Skip main navigation

Heterogeneity of Platelet Functional Alterations in Patients With Filamin A Mutations

Originally publishedhttps://doi.org/10.1161/ATVBAHA.112.300603Arteriosclerosis, Thrombosis, and Vascular Biology. 2013;33:e11–e18

Abstract

Objective—

We examined platelet functions in 4 unrelated patients with filaminopathy A caused by dominant mutations of the X-linked filamin A (FLNA) gene.

Methods and Results—

Patients P1, P2, and P4 exhibited periventricular nodular heterotopia, heterozygozity for truncating FLNA mutations, and thrombocytopenia (except P2). P3 exhibited isolated thrombocytopenia and heterozygozity for a p.Glu1803Lys FLNA mutation. Truncated FLNa was undetectable by Western blotting of P1, P2, and P4 platelets, but full-length FLNa was detected at 37%, 82%, and 57% of control, respectively. P3 FLNa (p.Glu1803Lys and full-length) was assessed at 79%. All patients exhibited a platelet subpopulation negative for FLNa. Platelet aggregation, secretion, glycoprotein VI signaling, and thrombus growth on collagen were decreased for P1, P3, and P4, but normal for P2. For the 2 patients analyzed (P1 and P4), spreading was enhanced and, more markedly, in FLNa-negative platelets, suggesting that FLNa negatively regulates cytoskeleton reorganization. Platelet adhesion to von Willebrand factor under flow correlated with platelet full-length FLNa content: markedly reduced for P1 and P4 and unchanged for P2. Interestingly, von Willebrand factor flow adhesion was increased for P3, consistent with a gain-of-function effect enhancing glycoprotein Ib-IX-V/von Willebrand factor interaction. These results are consistent with a positive role for FLNa in platelet adhesion under high shear.

Conclusion—

FLNA mutation heterogeneity correlates with different platelet functional impacts and points to opposite regulatory roles of FLNa in spreading and flow adhesion under shear.

Introduction

Filamins (FLN) are large dimeric actin-binding proteins that stabilize actin filament networks with which they connect to the cellular membrane.1 FLN also bind to a large number of signaling proteins (>70).2 The FLN family consists of 3 dimeric proteins (FLNa, FLNb, and FLNc) produced by distinct genes. FLNa, the most abundant isoform, is encoded by FLNA located on chromosome Xq283 and is composed of an N-terminal actin-binding domain followed by 24 Ig-like repeats and the C-terminal domain that mediates the dimerization.4

FLNA mutations produce a wide spectrum of rare developmental diseases and cause various malformations of brain, heart, and muscle.5 The most frequent phenotype is periventricular nodular heterotopia (FLNA-PVNH), which can be associated with other features, including thrombocytopenia, patent ductus arteriosus, and Ehlers-Danlos syndrome. Skeletal dysplasia, including the oto-palato-digital syndrome spectrum disorders and terminal osseous dysplasia, familial cardiac valvular dystrophy, and congenital intestinal pseudo-obstruction, has also been described.6,7

Platelets express predominantly FLNa. FlnAloxP GATA1-Cre mice that exclusively lack FLNa in platelets are characterized by a macrothrombocytopenia and increased tail bleeding time.8 A thrombopathy was found to alter α-granule secretion, integrin αIIbβ3 activation, and signaling through the collagen receptor glycoprotein VI (GPVI) and the C-type lectin-like receptor 2. In human platelets, FLNa was shown to interact with glycoprotein Ibα (GPIbα) through the Ig-like repeat 17,9 the principal adhesion receptor of von Willebrand factor (VWF). The association of FLNa with GPIbα requires amino acids 563–571 localized in the cytoplasmic tail of GPIbα.10 Using transgenic mice expressing human GPIbα mutated at p.Phe568 and p.Trp570, a recent report showed that FLNa plays an important role in platelet adhesion and plasma membrane stability under pathological shear rates.11 Finally, the coordinated expression of GPIbα and FLNa seems to be essential for the production of normal-size platelets.12

Because hemorrhage and thrombocytopenia have been reported in patients with FLNa-PVNH,6,13 we have recently defined the platelet structural characteristics of 2 patients displaying this phenotype (P1 and P2) and of 1 patient (P3) exhibiting an isolated macrothrombocytopenia caused by a novel FLNA mutation. Anisocytosis with both giant and normal platelets was observed in all 3 patients, as well as an abnormal platelet production.14

Filaminopathy A thus seems as a unique way to examine the role of FLNa in human platelet functions. We now report 4 patients exhibiting FLNA mutations and altered platelet functions. The first 3 patients are the same as in our first study, a fourth patient with FLNA-PVNH being added (P4). For the 3 patients in whom the FLNa mutation is predicted to be truncating, we show that platelets exhibit wild-type FLNa at variable levels and that truncated FLNa is undetectable. Abnormal responses to collagen, including aggregation, secretion, the GPVI signaling pathway, and thrombus formation under flow on collagen, were observed in 3 patients with thrombocytopenia. Similarly, spreading was increased especially for platelets without FLNa, whatever the matrix type. Finally, we found that altered adhesion and altered thrombus growth at low- and high-shear rates paralleled the levels of wild-type FLNa. We thus conclude that in patients with FLNa-truncated molecules, the platelet functional defects depend essentially on the levels of the remaining full-length FLNa. In addition, the mutation for the fourth patient appears as a gain-of-function mutation. Our study thus confirms a central role for FLNa in platelet-adhesive functions.

