Inhibition of transcription factor NFAT activity in activated platelets enhances their aggregation and exacerbates gram-negative bacterial septicemia

Abstract

During gram-negative septicemia, interactions between platelets and neutrophils initiate a detrimental feedback loop that sustains neutrophil extracellular trap (NET) induction, disseminated intravascular coagulation, and inflammation. Understanding intracellular pathways that control platelet-neutrophil interactions is essential for identifying new therapeutic targets. Here, we found that thrombin signaling induced activation of the transcription factor NFAT in platelets. Using genetic and pharmacologic approaches, as well as iNFATuation, a newly developed mouse model in which NFAT activation can be abrogated in a cell-specific manner, we demonstrated that NFAT inhibition in activated murine and human platelets enhanced their activation and aggregation, as well as their interactions with neutrophils and NET induction. During gram-negative septicemia, NFAT inhibition in platelets promoted disease severity by increasing disseminated coagulation and NETosis. NFAT inhibition also partially restored coagulation ex vivo in patients with hypoactive platelets. Our results define non-transcriptional roles for NFAT that could be harnessed to address pressing clinical needs.

Keywords: NETs; coagulation; gram-negative bacteria; neutrophil extracellular traps; neutrophils; platelets; sepsis; septic shock.

Figure 1.. NFAT is activated upon stimulation of the thrombin receptor and modulates platelet aggregation

(A) Western blot analysis of NFAT1, NFAT2, NFAT3, and NFAT4 was performed on lysates from murine platelets or the uterine tissue. One experiment representative of three experiments. (B) Phos-tag western blot analysis of mouse platelets pretreated, or not, with myristoylated VIVIT (MyrVIVIT) (10 μM) for 30 min and then stimulated with PAR4-AP (100 μM) for the indicated time points. One experiment representative of three experiments. (C) Murine platelets were pretreated, or not, with vehicle (DMSO), the myristoylated control peptide (MyrVEET), or MyrVIVIT (10 μM) for 10 min and then activated with PAR4-AP (75 μM). Aggregation was monitored for 5 min. Histograms represent the percentage of maximal aggregation and area under curve (AUC). Data represent mean ± SEM. n ≥ 3 independent experiments. (D) Human platelets were treated as in (C) and stimulated with TRAP (5 μM). Histograms represent the percentage of the maximal aggregation and AUC. Data represent mean ± SEM. n ≥ 3 independent experiments. (E) Platelet-rich plasma (PRP) from WT or Nfat1^−/− mice was stimulated with PAR4-AP (150 μM). Histograms represent the percentage of maximal aggregation and AUC. Each dot represents an independent test and lines highlight measurements performed concomitantly. (F) PRP from WT or Nfat1^−/− mice was stimulated with PAR4-AP (150 and 100 μM). Integrin αIIbβ3 activation was analyzed by cytofluorimetry. Bars represent the mean fluorescence intensity (MFI). Data represent mean ± SEM. n ≥ 3 independent experiments. (G) Representative phos-tag western blot analysis of murine platelets pretreated, or not, with 1 μM FK506 for 1 h and then stimulated with PAR4-AP (100 μM) for the indicated time. (H) Murine PRP was treated as in (G). Aggregation curves and histograms of maximal aggregation and AUC are shown. Each dot represents an independent test. (I) PRP was pretreated as in (G) and then stimulated with different concentration of PAR4-AP (150 and 100 μM). Integrin αIIbβ3 activation was analyzed by cytofluorimetry. Bars represent the MFI. Data represent mean ± SEM. n ≥ 3 independent experiments. (J) PRP from WT or Nfat1^−/− mice was pretreated as in (G), and integrin αIIbβ3 activation was analyzed by cytofluorimetry. Data represent mean ± SEM. n ≥ 3 independent experiments. One-way ANOVA (C and D), unpaired two-tailed t test (E and H), or two-way ANOVA (F, I, and J) were used for statistics. n.s., not significant (p > 0.05), **p < 0.01, ***p < 0.001, and ****p < 0.0001. See also Figure S1.

Figure 2.. NFAT inhibition potentiates the aggregation of human hypoactive platelets

(A) PRP from healthy donors was pretreated, or not, with 1 μg/mL cyclosporin A (CsA) for 2 h and then treated with TRAP (3 μM). Aggregation was monitored for 5 min. Histograms of maximal aggregation and AUC are shown. Each dot represents a different donor. (B) Human PRP was pretreated, or not, with 1 μg/mL CsA (CsA hi) or 500 ng/mL CsA (CsA lo) and then stimulated with TRAP (1.5 μM). Integrin αIIbβ3 activation was analyzed by cytofluorimetry. Data show the percentage of the increase of the αIIbβ3 MFI of CsA-treated PRP compared with the αIIbβ3 MFI of PRP not pretreated with CsA. Each dot represents a different donor. (C) PRP from healthy individual or patients with type 2 or type 1 HPS, as indicated, were pretreated, or not, with 1 μg/mL CsA and then stimulated with TRAP (3 μM). Histograms show platelets’ maximal aggregation. Each dot represents an independent test performed on one HPS type 2 patient (left panel) or on two HPS type 1 patients (pz1 and pz2, right panel). Paired t test (A and C) or one-sample t test (B) were used for statistics. *p < 0.05, **p < 0.01, ***p < 0.001. See also Figure S2.

