Fig 2: Hsa-miR-449c-5p suppresses PIGV mRNA level.A, diagram to show the number of shared microRNAs using ENCORI program predicting microRNAs regulating PIGV and deep sequencing of microRNAs with increased abundance in Tg-treated AsPC-1 cells. B, statistical analysis of QPCR results of microRNAs 103a-3p, 449a, 449c-5p, and 532-5p levels in DMSO, TG, and RE AsPC-1 cells. C, statistical analysis of QPCR results of the effects of mimic or inhibitor of microRNAs 103a-3p, 449c-5p, and 532-5p treatment on the mRNA levels of PIGV in AsPC-1 cells. D, immunoblot analysis of the effects of mimic or inhibitor of microRNAs 103a-3p, 449c-5p, and 532-5p on the protein levels of PIGV in AsPC-1 cells. E, flow cytometry analysis of the sensitivity of AsPC-1 or TG cell surface PrP to PI-PLC treatment when cells are treated with 449c-5p mimic or inhibitor, respectively. F, luciferase activity assay of PIGV 3'-UTR (diagram row 1) or mutated 3'-UTR (diagram row 3) when 293T cells are treated with mimic or inhibitor of microRNA449c-5p or control. G, statistical analysis of wound healing and Matrigel results of mimic or inhibitor of microRNA449c-5p treatment on the migration and invasion of AsPC-1 or TG cells, respectively. The data were expressed as mean ± SD and were analyzed by two-tailed double sample heteroscedasticity Student's t test. *p< 0.05, **p< 0.01 and ***p < 0.001. NC, non-targeting miRNA mimic or inhibitor control; NS, not significant; PIGV, phosphatidylinositol-glycan biosynthesis class V protein; PI-PLC, phosphoinositide-specific phospholipase C; PrP, prion protein; Tg, thapsigargin.

Fig 3: Tg treatment increases ATF6 to reduce PIGV, resulting in the accumulation of pro-PrP.A, immunoblot analysis of BiP, XBP-1s, ATF6, c-ATF6, p-PERK, PERK, p-eIF2a, eIF2a, and ATF4 for cell lysates made from DMSO, TG, and RE AsPC-1 cells. B, immunoblot analysis of XBP-1s, ATF6, and PIGV (left panels) and statistical analysis (right panel) of PIGV after siRNA silencing of XBP-1 and ATF6 in TG cells. C, statistical analysis of QPCR results of the effects of XBP-1 and ATF-6 silencing on PIGV mRNA in TG cells. D, flow cytometry analysis of the effects of XBP-1 and ATF-6 silencing on the sensitivity of cell surface PrP to PI-PLC treatment. E, statistical analysis of wound healing and Matrigel results of ATF-6 silencing on the migration and invasiveness of TG cells. The data were expressed as mean ± SD and were analyzed by two-tailed double sample heteroscedasticity Student's t test. ATF6, activating transcription factor 6; BiP, binding immunoglobulin protein; c-ATF6, cleaved activating transcription factor 6; eIF2a, eukaryotic translation initiation factor 2 alpha; NC, non-targeting siRNA control; NS, not significant; PIGV, phosphatidylinositol-glycan biosynthesis class V protein; PI-PLC, phosphoinositide-specific phospholipase C; PERK, pancreatic ER kinase-like ER kinase; p-eIF2a, phosphorylated eIF2a; PrP, prion protein; Tg, thapsigargin.

Fig 4: Tg treatment reduces PIGV resulting in the accumulation of pro-PrP.A, QPCR analysis of GPI-anchor synthesis–related gene expression of DMSO, TG, and RE AsPC-1 cells. Genes with significant higher abundance of two-fold differences are circles in red rectangle with dashed line. B, immunoblotting analysis of PIGV, PIGZ, and GDPD5 of DMSO, TG, and RE AsPC-1 cells (upper panel). Statistical analysis of immunoblot results (bottom panel). C, immunoblotting analysis of siRNA-silenced AsPC-1 cells. D, flow cytometry analysis of cell surface PrP after siRNA silencing PIGV and PIGZ in AsPC-1 cells treated with or without PI-PLC. E, flow cytometry analysis of cell surface PrP after siRNA silencing PIGV in MNC with or without PI-PLC. Statistical analysis of flow cytometry results is shown. Relative surface PrP after PI-PLC treatment defined as: (geometry mean of PrP intensity after PI-PLC treatment - background geometry mean of PrP intensity)/(geometry mean of PrP intensity before PI-PLC treatment - background geometry mean of PrP intensity). F, flow cytometry analysis of cell surface PrP after expressing an HA-tagged PIGV or PIGZ in TG cells with or without PI-PLC. G, statistical analysis of motility and invasiveness assays after silencing PIGV in AsPC-1 cancer cells. H, statistical analysis of motility and invasiveness assays after expressing an HA-tagged PIGV in AsPC-1-Tg cells. The data were expressed as mean ± SD and were analyzed by two-tailed double sample heteroscedasticity Student's t test. *p< 0.05 and **p< 0.01. GDPD5, glycerophosphodiester phosphodiesterase domain-containing protein 5; NS, not significant; NC, non-targeting siRNA control; MNC, mouse neural cell line; PrP, prion protein; PIGV, phosphatidylinositol-glycan biosynthesis class V protein; PI-PLC, phosphoinositide-specific phospholipase C; PIGZ, , phosphatidylinositol-glycan biosynthesis class Z protein; Tg, thapsigargin; Vec, empty vector control.

Fig 5: ATF6–miR449c-5p–PIGV axis contributes to PDAC in humans and marks prognosis in patients with PDAC.A, immunohistochemical analysis of ATF6 and PIGV and in situ hybridization analysis of microRNA449c-5p expression in different stages from matched PDAC biopsies (upper panels). Statistical results of the immunohistochemical and in situ hybridization results between stage I+II and stage III PDAC are shown (bottom panels). Scale bars represent 200 µm. B, Kaplan–Meier survival plot of PDAC patients based on the immunohistochemical and in situ hybridization analysis results. p-values and number of biopsies are indicated. The statistical significance of prognosis analysis was assessed using the Mantel-Cox test. The other data were expressed as mean ± SD and were analyzed by two-tailed double sample heteroscedasticity Student's t test. ATF6, activating transcription factor 6; PDAC, pancreatic ductal cell adenocarcinoma; PIGV, phosphatidylinositol-glycan biosynthesis class V protein.