Materials and Methods

Patients and Platelet Preparation

Four female patients with FLNA mutations were enrolled in this study after informed consent, and blood samples were obtained in accordance with the Declaration of Helsinki. Ethical approval was obtained from INSERM N° RBM 01-14 for the project Network on the inherited diseases of platelet function and platelet production. For P1, with PVNH and a low platelet count (80×109 platelets/L), a heterozygous frameshift point mutation identified in exon 27 of FLNA associated with PVNH was detected.15 For P2 who is affected with PVNH–Ehlers-Danlos syndrome, an heterozygous intragenic deletion involving FLNA exons 31 to 48 was observed.14 Her platelet count was normal (238×109 platelets/L). For P3, a heterozygous missense p.Glu1803Lys mutation affecting a highly conserved glutamic acid residue localized in Ig-like repeat 16 was detected14 and was associated with an isolated low platelet count (40–60×109 platelets/L). P4 is a 12-year-old teenager affected with PVNH associated with a moderately low platelet count (110×109 platelets/L). Gene dosage analysis using array comparative genomic hybridization revealed a heterozygous intragenic deletion involving FLNA exons 20 to 48 (Figure I in the online-only Data Supplement).16 All patients underwent analysis for the skewing of X inactivation in lymphocytes: none of them had skewed X inactivation (>70/30; Table I in the online-only Data Supplement).

Platelet Preparation

See Methods in the online-only Data Supplement.

Platelet Aggregation

Platelet aggregation of washed platelets was monitored by measuring light transmission through the stirred suspension of platelets using a Chronolog aggregometer (Coultronics, Margency, France). See Methods in the online-only Data Supplement.

Platelet Dense Granule Secretion

Dense granule secretion was quantified by measuring ATP release during platelet aggregation using a luminometer (Fluostar Optima; BMG Labtech). See Methods in the online-only Data Supplement.

Platelet Spreading

Spreading of platelets was performed as described in Methods in the online-only Data Supplement.

Immunoblotting

Immunoblotting was performed as described in Methods in the online-only Data Supplement.

Thrombus Formation Under Flow

Blood perfusion experiments were as described in Methods in the online-only Data Supplement.

Statistical Analyses

Results were analyzed using the Student t test or 1-way ANOVA followed by a least significant difference multiple comparisons as indicated.

Results

Only full-length FLNa is detected in platelets of patients with truncating mutations (P1, P2, P4). We first quantified the level of full-length and mutated FLNa in platelets for all patients. Because the patients are heterozygous for FLNA mutations, their platelets may harbor both the full-length and the mutated forms of FLNa. Because P1, P2, and P4 mutations are putatively truncating for FLNa (Figure I in the online-only Data Supplement), a search for such truncated FLNa was initiated by Western blotting. No polypeptides of the sizes predicted for truncated FLNa were detected in P1 (162 kDa), P2 (184 kDa), or P4 (104 kDa) platelets (Figure 1A). In contrast, full-length 280 kDa FLNa was detected in all 3 patients, at different levels, compared with control: 37% for P1, 57% for P4, and 82% for P2 (Figure 1B). Polypeptides below normal FLNa are likely to correspond to degradation products of normal FLNa, because the same patterns are observed for all patients. For P3, who carries the nontruncating missense p.Glu1803Lys mutation, a 79% FLNa content was found in P3 platelets, but the relative ratio of mutated versus full-length FLNa is unknown. Finally, FLNa expression determined by flow cytometry showed that 10% to 20% of FLNa-negative platelets were found in patients P1, P2, and P3, as previously described.14

Figure 1.

Figure 1. A and B, Level of wild-type and mutated filamin A (FLNa) in platelets for all patients. This figure represents a Western blotting of FLNa using a polyclonal antibody specific for the C-terminal FLNa and a monoclonal antibody specific for the N-terminal FLNa. These results are representative of at least 3 independent experiments.

Collagen- and Convulxin-Induced Platelet Aggregation and Platelet Secretion Are Decreased in Patients With FLNA Mutations

Because the signaling pathway of GPVI was shown to be impaired in FLNa-deficient mouse platelets,8 we next investigated platelet aggregation and secretion induced by type I collagen and convulxin (Cvx) in all patients. Platelet aggregation induced by various concentrations of type I collagen (0.25–2 μg/mL) and Cvx (400 and 800 pmol/L) was either absent or drastically decreased in P1, P3, and P4 platelets (Figure 2A, 2C, and 2D). Interestingly, at the same concentrations of Cvx, the intensity of platelet aggregation was normal in P2 (Figure 2B).

