Skip main navigation

Gain-of-Function Mutation in Filamin A Potentiates Platelet Integrin αIIbβ3 Activation

Originally publishedhttps://doi.org/10.1161/ATVBAHA.117.309337Arteriosclerosis, Thrombosis, and Vascular Biology. 2017;37:1087–1097

Abstract

Objective—

Dominant mutations of the X-linked filamin A (FLNA) gene are responsible for filaminopathies A, which are rare disorders including brain periventricular nodular heterotopia, congenital intestinal pseudo-obstruction, cardiac valves or skeleton malformations, and often macrothrombocytopenia.

Approach and Results—

We studied a male patient with periventricular nodular heterotopia and congenital intestinal pseudo-obstruction, his unique X-linked FLNA allele carrying a stop codon mutation resulting in a 100–amino acid–long FLNa C-terminal extension (NP_001447.2: p.Ter2648SerextTer101). Platelet counts were normal, with few enlarged platelets. FLNa was detectable in all platelets but at 30% of control levels. Surprisingly, all platelet functions were significantly upregulated, including platelet aggregation and secretion, as induced by ADP, collagen, or von Willebrand factor in the presence of ristocetin, as well as thrombus formation in blood flow on a collagen or on a von Willebrand factor matrix. Most importantly, patient platelets stimulated with ADP exhibited a marked increase in αIIbβ3 integrin activation and a parallel increase in talin recruitment to β3, contrasting with normal Rap1 activation. These results are consistent with the mutant FLNa affecting the last step of αIIbβ3 activation. Overexpression of mutant FLNa in the HEL megakaryocytic cell line correlated with an increase (compared with wild-type FLNa) in PMA-induced fibrinogen binding to and in talin and kindlin-3 recruitment by αIIbβ3.

Conclusions—

Altogether, our results are consistent with a less binding of mutant FLNa to β3 and the facilitated recruitment of talin by β3 on platelet stimulation, explaining the increased αIIbβ3 activation and the ensuing gain-of-platelet functions.

Introduction

Filamins (FLN) are dimeric actin-binding proteins that stabilize the actin skeleton. Three proteins (FLNa, FLNb, and FLNc) are the products of the corresponding genes FLNA, FLNB, and FLNC. FLNA that encodes FLNa, the most abundant isoform, is located on chromosome Xq28. FLNa is composed of an N-terminal actin-binding domain followed by 24 Ig-like repeats and the C-terminal domain that mediates dimerization. Platelets express predominantly FLNa.

FLNA mutations produce a wide spectrum of rare developmental disorders and cause various malformations of the brain, skeleton, and heart. The most frequent brain abnormalities are periventricular nodular heterotopia (FLNA-PVNH), which can be associated with other features including Ehlers–Danlos syndrome, although skeletal dysplasia including the otopalatodigital syndrome spectrum disorders and terminal osseous dysplasia, congenital intestinal pseudo-obstruction, and familial cardiac valvular dystrophy have also been described.1,2

FlnAloxP GATA1-Cre mice that lack FlnA exclusively in platelets are characterized by a macrothrombocytopenia with larger platelets and an increased tail-bleeding time.3 Moreover, α-granule secretion, integrin αIIbβ3 activation, and the signaling pathways depending on the collagen receptor glycoprotein VI (GPVI) were defective.4 Indeed, loss of the interaction between the tyrosine kinase Syk and FLNa results in a decrease in Syk tyrosine phosphorylation, which is required for ITAM receptor signaling. In platelets, FLNa interacts, through Ig-like repeat 17, with glycoprotein Ibα (GPIbα), the principal adhesion receptor of von Willebrand factor (VWF)5 This interaction that requires amino acids 563-571 of GPIbα plays an important role in platelet adhesion and plasma membrane stability under pathological shear rates.68 Finally, in unstimulated platelets, FLNa is constitutively linked to the cytoplasmic domain of the integrin β3 subunit, a component of the αIIbβ3 integrin, which on activation binds fibrinogen and VWF. The FLNa-β3 interaction requires FLNa repeat 21 and β3 amino acids 747–755.9 It has been proposed that activation of integrin αIIbβ3 requires the dissociation of FLNa from the β3 cytoplasmic domain.10,11 However, more recently, based on structural analysis, a new molecular mechanism for FLNa-mediated retention of integrin in a resting state has been proposed.12 However, this mechanism awaits experimental evidence in platelets.

In our previous studies, we have characterized the abnormal platelets in French patients with filaminopathies A, showing giant and normal platelets as well as an abnormal granule distribution and abnormal platelet production.13 In doing so, we correlated the platelet structural characteristics and the platelet functions of 4 female patients with heterozygous FLNA mutations: 3 patients with PVNH, whose FLNa mutations led to premature termination codons predicted to yield truncated FLNa-polypeptides for 2 of them and 1 patient without PVNH but exhibiting an isolated macrothrombocytopenia associated with a novel FLNa missense p.Glu1803Lys mutation within Ig-like repeat 16. We described how FLNa mutations alter platelet production leading to thrombocytopenia. We also found that no truncated mutant FLNa was detectable in the platelets of PVNH female patients with premature termination codons and that the severity of thrombocytopenia correlated with the residual FLNa expressed by the normal allele in these heterozygous patients: low platelet counts paralleled low full-length FLNa contents, whereas near-normal FLNa content correlated with the normal platelet count.14 Most importantly, platelet functions were always decreased compared with controls, paralleling the level of expressed wild-type FLNa.

