Translational fidelity and growth of Arabidopsis require stress-sensitive diphthamide biosynthesis

Abstract

Diphthamide, a post-translationally modified histidine residue of eukaryotic TRANSLATION ELONGATION FACTOR2 (eEF2), is the human host cell-sensitizing target of diphtheria toxin. Diphthamide biosynthesis depends on the 4Fe-4S-cluster protein Dph1 catalyzing the first committed step, as well as Dph2 to Dph7, in yeast and mammals. Here we show that diphthamide modification of eEF2 is conserved in Arabidopsis thaliana and requires AtDPH1. Ribosomal -1 frameshifting-error rates are increased in Arabidopsis dph1 mutants, similar to yeast and mice. Compared to the wild type, shorter roots and smaller rosettes of dph1 mutants result from fewer formed cells. TARGET OF RAPAMYCIN (TOR) kinase activity is attenuated, and autophagy is activated, in dph1 mutants. Under abiotic stress diphthamide-unmodified eEF2 accumulates in wild-type seedlings, most strongly upon heavy metal excess, which is conserved in human cells. In summary, our results suggest that diphthamide contributes to the functionality of the translational machinery monitored by plants to regulate growth.

Fig. 1. Identification and phenotypes of Arabidopsis dph1 mutants.

a Scheme of first step of diphthamide biosynthesis in yeast and human. SAM S-adenosylmethionine, MTA 5′-methylthioadenosine. b Arabidopsis hygromycin sensitivity. Photograph of 12-day-old seedlings of the wild-type (WT), dph1-1, dph1-2, and dph1-1 pDPH1:DPH1-GFP complemented line Y22-10 cultivated without (control) or with 2 μM hygromycin B (hyg). c–e Growth of WT, dph1, and complementation lines on 0.5× MS medium, shown as photograph (c), fresh biomass (d), and primary root length (e) of 12-day-old seedlings. Scale bar, 10 mm (c). Data are mean ± s.d., n = 4 pools of between 5 and 6 seedlings, with each pool sampled from an independent petri plate (d, e). Significant differences from WT: *P < 0.05, ***P < 0.001 (one-way ANOVA, Tukey’s test).

Fig. 2. Diphthamide modification of eEF2 and translational reading frame accuracy depend on DPH1 that localizes to the cytosol.

a Representative confocal laser scanning microscopic image of a root tip of an 8-day-old dph1-1 pDPH1:DPH1-GFP (line Y22-10) seedling stained with propidium iodide (PI, red). Scale bars, 50 μm. b Immunoblot probed with antibodies specific for detection of eEF2 lacking the diphthamide modification (unmodified), recognizing all forms of eEF2 (global), or serving as a loading control (anti-β-Actin for MCF-7, anti-plant ACTIN for A. thaliana), respectively. Wild-type (WT) and DPH1^−/− mutant human cell line MCF-7 served as controls (left). c MS/MS spectra of eEF2 peptide 684-GICFEVCDVVLHSDAIHR-701 from WT (top) and dph1-1 mutant (bottom), with diphthamide (_D) or without diphthamide (_N) modification on H700 respectively (_C: carbamidomethylation). The selected monoisotopic precursor m/z (z = +4) is given in each diagram; all detected y fragment ions are shown in blue. Red/pink lines visualize the consistent m/z difference of 71.055 between equivalent y_n²⁺ ions of the two genotypes. Tissues were shoots of 16-day-old 0.5× MS-grown seedlings (b, c). d Rates of ribosomal −1 frameshifting error. We normalized the ratio of firefly/renilla luciferase activity for a test −1 frameshift reporter construct to the ratio for an in-frame control reporter construct, as measured in transiently transfected Arabidopsis leaf mesophyll protoplasts prepared from 4-week-old soil-grown dph1 mutants and wild-type plants. Mean ± s.d., n = 3 independently transfected replicate aliquots of protoplasts, shown normalized to the wild-type (see Supplementary Fig. 4b-d). Significant differences from WT: **P < 0.01 (one-way ANOVA, Tukey’s test).

Fig. 3. Decreased organ size, cell number, and TOR activity in dph1 mutants.

