Seed DNA damage responses promote germination and growth in Arabidopsis thaliana

Abstract

The desiccated, quiescent state of seeds confers extended survival of the embryonic plant. However, accumulation of striking levels of genome damage in quiescence impairs germination and threatens plant survival. The mechanisms by which seeds mitigate this damage remain unclear. Here, we reveal that imbibed Arabidopsis seeds display high resistance to DNA damage, which is lost as seeds advance to germination, coincident with increasing cell cycle activity. In contrast to seedlings, we show that seeds minimize the impact of DNA damage by reducing meristem disruption and delaying SOG1-dependent programmed cell death. This promotes root growth early postgermination. In response to naturally accumulated DNA damage in aging seeds, SOG1 activates cell death postgermination. SOG1 activities are also important for promoting successful seedling establishment. These distinct cellular responses of seeds and seedlings are reflected by different DNA damage transcriptional profiles. Comparative analysis of DNA repair mutants identifies roles of the major genome maintenance pathways in germination but that the repair of cytotoxic chromosomal breaks is the most important for seed longevity. Collectively, these results indicate that high levels of DNA damage incurred in seeds are countered by low cell cycle activity, cell cycle checkpoints, and DNA repair, promoting successful seedling establishment. Our findings reveal insight into both the physiological significance of plant DNA damage responses and the mechanisms which maintain seed longevity, important for survival of plant populations in the natural environment and sustainable crop production under changing climates.

Keywords: DNA repair; genome stability; germination; seed; seed quality.

Fig. 1.

Imbibed seeds progressively lose radioresistance through germination. (A) Root growth sensitivity of wild-type Col-0 Arabidopsis seeds in response to X-irradiation. Seeds were stratified for 2 d at 4 °C and exposed to X-rays (100 Gy at 2 Gy/min) at the stated time point after transfer to 23 °C. Irradiated seeds and unirradiated controls were plated onto half MS plates postirradiation, grown vertically at 23 °C 16-h day, and root growth quantified over 7 d. Data were analyzed using one‐way ANOVA with Tukey's honestly significant difference post hoc test. Letters denote homogeneous subsets (P < 0.05). Error bars show SEM of >15 roots. (B) X-ray–induced PCD in Arabidopsis Col-0 seeds and seedlings. Seeds were stratified for 2 d at 4 °C and either exposed to 100-Gy X-rays immediately poststratification (0-d seeds) or after 7 d growth on half MS at 23 °C 16-h day (7-d seedlings). The appearance of cell death was monitored over 96-h recovery from irradiation. Confocal images of PI-stained Col-0 roots expressing the PWOX5:GFP QC marker. (Scale bar, 50 µm.) A bright field image is also shown at 96 h. (Scale bar, 500 µm.) See SI Appendix, Fig. S3 for unirradiated controls. (C) Maximum width of the root tip (n = 10) and (D) area of WOX5-GFP expression in plants 96 h after exposure to 100-Gy X-rays at 0 d or 7 d poststratification or unirradiated controls (n = 10). Letters denote homogeneous subsets (P < 0.001). (E) Timing of S-phase (DNA replication) in germinating Col-0 Arabidopsis seeds. EdU labeling and confocal microscopy of germinating seeds indicating the onset of S-phase by 16 h. (F) Timing of G2/M-phase in germinating Col-0 Arabidopsis seeds. β-glucuronidase (GUS) reporter analysis of PCYCLINB1;1-GUS lines indicating G2/M-phase cells postgermination at 2 d poststratification. (G) Representative brightfield images of seeds at different time points poststratification. (Scale bar, 50 µm.)

Fig. 2.

