SEARCH

SEARCH BY CITATION

Keywords:

  • GDF-15;
  • integrin;
  • platelets;
  • thrombosis

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Sources of funding
  9. Disclosure of Conflict of Interest
  10. References

Background

Integrin-mediated platelet function plays an important role in primary hemostasis. Growth-differentiation factor 15 (GDF-15) has been shown to inhibit β2-integrin activation in leukocytes.

Methods

We investigated the effect of GDF-15 on platelet integrin activation in vitro and in different in vivo models of thrombus formation.

Results

GDF-15-/- mice showed an accelerated thrombus formation and a reduced survival rate after collagen-induced pulmonary thromboembolism. In reconstitution experiments, recombinant GDF-15 decelerated thrombus formation and prolonged the bleeding time. In vitro experiments demonstrated that GDF-15 pretreated, agonist-stimulated platelets showed decreased binding to fibrinogen in flow chamber assays and reduced activation of β1- and β3-integrins in flow cytometry experiments. Pretreating human and mouse platelets with GDF-15 reduced platelet aggregation. Mechanistically, GDF-15 prevents agonist-induced Rap1- dependent αIIbβ3 activation by activating PKA. Platelet P-selectin expression and dense granule secretion after stimulation were unaffected by GDF-15, indicating a specific effect of GDF-15 on integrin activation.

Conclusion

GDF-15 specifically inhibits platelet integrin activation. These findings may have profound clinical implications for the treatment of hemostatic conditions involving platelets.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Sources of funding
  9. Disclosure of Conflict of Interest
  10. References

Controlling integrin activation is essential for all cells. Integrin-mediated platelet function is important for sealing injured blood vessels and preventing blood loss, and excessive platelet adhesion and aggregation can initiate arterial thrombosis, causing a stroke and heart attacks. At sites of injury, the platelet receptors GPIb and GPVI bind to von Willebrand factor (VWF) and collagen, respectively [1, 2]. This interaction, together with locally produced thrombin, triggers the activation of the platelet-specific integrin αIIbβ3 [platelet glycoprotein (GP) IIb/IIIa] and the release of soluble platelet agonists including thromboxane A2 (TxA2) and ADP. Activated αIIbβ3 binds VWF, fibrinogen and fibronectin, thus allowing platelet adhesion, spreading and aggregation.

Regulation of the affinity of the platelet integrin αIIbβ3 for adhesive ligands is central to the control of platelet aggregation [3]. Under physiological conditions, the integrin αIIbβ3 rests in a low-affinity state on the platelet membrane. The initiation of an intracellular signaling cascade induced by agonists leads to an increased affinity of αIIbβ3, often referred to as ’activation‘ [4]. Agents that block signaling through the generation of thromboxane A2 (e.g. aspirin) or through ADP receptors (e.g. clopidogrel) are moderately effective in the chronic prevention of arterial thrombosis [5], suggesting that specific inhibition of αIIbβ3 activation might be a useful antithrombotic strategy. However, only little is known about inhibition of platelet integrin activation. A recently published study showed that the endogenous cytokine growth-differentiation factor 15 (GDF-15), a member of the transforming growth factor (TGF)-β super family, directly inhibits neutrophil adhesion and arrest under flow and neutrophil transmigration into inflamed tissue [6]. Mechanistically, GDF-15 counteracted chemokine-triggered conformational activation and clustering of the β2-integrin αLβ2 (LFA-1) on neutrophils. GDF-15 is the first cytokine identified as an inhibitor of neutrophil recruitment by direct interference with chemokine-stimulated activation of leukocyte integrins [6].

Several compounds have been developed for specific blockade of ligand binding to αIIbβ3 leading to near complete inhibition of aggregation. These compounds have been proven to be effective for the prevention and treatment of arterial thrombosis in acute settings of percutaneous coronary intervention [7]. In spite of the compelling mechanistic rationale for the blockade of αIIbβ3 by agents, the chronic administration of oral αIIbβ3 antagonists has not been proven to be beneficial in preventing recurrent thrombotic events [8]. One plausible explanation for this negative result is the relatively narrow therapeutic window for these agents because pathological bleeding associated with complete loss of αIIbβ3 function necessitates maintenance of less than maximal blockade of the integrin [7]. Although, blockade of ligand binding to αIIbβ3 is possible, a direct and specific inhibitor of platelet integrin activation is not known.

In the present study, we investigated the effects of GDF-15 on agonist-induced platelet integrin activation using different in vivo models of hemostasis. In order to investigate the molecular mechanism of altered hemostasis by GDF-15, flow chamber assays, reporter antibody binding experiments and biochemistry assays were employed. Our in vivo and in vitro data show that GDF-15 prevents platelet-mediated thrombus formation by specifically inhibiting β1- and β3-integrins on platelets.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Sources of funding
  9. Disclosure of Conflict of Interest
  10. References

Reagents and animals

Unlike otherwise stated, all reagents were obtained from Sigma-Aldrich (Taufkirchen, Germany). We used 8- to 12-week-old C57Bl/6 wild-type (WT) mice (Charles River Laboratories, Wilmington, MA, USA) and GDF-15-/- mice [9]. Mouse colonies were maintained under specific pathogen-free conditions. All animal experiments were approved by local government authorities and were in agreement with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Murine model of acute pulmonary thromboembolism

A pulmonary thromboembolism was performed as described previously [10], with some modifications. Mice were anesthetized and received 1.6 mg kg−1 collagen (BD Biosciences, Franklin Lakes, NJ, USA) and 60 mg kg−1 epinephrine (Sanofi Aventis, Frankfurt, Germany) via tail vein injection. Some mice received 4 μg GDF-15 (R&D Systems, Minneapolis, MN, USA) 15 min prior to induction of thromboembolism. Animals were monitored and death was recorded as cessation of breathing. The experiment was terminated after 30 min and the surviving animals were euthanized. The lungs were fixed with 4% paraformaldehyde, removed and embedded in paraffin for histological hematoxylin and eosin staining.

