Fig 2: Raloxifene delays resumption of protein translation post-hypoxia in correlation with increased eIF2α phosphorylation and decreased mTOR signaling.A and B Puromycin incorporation assay. U251 cells were treated with 40 μM raloxifene or DMSO vehicle control for 1 h prior to 2 h hypoxic A or non-hypoxic B incubation. At various times post-hypoxic or normoxic incubation cells were treated with 10 μg/mL puromycin for 10 min and puromycin incorporation into nascent proteins was detected by anti-puromycin blot. Blots were normalized to total lane protein, and represented as ratios with DMSO time 0 normalized to 1 (bottom quantification panels). Data is presented as the mean of triplicates ± SEM, unpaired t-test *p < 0.05; **p < 0.01. C and D U251 cells were treated as in A but without addition of puromycin. Cell lysates were probed for eIF2α, phospho-eIF2α, GADD34, rpS6, and phospho-rpS6 at various times post-hypoxic C or normoxic D incubation. All blots were normalized to total lane protein with DMSO time 0 set to 1 (bottom quantification panels). The level of eIF2α phosphorylation is presented as the ratio of p-eIF2α to total eIF2α (bottom quantification panels). The level of rpS6 phosphorylation is similarly presented. Data is presented as the mean of triplicates ± SEM, unpaired t-test, and false discovery rate of 1%. *p < 0.05; **p < 0.01.

Fig 3: Hypoxic stress responses are activated in low-grade astrocytoma and GBM and correlate to poor outcomes.A TCGA database was interrogated with GlioVis for mRNA expression levels of genes involved in activation of the ISR (PERK, GCN2), in nucleating SGs (G3BP1) and terminating the ISR and triggering SG disassembly (GADD34) in grade II, III, and IV astrocytoma. Statistical analysis using Tukey’s honest significant difference and performed on GlioVis, ***p < 0.001, **p < 0.01, and *p < 0.05. B Kaplan–Meir survival curves for high and low mRNA expression of PERK, GCN2, G3BP1, GADD34 in low-grade astrocytoma. p-values adjusted for multiple comparisons. C Graphical representation of IHC cytoplasmic granular scoring of a TMA consisting of 90 GBM cores from 45 patients stained with SG markers G3BP2 and TIAR. Grading was assigned 0–4, no staining, minority of cells, ~50% of cells, majority of cells or all cells (excluding vascular endothelium), respectively. Example of staining grades on left panel. D Examples of IHC G3BP2 and TIAR-stained GBM and normal cortex showing cytoplasmic granular staining (blue arrowheads). The vascular endothelium encasing tumor is visible (red arrow) with negative staining of G3BP2 and TIAR. White boxes denote magnified areas. Black line represents 100 μm. E IHC example of GBM core stained for TIAR and G3BP2 demonstrating necrosis (right of orange dotted line) and increased granular staining in the cells adjacent to the necrotic core. Black boxes magnified and granular punctate demonstrated by blue arrow heads. Black line is 100 μm.

Fig 4: GADD34 activity is required for lysosomal biogenesis following starvation.(A) Lysosomal biogenesis assay in a model of Fabry disease, i.e., HeLa cells knocked out for a-galactosidase (GLA) gene. Storage accumulation [globotriaosylceramide (Gb3)] discriminates small newly formed storage-free lysosomes from older and larger storage-filled lysosomes. (B) Lysosomes and Gb3 labeled with anti-LAMP1 antibody (green) and CY3-conjugated Shiga Toxin Subunit B (CY3-ShTxB) in GLA-KO cells in growth medium or starved (HBSS) for 24 hours in the absence or presence of salubrinal (50 µM). (C) Total number of lysosomes in GLA-KO cells across conditions in (B); means ± SD. n = 60 cells per condition from three experiments; P values from t tests. (D) Distribution of lysosome size and storage (Gb3) content across the indicated conditions as a percentage of their total number; n = 60 cells per condition from three experiments. Lysosomes were binned into three classes according to their area [small, S = area < 0.1 µm2; medium, M = area in (0.1 µm2, 1 µm2); large, L = area > 1 µm2] (P = 7.06 × 10-16 from Fisher’s exact test). (E) Lysosomes were labeled with anti-LAMP1 antibody in CHO-WT and GADD34 KO cells in growth medium or starved (HBSS) for 24 hours. (F) Total number of lysosomes in WT and GADD34 KO cells across conditions in (E); means ± SD. n = 50 cells per condition from three experiments; P value from t tests. *P = 0.05; **P = 0.01; ***P = 0.001; ****P = 0.0001.

Fig 5: Summary of the predicted modulation of eIF2α phosphorylation by PPRV. After viral entry, the viral phosphoprotein (P) is produced through transcription and translation process. The P protein expression prevents PERK activation and activates cellular GADD34 to block eIF2α phosphorylation through its negative feedback loop that attenuates signaling of integrated stress, thus leading to the cell survival and translation recovery. Translation recovery may then facilitate PPRV replication and viral protein synthesis. The mechanism by which PPRV P protein induces upregulation of ATF4 in the absence of PERK/eIF2α phosphorylation requires further investigation. In fact, ATF4 was associated with cell death or survival decision through a selection of autophagy or apoptosis with or without CHOP expression.