Dengue virus nonstructural protein 1 activates platelets via Toll-like receptor 4, leading to thrombocytopenia and hemorrhage

Abstract

Dengue virus (DENV) infection, the most common mosquito-transmitted viral infection, can cause a range of diseases from self-limiting dengue fever to life-threatening dengue hemorrhagic fever and shock syndrome. Thrombocytopenia is a major characteristic observed in both mild and severe dengue disease and is significantly correlated with the progression of dengue severity. Previous studies have shown that DENV nonstructural protein 1 (NS1), which can be secreted into patients' blood, can stimulate immune cells via Toll-like receptor 4 (TLR4) and can cause endothelial leakage. However, it is unclear whether DENV NS1 can directly induce platelet activation or cause thrombocytopenia during DENV infection. In this study, we first demonstrated that DENV but not Zika virus cell culture supernatant could induce P-selectin expression and phosphatidylserine (PS) exposure in human platelets, both of which were abolished when NS1 was depleted from the DENV supernatant. Similar results were found using recombinant NS1 from all four serotypes of DENV, and those effects were blocked in the presence of anti-NS1 F(ab')2, anti-TLR4 antibody, a TLR4 antagonist (Rhodobacter sphaeroides lipopolysaccharide, LPS-Rs) and a TLR4 signaling inhibitor (TAK242), but not polymyxin B (an LPS inhibitor). Moreover, the activation of platelets by DENV NS1 promoted subthreshold concentrations of adenosine diphosphate (ADP)-induced platelet aggregation and enhanced platelet adhesion to endothelial cells and phagocytosis by macrophages. Finally, we demonstrated that DENV-induced thrombocytopenia and hemorrhage were attenuated in TLR4 knockout and wild-type mice when NS1 was depleted from DENV supernatant. Taken together, these results suggest that the binding of DENV NS1 to TLR4 on platelets can trigger its activation, which may contribute to thrombocytopenia and hemorrhage during dengue infection.

Fig 1. DENV NS1 is critical for DENV-induced P-selectin expression and PS exposure in platelets.

(A) (B) Platelets were incubated for 1 h with DENV or ZIKV supernatant that contained different concentrations of NS1 as indicated (n = 4). (C) (D) Platelets were incubated with medium, 10 μg/ml NS1-containing DENV supernatant or NS1-depleted DENV supernatant (NS1Δ) for 1 h (n = 3). The percent fluorescence of P-selectin surface expression on platelets and annexin V binding to platelets were analyzed by FACSCalibur flow cytometry. *P<0.05, ***P<0.001; unpaired t-test (panels A and B), Kruskal-Wallis ANOVA (panels C and D).

Fig 2. DENV NS1 directly induces platelet activation in a dose- and time-dependent manner.

(A) Platelets were treated with BSA or DENV NS1 recombinant proteins (10 μg/ml) for the indicated time (n = 3). (B) Platelets were treated with different concentrations of DENV NS1 recombinant proteins for 1 h (n = 3). (C) Platelets were treated with BSA or DENV NS1 (10 μg/ml) (cotreated with or without 25 μg/ml anti-NS1 specific antibodies F(ab’)₂ fragment, 33D2 F(ab’)₂, or the isotype-matched mouse antibodies F(ab’)₂ and cmIgG F(ab’)₂) for 1 h (n = 3). (D) Platelets were treated with BSA or DENV1-4 NS1 (10 μg/ml) for 1 h (n = 3). The percent fluorescence of P-selectin surface expression on platelets was analyzed by FACSCalibur flow cytometry. ***P<0.001; unpaired t-test (panels A and B), Kruskal-Wallis ANOVA (panels C and D).

Fig 3. DENV NS1 directly induces platelet apoptosis in a dose- and time-dependent manner.

(A) Platelets were treated with BSA or DENV NS1 recombinant protein (10 μg/ml) for the indicated time (n = 3). (B) Platelets were treated with different concentrations of DENV NS1 recombinant protein for 1 h (n = 3). (C) Platelets were treated with BSA or DENV NS1 (10 μg/ml) (cotreated with or without 25 μg/ml 33D2 F(ab’)₂, or cmIgG F(ab’)₂) for 1 h (n = 3). (D) Platelets were treated with BSA or DENV1-4 NS1 (10 μg/ml) for 1 h (n = 3). The percent fluorescence of annexin V binding to platelets was analyzed by FACSCalibur flow cytometry. *P<0.05, **P<0.01, ***P<0.001; unpaired t-test (panels A and B), Kruskal-Wallis ANOVA (panels C and D).

Fig 4. DENV NS1 binds to platelets and induces activation through TLR4 signal transduction.

(A) The protein expression levels of TLR4 and β actin, an internal control, in human-isolated platelets from 3 different donors were detected using Western blotting (50 μg protein/lane). (B) Human-isolated platelets were preincubated with αTLR4 or a control Rabbit IgG for 1 h, and the binding of NS1 on platelet surfaces was determined by flow cytometry using FITC-conjugated anti-NS1 monoclonal antibodies (33D2-FITC) (n = 3). (C) The binding of NS1 on platelets isolated from wild-type or TLR4 knockout mice was determined by flow cytometry (n = 4). Platelets were preincubated with or without different concentrations of (D) αTLR4 (E) the TLR4 antagonist LPS-Rs, (F) the TLR4 signaling inhibitor TAK242, or the LPS inhibitor polymyxin B (10 μg/ml) for 30 min, followed by BSA or NS1 (10 μg/ml) stimulation for 1 h (n = 4). The percent fluorescence of NS1 binding and the P-selectin surface expression on platelets were analyzed by FACSCalibur flow cytometry. *P<0.05, **P<0.01; Kruskal-Wallis ANOVA (panels B to panel F).