Figure 2.

Figure 2. Platelet aggregation and secretion induced by type I collagen and convulxin (Cvx). Aggregation of washed platelets and secretion were initiated by adding various concentrations of type I collagen (0.25–2 μg/mL) or Cvx (400 or 800 pmol/L) during 3 minutes. Aggregations of P1 (A), P2 (B), P3 (C), and P4 (D) were expressed as the percentage change in light transmission, with the value of the blank (buffer without platelets) set at 100%. Traces are representative of at least 2 experiments. Dense granule secretion of P1 (A), P2 (B), P3 (C), and P4 (D) was evaluated by assessment of ATP release after aggregation. Results are expressed as the amount of ATP released (pmol).

Because, at the concentrations used for collagen and Cvx, platelet aggregation is dependent on ADP (results not shown), dense granule secretion was quantified by measuring ATP released after 3 minutes of aggregation. At low doses of type I collagen or Cvx, P1, P3, and P4 platelets released 20% to 95% less ATP than control platelets (Figure 2A, 2C, and 2D). In contrast, Cvx-induced ATP release was normal in P2 platelets (Figure 2B). Interestingly, in patients’ platelets where truncated FLNa was undetectable, the level of remaining full-length FLNa seemed to correlate with the functional effect: weak aggregation/secretion for low FLNa patients (P1, 37% and P4, 57%), whereas normal aggregation/secretion for P2 (82%). P3 platelets, despite a near-normal FLNa content (79%), exhibited low aggregation/secretion, consistent with a dominant-negative effect of the p.Glu1803Lys mutation.

Platelet Signaling Induced by Cvx Is Decreased in Patients P1 and P4 With Low FLNa Content or Patient P3 With p.Glu1803Lys FLNa

We next characterized the signaling pathway of GPVI induced by Cvx (400–800 pmol/L) in unstirred platelets, that is, in the absence of integrin αIIbβ3 signaling. The expression of GPVI and FcRγ chain being normal for all patients (Figure II in the online-only Data Supplement), we tested the tyrosine kinase Syk phosphorylation, previously shown to be dependent on FLNa.8 Cvx activation of platelets induced Syk phosphorylation (Syk-P) in nonstirring conditions (Figure 3). Quantification of Syk-P showed it was low in P1, P3, and P4 platelets (48%, 59%, and 40% of control, respectively) but was normal in P2 platelets (103%). The phosphorylation of LAT (LAT-P), a direct substrate of Syk and a key actor in phospholipase C γ2 activation,17 was also low in P1 (49%), P3 (44%), and P4 platelets (52%), but not in P2 platelets (99%; Figure 3). The level of Syk-P or LAT-P parallels the level of wild-type FLNa in patients with FLNA deletions or nonsense mutation: low for P1 and P4 (37% and 57% wtFLNa) and normal for P2 (82%). The low Syk-P/LAT-P in P3 platelets is consistent with a dominant-negative effect of p.Glu1803Lys mutation.

Figure 3.

Figure 3. AD, Platelet signaling induced by convulxin (Cvx).Washed platelets in suspension were activated by Cvx (400 or 800 pmol/L) for 3 minutes in the absence of stirring. Tyrosine phosphorylation of Syk (Syk-P) and LAT (LAT-P) was assessed by immunoblotting with an anti–Syk-P and anti–LAT-P, respectively. The membranes were reprobed with an anti-Syk and anti–β-actin. These results are representative of at least 2 independent experiments.

Platelet Aggregation Induced by Other Agonists

We next examined in the 4 patients the effect of FLNA mutations on platelet aggregation and secretion induced by other agonists, such as thrombin (0.1–0.2 U/mL), proteinase-activated receptor 1-activating peptide (PAR1-AP) (5–10 μmol/L), and ADP (5–20 μmol/L). No significant difference between control and patients was observed in platelet aggregation and secretion, whatever the G protein–coupled receptor agonists (Figure III in the online-only Data Supplement), consistent with a specific defect in the GPVI-dependent activation pathway in these patients.

Platelet Spreading on Collagen, VWF, and Fibrinogen

FLNa being central in platelet spreading and cytoskeleton reorganization, we next analyzed the spreading of patients’ platelets on different matrices (collagen, VWF, and fibrinogen). Contrary to control platelets, the spreading of patients’ platelets over type I collagen (Figure 4A) was heterogeneous, with both large and small platelets present. The mean platelet area (versus control, 100%) was increased in P1, P3, and P4 reaching 196±25% (P<0.01), 167±12% (P<0.001), and 174±12% (P<0.001), respectively, but was unchanged in P2 (93±3%; Figure 4A). The absence of increased platelet spreading for P2 is consistent with the high amount (82%) of residual wild-type FLNa. For P3, the increased spreading might reflect the mean platelet area 1.5× larger than normal control. Conversely, the increased spreading for P1 and P4 is the likely result of low amounts of FLNa (P1 and P4). The defect may be a direct functional consequence of altered (quantitatively or qualitatively) FLNa and GPVI (collagen-specific) signaling or may be the indirect consequence of comparatively higher adhesion of large versus small platelets.

