Profilin 1–mediated cytoskeletal rearrangements regulate integrin function in mouse platelets

Platelet adhesion and aggregation at sites of vascular injury is essential for hemostasis but may also cause thrombosis.¹  Firm platelet adhesion is mediated by heterodimeric receptors of the β1- and β3-integrin families, which upon activation reversibly shift to a high-affinity state and efficiently bind their ligands, most notably components of the extracellular matrix and other receptors.^2-4  Binding of talin-1 (Tln1) and kindlins to the intracellular tail of the integrin β-subunit triggers the switch to the high-affinity state, whereas their dissociation results in integrin closure, the switch back to the low-affinity state.^5-8  Tln1 and kindlins connect integrins to the actin cytoskeleton, thereby enabling the sensing and exertion of mechanical forces as well as regulating adhesion formation and turnover.^9-12  Consistently, defects in actin-regulating proteins result in altered platelet and megakaryocyte integrin function.^13-19  Furthermore, we have recently shown a critical role of twinfilin 2a (Twf2a) and the cortical cytoskeleton in regulating platelet integrin turnover in a profilin 1 (Pfn1)–dependent manner.¹⁰ 

The small actin-binding protein Pfn1 is central for actin dynamics by mediating the nucleotide exchange on G-actin monomers, thereby promoting filament assembly with implications for platelet biogenesis.^13,20  Megakaryocyte-specific Pfn1 deficiency resulted in microthrombocytopenia because of cytoskeletal alterations and accelerated platelet clearance (supplemental Figure 1).¹³  However, the precise role of Pfn1 for platelet function is unknown.

Here, we report that the lack of Pfn1 in platelets (Pfn1^fl/fl-Pf4Cre) perturbs the organization of the adhesion-dependent circumferential actin network and thereby results in accelerated integrin inactivation and hence impaired platelet function in vitro and in vivo.

Animals, flow cytometry, aggregation, immunostaining, adhesion under flow, clot retraction, immunoblotting, tail bleeding assay, in vivo model, Ca²⁺ measurements, atomic force microscopy, and statistical analyses are described in the supplemental Methods.¹³  Animal studies were approved by the district government of Lower Franconia (Bezirksregierung Unterfranken).

To study the role of Pfn1 in integrin function, platelet β1- and β3-integrin activation was assessed by flow cytometry at different time points. Strikingly, Pfn1^fl/fl-Pf4Cre platelets showed impaired activation of β1- and β3-integrins (Figure 1A-B; supplemental Figures 2 and 3), whereas α-granule release (assessed by P-selectin exposure; Figure 1C), as well as surface expression of β1- and β3-integrins, was only marginally altered (supplemental Figure 4).

The β3-integrin activation defect was further revealed by a decreased binding of Alexa-F488-conjugated fibrinogen to Pfn1^fl/fl-Pf4Cre platelets at all tested time points in response to agonist stimulation (Figure 1D; supplemental Figure 5). Of note, irreversible fibrinogen binding was almost indistinguishable between WT and Pfn1^fl/fl-Pf4Cre platelets, suggesting that receptor clustering is not altered by the Pfn1 deficiency (supplemental Figure 6). Surprisingly, aggregation responses of Pfn1^fl/fl-Pf4Cre platelets were only slightly affected by the impaired integrin activation (Figure 1A,D-E; supplemental Figures 2, 5, and 7), whereas the adhesion and aggregate formation of Pfn1^fl/fl-Pf4Cre platelets on collagen I under flow were markedly reduced (Figure 1F-G; supplemental Figure 8). We speculate that the reduced integrin function of Pfn1^fl/fl-Pf4Cre platelets is still sufficient to mediate platelet aggregation under static conditions, but insufficient to resist shear forces generated under flow because of an impaired connection of integrins and the actin cytoskeleton.

