Mechanisms of anti-GPIbα antibody-induced thrombocytopenia in mice

Abstract

Immune thrombocytopenia (ITP) is an acquired bleeding disorder characterized by antibody-mediated platelet destruction. Different mechanisms have been suggested to explain accelerated platelet clearance and impaired thrombopoiesis, but the pathophysiology of ITP has yet to be fully delineated. In this study, we tested 2 mouse models of immune-mediated thrombocytopenia using the rat anti-mouse GPIbα monoclonal antibody 5A7, generated in our laboratory. After a single IV administration of high-dose (2 mg/kg) 5A7, opsonized platelets were rapidly cleared from the circulation into the spleen and liver; this was associated with rapid upregulation of thrombopoietin (TPO) messenger RNA. In contrast, subcutaneous administration of low-dose 5A7 (0.08-0.16 mg/kg) every 3 days gradually lowered the platelet count; in this case, opsonized platelets were observed only in the spleen, and TPO levels remained unaltered. Interestingly, in both models, the 5A7 antibody was found on the surface of, as well as internalized to, bone marrow megakaryocytes. Consequently, platelets generated in the chronic phase of repeated subcutaneous 5A7 administration model showed reduced GPIbα membrane expression on their surface. Our findings indicate that evaluation of platelet surface GPIbα relative to platelet size may be a useful marker to support the diagnosis of anti-GPIbα antibody-induced ITP.

Conflict of interest statement

Conflict-of-interest disclosure: Z.M.R. is founder, president, and CEO of MERU-VasImmune Inc, which may develop commercial products based on methodologies presented in this article. S.K. and T.K. have equity interest in MERU-VasImmune Inc. The remaining authors declare no competing financial interests.

© 2020 by The American Society of Hematology.

Figures

Graphical abstract

Figure 1.

Generation and analysis of an acute platelet-depletion model induced by single high-dose injection of an anti-GPIbα antibody. (A) An anti-GPIbα antibody (5A7 or R300, 2 mg/kg) was administered to C57BL/6J wild-type (WT) mice by retro-orbital injection, and platelet counts were monitored. (B-C) Platelet distribution was visualized by wide-field immunofluorescence analysis of cryosections prepared from spleen (B) and liver (C) harvested from untreated mice or from treated mice 30 minutes or 24 hours after antibody administration (BZ-X700 microscope). Bars represent 100 µm. Tissues were stained with DAPI (blue), anti-rat IgG (green; detecting administered 5A7 antibody), and anti-αIIb antibody (purple). (D) Immunofluorescence staining of cryosectioned liver. Opsonized platelets were stained with anti-rat IgG (green) and macrophages were stained with anti-F4/80 antibody (red). Hepatocytes were stained with anti-ASGPR1 antibody (red; bottom). Images were obtained by confocal microscopy (LSM 880 microscope). Bars represent 20 µm. (E) IVIG was administered via intraperitoneal injection (2 g/kg) 24 hours before 5A7 IV injection (0.6 mg/kg). Spleens harvested 20 minutes after 5A7 administration were cryosectioned and immunostained with anti-rat IgG (green; detecting administered 5A7 antibody) and anti-F4/80 antibody (red), along with costaining with DAPI (blue). (F) Images were binarized, and colocalization of opsonized platelets (stained by anti-rat IgG) with F4/80 was calculated as a ratio by quantifying anti-rat IgG and F4/80 costained area divided by total of anti-rat IgG–stained area, by using Fiji image-analysis software. ***P < .001 by unpaired Student t test.

Figure 2.

Upregulation of TPO in the acute platelet-depletion model. (A) Liver TPO mRNA was quantified by real-time quantitative PCR before and at the indicated time points after IV 5A7 injection. (B-C) Plasma TPO was measured by enzyme-linked immunosorbent assay before (B) and after (C) antibody injection, as indicated. ***P < .001 by 1-way analysis of variance, with Dunn’s multiple-comparison test.

Figure 3.

Analysis of BM MKs in the acute platelet-depletion model. (A-C) BM sections harvested at the indicated time points after IV 5A7 injection were stained with DAPI (blue), anti-rat IgG (green; detecting administered 5A7 antibody), and anti-αIIb antibody (purple) and visualized by BZ-X700 microscope (bars represent 100 µm) (A) or confocal microscopy (LSM 880 microscope; bars represent 10 µm) (B). (C) Images were binarized, and colocalization of 5A7 with integrin αIIb was calculated as a percentage by quantifying the anti-rat IgG–stained area and dividing by the anti-integrin αIIb antibody–stained area, by using Fiji image-analysis software. ***P < .001 by one-way analysis of variance with Dunn’s multiple-comparison test. (D) BM cells were harvested 24 hours after IV 5A7 injection and analyzed by FACS for GPIbα expression by staining with monoclonal anti-GPIbα antibody (Xia.G5) which does not compete with 5A7 for binding to GPIbα. *P < .05, ***P < .001 by 2-way analysis of variance with Sidak’s multiple-comparison test.