Figure 3.. NFAT inhibition increases platelet aggregation in vivo

(A) PRP from mice treated, or not, with 4 mg/kg of FK506 for 14 days were stimulated with PAR4-AP (150 μM). Maximal aggregation and AUC are shown. Each dot represents an independent test and lines highlight measurements performed concomitantly. (B) Tail bleeding test was performed on mice treated, or not, with FK506 (left panel) or on WT and Nfat1^−/− mice (right panel). Each dot represents a mouse (n = 29 for FK-treated mice, n = 16 for Nfat1^−/− mice). (C) Mice treated, or not, with 4 mg/kg of FK506 for 14 days (left) or WT and Nfat1^−/− mice (right) were intravenously injected with a mixture of collagen (500 μg/kg) and epinephrine (3 μg/kg) and were scored for a pathology index. Each dot represents a mouse (n = 9 for FK-treated mice, n = 5 for Nfat1^−/− mice). (D) Tail bleeding test (left panel) and thromboembolism test (right panel) were performed on HPS mice treated, or not, with FK506. Each dot represents a mouse (n = 6). Unpaired two-tailed t test was used for statistics. *p < 0.05, **p < 0.01, and ***p < 0.001. See also Figure S3.

Figure 4.. NFAT function governs the interaction between platelets and neutrophils

(A) Whole blood from mice treated, or not, with 4 mg/kg of FK506 for 14 days was stimulated, or not, with PAR4-AP (100 μM). The formation of complexes between platelets and neutrophils was analyzed by cytofluorimetry. Results show the MFI for the platelet marker CD41 on neutrophils, identified as Ly6G⁺CD11b⁺ cells. Data represent mean ± SEM. n ≥ 3 independent experiments. (B) Platelet-neutrophil complex formation was analyzed by ImageStream. Platelets treated, or not, with 1 μM FK506 for 1 h and identified with different fluorochromes, were washed, mixed with the blood from platelet-depleted mice, and stimulated (PAR4-AP), or not (Unstim), with PAR4-AP (25 μM). The percentage of neutrophils (identified as Ly6G⁺ cells in the whole blood) that did not bind any platelets (0), or that bound 1 or 2 platelets (1–2), or 3 or more platelets (>3) was measured for both platelets pretreated (FK-506) or not (NT) with FK506. Data represent mean ± SEM. n ≥ 3 independent experiments. (C) PRP was pretreated, or not, with 1 μM FK506 for 1 h and stimulated with PAR4-AP (150 μM) for 5 min. PAR4-AP-activated platelets were then used to induce NETosis. Histograms represent the percentage of neutrophils positive for the citrullinated histone H3 (left). Each dot represents an independent test. A representative image of NET released by neutrophils is shown (right). Scale bars, 10 μm. (D) Murine PRP was pretreated, or not, with 1 μM FK506 for 1 h and then stimulated with different concentration of PAR4-AP (150 and 100 μM). P-selectin exposure was analyzed by cytofluorimetry and P-selectin MFI is shown. Data represent mean ± SEM. n ≥ 3 independent experiments. (E) Neutrophils were stimulated as in (C) in the presence of the anti-P-selectin antibody or respective isotype control. Data represent mean ± SEM. n ≥ 3 independent experiments. One-way ANOVA (A, D, and E), two-way ANOVA (B), or unpaired two-tailed t test (C) were used for statistics. n.s., not significant (p > 0.05), *p < 0.05, **p < 0.01, and ****p < 0.0001. See also Figure S4.

Figure 5.. NFAT modulates release of platelet granules

(A) PRP was pretreated, or not, with 1 μM FK506 for 1 h and incubated, or not, for 10 min with 3 μM ARC-66096 (ARC) then stimulated with different concentration of PAR4-AP (150 and 100 μM). Integrin αIIbβ3 activation and P-selectin exposure were analyzed by cytofluorimetry. Bars represent the MFI. Data represent mean ± SEM. n ≥ 3 independent experiments. (B) PRP was pretreated, or not, with 1 μM FK506 for 1 h and the release of ATP was measured upon stimulation with PAR4-AP (150 μM). Histogram represents maximal ATP release. Data represent mean ± SEM. n ≥ 3 independent experiments. (C and D) PRP was pretreated as in (B) and then stimulated with ADP (20 μM). Aggregation (C) and integrin αIIbβ3 activation and P-selectin exposure were analyzed (D). Data represent mean ± SEM. n ≥ 3 independent experiments. (E) PRP was pretreated as in (B) and then stimulated with PAR4-AP (150 μM) alone or in combination with ADP (10 nM). Integrin αIIbβ3 activation and P-selectin exposure were analyzed by cytofluorimetry. Data show mean ± SEM. n ≥ 3 independent experiments. Two-way ANOVA (A, D, and E) or unpaired two-tailed t test (B) were used for statistics. n.s., not significant (p > 0.05), *p < 0.05, **p < 0.01, and ****p < 0.0001. See also Figure S5.

