Fig 2: PAUF activates the MyD88-dependent, not the TRIF-dependent, TLR4 downstream pathway. (A) Panc-1_Mock and Panc-1_TLR4OE cells were treated with rPAUF (0.1 μg/mL) for 0, 10, 30, and 60 min. Collected proteins were immunoprecipitated with MyD88 antibody (α-MyD88) using protein A agarose beads. Immunocomplexes were determined by Western blot against MyD88 and TLR4. (B) Panc-1_Mock and Panc-1_TLR4OE cells were treated with rPAUF (0.1 μg/mL) for 0, 60, and 120 min. Collected proteins were immunoprecipitated with TRIF antibody (α-TRIF) using protein A agarose beads. Immunocomplexes were determined by Western blot against TRIF and TLR4. (C) Western blot showed that Panc-1_trans TLR4OE cells were successfully engineered to transiently overexpress TLR4, compared to the control, Panc-1_trans Ctrl cells. (D) Panc-1_trans Ctrl and Panc-1_trans TLR4OE cells were pretreated with or without IKK inhibitor XII (5 μM) for 4 h before rPAUF (0.1 μg/mL) or PBS treatment. After overnight culture, cells were collected for dual-luciferase activity assay. Data are represented as a mean ± SD from triplicates (**** p < 0.0001 were obtained from two-way ANOVA and post-hoc multiple comparisons with Bonferroni correction). (E) Panc-1_Mock and Panc-1_TLR4OE cells were treated with rPAUF (0.1 μg/mL) for 0, 10, 30, 60, 120, and 240 min, and activation of TRIF-dependent TLR4 downstream molecules (p-TBK1 and p-IRF3) was estimated using Western blot.

Fig 3: Positive TLR4 expression is detected in six different pancreatic cancer cell lines but not the HPDE normal pancreatic cell line. (A) mRNA expression levels of TLR4 quantified by RT-qPCR from the current study and retrieved from the Cancer Cell Line Encyclopedia (CCLE) Expression 22Q2 Public database. (B) Protein expression level of TLR4 assessed by Western blot analyses, the two bands of TLR4 appeared to be glycosylated (130 kDa) and deglycosulated (100 kDa) TLR4 [9,10]. (C) Relative mRNA levels of TLR1~9 in two representative pancreatic cancer cell lines. The mRNA expression levels were adjusted based on the two cell lines’ TLR4 data from (A). Data are represented as a mean ± SD from triplicates, ** p < 0.01, **** p < 0.0001 (indicating differences between Panc-1 and BxPC-3) were obtained from two-way ANOVA and post-hoc multiple comparisons with Bonferroni correction.

Fig 4: TLR4 inhibitor (TAK-242), at non-cytotoxic concentrations, reduces migration of pancreatic cancer cells expressing high level of TLR4 (BxPC-3). (A) Migration of BxPC-3 and Panc-1 cells, and (B) cell viability of BxPC-3 and Panc-1 cells measured after treated with TAK-242, a specific TLR4 inhibitor, at different concentrations for 24 h. Migration was presented as % of control, and cell viability was determined using WST-1 assay. Data are represented as mean ± SD from multiple replicates. The Jonckheere-Terpstra test was conducted to indicate the trend of cell migration’s change with increasing TAK-242 concentration after a significant multiple comparisons test (* p < 0.05, *** p < 0.001, **** p < 0.0001, compared to control, obtained from one-way ANOVA and post-hoc multiple comparisons with Dunnett correction). Scale bars, 50 μm.

Fig 5: PAUF’s impacts on pancreatic cancer cell migration and invasion are exclusively dependent on TLR4. Migration of Panc-1_Mock and Panc-1_TLR4OE were estimated using transwell assays after treatment of (A) recombinant PAUF (rPAUF, 0.5 μg/mL) or (B) PAUF antibody (α-PAUF, 20 μg/ml) for 24 h. (C) Invasion of Panc-1_Mock and Panc-1_TLR4OE were determined using transwell assays after treatment of rPAUF (0.5 μg/mL) or α-PAUF (20 μg/ml) for 24 h. (D) Migration of pancreatic cancer cells induced by rPAUF in presence or absence of TAK-242, a TLR4 inhibitor. Data are represented as mean ± SD from multiple replicates. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 were obtained from two-way ANOVA and post-hoc multiple comparisons with Bonferroni correction. Scale bars, 50 μm.