Bleeding time

The bleeding time was measured as described previously [11]. Briefly, WT mice pretreated with 4 μg GDF-15 or saline 15 min before the experiments were anesthetized and placed on a heating pad to maintain a constant body temperature of 37 °C. A 3-mm piece of the tail-tip was cut off with a sharp scalpel. The tail was immersed into physiologic saline solution preheated to 37 °C and the time until cessation of bleeding for more than 5 s was recorded.

Ferric chloride-induced mesenteric thrombosis

Ferric chloride-induced injury was performed as described previously [12]. Briefly, after induction of anesthesia, mice received an intravenous injection of DIOC6 (3,3′-dihexyloxacarbocyanine iodide, 1.25 μm kg−1) to fluorescently label platelets and facilitate the visualization of in vivo thrombus formation. Mice were placed on a heating pad and a ventral midline incision was performed followed by gentle exteriorization of the mesentery. A single arteriole was selected (diameter between 120 and 160 μm). A strip of filter paper was soaked with 10% aqueous ferric chloride solution and placed perpendicular to the vessel. After 5 min, the filter paper was removed and the mesentery was constantly rinsed with preheated, physiologic saline solution. Thrombus formation was observed using a 20/0.5 objective on an AxioScope microscope (Carl Zeiss, Jena, Germany). Digital images were recorded using a digital camera (Sensicam QE, Cooke Corporation, Romulus, MI, USA). For some experiments, WT or GDF-15-/- mice were injected with 1 or 4 μg GDF-15 and/or Tirofiban (Iroko Cardio, London, UK) 15 min prior to experiments.

Flow cytometric analysis of β1- and β3-integrin activation and P-selectin expression

Citrate-anticoagulated whole blood samples were withdrawn from wild-type mice. Platelet-rich plasma was separated by centrifugation at 68 × g. Isolated platelets were resuspended in Tyrode's buffer and incubated with or without GDF-15 (20 ng mL−1) at 37 °C for 20 min prior to experiments. Platelets were stimulated with 10 μm ADP, 0.1 U mL−1 Thrombin or 3 μm U46619 in the presence of a reporter antibody specific for the high-affinity conformation of the mouse integrin αIIbβ3 (clone JON/A-PE, Emfret Analytics, Würzburg, Germany), a reporter antibody recognizing specifically the activated form of β1-integrins (clone 9EG7), or a P-selectin antibody (clone RB40.34) at 37 °C for 5 min. The reaction was stopped with the addition of 400 μl phosphate-buffered saline (PBS) and the samples were analyzed on a FACSCan to flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA) immediately. FACS data were processed using FlowJo (version 7.5.5; Tree Star Inc., Ashland, OR, USA).

In vitro platelet flow chamber

In order to investigate the adhesion of activates platelets under flow conditions we developed a microflow chamber assay. Rectangular glass capillaries (20 × 200 μm) were coated with fibrinogen (1500 μg mL−1) for 2 h followed by blocking of unspecific binding sites with casein 1% (Thermo Fisher Scientific, Waltham, MA, USA) for 1 h. For some experiments, flow chambers were coated with VWF (10 μg mL−1; Loxo, Dossenheim, Germany) or collagen (200 μg mL−1; BD Biosciences) for 2 h. One end of the glass capillary was connected to a PE50 tubing (BD Biosciences) and used to control the wall shear stress in the capillary as described previously [13, 14]. Wall shear stress was adjusted to ~20 dyne cm−2 resembling arterial flow conditions. Anticoagulated whole blood samples were withdrawn from WT mice and incubated at 37 °C for 20 min in the presence or absence of GDF-15 (20 ng mL−1) and/or Tirofiban (100 μm) prior to experiments. Platelets were stimulated with 10 μm ADP, 0.1 U mL−1 Thrombin or 3 μm U46619 solved in DMSO at 37 °C for 5 min. The chamber was perfused for 2 min and washed with PBS for 1 min. Representative fields of view were recorded using an SW40/0.75 objective and a digital camera. The numbers of adherent platelets per mm2 were calculated.

Platelet spreading

Circular cover slips were coated with fibrinogen (200 μg mL−1) overnight, followed by blocking with 1% BSA/PBS for 1 h. Isolated murine platelets (3 × 106 in 100 μL) were incubated with 0.1, 1 or 5 μg mL−1 convulxin (Santa Cruz Biotechnology, Santa Cruz, CA, USA) and allowed to spread for 30 min before fixation with 4% PFA. Platelets spreading was visualized using a Zeiss AxioImager (Zeiss, Oberkochen, Germany) using a 63 × /1.4 oil immersion objective.