We now report the rare case of a male patient associating PVNH with congenital intestinal pseudo-obstruction and gain-of-platelet functions. His platelets express a novel hemizygous mutant FLNa; the resulting mutation deleted the TG dimer of the TGA stop codon of FLNa leading to an extended (100 amino acids) C-terminal region (NP_001447.2: p.Ter2648SerextTer101). The platelet counts were normal, and only a small subpopulation of platelets exhibited an abnormal granule distribution and vacuoles. Importantly, aggregation, secretion, adhesion, and thrombus formation dependent on ADP (on collagen and VWF under flow conditions) were found significantly increased. Although Rap1 activation was normal, αIIbβ3 integrin activation and talin recruitment to β3 induced by ADP were upregulated. When compared with wild-type FLNa, the expression of mutant FLNa in the HEL megakaryocytic cell line enhanced αIIbβ3 integrin activation, talin-β3 and kindlin-3-β3 association assessed by a proximity ligation assay, indicating that mutant FLNa per se is the cause of the gain-of-platelet functions in the patient.

Materials and Methods

Materials and Methods are available in the online-only Data Supplement.

Results

Mutation Analysis

A patient (P) carrying a FLNA mutation, with a familial history of early deaths of several siblings (older brothers, uncles) linked to congenital intestinal occlusions, was studied. Only males were affected, consistent with an X-linked transmission (Figure 1A). DNA sequencing of the FLNA gene of the proband (III.6) identified an hemizygous 2 base-pair deletion in exon 48 (NM_001110556.1(FLNA):c.7941_7942delCT). The same mutation was recently reported in another family.15 This mutation substitutes the natural TGA stop codon for Ser extending the FLNa C-terminus by 100 amino acids (p.Ter2648SerextTer101). His mother (II.3) was found heterozygous for the mutation, whereas his 2 unaffected sisters (III.2 and III.5) were not carriers (Figure 1B).

Figure 1.

Figure 1. Pedigree of the proband and familial analysis of FLNA. A, All affected males, except the proband (III.6, arrow), died perinatally from severe intestinal alterations. DNA was not available for genetic testing in I.1 (nt). B, Electrophoregrams from FLNA exon 48 sequencing showing the c.7941_7942del mutation in the proband (III.6) and his mother (II.3). Nucleotide and corresponding codons are noted above electrophoregrams. Note the double sequence for the mother (II.2), heterozygote for the mutation. Wild-type electrophoregram is shown for III.2 (natural stop codon is underlined).

Patient Platelets Are Essentially Normal in Size

Because giant and normal platelets were previously described in patients with FLNA-PVNH,13 we analyzed the patient platelet characteristics by electron microscopy (Figure I in the online-only Data Supplement). The resulting images show that the majority of the platelets was normal in size. Only few platelets appeared larger than normal and exhibiting vacuoles and an abnormal distribution of granules. These platelets represented a minor fraction of total platelets (7.4% of the patient platelets versus 3.7% of control platelets) and did not affect the platelet count (220–300×109 platelets per liter) and the mean platelet volume (11.4–11.8 fL).

FLNa Level Is Low in the Patient Platelets

We then quantified platelet FLNa by flow cytometry and by Western blotting. Because the FLNA gene is located on the X chromosome and the proband is a male, only the mutant allele was expected to be detected. Quantification by flow cytometry using an FLNa-specific antibody showed that the patient platelet FLNa level was only 30% of control platelets (Figure 2A). Western blotting showed that the mutant FLNa migrated with a slightly higher apparent molecular mass than the wild-type FLNa (Figure 2B), consistent with the predicted extended C-terminal tail (≈1 kDa). Note the absence of a band comigrating with wild-type FLNa. Assessment of band intensities confirmed a 33±1% content in FLNa in patient platelets compared with control platelets. Using an antibody specific for the putative extended C-terminal sequence of patient FLNa (see online-only Data Supplement), mutant FLNa was detected in the patient platelets and not in control platelets, confirming expression of the extended form of FLNa in the patient platelets (Figure 2C). FLNa-platelet distribution was then examined after platelet spreading on human VWF (10 µg/mL). Staining of FLNa showed that, as for control platelets, FLNa was present at approximately the same level in all patient platelets (Figure II in the online-only Data Supplement). Close examination of subcellular distribution by confocal microscopy showed that contrasting with control platelets, mutant FLNa seemed more central (Figure 2D). Altogether, only 30% of mutant FLNa is present in patient platelets, but in contrast to female PVNH patients,14 FLNa is present in all platelets.

Figure 2.

Figure 2. Mutant FLNa level is decreased in the patient platelets. A, Permeabilized control and patient P platelets were analyzed by flow cytometry using a FLNa-specific antibody. The level of FLNa was also quantified by immunoblotting using polyclonal antibodies directed against (B) wild-type FLNa or (C) the specific extended C-terminal domain of FLNa-P of the patient. D, Subcellular localization of FLNa was visualized by confocal microscopy (Zeiss LSM700 inverted, Center for Innovation and Technology Research, Pasteur Institute, Paris, France) using an FLNa-specific antibody and phalloidin (which visualizes F-actin).

