγRIII blockade as a promising strategy to prevent platelet clearance in this disease.

Introduction

Fetal and neonatal alloimmune thrombocytopenia (FNAIT) is a serious pregnancy disorder caused by maternal alloimmunization against paternally inherited fetal platelet antigens.¹ Maternal immunoglobulin G (IgG) antibodies cross the placenta, bind fetal platelets, and mediate their destruction, leading to fetal thrombocytopenia and potentially severe bleeding, including intracranial hemorrhage.2, 3, 4 The predominant antigen implicated in FNAIT in White populations is human platelet antigen 1a (HPA-1a), located on the β3 integrin subunit, which can pair with either αIIb or αv.⁵ Clinical heterogeneity in disease severity suggests that distinct immune effector mechanisms contribute to thrombocytopenia.⁶

The engagement of Fc gamma receptors (FcγRs) on immune effector cells plays a pivotal role in the recognition and clearance of opsonized platelets.⁷ These receptors are expressed as early as 18 weeks of pregnancy and are categorized as high-affinity (FcγRI) and low-affinity (FcγRIIa, FcγRIIb, FcγRIIc, and FcγRIII) subtypes.^8,9 The effectiveness of IV immunoglobulin in FNAIT postulates FcγR involvement, given one of IV immunoglobulin’s proposed mechanisms of action is competitive FcγR blockade or inhibition.^10,11 Supporting this hypothesis, previous in vitro studies have demonstrated that individually blocking FcγRI and FcγRIII in an FNAIT model influenced binding and platelet phagocytosis.^12,13 Whether blocking these receptors in an animal model of FNAIT ameliorates the platelet count is unknown.

Complement activation may be another potential mechanism of platelet clearance. Activation of the classical complement pathway can accelerate platelet opsonization by C1q-mediated downstream mechanisms leading to phagocytosis through complement receptors or cell lysis.⁶ In addition, a recent observational study reported a higher degree of C4d deposition in newly diagnosed FNAIT patients.¹⁴

Recent studies in immune thrombocytopenia (ITP) have shown that platelet autoantibodies can mediate Fc-independent clearance through desialylation-dependent recognition by Ashwell-Morell receptors on hepatocytes or Kupffer cells.^15,16 In FNAIT, limited evidence suggests that anti–HPA-1a antibodies may also induce platelet activation in a subset of patients.¹⁷ However, the relevance of platelet activation and desialylation in FNAIT remains poorly defined. These pathways may represent additional mechanisms of platelet clearance, particularly in patients refractory to IV immunoglobulin.

Given the limitations of current prophylactic and therapeutic strategies in FNAIT, there is a critical need to elucidate the molecular mechanisms that drive fetal platelet clearance and to support the development of more targeted interventions. In 2012, the European Union–funded PROFNAIT Consortium initiated efforts to create a hyperimmune anti–HPA-1a IgG product aimed at preventing maternal HPA-1a alloimmunization.¹⁸ This product, termed NAITgam or RLYB-211, was manufactured from plasma collected from HPA-1a–immunized women. Early proof-of-concept studies have demonstrated that NAITgam/RLYB-211 can rapidly neutralize or eliminate HPA-1a–positive platelets.¹⁹ Our laboratory received aliquots of this pooled plasma (hereafter referred to as anti–HPA-1a plasma) for mechanistic studies.

In this study, we examined FcγR-mediated platelet phagocytosis using anti–HPA-1a pooled plasma and individual patient samples, in addition to assessing evidence of complement activation, platelet activation, and desialylation. To extend these findings in vivo, we then evaluated platelet clearance in a humanized murine model.

Materials and methods

For more information, see the supplemental Data, Materials and methods.

Cell line and animals

THP-1-CD16A cells were maintained in a 37°C, 5% CO₂ environment in complete RPMI 1640 medium.²⁰ FcγR humanized (H-2^d) mice were generously provided by Jeffery Ravetch from The Rockefeller University.²¹ These mice are genetically deficient for murine FcγRs and express human FcγRI, FcγRIIA, FcγRIIB, FcγRIIIA, and FcγRIIIB in a manner that recapitulates human FcγR patterns.²¹ APLDQ mice were generously provided by Peter Newman from the Versiti Blood Center of Wisconsin and are C57BL/6 mice expressing murine glycoprotein IIIa (GPIIIa) harboring T₃₀→A, S₃₂→P, Q₃₃→L, N₃₉→D, and M₄₇₀→Q (APLDQ) mouse-to-human amino acid substitutions within the plexin-semaphorin-integrin and adjacent epidermal growth factor 1 domains.²²

Collection of FNAIT pooled plasma and patient sera

Pooled anti–HPA-1a plasma (NAITgam) from the PROFNAIT project was obtained, and sera from 10 mothers whose children had FNAIT were received. A very large library of patient samples was extensively screened to identify those with antibodies exclusively targeting HPA-1a and free of other alloantibodies, including anti-HLA and non–HPA-1a specificities. Only samples meeting these strict criteria were included, resulting in a necessarily limited sample size (n = 10). Control sera were obtained from 4 to 5 healthy adult donors. For the animal experiments, pooled anti–HPA-1a plasma (NAITgam) was converted to a serum preparation by addition of CaCl₂ to induce clotting factor depletion before use.