Pfn1^fl/fl-Pf4Cre platelets displayed impaired integrin outside-in signaling, as evidenced by almost abolished lamellipodia (WT, 58.0% ± 3.0% vs Pfn1^fl/fl-Pf4Cre, 18.5% ± 4.3%; ***P < .001) and filopodia formation (WT, 46.9% ± 1.4% vs Pfn1^fl/fl-Pf4Cre, 11.6% ± 1.5% ; ***P < .001) when allowed to spread on fibrinogen (supplemental Figure 9).^13,21  Similarly, Pfn1^fl/fl-Pf4Cre platelets showed severely delayed and less effective clot retraction (Figure 1H), further reflecting impaired integrin function and actin dynamics.^10,13  Because Pfn1 is an effector of the RhoA/Rho-associated kinase (ROCK) pathway, these findings are in agreement with reports on RhoA-deficient platelets that show abolished clot retraction.²²  Furthermore, the strongly impaired filopodia and lamellipodia formation of Pfn1^fl/fl-Pf4Cre platelets suggests that Pfn1 acts as a key effector in actin dynamics and integrin function downstream of Rho guanosine triphosphatases (GTPases).²³ 

In agreement with the in vitro data, Pfn1^fl/fl-Pf4Cre mice displayed prolonged bleeding times (13.3 ± 2.7 minutes) as compared with controls (7.2 ± 2.3 minutes). Two out of 10 Pfn1^fl/fl-Pf4Cre mice even failed to cease bleeding within the observation period (Figure 1I). Consistently, upon induction of arterial thrombosis by mechanical injury of the abdominal aorta, all control mice formed occlusive thrombi (4.0 ± 1.2 minutes), whereas 5 out of 6 Pfn1^fl/fl-Pf4Cre mice did not form stable occlusions (Figure 1J; supplemental Figure 10), thus revealing Pfn1 as a critical factor for platelet function in hemostasis and thrombosis. These defects appeared to be caused by the defective integrin function of Pfn1^fl/fl-Pf4Cre platelets, because degranulation was normal (Figure 1C) and alterations in coagulation factors or other cell types could be excluded because we capitalized on a megakaryocyte- and platelet-specific conditional knockout mouse model.

Ca²⁺ signaling is essential for platelet integrin activation; however, contrary to a functional impairment we found an increased basal [Ca²⁺]_i and enhanced store-operated Ca²⁺ entry in Pfn1^fl/fl-Pf4Cre platelets (Figure 1K-L). This was most likely due to impaired actin dynamics in the absence of Pfn1,¹³  because inhibition of activation-dependent actin-polymerization was shown to increase store-operated Ca²⁺ entry.²⁴ 

Besides Ca²⁺ signaling, Tln1 recruitment to β-integrin tails represents a key step in integrin activation. Whereas in spread control platelets Tln1 and β3-integrins colocalized at the cell cortex, we found a diffuse colocalization pattern in Pfn1^fl/fl-Pf4Cre platelets, suggesting impaired β3-integrin localization or Tln1 recruitment as cause for the observed defects (Figure 2A-B). Strikingly, the localization of Tln1 and β3-integrins at the cell cortex of control platelets coincided with a high cellular stiffness, reminiscent of the cortical cytoskeleton.²⁰ Pfn1^fl/fl-Pf4Cre platelets, however, only displayed focal spots of increased stiffness on the cell cortex, strongly resembling the random colocalization pattern of Tln1, β3-integrins, and the disrupted cortical actin cytoskeleton (Figure 2A-D; supplemental Figure 11). These results indicate that the impaired cytoskeletal dynamics in Pfn1^fl/fl-Pf4Cre platelets may account for the altered recruitment to, or stabilization of, integrins and/or Tln1 at the leading edge of spread platelets.^9-13 

