Concluding Remarks

The extent to which the environment could contribute to transgenerational epigenetic inheritance and adaptive responses in plants is a debatable topic and a central question in genetics and evolution. Various studies in Arabidopsis suggest that environmental stimuli could direct global changes in DNA methylation and providing plant with novel adaptive benefits (Bilichak et al., 2012; Dowen et al., 2012; Jiang et al., 2014). However, these studies have not been able to fully explain the extent of the proposed epigenetic changes, their mode of inheritance, or their adaptive value to the progeny. Moreover, results from various studies in epigenetic inheritance are conflictive to each other and inconclusive (summarize in Table 3).

Table 3. Summary of studies in stress-induced epigenetic inheritance

Results from existing studies Source

Several studies have proposed that stress triggers global DNA methylation and that stress-induced epigenetic changes can be stably inherited by the non-stressed.

Molinier et al., 2006; Bilichak et al., 2012; Luna et al., 2012; Rasmann et al., 2012; Slaughter et al., 2012; Jiang et al., 2014. Other studies have reported that stress does not induce heritable

changes in DNA methylation and that the primary effects of

stress to plant’s methylation are transient.

Lang-Mladek et al., 2010; Pecinka et al., 2010; Pecinka and Mittelsten Scheid, 2012; Sani et al., 2013. A few studies have reported genome-wide DNA methylation

analyses following stress treatment and showed that stress could induce genome wide methylation changes in plants. However these methylome studies were not combined with transgenerational design and transcriptional or phenotypic analyses, thus the significance of the proposed changes remains

Dowen et al., 2012; Jiang et al., 2014.

The above conflicting and inconclusive conclusions are being repeatedly questioned as some of the studies contain deficiencies in experimental design and analysis, such as: · Stress was applied to plants over their entire lifetime, which could directly impact the developing embryos (offspring) in these plants.

· Progeny where not grown in the absence of stress over generations, thus the stability and heritability of the epigenetic changes could not be assessed.

· Statistical analysis of methylation data was not robust enough (the analysis was performed in single plants) thus it is impossible to distinguish between individual-stochastic changes and concrete changes that shared between individual.

Paszkowski and Grossniklaus, 2011; Pecinka and Mittelsten Scheid, 2012; Heard and Martienssen, 2014; Kinoshita and Seki, 2014.

To elucidate the role of stress in transgenerational adaptation a much more robust and systematic approach is needed, which applies to both the experimental design (ensuring that that developing embryos are not directly exposed to stress, following multiple plants over several generations, including generations without stress exposure) and the analyses (integration of methylome data with phenotypic, transcriptomic and mutations studies). In this study, I performed a systematic DNA methylation analysis of plant populations exposed to salt stress for five consecutive generations followed by non-stress exposure for a further two generations. In addition, I have combined this methylome analysis with phenotypic analyses and molecular studies of the affected loci.

My work has revealed that plants possessed highly dynamic short-term stress memory in the form of DNA methylation changes that affect the expression of stress responsive genes and confers phenotypic-plasticity to the immediate progeny. I have found that repeated exposure to salt stress in the parental lineage could lead to increased tolerance to salt stress in the direct progeny. Some studies proposed that stress-induced epigenetic changes and adaptive

response can be stably inherited across generations and I hypothesize that multigenerational salt stress treatments could lead to heritable methylation changes and adaptive traits. In contrary to my hypothesis and some studies, I found that adaptive response to salt only lasted for one generation and the subsequent non-stressed generation did not shows any improved tolerance to stress. Nevertheless, in accordance to previous studies by Boyko et al (2010) and Luna et al (2012), I have shown that this adaptive response is abolished in mutants defective in the non-CG methylation and DNA demethylation pathways, suggesting that the intergenerational response to salt stress is regulated epigenetically. Whole-genome bisulfite sequencing revealed that salt stress alters CHG and CHH methylation, mostly in form of hypermethylation. Consistent with the observed phenotypes, these methylation changes are present in the immediate progeny but are gradually erased once the stress is alleviated.

I hypothesize that salt-induced methylation marks and adaptive traits will be inherited maternally due to the active resetting of methylation marks in the male gametes. In this study I found that indeed the adaptive responses to salt stress are primarily inherited through maternal transmission. Whole-genome bisulfite sequencing on the male gametes revealed that salt-induced DMRs are being reset in the male germline by the activity of DNA glycosylases (DME). Previous work published by Calarco et al. (2012) had shown that DME is involved in resetting of DNA methylation in the male germline, however the implications of such resetting were not clear. My data reveal a significant biological role for this mechanism: the resetting of epigenetic changes induced by stress. This finding highlight the significant differences in the way males and females transfer newly acquired epigenetic changes to offspring in plants, and this may extend to other sexually reproducing organisms.

Finally, I hypothesize that stress-induced epigenetic changes are not occurred stochastically but targeted to certain stress-responsive genic regions or TEs adjacent to genes. These

salt-induced DMRs identified in this study are overlapped with spontaneous DMRs present in plants grown under controlled environment conditions, thus suggesting that certain regions of the genome are epigenetically labile and prone to methylation changes. I found that salt-induced DMRs are primarily targeted to TEs sequences located nearby protein- coding genes. I have provided evidence that salt-induced DMRs affect the transcriptional responsiveness of salt-regulated genes and that an integral component of this regulation is mediated by the activity of antisense lncRNAs.

In summary, this study clarified how plants are able to respond and adapt to stress and explain the regulatory mechanisms by which these stress response occurs. As proposed in the aims, I have addressed three important questions in the field of epigenetic inheritance: the extent of stress-induced epigenetic changes, their mode of inheritance and their adaptive value to the progeny. The robust conclusions drawn from this study hopefully would provide the necessary insight to understand stress memory mechanism in plants and have immense implications for future studies in plant and animal assisted breeding and reproduction.