Fig 2: Functional relevance of rs372883 variants. (A) Relative reporter gene activity bearing BACH1 3'UTR fragment with the rs372883T or rs372883C allele in CFPAC-1 and BXPC-3 cells. Results are mean ± SEM from three experiments and each had six replicates. The P-values are for Student's t-test. (B) In silico prediction of interaction between miR-1257 and BACH1 3'UTR showing differences in binding within the seed region. (C) Relative reporter gene activity of the psiCHECK2-rs372883T and psiCHECK2-rs372883C constructs cotransfected with 1.0, 5.0 and 10.0 pmol of miR-1257 or its inhibitor in CFPAC-1 and BXPC-3 cells. Results are mean ± SEM from three experiments and each had six replicates; P-values were for Student's t-test. (D-F) Levels of BACH1 mRNA, miR-1257 and HMOX1 mRNA in normal pancreatic tissues adjacent to tumors of subjects with the rs372883 CC (N=11), CT (N=38) or TT (N=26) genotype. Results are mean ± SEM relative to GAPDH or U6. (G) Western blot analysis of AKT and ERK signaling modules downstream of HO-1 in surgically removed PDAC specimens from subjects with the rs372883CC, CT or TT genotype.

Fig 3: Expression of BACH1 in PDAC. (A) BACH1mRNA levels determined by qRT-PCR were significantly lower in PDAC compared with paired adjacent normal tissues (N=75). Results represent mean ± SEM normalized to GAPDH and P-values are for Student's t-test. (B) BACH1 protein levels determined by Western blot were significantly lower in PDAC (T) than in paired normal tissues (N). Shown is a representative picture selected from 35 pairs of clinical specimens. (C) Representative IHC images (200×) showing BACH1 protein expression in normal pancreatic tissues (left) and PDAC (right). Bar scale, 100 μm. (D-E) Expression of BACH1 in published data sets of Pei et al.27 and Badea et al. 28 (Oncomine, https://www.oncomine.org/resource/main.html). The line in the middle of the box represents the median; bars represent 10th and 90th percentiles and greater values were plotted as individual points. P-values are for unpaired Wilcox rank-sum test.

Fig 4: BACH1 expression was regulated by circ_0081001/miR-494-3p axis. (A) The binding site between miR-494-3p and the 3’UTR of BACH1 mRNA. (B) The binding relationship between miR-494-3p and BACH1 mRNA 3’UTR was verified by dual-luciferase reporter gene experiment. (C–F) The effects of circ_0081001 and miR-494-3p on the expression levels of BACH1 mRNA and protein were examined by qRT-PCR and Western blot after the transfection. ** and *** represent P<0.01 and P<0.001, respectively.

Fig 5: Circ_0081001, miR-494-3p, and BACH1 were potential regulators in OS. (A) Volcano map was used to indicate the difference of the expression levels of miRNAs between normal bone tissues and OS tissues in GSE28423, and the upregulated and downregulated miRNAs expressions were marked with red and green, respectively. (B) Venn diagram showed the intersection of predicted miRNA targets of circ_0081001 and lowly expressed miRNAs in OS tissues in GSE28423. (C) MiRmap, microT, miRanda, and TargetScan were employed for predicting the downstream target genes of miR-494-3p. (D) The potential biological function of miR-494-3p was predicted utilizing DAVID GO enrichment analysis. (E) Volcano map was used to indicate the difference of the gene expression profile between human normal osteoblasts and OS cell line in GSE11414, and the upregulated and downregulated gene expressions were marked with red and green, respectively. (F) Venn diagram showed the intersection of predicted target genes of miR-494-3p and highly expressed genes in OS tissues in GSE11414.