Supplemental Table and Figure

Table S1 Predicted miRNA clusters targeting ER stress regulatory genes

miRNA name / Positions of miRNA binding sites
HSPA5
(GRP78) / EIF2AK3
(PEAK) / ERN1
(IRF1a) / ATF6 / ATF4 / DDIT3
(CHOP) / XBP1
hsa-miR-199 / 345-351 / 429-435 / 186-192
hsa-miR-181 / 84-90
hsa-miR-329 / 240-246 / 878-884
hsa-miR-379 / 309-315
hsa-miR-30-5p / 142-148 / 590-596
hsa-miR-224 / 947-953
hsa-miR-488 / 954-960 / 346-352
hsa-miR-495 / 334-341
hsa-miR-33a / 980-986
hsa-miR-124 / 113-119
hsa-miR-200a / 248-255
hsa-miR-1283 / 70-76
hsa-miR-96 / 80-86

MiRNAs that target the 3’UTR of these ER stress regulatory genes’ mRNA were predicted using the bioinformatics tool Targetscan 6.2

Figure S1:miR-199a-5p expression in mouse hepatocyte cell line and liver

(A). The relative level of miR-199a-5p and miR-199b-5p in mousetransformedhepatocyte BNL-CL2 cells after thapsigargin (TG) or deoxycholic acid (DCA) or dimethyl sulfoxide (DMSO) treatements. (B). The relative level of miR-199a-5p and miR-199b-5p in the mouse liver 24 hr after bile duct ligation (BDL) or sham operation or no treatment. The relative expression levels of RNA were quantified by qRT-PCR and were normalized to U6. Data are shown as means ± SD of three independent experiments. ** P<0.01.

Figure S2:The basal expression efficiency of 3’UTR reporter vectors.

The expression efficiency of 3’UTR reporter constructs were determined byluciferase activity 24 hr after transfection intoHEK293 cells. WT: wild-type; Mut: mutant 3’UTR reporter. Data are shown as means ± SD of three independent experiments.

Figure S3:Luciferase mRNA expression levels after miR-199a-5p trasnsfection

Luciferase reporters that had either the wild-type 3’ UTRs of GRP78, ATF6 and IRE1αor mutant UTRs with a 6-base pair (bp) deletion in the target sites were co-transfected with miR-199a-5p mimics or negative control (NC) RNA into HEK293 cells. The luciferase mRNA levels were assayed 48 hr after co-transfection with qRT-PCR (By adding DNAse into purified RNA to remove possibly residual plasmids) and were normalized to GAPDH. Data are shown as means ± SD of three independent experiments. ** P<0.01.

Figure S4 RNA interfering efficiency of siRNAs (si-) targeting to GRP78, ATF6 or IRE1α

RNA interfering efficiency of siRNAs (si-) targeting to GRP78, ATF6 or IRE1α in HL-7702/L02 cells were accessed by qRT-PCR. Data are shown as means ± SD of three independent experiments. ** P<0.01.

Figure S4: Cell death levels of siRNAs- or NC-RNA-transfected HL-7702/L02 cells after DCA or TG treatment.

Cell death levels of siRNAs- (si-) or NC-RNA-transfected HL-7702/L02 cells were assayed by LDH release after DCA or TG treatment for 48 hr. Data are shown as means ± SD of three independent experiments. ** P<0.01, ref to NC-RNA-transfected cells.

Figure S5:Predicted AP-1-binding sites in the miR-199A2 promoter

Predicted AP-1-binding sites in the miR-199A2 promoter are shown. Reporter vectorΔ-1 deleted AP-1 binding site 1. Reporter vectorΔ-2 containing mutant AP-1 binding site 2 and the mutant 7 bases were shown.

Figure S6: The absoluteluciferase activity values of MIR199A2 promoter assays

Reporter vectors containing wild-type (WT) or AP-1-site-1-deleted (Δ1) or AP-1-site -2-mutant (Δ2)promoters were constructed, and absolute luciferase activities were detected in hepatocytes after thapsigargin (TG), deoxycholic acid (DCA) or dimethyl sulfoxide (DMSO) treatment. Data are shown as means ± SD of three independent experiments.