Plant Senescence

Last update: Oct 2020

This section provides an overview of the study conducted by the scientists from Daegu and Jeju in South Korea, and Beijing and Shenzhen in China. The team investigated aging mechanisms in plants, namely the influence of DNA double-strand breaks (DSB) on plant senescence, characterized by leaf yellowing. In their study, Li et al. showed that some aging mechanisms are probably conserved between plants and animals, supporting the DNA-damage theory of aging.

The team demonstrated that deficiency in gene ATAXIA TELANGIECTASIA MUTATED (ATM) adversely impacts the leaf lifespan in Arabidopsis due to DSB accumulation. They showed that levels of DSBs increased, and DNA repair capacity decreased with plant’s age. To test the hypothesis stating that DSBs cause senescence, the team used blenomycin (BLM) treatment and endonuclease I-PpoI activation to generate DSBs in young plants, both methods resulting in premature senescence. The team also showed that ATM repressed DSB-induced expression of Senescence-Associated Genes (SAGs). Interestingly, researchers found that the SUPPRESSOR OF VARIEGATION 3–9 HOMOLOG 2 (SUVH2) was the downstream component of ATM in DSB-induced leaf senescence.

The work of Zhonghai Li, Jin Hee Kim, Jeongsik Kim, Jae Il Lyu, Yi Zhang , Hongwei Guo, Hong Gil Nam and Hye Ryun Woo titled "ATM suppresses leaf senescence triggered by DNA double-strand break through epigenetic control of senescence-associated genes in Arabidopsis" was published in 2020 in The New Phytologist. The PDF file can be accessed by clicking this button (offsite link to Research Gate, works as of Oct 2020).


All life on Earth is at risk of experiencing endogenous and environmental stresses, which can eventually lead to DNA damage such as double-strand breaks (DSBs). While it was demonstrated that DNA damage drives aging in animals, less focus was given to elucidating this relationship in plants. Li et al. demonstrated that, in Arabidopsis thaliana, levels of DSBs increase, and the DNA repair efficiency decreases as its leaves age. This finding was further consolidated by showing that when DSBs were generated via expression of I-PpoI nuclease, the plant exhibited a premature senescence phenotype.

The team also created loss-of-function mutants for 13 genes involved in the DNA repair pathway, showing that the deficiency in ATAXIA TELANGIECTASIA MUTATED (ATM), which is activated in response to DSBs and mobilizes the DNA damage response (DDR) network, causes premature senescence in Arabidopsis. The scientists showed that ATM can repress DSB-induced expression of senescence-associated genes, such as genes encoding WRKY and NAC transcription factors, which orchestrate the leaf senescence process, by changing histone lysine methylation.

Overall, Li et al. demonstrated the importance of ATM in controlling leaf senescence, which hints that some aging mechanisms could be conserved in animals and plants.


Plants, similarly to other living organisms, suffer from DNA damage caused, for example, by DNA replication errors and exogenous genotoxic stresses. DNA damage can also be induced by environmental hazards (drought, UV light) and metabolic by-products (reactive oxygen species). DNA damage is detrimental to plants, as it can cause mutations, genome instability, cell death, and premature aging. It is, therefore, in the plant’s best interest to detect and repair the damaged DNA, especially DSBs.

The system controlling this process is called the DNA damage response (DDR) signaling pathway. DDR network is orchestrated by ATAXIA TELANGIECTASIA MUTATED (ATM)- a highly conserved protein kinase that is preferentially activated when DSBs are detected. ATM acts upstream of p53, activating cell cycle checkpoints in response to DSBs. It phosphorylates TP53, BRCA1, and many other targets.

Apart from ATM, ATM AND RAD3-RELATED (ATR) protein kinase is also activated in response to DNA damage, particularly in response to single-strand breaks (SSBs), which are usually found at stalled replication forks. ATR phosphorylates CHK1 (checkpoint kinases 1), leading to cell cycle arrest.

ATM and ATP can phosphorylate an overlapping set of DNA repair or checkpoint targets, such as the aforementioned p53. They can both induce G1/S, G2/M, and S-phase checkpoints. Thus, they ensure that DNA damage is repaired prior to DNA replication or cell division. If the damage is not repaired successfully, cell death occurs. In this way, ATM and ATR can determine cell fate.

Most of the orthologues of DDR components found in animals have also been identified in plants. This includes ATM and ATR. However, some DDR components are plant-specific, such as the SUPPRESSOR OF GAMMA RESPONSE 1 (SOG1). SOG1 is a transcription factor regulating the transcriptional activation response to gamma irradiation, participating in the DDR pathways governed by both ATM and ATR. SOG1 is a functional homolog of animal p53.