Aggregometry

Platelet aggregation in murine and human platelet-rich plasma (PRP) was measured using a lumi-aggregometer (Model 700; Chrono-Log, Havertown, PA, USA) according to the manufacturer's instructions.

Rap1-activation assay

For biochemical assays, anticoagulated whole blood samples were withdrawn from WT mice. PRP was separated by low speed centrifugation at 68 × g. Isolated platelets were resuspended in Tyrode's buffer and incubated with or without GDF-15 (20 ng mL−1) at 37 °C for 20 min prior to experiments. Platelets were stimulated with 10 μm ADP, 0.1 U mL−1 Thrombin or 3 μm U46619 at 37 °C for 5 min and immediately lysed with EDTA-free ice-cold lysis buffer [15]. The Rap1 pull-down assay was performed as described previously [16].

Statistics

Statistical analysis was performed with PASW (version 18.0, SPSS Inc., Chicago, IL, USA) using one-way anova, the Student–Newman–Keuls test, post hoc correction or t-test where appropriate. Differences in survival between groups were detected by chi-square tests. All data are represented as means ± standard error of the mean (SEM). A P-value < 0.05 was taken as statistically significant.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Sources of funding
  9. Disclosure of Conflict of Interest
  10. References

GDF-15 decreases platelet aggregation in vitro

To address the role of GDF-15 on in vitro platelet aggregation, we performed aggregometry with PRP from WT mice and GDF-15-/- mice. After ADP (Fig. 1A) or U46619 (Fig. 1C) stimulation, aggregation was increased slightly in GDF-15-/- mice compared with WT mice. GDF-15 reduced platelet aggregation in a dose-dependent manner. The same effect on aggregation was observed in human PRP pretreated with GDF-15 at 10 or 20 ng mL−1 compared with the untreated control PRP (Fig. 1B+D).

image

Figure 1. Growth-differentiation factor 15 (GDF-15) decreases platelet aggregation in vitro. Aggregometry in murine and human PRP pretreated with GDF-15 at concentrations of 0, 10 or 20 ng mL−1 was analyzed after stimulation with 10 μm ADP (A+B) or 3 μm U46619 (C + D) (exemplary graph from three experiments).

Download figure to PowerPoint

GDF-15 reduces agonist-induced activation of β1- and β3-integrins on platelets

We evaluated the role of GDF-15 on agonist-induced activation of β1- and β3-integrins on platelets by employing reporter antibody binding and flow chamber assays. Direct assessment of αIIbβ3 activation in murine platelets stimulated with different agonists, using an antibody against the activated form of the receptor (JON/A-PE) [17], showed the same activation status of αIIbβ3 integrin on platelets in the presence or absence of GDF-15. Stimulation of WT platelets with ADP, thrombin or U46619 increased reporter antibody binding (Fig. 2A). Pretreating WT platelets with GDF-15 before stimulation with ADP, thrombin or U46619 significantly reduced reporter antibody binding (Fig. 2A). After incubating murine whole blood from WT mice with different agonists (ADP: 10 μm, thrombin: 0.1 U mL−1, thromboxane A2-analogue U46619: 3 μm) for 5 min, the number of adherent platelets in a fibrinogen-coated flow chamber under physiological flow conditions significantly increased compared with control or dimethylsulfoxide-treated blood (Fig. 2B). Pretreatment with GDF-15 significantly decreased the number of adherent platelets after stimulation (Fig. 2B). To directly show that the decreased platelet adhesion is as a result of a GDF-15-mediated impaired activation of αIIbβ3, we treated whole blood with a specific αIIbβ3 antagonist (Tirofiban: 100 μm) before stimulating the cells (Fig. 2B). Blocking of αIIbβ3 by Tirofiban reduced platelet adhesion to fibrinogen to a level seen after pretreatment with GDF-15 (Fig. 2B). Blocking αIIbβ3 by Tirofiban in the presence of GDF-15 had no additional effect on the number of adherent platelets, except for thrombin, suggesting that GDF-15 abolishes agonist-induced αIIbβ3 activation.

image

Figure 2. Growth-differentiation factor 15 (GDF-15) reduces agonist-induced activation of β1- and β3-integrins on platelets. GDF-15 reduces αIIbβ3 activation under static conditions (A) and impairs platelet adhesion to fibrinogen under flow conditions in murine (B) and human samples (C) (n = 3–4). Rap1 activation in stimulated murine (D) and human (E) platelets (exemplary blots from three experiments). (F) Activation of β1-integrins after thrombin stimulation under static conditions is reduced in presence of GDF-15 (n = 3–4). (G) P-selectin expression of isolated platelets after stimulation (n = 3). (H) Integrin αIIbβ3 activation in presence of Mn2+ and Tat-Rap1-wild-type (WT)/ constitutive-active (CA) peptides (n = 3). Binding of platelets to von Willebrand factor (VWF) (I) and collagen (J) under flow conditions (n = 3). (K) Platelet spreading on fibrinogen after stimulation with convulxin (n = 3). *P < 0.05.

Download figure to PowerPoint

In order to investigate the effect of GDF-15 on platelet integrin activation in the human system, flow chamber experiments with human samples were conducted. For the flow chamber assays, human whole blood was stimulated with ADP (10 μm), thrombin (0.1 U mL−1) or U46619 (3 μm) for 5 min at 37 °C. Pretreatment with GDF-15 significantly decreased the number of adherent human platelets after stimulation (Fig. 2C). Blocking of αIIbβ3 by Tirofiban reduced platelet adhesion to fibrinogen to a level seen after pretreatment with GDF-15 (Fig. 2C). Blocking αIIbβ3 by Tirofiban in the presence of GDF-15 had no additional effect on the number of adherent platelets, except for thrombin, suggesting that GDF-15 abolishes agonist-induced αIIbβ3 activation (Fig. 2C).