Convulxin-Induced Functions of the Patient Platelets Are Increased and Dependent on Secreted ADP

FLNa being central to GPVI signaling, we next examined the response of the patient platelets to convulxin, a potent GPVI-dependent platelet agonist. First, the levels of GPVI, FcR γ-chain, integrin β1, and αIIbβ3 were assessed and found to be normal in the patient platelets by Western blotting and flow cytometry (Figures 3A and 3B). Surprisingly, aggregation intensity of the platelets was higher than that for controls at intermediate concentrations of the agonist (convulxin: 400 pmol/L) but was normal at higher concentrations (800 pmol/L; Figure 3C). Furthermore, dense granule secretion, as measured by quantification of released ATP, was markedly increased at 400 pmol/L of convulxin (560% of control; Figure 3D). We next assessed GPVI signaling after platelet activation by convulxin (400 and 800 pmol/L) in unstirred conditions to prevent αIIbβ3 engagement. Tyrosine kinase Syk phosphorylation previously shown to be dependent on FLNa4 was low in the patient platelets (58% of control at 400 pmol/L; Figure 3E), probably because of the lower amount of FLNa in the patient platelets (30% of control). In contrast, PLCγ phosphorylation was normal whatever the concentrations of convulxin used, strongly suggesting that the gain-of-platelet dense granule secretion is not the result of an increased Syk recruitment by FLNa. Finally, we investigated adhesion and thrombus formation on collagen at 300 s−1; both are dependent on collagen receptors but independent of the VWF/glycoprotein Ib-IX-V (GPIb-IX-V) interaction. After 5 minutes, the area covered by the patient platelets was greater than the control (138%; P<0.001; Figure 3F). Interestingly, removal of ADP by the ADP/ATP scavenger apyrase (2 U/mL) totally abolished the adhesion increase of the patient platelets, which adhered even less than control platelets (≈50%). Likewise, thrombus size, measured by mean fluorescence intensity, with the patient platelets was significantly increased compared with controls (161%; P<0.001) in a manner totally dependent on secreted ADP (Figure 3F). Altogether, these results lead to the conclusion that mutant FLNa correlates with a gain-of-adhesion and -thrombus formation dependent on secreted ADP in conditions of activation via GPVI.

Figure 3.

Figure 3. Platelet activation is increased after stimulation with convulxin (Cvx). The levels of (A) collagen receptors (β1 subunit integrin, glycoprotein VI [GPVI], and FcRγ-chain) and of (B) αIIbβ3 were assessed by flow cytometry and Western blotting. C, Platelet aggregation and (D) secretion were initiated by adding various concentrations of Cvx (0.2–0.8 nmol/L) for 3 min. Results are expressed as the amount of ATP released (pmoles). Gray bars corresponding to control and hatched bars to patient. E, GPVI signaling pathway activated by various concentrations of Cvx (0.2–0.8 nmol/L) in the absence of stirring for 3 min was assessed by immunoblotting using an anti-Syk-P and anti-PLCγ-P. Results are representative of 4 independent experiments with different donors as control. F, Thrombus formation was assessed in blood flow on collagen matrix (50 µg/mL) at 300 s−1 in whole blood after normalization of platelet counts, in the presence (+, hatched bars) or the absence (−, plain bars) of apyrase for both control (white bars) and the patient (gray bars). Adhesion and thrombus formation were evaluated by fluorescence microscopy. The total area covered by platelets (Adhesion) was expressed as the mean±SEM of 3 independent experiments (***P<0.001, paired Student t test), relative to control platelets given as 100%.

VWF/GPIb-IX-V Interaction Increases Patient Platelet Functions

Because a defect in FLNa has been shown to affect GPIb-IX-V surface expression,4 we next tested GPIb-IX-V–dependent functions of the patient platelets. First, platelet surface glycoprotein Ibα level was normal as assessed by flow cytometry (Figure 4A). Platelet agglutination and aggregation were next assessed in the presence of different concentrations of VWF (0.5–2.5 µg/mL) in the presence of ristocetin (0.8 mg/mL). Aggregation of the patient platelets was slightly increased at intermediate concentrations of VWF (0.5 and 1 µg/mL) and normal at higher concentration (2.5 µg/mL; Figure 4B). An increased secretion of ATP was observed at higher concentrations of VWF (1 and 2.5 µg/mL; Figure 4C). Finally, under blood flow conditions at 1500 s−1, adhesion was similar for controls and the patient (Figure 4D). In contrast, thrombus size was significantly increased to 231% (P<0.001) of the control and was essentially independent of ADP (Figure 4D). Altogether, these results indicate that the patient platelets exhibit a minor GPIb-IX-V–dependent gain-of-platelet functions, not affecting GPIb-IX-V/VWF interaction under high shear but αIIbβ3-dependent thrombus growth and independently from ADP secretion, pointing to direct upregulation of αIIbβ3 activation via GPIb-IX-V/VWF signaling.

