Fig 2: The novel role of ATF5 in the HIF1 signaling pathway. a Detection of the expression level of identified proteins in shATF5 versus shnc cells by western blot in hypoxia. b Detection of the expression level of identified mRNAs in shATF5 versus shnc cells by RT-PCR; p < 0.01 in hypoxia. c Evaluating the secretion of VEGFA in the identified cells by ELISA; p < 0.01 in hypoxia. d Dual-luciferase assays were performed to detect the luciferase activation of identified genes in ESCA cells transfected with shnc and shKDM4C; p < 0.01 in hypoxia. e Investigating the tube formation ability of HUVECs induced by supernatant from medium of identified cells in hypoxia; Scale bar 100 µm; p < 0.01

Fig 3: Osteogenesis was impaired in Mut-H USCs but could be alleviated by ATF5 knockdown.A, B ATF5 knockdown increased WNT7B and GSK3B gene expression. ATF5 was knockdown by siRNA for 48 h and the mRNA levels of WNT7B and GSK3B were quantified through RT-qPCR. C–E Expressions of osteogenesis-related markers were increased by ATF5 knockdown in all USCs. USCs were transfected with siRNA specific to ATF5 for 48 h and cultured in osteogenic induction medium for 48 h. mRNA levels of RUNX2, OCN and BMP2 were quantified through RT-qPCR. F Alizarin Red S staining of calcium levels in Ctr, Mut-L, and Mut-H USCs. USCs were cultured in osteogenic induction medium with or without siRNA transfection specific to ATF5 for 48 h and cultured in osteogenic induction medium for 21 days. Cells were stained with Alizarin Red S and imaged with dissecting microscope under bright field. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Scale bar = 50 µm.

Fig 4: ATF5 knockdown reversed mitochondrial function in Mut-H USCs.A ATF5 knockdown decreased UPRmt related gene expression. USCs were transfected with siRNA specific to ATF5 for 48 h and the mRNA levels of UPRmt related genes (ATF5, HSP70, HSP60 and LONP1) expression were quantified by RT-qPCR. B Mitochondrial membrane potential was increased by ATF5 knockdown. USCs were transfected with siRNA specific to ATF5 for 48 h and mitochondrial membrane potential was measured using fluorescence probe JC-1 assay system. Red color indicates cells with normal mitochondrial membrane potential while green indicates cells with loss of mitochondrial membrane potential. C Quantification plot for B. D Intracellular ROS was decreased by ATF5 knockdown. USCs were transfected with siRNA specific to ATF5 for 48 h and stained with DCFH-DA. Green color indicate ROS-positive cell population. E Quantification plot for D. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Scale bar = 50 um.

Fig 5: ATF5 promotes odontoblast terminal differentiation by activating a Dmp1-related enhancer. Primary mDPCs of E18.5 mice cultured in mineralization induction medium for 0 days (D0) and 9 days (D9) after infection with lentivirus-expressing shRNA against Atf5 and empty control. (A) qRT-PCR and Western blot showing expression of related odontoblast markers in different conditions. (B) Alizarin red staining showing fewer mineralization nodules on D9 in shAtf5 group compared with control. (C) A chromatin region in the downstream of Dmp1 (Dmp1+13) gradually gained accessibility (belonged to cluster 3 in Figure 3) in time–course ATAC-seq with a typical binding motif for the ATF family. (D) Tn5-transposase footprint based on CENTIPEDE revealed the ATF-binding motif in (C) “protected” on D9 ATAC-seq. (E) Dual-luciferase assay showing the enhancer activity of Dmp1+13 could be promoted by the overexpression of ATF5 in mDPC cells. *p < 0.05, **p < 0.01, ***p < 0.001.