Validation of Grin1 knockout (KO) in mouse MKs. (A) Schematic demonstrating the Grin1 KO strategy. (A) The Grin1 gene locus (i) and the GluN1 protein (ii) are drawn showing the targeted portion for deletion in blue. Exons 11 to 22 surrounded by the loxP sites are shown in panel Ai; black arrows indicate location of the genotyping primers. The protein was truncated approximately at amino acid (aa) 400 (ii, blue arrow). The deletion encompassed the transmembrane (TM) region containing 4 TM domains (TM I-IV) and the C-terminus of GluN1. (B-C) Bar graphs showing the relative level of Grin1 deletion by Cre recombinase (B) and Grin1 messenger RNA expression (C) examined by quantitative PCR in Pf4-Grin1^−/− MKs, calculated relative to WT MKs. Genomic DNA was amplified using primer sets targeting a nondeleted region of Grin1 (exons 7-8) and the loxP-flanked deleted region of Grin1 (exons 13-14); levels of deleted Grin1 were normalized to nondeleted Grin1 and made relative to WT (B); transcript levels of Grin1 were normalized to GusB and made relative to WT (C). Four independent experiments were performed, in triplicate. Bar graphs show mean ± SEM. Statistical significance is shown (2-tailed Student t test). (D) RT-PCR results showing expression of NMDAR subunit transcripts Grin2A, Grin2D, and Grin3B in WT and Pf4-Grin1^−/− MKs; transcripts for other NMDAR subunits were not expressed. Mouse brain tissue was used as positive control. Negative control contained no DNA template (ie, no template control [NTC]). Actin B (Actb) served as a loading control. (E) Immunostaining of BM MKs (formalin-fixed paraffin-embedded sections) for the MK marker CD61 (red) and the GluN1 protein (green). An overlap between CD61 and GluN1 staining was seen in WT but not Pf4-Grin1^−/− mice (an example is indicated by arrows). Cell nuclei were counterstained with Hoechst 33358 (blue). Three independent experiments were performed. Scale bar is shown. (Fi) Semiquantitative western blotting results demonstrating the level of the GluN1 signal estimated relative to β-actin in washed platelets from WT and Pf4-Grin1^−/− mice. Results from 4 experiments are shown, each relative to the mean of controls within that run (normalized to 100%). Green symbols mark platelet samples processed by negative immunomagnetic selection procedure to deplete them of red cells (using anti–TER-119) and white cells (using anti-CD45 antibodies). (Fii) Western blot examples showing GluN1 and β-actin signals in washed platelets (lanes 1-5; 4 and 5 were loaded in duplicate) and in negatively selected platelets (lanes 6-9). (G-H) Ca²⁺ responses to NMDAR agonists in WT and Pf4-Grin1^−/− platelets loaded with fura-2-AM. Baseline values were recorded for 30 seconds, and cells were stimulated by the addition of NMDA (100 μM) (G) and glutamate (500 μM) (H). Values are reported ratiometrically as 340 nm (Ca²⁺ bound) over 380 nm (Ca²⁺ free) (F340/F380) after being normalized to (1) at baseline. (Gi-Hi) Line graphs showing the mean relative levels of intracellular Ca²⁺ recorded over 120 seconds. (Gii-Hii) Corresponding scatterplots showing the peak relative levels of intracellular Ca²⁺ during the observation period (mean ± SEM). Data represent triplicate values from 3 biologic replicates per genotype. Statistical significance is shown (1-way analysis of variance with Dunnett post hoc for panels Gi and Hi and 2-tailed Student t test for panels Fi, Gii, and Hii). gDNA, genomic DNA; MW, molecular weight; ns, not significant.