Fig 2: Analysis of minigene induction and splicing. (A) Schematic of primer target locations for analysis of minigene-derived RNA. (B) Quantitative RT-PCR (qRT-PCR) with primers directed against Tnp1 exon 2 and the Pfn3 ORF following 4 h of treatment with Dox or vehicle. Minigene expression was normalized to the average of three reference genes (ALDO, GAPDH and RPS16). Data shown are from three independent, biological replicates; error bars depict mean ± SEM. (C) Phosphorimaging analysis of 4 h induced RT-PCR products generated with intron-flanking Tnp1 primers or 5' and 3'-directed Pfn3 ORF primers in the presence of dCTP a32P. Percent-spliced-in (PSI) was determined as the signal intensity of the spliced product versus the sum of spliced and unspliced products. Amplification of ß-tubulin (TUBB) was performed in a multiplex reaction as an internal control for loading (bottom). (D) 3'-end RT-PCR to assess minigene polyadenylation. Distinct products reflect use of either the endogenous polyadenylation sites-encoded within the Tnp1 and Pfn3 minigenes or the vector-encoded site located within the BGH cassette. ß-tubulin (TUBB) served as a loading control.

Fig 3: Methionine deficiency leads to induction of the PINK11/PARKIN axis. Western blot analysis of (A) PARKIN, (B) Ubiquitin-pSer65 and (C) Mfn2 in the liver of trout fed the control diet (CTL) or methionine-deficient diet (MD) and sampled 16 h after feeding the last meal. Western blot analysis was carried out on six individual samples per treatment, and a representative blot is shown (Source data for each blot are available in Supplementary Figs 6 and 7). Graphs show the ratio of the targeted protein: ß-tubulin (TUBB) or the total amount of the targeted protein used as a loading control. Values are means (n = 6), with standard error of the mean represented by vertical bars. *was used to indicate significant difference between treatment among the two dietary group (P < 0.05; t-test).

Fig 4: Methionine deficiency and mitochondrial mass. (A) Relative mtDNA copy number in the liver of trout fed the control diet (CTL) or the methionine deficient diet (MD) and sampled 16 h after the last meal. (B) TIMM23 levels in the liver of trout fed the control diet (CTL) or the methionine deficient diet (MD) and sampled 16 h after the last meal. Western blot analysis was carried out on six individual samples per treatment, and a representative blot is shown (Source data are available in Supplementary Fig. 4). Graphs show the ratio of the amount of TIMM23: ß-tubulin (TUBB) used as a loading control. (C) Citrate synthase activity in the liver of trout fed the control diet (CTL) or the methionine deficient diet (MD) and sampled 16 h after the last meal. Analysis was carried out on six individual samples per treatment. (D) mRNA level of pgc1a in the liver of trout fed the control diet (CTL) or the methionine deficient diet (MD) and sampled 16 h after the last meal. Expression values are normalized with the eukaryotic translation elongation factor 1 a 1 (eef1a1) mRNA. Values are means (n = 6), with standard error of the mean represented by vertical bars. *was used to indicate significant difference between treatment among the two dietary group (P < 0.05; t-test or Wilcoxon test).

Fig 5: Methionine deficiency lead to induction of mitophagy. Electron microscopy (EM) analysis of liver section of trout fed the control diet (CTL) or the methionine deficient diet (MD). Picture (A) represents liver sections of trout fed MD diet and picture (B–D) liver section of trout fed deficient diet. N, nucleus; m, mitochondria; v, autophagic vacuoles and L, lysosome. Mitophagy is demonstrated by the presence of mitochondria engulfed and/or being engulfed inside autophagosome-related vacuoles. The graphe (E) represents the quantity of mitochondria-autophagic vacuoles in contact and (F) the quantity of autophagic vacuoles per µm2 in the EM images (n = 3 samples with 8 to 10 micrographs per sample). (G) LC3II protein in liver of trout fed CTL diet or MD diet and sampled 4 h after the last meal. Western blot analysis was carried out on six individual samples per treatment, and a representative blot is shown (Source data are available in Supplementary Fig. 5). Graphs show the ratio of the amount of LC3II: ß-tubulin (TUBB) used as a loading control. Values are means (n = 6), with standard error of the mean represented by vertical bars. *was used to indicate significant difference between treatment among the two dietary group (P < 0.05; t-test or Wilcoxon test).