Senescence in Animals

DNA damage theory of aging postulates that the age-associated accumulation of DNA damage drives the aging process in animals. As a result of increased genotoxic stress and decreased DNA repair capacity, DNA damage accumulates throughout normal aging. DSBs, being difficult to repair, are believed to be key in the mammalian aging process. This theory is supported by the fact that deficiencies in DSB repair characterize some human diseases or disease models. When there is a deficiency in ATM, which normally controls the DDR pathway, a condition called ataxia-telangiectasia (AT) emerges. AT is characterized by the pleiotropic phenotype, which includes the appearance of age-related alterations such as hair graying, hair loss, and osteoporosis in humans and mice disease models. The AT patients have trouble with coordinating movements, and their life expectancy is notably shortened.

In mice, increased ATM activity was shown to increase lifespan via stabilizing the pro-longevity enzyme SIRTUIN 6, which is required for maintaining genome stability.

Generation of DSBs by the inducible expression of a restriction endonuclease SacI can result in premature aging phenotypes, which suggests that the importance of ATM may lie in its activity resulting in DSB repair. It could also mean that DSBs themselves are the factor that promotes aging.

Senescence in Plants

Similarly to animals, plants are not exempt from the aging process. Senescence is the last phase of plant growth and is associated with degenerative events that result in a decrease in metabolic activities and the death of cells, tissues, and organs. Senescence in plants is characterized by leaf yellowing, which is very easy to spot even with the naked eye. Yellowing occurs due to the loss of green pigments called chlorophylls. Other common events accompanying senescence include alteration in mitochondrial and thylakoid structure and shrinkage of the chloroplasts and the nucleus. During senescence, essential macromolecules such as proteins and photosynthetic pigments are degraded, and their components are often recycled in other parts of the plant. This process is mediated by senescence-associated genes (SAGs).

Leaf senescence is believed to be a highly organized, genetically controlled, programmed cell death process. During the process, gene expression is altered, followed by a decline in photosynthetic activity and the aforementioned degradation of macromolecules. Over the past years, leaf senescence mechanisms have been extensively studied in many plants. However, the influence of DSBs and their repair on leaf lifespan has not been elucidated.

Methods and Results

Shortened Lifespans of ATM-deficient Leaves

First, the team wanted to test the hypothesis that plants use the age-associated activity of DNA repair systems as an important, conserved anti-aging mechanism countering senescence. This hypothesis was based on previous findings that dysfunctions of DNA repair pathways in many organisms led to premature aging, and that plants employ many DNA repair pathways to fix the DNA damage.

The team tested whether DNA repair pathway genes influenced lifespan in Arabidopsis leaves. They created loss-of-function mutants for 13 key genes involved in DNA repair and examined their phenotypes. Of those 13 mutations, only the mutation in the ATM gene resulted in notable premature senescence exhibited by leaf yellowing. As you can see in Figure 1.a (below), the atm mutant, which was deficient in ATM activity, demonstrated early senescence compared to wild type Col-0 plant. Col-0 only exhibited significant yellowing at 41 days old, while one could already notice many yellow leaves on the atm mutant plant when it was only 37 days old.

Next, the team observed the 3rd and 4th leaves of all the plants at different points during their lives, day 0 marking the emergence of individual leaves. You can see the representative leaf samples for Col-0 and atm in Figure 1.b. Note that Col-0 leaves only exhibited yellowing when 32 days old, while atm mutant leaves already suffered from senescence when 28 days old.

After the visual assessment, the team assessed the molecular performance of the Arabidopsis plants. Figure 1.c shows the photochemical efficiency of photosystem II (PSII), a crucial component of the plant’s photosynthetic pathway. By assessing the photochemical efficiency of PSII, one can indirectly assess the physiological state of the leaves. Senescent leaves tend to have lower photochemical efficiency than younger or mature leaves. Photochemical efficiency, denoted as Fv/Fm, was measured using IMAGING-PAM Chlorophyll Fluorometer that employs Saturation Pulses. Fv/Fm is also denoted as a maximal PS II quantum yield. It is calculated from formula (Fm-Fo)/Fm, where Fo is the dark fluorescence yield, and Fm is the maximal fluorescence yield.