The Arabidopsis transcriptional response to DNA damage differs between developmental stages. RNAseq analysis of Arabidopsis Col-0 seed and seedling responses to X-rays. Seeds (0 d) were stratified on half MS agar for 2 d at 4 °C before X-irradiation (100 Gy at 2 Gy/min at RT). Seedlings (7 d) were grown vertically on half MS plates at 23 °C 16-h day for 7 d before X-irradiation as described for seeds. Unirradiated controls were maintained at RT for 30 min in place of X-ray treatment. RNA was extracted from stratified seeds or seedlings 6 h after the end of the irradiation treatment, and transcripts were quantified by sequencing. (A) Comparison of the log2 fold change (FC) in transcript levels of genes differentially expressed 6 h post 100-Gy X-irradiation in 7-d seedlings and stratified seeds (0 d) by RNAseq. X-ray–responsive genes with no significant difference in transcript levels between 0-d and 7-d irradiated samples (black). Genes with lower (<−1 log2 fold change) (blue) and higher (>1 log2 fold change) (red) abundance postirradiation and which are differentially expressed (P < 0.05 Holm-Bonferroni correction) between irradiated 0-d and 7-d developmental stages are indicated. (B) Venn diagram displaying the irradiation response in 0-d seeds or 7-d seedlings. Genes that are induced by irradiation are indicated (↑) in seeds (pink circle, 0 d) or seedlings (red circle, 7 d) or at both stages. Genes with lower (↓) abundance after irradiation in seeds (0 d) are indicated by the purple circle and in seedlings (7 d) by the blue circle. The gene ontology enrichment of highest statistical significance is indicated in green text, and numbers of genes in each section of the Venn diagram are indicated. (C) qPCR confirmation of RNAseq expression patterns for transcripts with higher abundance after irradiation, including XRI1, induced in 0-d seeds and 7-d seedlings in WT but not in sog1-2 mutants, compared to AT4G05380 and AT2G18193 that display significant induction only in 7-d WT plants. (D) qPCR analysis AT2G25060 and AT5G06150 that display significant X-ray–induced reduction in transcript levels only in 7-d WT seedlings. Data were analyzed by ANOVA with Tukey’s post hoc correction for multiple testing (P < 0.01) for each gene. All unlabeled data points belong to group “a.”

Fig. 3.

Germination and aging-induced cell death in Col-0 and sog1 mutant lines. Germination of sog1 mutant seed is resistant to aging. Germination of sog1-2, sog1-3, and Col-0 seeds was analyzed after accelerated aging at 35 °C and 83% RH relative to unaged control seed. Seeds were stratified at 4 °C for 2 d before transfer to 23 °C 16-h day and scored for radicle emergence each day poststratification. (A) Mean germination time (MGT) and (B) viability. Error bars represent the SEM of the mean of three replicates of 30 seeds (*P < 0.05; **P < 0.01, t test of the sog1 mutant against Col-0 for each aging regime). (C) Seedling establishment is lower in seedlings germinated from aged sog1 mutants relative to wild type. Analysis of survival of seeds transferred from germination to soil at 7 d postgermination. Error bars are SEM of three replicates of 20 plants (*P < 0.05; **P < 0.01, t test). (D) PCD in wild-type and DNA repair mutant seed was analyzed by viability staining. Embryos were isolated from aged seeds for 10 d and unaged seeds 2 d postgermination and roots analyzed by PI staining and confocal microscopy. Col-0 images include GFP signal from the PWOX:GFP QC reporter. Bar is 50 μm. (E) PCD was quantified as the percentage of roots displaying one or more dead cells in unaged seeds (black bars) and after 10 d aging (yellow bars). Significance groups are indicated by letters (P < 0.05, n > 40 per treatment, Fisher’s exact test with post hoc analysis).

Fig. 4.

Germination performance of DNA repair mutant and wild-type lines. Aging sensitivity of independent alleles of mutants in the major plant DNA repair pathways relative to wild-type lines. Germination of Col-0 and mutant lines was analyzed after accelerated aging at 35 °C and 83% RH relative to unaged control seed. Seeds were stratified at 4 °C for 2 d before transfer to 23 °C 16-h day and scored for radicle emergence each day poststratification. (A) Germination of Col-0, arp1-1, xrcc2-1, ercc1-1, and ku70-1 mutant alleles. (B) Germination of Col-0 and mutant alleles after aging for 1 wk . (C) MGT of control seed lots. (D) Mean germination time of 7-d aged seed lots. (E) Mean viability of control seed lots. (F) Mean viability of 7-d aged seed lots. Error bars represent the SEM of the mean of three replicates of 30 seeds. Data were analyzed by ANOVA with Tukey’s post hoc correction for multiple testing (P < 0.05).