Figure 4.

Figure 4. von Willebrand factor (VWF)–induced platelet activation is increased. A, The expression level of GPIbα was quantified by flow cytometry. B, Platelet aggregation and (C) secretion were initiated by adding various concentrations of VWF (0.2–2.5 µg/mL) in the presence of ristocetin (0.8 mg/mL) during 3 min. Results are expressed as the amount of ATP released (pmoles), gray bars corresponding to control and hatched bars to patient. Results are representative of 4 independent experiments with different donors as control. D, Thrombus formation was assessed on a VWF matrix (10 µg/mL) under blood flow at 1500 s−1. Platelet counts in whole blood were normalized. After 3 min, adhesion and thrombus formation were evaluated by fluorescence microscopy (original magnification ×20). The total area covered by platelets (Adhesion) and the fluorescence of each thrombus that allows evaluation of the extent of thrombus formation were achieved using Image J software and expressed as the mean±SEM of 3 independent experiments (***P<0.001, paired Student t test).

ADP-Induced Platelet Aggregation and Secretion Are Enhanced

Because all results pointed to upregulation of αIIbβ3 engagement, we next examined ADP aggregation of the patient platelets, the best known activation pathway of αIIbβ3. Aggregation to ADP was significantly increased over control platelets whatever the ADP concentration (Figure 5A). Simultaneous, measurements of platelet secretion from dense granules (ATP release) and from α granules (VWF release) showed increases for both compared with control platelets (Figures 5B and 5C). Most importantly, αIIbβ3 activation as assessed by PAC1 binding by flow cytometry was markedly enhanced (2-fold) for the patient platelets compared with controls (Figure 5D), whereas total αIIbβ3 was normal (Figure 3A). Strikingly increased αIIbβ3 activation was correlated with increased talin recruitment by β3 in patient platelets (180% of control) as assessed by coimmunoprecipitation of integrin αIIbβ3 and talin after ADP stimulation (Figure 5E). In contrast, αIIbβ3 activation signaling pathways including Rap1 activity, which is required for the recruitment of talin to β3 and αIIbβ3 activation (Figure 5F), or Ca2+ store mobilization, Ca2+ influx, and PI3-kinase activity (Figure III in the online-only Data Supplement) were normal. Altogether, these results show that upregulation of αIIbβ3 activation is driven by the last step of αIIbβ3 activation pathway and that the mutant FLNa potentiates talin recruitment to integrin αIIbβ3.

Figure 5.

Figure 5. ADP-induced platelet activation is increased in patient platelets. A, Platelet aggregation was initiated by adding various concentrations of ADP (2.5–20 µmol/L final) and fibrinogen (0.2 mg/mL) during 3 min. Aggregation was expressed as % change in light transmission with the value of the blank (buffer without platelets) set at 100%. B, Dense granule secretion was evaluated by assessment of ATP release after platelet aggregation. Results are expressed as the amount of ATP released (pmoles), gray bars corresponding to control and hatched bars to patient. C, Alpha granule secretion was assessed by quantification of VWF release by immunoblotting after platelet aggregation induced by ADP (10 µmol/L) and at various time points (0–10 min). D, Integrin αIIbβ3 activation induced by various concentrations of ADP (2.5–20 µmol/L final) was assessed by flow cytometry using the monoclonal antibody PAC1 specific for the active conformation of αIIbβ3. The level of integrin activation is expressed as the mean fluorescence intensity (MFI). E, Talin recruitment by integrin αIIbβ3 induced by ADP (10 µmol/L) was examined. Coimmunoprecipitation of αIIbβ3 integrin and talin was performed with the αIIbβ3-specific P2 monoclonal antibody (4 µg/mL) followed by immunoblotting using anti-talin and anti-β3 antibodies. F, Rap1 activation that governs αIIbβ3 integrin activation was analyzed after ADP (10 µmol/L) induction, and GTP-bound Rap1 was assessed by pull-down (see Materials and Methods section) followed by immunoblotting. Total Rap1 was assessed in the same amount of lysate by use of an anti-Rap1 antibody. Results are representative of 4 independent experiments with different donors as control.

Expression of Mutant FLNa in HEL Cells Induced an Increase in αIIbβ3 Activation and in the Association of Talin and Kindlin-3 With the Integrin