In order to reveal the molecular mechanism by which GDF-15 reduces αIIbβ3 activation, we investigated the activation of Rap1, a molecule which is known to be involved in agonist-induced αIIbβ3 activation [18]. Isolated murine platelets where pretreated with or without GDF-15 (20 ng mL−1) and consecutively stimulated with different agonists (ADP: 10 μm, thrombin: 0.1 U mL−1 and U46619: 3 μm). Rap1 activation was increased after stimulation with ADP, thrombin and U46619 compared with unstimulated control samples (Fig. 2D). In lysates from platelets pretreated with GDF-15 before stimulation, Rap1 activity was decreased (Fig. 2D), suggesting that GDF-15 reduces αIIbβ3 activation by inhibiting Rap1.

Additionally, we investigated the activation of Rap1 in isolated human platelets pretreated with or without GDF-15 (20 ng mL−1) and subsequently stimulated the cells with ADP (10 μm), thrombin (0.1 U mL−1) or U46619 (3 μm). In untreated platelets, Rap1 activation was increased after stimulation with ADP, thrombin and U46619 compared with unstimulated control samples (Fig. 2E). In lysates from platelets pretreated with GDF-15 before stimulation, Rap1 activity was decreased, suggesting that GDF-15 reduces αIIbβ3 activation by inhibiting Rap1 (Fig. 2E). Similar to static conditions, GDF-15 pretreatment also abolished agonist-induced activation of Rap1 in stirred conditions (data not shown).

Furthermore, agonist-stimulated GDF-15 pretreated platelets failed to bind the monoclonal anti-β1-integrin antibody 9EG7 (Fig. 2F), which specifically recognizes the activated form of β1-integrins [19].

Resting platelets store P-selectin in α-granules, which fuse with the plasma membrane after agonist-induced platelet activation [20]. ADP, thrombin and U46619 induced P-selectin translocation on untreated or GDF-15 pretreated platelets (Fig. 2G).

To investigate whether GDF-15 is an inhibitor of platelet dense granule secretion, we measured the ADP and ATP levels in PRP obtained from WT mice with and without GDF-15 pretreatment (20 ng mL−1, 37 °C, 20 min). No significant difference in the ADP release without stimulation was detected between the groups (data not shown). After stimulation with ADP (10 μm), thrombin (0.1 U mL−1) or U46619 (3 μm) we did not detect a difference in ATP release between the two groups.

To assess a possible alteration of surface adhesion molecule expression on platelets by GDF-15, we conducted flow cytometry with isolated platelets pretreated with and without GDF-15 under baseline conditions and after stimulation with ADP (10 μm), thrombin (0.1 U mL−1) or U46619 (3 μm). No significant differences in the surface expression of αIIbβ3, GPIb, GPIX or β1-integrin were observed between the groups (Table 1). However, the expression of the integrin αIIbβ3 increased after stimulation, although there was no significant difference between the untreated and the GDF-15-pretreated group. These data show that GDF-15 specifically disables the activation of β1- and β3-integrins without affecting other platelet functions.

Table 1. Growth-differentiation factor 15 (GDF-15) does not affect platelet integrin expression. Isolated human platelets were pretreated with or without GDF-15 (20 ng mL−1) at 37 °C for 20 min. The expression of β1-integrins, αIIbβ3, GPIb and GPIX on the platelet cell surface was quantified by flow cytometry using FITC- or PE-conjugated antibodies at baseline or after 5 min of stimulation with ADP (10 μm), thrombin (0.1 U mL−1) or U46619 (3 μm)
 ControlGDF-15P- value
BaselineADPThrombinU46619BaselineADPThrombinU46619
  1. The data are expressed as mean fluorescence intensity (MFI) ± standard error of the mean (SEM) (n = 3).

β13184 ± 763227 ± 833181 ± 1323203 ± 903114 ± 1443122 ± 853160 ± 783130 ± 950.999
αIIb β31554 ± 282111 ± 792104 ± 622132 ± 1081555 ± 72159 ± 172207 ± 522207 ± 20 < 0.001
GPIb9669 ± 979611 ± 1089790 ± 1069707 ± 599801 ± 1099774 ± 1299768 ± 1279798 ± 1290.680
GPIx1729 ± 641704 ± 261703 ± 331732 ± 371777 ± 301775 ± 341766 ± 321768 ± 310.887

To demonstrate that the effect of GDF-15 is mediated by inhibiting integrin activation, we pretreated stimulated isolated murine platelets with or without GDF-15 (20 ng mL−1) at 37 °C for 20 min prior to stimulation with ADP (10 μm). Furthermore, one sample was stimulated in the presence of Mn2+ (1 mM) and GDF-15. The activation of the integrin αIIbβ3 was assessed by binding of the reporter antibody JON/A-PE and quantified by flow cytometry. The inhibitory effect of GDF on αIIbβ3 activation after stimulation with ADP could be reversed in the presence of Mn2+ (Fig. 2H), indicating that GDF-15 primarily acts by inhibiting inside-out integrin activation.

