IFITM3 regulates fibrinogen endocytosis and platelet reactivity in nonviral sepsis
Robert A. Campbell, Bhanu Kanth Manne, Meenakshi Banerjee, Elizabeth A. Middleton, Abigail Ajanel, Hansjorg Schwertz, Frederik Denorme, Chris Stubben, Emilie Montenont, Samantha Saperstein, Lauren Page, Neal D. Tolley, Diana L. Lim, Samuel M. Brown, Colin K. Grissom, Douglas W. Sborov, Anandi Krishnan, Matthew T. Rondina
Robert A. Campbell, Bhanu Kanth Manne, Meenakshi Banerjee, Elizabeth A. Middleton, Abigail Ajanel, Hansjorg Schwertz, Frederik Denorme, Chris Stubben, Emilie Montenont, Samantha Saperstein, Lauren Page, Neal D. Tolley, Diana L. Lim, Samuel M. Brown, Colin K. Grissom, Douglas W. Sborov, Anandi Krishnan, Matthew T. Rondina
View: Text | PDF
Research Article Hematology Inflammation

IFITM3 regulates fibrinogen endocytosis and platelet reactivity in nonviral sepsis

Abstract

Platelets and megakaryocytes are critical players in immune responses. Recent reports suggest infection and inflammation alter the megakaryocyte and platelet transcriptome to induce altered platelet reactivity. We determined whether nonviral sepsis induces differential platelet gene expression and reactivity. Nonviral sepsis upregulated IFN-induced transmembrane protein 3 (IFITM3), an IFN-responsive gene that restricts viral replication. As IFITM3 has been linked to clathrin-mediated endocytosis, we determined whether IFITM3 promoted endocytosis of α-granule proteins. IFN stimulation enhanced fibrinogen endocytosis in megakaryocytes and platelets from Ifitm+/+ mice, but not Ifitm–/– mice. IFITM3 overexpression or deletion in megakaryocytes demonstrated IFITM3 was necessary and sufficient to regulate fibrinogen endocytosis. Mechanistically, IFITM3 interacted with clathrin and αIIb and altered their plasma membrane localization into lipid rafts. In vivo IFN administration increased fibrinogen endocytosis, platelet reactivity, and thrombosis in an IFITM-dependent manner. In contrast, Ifitm–/– mice were completely rescued from IFN-induced platelet hyperreactivity and thrombosis. During murine sepsis, platelets from Ifitm+/+ mice demonstrated increased fibrinogen content and platelet reactivity, which was dependent on IFN-α and IFITMs. Platelets from patients with nonviral sepsis had increases in platelet IFITM3 expression, fibrinogen content, and hyperreactivity. These data identify IFITM3 as a regulator of platelet endocytosis, hyperreactivity, and thrombosis during inflammatory stress.

Authors

Robert A. Campbell, Bhanu Kanth Manne, Meenakshi Banerjee, Elizabeth A. Middleton, Abigail Ajanel, Hansjorg Schwertz, Frederik Denorme, Chris Stubben, Emilie Montenont, Samantha Saperstein, Lauren Page, Neal D. Tolley, Diana L. Lim, Samuel M. Brown, Colin K. Grissom, Douglas W. Sborov, Anandi Krishnan, Matthew T. Rondina

×

Figure 2

IFN-induced IFITM3 expression is dependent on STAT1 and mTOR pathways.

Options: View larger image (or click on image) Download as PowerPoint
IFN-induced IFITM3 expression is dependent on STAT1 and mTOR pathways. (...
(A–D) CD34+-derived cells were transfected on day 5 of culture with negative (Neg) control or STAT1 crRNA. Megakaryocytes were then stimulated with IFN-α (500 or 1000 U/mL, final) or no treatment (NT) on day 13. After 24 hours, megakaryocytes were lysed and probed to examine total STAT1 (B), p-STAT1 (C), and IFITM3 expression by immunoblot (D). β-Tubulin was used as a loading control. Representative blots of total STAT1, p-STAT1, and IFITM3 expression are shown (A–D). Statistical analysis used was a mixed effect analysis with Tukey’s multiple comparison test with values normalized to NEG NT (n = 5–6). **P ≤ 0.01; ****P ≤ 0.0001. (E–H) CD34+-derived cells were transfected on day 5 of culture with negative control or mTOR crRNA. Megakaryocytes were then stimulated with IFN-α (500 or 1000 U/mL, final) or no treatment on day 13. After 24 hours, megakaryocytes were lysed and probed to examine mTOR expression after IFN-α treatment (E and F), mTOR deletion after CRISPR (E and G), and IFITM3 expression by immunoblot (E and H). β-Actin was used as a loading control. Representative blots are shown. Statistical analysis used was Wilcoxon’s test with values normalized to negative (G), Kruskal-Wallis test with values normalized to negative (F), and 2-way ANOVA with Šidák’s multiple comparison test with values normalized to negative vehicle (H) (n = 6–7). *P ≤ 0.05. The β-actin in Figure 2E is the same β-actin as in Supplemental Figure 6A.