To determine whether mutant FLNa per se was responsible for the increased activation of αIIbβ3, wild-type FLNa (FLNa-WT) and mutant FLNa (FLNa-P) constructs were overexpressed in HEL cells. The efficiency of transfection was between 20% and 35% for FLNa-WT and FLNa-P, with 5- to 8-fold enhanced expression for both FLNa-WT and FLNa-P. Among several hematopoietic cell lines with megakaryocytic potential, HEL cells have been particularly used to study integrin αIIbβ3 functions.16 Activation of αIIbβ3 was evaluated by binding of Oregon Green-labeled fibrinogen to HEL cells stimulated with 800 nmol/L PMA (an activator of protein kinase C, a pathway known to activate the integrin) and analyzed by flow cytometry. αIIbβ3-binding specificity was demonstrated by complete inhibition with the αIIbβ3-specific P2 monoclonal antibody (10 µg/mL).17 PMA-induced fibrinogen binding to αIIbβ3 was unaffected by expression of FLNa-WT. After nucleofection of FLNa-WT and using the fraction of HEL cells not expressing recombinant FLNa as an internal control (rFLNa), no difference in fibrinogen binding was found between FLNa-WT-expressing (rFLNa+) and rFLNa HEL cells (Figure 6A). This demonstrated that expression of FLNa-WT per se did not impact αIIbβ3 activation. In contrast, expression of FLNa-P in transfected cells (rFLNaP+) on PMA activation significantly increased fibrinogen binding compared with rFLNa-P HEL cells (240±19%; P<0.001; Figure 6A). In parallel, the level of αIIbβ3 in HEL cells, expressing or not recombinant FLNa, remained unchanged in all conditions (Figure 6B). Thus, these results clearly show that mutant FLNa-P per se is responsible for the increase in αIIbβ3 activation. We next explored talin and kindlin-3 recruitment by αIIbβ3 in recombinant FLNa-expressing HEL cells stimulated with PMA. Talin-β3 and kindlin-3-β3 associations were assessed by proximity ligation assay that generates a signal (visualized as fluorescent dots) only when 2 proteins are close enough (≤40 nm). Bright red fluorescent dots, indicating talin-β3 and kindlin-3-β3 complexes, were detected in HEL cells, expressing or not recombinant FLNa (Figures 6C through 6E). Quantification showed that in FLNa-WT–transfected HEL cells, talin-β3 and kindlin-3-β3 were detected at the same level in rFLNa+ and rFLNa cells (Figure 6D and 6E). In contrast, in FLNa-P–transfected HEL cells, the talin-β3 and kindlin-3-β3 complexes were significantly higher (talin-β3: 227±53%; P<0.05 and kindlin-3-β3: 168±14%; P<0.001) in rFLNaP+ versus rFLNaP cells. We next performed immunoprecipitation of αIIbβ3 followed by detection of rFLNa (WT or P) by Western blotting using an anti-Myc tag antibody. Coimmunoprecipitation of FLNa-P with αIIbβ3 was significantly lower than that of FLNa-WT (Figure 6F). To check that the difference in FLNa recovery was not because of a difference in FLNa content available to αIIbβ3 for binding, FLNa was assessed in the detergent-soluble and -insoluble fractions of HEL lysates. Surprisingly, although most FLNa-WT was found in the soluble fraction, only a minor fraction (30%) of FLNa-P was recovered in the soluble fraction (Figure 6G). This suggests that the lower recovery of FLNa-P with αIIbβ3 is secondary to its low concentration in the soluble fraction.

Figure 6.

Figure 6. Fibrinogen binding, αIIbβ3-talin, and αIIbβ3-kindlin-3 associations are enhanced in HEL cells expressing mutant. A, HEL cells were transiently transfected with plasmids encoding myc-tagged wild-type FLNa (FLNa-WT) or myc-tagged mutant FLNa (FLNa-P). Cells were stimulated with 800 nmol/L PMA in the presence of Oregon Green 488–labeled fibrinogen (20 µg/mL). Fibrinogen binding was measured by flow cytometry in myc-FLNa-negative (rFLNa, gray bars) and myc-FLNa–positive cells (rFLNa+, hatched bars; assessed by detection of the myc tag on expressed recombinant FLNa). Increase in mean fluorescence intensity (MFI) induced by PMA in rFLNa cells was set at 100%. The error bars represent means±SEM of 3 determinations (***P<0.001, unpaired Student t test). B, Expression of αIIbβ3 in rFLNa and rFLNa+ was measured by cytometry. Transfected cells were stimulated by PMA (20 nmol/L) and plated on fibrinogen-coated coverslips for 30 min before analysis of (C and D) talin-αIIbβ3 and (E) kindlin-3-αIIbβ3 association by proximity ligation assay (PLA) and fluorescence microscopy (epifluorescence microscope Nikon Eclipse E600). Red spots correspond to PLA-positive signals. rFLNa are the subpopulation of FLNa-transfected HEL cells not expressing recombinant FLNa, as judged by absence of signal for the myc tag. No specific signal was detected when HEL cells were incubated without primary antibodies (negative control). D and E, PLA signal/cell was quantified, and the average of signal/FLNa cell was set at 100%. For each experiment, 50 cells per condition were analyzed. Gray bars correspond to rFLNa cells and hatched bars to rFLNa+ cells. Data are representative of 3 independent experiments (*P<0.05, ***P<0.001, unpaired Student t test) F, Coimmunoprecipitation of αIIbβ3 integrin and rFLNa in transfected cells was performed with the αIIbβ3-specific P2 monoclonal antibody (4 µg/mL) followed by Western blotting using antimyc and anti-β3 antibodies. G, FLNa-WT and FLNa-P levels were determined by Western blotting using an anti-Myc antibody in total and detergent lysates.

Taken together, these results clearly show that FLNa-P is associated with enhancement of αIIbβ3 activation via increased talin and kindlin-3 recruitment and suggested that a decrease of FLNa affinity for αIIbβ3 was not involved.