Figure 4.

Figure 4. A and B, Platelet spreading on collagen matrix under static conditions. Washed platelets (107 platelets/mL; 150 μL) were plated on glass coverslips precoated with type I collagen (50 μg/mL) at room temperature for 30 minutes. Surface area of platelets spread on collagen was quantified as described in Material and Methods section. Data are expressed as the mean±SEM of 3 independent experiments. Bar, 10 μmol/L. **P<0.01; ***P<0.001 (paired Student t test).

To test these hypotheses, we examined platelet spreading on VWF and fibrinogen. On both matrices, platelet area versus control (100%) was increased for P1 and P4: 159±11% (P<0.001) and 181±11% (P<0.01) for fibrinogen and 133±9% (P<0.01) and 181±16% (P<0.01) for VWF (Figure 5A). In contrast, the spreading of P2 and P3 platelets over VWF and fibrinogen was comparable with control. For P3, the normal spreading over VWF or fibrinogen, despite an enhanced mean platelet area (×1.5), might reflect a lower capacity of spreading compared with control. The absence of increased spreading for P2, whether on VWF, fibrinogen, or collagen, remains consistent with its near-normal content in FLNa. The increased spreading for P1 and P4 platelets, whatever the matrix type, confirms the effect of low FLNa content.

Figure 5.

Figure 5. Platelet spreading on von Willebrand factor (VWF), fibrinogen, and collagen matrix under static conditions.Washed platelets (107 platelets/mL; 150 μL) were plated on glass coverslips precoated with VWF (10 μg/mL) or fibrinogen (100 μg/mL) or collagen (50 μg/mL) at room temperature for 30 minutes. A, Surface area of platelets spread on VWF and fibrinogen was quantified as described in Material and Methods section. Bar, 10 μm. **P<0.01 (paired Student t test). B and C, Platelets were stained with Alexa Fluor488–labeled phalloidin (1/500) and an anti–C-terminal filamin A (FLNa; 1/500). Surface area of platelets expressing FLNa (FLNa+) or nonexpressing FLNa (FLNa–) was quantified as described in Material and Methods section. These results are representative of at least 2 independent experiments. Data are expressed as the mean±SEM of 3 independent experiments, and statistical significance was determined by 1-way ANOVA followed by a least significant difference multiple comparison (*P<0.05).

We thus examined platelet spreading with regard to FLNa content, as determined by immunofluorescence with an anti–C-terminal FLNa antibody, specific for full-length FLNa and Alexa Fluor488–labeled phalloidin. P4 patient was examined. Truncated FLNa was not detected by double labeling with anti–N-terminal and anti–C-terminal FLNa antibodies. Interestingly, the surface area of FLNa-negative P4 platelets (50–60% of adhering platelets) was markedly increased on all matrix types: 201±14% compared with control (100%) on collagen, and 226±22% and 225±16% on VWF and fibrinogen, respectively (Figure 5B). In contrast, the surface area of FLNa-positive P4 platelets was not different from control on collagen, VWF, or fibrinogen. In parallel, the size of FLNa-negative or FLNa-positive P4 platelets was measured by fluorescence-activated cell sorter analysis. No significant difference was observed between these 2 populations (FLNa+: 133% of control; FLNa–: 142% of control; Figure 5C). Altogether, these results suggest that absence of FLNa drastically augments spreading, consistent with FLNa exhibiting a negative regulation on cytoskeletal reorganization during spreading. In addition, the specific increase in spreading on collagen of FLNa-positive P3 platelets suggests that the GPVI/Syk pathway is more dependent on FLNa than the GPIb/VWF or the αIIbβ3/fibrinogen pathways.

Thrombus Formation Under Flow

Platelet cytoskeletal regulation is likely to play a role in thrombus formation.1118 We thus investigated thrombus formation on type I collagen at 300 s–1 and 1500 s–1 (Figure 6A and 6B). For these experiments, platelet counts were normalized. After 5 minutes of perfusion at 300 s–1, the area covered by P1, P3, and P4 platelets was lower than control (100%), reaching 54.4±6.0% (P<0.05), 45.2±2.5% (P<0.05), and 62.6±4.4% (P<0.01), respectively (Figure 6A). For P1 and P3, lower thrombus area correlated with defective platelet secretion as indicated by the absence of additive effect of the ADP or ATP scavenger apyrase on thrombus formation (Figure IV in the online-only Data Supplement). P4 was not tested. In contrast, the area covered by P2 was normal (115±5%).

Figure 6.

