Fig 2: HIFa downregulation by tunicamycin is UPR independent(A, C)PERK+/+, PERK-/-, U87 scrambled (Scr) and U87 IRE1a knock down cells were cultured under normoxic conditions (16% O2) for 24 h, then treated with Tu (10 µg/ml) and further cultured either under normoxia or hypoxia (5% O2) for 4 h. HIF-1a or HIF-2a levels measured under normoxia (16% O2) were set to 1. *significance 16% O2 vs 5% O2, **significance 16% O2 vs 16% O2 + Tu, #significance 5% O2 vs 5% O2 + Tu, p = 0.05, n = 3. (B, D) Representative Western blot of total proteins analyzed with antibodies against HIF-1a, HIF-2a, total PERK, IRE1a and a-tubulin. (E) HepG2 cells were treated with Tu (10 µg/ml) and further cultured for 4 h under normoxic (16% O2) or hypoxic (5% O2) conditions. The GRP78, phospho PERK (pPERK), total PERK (PERK), ATF-6a, IRE1a, and a-tubulin protein levels were measured by Western blot analysis. In each experiment the protein levels under 16% O2 were set to 1. *significance 16% O2 vs 5% O2, **significance 16% O2 vs 16% O2 + Tu, #significance 5% O2 vs 5% O2 + Tu, p = 0.05, n = 3. (F) Representative Western blots of total protein analyzed with antibodies against GRP78, pPERK, PERK, ATF-6a, IRE1a and a-tubulin. (G) HEK 293 cells were cultured under normoxia (16% O2) for 24 h, then treated with Tu (10 µg/ml) and further cultured either under normoxia or hypoxia (5% O2) for 4 h. Total RNA was extracted, and XBP1 mRNA was detected by RT-PCR. XBP1u was observed as a 442-bp band, and XBP1s was observed as a 416-bp band. ß-actin served as a loading control.

Fig 3: Association of BiP differential expression with breast cancer clinical factors and pathways. (A) Analysis of BiP mRNA expression in breast tumor tissue compared to normal and metastatic tissues using gene-chip- or RNA-seq-based data available in TNMplot; (B) Analysis of BiP mRNA expression across PAM50 subtypes using TCGA-BRCA data in R. ANOVA followed by Tukey’s test: significant difference at 5% level of significance. (C) Analysis of BiP protein levels across breast cancer subtypes using the CPTAC dataset in cBioPortal. (D) BiP correlation with clinical factors. BiP protein abundance in the CPTAC dataset was used to stratify the BiP higher and lower quartiles (BiP-H and BiP-L) and an analysis carried out using cBiopPotal. ER—estrogen receptor, PR—progesterone receptor, TNBC—triple-negative breast cancer, and ERBB2—receptor tyrosine-protein kinase erbB-2.

Fig 4: Association of BiP differential expression (high vs. low; BiP-H and BiP-L, respectively) with tissue-infiltrating immune and stromal cell populations using the mcp_count method from the immunodeconv package and TCGA-BRCA data overall.

Fig 5: Representative pictures of BiP immunostaining and respective H&E in treatment-naïve primary tumors vs. their paired metastasis. Arrows indicate cytosolic BiP expression while arrowheads show BiP nuclear localization in neoplastic cells. All pictures were from the same invasive ductal carcinoma case, as seen in the corresponding H&E slide showing irregular ducts infiltrating the breast stroma. Scale bars: 50 µm. The bar graphs comparing %BiP positive cells, BiP IHC score, and BiP IHC intensity of staining in paired treatment untreated tumors vs. post-treatment metastases (n = 12). *: p < 0.05, **: p < 0.01; Wilcoxon paired test.