Discussion

Evidence is growing that FLNa plays major roles in platelet functions. Accordingly, the constitutive binding of FLNa to GPIb-IX-V, the receptor for VWF, has been shown to strengthen adhesion of platelets onto VWF at high shear rates.6 FLNa is also constitutively bound to αIIbβ3 integrin (a receptor for fibrinogen and VWF), and maintains it in a resting state.12 The dissociation FLNa from the integrin is required for receptor activation.911 Finally, FLNa is involved in GPVI signaling, through FLNa/Syk tyrosine kinase interaction.4 In addition, recent studies on filaminopathy A patients have shown that FLNa mutations in female patients are associated with thrombocytopenia with giant platelets, alteration of platelet production, platelet morphology, and abnormal granule distribution.13 Analysis of platelet functions of 3 of these female patients, all with a putatively truncating mutations of FLNa, showed impaired platelet functions correlating with low levels of residual full-length FLNa in platelets.14 Of note, functional alteration of these patients’ platelets was not secondary to the mutant truncated FLNas, not expressed in their platelets. The present study addresses the functional alterations of platelets from a male patient, thus carrying only one X-linked FLNA allele, and exhibiting a hemizygous FLNa mutation (FLNa-P) with an extended C-terminal sequence as a consequence of a stop codon mutation.

Unlike female PVNH patients analyzed previously,14 a normal count of platelets was observed in this male patient exhibiting both PVNH and congenital intestinal pseudo-obstruction. Morphology analysis showed that the majority of his platelets was normal. Only a minor subpopulation of platelets with vacuoles appeared larger but did not affect platelet count and platelet volume. In contrast to PVNH female patients, exhibiting low levels of full-length FLNa and no truncated FLNa detectable in their platelets, the mutant FLNa of this male patient was present in all platelets at 30% of the normal level. Although not fully quantitative, immunofluorescence imaging suggested that subcellular distribution of FLNa-P between the periphery and the center of platelets was altered. Indeed 38% of FLNa-P versus 9% of FLNa-WT being present at the center of platelets suggesting either higher interaction of FLNa-P with the cytoskeleton or lower interaction with its peripheral partners.

Surprisingly, platelet aggregation and secretion induced by various agonists such as convulxin or VWF in the presence of ristocetin (online-only Data Supplement) were upregulated. This gain-of-platelet function is unlikely to be a consequence of the low level in platelet FLNa (30%), because in PVNH female patients, the same low levels of FLNa (30%, WT) correlated with loss- and not gain-of-platelet functions.14 Unlike FLNa-defective platelets from heterozygous female patients, platelets from male patient expressed only the mutant FLNa-P and not the WT allele. Thus, upregulation of platelet aggregation and secretion is the most likely direct effect of mutant FLNa-P per se and not its low level. In addition, the increase in platelet aggregation induced by convulxin was not the consequence of an upregulation of the first steps of the GPVI-dependent signaling pathway because Syk phosphorylation was diminished, and PLCγ phosphorylation was normal in the patient platelets. Enhanced secretion was not the consequence either of an increase in granule number or content because full secretion of the patient platelets after stimulation with high concentrations of agonists (convulxin or VWF in the presence of ristocetin) was normal. In blood flow conditions, enhanced thrombus formation on collagen (300 s−1) essentially secondary to enhanced dense granule secretion (ATP, ADP) of the patient platelets confirmed the gain-of-platelet functions. Taken as a whole, these results are consistent with secretion enhancement to be most likely secondary to enhanced αIIbβ3 engagement. Increased ADP-induced aggregation of patient platelets correlated with an increase in surface exposure of activated αIIbβ3 (detected by the specific monoclonal antibody PAC-1) and ATP secretion. Conversely, in unstirred platelets, that is, in conditions of absence of αIIbβ3 engagement, ADP-stimulated platelets did not induce augmented P-selectin expression, that is, secretion (Figure VI in the online-only Data Supplement). In conditions of blood flow, thrombus formation on VWF and fibrinogen (results not shown) is known to be because of αIIbβ3 engagement. The enhancement shown is mostly independent of ADP, suggesting that activation enhancement in the patient platelets is not specific to the ADP pathway. Moreover, Akt phosphorylation, which reflects PI3-kinase activity, and Ca2+ mobilization, both required for αIIbβ3 activation induced by ADP, were normal in unstirred platelets. Furthermore, in the ultimate step of αIIbβ3 integrin activation, Rap1 activation, which controls recruitment of talin to αIIbβ3, was normal, in apparent contradiction with the increased talin recruitment by αIIbβ3. Finally, expression of mutant FLNa in HEL cells confirmed the enhanced talin-αIIbβ3 and kindlin-3-αIIbβ3 association and αIIbβ3 activation measured by the soluble fibrinogen binding induced by PMA. Together, these results indicate that mutant FLNa-P is involved in αIIbβ3 activation enhancement, explaining the gain of functions of the patient platelets.