To prove that Rap1 is directly involved in GDF-15-dependent inhibition of integrin activation, we incubated isolated murine platelets with a WT or constitutive-active (CA) Tat-fusion mutant of Rap1a (1 μm, 37 °C, 30 min) and GDF-15 (20 ng mL−1, 37 °C, 20 min), followed by stimulation with ADP (10 μm) at 37 °C for 5 min. Treatment with Tat-Rap1a-CA reversed the effect of GDF-15 on integrin αIIbβ3 activation, quantified as binding of the reporter antibody JON/A-PE in flow cytometry (Fig. 2H). This indicates that Rap1a is involved in the pathway leading to inhibition of integrin activation by GDF-15.

To assess the effect of GDF-15 on GPIb and platelet collagen receptor function, we conducted flow chamber experiments with unstimulated murine whole blood in flow chambers coated with either VWF (10 μg mL−1) or collagen (200 μg mL−1). There was no significant difference in the number of adherent platelets pretreated with or without GDF-15 on VWF (Fig. 2I) or collagen (Fig. 2J). These data suggest that GDF-15 does not affect the function of GPIb or platelet collagen receptors.

To investigate the role of GDF-15 in ITAM-mediated signaling events, we performed platelet spreading experiments with isolated murine platelets after stimulation with convulxin at concentrations of 0.1, 1 or 5 μg mL−1. After stimulation, platelets were allowed to adhere to fibrinogen-coated glass cover slips for 30 min. The stimulation with convulxin induced a dose-dependent increase in the percentage of fully spread platelets after 30 min (Fig. 2K). Pretreatment of the platelets with GDF-15 prior to stimulation did not significantly alter the percentage of spread platelets (Fig. 2K).

GDF-15 reduces agonist-induced activation of αIIbβ3 by activation of PKA

It is known that the activation of Rap1 in platelets can be decreased by Rap1GAP2 [21], which can be activated by the protein kinase A (PKA). To investigate whether GDF-15 pretreatment affects the activation of PKA, we investigated the serine and threonine phosphorylation of PKA substrates. GDF-15 pretreatment of platelets caused increased phosphorylation of PKA substrates, indicating the activation of PKA by GDF-15 (Fig. 3A). The overall platelet tyrosine phosphorylation was mostly unchanged, with exception of a single band with a size between 36 and 55 kDa (Fig. 3B). Inhibition of PKA by a specific inhibitor (H89) abolished the inhibitory effect of GDF-15 on Rap1 activation in pull-down assays after stimulation with ADP or U46619 (Fig. 3C). Functionally, the inhibition of PKA by H89 also diminished the inhibitory effect of GDF-15 on activation of the integrin αIIbβ3 after stimulation in flow cytometric experiments (Fig. 3D)

image

Figure 3. Growth-differentiation factor 15 (GDF-15) reduces agonist-induced activation of αIIbβ3 by activation of protein kinase A (PKA). (A) The phosphorylation of PKA substrates was investigated in platelet lysates after GDF-15 pretreatment. (B) Total tyrosine-phosphorylation in platelet lysates after GDF-15 pretreatment (exemplary blots of 3 independent experiments). Rap1 activation (C) and Integrin αIIbβ3 activation (D) after stimulation after pretreatment with GDF-15 and/or H89 (n = 3). *P < 0.05.

Download figure to PowerPoint

GDF-15 modulates platelet-dependent hemostasis in wildtype mice

A bleeding time model was used to assess the effect of GDF-15 on platelet-dependent hemostasis [11]. A prolonged bleeding time can be a result of defective platelet function or low platelet count. Indeed, β3-null mice have prolonged bleeding times similar to that observed in humans lacking αIIbβ3 expression [22]. The bleeding time was reduced in GDF-15-/- mice compared with WT mice. Pretreating WT mice with recombinant GDF-15 significantly prolonged the bleeding time (Fig. 4), suggesting that GDF-15 may modulate platelet-dependent hemostasis after tail transection in the mouse.

image

Figure 4. Growth-differentiation factor 15 (GDF-15) modulates platelet-dependent hemostasis under physiological conditions in wild-type (WT) mice. (A) Bleeding time in WT (WT) mice with or without GDF-15 pretreatment and GDF-15-/- mice was recorded as cessation of bleeding for more than 5 s (n = 13–14). *P < 0.05.

Download figure to PowerPoint

Collagen-induced pulmonary thromboembolism is increased in GDF-15-/- mice

As αIIbβ3 is required for platelet aggregation, we examined the effect of GDF-15 on the formation of occlusive platelet thrombi in a pulmonary thromboembolism model in which platelet activation was induced by tail vein infusion of epinephrine and collagen [10]. Upon intravenous injection of a mixture of these platelet agonists, 30% of WT mice (6 of 20 mice) died within 30 min (Fig. 5A). Histological examination of lung tissue from these animals revealed platelet thrombi throughout the pulmonary vasculature (Fig. 5B). In contrast, 73% of GDF-15-/- mice (14 of 22 mice) died after the infusion of epinephrine and collagen (< 0.05; Fig. 5A). The lungs of GDF-15-/- mice showed an increased number of platelet thrombi (Fig. 5B). We conducted the acute pulmonary thrombembolism model in WT mice pretreated with GDF-15 (4 μg/mouse). Almost all WT mice pretreated with GDF-15 survived 30 min after the induction of acute pulmonary thrombembolism (Fig. 5A).

image

Figure 5. Collagen-induced pulmonary thromboembolism is increased in GDF-15-/- mice. Acute pulmonary thromboembolism was induced by intravenous injection of collagen and epinephrine. (A) Percentage of surviving wild-type (WT) mice, GDF-15-/- mice and WT mice pretreated with GDF-15 after 30 min (n = 15–22). (B) Representative pictures of hematoxylin and eosin (H&E) stained lung sections from WT and GDF-15-/- animals. Arrows indicate microthrombi. Scale bars equal 100 μm.