The y-axis in Figure 1.c is the photochemical efficiency of 3rd and 4th leaves. The x-axis denoting leaf age is shared between Figure1.c, d, and e. In Figure 1.c, you can see that the photochemical efficiency decreased with leaf age both for Col-0 and atm. This could be because the fitness of an organism generally tends to decrease with age, which is caused, among other things, by DNA damage accumulation and decreased DNA repair capacity. However, for 28-day-old leaves, there was a statistically significant difference between the photochemical efficiency of Col-0 and atm mutant samples, which suggests that the deficiency in ATM may reduce the fitness of the leaves.

Figure 1.d assesses the chlorophyll (Chl) content of 3rd and 4th leaves of different ages in arbitrary units (a.u.). Chlorophyll breakdown is an essential catabolic process of leaf senescence. Therefore, a decrease in chlorophyll content is considered to be an important marker of leaf senescence. Chlorophyll content was measured using the atLEAF CHL PLUS handheld device that measures the optical density difference at 2 wavelengths (640nm and 940nm) and infers chlorophyll content from those values. In Figure 1.d, you can observe that the significant difference between Col-0 and atm emerged earlier than in Figure 1.c. We can say with 95% confidence that the chlorophyll content of atm mutant leaves was lower than that of Col-0 when leaves were 20 days old, although the difference was not significant for 24-day-old leaves. After day 28, we can say with 99% confidence that atm chlorophyll content was significantly lower than that of Col-0 samples. Again, this suggests that atm mutants suffer from leaf senescence earlier than Col-0 leaves.

To further quantify senescence in the leaf samples, the team investigated the transcription levels of Senescence-Associated Gene 12 (SAG12), which is a senescence marker gene. They used reverse transcription quantitative PCR (RT-qPCR) analysis on 3rd and 4th leaf samples. The transcript level of SAG12 was normalized to actin2 (ACT2, At3g18780) internal control. The y-axis in Figure 1.e represents the SAG12 expression levels as the leaves age. The expression levels are significantly higher for atm mutant samples as compared to Col-0 samples for 24-day-old and older leaves.

Figure.1 The influence of ATM deficiency on Arabidopsis leaf fitness

Also, Figure 2 (from supplemental materials) shows that the ATM expression in Col-0 (normalized to ACT2) decreases with leaf age. This observation is in line with the hypothesis that ATM could act as a suppressor of leaf senescence, and that the reduction of its expression, whether by normal aging of loss-of-function mutations, leads to leaf senescence. It suggests that ATM might regulate the lifespan of plant leaves.

Given that ATM probably has an important function in DSB repair, Li et al. decided to examine whether the atm mutant had higher levels of DSBs as leaves aged. The team performed a single-cell electrophoresis assay (comet assay) for 8- and 20-day-old leaves of Col-0 and atm in order to assess the extent of their DNA damage. DSB levels were calculated as the mean percentage of DNA in the comet tail, which contains short strands of DNA that have "broken off" (because of DSBs) the large DNA strands located in the comet head. The results of this experiment are presented in Figure 1.f. You can see that while there was no significant difference in DSB levels between Col-0 and atm samples when the leaves were 8 days old, the difference becomes apparent for 20-day-old leaves. Also, note that even for Col-0, DSB levels were 2.5-times higher in 20-day-old leaves than in 8-d-old leaves, which suggests that DSBs accumulate as part of the normal aging process. As hypothesized, there was more DSBs in atm mutant samples than in Col-0 samples on day 20. When we combine this result with the fact that no significant difference between the two samples was observed on day 8, we could deduce that the shortened leaf lifespan of the atm mutant might be due to age-dependent excessive accumulation of DSBs.

The experimental results of Li et al. hint that ATM could be a conserved regulator of aging not only in animals (as demonstrated by other studies) but also in plants. This aging regulation could take place via suppression of age-related DSB accumulation.

Figure 2. Expression of ATM gene with age of Columbia ecotype of Arabidopsis

DSBs Can Induce Premature Senescence

Seeing that DSBs were accumulated in older leaves, Li et al. decided to check whether the DNA repair capacity decreased with leaf age. They performed a comet assay in 8- and 20-day-old 3rd and 4th leaves of Col-0 with and without 1hr treatment with 10 μgml-1 bleomycin (BLM), which generates DSBs. Afterward, the leaves were given time to repair the DNA damage for 0, 20, 40, or 60 minutes, and then subjected to the comet assay (Fig.3.a).

