Fig 2: Kupffer cells are the primary cell type phagocytizing desialylated platelets in the liver.A, B Representative confocal microscopic images. CD41 marks platelets, F4/80 marks Kupffer cells, CD31 labels sinusoidal endothelium, and hepatocytes were detected by autofluorescence. C Pseudocolored TEM micrographs of WT or Clec4f-/- ultrathin sections. Red arrow marks a phagocytized platelet. Blue arrows mark potential interactions between platelets and sinusoidal endothelium.

Fig 3: Kupffer cell CLEC4F is a dominant receptor in mediating the rapid clearance of desialylated platelets in the liver.A Flow cytometry analysis of galactose exposure (RCA I binding) on platelets of either WT or Clec4f-/- recipients treated with Arthrobacter ureafaciens-derived neuraminidase at different time points. MFI, Mean fluorescence intensity. B Percentage of platelets of either WT or Clec4f-/- recipients treated with Arthrobacter ureafaciens-derived neuraminidase at different time points. Platelet count was measured using flow cytometric analysis of peripheral blood based on anti-CD41. C Representative immunofluorescent images showing association of platelets (CD41, red) with Kupffer cells (F4/80, blue) in the liver sections. Endothelium, (CD31, green). D, E Quantification of association of platelets with liver Kupffer cells or Kupffer cell (KC) number at 2 h after a neuraminidase injection. n = 5–10 images/10 × fields/experiment. Data represent means ± SEM from three experiments. *P < 0.05.

Fig 4: CLEC4F is critical to the Kupffer cell-mediated phagocytosis of desialylated platelets.A Representative confocal microscopic images of desialylated platelets (CD41, red) and Kupffer cells (F4/80, blue) in liver sections of different mouse models. B Quantification of platelets with liver Kupffer cells of different mouse models 2 h after Clostridium perfringens-derived neuraminidase injection. n = 5–10 images/20 × fields/mouse. Data represent means ± SEM from three experiments. *P < 0.05. ns, not significant.

Fig 5: TPO treatment and inhibition of TGF-ß1 activity increase MK numbers in BM culture in vitro. BM from PF4Cre;Tgfb1flox/flox or WT mice were cultured with and without TPO for up to 5 days. BM cells from WT mice were cultured with TPO in the presence or absence of anti–TGF-ß1–neutralizing antibody (a-TGF-ß1; 2 µg/mL; AF101; R&D Systems). (A-B) MK numbers (large cells indicated by yellow arrows) were counted manually under a light microscope. (C-D) MK numbers were counted by flow cytometry after staining with CD41 antibody, and nuclei were labeled with Hoechst. (D) Representative flow cytometric dot plot analysis of gating of MKs with double-positive CD41+Hoechst+. (E) Plot shows MK numbers counted for indicated conditions. FSC, forward scatter; SSC, side scatter.