Figure 6. Thrombus formation on various matrix under flow. Whole blood was perfused (A) over a collagen matrix (50 μg/mL) at a shear rate of 300 s–1 (A) or 1500 s–1 (B) or over human von Willebrand factor (VWF; 50 μg/mL) at 1500 s–1 (C) and 5000 s–1 (D) in a parallel plate chamber. After 5 minutes (300 s–1) and 3 minutes (1500–5000 s–1), platelet adhesion and thrombus formation were evaluated by fluorescence microscopy (original magnification ×20). Total area covered by platelets was expressed as the mean±SEM of 3 independent experiments. *P<0.05; **P<0.01; ***P<0.001 (paired Student t test).

Under the shear conditions of arterial flow (1500 s–1), thrombus formation on collagen-coated surface was comparable with 300 s–1 for all patients’ platelets (Figure 6B). Altogether, collagen-induced thrombus formation correlates with the defect in FLNa: altered for patients exhibiting a quantitative (P1 and P4) or qualitative (P3) defect in platelet FLNa and unaltered when near-normal level of FLNa is observed (P2).

Because of the known interaction of FLNa with GPIbα, we extended our study to the investigation of thrombus formation on a VWF matrix. Shear rates at 1500 s–1 and 5000 s–1 were tested to mimic normal and pathological arterial flows, respectively. Total GPIbα expression assessed by flow cytometry analysis was normal for P1, P2, and P4 (results not shown), but increased for P3.14 At 1500 s–1, thrombus formation was similar for P1 and P2 compared with control donors (115.8±10.0% and 115.2±3.9%), moderately decreased for P4 (75.9±3.8%; P<0.01), but was significantly increased for P3 (152.6±11.0%; P<0.01; Figure 6C). At 5000 s–1, corresponding to a pathological shear rate, platelets from P1 and P4 covered only 29.2±3.4% (P<0.001) and 48.9±2.1 (P<0.001) of the VWF-coated surface covered by control platelets, whereas the covered surface was increased, up to 171.5±20.3% (P<0.01), for P3 and was normal for P2 (104±2.7%; Figure 6D).

Altogether, these results lead to the conclusion that (1) the status of thrombus formation, particularly at 5000 s–1 parallels that of platelet wild-type FLNa: low for P1 and P4 and normal for P2; and (2) the increased thrombus formation for P3 may be the consequence of increased GPIb exposure at the platelet surface. Another hypothesis, not exclusive from the first one, is that increased thrombus formation is the result of a gain-of-function mutation (pGlu1803Lys), possibly linked to the proximity of the interaction site betweeen FLNa and the GPIbα cytoplasmic domain.

Discussion

Recent results obtained using transgenic mice support a major role for FLNa in platelet functions. However, because these results were obtained in mouse models, we wished to test the role of FLNa in a human platelet context by examining the impact on platelet functions of FLNA gene mutations responsible for filaminopathy A. The 4 patients examined in this report are women, heterozygous for the FLNA mutations, 3 of which predicted to lead to truncated FLNa polypeptides, because of premature termination induced by a frameshift point mutation (P1) or by a gene deletion (P2 and P4). A fourth mutation (P3), p.Glu1803Lys, is predicted not to affect FLNA length (Figure I and Table I in the online-only Data Supplement). Because patients are women and FLNA is located on the X chromosome, both wild-type and mutant alleles are expected to be coexpressed. Interestingly, no truncated peptide was detected in platelets. This may be consistent with a nonsense mRNA decay mechanism that may be enhanced by the nonrenewal of transcripts in anucleated platelets.19 In addition, the long half-life of circulating platelets may accelerate the decay of truncated polypeptides that may have been translated in megakaryocytes (MK). Platelet proteins exhibit a degradation pattern (Figure 1). However, this seems to correspond to degraded normal FLNa since (1) seen with both N terminus–specific or C terminus–specific antibodies, whereas all truncations delete the C-terminal end of FLNa, (2) patterns are nonspecific, because observed with different mutations or even with some controls (see Figure 1 and P1, P4, and control), and (3) also visible in P3, whose mutation is nontruncating (p.Glu1803Lys).

A question raised by our observation is the varying levels of wild-type FLNa among patients. Because of X inactivation and X clonal expression in MK lineages,20 the FLNA gene in our female patients should undergo allelic exclusion. This would lead to 2 populations of platelets, each containing an FLNa originating either from the normal (FLNa+) or the defective allele (FLNa–), but not both. Although this seems to be the case for patient P4, where 50% to 60% of adhesion and spreading of platelets do not contain FLNa (FLNa–), only 20% of FLNa-negative platelets were observed for P1 or P2 platelets.14 This suggests that at least in those 2 patients, FLNa expression is skewed toward the normal allele, because the defective allele leads to less-efficient MK maturation or platelet production or to higher platelet clearance. Indeed, in mice, 95% of platelets isolated from FlnA+/– females contained FLNa.8 The reason why platelet numbers are normal for P2, with a total amount of FLNa near normal (82%), is that, in this patient, there is a higher expression of the unaffected FLNA allele compared with controls. Alternatively, platelet production from the mutant FLNA allele MK lineage is severely altered, but P2 platelet counts remain within the normal range because the MK lineage containing the wild-type FLNA allele happens to produce high platelet numbers. This is clearly not the case for P1, because total FLNa is low (37%) despite an 80% ratio of FLNa-containing platelets, as previously described14: this indicates that in this patient, the expression and stability of normal FLNa are diminished compared with controls. These marked differences in FLNa expression levels are likely because of differential consequences of the mutations at the level of platelet production or clearance.

