Platelets lacking PIP5KIγ have normal integrin activation but impaired cytoskeletal-membrane integrity and adhesion

Abstract

Three isoforms of phosphatidylinositol-4-phosphate 5-kinase (PIP5KIα, PIP5KIβ, and PIP5KIγ) can each catalyze the final step in the synthesis of phosphatidylinositol-4,5-bisphosphate (PIP2), which in turn can be either converted to second messengers or bind directly to and thereby regulate proteins such as talin. A widely quoted model speculates that only p90, a longer splice form of platelet-specific PIP5KIγ, but not the shorter p87 PIP5KIγ, regulates the ligand-binding activity of integrins via talin. However, when we used mice genetically engineered to lack only p90 PIP5KIγ, we found that p90 PIP5KIγ is not critical for integrin activation or platelet adhesion on collagen. However, p90 PIP5KIγ-null platelets do have impaired anchoring of their integrins to the underlying cytoskeleton. Platelets lacking both the p90 and p87 PIP5KIγ isoforms had normal integrin activation and actin dynamics, but impaired anchoring of their integrins to the cytoskeleton. Most importantly, they formed weak shear-resistant adhesions ex vivo and unstable vascular occlusions in vivo. Together, our studies demonstrate that, although PIP5KIγ is essential for normal platelet function, individual isoforms of PIP5KIγ fulfill unique roles for the integrin-dependent integrity of the membrane cytoskeleton and for the stabilization of platelet adhesion.

Figure 1

PIP5KIγ colocalizes with talin in primary megakaryocytes. (A) Hematopoietic progenitor cells were differentiated into megakaryocytes and infected with a retrovirus that directed the expression of the GFP-p90 PIP5KIγ fusion protein. Shown are confocal micrographs demonstrating the intracellular distribution of (top) GFP-PIP5KIγ and (middle) talin. The bottom panel is a merged image of GFP-PIP5KIγ and talin suggesting that PIP5KIγ colocalizes with talin in primary megakaryocytes. (B) Megakaryocytes expressing either p87 or p90 GFP-PIP5KIγ fusion protein were stained with mouse anti-talin and rabbit anti-GFP antibodies followed by both secondary immunogold particles, conjugated anti-rabbit IgG (10 nm), and anti-mouse IgG (5 nm). The small gold particles demonstrate the localization of talin (blue circles), and the big gold particles (red squares) demonstrate the distribution of the GFP-PIP5KIγ fusion protein. The magnification bar is 100 nm. (C) The frequency of talin colocalizing with p87 or p90 PIP5KIγ.

Figure 2

Generation of mice lacking either p90 PIP5KIγ alone or lacking both PIP5KIγ isoforms. (A) The PIP5KIγ E17 conditional targeting vector contains loxP sites flanking exon 17. These mice were crossed with PF4 or CMV promoter-driven Cre recombinase. (B) RT-PCR and immunoblotting confirmed that platelets from homozygous floxed PIP5KIγ E17^−/− PF4 Cre⁺ mice did not express detectable p90 PIP5KIγ mRNA or protein. Similar results were obtained using platelet mRNA or lysates derived from PIP5KIγ E17^−/− CMV Cre⁺ mice (not shown). (C) Conditional rescue of the PIP5KIγ gene trap. The diagram shows the location of the β-geo trap within the first intron of the PIP5KIγ gene. The insertion leads to a read-through mutation within the first intron of the targeted gene and truncated PIP5KIγ after the 32nd amino acid. The location of the LoxP sites flanking the splice acceptor is shown. (D) PIP5KIγ expression is seen in brain and in platelet lysates. The anti-PIP5KIγ immunoblot shows the complete loss of protein in all analyzed tissues derived from the PIP5KIγ^−/− MLC-2v Cre⁺ mice.

