Fig 2: Calcium chelation prevents the loss of epithelial phenotype in primary rat alveolar epithelial cells exposed to hypoxia. Primary rat AECs were treated or not with 1 µM 1,2-bis(o-aminophenoxy)ethane-N,N,N,N-tetraacetic acid (BAPTA-AM) 90 min before exposure to normoxia (Nx) (21% O2) or hypoxia (Hx) (1.5% O2) for increasing times. (A) ZO-1 (magenta) and TTF1 (cyan) immunostainings were performed after a 6-day exposure. A representative picture of at least n = 4 independent experiments for each condition has been presented and scale bar represents 50 µm. (B) Expression levels of Tgf-β1, (C) Ctgf, (D) Zeb1 and (E) Twist1 mRNA transcripts were assessed after a 48 h exposure. mRNA levels were quantified by qRT-PCR using 2−∆∆CT method and values in hypoxic cells were reported to values obtained in normoxic condition (n = 4 experiments). Raw data were submitted a Kruskal-Wallis test. ** indicate a significant difference as compared with normoxic condition with p < 0.01, respectively. # and ## indicate a significant difference as compared with value in untreated hypoxic cells with p < 0.05 and p < 0.01, respectively.

Fig 3: Hypoxia induces UPR pathways and modulates epithelial and mesenchymal markers expression in lungs of rats exposed to hypoxia. (A) Lungs of rats stabulated in normoxia (Nx) (21% O2) or exposed to hypoxia (Hx) (equivalent to 8% FiO2) during 16 h, 24 h, 48 h or 72 h were isolated and used for western blotting, immunohistochemistry and RT-qPCR analyses. Western blot of ATF4 protein, (B) ATF6N/ATF6 and (C) spliced XBP1 (sXBP1)/XBP1 were performed in lung homogenates. Representative blot of n = 8 experiments is shown. Quantification has been done and expression levels of ATF4, ATF6N or sXBP1 was reported to the ß-actin expression level for each condition. Raw data were submitted to a Kruskal-Wallis one-way analysis of variance. * indicate a significant difference as compared with normoxic condition (p < 0.05). (D) Western blot of a-SMA protein and (E) vimentin were performed in lung homogenates from rat exposed to a 72 h-hypoxia. Representative blot of n = 5 experiments is shown. Quantification has been done and the expression of a-SMA (D) and vimentin was reported to the ß-actin expression for each condition. Raw data (n = 5 rats in each group) were submitted to a Mann-Whitney test and relative expression was represented. (F) mRNA transcript expression levels of Tgf-ß1, Ctgf, Zeb1 and Twist1 in lungs homogenates of rats exposed 48 h to hypoxia were quantified by qRT-PCR using 2-??CT method and reported to the normoxic condition. Raw data (n = 5 rats in each group) were submitted to a Mann-Whitney test and relative expression was represented. (G) Immunostaining of a-SMA or TTF1 were performed on 5 µm slices of paraffin-embedded from the left lobe of rat lung exposed to 72 h normoxia or hypoxia. A representative picture of at least n = 5 independent experiments for each condition has been presented. Original magnification: × 200 and scale bars represent 20 µm. Raw data were submitted to a Mann-Whitney test and relative expression was represented. (H) Co-staining for a-SMA (magenta), TTF1 (yellow), and DAPI to localize nuclei (in cyan) is shown (original magnification: × 400 or × 1000). Percentage of a-SMA or TTF1 positive cells in the lung of rats exposed 72 h to hypoxia as compared to normoxia (n = 5 rats in each group). Raw data were submitted to a Mann-Whitney test * and ** indicate a significant difference as compared with normoxic condition with a p < 0.05 and p < 0.01, respectively.

Fig 4: Inhibition of UPR pathways prevents loss of alveolar epithelial cells phenotype in primary rat alveolar epithelial cells exposed to hypoxia. Isolated primary rat AECs were cultured in normoxia (Nx) (21% O2) or hypoxia (Hx) (1.5% O2) during various periods in the presence or absence of 100 mM 4-phenylbutyrate (4-PBA). (A) Western blot of ATF4 protein and (B) ATF6N/ATF6 ratio were performed primary AECs exposed 6 h to hypoxia. Representative blot of n = 5 experiments is shown. Quantification has been done and the expression levels of ATF4 and ATF6 were reported to the β-actin expression for each condition. (C) Primary rat AECs transfected with plasmid coding for luciferase reporter activity of amino acid response element (AARE: i.e., ATF4-luc) or (D) endoplasmic reticulum stress element (ERSE: i.e., ATF6N/sXBP1-luc), were treated or not with 100 mM 4-PBA and exposed 6 h in hypoxic condition. (C) Luciferase activity corresponding to the transcriptional capacity of ATF4 or (D) ATF6N was measured. (E) ZO-1 (magenta) and TTF1 (cyan) immunostaining were performed on rat AECs cultured on filter and exposed for 6-days to hypoxia in the presence or absence of 100 mM 4-PBA. A representative picture of at least n = 5 independent experiments for each condition has been presented. Scale bar represents 50 µm. (F) mRNA expression levels of Tgf-β1, (G) Ctgf, (H) Zeb1 and (I) Twist1 were quantified by qRT-PCR using 2−∆∆CT method in rat AECs cultured in the presence or absence of 100 mM 4-PBA and exposed 48 h to normoxia or hypoxia. mRNA levels under hypoxic condition were reported to the normoxic condition (n = 5 experiments). Raw data were submitted a Kruskal-Wallis test. *, ** and *** indicate a significant difference as compared with normoxic value with p < 0.05, p < 0.01 and p < 0.001 respectively. # and ### indicate a significant difference as compared with value in untreated hypoxic cells with p < 0.05 and p < 0.001, respectively.

Fig 5: ER stress: a major regulator of the loss of epithelial cell phenotype in hypoxic environment. Loss of alveolar phenotype is a consequence of a long term exposure of the cells to a hypoxic microenvironment. Hypoxia leads to the activation of the PERK/ATF4 and ATF6a/ATF6N branches of the UPR, to the upregulation of Tgfß1, Ctgf, Zeb1 and Twist1 gene expression, to a decrease in TTF1 expression, and finally to a disruption of ZO-1 expression (black arrows). In vitro, pharmacological inhibitors (dotted arrow) revealed that this effect is partly mediated by the stabilization of HIF-1a. Calcium chelation induced by BAPTA-AM prevents the loss of epithelial phenotype observed under hypoxia, by blunting the induction of the UPR and HIF-1 transcriptional activity.