Platelets from P1, P3, and P4 patients exhibited altered aggregation and secretion depending on collagen receptors but not on G protein–coupled receptor (thrombin or ADP receptors). This defect on platelet secretion is the consequence of FLNA mutations and not of an abnormal dense granule storage pool (results not shown). Furthermore, thrombus formation on collagen that is dependent on secretion in conditions of low shear rates (Figure IV in the online-only Data Supplement) was also diminished. In these cases, the low level of full-length FLNa in P1 and P4 platelets may explain altered GPVI signaling, because FLNa acts as a signaling scaffold for GPVI through interaction with Syk.21 For P3, it is possible that p.Glu1803Lys, an Ig repeat 16 mutation, interferes with the Ig repeat 5 engaged in signaling of GPVI–collagen interaction. Our data are consistent with previous reports in mouse platelets shown in FlnA-deficient mice that the presence of FLNa is essential for GPVI signaling.8 Of note, the limited platelet functional alterations (aggregation, secretion, thrombus formation) for P2 are consistent with the near-normal levels of full-length FLNa (82%).

Several lines of evidence have demonstrated that collagen-induced tyrosine phosphorylation of proteins, including immunoreceptor tyrosine-based activation motif (ITAM)-containing FcRγ chain22 and Syk tyrosine kinase,23 is essential for platelet activation. Using the model of FlnA-deficient mice, FLNa has been shown to contribute to Syk spatial distribution at the cytoplasmic surface of the plasma membrane.8 For all patients, the platelet functional defect is not a consequence of a defect in collagen receptors, because the levels of GPVI and FcRγ were normal. In contrast, the GPVI signaling pathway was affected in P1, P3, and P4 platelets, as confirmed by the drastic decrease in Cvx-induced phosphorylation of Syk and LAT. Our hypothesis is that the low level of full-length FLNa for P1 and P4 leads to low level of FLNa-associated Syk that, in turn, poorly recruits ITAM motifs of FcRγ. For P3, either the p.Glu1803Lys mutation or the abnormal localization of FLNa found for this patient in our previous study,14 or both, affects GPVI signaling.

Surprisingly, spreading of P1 and P4 platelets was increased on collagen, VWF, and fibrinogen. Platelet distribution of FLNa on collagen shows that the surface area was largely increased in FLNa-negative platelets for P4. Unexpectedly, FLNa-positive platelets also exhibited increased surface area, although to a lesser extent than FLNa-negative platelets, suggesting that the level of full-length FLNa in FLNa-positive platelets is probably lower than in control platelets. The other possibility is that there is functional interference between mutant and wild-type allele-bearing MK and platelets within the bone marrow or circulating blood. Furthermore, Syk has been reported to be critical for lamellipodia formation on various matrices, such as collagen, VWF, and fibrinogen.24 The impairment in Syk activation raises the question of lamellipodia formation in our patients’ platelets (P1, P3, and P4). In fact, full lamellipodia formation occurred in FLNa-negative platelets (results not shown), suggesting that lamellipodia formation (in addition to not requiring FlnA) does not require full phosphorylation of Syk. The other hypothesis is that the engagement of Syk in spreading is different in mouse24 and human platelets, for example, through different signaling pathways. Finally, a normal phosphorylation of Syk cannot be completely excluded in conditions of platelet spreading (not explored in our conditions), which would suggest that an alternate Syk-P pathway bypasses FLNa in human (but not in mouse) platelets during spreading.