The higher αIIbβ3 integrin activation and talin recruitment to β3 in the patient platelets on ADP stimulation are not secondary to increased stimulation of the Rap1-signaling pathway because Rap1 activity is normal. More specifically, overactivation of αIIbβ3 in FLNa-P–expressing HEL cells correlating with over-recruitment of talin and kindlin-3 by the integrin strongly argues in favor of FLNa-P being responsible for increased talin and kindlin-3 recruitment. This is consistent with the model of αIIbβ3 activation by the coordinated dissociation of FLNa from αIIbβ3 cytoplasmic domain and the concurrent binding of talin and kindlins.10,11,18 A recent elegant structural study has shown that the Ig-like domain 21 (and possibly Ig-9, -12, -17, and -19) of FLNa clasps together the cytoplasmic domains of both αIIb and β3 thereby stabilizing the integrin in an inactive state.12 A central point in our data is that the basal level of activated αIIbβ3 in resting platelets of the patient was low, identical to control platelets. This suggests, based on the model of Liu et al, that FLNa-P is constitutively bound to αIIbβ3, maintaining αIIbβ3 in a resting conformation. This was confirmed in unstimulated transfected HEL cells, where FLNa-P coimmunoprecipitated with αIIbβ3. Smaller amounts of FLNa-P compared with FLNa-WT were recovered but paralleled the lower concentration of FLNa-P in the detergent-soluble fraction of HEL cells, thus suggesting that FLNa-P and FLNa-WT have a similar affinity for αIIbβ3.

FLNa-P in patient platelets was expressed at only 30% the normal level: it is tempting to speculate on a similar lower interaction with αIIbβ3. Unfortunately, detection of FLNa-P in the insoluble fraction of patient platelets was unsuccessful, presumably because of too low amounts of FLNa-P for detection.

The question remains as to how does FLNa-P induce over-recruitment of talin/ kindlin-3 (and overactivation of the integrin)? The simplest explanation could be that the smaller amount of FLNa-P associated with αIIbβ3 leaves a significant part of αIIbβ3 free of FLNa and therefore easily available to talin/kindlin-3. However, the corollary is that FLNa-free αIIbβ3 remains in a resting state because the basal level of PAC1 binding to αIIbβ3 was not elevated in the patient platelets. This contention is contradictory with the model of Liu et al12 of regulation of αIIbβ3 activation by FLNa, as well as with experiments from Das et al,19 showing that transfected αIIbβ3 exhibited a significant basal level of activation, downregulated by overexpression of FLNa. This argues against the FLNa-free αIIbβ3 activation hypothesis. Furthermore, this hypothesis would not be consistent with the observation that PVNH platelets from patients of our first study, with low levels of normal FLNa,14 did not exhibit αIIbβ3 activation but instead αIIbβ3 inhibition.

Other hypotheses are, for example, that the 100 amino acid extension interferes with the Ig-like 24 domain and thus with FLNa dimerization itself involved (undocumented yet) in αIIbβ3 activation or that the high affinity of FLNa-P for polymerizing actin cytoskeleton (consistent with FLNa-P centralization in spread platelets or FLNa-P cosedimentation with detergent-insoluble material in HEL cells) facilitates an unknown mechanism of αIIbβ3 activation. However, sorting between these hypotheses and elucidating the exact mechanism at play will require extended studies.

Overactivation of patient αIIbβ3 leads to the possibility that patients experience thrombosis. Fortunately, this has not been the case. This is in fact consistent with our finding that in the patient platelets, αIIbβ3 is in a resting state. However, in a context favoring platelet activation such as, for example, atherosclerosis or diabetes mellitus, there is a possibility that this patient develops severe thrombosis. Peculiar attention is, thus, required for the follow-up of this patient, as well as healthcare.

In conclusion, this study shows evidence for gain-of-platelet functions of a mutant FLNa and demonstrates in human platelets the role of FLNa in the negative control of αIIbβ3 activation.

Nonstandard Abbreviations and Acronyms

αIIbβ3

integrin αIIbβ3

β3

β3 subunit of integrin αIIbβ3

FLNa

filamin A

FLNA

human gene for FLNa

GPIb-IX-V

glycoprotein Ib-IX-V

GPIbα

α chain of the subunit GPIb of GPIb-IX-V

GPVI

glycoprotein VI

PVNH

periventricular nodular heterotopia

VWF

von Willebrand factor

Acknowledgments

We wish to thank the patient who participated in this study.

Footnotes

*These authors contributed equally to this article.

The online-only Data Supplement is available with this article at http://atvb.ahajournals.org/lookup/suppl/doi:10.1161/ATVBAHA.117.309337/-/DC1.