In Figure 3.b, we can see the results of this experiment, with young leaves on the left and 20-day-old ones on the right. The y-axis represents DSB levels. The letters on top of the bars represent significant differences. The columns that are not marked with a common letter exhibit a statistically significant difference, with P below 0.05. For both 8- and 20-day-old samples, you can see a sharp increase in DSB levels compared to controls (-, no BLM treatment), which indicates that BLM was successful in generating DSBs. For 8-day-old leaves right after BLM treatment (repair time= 0 min), DSB levels increased to around 78%. However, when the leaves were given 1 hour to repair the damage, DSB levels were down to 35%. Notably, 20-day-old leaves were not able to repair the damage with similar efficiency. This suggests that DNA repair capacity decreases with increasing leaf age, which might cause the accumulation of DSBs as the leaves age.

Next, the team tried to elucidate the relationship between DSBs and leaf lifespan by comparing the DSB levels in Col-0, ore1 mutant, and nore1 mutant. ore1 is a delayed senescence mutant oresara1, while nore1 is an earlier senescence mutant not oresara1. Figure 3.c shows the DSB levels for those three in 8- and 20-day-old leaves. While the difference between the plants was not statistically significant for 8-day-old leaves, it was pretty apparent for 20-day-old ones. As hypothesized, delayed senescence mutant ore1 displayed lower, and the early senescence mutant nore1 displayed higher DSB levels than Col-0. This data suggests that there is a negative correlation between DSB level and leaf lifespan.

Figure 3. DSB generation by blenomycin (BLM) treatment leads to premature senescence

Generation of DSBs through Inducible Activation of I-PpoI Endonuclease Results in Premature Senescence

To confirm that DSBs can lead to leaf senescence in other experimental settings (not only by BLM treatment), Li et al. employed transgenic Arabidopsis plants expressing Intron-encoded endonuclease I-PpoI fused to a rat glucocorticoid receptor to generate DSBs. In this experimental design, expression of the I-PpoI can be induced by treating plant leaves with dexamethasone (Dex), which is a corticosteroid that binds to and activates the glucocorticoid receptor. Activated I-Ppol can then generate DSBs.

The reason why the team used other agents than BLM to generate DSBs is to check whether premature aging phenotype obtained in the last set of experiments is indeed due to DSBs, which can be generated by either BLM or endonucleases, or if it could by due to some other effect that BLM alone has on plant cells. If I-Ppol activation generated the same phenotype as BLM treatment, the research team would have a stronger argument advocating that it is the DSBs (and not a side-effect of the agent they used to generate them) that are a driver of aging.

Figure 4.a shows the expression analysis of I-PpoI in Col-0 and two I-PpoI transgenic plants with and without a 6-hour treatment with 20 μM Dex. ACT2 was used as an internal control. You can see that for the two transgenic plants, I-PpoI was activated after applying Dex (upper bands appeared), which confirmed that the experimental design worked as expected and that the researchers could proceed with their study.

The team wanted to know if DSBs created by I-PpoI will influence leaf senescence. They applied Dex to the leaves of 12-day-old plants (not leaves this time) every two days and inspected the change in appearance. Figure 4.b shows the 28-day-old plants. The transgenic plants treated with Dex exhibited some yellowing at the tips, while the untreated one remained green. Figure 4.c shows the 40-day-old plants, which exhibit a clear difference between the Dex-treated and -untreated plants. Dex-treated I-PpoI plant leaves became visibly yellower.

The team then investigated the age-dependent senescence phenotypes in 3rd and 4th leaves of mock and Dex-treated I-PpoI plants. As shown in Figure 4.d and 4.e, both the photochemical efficiency of PSII and Chlorophyll content decrease markedly for the Dex treated I-PpoI leaves compared to other samples.

Figure 4.f shows the RT-qPCR analysis performed on 28-day-old leaves of the I-PpoI#3 line. It investigated the expression of 4 senescence-associated transcription factors (Sen-TFs), and one Senescence-Associated Gene SAG12, normalized to ACT2. As you can see, the expression of all 5 is significantly higher for leaves treated with Dex.

These results suggest that DSBs can induce premature leaf senescence. Also, they show that DSB accumulation, which results from the decrease in DNA repair capacity as leaves age, could be one of the causes of leaf senescence.