For P3, it is possible that the limited increase in spreading on collagen is only the consequence of the large size of platelets (×1.5).14 In turn, because spreading on VWF and fibrinogen is comparable with control, and thus not increased proportionally to size, this suggests that p.Glu1803Lys FLNa exhibits a stronger negative regulation than wild-type FLNa on spreading on VWF and fibrinogen. The differential effect between collagen and VWF or fibrinogen may point to a differential engagement of FLNa in these conditions. Next, thrombus formation on a VWF matrix in physiological and pathological conditions was examined. In contrast to a recent report showing that GPIbα was degraded in platelets from FlnAloxP PF4-Cre mice,8,21 the cell surface expression of GPIbα assessed by flow cytometry analysis was normal for P1, P2,14 and P4 (results not shown) and was increased for P3,14 suggesting that the level of remaining FLNa is sufficient to promote normal cell surface expression of the GPIb complex. In agreement with the model of mice expressing human GPIbα mutated in the FLNA interaction site,11 thrombus formation on VWF was normal at 1500 s–1 for P1 and P2 platelets and only slightly decreased for P4 platelets. Thus, low FLNa content does not preclude normal adhesion in conditions of arteriolar shear. In contrast, the decrease in thrombus formation at 5000 s–1 for patients with a low level of full-length FLNa (P1 and P4) confirms that a normal level of FLNa is essential for platelet adhesion under pathological shear rates.11 Surprisingly, for P3, thrombus formation on a VWF matrix was increased in both physiological and pathological conditions. This gain-of-function effect might be the consequence of increased GPIbα expression (×2.2).14 The other possibility is that the mutation (Ig-like repeat 16) located between the site of FLNa–GPIb/VWF interaction (Ig-like repeat 17) and the calpain site affects the binding to GPIbα. The less peripheral distribution of FLNa in large versus normal-size P3 platelets, suggesting that GPIb–Glu1803Lys–FLNa would be less available to an interaction with GPIb, may be compensated for by the likely excess of FLNa over GPIb, leaving enough peripheral molecules to interact with membrane GPIb.

This study clearly demonstrates that platelet functions are altered in patients with FLNA mutations. Abnormal response to collagen affecting aggregation, secretion, thrombus formation, and GPVI signaling was associated with a low level of full-length FLNa. Spreading on various matrix types was also abnormal, especially in platelets devoid of FLNa. Finally, thrombus growth at low and high shear rates was associated with the normal level of FLNa. In conclusion, FLNa is essential for human platelet adhesive functions.

Acknowledgments

We acknowledge Hervé Falet for the critical revision of the manuscript. We gratefully acknowledge Anne-Cécile Pons. We also thank the patients who participated in this study.

Footnotes

*These authors contributed equally to this work.

The online-only Data Supplement is available with this article at http://atvb.ahajournals.org/lookup/suppl/doi:10.1161/ATVBAHA.112.300603/-/DC1.

Correspondence to Marijke Bryckaert, PhD, INSERM U770 Hôpital Bicêtre, 80 rue du Général Leclerc, 94276 Le Kremlin Bicêtre Cedex, France. E-mail