FcγR-blocking reagents

Antibodies 10.1 (anti-FcγRI), AT10 (anti-FcγRIIA/B/C), 3G8 (anti-FcγRIII), and the monovalent FcγRIII-blocking reagents 17C02-albumin and 17C02-IgG1_OA (a 1-armed human IgG1-based construct with a LALA-mutated Fc region) were used. The 17C02-albumin and 17C02-IgG1_OA constructs have previously been described.²³ Antibodies 10.1, AT10, and 3G8, and isotype controls used in the blocking studies were fully Fc-region deglycosylated as previously described.^23,24

In vitro platelet phagocytosis assays

Platelet phagocytosis assays were performed as previously described.^25,26 In the human platelet assay, platelets isolated from citrate-anticoagulated healthy donor blood were labeled with cytoplasmic dye 5-chloromethylfluorescein diacetate (CMFDA), sensitized with patient or control serum, and incubated with phorbol 12-myristate 13-acetate–differentiated THP-1-CD16A macrophages. When indicated, macrophages were pretreated with deglycosylated FcγR-blocking antibodies or isotype controls. Surface-bound nonphagocytosed platelets were identified using anti-GPIX–Alexa Fluor 647, and phagocytosis was quantified from confocal microscopy images as phagocytic index, defined as engulfed platelets per 100 macrophages.

In the murine assay, CMFDA-labeled APLDQ platelets isolated by cardiac puncture were sensitized with anti–HPA-1a or control (normal human serum) and incubated with peritoneal macrophages from FcγR-humanized mice. When indicated, macrophages were pretreated with FcγR-blocking antibodies or isotype controls. Surface-bound platelets were detected using anti-GPV followed by Alexa Fluor 647–conjugated secondary antibody, and phagocytosis was quantified by confocal microscopy as described earlier. Additional procedural details are provided in the supplemental Methods.

Protein G fractionation of pooled anti–HPA-1a (NAITgam)

Pooled anti–HPA-1a plasma (NAITgam) was fractionated into IgG-enriched and IgG-depleted fractions using protein G affinity purification, as previously described. Additional details are provided in the supplemental Methods.

Complement fixation assay

Complement fixation was assessed in all 10 heat-inactivated patient sera using genotype-specific platelets and guinea pig complement, as previously described.²⁷ Residual free complement was measured using an indicator sheep erythrocyte lysis assay. Additional details are provided in the supplemental Methods.

Monoclonal antibody–specific immobilization of platelet antigens assay

The indirect monoclonal antibody–specific immobilization of platelet antigens assay, as described by Kiefel et al, was used to detect HPA-1a IgG alloantibodies in maternal sera.²⁸ Monoclonal antibody AP3 (GPIIIa-specific) was used to immobilize GPIIb/IIIa. Alloantibodies to major histocompatibility complex class I were screened using a Luminex bead assay system.

Platelet activation and desialylation

In vitro platelet activation and desialylation assays were performed as previously described.^29,30 Gel-filtered platelets were left untreated or incubated with sera from patients with FNAIT or control reagents, including normal pooled plasma IgG and IV immunoglobulin. IV immunoglobulin was used as a control condition and not as a pretreatment before anti–HPA-1a stimulation. Platelets were stained with anti-CD62P, Ricinus communis agglutinin I (RCA-I), and anti-CD41, and analyzed using a Sony SP6800 spectral cytometer. CD41 was used exclusively as a platelet lineage/gating marker, and RCA-I binding was quantified within gated CD41⁺ platelet events.

Platelet labeling and in vivo clearance study

APLDQ platelets were collected, labeled with CMFDA, and sensitized with anti–HPA-1a or control serum at a 1:8 dilution. FcγR-humanized mice were treated with 53.1 μg of the monovalent FcγRIIIA-blocking 1-armed human IgG1 antibody 17C02-IgG1_OA or 79.7 μg of human IgG1 isotype control at equimolar dose, followed 2 hours later by IV injection of sensitized platelets. Circulating platelet levels were monitored for 6 hours by blood sampling from the saphenous vein at 1 minute and every 2 hours thereafter. The 1-minute sample was used as baseline to balance systemic distribution of transfused platelets with minimizing the impact of very early clearance.^31,32 Platelet counts were quantified using a Guava easyCyte flow cytometer.