Figure 4. I-Ppol induced by Dex treatment generates DSBs that lead to premature senescence

DSBs Lead to Senescence-Like Transcriptome Changes

In this experiment, Li at al. wanted to investigate the molecular regulation of DSB-mediated leaf senescence. They compared the transcriptomes (all mRNA) of BLM-responsive and age-dependent plant samples. In the case of BLM-responsive plants, the specimens were grown on media with or without BLM until they were 3 weeks old. In the age experiments, the transcriptomes were assessed for 3rd and 4th leaf samples of 14- and 28-day-old leaves. As shown in Figure 5.a, almost 3200 genes were upregulated (red), and 2400 downregulated (blue) due to BLM treatment, the cut-off value being a twofold change in expression (FC). On the other hand, 4000 genes were upregulated and 2400 downregulated in response to aging. 47% of the BLM-responsive genes were also upregulated in leaf aging. Moreover, 40% of the genes downregulated due to BLM treatment were also downregulated in leaf aging. FE under the Venn diagrams stands for fold enrichment, and it denotes how many folds higher or lower was the expression than what it was expected to be by random chance. So in a way, FE is a way to assess our confidence level in a particular result.

Figure 5.b illustrates the expression levels of all the genes whose expression increased or decreased twofold compared to each control. You can see that the boxes, which encompass 25th to 75th percentiles of genes, span similar gene expression levels for BLM and aging. This shows that many genes are affected to a similar extent by both of those factors. For example, many genes increase around 1.5 fold for both BLM exposure and age.

Overall, these results suggest that DSB-inducing condition such as BLM treatment causes similar changes in transcriptome to the changes induced during age-dependent leaf senescence.

Figure 5. BLM-responsive and age-dependent transcriptome changes in Arabidopsis partially overlap

The next experiment uses Gene Ontology (GO) enrichment analysis of transcriptome in response to BLM and age. Gene Ontology is an ambitious bioinformatics project aiming to connect the genes and their protein products to their biological functions in different model organisms. GO analysis facilitates the interpretation of the vast sets of data generated in high-throughput experiments such as RNA-seq. Those experimental results consist of massive lists of genes (or their protein products) that are over- or under-represented. In GO, genes have function terms assigned (annotated) to them. For example, a specific gene can be involved in photosynthesis or carotenoid metabolism, and therefore has this GO term assigned to it. GO analysis can be employed on a set of genes to find out which GO terms associated with them are enriched. That is to say, which terms relating to a specific function are observed more frequently than expected by random chance. For instance, one can compare the transcriptome of cell samples collected from individuals with and without a particular condition (disease). We can use GO enrichment analysis to determine which biological processes or molecular functions are related to that condition.

Figure 5.c shows the result of the GO enrichment analysis of transcriptome in response to BLM and age that used the FC values of over 15 thousand genes with valid expression levels. Red and blue boxes indicate respectively significantly upregulated and downregulated GOs. Statistical significance is expressed as the Z-score. As you can see, the expression of genes involved in photosynthesis and carotenoid metabolism significantly decreased in both BLM and age-regulated transcriptomes, as indicated by the similar shade of blue in both gene types. Both transcriptomes also displayed similar levels of AUX/IAA transcription factors downregulation.

Notably, a major upregulation can be observed in the WRKY transcription factors (TFs) family for both BLM- and age-regulated transcriptomes. This is especially striking when compared to the NAC TF family, which is also known to exhibit upregulated expression during leaf senescence. In Figure 5.d, each cell indicates the expression of a gene belonging to the WRKY or NAC families. Heatmaps show FC levels. As we can see, the expression of most WRKY genes was increased correspondingly for BLM and aging treatments, while it certainly was not the case for NAC genes. While Pearson’s correlation coefficient value between BLM- and aging-induced expression change was 0.65 in the case of WRKY TFs, it was only 0.36 for NAC TFs. These results suggest that WRKY TFs could have regulatory roles in common leaf senescence pathways induced by DSBs (BLM) and age alike.

DSBs Induce the Expression of Senescence-Associated Genes

In order to test if DSBs can directly regulate the expression of senescence-associated genes, Li et al. examined the expression of SAGs following BLM treatment. Nine SAGs were selected from almost 2500 genes over- or underexpressed both in BLM- and age-regulated transcriptomes. Their transcription levels were measured in 12-day-old Col-0 leaves, 3 and 6 hours after 10 μgml-1 BLM treatment. As shown in Figure 6.a, five senescence-associated transcription factors (Sen-TF) genes (ANAC016, ANAC019, WRKY6, WRKY53, and WRKY75) and two senescence marker genes (BIFUNCTIONAL NUCLEASE 1 and SAG13) were significantly upregulated after exposure to DSB-inducing BLM. The increase was especially notable for WRKY53 Sen-TF. On the other hand, the two genes related to photosynthesis (LHCB2.1 and LHCB2.4) were downregulated following treatment, which could be another marker of leaf senescence.