We recently showed that the organization of the circumferential actin-cytoskeleton modulates calpain-mediated inactivation of activated platelet β3-integrins and thereby also controls their localization.¹⁰  Based on this, together with the disrupted cortical actin-cytoskeleton and the enhanced Ca²⁺ signaling in Pfn1^fl/fl-Pf4Cre platelets, we hypothesized that Tln1 and β3-integrins might be more prone to calpain-mediated cleavage and that this could account for the altered localization and impaired function of integrins in Pfn1^fl/fl-Pf4Cre platelets. In support of this, we found a significantly increased percentage of phosphatidylserine-positive Pfn1^fl/fl-Pf4Cre platelets with inactivated integrins (JON/A⁻ Anxa5⁺; lower right quadrant) after both 5 minutes (WT, 18.2% ± 5.6% vs Pfn1^fl/fl-Pf4Cre, 32.6% ± 4.1%; ***P < .001) and 30 minutes (WT, 50.9% ± 8.7% vs Pfn1^fl/fl-Pf4Cre, 65.6% ± 4.3%; **P < .01) upon stimulation with thrombin and CRP (Figure 2E-F; supplemental Figure 12A,C). Similar observations were made upon stimulation with CRP alone, but not when using the A23187 ionophore that efficiently induces phosphatidylserine exposure without integrin activation (Figure 2G; supplemental Figure 12A-B,D-F). Remarkably, stimulated Pfn1^fl/fl-Pf4Cre platelets displayed an increased loss of β3-integrin, Tln1, and filamin A as compared with control that could be fully rescued by chemical inhibition of calpain or actin assembly, suggesting that enhanced calpain-mediated integrin inactivation due to an insufficient cytoskeletal linkage may account for the observed defects in Pfn1^fl/fl-Pf4Cre platelets (Figure 2H-K; supplemental Figures 13 and 14). In support of this, pretreatment of platelets with the actin polymerization inhibitor cytochalasin D overall reduced β3-integrin activation but, most importantly, diminished the differences between WT and Pfn1^fl/fl-Pf4Cre platelets (Figures 1A and 2L; supplemental Figure 2). Moreover, chemical inhibition of calpain improved the spreading of Pfn1^fl/fl-Pf4Cre platelets and restored the localization as well as activation of β3-integrins and Tln (Figure 2M-N).

In summary, these results revealed that the defective organization of the cortical cytoskeleton in Pfn1^fl/fl-Pf4Cre platelets leads to accelerated integrin inactivation and hence impaired platelet function. Based on the patchy colocalization pattern of Tln1, β3-integrins, and the disrupted cytoskeleton in Pfn1^fl/fl-Pf4Cre platelets, it is tempting to speculate that Pfn1 might act as an organizer of the adhesion-dependent, circumferential actin network, thereby linking integrins to the underlying cytoskeleton and limiting calpain-mediated integrin inactivation.²⁴  In support of this, we have found increased Pfn1 activity in Twf2a^−/− platelets, which resulted in a thickened cortical cytoskeleton and sustained integrin activation due to limited calpain-mediated inactivation.¹⁰  However, an increased calpain activity due to the impaired actin dynamics in Pfn1^fl/fl-Pf4Cre platelets also needs to be considered.

In summary, our findings highlight a central role of Pfn1-mediated actin rearrangements for normal platelet integrin function and turnover with implications for hemostasis and thrombosis.

The authors thank Stefanie Hartmann for excellent technical assistance and the microscopy platform of the Bioimaging Center (Rudolf Virchow Center) for providing technical infrastructure and support.

This work was supported by the Deutsche Forschungsgemeinschaft (SFB688 and NI 556/11-1) (B.N.). S. Stritt was supported by a research fellowship of the Deutsche Forschungsgemeinschaft (STR 1538/1-1) and a nonstipendiary long-term fellowship of the European Molecular Biology Organization (ALTF 86-2017). M.B. is supported by an Emmy Noether grant of the Deutsche Forschungsgemeinschaft (BE5084/3-1).

Contribution: S. Stritt designed research, performed experiments, analyzed data, and wrote the manuscript; I.B. performed experiments, analyzed data, and contributed to the writing of the revised manuscript; S.B., S. Sorrentino, K.T.S., J.H., J.H.H., A.B., and M.B. performed experiments, analyzed data, and commented on the manuscript; H.S. and O.M. analyzed data and commented on the manuscript; J.V., F.G.-I., and X.D. provided vital reagents and commented on the manuscript; and B.N. designed and supervised research, analyzed data, and wrote the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

The current affiliation for S. Stritt is Department of Immunology, Genetics and Pathology, Uppsala University, Uppsala, Sweden.

Correspondence: Bernhard Nieswandt, Institute of Experimental Biomedicine, University Hospital and Rudolf Virchow Centre, University of Würzburg, Josef-Schneider-Str 2, 97080 Würzburg, Germany; e-mail: bernhard.nieswandt@virchow.uni-wuerzburg.de.