Fig 2: Analysis of structure and similarity within Arabidopsis ADH1 homologues. (A) A Phylogenetic tree depicting the relatedness of ethanol-processing dehydrogenases (At_ADH1, PhADH1, ZmADH1, and Hs_ADH1), GSNO reductases (At_GSNOR and Hs_GSNOR), and ADH-like sequences of unknown function (At5g24760, At1g64710, At1g32780, At5g42250, At5g42250, At1g22440, At1g22430, and At4g22110). Bootstrap confidence levels (n=500) are shown at each branching site. The complete alignment from which the tree derives is contained in Supplementary Dataset S2. (B) Multialignment of protein regions that contain ADH1/GSNOR distinguishing motifs (coloured shading). Black–grey shadings indicate conserved domains among all ADH proteins considered. (C) Three-dimensional representation of the substrate-binding domain (1–184) of the Arabidopsis ADH1 (PDB ID: 4rqt). Distinctive ADH1/GSNOR motifs are shown in the same colours used to highlight the sequences in (B). Ethanol (brown for carbon, white for hydrogen, and red for oxygen atoms) and zinc ions (pale green) are shown using spherical representations. The side chains of His and Cys residues involved in catalytic zinc co-ordination are shown.

Fig 3: Phylogenetic tree illustrating the relatedness of ADH proteins identified in species representing land plant evolution (Arabidopsis thaliana, Zea mays, Petunia hybrida, Solanum lycopersium, Pinus pinea, Pteris vittata, Selaginella moellendorffii, Physcomitrella patens, and Marchantia polymorpha). MEGA7 (Kumar et al., 2016) was used to generate an unrooted maximum-likelihood tree. The original alignment is shown in Supplementary Dataset S1. Red and blue shadings indicate significantly reproducible (bootstrap value >90) clustering of ADH1-like sequences and GSNOR-like sequences, respectively.

Fig 4: Loss of function of ADH1 attenuates submergence tolerance and autophagosome formation. (A) Phenotypes of 4-week-old wild-type (WT), adh1-4, adh1-8, and adh1-16 plants before treatment (Air) and following 6 d of recovery after 7 d of light (normal light/dark cycle) submergence (LS) treatment (LS+R6). (B) ATG7 and ATG8a protein levels in 4-week-old WT, adh1-4, adh1-8, and adh1-16 plants before treatment (Light) and after 24 h of LS treatment. Coomassie blue-stained total proteins are shown below the blots to indicate the amount of protein loaded per lane. (C) Monodansylcadaverine (MDC) staining of mature root cells from one-week-old WT, adh1-4, adh1-8, and adh1-16 seedlings under normal light/dark (Light) conditions or following 24 h of LS treatment. Red arrows indicate labeled autophagosomes. Bars = 50 µm. (D) Number of puncta per root section in mature root cells from the WT, adh1-4, adh1-8, and adh1-16 seedlings shown in (C). Data are average values ± SD of three biological replicates. For each experiment, 15 sections were analyzed per genotype. Asterisks represent significant differences between WT and adh1 samples, as determined by Student’s t-test (* p < 0.05 and ** p < 0.01).

Fig 5: Assessment of ethanol fermentation in ADH and PDC mutants. (A) ADH activity in 6-day-old Arabidopsis seedlings (Col-0 wild type, and adh1, adh1adh2, and pdc1pdc2 mutants) after incubation under normoxic (21% O2 v/v) or hypoxic (1% O2 v/v) conditions for 12 h. The box plots are built on the bases of four replicates. (B) ADH1 expression in the wild type (Col-0), and adh1 and pdc1pdc2 mutants. The box plots are built on the bases of four replicates. (C) Immunodetection of ADH1 protein in aerobic wild-type (Col-0), adh1, and pdc1pdc2 seedlings, grown in liquid MS medium for 4 d, in the presence of 1% sucrose (w/v). Three biological replicates for each genotype are shown. Gel staining with Coomassie G-250 is shown as a protein loading control (D) Ethanol production in the same genotypes used for the ADH activity quantification shown in (A). The box plots are built on the bases of four replicates.