Interplay between the tyrosine kinases Chk and Csk and phosphatase PTPRJ is critical for regulating platelets in mice

Abstract

The Src family kinases (SFKs) Src, Lyn, and Fyn are essential for platelet activation and also involved in megakaryocyte (MK) development and platelet production. Platelet SFKs are inhibited by C-terminal Src kinase (Csk), which phosphorylates a conserved tyrosine in their C-terminal tail, and are activated by the receptor-type tyrosine phosphatase PTPRJ (CD148, DEP-1), which dephosphorylates the same residue. Deletion of Csk and PTPRJ in the MK lineage in mice results in increased SFK activity, but paradoxically hypoactive platelets resulting from negative feedback mechanisms, including upregulation of Csk homologous kinase (Chk) expression. Here, we investigate the role of Chk in platelets, functional redundancy with Csk, and the physiological consequences of ablating Chk, Csk, and PTPRJ in mice. Platelet count was normal in Chk knockout (KO) mice, reduced by 92% in Chk;Csk double KO (DKO) mice, and partially rescued in Chk;Csk;Ptprj triple KO (TKO) mice. Megakaryocyte numbers were significantly increased in both DKO and TKO mice. Phosphorylation of the inhibitory tyrosine of SFKs was almost completely abolished in DKO platelets, which was partially rescued in Src and Fyn in TKO platelets. This residual phosphorylation was abolished by Src inhibitors, revealing an unexpected mechanism in which SFKs autoinhibit their activity by phosphorylating their C-terminal tyrosine residues. We demonstrate that reduced inhibitory phosphorylation of SFKs leads to thrombocytopenia, with Csk being the dominant inhibitor in platelets and Chk having an auxiliary role. PTPRJ deletion in addition to Chk and Csk ameliorates the extent of thrombocytopenia, suggesting targeting it may have therapeutic benefits in such conditions.

Graphical abstract

Figure 1.

Severe macrothrombocytopenia in DKO mice is partially rescued in TKO mice. (A) Representative blots of capillary-based immunoassays on platelet lysates with the indicated antibodies. (B) Representative data from panel A displayed as electropherograms. (C) Quantification of peak areas of Csk signal from (A-B), n = 6 mice/genotype. (D) Median fluorescence intensity (MFI) measured in αIIb⁺ cells costained for PTPRJ in blood, n = 6 mice/genotype. (E) Platelet counts, n = 14 mice/genotype. (F) Platelet volumes, n = 14 mice/genotype. (G) Spleen/body weight ratio, n = 15 mice/genotype. Chk^−/− (CKO), Chk^−/−;Csk^fl/fl;Pf4-Cre⁺ (DKO), Chk^−/−;Csk^fl/fl;Ptprj^fl/fl;Pf4-Cre⁺ (TKO). *Significant difference compared with control, or significant difference between DKO and TKO. **P < .01, ***P < .001, 1-way ANOVA with Sidak test; mean ± SD. MW, molecular weight.

Figure 2.

Aberrant megakaryocyte counts and myelofibrosis in DKO and TKO mice. Histological sections of mouse bone marrow (A), spleen (B), and lung (C) of the indicated genotypes stained with hematoxylin and eosin (H&E), the megakaryocytic marker CD42b, or reticulin. MKs are highlighted by yellow arrows. (D) Total number of MKs per section in the indicated organs, n = 5 mice/genotype. (E) MK numbers per field of view in the indicated organs, n = 5 mice/genotype (mean of 15 fields/mouse). Analysis of MK numbers was performed in a double-blinded manner. *P < .05, **P < .01, ***P < .001, 1-way ANOVA with Sidak test; mean ± SD.

Figure 3.

