Fig 2: Hypoxia-induced changes in expression of FAM13A, IREB2 and HIF1α in lung cancer cell lines and human lung fibroblasts. A. Quantitative analysis of FAM13A, IREB2 and HIF1α mRNA expression in 3 cell lines. Cells were cultured in hypoxia or ambient oxygen (normoxia) for 72 hours. Gene expression was normalized to the endogenous control, GUSB. Mean normalized expression levels in hypoxia and normoxia are shown. Data are expressed as the mean ± SEM of triplicate values. Results are mean from three independent experiments. A549, human non-small cell lung carcinoma cell line; CORL-105, Caucasian lung adenocarcinoma cell line; HL, human lung fibroblast; N, normoxia (blue); H, hypoxia (red). *P < 0.001, **P < 0.0001. B. Western blot analysis of FAM13A, IREB2 and β-actin proteins expression in 3 cell lines, from two separate experiments. Cells were cultured in hypoxia or ambient oxygen (normoxia) for 72 hours and total proteins were extracted. A549, human non-small cell lung carcinoma cell line; CORL-105, Caucasian lung adenocarcinoma cell line; HL, human lung fibroblast; Hypoxia exposition +; Normoxia, -.

Fig 3: Genetic repression of miR-29 chronically affects iron homeostasis. a, b Western blot of IRP2 and TFR1A in 12-week-old tg(kif5aa:eGFP-sponge-29) (n = 4) and wild-type (n = 4) fish brain extracts and (b) relative densitometric analysis (*P < 0.05; **P < 0.01, Mann–Whitney U-test). c Representative Western blot of IRP2 in brain extracts of kif5a:sponge-29 at different ages (5, 12, 20, 27). In a–c α-TUBULIN was used as loading control. d Comparison of densitometric analysis calculated on Additional file 8a and b. The blue line represents IRP2 expression in tg(kif5aa:eGFP-sponge-29) animals, gray line represents expression in wild-type fish. Values were normalized to the mean of 5 weeks value (Kruskal–Wallis test, P = 0.0939, n = 3 animals for each age point). e, f Expression level at 24 hpf of Ireb2 and Tfr1a upon miR-29 mimics injection. The expression level was determined by RT-qPCR. Statistical significance was assessed by one-way ANOVA with post-hoc Tukey’s test (*P < 0.05), the analysis was performed on total RNA extraction from 30–40 embryos for each condition. Error bars indicate standard errors of means. g Brain non-heme iron content (μg/g wet tissue) in kif5a:eGFP-sponge-29 (n = 5) compared to wild-type (n = 6) at age 12 weeks (*P < 0.05, Mann–Whitney U-test). h Representative images of lipofuscin accumulation in the optic tectum of 12-week-old kif5a:eGFP-sponge-29 and wild-type fish brains. Lipofuscin auto-fluorescent granules (green) were detected with ApoTome microscope, counterstained with DAPI (blue). Scale bar: 50 μm. i Quantification of lipofuscin density based on percentage of area over threshold, n = 6 (*P < 0.05; Mann–Whitney U-test)

Fig 4: MiR-29 family targets Ireb2 mRNA. a Presence of a putative binding site for miR-29 family in the Ireb2 mRNA 3′-UTR sequence in several vertebrate species (H. sapiens: ENSG00000136381; M. musculus: ENSMUSG00000032293; R. norvegicus: ENSRNOG00000013271; C. familiaris: ENSCAFG00000001766; G. gallus: ENSGALG00000003171; D. rerio: ENSDARG00000021466; N. furzeri: Nofu_GRZ_cDNA_3_0193494), mammalian sequences are in black, birds in purple, and teleost fish in blue. Perfect match to positions 2–8 of the miR-29 seed sequence is highlighted in green and is present in all vertebrate sequences shown. b Expression of GFP assessed by cytofluorometric analysis. Fusion with Ireb2 3′-UTR of N. furzeri. The gray column reports the baseline fluorescence intensity of the construct without miR-29 mimic. The middle blue column reports the fluorescence with miR-29 mimic and the red column the fluorescence of a construct (Δ) where the putative binding site for miR-29 in the Ireb2 3′-UTR was mutated to destroy complementarity. Statistical significance of fluorescence difference between baseline and co-injection with miR-29 mimics was evaluated by Student’s t-test (* P < 0.05). c Age-dependent regulation of transcripts coding for key genes of iron metabolism, from Baumgart et al. [46]. One-way ANOVA for linear trend is reported for each gene: Tfr1a (R = 0.5009; P < 0.0001), Scl11a2 (R = 0.3892; P < 0.01), Fth1a (R = 0.067; P = 0.11), Slc40a1 (R = 0.078; P = 0.11), Ireb2 (R = 0.141; P = 0.069). Expression values (in RPMKs) for each age were centered and scaled to the mean, n = 5 animals for age point. d Representative Western blot of IRP2 in brain extracts and densitometric analysis relative to Additional file 8a (Kruskal–Wallis test, P = 0.0216, n = 3 animal for each age point). α-TUBULIN was used as loading control. e Brain non-heme iron content (μg/g wet tissue) in fish of different ages. (One-way ANOVA: P < 0.0001). f DAB-enhanced Perls staining of the cerebellum of young and old fish. The black arrows point to labelled Purkinje cells. The inset in the bottom right corner of each picture shows a magnification of the Purkinje cell bodies. g, h Age-dependent expression of Fbxl5 mRNA (ANOVA for linear trend: R = 0.1922, P = 0.0688) and western blot of FBXL5 in brain lysates at three age points (5, 12, and 27 weeks). α-TUB was used as loading control. For b, c, d, e, g mean ± standard errors of means is reported

Fig 5: Quantitative analysis of FAM13A, IREB2 and HIF1α mRNA expression in cultured lung cancer fragments. Non-small cell lung cancer (NSCLS) samples from 9 donors (different colors) were cultured in hypoxia or ambient oxygen (normoxia) for 72 hours. Data are presented as fold changes (FCh) in gene expression. FCh were calculated using the 2-ΔΔCT method as described by Livak et al. Briefly, mean Ct values for the gene of interest and a reference gene in hypoxia (CtT,GOI and CtT,Ref, respectively) and normoxia (CtN,GOI and CtN,Ref, respectively) were used to calculate ΔCt (CtGOI - CtRef) for each sample. Then to derive the fold change (FCh) of the gene in the hypoxia compared with the normoxia FCh = 2-ΔΔCt, where ΔΔCt = ΔCtH - ΔCtN. An FCh of >2 was defined as overexpression, and an FCh of <0.5 was defined as under-expression of the gene.