Results

The phagocytic activity of anti–HPA-1a plasma can be mediated through IgG and FcγRs

The phagocytic potential of pooled anti–HPA-1a plasma was first assessed using human donor–derived platelets and THP-1-CD16A–derived macrophages.²⁰ Platelets were opsonized with either the benchmark reagent; mouse monoclonal antibody (clone A2A9/6); or a negative control, normal human serum (control), or anti–HPA-1a from the PROFNAIT project. Platelet internalization by macrophages was visualized and quantified using confocal microscopy. Nonphagocytosed platelets were distinguished by surface labeling with a red fluorescent anti-GPIX antibody (Figure 1A). The pooled anti–HPA-1a significantly enhanced platelet phagocytosis compared with controls (Figure 1B).

Figure 1.

THP-1-CD16A macrophages phagocytose anti–HPA-1a IgG-opsonized platelets predominantly through FcγRI and FcγRIII. (A) Phagocytosis of anti–HPA-1a–opsonized platelets by THP-1 CD16A cells. THP-1 CD16A cells were visualized by spinning-disc confocal microscopy under 63× objective with (i) differential interference contrast and laser fluorescence. Macrophage nuclei were distinguished with DAPI (4′,6-diamidino-2-phenylindole; blue) (ii). Platelets were labeled with the cytoplasmic dye CMFDA (green) (iii). Platelets that were not internalized were identified after phagocytosis using as Alexa Flour 647–conjugated anti-CD42a (GPIX) antibody (red) (iv). Merged channels with differential interference contrast images are shown in the bottom panel (v). Areas of red and green colocalization in the merge appear as yellow (ie, phagocytosed platelets). A total of 4 to 5 images were taken with 40 to 60 macrophages per image, and these images were Z-stacked to allow 3-dimensional reconstruction using Imaris version 10.0.0 to help visualize the internal vs external platelets (not shown). Platelets were defined by size (1.5-3.5 μm) to differentiate them from internalized microparticles or platelet aggregates. Arrows indicate example of phagocytosed platelets (in panels Aiii-v). Scale bar, 10 μm. (B) THP-1-CD16A macrophage phagocytosis of platelets from A⁺ and O⁺ donors sensitized with normal human plasma (control), with a benchmark reagent A2A9/6 (mouse IgG2a) or pooled anti–HPA-1a plasma (NAITgam) as indicated. Donor and plasma blood type was matched (n = 6). Statistical significance was calculated by the nonparametric Kruskal-Wallis test with the Dunn multiple comparisons test. ∗∗∗P < .001. (C) Phagocytosis mediated by pooled anti–HPA-1a plasma, the protein G–enriched IgG fraction from the HPA-1a plasma and the IgG-depleted fraction are shown (n = 4). Control (normal human serum) was used to sensitize donor platelets as the negative control. Significance: the nonparametric Kruskal-Wallis test with the Dunn multiple comparisons test. ∗P < .05. (D) FcγRI and FcγRIII on THP-1 CD16A macrophages mediate phagocytosis of anti–HPA-1a opsonized platelets. FcγRs were blocked using deglycosylated antibodies to FcγRI (10.1), FcγRIIA/B/C (AT10), or FcγRIII (3G8), vs an isotype control (mouse IgG1), final concentration of 10 μg/mL; 0.07μM each. Healthy donor platelets from A⁺ as well as O⁺ donors were opsonized with anti–HPA-1a pooled antibody (n = 5-6 experiments). Statistical significance was calculated by the nonparametric Kruskal-Wallis test with the Dunn multiple comparisons test. ∗∗∗P < .001; ∗∗∗∗P < .0001. Data error: mean ± 1 standard deviation (SD). (E) THP-1-CD16A macrophage phagocytosis of human platelets sensitized with normal human serum (control), with a benchmark reagent A2A9/6 (mouse IgG2a) or pooled anti–HPA-1a. Blocking FcγRIII was tested using 3G8- and 17C02-based molecules (17C02-albumin and 17C02-IgG1_OA) using the same comparative final molar concentration (0.07μM). Statistical significance was calculated by 1-way analysis of variance (ANOVA) with multiple comparisons against all means with the Šidák multiple comparisons test. ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. Data are presented as mean ± SD (n = 3-4). Phagocytic index: number of phagocytosed platelets per 100 counted macrophages. ns, nonsignificant.