a Photographs of 4-week-old wild-type (WT) and dph1 plants cultivated in soil. b Sixth oldest leaves of the plants as shown in (a). Scale bars, 10 mm (a, b). c–e Leaf area (c), cell area (d), and cell number (e) of the sixth oldest leaves as shown in (b). f Number of leaves of plants as shown in (a). Data are mean ± s.d., n = 5 plants (c,e, f), and a boxplot, n = 272 (WT), n = 287 (dph1-1), and n = 255 (dph1-2) cells from a total of 5 plants (d). g Number of leaves at bolting. Data are mean ± s.d., n = 10 plants, with cultivation in neutral days (12 h light). h Confocal laser scanning microscopic images of PI-stained root meristematic zones of 9-day-old WT and dph1 mutant seedlings cultivated on 0.5× MS medium. Scale bars, 50 μm. Arrows mark the cortex transition zones. i, j Meristem length (i) and meristem cell number (j) of roots as shown in h. Shown are boxplots, n = 12 seedlings per genotype. k Sensitivity to TOR kinase inhibitor. Data are mean ± s.d., n = 10 and 15 seedlings for WT and dph1 mutants, respectively, from day 4 to day 7 of AZD-8055 treatment (Supplementary Fig. 8). l, m TOR activity. Total protein extracts of 14-day-old seedlings grown in liquid 0.5× MS medium with 0.5% (w/v) sucrose were resolved by SDS-PAGE, blotted, and the membrane probed with antibodies against S6K-P and S6K or stained with Coomassie Brilliant Blue (CBB) as a loading control. m S6K-P/S6K ratio. Data are mean ± s.d., n = 4 independent experiments (l and Supplementary Fig. 9). Violin/boxplots show median (center), 1st and 3rd quartile (gray lines/box), minimum and maximum (drawn lines/whiskers) (d, i, j). Significant differences from WT: *P < 0.05, **P < 0.01, ***P < 0.001 (one-way ANOVA with Games-Howell test). Different characters reflect significant differences between means (two-way ANOVA, Scheffé test, P < 0.05, k).

Fig. 4. Comparative transcriptomics of dph1 mutants and wild-type plants.

a Principal Component Analysis of normalized transcriptome data (transcripts per kilobase million, TPM). b Venn diagrams showing transcripts present at higher (upregulated) and lower (downregulated) levels in dph1-1 and dph1-2 mutants, compared to the wild-type (WT). c, d Representative GO terms significantly enriched among genes of which transcript levels are upregulated (c) and downregulated (d) in both dph1-1 and dph1-2 mutants compared to WT (FDR < 0.05). e Normalized transcript levels of DPH1. Red horizontal line marks threshold TPM of 2.46 used to filter out non-expressed genes. f, g TPM of TCTP1 (f) and TCTP1 protein levels determined by immunoblot using an anti-TCTP1 antibody or stained with Coomassie Brilliant Blue (CBB) as a loading control; numbers are band intensities relative to those in WT (g). Data are mean ± s.d. (e, f), n = 3 independent experiments, with leaves sampled from 4-week-old soil-grown plants (a–g). Significant differences from WT: ***P < 0.001 (one-way ANOVA, Tukey’s test).

Fig. 5. Copper and cadmium toxicity correlate with accumulation of diphthamide-unmodified eEF2 protein.

a Shoot fresh biomass of 18-day-old seedlings. Wild-type (WT; n = 5), iron uptake-defective irt1 (n = 3), sulfate uptake-defective sultr1;1 sultr1;2 (sultr)(n = 3), dph1-1 (n = 5), and dph1-2 (n = 7) mutants were cultivated in modified Hoagland medium. Additionally, WT was cultivated in modified Hoagland medium supplemented with 5 µM Cu (n = 3), 80 µM Zn (n = 4), 2 µM Cd (n = 3), 5 nM methyl viologen (MV; n = 4), 25 µM Pb (n = 4), and a specific Pb-free control medium (C_Pb; n = 3). Shown are mean ± s.d., n biologically independent pools of 17 seedlings, with each pool sampled from a replicate petri plate, from one experiment. *P < 0.05, **P < 0.01, ***P < 0.001 (one-way ANOVA with Games-Howell test, or two-tailed Student’s t-test compared with the respective control). b Immunoblot detection of unmodified and global eEF2 protein in total protein extracts from shoot tissues of the seedlings shown in a. c, d Photographs of 18-day-old WT seedlings cultivated on series of CuSO₄ (c) or CdCl₂ (d) concentrations in modified Hoagland medium (Controls: 0.5 μM Cu, 0 µM Cd). Scale bars, 10 mm. e, f Immunoblots (as in b for shoots of Cu-exposed (e) and Cd-exposed seedlings (f), respectively. Numbers indicate the fold increase in unmodified eEF2 protein amount relative to controls, after normalization to global eEF2 protein levels. Protein extracts from dph1-1 and dph1-2 seedlings served as controls. g, h Relationships between the amount of unmodified eEF2 protein and shoot growth inhibition in Cu-exposed (g, y = 1.80 x − 1.15, R² = 0.9995) and Cd-exposed (h, y = 3.68 x + 5.00, R² = 0.9384) WT seedlings. A.U. arbitrary units. i Immunoblot detection of unmodified and global eEF2 protein in total protein extracts from DPH1^−/− mutant and WT of human MCF-7 cell line cultivated in liquid medium without or with addition of 125 μM CuSO₄ for 3 days.