RAB31-Mediated Endosomal Trafficking Is Defective in RUNX1 Haplodeficiency

RAB GTPases are key players in vesicle trafficking, granule targeting of proteins, granule biogenesis and secretion. RAB31 (Ras-related protein 31) has been implicated in the regulation of vesicular trafficking between the Golgi/TGN and endosomes, and from early endosome (EE) to late endosome (LE). Studies in neuronal cells have implicated RAB31 in the transport of epidermal growth factor receptor (EGFR) from EE to LE. RUNX1, a major hematopoietic transcription factor, plays a critical role in megakaryocyte (MK) maturation and platelet production. Patients with RUNX1 haplodeficiency have thrombocytopenia, abnormal platelet function, and impaired granule contents and secretion. Our studies in a patient with a heterozygous RUNX1 mutation (c.969-323G>T ) (Sun et al, Blood 103: 948-54, 2004) associated with thrombocytopenia, platelet dysfunction, granule deficiency and impaired platelet responses revealed decreased platelet expression of RAB31. The role of RAB31 in MK and platelets is unknown. We addressed the hypothesis that RAB31 is a transcriptional target of RUNX1. We studied the role of RAB31 on vesicle transport in MK cells in human erythroleukemia (HEL) cells transformed into megakaryocytic cells by PMA (phorbol 12-myristate 13-acetate). RAB31 mRNA was decreased on platelet expression profiling of the patient (fold change: 0.28, p<0.0076, Sun et al J Thromb Haemost 5: 146-154). With real-time PCR platelet platelet RAB31 mRNA was decreased compared to 5 healthy controls by 60-80% in our patient (P1) and in two additional unrelated patients (siblings, P2 and P3) with RUNX1 mutation (c.508+1G>A). Platelet RAB31 protein was decreased compared to that in 5 healthy controls in patients P2 and P3. RAB31 promoter region (-2023/-1bp from the ATG) revealed 4 RUNX1 consensus sites: site I (-813/-808), site II (-972/-967), site III (-1500/-1495) and site IV (-2007/-2002). Chromatin immunoprecipitation (ChIP) revealed RUNX1 binding to RAB31 sites II and IV but not to other sites. Electrophoretic mobility shift assays using HEL cell proteins showed RUNX1 binding to sites II and IV. In luciferase reporter assays, mutation of individual sites II and IV decreased promoter activity indicating that they are functional sites. RAB31 promoter activity and protein expression were inhibited by RUNX1 siRNA and enhanced by RUNX1 overexpression. These indicate that RAB31 is a direct RUNX1 target, providing a mechanism for decreased RAB31 in patient platelets. We investigated putative roles of RUNX1 and its target RAB31 in endosomal dynamics in PMA treated HEL cells. We used immunofluorescent staining for markers of early endosomes (EE, EEA1) and late endosomes/multivesicular bodies (LE/MVB, CD63) in RUNX1- or RAB31-depleted cells by siRNA transfection. Either RUNX1 or RAB31 siRNA yielded a striking enlargement of early endosomes, as indicated by the EE marker EEA1. This finding suggests a role for RUNX1/RAB31 in EE maturation, either by mediating vesicle fission, or maturation to late endosomes by fusion with other endosomal vesicles. This effect of RUNX1 knockdown on EE enlargement was partially reversed by reconstitution of RAB31 by plasmid co-transfection, indicating that RAB31 is a significant but non-exclusive contributor to this RUNX1 function.

Conclusions: These studies provide the first evidence that RAB31 is a direct transcriptional target of RUNX1 and a mechanism for RAB31 downregulation in RUNX1 haplodeficient patients. Downregulation of RAB31 or RUNX1 results in impaired endosomal maturation/trafficking, and this may contribute to the defective handling of α-granule proteins recognized in patients with RUNX1 mutations.