Download figure to PowerPoint

GDF-15 decreases FeCl3-induced arterial thrombus formation

To examine the effects of GDF-15 in a model of arterial thrombosis initiated by vascular injury, we measured the time course of thrombus formation and vessel occlusion in response to ferric chloride-induced injury of the mesenteric artery [12]. Time to thrombus formation (defined by occurrence of a thrombus > 30 μm) and vessel occlusion were significantly prolonged in WT mice pretreated with recombinant GDF-15 compared with control mice receiving saline (Fig. 6A-C). Pretreating WT mice with GDF-15 and Tirofiban did not further increase the time to thrombus formation and vessel occlusion, suggesting that GDF-15 exerts its antithrombotic effect by affecting the integrin αIIbβ3. GDF-15-/- mice exhibited a significantly accelerated time to thrombus formation and vessel occlusion compared with WT mice (Fig. 6A and B). Reconstitution of recombinant GDF-15 in GDF-15-/- mice decelerated the time to thrombus formation and vessel occlusion (Fig. 6A and B), whereas these effects were dose dependent in GDF-15-/- mice. The diameter of the recorded vessels did not significantly differ between the groups (data not shown).

image

Figure 6. Growth-differentiation factor 15 (GDF-15) decreases FeCl3-induced arterial thrombus formation in vivo. Mesenteric thrombosis was induced in wild-type (WT) and GDF-15-/- mice by topical application of FeCl3. (A) Time to thrombus formation (TF) in WT mice and GDF-15-/- mice (n = 4). (B) Time to vessel occlusion (VO) in WT and GDF-15-/- mice (n = 4). (C) Representative brightfield and epifluorescence images. *P < 0.05.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Sources of funding
  9. Disclosure of Conflict of Interest
  10. References

This study shows that GDF-15 prevents platelet integrin activation and platelet aggregation. The absence of GDF-15 in knockout mice leads to an increased mortality in a pulmonary thromboembolism model and accelerates thrombus formation after FeCl3-induced injury of mesenteric arterioles. GDF-15 abolishes the agonist-induced conformational change of β1- and β3-integrins on platelets to its high affinity state. Mechanistically, GDF-15 reduces agonist-induced Rap1 activation in platelets. Pretreating human and mouse platelets with GDF-15 reduced platelet aggregation.

Thrombus formation requires the activation of the platelet integrin αIIbβ3 [23]. αIIbβ3 deficiency or blocking this molecule leads to complete inhibition of platelet aggregation and adherence to sites of vascular injury [20, 22]. In humans, integrin αIIbβ3 deficiency causes Glanzmann thrombasthenia, which is characterized by a severe bleeding tendency and defective platelet aggregation [25]. In this study, we have shown that GDF-15 potentially inhibits agonist-induced αIIbβ3 activation, which prevents platelet-mediated thrombosis formation in vivo. In contrast to αIIbβ3 deficiency, where spontaneous bleeding can occur [25], we did not observe spontaneous bleeding at the site of the surgical procedure (preparation of the M. cremaster) within 3 h after the application of recombinant GDF-15, suggesting that GDF-15 is potent enough to block αIIbβ3 activation and prevent thrombus formation under pathological conditions, but does not induce spontaneous bleeding. One possibility how this observation can be explained is that GDF-15 does not completely inhibit αIIbβ3 activation and this is in accordance with our data (Fig. 2B) which show that GDF-15 does not completely inhibit thrombin-induced αIIbβ3 activation. In myeloid cells, we have shown that GDF-15 inhibits chemokine-triggered β2-integrin activation by inhibiting the small GTPase Rap1, leading to the inhibition of conformational activation and clustering of β2-integrins [6]. Rap1 is also known as an essential signaling molecule involved in integrin activation in platelets after activation with several pro-thrombotic stimuli, including ADP, thrombin, thromboxane A2 and collagen [26, 27]. However, αIIbβ3 activation after thrombin stimulation is mediated by a Rap1-dependent and -independent pathway [26]. The partial inhibition of αIIbβ3 activation after thrombin stimulation may be explained by the fact that GDF-15 only inhibits Rap1 activation without affecting the Rap1-independent signaling pathway. It is still unknown whether blocking Rap1 has the same phenotype such as blocking kindlin-3 or talin-1. The authors would assume that blocking Rap1 does result in such a strong phenotype like blocking kindlin-3 or talin-1, because Rap1 is located upstream of kindlin-3 or talin-1 and kindlin-3 or talin-1 may also be activated by other molecules. It will be interesting and important to reveal further details of the GDF15-driven signaling pathway in the future.