Figure 4.

Thrombocytopenia, induced by repeated subcutaneous administration of anti-GPIbα antibody 5A7. (A) The antibody 5A7 (0.08 mg/kg) was injected subcutaneously (s.c.) every 3 days. Blood platelet counts were measured before and at the indicated time points after antibody injection. (B) Plasma TPO level was measured by enzyme-linked immunosorbent assay. (C) Immunostaining analysis of organs harvested on day 8 performed as described for the corresponding images in Figure 1. Bars represent 100 µm. (D) Immunostained BM sections were visualized by fluorescence microscopy (top) and confocal microscopy (bottom). Bars represent 100 μm (upper panels) and 10 μm (lower panels). (E) Plasma samples collected from mice treated with single dose (2 mg/kg) of IV 5A7 injection (left), and 3 doses of subcutaneous injection (0.08 mg/kg each) were analyzed by enzyme-linked immunosorbent assay to measure plasma concentration of administered 5A7.

Figure 5.

Analysis of platelets in thrombocytopenia mice induced by subcutaneous 5A7 injection. (A) Schematic diagram of platelet staining for FACS analysis. In this experiment, AlexaFluor 488–labeled 5A7 was administered in vivo so it could be detected regardless of subcellular localization. The platelet population was gated using a Brilliant Violet (BV) 421–labeled anti-αIIb antibody. PE-Cy7 labeled anti-rat IgG was used to detect 5A7 bound to GPIbα on the platelet surface. DyLight649 anti-GPIbα antibody (Xia.G5) was used to stain GPIbα on the platelet surfaces. (B) MFI of platelet GPIbα and αIIb corrected for platelet size (FSC) and expressed relative to normal control. Total results of 2 independent experiments are shown (n = 10 in each group); ***P < .001 determined by Student t test. NS, not significant. (C) Platelet subpopulations identified by presence/absence of in vivo–administered AlexaFluor 488–labeled 5A7 and their subcellular location. Positive signals on the abscissa indicate the presence of 5A7 either inside or on the surface of platelets. Positive signals on the ordinate indicate the presence of 5A7 on the platelet surface. The criteria of the quadrant were determined by a control blood sample collected from an untreated mouse stained with PE-Cy7 labeled anti-rat IgG (left). Mean percentage and standard deviations were calculated for the 10 mice tested. (D) GPIbα expression on platelet surface was determined by staining with DyLight 649 anti-GPIbα antibody (Xia.G5) which does not compete with 5A7 for GPIbα binding. The MFI of each population is indicated as relative to normal controls. MFI of GPIbα corrected by platelet size (FSC) is shown in the bottom row of the table. Results are obtained from 10 mice. (E) Confocal analysis of platelets in Q2 (top) and Q3/Q4 (bottom) subpopulations. Bar represents 5 µm. (F) During the course of chronic thrombocytopenia induced by 5A7 subcutaneous administration every 3 days, the percentage of platelets negative for surface GPIbα expression was monitored by FACS.

Figure 6.

Schematic presentation showing different mechanisms of platelet clearance in 2 mouse models of antibody-induced thrombocytopenia. (A) In the acute platelet-depletion model (left), high-dose antibody causes platelet activation, aggregation, and desialylation. Aggregates and desialylated platelets are stuck in the microvasculature of the liver by binding to AMR. Opsonized platelets are also cleared by splenic macrophages via Fc receptor. Excess antibody is accumulated in MK in BM, which induces internalization of GPIbα. In the chronic thrombocytopenia model (right), 5A7 accumulates in MK in the BM and surface expression of GPIbα of newly produced platelets is decreased by shedding and internalization. Antibody concentration in plasma is not high enough to induce platelet aggregation, and thus, opsonized platelets are primarily cleared by splenic macrophages. (B) In the spleen, the antibody-bound platelets or platelet aggregates are cleared by macrophages via Fc receptor in both models. (C) In the acute platelet-depletion model, platelets are desialylated and aggregated. These platelets interact with hepatocytes through fenestrations in liver sinusoidal endothelial cells, transduce signal through AMR to upregulate TPO mRNA expression. Trapped desialylated platelets and aggregates are cooperatively cleared via hepatocyte capture and Kupffer cell phagocytosis. C-type lectins expressed on Kupffer cells, such as CLEC4F and macrophage galactose-type lectin,, which have high affinity for desialylated glycoproteins, may contribute to this process.