To determine whether IgG was responsible for the observed phagocytic activity, the pooled anti–HPA-1a plasma was fractionated into IgG-purified and IgG-depleted fractions using protein G affinity chromatography. Successful separation was confirmed via sodium dodecyl sulfate–polyacrylamide gel electrophoresis and western blotting (supplemental Figure 1). Phagocytosis assays revealed that the IgG-depleted fraction exhibited a near-complete loss of activity, supporting the essential role of IgG in phagocytosis (Figure 1C). However, the IgG-purified fraction alone demonstrated a tendency toward reduced phagocytic activity relative to unfractionated plasma. A possible explanation for this observation could be the effects of the low pH during the IgG purification process or potentially implicating non-IgG components in phagocytic function,³³ although it did not reach statistical significance. Consistent with this, heat inactivation of the pooled anti–HPA-1a plasma also led to a reduction in phagocytic activity, implicating the potential involvement of heat-labile plasma factors (supplemental Figure 2).

To evaluate the requirement for specific FcγRs involved in anti–HPA-1a–mediated phagocytosis, receptor-blocking antibodies targeting FcγRI, FcγRIIA/B/C, and FcγRIII were used. Platelets opsonized with pooled anti–HPA-1a plasma or control plasma were incubated with macrophages pretreated with individual FcγR blockers. Blockade of FcγRI and FcγRIII significantly reduced platelet uptake (Figure 1D), whereas FcγRII blockade had only a marginal effect, consistent with previous findings in ITP models.²⁵

As blockade of FcγRIII provided the most efficacious increase in platelet counts, FcγRIII-blockade was further explored comparing 3 previously described FcγRIII-blocking biologics: deglycosylated antibody 3G8, monovalent 17C02-albumin, and monovalent 17C02-IgG1_OA.^20,23 Platelets were sensitized with either anti–HPA-1a, A2A9/6, or control (normal human serum). All 3 FcγRIII-blocking biologics significantly reduced phagocytosis compared with the untreated control (Figure 1E).

Platelet phagocytosis predominates in samples from patients with FNAIT with complement fixation in a subset

Ten sera positive for anti–HPA-1a antibodies were selected from individuals who had pregnancies affected by FNAIT and tested negative for anti-HLA antibodies and other HPA antibodies. Available patient information is provided in supplemental Table 1. To assess the capacity of FNAIT sera to mediate platelet clearance, we evaluated their ability to induce phagocytosis of donor platelets by THP-1-CD16A–derived macrophages. Phagocytosis was performed blinded to patient clinical information and all characterization of sera. Sera are displayed in ascending order of phagocytic index (Figure 2A). Samples were considered positive for phagocytosis if the mean phagocytic index exceeded 2 standard deviations above the mean of the control (normal human serum). Each healthy platelet donor was color-coded and grouped by blood type to assess any potential influence on phagocytic response.

Figure 2.

Anti-HPA-1a antibodies from sera from patients with FNAITtrigger platelet phagocytosis in vitro. (A) The ability of FNAIT sera to trigger THP-1-CD16A macrophage phagocytosis of platelets was evaluated (n = 5-6). Normal human serum (control) from healthy donors was used to opsonize platelets as a negative control. Platelets from 4 healthy donors were included and represented by blue, green, orange, and pink data points. These donors had 3 different blood types, as indicated by the symbols: ■, ▲, and ▼. The cutoff to determine if phagocytosis was “positive” for triggering phagocytosis was if the phagocytic index value of the patient sera mean was higher than 2× the SD of the mean of the control (normal human serum). Phagocytic index: number of phagocytosed platelets per 100 counted macrophages. Error is presented as mean ± 1 SD. FNAIT sera are displayed in ascending order based on phagocytic index. Thrombocytopenic severity is indicated by the color of the bar based on the fetal platelet count. The set of “+” or “–“ below the x-axis display each sera’s ability to fix complement. The second row below the x-axis indicates the sera’s antibody titer. (B) Healthy platelets from third-party donors were opsonized with sera from 3 different patients (patients B, C, and G) all with severe thrombocytopenia (n = 4-7 experiments depending on sera availability). Control (normal human serum) was used with platelets as negative controls. ABO blood group information for the maternal sera was not available; therefore, donor-serum ABO compatibility was not assessed in this experiment. FcγRs were blocked using deglycosylated antibodies to FcγRI (clone 10.1) or FcγRIII (3G8), and isotype control (mouse IgG1). Significance: 1-way ANOVA with multiple comparisons against all means with a Tukey post hoc test. ∗∗∗∗P < .0001. Data error: mean ± SD.