Intravenously administered pharmacological blockers of αIIbβ3 (such as eptifibatide and tirofiban) have been proven to be effective in the secondary prevention of ischemic diseases [28]. In contrast, orally administered αIIbβ3-antagonist failed in clinical trials owing to increased overall mortality, possibly as a result of dosing difficulties and paradoxical receptor activation by induction of a ligand-bound conformation of αIIbβ3 together with clustering and platelet prestimulation [8, 29].

In order to extend our findings, we also investigated agonist-induced activation of β1-integrins on platelets. In accordance with our previously published study, where we demonstrated that GDF-15 reduces chemokine-induced β1-dependent adhesion of monocytes under flow [6], we here directly show that GDF-15 prevents the agonist-induced conformational change of β1-integrins, which can be detected by the binding of the reporter antibody 9EG7. GDF-15 inhibits agonist-induced integrin activation in platelets, neutrophils [6], monocytes [6] and lymphocytes (unpublished data, personal communication, D. Vestweber), suggesting an important role of GDF-15 for different hematopoietic cell types. Importantly, GDF-15 specifically prevents agonist-induced activation of β1-, β2 - and β3- integrins without affecting other cell functions, such as integrin-mediated outside-in signaling and chemokine induced Ca++-transients in neutrophils [6] or platelet activation as monitored by P-selectin translocation to the plasma membrane (Fig. 2G). Thus, this cytokine may provide a valuable therapeutic approach for preventing and treating thrombus formation and inflammation by targeting specifically the activation of integrins. However, further research on the pharmacokinetic and dynamic effects of GDF-15 as well as studies emphasizing on the safety of inhibition of integrin inhibition in various cell types in vivo is necessary to gauge the value of GDF-15 for the treatment of thrombotic and inflammatory diseases compared to established treatment alternatives.

In conclusion, our results indicate a crucial role for GDF-15 in preventing agonist-induced platelet integrin activation leading to diminished platelet in response to pathologic stimuli. The newly identified pathophysiological role of GDF-15 in regulating the activation state of β1- and β3-integrins on platelets may have profound clinical implications and a promising potential for the development of new treatment strategies and warrants further research within this field.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Sources of funding
  9. Disclosure of Conflict of Interest
  10. References

The authors are grateful to Helena Block and Dagmar Rademaekers for help and technical advice. The authors thank Carlo Laudanna for providing the Tat-peptides.

Sources of funding

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Sources of funding
  9. Disclosure of Conflict of Interest
  10. References

This study was supported by grants from the German Research Foundation (AZ 428/3-1, AZ 428/6-1, AZ 428/8-1 to A.Z.), Else Kröner-Fresenius-Stiftung (grant A69/07 to A.Z.), Innovative Medizinische Forschung (IMF Münster, Germany, A.Z.) and the Interdisciplinary Clinical Research Center (IZKF Münster, Germany, SEED01/12 to J.R.)

