Fig 2: Calibration curves and evaluation of AP-F13A1 fragments by LC-PRM assays. a, b Graphical plot representation of theoretical concentration spanning according to the experimental peak ratio light/heavy peptides for AVPPNNSNAAEDDLPTVELQGVVPR and AVPPNNSNAAEDDLPTVELQGLVPR. c Coefficients of variations (CVs) of the light/heavy ratios distribution for the two peptides, over the 6 concentration points. Serum concentration of AP-F13A1 in healthy patients and patients with declared CRC from the first sample bank (d) and the second sample bank (e). The calculated means were 181.8 ± 71.0 and 187.2 ± 92.2 ng/mL for the controls, and 59.4 ± 33.3 and 70.4 ± 38.4 ng/mL for the cancer patients, both showing significant Wilcoxon tests (****P value = 0.0001). Receiver Operating Characteristic curves (ROC) for AP-F13A1 from first sample bank (f) and the second one (g). AUC computing were 0.95 (0.89–1.00) and 0.93 (0.87–0.98) and calculated values of sensitivity/specificity were 75%/90% and 71%/95%, respectively

Fig 3: Clot contraction in megakaryocytic HEL cells and CD34+ megakaryocytes. (A) Clot contraction in HEL cells without (−) or with (+) PMA treatment. HEL cells grown with 30 nM PMA for 4 days were used to evaluate clot contraction for up to 120 minutes, representative of 2 separate experiments. (B) Clot contraction in phorbol 12-myristate 13-acetate (PMA)-treated human erythroleukemia (HEL) cells following small inhibitor RNA (siRNA) knockdown (KD) of RUNX1 or F13A1. Cells were suspended in buffer with added red blood cells and 50 mg/mL fibrinogen, and clotting was initiated with 5 U/mL thrombin and 10 mM CaCl2. The line graph represents percent contraction up to 120 minutes; shown are the mean ± SEM of 4 independent experiments. The extent of contraction was determined from the pictures using ImageJ software. (C) Clot contraction in PMA-treated HEL cells was suspended in platelet-poor plasma (PPP) with added red blood cells, and clotting was initiated with 5 U/mL thrombin and 10 mM CaCl2. Shown is a representative experiment. The line graph shows the mean ± SEM of the extent of clot contraction over 120 minutes from 4 independent experiments. (D) Immunoblots of HEL cell lysates for RUNX1 and factor (F)XIII-A in control cells and following siRNA KD of RUNX1 or F13A1 with β-actin as loading control. The bar graphs show quantification of protein bands normalized to β-actin (n = 4). (E) FXIII-A surface expression in resting and convulxin + thrombin-activated HEL cells in RUNX1 and F13A1 siRNA KD by flow cytometry. The bar graph shows the mean fluorescent intensity (MFI) from 3 independent experiments (mean ± SEM). (F) Shown are immunoblots of cell lysates for RUNX1, FXIII-A, and β-actin as a loading control and quantification of protein bands normalized to β-actin (n = 3). NS, not significant.

Fig 4: Binding of RUNX1 to F13A1 promoter region and functional activity in phorbol 12-myristate 13-acetate (PMA)-treated human erythroleukemia (HEL) cells. (A) The 5’ upstream region of F13A1 with 7 RUNX1 consensus binding sites. TSS, transcription start site. (B) Binding of RUNX1 to F13A1 promoter region by chromatin immunoprecipitation using control immunoglobulin (IgG) or RUNX1 antibody (Ab). Polymerase chain reaction was performed with primers designed to amplify the F13A1 promoter region encompassing RUNX1 binding sites 1 through 7 and GAPDH. Shown are polymerase chain reaction amplification of control IgG, RUNX1, total input DNA, and genomic DNA (gDNA; columns 1 to 4, respectively). (C) Binding of RUNX1 to F13A1 promoter region by electrophoretic mobility shift assay. Shown is the electrophoretic mobility shift assay using an oligonucleotide probe with RUNX1 binding sites 1 and 4, a probe with sites 5, 6, and 7, and nuclear extracts from HEL cells. For each panel: lane 1, probe alone; lane 2, probe with nuclear extract; lane 3, competition with ×100 molar excess unlabeled probe; lane 4, the effect of control IgG; and lane 5, the effect of RUNX1 Abs. Arrows indicate areas where the band was supershifted or competed by the RUNX1 Abs. The nucleotide sequences of the probes are shown in Supplementary Table S5. (D) Luciferase reporter studies on the F13A1 promoter. Shown are luciferase promoter activity at 24 hours of the wild-type construct and the effect of mutating each of the 7 RUNX1 consensus binding sites in the 5’ upstream region of F13A1 (Supplementary Table S6). The ratios of luciferase to renilla activity of the wild-type construct and constructs with each RUNX1 consensus site mutated are shown. Site-directed mutagenesis of the RUNX1 binding sites on F13A1 promoter decreased promoter activity with the exception of site 5, which seemed to exhibit negative regulation. (E) Effect of RUNX1 overexpression on F13A1 promoter activity. PMA-treated HEL cells were transfected with the empty pCMV6 vector or RUNX1B-pCMV6 expression vector together with the F13A1-promoter-PGL3-luciferase vector. Shown are luciferase promoter activities at 24 hours compared with that of cells transfected with promoterless PGL3 basic vector (mean ± SEM of 3 experiments) and immunoblots of cell lysates for factor (F)XIII-A and RUNX1 with β-actin as a loading control with quantification bar graphs at the bottom. (F) Effect of RUNX1 small inhibitor RNA (siRNA) on F13A1 promoter activity. PMA-treated HEL cells were cotransfected with control or RUNX1 siRNA together with F13A1-promoter-PGL3 luciferase reporter constructs. Reporter activity was measured at 24 hours. The mean ± SEM of 3 experiments performed in triplicate are shown. Also shown are immunoblots of cell lysates for FXIII-A and RUNX1 with β-actin as loading control; bar graphs show quantification from 3 experiments.

Fig 5: F13A1 expression in platelets and plasma in familial platelet disorder with predisposition to myeloid malignancy (FPDMM) patients. (A) Platelet F13A1 mRNA levels by quantitative polymerase chain reaction in healthy controls and patient P1 (father) and 2 unrelated patients (brother and sister). Shown are mRNA levels normalized to GAPDH. (B) Factor (F)XIII-A by immunostaining in platelets from patient P1 and a healthy control. The bar graph shows normalized corrected total cellular fluorescence (mean ± SEM) calculated from images. (C) Immunoblot analysis showing FXIII-A expression in platelet lysates from patient P1 and 4 healthy controls (N1-N4) with β-actin as a loading control. The bar graphs below show FXIII-A quantification normalized to actin. (D) Immunoblot analysis showing FXIII-A expression in platelet lysates from 2 unrelated siblings and 4 healthy controls. The bar graphs below show FXIII-A quantification normalized to actin.