As the next part of the experiment, the team decided to determine how DSBs regulated the expression of the nine SAGs. Multiple sources state that DSBs could induce changes in various histone modifications, such as lysine methylation, which is a particularly prominent factor in the epigenetic regulation of gene expression during leaf senescence. Thus, Li et al. investigated the histone lysine methylation status at specific SAG loci in response to DSBs. 12-day-old Col-0 leaves were treated with 10 μgml-1 BLM for 6 hours to induce DSBs, then the ChIP analysis of histone lysine methylation followed by qPCR was performed for four Sen-TF genes. Histone 3 lysine 4 trimethylation (H3K4me3) and histone 3 lysine 27 trimethylation (H3K27me3) antibodies were used for the experiment. Recall that H3K4me3 is generally associated with the transcription activation of nearby genes, while H3K27me3 has been linked to gene repression/silencing.

Figure 6.d shows the H3K4me3 and H3K27me3 enrichment at the four promoters of Sen-TFs. Note that the main text of the paper says Figure b, but it described Figure d, so the graphs are probably mislabelled. As displayed in Figure 6.d, the enrichment of H3K4me3 associated with gene activation increases for Sen-TF samples treated with BLM compared to mock treatments. Conversely, enrichment of H3K27me3 inked to gene silencing decreases for all four Sen-TFs investigated. This observation shows that the increase in DSB levels induced by BLM treatment leads to the upregulation of Sen-TFs, possibly by altering histone lysine methylation.

If a change in expression of Sen-TFs through histone lysine methylation is a prerequisite for correct regulation of DSB-induced leaf senescence, then it follows that the atm mutant exhibiting leaf senescence should be characterized by altered histone lysine methylation of Sen-TFs. To test this hypothesis, the team set out to investigate the histone lysine methylation of Sen-TFs in atm mutant leaves.

First, they quantified the regular expression of four Sen-TFs in 8- and 20-day-old Col-0 and atm leaves. Figure 6.b shows the results for 8-day-old, and Figure 6.e for 20-day-old 3rd and 4th leaves. While there was no significant difference between Sen-TFs expression in 8-day-old leaves when we compare Col-0 and atm, the discrepancy becomes apparent in 20-day-old leaves. Figure 6.e shows that Sen-TFs expression is significantly higher in atm mutant samples.

Equipped with the knowledge of transcription levels, the team investigated the enrichment of H3K4me3 and H3K27me3 in Sen-TF gene promoters. Figure 6.c shows the enrichment levels for 8-day-old, and Figure 6.f for 20-day-old leaves. Again, there was no significant difference in enrichment levels between Col-0 and atm in the case of 8-day-old samples. Also, consistent with the gene expression patterns, 20-day-old atm mutant leaves exhibited significantly higher levels of H3K4me3 enrichment (activation) than Col-0 leaves, and lower levels of H3K27me3 enrichment (repression) for WRKY53 and WRKY75 (Fig.6.f).

Taken together, the experimental results suggest that DSB accumulation in the atm mutant might induce SAG expression through changes in H3K4me3/H3K27me3 enrichment as the leaves age. This epigenetic change could be the cause of premature senescence.

Figure 6. DSBs regulate the expression of SAGs by modifying histone lysine methylation state

SUVH2 Functions Downstream of ATM in the Suppression of DSB-Induced Senescence

In the last series of experiments, Li et al. examined which epigenetic regulators are involved in ATM-dependent changes in histone methylation of SAGs during leaf senescence. Out of many epigenetic regulators affecting histone methylation in Arabidopsis, the team chose the SUVH2-overexpressing (SUVH2ox) transgenic line as their candidate regulator and performed their experiments using this particular line. They chose SUVH2ox because it is the only documented line that shows altered age-dependent leaf senescence phenotype and histone methylation status of WRKY53. SUVH2 itself is also a Sen-TF.

In the experiment, the team used Col-0, atm, and SUVH2ox leaves subjected to DSB generation resulting in leaf senescence by treatment with mock and 1 or 10 μgml-1 BLM for 5 days. The photochemical efficiency of PSII was measured, and the results are summarized in Figure 7.a-c. Figure 7.a shows no significant difference in photochemical efficiency for mock treatment. However, for both BLM concentrations (Fig.7.b,c), SUVH2ox plants exhibited significantly higher efficiencies than either atm mutant or Col-0, which illustrates their resistance to premature leaf senescence.

