Strategies for Analysis of Novel Molecular Variants in the RUNX1 Gene As a Cause of Familial Platelet Disorder with Predisposition to Acute Myeloid Leukemia (FPD/AML)

Introduction: Runx 1 is a key transcription factor (TF) in hematopoiesis. Germline mutations in RUNX1 (≈40 described) are associated with FPD/AML, an autosomal dominant disorder characterized by moderate thrombocytopenia, platelet dysfunction and a high risk (40% before 35 yr) of developing acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS). Overestimating the pathogenicity of new molecular variants detected by high throughput sequencing (HTS) in the diagnosis of inherited thrombocytopenias (IT), even affecting well established candidate genes, could be more negative for the management and quality of life of patients than lacking such diagnosis (Lentaigne, Blood 2016). This is especially relevant for variants in TF such as Runx1, due to the risk of developing myeloid malignancies.

Objective: To evaluate the pathogenicity of new variants in RUNX1 identified in patients enrolled in the Spanish multicentric project "Functional and molecular characterization of patients with Congenital Platelet Disorders" (TPC-GT-PH-SETH).

Methods: Three unrelated women aged 27 (P1), 46 (P2) and 50 (P3) years, were included in TPC-GT-PH-SETH because of their history of mild thrombocytopenia (≈10⁵pl/uL) with normal volume (MPV) and moderate bleeding. A daughter and a brother of P2 died after developing MDS and LAM, respectively, and a sister of P3 had breast cancer. Only P1 and P2 had previously showed moderate functional platelet abnormalities (reduced aggregation, P-selectin secretion and number of δ-granules). By HTS (Bastida, Haematologica 2018) we identified new heterozygous variants in RUNX1: P1, de novo c.802C<T [p.Gln.268*]; P2, c.586A>G [p.Thr196Ala], segregated with thrombocytopenia and neoplasia in the family; P3, c.476A>G; [p.Asn159Ser], no segregation with thrombocytopenia, with 4 of 6 carriers in the family having normal platelet numbers.

For this study, we obtained blood from patients and three controls, and isolated ultrapure platelets by filtration+immunoselection. Platelet RNA was analyzed using Clariom-D Array (≈540 000 transcripts) (Caparrós-Pérez, PLoS One 2017). In P1 we isolated CD34+ cells from blood, differentiated them in vitro to megakaryocytes (Mks), and evaluated the formation of proplatelets.

Results: Analysis of transcriptomes showed that when considering the 13 genes well recognized as targets of RUNX1 (including MYL9, MYH9, ALOX12), the percentage of those with altered expression in patients, vs. controls, was 69.2% and 61.5% in P1 and P2, respectively, and only 7.7% in P3. Previously, the transcriptome of a single patient with RUNX1 variant (83aa deletion) (Sun, JTH 2007) has reported the infraexpression of 100 genes. Our cases showed low expression of 74.7%, 67.7% and 7.1% (P1, P2 and P3, respectively) of those genes. Principal component analysis of the array intensity signals, grouped together controls and the P3 case, while the P1 and P2 cases were clustered together. Additionally, in vitro culture of CD34-Mks from patient P1 showed reduced pro-platelet formation.

Conclusion: Platelet phenotyping, family segregation studies, platelet transcriptome and in vitro pro-platelet formation, support the pathogenicity of Runx 1 variants p.Gln268* and p.Thr196Ala, but not of p.Asn159Ser. Platelet transcriptome analysis is a useful tool in the pathogenic characterization of new molecular variants affecting TF genes in patients with IT, such as RUNX1 and can help to identify new target genes of these TF.