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Sources of funding
  9. Disclosure of Conflict of Interest
  10. References
  • 1
    Savage B, Almus-Jacobs F, Ruggeri ZM. Specific synergy of multiple substrate-receptor interactions in platelet thrombus formation under flow. Cell 1998; 94: 65766.
  • 2
    Nieswandt B, Watson SP. Platelet-collagen interaction: is GPVI the central receptor? Blood 2003; 102: 44961.
  • 3
    Shattil SJ, Newman PJ. Integrins: dynamic scaffolds for adhesion and signaling in platelets. Blood 2004; 104: 160615.
  • 4
    Ginsberg MH, Partridge A, Shattil SJ. Integrin regulation. Curr Opin Cell Biol 2005; 17: 50916.
  • 5
    Bhatt DL, Topol EJ. Scientific and therapeutic advances in antiplatelet therapy. Nat Rev Drug Discovery 2003; 2: 1528.
  • 6
    Kempf T, Zarbock A, Widera C, Butz S, Stadtmann A, Rossaint J, Bolomini-Vittori M, Korf-Klingebiel M, Napp LC, Hansen B, Kanwischer A, Bavendiek U, Beutel G, Hapke M, Sauer MG, Laudanna C, Hogg N, Vestweber D, Wollert KC. GDF-15 is an inhibitor of leukocyte integrin activation required for survival after myocardial infarction in mice. Nat Med 2011; 17: 5818.
  • 7
    Quinn MJ, Byzova TV, Qin J, Topol EJ, Plow EF. Integrin alphaIIbbeta3 and its antagonism. Arterioscler Thromb Vasc Biol 2003; 23: 94552.
  • 8
    Chew DP, Bhatt DL, Sapp S, Topol EJ. Increased mortality with oral platelet glycoprotein IIb/IIIa antagonists: a meta-analysis of phase III multicenter randomized trials. Circulation 2001; 103: 2016.
  • 9
    Hsiao EC, Koniaris LG, Zimmers-Koniaris T, Sebald SM, Huynh TV, Lee SJ. Characterization of growth-differentiation factor 15, a transforming growth factor beta superfamily member induced following liver injury. Mol Cell Biol 2000; 20: 374251.
  • 10
    Petrich BG, Fogelstrand P, Partridge AW, Yousefi N, Ablooglu AJ, Shattil SJ, Ginsberg MH. The antithrombotic potential of selective blockade of talin-dependent integrin alpha IIb beta 3 (platelet GPIIb-IIIa) activation. J Clin Invest 2007; 117: 22509.
  • 11
    Zarbock A, Bishop J, Muller H, Schmolke M, Buschmann K, van AH, Singbartl K. Chemokine homeostasis vs. chemokine presentation during severe acute lung injury: the other side of the Duffy antigen receptor for chemokines. Am J Physiol Lung Cell Mol Physiol 2010; 298: L46271.
  • 12
    Marx I, Christophe OD, Lenting PJ, Rupin A, Vallez MO, Verbeuren TJ, Denis CV. Altered thrombus formation in von Willebrand factor-deficient mice expressing von Willebrand factor variants with defective binding to collagen or GPIIbIIIa. Blood 2008; 112: 6039.
  • 13
    Kuwano Y, Spelten O, Zhang H, Ley K, Zarbock A. Rolling on E- or P-selectin induces the extended but not high-affinity conformation of LFA-1 in neutrophils. Blood 2010; 116: 61724.
  • 14
    Zarbock A, Lowell CA, Ley K. Spleen tyrosine kinase Syk is necessary for E-selectin-induced alpha(L)beta(2) integrin-mediated rolling on intercellular adhesion molecule-1. Immunity 2007; 26: 77383.
  • 15
    Zarbock A, Abram CL, Hundt M, Altman A, Lowell CA, Ley K. PSGL-1 engagement by E-selectin signals through Src kinase Fgr and ITAM adapters DAP12 and FcR gamma to induce slow leukocyte rolling. J Exp Med 2008; 205: 233947.
  • 16
    Stadtmann A, Brinkhaus L, Mueller H, Rossaint J, Bolomini-Vittori M, Bergmeier W, Van AH, Wagner DD, Laudanna C, Ley K, Zarbock A. Rap1a activation by CalDAG-GEFI and p38 MAPK is involved in E-selectin-dependent slow leukocyte rolling. Eur J Immunol 2011; 41: 207485.
  • 17
    Lenter M, Uhlig H, Hamann A, Imhof B, Vestweber A. A monoclonal antibody against an activation epitope on mouse integrin chain beta 1 blocks adhesion of lymphocytes to the endothelial integrin alpha 6 beta1. Proc Natl Acad Sci U S A. 1993; 90: 90515.
  • 18
    de Bruyn K, Zwartkruis FJ, de Rooij J, Akkerman JW, Bos JL. The small GTPase Rap1 is activated by turbulence and is involved in integrin [alpha]IIb[beta]3-mediated cell adhesion in human megakaryocytes. J biol chem 2003; 278: 2241217.
  • 19
    Bergmeier W, Schulte V, Brockhoff G, Bier U, Zirngibl H, Nieswandt B. Flow cytometric detection of activated mouse integrin alphaIIbbeta3 with a novel monoclonal antibody. Cytometry 2002; 48: 806.
  • 20
    Ruggeri ZM. Platelets in atherothrombosis. Nat Med 2002; 8: 122734.
  • 21
    Hoffmeister M, Riha P, Neumuller O, Danielewski O, Schultess J, Smolenski AP. Cyclic nucleotide-dependent protein kinases inhibit binding of 14-3-3 to the GTPase-activating protein Rap1GAP2 in platelets. J biol chem 2008; 283: 2297306.
  • 22
    Hodivala-Dilke KM, McHugh KP, Tsakiris DA, Rayburn H, Crowley D, Ullman-Cullere M, Ross FP, Coller BS, Teitelbaum S, Hynes RO. Beta3-integrin-deficient mice are a model for Glanzmann thrombasthenia showing placental defects and reduced survival. J Clin Investig 1999; 103: 22938.
  • 23
    Shattil SJ. Function and regulation of the beta 3 integrins in hemostasis and vascular biology. Thromb Haemost 1995; 74: 14955.
  • 24
    Ni H, Denis CV, Subbarao S, Degen JL, Sato TN, Hynes RO, Wagner DD. Persistence of platelet thrombus formation in arterioles of mice lacking both von Willebrand factor and fibrinogen. J Clin Investig 2000; 106: 38592.
  • 25
    Nurden AT. Glanzmann thrombasthenia. Orphanet J Rare Dis 2006; 1: 10.
  • 26
    Crittenden JR, Bergmeier W, Zhang Y, Piffath CL, Liang Y, Wagner DD, Housman DE, Graybiel AM. CalDAG-GEFI integrates signaling for platelet aggregation and thrombus formation. Nat Med 2004; 10: 9826.
  • 27
    Bergmeier W, Goerge T, Wang HW, Crittenden JR, Baldwin AC, Cifuni SM, Housman DE, Graybiel AM, Wagner DD. Mice lacking the signaling molecule CalDAG-GEFI represent a model for leukocyte adhesion deficiency type III. J Clin Investig 2007; 117: 1699707.
  • 28
    Kristensen SD, Wurtz M, Grove EL, de Caterina R, Huber K, Moliterno DJ, Neumann FJ. Contemporary use of glycoprotein IIb/IIIa inhibitors. Thromb Haemost 2012; 107: 21524.
  • 29
    Bassler N, Loeffler C, Mangin P, Yuan Y, Schwarz M, Hagemeyer CE, Eisenhardt SU, Ahrens I, Bode C, Jackson SP, Peter K. A mechanistic model for paradoxical platelet activation by ligand-mimetic alphaIIb beta3 (GPIIb/IIIa) antagonists. Arterioscler Thromb Vasc Biol 2007; 27: e915.