Of the 10 sera from patients with FNAIT tested, 9 (90%) demonstrated a significant capacity to induce platelet phagocytosis (Figure 2A). These 9 patients had either mild, moderate, or severe neonatal thrombocytopenia, with only 1 (patient J) failing to elicit a phagocytic response. Complement fixation was assessed in all 10 sera from patients with anti–HPA-1a, of which 2 samples (patients B and J), demonstrated detectable complement-fixing activity. Notably, no correlation was observed between either phagocytic index and platelet count or anti–HPA-1a antibody titer (supplemental Figure 3A-B). A statistically significant correlation was found between antibody titer and platelet count (supplemental Figure 3C). Although patient J presented with severe thrombocytopenia, this serum did not meet the threshold for phagocytosis induction. Notably, patient J sera was 1 of the 2 that tested positive for complement fixation.

To further delineate the contribution of FcγRs in mediating phagocytosis, we evaluated the effect of receptor-specific blockade using sera from 3 patients with severe thrombocytopenia. Blocking using antibodies targeting FcγRI and FcγRIII significantly reduced platelet uptake in all individual patients (Figure 2B).

Anti–HPA-1a IgG can potentially induce platelet activation (CD62P expression) and desialylation (RCA-I binding)

To investigate whether additional mechanisms of potential platelet clearance should be considered in the evaluation of patients with FNAIT, we assessed the capability of anti–HPA-1a antibodies to induce platelet activation and desialylation. Platelet activation was evaluated by measuring surface expression of P-selectin (CD62P), a well-established marker of α-granule release.^29,30 Desialylation was assessed using RCA-I binding, which detects exposed β-galactose residues after sialic acid loss.^34,35

Anti-HPA-1a induced significantly greater CD62P expression (Figure 3A) and RCA-I binding (Figure 3B) than pooled normal human plasma, with effects becoming significant at 1:100. This dilution was therefore selected for subsequent analyses. At this fixed concentration, pooled anti–HPA-1a plasma induced clear platelet activation and desialylation (Figure 3C-D), providing a reference condition for comparison with individual patient sera. Purified anti–HPA-1a IgG also induced greater activation and desialylation than control IgG (Figure 3E-F).

Figure 3.

Anti-HPA-1a IgG can induce platelet activation and desialylation in vitro. (A) Surface P-selectin expression and (B) RCA-1 binding to gel-filtered human platelets either treated with normal pooled plasma or pooled anti–HPA-1a plasma and assessed by flow cytometry. Pooled anti–HPA-1a plasma induces surface P-selectin (CD62P) expression (C) and RCA-1 binding to gel-filtered human platelets (D) either untreated (resting) or treated with normal human plasma (control; 1:100 dilution), and pooled anti–HPA-1a plasma measured by flow cytometry. Surface P-selectin expression (E) and RCA-I binding (F) to gel-filtered human platelets treated separately with protein G–purified anti–HPA-1a IgG, normal pooled plasma IgG, or IVIG control. (G) Surface P-selectin expression and (H) RCA-1 binding to gel-filtered human platelets treated with normal human serum (control), and patient sera (1:100 dilution) measured by flow cytometry. Data were analyzed using FlowJo version 10. MFI fold change was calculated relative to healthy control sample within each assay. Data are presented as mean ± SD from 4 to 5 independent experiments, depending on sample availability. Within each experiment, samples were measured in duplicate and duplicate values were averaged. Statistical significance (panels A-D) was calculated by the nonparametric Kruskal-Wallis test with the Dunn multiple-comparison test (comparing to resting platelet values). Statistical analysis (panels E-H) was performed using a 2-way ANOVA and the Šidák multiple-comparison test, comparing the MFI values of the anti–HPA-1a samples with that of the control or IgG. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. MFI, mean fluorescence intensity.

Among 8 patient sera tested, 2 (patients C and I) induced both significant platelet activation and desialylation, as demonstrated by elevated CD62P expression (Figure 3G) and RCA-I binding (Figure 3H). Patient F was excluded from this analysis due to the absence of observable platelets in the “platelet gate” (not shown), potentially indicating clumping after activation.

Blocking FcγRIII ameliorates anti–HPA-1a–mediated clearance of APLDQ platelets in vivo

To translate our in vitro findings on FcγR-dependent phagocytosis into an in vivo model, we used a previously established transgenic mouse model expressing the human HPA-1a epitope on a murine GPIIIa backbone (APLDQ mice; Figure 4A).²² Before running in vivo experiments, we confirmed that APLDQ platelets were phagocytosed in vitro by peritoneal macrophages derived from FcγR-humanized mice. For the in vivo studies, pooled anti–HPA-1a plasma was converted to a serum preparation by addition of CaCl₂ before platelet opsonization.³⁶ APLDQ platelets were opsonized with either pooled anti–HPA-1a patient sera, a benchmark reagent (clone A2A9/6), or normal human serum (control). As expected, anti–HPA-1a sera significantly increased platelet phagocytosis relative to controls (Figure 4B). Pretreatment of macrophages with blocking reagents against FcγRI (10.1) or FcγRIII (17C02-albumin) reduced platelet uptake, with FcγRIII blockade reaching statistical significance and FcγRI showing a clear trend toward inhibition (Figure 4B).

