The mammalian TRAP/Mediator complex is a master transcriptional regulatory complex that integrates signals of diverse activators and recruits RNA polymerase II and other general factors to activate transcription. The TRAP220/MED1 subunit was originally identified as a ligand-dependent coactivator specific for nuclear receptors. We have previously shown through biochemical and mouse genetic studies that MED1 is essential for embryogenesis, cell growth/differentiation and homeostasis, and that it stimulates nuclear receptor-mediated myelomonopoiesis. MED1 also integrates other activators such as GATA-1 and C/EBPβ and appears to mediate erythropoiesis as well. The niche cells in the bone marrow plays a pivotal role in the maintenance of hematopoietic stem/progenitor cells (HSPCs). In this study, we employed mouse embryonic fibroblasts (MEFs) as a model to analyze the role of MED1 in the niche, since MEFs have a mesenchymal feature with the osteoblastic precursor lineage and are known to support HSPCs. To establish an experimental system, we crossed Med1 and p53 double knockouts to obtain Med1+/+/p53−/− and Med1−/−/p53−/− E10.0 embryos from a single female and prepared stable MEF lines. Then the Med1−/−/p53−/− MEFs were stably transfected with a MED1 expression vector (Rev-Med1−/− MEFs) or a control empty vector. When normal mouse bone marrow cells were cocultured with these MEFs treated with mitomycin C for a short period of 2 weeks, cell counts, live cells (MTT assay) and a DNA synthesis (BrdU incorporation) of marrow cells were measured. The number of live cells as well as DNA synthesis on Med1−/− MEFs was significantly decreased during this period, but those on Rev-Med1−/− MEFs recovered to the control levels. Thus the growth stress on MEFs appears to be attenuated on Med1−/− MEFs. When apoptosis of the marrow cells was measured, both the FITC-dUTP incorporation by TdT and annexin V/PI double positive cells were lower for Med1−/− MEFs, indicating that apoptosis was also attenuated. We next assessed the role of MED1 in MEFs to support long-term bone marrow culture. After bone marrow cells were cultured on mitomycin C-treated MEFs for 8 weeks in Myelocult M5300 (StemCell Technologies) or IMDM supplemented with BIT9500 (StemCell Technologies) and LDL, progenitor cells (adherent and nonadherent) were collected and cultured in complete methylcellulose (Methocult M3434; StemCell Technologies), and colonies were counted. The number of both myeloid and erythroid colonies were significantly attenuated (0 to 40% depending on experimental conditions) for cells on Med1−/− MEFs, but colonies for cells cultured on Rev-Med1−/− MEFs recovered to the control level. In order to exclude the possibility that lot differences among MEFs or p53 depletion might have affected the results, we next prepared primary Med1+/+ and Med1−/− MEFs by crossing Med1+/− mice and conducted the long-term culture experiments using these MEFs. The attenuated number of colonies for cells cocultured with Med1−/− MEFs (circa 10% of the control) was reproduced repeatedly, indicating that the observed role of MED1 in MEFs to support HSPCs is intrinsic. Since MED1 converges signals from a series of activators on specific promoters and activates transcription, one or some products of the downstream target genes in MEFs may be responsible for the observed activity to maintain HSPCs. In search for candidate MED1 target gene products among a series of known molecules that possess an activity on HSPCs, only the expression of osteopontin was found to be attenuated in Med1−/− MEFs and reverted in Rev-Med1−/− MEFs. Other factors including Angiopoietin-1 and Jagged-1 were comparable. This fact contrasts with the previous observation of osteopontin knockouts where the null niche cells that restricted the size of HSPC number overexpressed these factors. We next assessed the role of MED1 on the osteopontin promoter. We focused on vitamin D receptor (VDR) and Runx2 among the activators and tested MEFs by luciferase reporter assays. The basal level of transcription without any activators in Med1−/− MEFs was about half of the control. Moreover, both the activation by Runx2 and the liganddependent VDR function were significantly attenuated in Med1−/− MEFs. These results indicate that transcriptional coactivator MED1 in niche cells plays an important role in HSPCs support, and that osteopontin may be one of the downstream candidate target genes for MED1.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author