Fig 2: Proposed model showing the intriguing functional DEC2-HIF-1 association in the progression of osteosarcoma. In osteosarcomas, a combination of hypoxia and oncogenic signaling may increase the levels of HIF-1a, which upregulates DEC2 expression. DEC2 in turn facilitates the stabilization of HIF-1a, sensitizing osteosarcoma cells to hypoxic or oncogenic signaling and rapidly activating HIF-1 function.

Fig 3: Invasive U2OS-M cells have higher levels of DEC2 and accumulate HIF-1a more rapidly in response to hypoxia. A. In vitro migration and invasion activity of U2OS and U2OS-M. Migration and invasion activities were measured in vitro with trans-well chambers, as described in Materials and Methods. Photos are representative fields of invasive cells on the membrane. Magnification: 100 ×. B. Quantification of data shown in A. Bar graphs represent the average number of cells on the underside of the membrane (means ± s.e.m.) C. U2OS-M cells express higher level of DEC2 mRNA than the parental U2OS, while the HIF-1a mRNA level was not significantly affected as determined by qRT-PCR (P > 0.05). D. U2OS-M cells express higher level of DEC2 protein than U2OS cells. Cells were cultured under regular culture conditions or hypoxic condition (1 h) prior to harvest. Protein levels were determined by Western blotting. Under regular culture conditions, neither U2OS nor U2OS-M showed detectable HIF-1a. After 1 h hypoxic exposure, U2OS-M cells, which express higher level of DEC2, accumulated more HIF-1a than U2OS. E. U2OS-M Cells were transfected with either control or DEC2 siRNAs for 48 h. Migration and invasion activities were measured in vitro with trans-well chambers, as described in Materials and Methods. Photos are representative fields of invasive cells on the membrane. Magnification: 100 ×. F. Quantification of data shown in E. Bar graph presents the average numbers of cells on the underside of the membrane (means ± s.e.m.). G. DEC2 knockdown in U2OS-M slowed down HIF-1a accumulation. Cells were transfected with indicated siRNAs and cultured for 48 h. Cells were exposed to hypoxia for 1 h prior to protein sample collection. HIF-1a and DEC2 levels were determined by Western blotting, and ß-actin was determined for normalization of loading. ***: P < 0.001.

Fig 4: HIF-1 activation is sufficient to upregulate DEC2 in osteosarcomas under normoxic condition. A. Bioinformatics analysis of DEC2 promoter reveals that a region with highly acetylated histone H3 at K27 contains potential binding sites for multiple transcription regulators including HIF-1, c-Fos/Jun, c-Myc/Max and E2F. B. HIF-1 activation is sufficient to upregulate DEC2 levels under normoxic condition. 143B cells were cultured in hypoxia workstation or treated with 100 µM of DFX for 6 h prior to harvest to stabilize HIF-1a. Western blots show HIF-1a stabilization up-regulates DEC2.

Fig 5: Identification of the BHLHE40/41-responsive site in the promoter of the MIR301B-MIR130B cluster (A) Expression levels of MIR130B and MIR301B in HEC-1 and HEC-6 cells transfected with vectors to express HA-BHLHE40 and/or FLAG-BHLHE41. The lower panels show the protein expression of HA-BHLHE40 and FLAG-BHLHE41 using anti-HA or -FLAG antibody. (B) Expression levels of MIR130B and MIR301B in HHUA cells knocked-down with shBHLHE40 and/or shBHLHE41. The lower panels show the protein expression levels of BHLHE40/41. (C) Reporter analysis of the MIR301B-MIR130B promoter in HEC-6 cells transfected with HA-BHLHE40 and/or FLAG-BHLHE41. The control activity of the mutant reporter was adjusted to the same value as that of the wild-type reporter to evaluate the effects of BHLHE40/41 expression (C, white bars). (D) Reporter analysis of the MIR301B-MIR130B promoter in HHUA cells knocked-down with both shBHLHE40 and shBHLHE41. (E, upper panel) EMSA using nuclear extract from 293T cells transfected with HA-BHLHE40 and/or HA-BHLHE41. The nuclear extracts were incubated with labeled E-box2 probe (Supplementary Table 2). (E, lower panel) Nuclear extracts from 293T cells used for the EMSA were immunoblotted with an anti-HA antibody. (F) Nuclear extracts from 293T cells transfected with HA-BHLHE40 and FLAG-BHLHE41 were incubated with labeled E-box2 probe (Supplementary Table 2). Anti-HA or -FLAG antibody was used to form supershifted bands. An anti-SRF antibody was used as a negative control. Chromatin immunoprecipitation assay using 293T cells transfected with HA-BHLHE40 and FLAG-BHLHE41. Protein-DNA complexes immunoprecipitated with each of the anti-HA (G), -FLAG (H), -acetylated Histone H3 (I), and -HDAC1 (J) antibodies were used to amplify the E-box by PCR. The 10% input samples were used to calculate the occupancy ratio (%) from the values measured by real-time PCR. Data were representative from at least three experiments. FL, FLAG; E40, BHLHE40; E41, BHLHE41; AcH3, acetylated Histone H3; n.s., not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001.