Figure 4.

FcγRIII blockade in vivo ameliorates anti–HPA-1a–mediated clearance of platelets. (A) Platelets collected from APLDQ mice (ie, HPA-1a–positive platelets) were fluorescently labeled and sensitized with NAITgam-derived anti–HPA-1a serum or normal human serum (control). These platelets were transfused into FcγR-humanized mice that have been either treated with FcγRIII blocking antibody (17C02-IgG1_OA) or a human IgG1 isotype control. Platelet clearance was tracked in vivo over the span of 6 hours. (B) Peritoneal macrophages derived from FcγR-humanized mice were used to assess phagocytosis of APLDQ platelets either untreated (nonopsonized) or sensitized with either normal human serum (control), a benchmark reagent A2A9/6 (mouse IgG2a) or NAITgam-derived anti–HPA-1a serum (n = 6). FcγRs were blocked using an antibody to FcγRI (10.1) or FcγRIII (17C02-albumin). Statistical significance was calculated by the nonparametric Kruskal-Wallis test with the Dunn multiple-comparison test. ∗P < .05. Data are presented as mean ± SD. (C) Dilution curve of NAITgam-derived anti–HPA-1a sera binding to APLDQ platelets. Stained and sensitized platelets were washed and analyzed by flow cytometry using the BD LSR Fortessa X-20. Data analysis was performed using FlowJo version 10. Data are presented as mean ± SD (n = 6). The solid and dashed lines represent the MFI (arbitrary units) value for the corresponding nonstaining group and secondary antibody alone at 5 μg/mL, respectively. (D) APLDQ platelets labeled with green CMFDA and opsonized with NAITgam-derived anti–HPA-1a serum or normal human serum (control serum). Mice were pretreated with the FcγRIII-blocking 17C02-IgG1_OA or an IgG1 isotype control at equimolar concentration. Statistical significance was calculated by 2-way ANOVA with multiple comparisons against all means with the Tukey multiple-comparison test. ∗P < .05; ∗∗P < .01. Data are presented as mean ± SEM (n = 5 mice). MFI, mean fluorescence intensity; NHS, normal human serum. Panel A created with biorender.com. Tawhidi Z. (2026) https://BioRender.com/u67rlwd.

To determine the optimal concentration of the NAITgam (anti–HPA-1a) sera for the in vivo tracking studies, we conducted a dilution series to assess binding to APLDQ platelets. Based on the resulting binding profile, a 1:8 dilution was selected (Figure 4C). APLDQ platelets were fluorescently labeled and opsonized with either anti–HPA-1a serum or control (normal human serum) at the 1:8 dilution. These sensitized-labeled platelets were transfused into FcγR-humanized mice and circulating platelet levels were monitored every 2 hours over a 6-hour period. Platelet levels at 1 minute after transfusion served as the baseline, a time point selected to allow systemic distribution of transfused platelets with the need to minimize the influence of very early clearance on baseline measurements. Mice were pretreated with either an isotype control (human IgG1) or 17C02-IgG1_OA to block FcγRIII. Starting at 2 hours after transfusion, mice treated with the 17C02-IgG1_OA monovalent FcγRIII blocker showed significantly higher levels of circulating platelets than isotype-treated controls receiving anti–HPA-1a serum (Figure 4D).

Discussion

FNAIT is a leading cause of severe thrombocytopenia in fetuses and neonates. Thrombocytopenia occurs in ∼1% to 4% of all newborns, but because routine blood counts using umbilical cord blood are not established and due to the lack of clinical symptoms in most neonates with thrombocytopenia, it frequently goes undiagnosed.³⁷ FNAIT is often considered the platelet counterpart to hemolytic disease of the fetus and newborn. However, key differences exist between the 2 conditions, particularly in the target antigen and the potential for maternal alloimmunization to affect the first pregnancy.^38,39 These distinctions have contributed to the absence of an approved prophylaxis for FNAIT, and the recent clinical development program for anti–HPA-1a monoclonal antibodies has been discontinued.⁴⁰

In this study, we provide novel insight into the immune mechanisms that mediate fetal platelet destruction in FNAIT, focusing on the effector functions of anti–HPA-1a IgG antibodies. Our data demonstrate that phagocytosis is the predominant functional readout across patient samples, whereas complement fixation, platelet activation, and desialylation are observed only in a subset. In addition, antibody titer is correlated with neonatal platelet nadir but not with phagocytic activity, suggesting that antibody quantity alone does not fully explain pathogenicity and that qualitative features of the antibody response may modulate effector function.