Fig 5. DENV NS1 enhances platelet aggregation ability after activation.

(A) Platelets were plated on poly-L-lysine-coated coverslips and treated with BSA or DENV NS1 (10 μg/ml) for 1 h. After fixation and permeabilization, platelets were stained with anti-CD61 polyclonal antibody and an anti-rabbit Alexa 594-conjugated antibody. (B) The light transmission of platelet-rich plasma (PRP) was measured in a Chrono-log aggregometer for 5 min after ADP (10 μM) or DENV NS1 (10 μg/ml) was added. (C) PRP was treated with BSA or DENV NS1 (10 μg/ml) (cotreated with or without 25 μg/ml 33D2 F(ab’)2, or cmIgG F(ab’)2) for 1 h and stimulated with ADP (2.5 μM). The light transmission of PRP was measured in a Chrono-log aggregometer, and (D) the maximum percentage of platelet aggregation was calculated by an AggRAM^™ platelet aggregation analyzer (n = 3). **P<0.01; Kruskal-Wallis ANOVA (panel D).

Fig 6. DENV NS1 increases platelet adherence to endothelial cells and phagocytosis by macrophages.

(A) DENV-NS1-activated platelets were cultured with confluent HUVEC monolayers for 1 h, followed by washing and fixation. The adherent platelets were examined by confocal microscopy. CD61, a platelet marker, is stained in green, and HUVEC nuclei are stained with Hoechst (blue). (B) The adherent platelets on HUVEC monolayers in (A) were quantified in 100 HUVECs from three independent experiments and expressed as the mean fluorescence intensity (MFI) with ImageJ. (C) For the phagocytosis assay, DENV-NS1-activated platelets were cultured with PMA-activated THP-1 cells for 4 h. After fixation and permeabilization, the cells were stained with CD61 (green, a marker of platelets) and CD14 (red, a marker of monocytes/macrophages) and examined by confocal microscopy. The yellow dots represent the engulfed platelets, and the green dots represent the adherent platelets. (D) The average number of engulfed platelets per THP-1 cell was determined by counting 100 THP-1 cells per sample. The images were further acquired by confocal microscopy. *P<0.05; Kruskal-Wallis ANOVA (panels B and D).

Fig 7. DENV NS1 causes the formation of aggregated complexes and cell death in a coculture system.

(A) PMA-activated THP-1 cells and isolated platelets were added to confluent HUVECs on coverslips in 24-well plates with different treatments. After 1 h of incubation, unbound cells were washed out, and images were obtained by optical microscopy. (B) After fixation and permeabilization, monolayers were stained for CD61 (green), CD14 (red) and nuclei (blue). (C) Supernatants from different coculture conditions were collected, and cell death was determined using an LDH release assay. The images were further acquired by confocal microscopy. *P<0.05; **P<0.01; unpaired t-test (panels C).

Fig 8. DENV NS1 is critical for DENV-induced thrombocytopenia, prolonged bleeding time, and hemorrhage in mice.

A hemorrhagic C3H/HeN mouse model was created as described in the Methods. (A) Mouse skin samples were removed to observe local hemorrhage on day 3 after DENV injection. The number of mice with hemorrhage divided by the total number of mice inoculated in each group is indicated. Yellow arrows indicate local skin hemorrhage. (B) The clinical hemorrhage score was quantified and determined as digital hemorrhage severity. (C) The tail bleeding time and (D) platelet counts were also determined on day 3 before sacrifice. *P<0.05; Tukey’s multiple comparison test (panel B); Kruskal-Wallis ANOVA (panel C and D).

Fig 9. TLR4 is involved in DENV-induced prolongation of bleeding time and hemorrhage in mice.

A hemorrhagic mouse model was performed with C57BL/6J mice (WT) and TLR4^-/- C57BL/6J background mice as described in the Methods. (A) Mouse skin samples were removed to observe local hemorrhage on day 3 after DENV injection. The number of mice with hemorrhage divided by the total number of mice inoculated in each group is indicated. Yellow arrows indicate local skin hemorrhage. (B) The clinical score of hemorrhage was quantified and determined as digital hemorrhage severity. (C) The tail bleeding time and (D) platelet counts were also determined on day 3 before sacrifice. *P<0.05, **P<0.01; Tukey’s multiple comparison test (panel B); Kruskal-Wallis ANOVA (panel C and D).

Fig 10. Proposed mechanisms of the contribution of DENV NS1 to cause thrombocytopenia and hemorrhage during DENV infection.

Circulating DENV NS1 binds to platelets via TLR4 or other molecules to induce the release of ADP, which in turn elevates P-selectin expression and PS exposure on platelet surfaces leading to platelet activation and enhancement of the platelet aggregation. In addition, NS1-activated platelets are prone to adhere onto endothelium or phagocytosis by macrophages. On the other hand, NS1 can also bind to endothelial cells and macrophage to cause their activation and cytokine release. All these effects induced by NS1 can contribute to the thrombocytopenia and hemorrhage during DENV infection.