Figure 6.. The iNFATuation model demonstrates that NFAT activation in platelets regulates platelet functions

(A) Schematic of the iNFATuation transgenic mouse model. (B) Blood from WT, littermates or iNFATuation-Pf4-cre mice was stained for the platelet marker CD41 and analyzed for expression of VIVIT-tdTomato signal. One experiment representative of three experiments. (C) PRP from littermates or iNFATuation-Pf4-cre mice was stimulated with PAR4-AP (150 μM). Aggregation was monitored for 5 min. Histograms of maximal aggregation and AUC are shown. Each dot represents an independent test. (D) PRP from littermates or iNFATuation-Pf4-cre mice was stimulated with PAR4-AP (150 μM). P-selectin exposure was analyzed by cytofluorimetry, and P-selectin MFI is shown. Data represent mean ± SEM. n ≥ 3 independent experiments. (E) Whole blood from littermates or iNFATuation-Pf4-cre mice was treated with PAR4-AP (100 μM). The formation of complexes between platelets and neutrophils was analyzed by cytofluorimetry. Results show the MFI for the platelet marker CD41 on neutrophils, identified as Ly6G⁺CD11b⁺ cells. Data represent mean ± SEM. n ≥ 3 independent experiments. (F) PRP from littermates or iNFATuation-Pf4-cre mice was stimulated with PAR4-AP (150 μM) for 5 min. PAR4-AP-activated platelets were then used to induce NETosis. Histograms represent the percentage of neutrophils positive for the citrullinated histone H3. Each dot represents an independent test. (G) Tail bleeding test (left panel) and thromboembolism test (right panel) were performed on littermates or iNFATuation-Pf4-cre mice. Each dot represents a mouse (n = 4 littermates, n = 6 iNFATuation-Pf4-cre). Data show mean ± SEM. Unpaired two-tailed t test (C, F, and G) and two-way ANOVA (D and E) were used for statistics. *p < 0.05, **p < 0.01, and ***p < 0.001. See also Figure S6.

Figure 7.. NFAT inhibition exacerbates inflammation during sepsis

(A) Schematic of the septic shock model. Mice were intraperitoneally (i.p.) injected, or not, with 25 mg/kg of GSK484 (−24 h and −6 h) and then were i.p. administered with two doses of LPS (−4 h, 1 mg/kg and 0 h, 2 mg/kg) to induce septic shock. Peritoneal lavage and blood were collected 2 h after the LPS challenge. (B) Schematic of gram-negative bacteria-induced sepsis. Mice were i.p. injected with 1 × 10⁸ CFU of E. coli. Peritoneal lavage and blood were collected 6 h after the E. coli challenge. (C) Littermates or iNFATuation-Pf4-cre mice were challenged with two doses of LPS (−4 h, 1 mg/kg and 0 h, 2 mg/kg). Temperatures were recorded at the indicated time points. n = 4 mice per group. Data show mean ± SEM. (D–I) Littermates or iNFATuation-Pf4-cre mice were pretreated, or not, with GSK484 and then challenged with two doses of LPS. 2 h after LPS challenge blood was collected, platelet number (D) was assessed by cytofluorimetry, and plasma fibrinogen content (E) and thrombin-antithrombin complexes (F) were measured by ELISA. Neutrophils positive for platelet marker CD41 (G) and for citrullinated histone H3 (H) were quantified in the peritoneal lavage by cytofluorimetry. Plasma IL-6 content was measured by ELISA 18 h post challenge (I). Each dot represents a mouse. “Saline” are WT mice not challenged with LPS used as control. (J) Littermates or iNFATuation-Pf4-cre mice were challenge with 1 × 10⁸ CFU of E. coli. 6 h post infection, plasma IL-6 was quantified by ELISA. Each dot represents a mouse. “Saline” are WT mice not challenged with E. coli used as control. (K–M) Platelet-depleted mice reconstituted with WT or Nfat1^−/− platelets were challenged with 2 doses of LPS. 2 h after LPS challenge temperatures were recorded (K), percentage of neutrophils positive for citrullinated histone H3 were analyzed from the peritoneum by cytofluorimetry (L), and plasma IL-6 concentration was measured by ELISA (M). Each dot represents a mouse. “Saline” are WT mice not challenged with LPS used as control. Two-way ANOVA (A), and one-way ANOVA (D–M) were used for statistics. n.s., not significant (p > 0.05), *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. In (D–M) gray stars represent statistics compared with saline group. See also Figure S7.