We confirmed that most of the anti–HPA-1a–positive samples robustly induced platelet phagocytosis in vitro via FcγRs, predominantly through FcγRI and FcγRIII. Our results also align with the findings from a recent study by Ames et al, demonstrating the involvement of these 2 receptors.¹³ Patient samples in our study were stringently selected based on the presence of only anti–HPA-1a antibodies, and an absence of HLA antibodies. FcγR expression in the THP-1-CD16A cell line used in this study has been previously characterized.²⁰ Briefly, FcγRIIIA (CD16A) expression is stably maintained after transduction, whereas endogenous expression levels of other FcγRs decreased slightly compared with those observed in wild-type THP-1 cells. Thus, FcγRIIIA transduction does not substantially alter the broader FcγR expression profile, supporting the interpretation of relative FcγR dependence observed in our model.

IgG-depleted fractions of plasma did not retain the ability to trigger macrophage phagocytosis of platelets, supporting a central role for IgG antibodies in this process. Consistent with previous reports, maternal anti–HPA-1a antibody titer correlated with neonatal platelet nadir in our cohort.⁴¹ In contrast, antibody titer did not correlate with phagocytic activity, suggesting that antibody quantity alone is insufficient to explain effector function. Other alloantibody characteristics, including epitope specificity and Fc glycosylation, may therefore contribute importantly to pathogenicity.^42,43 Phagocytosis was also reduced after IgG purification and heat inactivation of plasma, suggesting that additional plasma factors may modulate this response, although their relevance in vivo remains uncertain.

Given previous evidence implicating classical complement activation in FNAIT, we assessed whether anti–HPA-1a patient sera could fix complement on platelets. We observed that 2 of 10 (∼20%) patient samples demonstrated the ability to fix complement. This test assesses the antibody’s ability to initiate the classical complement pathway upon binding to its target. This finding is particularly interesting given previous reports of elevated C4d deposition in newly diagnosed FNAIT patients, as compared with the IV immunoglobulin–treated FNAIT patients or negative controls.¹⁴ Although further investigation of complement in vivo would be valuable, this could not be adequately performed in our model because it does not recapitulate pregnancy-induced alloimmunization. In ITP, the primary complement-fixing autoantibodies target GPIIb/III, the same complex that carries the HPA-1a antigen.²⁷ Preclinical and clinical evidence supports the involvement of classical pathway activation in ITP, with disease severity associated with complement deposition on platelets and reduced C4 levels.⁴⁴ Recent studies of humanized antibodies targeting C1s to selectively block classical pathway activation have also shown promising results.⁴⁵

From this subset of patients, 2 of 8 (∼25%) sera also induced platelet activation, as indicated by increased CD62P expression, and desialylation, as measured by RCA-I binding. Both these processes have been implicated in Fc-independent pathways of platelet clearance in ITP.^46,47 Increased desialylation of platelet surface glycans has been demonstrated in accelerated hepatic platelet uptake via the Ashwell-Morell receptor, bypassing the FcγR-mediated clearance in the spleen. In a pregnancy environment, these pathways may be influenced by developmental differences in fetal immune and hepatic functions. For instance, the fetal liver is an active site of hematopoiesis, which may alter the level of platelet clearance.⁴⁸ However, currently the reticuloendothelial system in fetuses remains an area largely unexplored. Interestingly, in all patients for whom a DNA sample was available, at least 1 copy of a high affinity FcγRIIa or FcγRIIIa polymorphism was observed (supplemental Table 1). Given that FcγRIIa is expressed on resting platelets, it is possible that this receptor could be involved in platelet activation in these patients; however, in contrast, the only patient with 2 copies of this high affinity receptor polymorphism (patient J) did not show evidence of involvement in this pathway.

To translate our in vitro findings on FcγR-dependent phagocytosis into an in vivo model, we used a combination of 2 transgenic mouse strains: 1 expressing all human FcγRs (FcγR-humanized), and the other expressing a humanized HPA-1a antigen on its platelets (APLDQ). The APLDQ model has previously been successfully validated for investigating prophylactic treatment strategies for FNAIT.²² Our approach involved transfusing APLDQ platelets, sensitized with anti–HPA-1a serum, into FcγR-humanized mice to track platelet clearance.