Li et al. also examined the transcription levels of the four senescence-associated transcription factors in Col-0 and SUVH2ox leaves with and without a 6-hour BLM treatment. In Figure 7.d, the expression level in the mock-treated Col-0 leaves was set to 1. You can see an increase in the expression levels of all four Sen-TF genes when you compare mock- and BLM-treated samples of Col-0 and SUVH2ox. However, there is also a massive, statistically significant difference between BLM-treated Col-0 and SUVH2ox leaves, with SUVH2ox having much lower expression levels of the four Sen-TFs. This suggests that SUVH2 can suppress DSB-triggered senescence by repressing the transcriptional activation of Sen-TFs.

To find out whether ATM and SUVH2 are indeed members of the same pathway, and which one is the upstream agent, the research team used a specific inhibitor of the ATM kinase called KU (KU60019) in combination with BLM treatment. In other words, the team decided to assess the physiological status of Arabidopsis leaves when its subjected to DSBs generation (by BLM), and when the ATM that drives the DNA Damage Response is being inhibited. Beforehand, they confirmed that 10 μM KU treatment alone does not cause a change in the photochemical efficiency of Col-0, atm, and SUVH2ox leaves.

Figure 7.e and 7.f show the photochemical efficiency of PSII in the leaves of Col-0, atm, and SUVH2ox plants treated with 1 μgml-1 or 10 μgml-1 BLM and 10 μM KU. Compared to the treatment with BLM only (DSB generation), co-treatment with BLM and KU (DSB generation and ATM inhibition) resulted in a statistically significant decrease in photochemical efficiency in Col-0 leaves at each time point (Fig.7.b vs. e and c vs. f, black lines). The co-treatment and BLM-only treatment did not yield statistically different results for the atm mutant. It was probably because it already has the loss-of-function atm mutation, so the additional treatment with ATM inhibitor would not make any difference to its performance. Notably, SUVH2ox leaves did not exhibit a significant decrease in photochemical efficiency when co-treated rather than treated with BLM only. Therefore, we can say that, for SUVH2ox, the photochemical efficiency was unaffected by the ATM inhibition. These results suggest that ATM requires SUVH2 to suppress DSB-induced leaf senescence.

Figure 7. SUVH2 is required for ATM-mediated suppression of DSB-induced senescence


This paper provided experimental results suggesting that the formation and repair of DSBs can determine the rate of aging in plants. Li et al. demonstrated the accumulation of DSBs in older Arabidopsis leaves, which mirrors the DNA damage accumulation in older mammals. The team showed that DSB repair capacity decreased with the age of leaves, which is also in line with the findings in mammals. Li et al. demonstrated that DSB levels in early/delayed leaf senescence mutants were negatively correlated with leaf lifespan. Collectively, this suggests that DSBs could be a common cause of aging both in plants and mammals. Deficiency in ATM gene, which alarms the DNA Damage Response of the DSBs, shortened leaf lifespan in Arabidopsis, which supports previous findings in animals. Induction of DSBs by the expression of I-PpoI was also shown to cause early leaf senescence phenotypes in Arabidopsis. However, further studies of the genome-wide changes caused by inducible activation of I-PpoI in Arabidopsis and their influences on leaf senescence are needed.

Overall, multiple experimental results in this study hint that DSBs could be an evolutionarily conserved driver of the aging processes in plants and animals.

Li et al. also established that the deficiency of ATM in Arabidopsis leads to a shortened leaf lifespan due to the age-associated accumulation of DSBs. This supports the theory listing ATM as a shared regulator of DSB-induced aging in animals and plants. ATM was shown to transduce DSB signaling in Arabidopsis, indicating that this mechanism is also conserved in plants and animals. Transcriptome analysis shows that ATM mediates DSB-regulated gene expression in Arabidopsis. DSBs were also documented as rate-limiting factors for seed germination, ATM being an important determinant of seed viability in Arabidopsis. However, atm mutant seeds were shown to germinate quicker than WT seeds. In this paper, atm mutant exhibited premature senescence due to BLM treatment and with age, as well as DSB accumulation with age, which stresses the importance of ATM in leaf aging regulation in plants.

Even though the team examined the phenotypes of loss-of-function mutants for 13 DDR genes, surprisingly, only one of them affected leaf longevity with age. This stands in striking contrast to animals. Though plants seemed mostly unaffected by those gene mutations, deficiencies in any of the nine homologous human and mouse genes, including ATR, BREAST CANCER SUSCEPTIBILITY1 (BRCA1), KU80, and Radiation sensitive 50 (Rad50), leads to the formation of premature aging phenotypes. What is more, the atm mutant exhibited only mild senescence phenotypes in Arabidopsis leaves, while it would show a much more severe phenotype in animals.