Figure 3

Platelets lacking p90 PIP5KIγ have impaired anchoring of their cell membrane with the cytoskeleton. (A) Platelets were analyzed for their ability to form membrane tethers by an optical trap. The figure illustrates that pulling on the membrane-attached fibrinogen-coated bead will only stretch the membrane when membrane-bound integrins are not anchored to the underlying cytoskeleton. (B) The ability of PIP5KIγ E17^−/− PF4 Cre⁺ platelets to form membrane tethers is shown. Tethers were rarely seen in PIP5KIγ E17^−/− PF4 Cre⁻ control platelets. (C-D) PIP5KIγ E17^−/− CMV Cre⁺ and PIP5KIγ E17^−/− CMV Cre⁻ platelets were perfused over immobilized (C) collagen or (D) fibrinogen at a wall shear stress of 1 dyn/cm² for 10 minutes. After changing to platelet-free buffer, the wall shear stress was increased every 20 seconds up to 40 dyn/cm², and the platelet covered area was measured. The data in panels C and D represent the mean ± standard error of the mean of 5 experiments. *P < .05 relative to wild-type platelets at the same wall shear stress.

Figure 4

Platelets lacking PIP5KIγ display normal integrin function. (A) Washed murine platelets lacking PIP5KIγ that were analyzed after agonist stimulation in a Lumi-dual aggregometer. Platelets lacking PIP5KIγ have normal aggregation in response to collagen, PMA, thrombin, and AYP, a peptide agonist of PAR4 (the dominant murine thrombin receptor). Results are representative of 3 experiments. (B) Loss of PIP5KIγ does not impair the binding of Jon/A, an antibody that only recognizes the activated form of αIIbβ3. The results are derived from 3 experiments. (C) Following stimulation of washed murine platelets with thrombin, platelets were fixed, permeabilized, and stained with fluorescent phalloidin. Flow cytometry was used to quantitate phalloidin binding in 100 000 cells, and analysis was performed using FlowJo software. Shown is the mean ± standard error of the mean for 3 experiments. (D) PIP5KIγ^−/− MLC-2v Cre⁺ platelets that were layered onto fibrinogen and allowed to adhere for the labeled time periods. Cells were lysed, and immunoblotted with anti-FAK or anti-phosphorylated FAK antibodies.

Figure 5

Complete loss of both PIP5KIγ isoforms disrupts the integrin-dependent integrity of the membrane cytoskeleton and the stabilization of platelet adhesion. (A) Differential interference contrast micrograph that shows that PIP5KIγ-null platelets form membrane tethers by pulling on a fibrinogen-coated bead that was touched to the surface of the cell and then moved apart. (B) PIP5KIγ^−/− platelets that were analyzed for their ability to form membrane tethers were quantified by an optical trap. The graph illustrates that pulling on the membrane-attached fibrinogen-coated bead will only stretch the membrane when membrane-bound integrins are not anchored to the underlying cytoskeleton. Comparable results are obtained with platelets lacking PIP5KIγ (PIP5KIγ^−/− MLC-2v Cre⁺) or wild-type platelets exposure to a pharmacologic G-actin sequestering agent, Latrunculin A. Tethers were rarely seen in wild-type platelets. (C-D) Wild-type platelets or platelets lacking PIP5KIγ were perfused over immobilized collagen at a wall shear stress of 1 dyn/cm² for 10 minutes. After changing to platelet-free buffer, the wall shear stress was increased every 20 seconds up to 40 dyn/cm², and the platelet covered area was measured. (C) Representative micrograph of platelet-covered areas remaining at a wall shear stress of 40 dyn/cm². (D) Data for all wall shear stresses. (E) Wild-type platelets or platelets lacking PIP5KIγ were perfused over immobilized von Willebrand factor at the indicated wall shear stress, and mean rolling velocities were measured. The data in panels D and E represent the mean ± standard error of the mean of 5 experiments. *P < .05; **P < .01 relative to wild-type platelets at the same wall shear stress.

Figure 6

Mice lacking PIP5KIγ form unstable thrombi in vivo. Ferric chloride was applied to an exposed carotid artery of anesthetized mice, and the formation of thrombi was monitored using a Doppler flowmeter. Mice lacking (A) p90 PIP5KIγ or (B) both PIP5KIγ isoforms form thrombi as rapidly as control mice. (C) However, mice lacking both p87 and p90 PIP5KIγ form thrombi that are frequently unstable.