Before initiating in vivo experiments, we evaluated three FcγRIII-blocking antibodies: 3G8, 17C02-albumin, and 17C02-IgG1_OA, in an in vitro phagocytosis assay. The 3G8 antibody was originally developed for clinical use in ITP but discontinued due to severe adverse events.⁴⁹ As a safer alternative, a monovalent single-chain variable fragment, 17C02, was engineered and expressed as either an albumin fusion protein (17C02-albumin) or a 1-armed human IgG1 antibody (17C02-IgG1_OA).²³ All 3 candidates significantly inhibited the phagocytosis of sensitized APLDQ platelets. Although it could be argued that the success of the 17C02-albumin molecule in a model of ITP would bode well for its use in the current model, the ability of an albumin-based molecule to enter the fetal circulation is not certain. For subsequent in vivo experiments, we therefore selected 17C02-IgG1_OA due to its inherent theoretical ability to engage the neonatal Fc receptor (FcRn) via its Fc domain, better mimicking maternal antibody transfer across the placenta in FNAIT.

Our in vivo findings revealed that APLDQ platelets sensitized with the NAITgam-based anti–HPA-1a serum underwent significantly faster clearance than those sensitized with control (normal human serum). Pretreatment of FcγR-humanized mice with 17C02-IgG1_OA markedly reduced this clearance, whereas no protective effect was observed with an IgG1 isotype control. Because NAITgam is derived from many donors, it likely contains antibodies capable of engaging multiple effector pathways, including FcγR-mediated phagocytosis, complement activation, and Fc-independent mechanisms such as platelet activation and desialylation. Despite this potential mechanistic diversity, FcγRIII blockade substantially prevented platelet clearance in vivo. This suggests that FcγRIII engagement is either the predominant mechanism driving anti–HPA-1a–mediated platelet destruction, or that other pathways may contribute but FcγRIII signaling remains a critical checkpoint for efficient platelet removal in this model. Although FcγRI also reduced phagocytosis in vitro, there are strong biological reasons to emphasize FcγRIII in FNAIT. FcγRIII is prominently expressed on macrophages, which are key effector cells in antibody-mediated platelet clearance, and its interaction with IgG is strongly enhanced by reduced Fc fucosylation.⁵⁰ This is particularly relevant because anti–HPA-1a antibodies with reduced Fc fucosylation have been associated with the most severe cases of FNAIT.⁴³ Thus, although FcγRI may also contribute to platelet destruction, our findings support FcγRIII as a compelling receptor to therapeutically target.

This study is, to our knowledge, the first to demonstrate the role of FcγRIII in mediating anti–HPA-1a–driven platelet clearance in an in vivo model. Additionally, we report, to our knowledge, for the first time, that platelet desialylation and activation may serve as novel biomarkers of disease activity in FNAIT, suggesting the involvement of Fc-independent mechanisms in a subset of patients.

There are several important limitations in this study. Our patient cohort was intentionally selected to exclude potential confounders, including anti–HLA or other HPA antibodies, to isolate the effects of anti–HPA-1a. Although this approach strengthened the mechanistic clarity, it limits the sample size and generalizability of our findings. Moreover, we lacked access to history of clinical treatment, which restricts our ability to further extrapolate our findings. In the in vitro assays, we used third-party donor platelets rather than patient-derived platelets, which does not fully recapitulate the in vivo interactions. In addition, although the transgenic mouse model used provides important insights into FcγR-mediated clearance, it does not mimic the natural process of maternal alloimmunization or the placental transfer of antibodies. Also, although the in vivo clearance data support a central role for FcγRIII-dependent effector mechanisms, the anatomical site and precise monocyte/macrophage subset responsible for anti–HPA-1a–mediated platelet clearance in this model were not determined. Additionally, the blockade of FcγRI and FcγRII in an in vivo model could be of interest and is a limitation of this study. Finally, we also did not assess platelet apoptosis, which could provide additional mechanistic insight into anti–HPA-1a–mediated platelet clearance and should be explored in future studies.

In conclusion, this study demonstrates that FcγR-mediated phagocytosis, particularly involving FcγRIII, is a central mechanism required for driving platelet clearance in this FNAIT model. Complement activation, platelet activation, and platelet desialylation, may contribute in some patients, but FcγRIII engagement appears to be a critical checkpoint for platelet destruction. These findings advance the mechanistic understanding of FNAIT and support FcγRIII as a promising therapeutic target.

Conflict-of-interest disclosure: A.H.L. and L.G.G. are inventors on pending patent application(s) related to the 17C02-based Fc-gamma receptor III–blocking molecules described in this study. A.H.L. has also received research funding from Argenx for work unrelated to this study. The remaining authors declare no competing financial interests.