Dengue induces platelet activation, mitochondrial dysfunction and cell death through mechanisms that involve DC-SIGN and caspases

Abstract

Background: Worldwide, dengue is the most prevalent human arbovirus disease. Dengue infection may cause a range of clinical manifestations from self-limiting febrile illness through to a life-threatening syndrome accompanied by both bleeding and shock. Thrombocytopenia is frequently observed in mild and severe disease; however, the mechanisms involved in DENV-induced platelet activation and thrombocytopenia are incompletely understood.

Patients and methods: Freshly isolated platelets from patients with dengue were evaluated for markers of activation, mitochondrial alteration and activation of cell death pathways. In parallel, we examined direct DENV-induced activation and apoptosis of platelets obtained from healthy subjects.

Results: We found that platelets from DENV-infected patients exhibited increased activation by comparison to control subjects. Moreover, platelets from DENV-infected patients exhibited classic signs of the intrinsic pathway of apoptosis that include increased surface phosphatidylserine exposure, mitochondrial depolarization and activation of caspase-9 and -3. Indeed, thrombocytopenia was shown to strongly associate with enhanced platelet activation and cell death in DENV-infected patients. Platelet activation, mitochondrial dysfunction and caspase-dependent phosphatidylserine exposure on platelets were also observed when platelets from healthy subjects were directly exposed to DENV in vitro. DENV-induced platelet activation was shown to occur through mechanisms largely dependent on DC-SIGN.

Conclusions: Together our results demonstrate that platelets from patients with dengue present signs of activation, mitochondrial dysfunction and activation of the apoptosis caspase cascade, which may contribute to the development of thrombocytopenia in patients with dengue. Our results also suggest the involvement of DC-SIGN as a critical receptor in DENV-dependent platelet activation.

Figure 1. Platelet activation is increased during dengue illness

The mean fluorescence intensity (MFI) of P-selectin expression (A) and the percentage of annexin V-binding platelets (B) in platelets freshly-isolated from healthy subjects (control), patients with non-dengue febrile illness (NDFI), and dengue-infected patients in febrile (Feb), defervescence (Def), and convalescence (Conv) phases. Boxes indicate median and interquartile ranges and whiskers indicate 5-95 percentile. *p<0.05 versus control; #p<0.05 versus NDFI.

Figure 2. Platelet counts in patients with dengue correlate with indices of platelet activation

(A) The mean fluorescence intensity (MFI) of P-selectin expression in thrombocytopenic (TCP) and non-thrombocytopenic (NTCP) dengue patients. Boxes indicate the median and interquartile ranges and whiskers indicate 5-95 percentile. (B-C) Platelet counts were plotted against the MFI of P-selectin expression (B) and the percent of annexin V-binding platelets (C) in febrile (Feb) defervescence (Def) and convalescence (Conv) phases. In Panels B and C the analysis was restricted to platelet counts obtained on the same day that P-selectin and phosphatidylserine were analyzed. (D) Platelet counts and the MFI of P-selectin expression were plotted against the day of illness in which each value was obtained. Non-linear regressions were traced according to the distribution of dots. *p<0.05 versus control; #p<0.01 TCP versus NTCP.

Figure 3. Mitochondrial function is impaired in platelets from dengue-infected patients

(A-C) The mean fluorescence intensity (MFI) for TMRE, (D) the Citrate-Synthase activity, and (E-F) the MFI of MitoSoxRed in platelets from healthy subjects (control), patients with non-dengue febrile illness (NDFI), and dengue-infected patients in febrile (Feb), defervescence (Def), or convalescence (Conv) phases. (A, C and E) Mitochondrial responses in platelets treated with FCCP or oligomycin. Bars (A) depict the mean±SEM of 7-9 healthy participants and dengue patients; dots (C and E) represent TMRE or MitoSoxRed fluorescence in platelets from control or dengue patients before (basal) and after the treatments. Boxes (B, D and F) indicate median and interquartile ranges and whiskers indicate 5-95 percentile. +p<0.05 versus basal; *p<0.05 versus control; #p<0.05 versus NDFI.

Figure 4. Platelets apoptosis in patients with dengue

(A) The mean fluorescence intensity (MFI) for TMRE was plotted against the percentage of annexin V-binding platelets in febrile (Feb), defervescence (Def) and convalescence (Conv) dengue phases. (B) The MFI of Caspase-9 activation in platelets freshly-isolated from healthy subjects (control), patients with non-dengue febrile illness (NDFI), and dengue-infected patients in Feb, Def, and Conv. Boxes indicate median and interquartile ranges and whiskers indicate 5-95 percentile. (C) Western analysis of pro- and cleaved caspase-9 (casp-9) and caspase-3 (casp-3), and β-actin in platelets isolated from control or dengue patients (representative of 5). *p<0.01 versus control; #p<0.05 versus NDFI.

Figure 5. DENV-2 induces platelet activation

The mean fluorescence intensity (MFI) of P-selectin expression (A, B and D) and the percentage (%) of annexin V-binding platelets (C and E) in platelets exposed to DENV in vitro. (A and C) Platelets were exposed to mock or DENV-2 for 1 min, 30 min, 1½ hour, 3 hours or 6 hours; or activated with thrombin for 1 min, 5 min, 15 min, 30 min or 1 hour. (B) Platelets were exposed (6 hours) to the filtrate or the retentate of DENV-2 purified through centrifugation in a Centricon Filter, or mock processed in parallel. (D and E) Platelets were exposed (6 hours) to mock, DENV or heat-inactivated DENV in the presence of the pan caspase inhibitor ZVAD-fmk. Dots and bars represent mean±SEM of 4 independent experiments from individual donors. *p<0.05 versus mock; #p<0.05 versus DENV filtrate or ZVAD-fmk treated platelets.

Figure 6. DENV-2 induces mitochondrial dysfunction and apoptosis in platelets

The mean fluorescence intensity (MFI) for TMRE (A), MitoSoxRed (B) and Caspase-9 activation (C) in platelets exposed for 6 hours to mock, DENV-2 or heat-inactivated DENV-2. The bars represent mean±SEM of 4 to 6 independent experiments from individual donors. (D) The MFI for TMRE was plotted against the percentage of annexin V-binding platelets. Representative density plots are shown. *p<0.01 versus mock.

Figure 7. DENV-2 activates platelets through mechanisms that involve DC-SIGN

Platelets were exposed for 6 hours to mock, DENV-2 or heat-inactivated DENV-2 in the presence or absence of neutralizing antibodies against DC-SIGN or the α_V integrin subunit. The percent increase in P-selectin expression (A) or percent decrease in TMRE fluorescence (B) related to mock values are shown. (C) Representative density plots showing the expression of DC-SIGN on platelets. (D) The percentage of DC-SIGN-expressing platelets exposed to mock or DENV. (E) Platelets exposed to mock or DENV-2 were stained for P-selectin and DC-SIGN. The mean fluorescence intensity (MFI) for P-selectin expression was assessed on platelets gated as DC-SIGN negative (DC-SIGN-) or positive (DC-SIGN+). Representative histograms and density plots are shown next to their corresponding graphs. Bars represent the mean±SEM of 4 independent experiments from individual donors. *p<0.05 versus mock; #p<0.05 between anti-DC-SIGN or IgG isotype (A and B); or between DC-SIGN- and DC-SIGN+ (E).