Why would some plants be so much more resistant to senescence than animals? One possible explanation is that plants have multiple DDR systems that can efficiently make up for the loss-of-function of a single gene. Also, while loss-of-function in ATR, BRCA1, KU80, and SOG1 does not induce the premature leaf senescence phenotype, it does affect processes such as response to c-irradiation and maintenance of genomic integrity, and transcription under genotoxic stress. Therefore, there may be another mechanism for repairing DSBs generated during age-dependent leaf senescence processes in plants that does not involve ATM. In order to better understand the regulation of leaf senescence through DNA damage repair, further studies of the influence of various DNA repair pathway genes on leaf senescence are needed.

Histone modifications are relevant to signaling and DSB repair in mammals and yeast. Acetylation and methylation in histone H3 are responsive to X-ray radiation in Arabidopsis, which suggests that changes in histone modification are a shared epigenetic response to DNA damage in animals and plants. Li et al. demonstrated the role of H3 methylation in regulating BLM-induced leaf senescence. By showing that the enrichment of H3K4me3 and H3K27me3 was altered in the promoters of Sen-TFs in older atm leaves and that the expression of Sen-TF genes decreased in BLM-treated SUVH2ox plants, Li et al. showed that ATM suppresses leaf senescence triggered by DSBs through modulating histone lysine methylation. Moreover, treating SUVH2ox plants with ATM inhibitor did not aggravate BLM-induced leaf senescence, suggesting that it acts downstream of ATM. However, how does SUVH2 influence the alterations in the methylation status of histones at Sen-TFs regions during DSB-induced leaf senescence? The researchers theorize that it could perhaps provide a scaffolding for the assembly of heterochromatin, which results in the repression of Sen-TFs, thus alleviating the senescence phenotype. One finding that supports this hypothesis is that SUVH2 was shown to associate with the chromatin-remodeling complex, acting as an adaptor protein to mediate polymerase V occupancy at RNA-directed DNA methylation loci.

Overall, the findings in this paper support the DNA-damage theory of aging. The work provides new insights on an evolutionarily conserved mechanism for DNA-damage-induced aging involving ATM in animals and plants.



[1] Li, Z., Kim, J. H., Kim, J., Lyu, J. I., Zhang, Y., Guo, H., Nam, H. G., & Woo, H. R. (2020). ATM suppresses leaf senescence triggered by DNA double-strand break through epigenetic control of senescence-associated genes in Arabidopsis. The New phytologist, 227(2), 473–484.

[2] Yossi Shiloh. Atlas of Genetics and Cytogenetics in Oncology and Haematology. 2016. ATM (ataxia telangiectasia mutated). [Accessed 25 October 2020].

[3] M. E. Gagou& M. Meuth. Atlas of Genetics and Cytogenetics in Oncology and Haematology. 2010. ATR (ataxia telangiectasia and Rad3 related). [Accessed 25 October 2020].

[4[ Heinz Walz GmbH. 2019. IMAGING-PAM M-Series Chlorophyll Fluorometer. Instrument Description and Information for Users. [Accessed 25 October 2020].

[5] atLEAF. 2020. atLEAF CHL PLUS chlorophyll meter. [Accessed 25 October 2020].

[6] Collins A. R. (2004). The comet assay for DNA damage and repair: principles, applications, and limitations. Molecular biotechnology, 26(3), 249–261.

[7] Ashburner, M., Ball, C. A., Blake, J. A., Botstein, D., Butler, H., Cherry, J. M., Davis, A. P., Dolinski, K., Dwight, S. S., Eppig, J. T., Harris, M. A., Hill, D. P., Issel-Tarver, L., Kasarskis, A., Lewis, S., Matese, J. C., Richardson, J. E., Ringwald, M., Rubin, G. M., & Sherlock, G. (2000). Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nature genetics, 25(1), 25–29.

[8] Pomaznoy, M., Ha, B., & Peters, B. (2018). GOnet: a tool for interactive Gene Ontology analysis. BMC bioinformatics, 19(1), 470.

[9] GO enrichment analysis. 2020. Gene Ontology. [Accessed 25 October 2020].

[10] Zhang, T., Cooper, S., & Brockdorff, N. (2015). The interplay of histone modifications - writers that read. EMBO reports, 16(11), 1467–1481.


Figures 1-7 Li, Z., Kim, J. H., Kim, J., Lyu, J. I., Zhang, Y., Guo, H., Nam, H. G., & Woo, H. R. (2020). ATM suppresses leaf senescence triggered by DNA double-strand break through epigenetic control of senescence-associated genes in Arabidopsis. The New phytologist, 227(2), 473–484.