References

  • 1. Fox JE. Identification of actin-binding protein as the protein linking the membrane skeleton to glycoproteins on platelet plasma membranes.J Biol Chem. 1985; 260:11970–11977.CrossrefMedlineGoogle Scholar
  • 2. Nakamura F, Stossel TP, Hartwig JH. The filamins: organizers of cell structure and function.Cell Adh Migr. 2011; 5:160–169.CrossrefMedlineGoogle Scholar
  • 3. Feng Y, Walsh CA. The many faces of filamin: a versatile molecular scaffold for cell motility and signalling.Nat Cell Biol. 2004; 6:1034–1038.CrossrefMedlineGoogle Scholar
  • 4. van der Flier A, Sonnenberg A. Structural and functional aspects of filamins.Biochim Biophys Acta. 2001; 1538:99–117.CrossrefMedlineGoogle Scholar
  • 5. Zhou X, Tian F, Sandzén J, Cao R, Flaberg E, Szekely L, Cao Y, Ohlsson C, Bergo MO, Borén J, Akyürek LM. Filamin B deficiency in mice results in skeletal malformations and impaired microvascular development.Proc Natl Acad Sci USA. 2007; 104:3919–3924.CrossrefMedlineGoogle Scholar
  • 6. Sheen VL, Dixon PH, Fox JW, et al. Mutations in the X-linked filamin 1 gene cause periventricular nodular heterotopia in males as well as in females.Hum Mol Genet. 2001; 10:1775–1783.CrossrefMedlineGoogle Scholar
  • 7. Sun Y, Almomani R, Aten E, Celli J, van der Heijden J, Venselaar H, Robertson SP, Baroncini A, Franco B, Basel-Vanagaite L, Horii E, Drut R, Ariyurek Y, den Dunnen JT, Breuning MH. Terminal osseous dysplasia is caused by a single recurrent mutation in the FLNA gene.Am J Hum Genet. 2010; 87:146–153.CrossrefMedlineGoogle Scholar
  • 8. Falet H, Pollitt AY, Begonja AJ, Weber SE, Duerschmied D, Wagner DD, Watson SP, Hartwig JH. A novel interaction between FlnA and Syk regulates platelet ITAM-mediated receptor signaling and function.J Exp Med. 2010; 207:1967–1979.CrossrefMedlineGoogle Scholar
  • 9. Nakamura F, Pudas R, Heikkinen O, Permi P, Kilpeläinen I, Munday AD, Hartwig JH, Stossel TP, Ylänne J. The structure of the GPIb-filamin A complex.Blood. 2006; 107:1925–1932.CrossrefMedlineGoogle Scholar
  • 10. Falet H. New insights into the versatile roles of platelet FlnA.Platelets. February 28, 2012. doi:10.3109/09537104.2011.654004. http://informahealthcare.com/doi/abs/10.3109/09537104.2011.654004. Accessed November 9, 2012.MedlineGoogle Scholar
  • 11. Cranmer SL, Ashworth KJ, Yao Y, Berndt MC, Ruggeri ZM, Andrews RK, Jackson SP. High shear-dependent loss of membrane integrity and defective platelet adhesion following disruption of the GPIbα-filamin interaction.Blood. 2011; 117:2718–2727.CrossrefMedlineGoogle Scholar
  • 12. Kanaji T, Ware J, Okamura T, Newman PJ. GPIbα regulates platelet size by controlling the subcellular localization of filamin.Blood. 2012; 119:2906–2913.CrossrefMedlineGoogle Scholar
  • 13. Parrini E, Ramazzotti A, Dobyns WB, et al. Periventricular heterotopia: phenotypic heterogeneity and correlation with Filamin A mutations.Brain. 2006; 129:1892–1906.CrossrefMedlineGoogle Scholar
  • 14. Nurden P, Debili N, Coupry I, et al. Thrombocytopenia resulting from mutations in filamin A can be expressed as an isolated syndrome.Blood. 2011; 118:5928–5937.CrossrefMedlineGoogle Scholar
  • 15. Solé G, Coupry I, Rooryck C, et al. Bilateral periventricular nodular heterotopia in France: frequency of mutations in FLNA, phenotypic heterogeneity and spectrum of mutations.J Neurol Neurosurg Psychiatr. 2009; 80:1394–1398.CrossrefMedlineGoogle Scholar
  • 16. Chardon JW, Mignot C, Aradhya S, Keren B, Afenjar A, Kaminska A, Beldjord C, Héron D, Boycott KM. Deletion of filamin A in two female patients with periventricular nodular heterotopia.Am J Med Genet A. 2012; 158A:1512–1516.CrossrefMedlineGoogle Scholar
  • 17. Pasquet JM, Gross B, Quek L, Asazuma N, Zhang W, Sommers CL, Schweighoffer E, Tybulewicz V, Judd B, Lee JR, Koretzky G, Love PE, Samelson LE, Watson SP. LAT is required for tyrosine phosphorylation of phospholipase Cgamma2 and platelet activation by the collagen receptor GPVI.Mol Cell Biol. 1999; 19:8326–8334.CrossrefMedlineGoogle Scholar
  • 18. Pleines I, Hagedorn I, Gupta S, May F, Chakarova L, van Hengel J, Offermanns S, Krohne G, Kleinschnitz C, Brakebusch C, Nieswandt B. Megakaryocyte-specific RhoA deficiency causes macrothrombocytopenia and defective platelet activation in hemostasis and thrombosis.Blood. 2012; 119:1054–1063.CrossrefMedlineGoogle Scholar
  • 19. Nicholson P, Yepiskoposyan H, Metze S, Zamudio Orozco R, Kleinschmidt N, Mühlemann O. Nonsense-mediated mRNA decay in human cells: mechanistic insights, functions beyond quality control and the double-life of NMD factors.Cell Mol Life Sci. 2010; 67:677–700.CrossrefMedlineGoogle Scholar
  • 20. Raslova H, Roy L, Vourc’h C, Le Couedic JP, Brison O, Metivier D, Feunteun J, Kroemer G, Debili N, Vainchenker W. Megakaryocyte polyploidization is associated with a functional gene amplification.Blood. 2003; 101:541–544.CrossrefMedlineGoogle Scholar
  • 21. Jurak Begonja A, Hoffmeister KM, Hartwig JH, Falet H. FlnA-null megakaryocytes prematurely release large and fragile platelets that circulate poorly.Blood. 2011; 118:2285–2295.CrossrefMedlineGoogle Scholar
  • 22. Gibbins J, Asselin J, Farndale R, Barnes M, Law CL, Watson SP. Tyrosine phosphorylation of the Fc receptor gamma-chain in collagen-stimulated platelets.J Biol Chem. 1996; 271:18095–18099.CrossrefMedlineGoogle Scholar
  • 23. Ichinohe T, Takayama H, Ezumi Y, Arai M, Yamamoto N, Takahashi H, Okuma M. Collagen-stimulated activation of Syk but not c-Src is severely compromised in human platelets lacking membrane glycoprotein VI.J Biol Chem. 1997; 272:63–68.CrossrefMedlineGoogle Scholar
  • 24. Hughan SC, Hughes CE, McCarty OJ, Schweighoffer E, Soultanova I, Ware J, Tybulewicz VL, Watson SP. GPVI potentiation of platelet activation by thrombin and adhesion molecules independent of Src kinases and Syk.Arterioscler Thromb Vasc Biol. 2007; 27:422–429.LinkGoogle Scholar