Reduced expression of (hemi-)ITAM-containing receptors, elevated immature platelet fraction and platelet activation markers in DKO and TKO mice. (A) Median fluorescence intensity (MFI) measured in αIIbβ3⁺ cells or αIIb⁺ cells costained for the indicated proteins in blood, n = 6 mice/genotype. (B) Representative image of reticulated platelet population as measured by flow cytometry. αIIb⁺ platelets in whole blood were gated and the percentage of reticulated platelets was determined in gate P1 (in red), in dot plot diagram forward scatter (FSC) vs RNA dye (Thiazole Orange). (C) Percentage of reticulated platelets as determined in panel B. (D) Representative image of desialylated platelet population. αIIb⁺ platelets in whole blood were gated and the percentage of desialylated platelets was determined in gate P1 (red), in dot plot diagram FSC vs peanut agglutinin lectin. (E) Percentage of desialylated platelets as determined in panel D. (F) Percentage of P-selectin⁺αIIb⁺, TLT-1⁺αIIb⁺, fibrinogen⁺αIIb⁺, or Annexin V⁺αIIb⁺ cells in blood as determined by flow cytometry, n = 6 mice/genotype. *P < .05, **P < .01, ***P < .001, 1-way ANOVA with Sidak test; mean ± SD.

Figure 4.

Defective platelet activation and hemostasis in DKO mice is restored in TKO mice. (A) Anti-P-selectin-FITC, anti-triggering receptor expressed on myeloid cells-like transcript 1 (TLT-1)-FITC and fibrinogen-488 binding to washed platelets (2 × 10⁷/mL) following stimulation with or without PAR4 peptide (AYPGKF) (100 μM or 500 μM, 20 minutes, room temperature) was measured by flow cytometry. The fold increase of the MFI relative to the corresponding unstimulated platelets was calculated, n = 5-7 mice/genotype. (B) Representative phalloidin-stained images of resting (basal) and thrombin-stimulated (0.1 U/mL, 5 minutes) platelets spread on fibrinogen-coated cover-slips (100 μg/mL, 45 minutes, 37°C, scale bar: 5 μm). (C) Mean surface area of individual platelets quantified by KNIME software, n = 6 mice/genotype (200 to 450 platelets/condition). (D) Hemostatic response was measured in saline tail bleeding assay by an excision of a 3-mm portion of the tail tip followed by immersion of the tail in 0.9% isotonic saline at 37°C. Plotted is the time to complete arrest of bleeding. Experiments were conducted in a double-blinded manner, n = 16 mice/genotype. *P < .05, **P < .01, ***P < .001, 1-way ANOVA with Sidak test; mean ± SD.

Figure 5.

Partial rescue of Src and Fyn inhibitory tyrosine phosphorylation in TKO platelets. (A) Representative blots of capillary-based immunoassays on lysates of resting platelets with the indicated antibodies. (B) Quantification of peak areas, n = 6 mice/genotype. *P < .05, **P < .01, ***P < .001, 1-way ANOVA with Sidak test; mean ± SD.

Figure 6.

SFKs auto-phosphorylate their C-terminal inhibitory tyrosine residues. (A) Representative blots of capillary-based immunoassays on platelet lysates with the indicated antibodies. Lysates were generated from platelets incubated with DMSO, 50 nM dasatinib, 10 μM PP1, or 3 μM PRT-060318 for 15 minutes, room temperature. (B) Quantification of peak areas, n = 6 mice/genotype. Asterisks refer to significant difference compared with DMSO-treated control samples within genotypes. *P < .05, **P < .01, ***P < .001, 2-way ANOVA with Sidak test; mean ± SD.

Figure 7.

Revised model of regulation of SFKs in the megakaryocyte lineage. (A) SFK activity is tightly regulated by the coordinated action of the tyrosine kinases Csk and Chk, the receptor-type tyrosine phosphatase PTPRJ, and SFKs themselves. Csk and Chk negatively regulate SFK activity, whereas PTPRJ and SFKs are dual positive and negative regulators of SFKs. (B) SFKs autoregulate their catalytic activity through the trans-phosphorylation of conserved tyrosine residues in the activation loop and C-terminal tail. (C) An equilibrium of different states of SFKs is established in resting and activated platelets through the interplay of these tyrosine kinases and phosphatase, as indicated. This is dependent on the concentrations, proximity, and catalytic activities of these enzymes.