Correspondence to Marijke Bryckaert, PhD, INSERM UMR_S 1176, Hôpital Bicêtre, 80 rue du Général Leclerc, 94276 Le Kremlin Bicêtre Cedex, France. E-mail

References

  • 1. 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 Psychiatry. 2009; 80:1394–1398. doi: 10.1136/jnnp.2008.162263.CrossrefMedlineGoogle Scholar
  • 2. Sheen VL, Jansen A, Chen MH, et al. Filamin A mutations cause periventricular heterotopia with Ehlers-Danlos syndrome.Neurology. 2005; 64:254–262. doi: 10.1212/01.WNL.0000149512.79621.DF.CrossrefMedlineGoogle Scholar
  • 3. Falet H. New insights into the versatile roles of platelet FlnA.Platelets. 2013; 24:1–5. doi: 10.3109/09537104.2011.654004.CrossrefMedlineGoogle Scholar
  • 4. 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. doi: 10.1084/jem.20100222.CrossrefMedlineGoogle Scholar
  • 5. Feng S, Reséndiz JC, Lu X, Kroll MH. Filamin A binding to the cytoplasmic tail of glycoprotein Ibalpha regulates von Willebrand factor-induced platelet activation.Blood. 2003; 102:2122–2129. doi: 10.1182/blood-2002-12-3805.CrossrefMedlineGoogle Scholar
  • 6. 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. doi: 10.1182/blood-2010-07-296194.CrossrefMedlineGoogle Scholar
  • 7. Cranmer SL, Pikovski I, Mangin P, Thompson PE, Domagala T, Frazzetto M, Salem HH, Jackson SP. Identification of a unique filamin A binding region within the cytoplasmic domain of glycoprotein Ibalpha.Biochem J. 2005; 387(pt 3):849–858. doi: 10.1042/BJ20041836.CrossrefMedlineGoogle Scholar
  • 8. 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. doi: 10.1182/blood-2005-10-3964.CrossrefMedlineGoogle Scholar
  • 9. Kiema T, Lad Y, Jiang P, Oxley CL, Baldassarre M, Wegener KL, Campbell ID, Ylänne J, Calderwood DA. The molecular basis of filamin binding to integrins and competition with talin.Mol Cell. 2006; 21:337–347. doi: 10.1016/j.molcel.2006.01.011.CrossrefMedlineGoogle Scholar
  • 10. García-Alvarez B, de Pereda JM, Calderwood DA, Ulmer TS, Critchley D, Campbell ID, Ginsberg MH, Liddington RC. Structural determinants of integrin recognition by talin.Mol Cell. 2003; 11:49–58.CrossrefMedlineGoogle Scholar
  • 11. 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; 117:2250–2259. doi: 10.1172/JCI31024.CrossrefMedlineGoogle Scholar
  • 12. Liu J, Das M, Yang J, Ithychanda SS, Yakubenko VP, Plow EF, Qin J. Structural mechanism of integrin inactivation by filamin.Nat Struct Mol Biol. 2015; 22:383–389. doi: 10.1038/nsmb.2999.CrossrefMedlineGoogle Scholar
  • 13. 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. doi: 10.1182/blood-2011-07-365601.CrossrefMedlineGoogle Scholar
  • 14. Berrou E, Adam F, Lebret M, Fergelot P, Kauskot A, Coupry I, Jandrot-Perrus M, Nurden A, Favier R, Rosa JP, Goizet C, Nurden P, Bryckaert M. Heterogeneity of platelet functional alterations in patients with filamin A mutations.Arterioscler Thromb Vasc Biol. 2013; 33:e11–e18. doi: 10.1161/ATVBAHA.112.300603.LinkGoogle Scholar
  • 15. Oegema R, Poulton CJ, Mancini GM. A single strand that links multiple neuropathologies in human disease.Brain. 2014; 137(pt 4):e266. doi: 10.1093/brain/awt197.CrossrefMedlineGoogle Scholar
  • 16. Bialkowska K, Byzova TV, Plow EF. Site-specific phosphorylation of kindlin-3 protein regulates its capacity to control cellular responses mediated by integrin αIIbβ3.J Biol Chem. 2015; 290:6226–6242. doi: 10.1074/jbc.M114.634436.CrossrefMedlineGoogle Scholar
  • 17. McGregor JL, Brochier J, Wild F, Follea G, Trzeciak MC, James E, Dechavanne M, McGregor L, Clemetson KJ. Monoclonal antibodies against platelet membrane glycoproteins. Characterization and effect on platelet function.Eur J Biochem. 1983; 131:427–436.CrossrefMedlineGoogle Scholar
  • 18. Lad Y, Jiang P, Ruskamo S, Harburger DS, Ylänne J, Campbell ID, Calderwood DA. Structural basis of the migfilin-filamin interaction and competition with integrin beta tails.J Biol Chem. 2008; 283:35154–35163. doi: 10.1074/jbc.M802592200.CrossrefMedlineGoogle Scholar
  • 19. Das M, Ithychanda SS, Qin J, Plow EF. Migfilin and filamin as regulators of integrin activation in endothelial cells and neutrophils.PLoS One. 2011; 6:e26355. doi: 10.1371/journal.pone.0026355.CrossrefMedlineGoogle Scholar

Highlights

  • A rare male patient with filaminopathy A is hemizygous for a C-terminal amino acid sequence extension of FLNa.

  • Platelets are heterogeneous in morphology, some being enlarged with vacuoles and abnormal granule distribution.

  • Mutant FLNa is expressed in all platelets; no wild-type FLNa is detected.

  • Patient platelets exhibit gain of functions on stimulation, including increased thrombus growth, aggregation, secretion, and αIIbβ3 integrin activation.

  • Platelet analysis and overexpression in the megakaryocytic cell line HEL demonstrate that mutant FLNa increases activation-dependent talin and kindlin-3 recruitment by and activation of αIIbβ3.