Saline conditions impose physiological constraints on plants through, firstly, oxidative stress, and secondly, ionic stress (Shabala 2017). Salt accumulating in the substrate surrounding the roots inhibits the capacity of roots to take up water (an osmotic effect). Lower water availability at this stage will lead to reduced plant growth, as has been shown for A. thaliana (Luo, et al. 2017), rice (Yeo, et al. 1991), maize (Frensch and Hsiao 1994; Rodriguez, et al. 1997), and barley (Munns, et al. 2000). Plant roots use the ionic composition of the substrate in which they are grown for turgor recovery, and thus can soon begin to take in water after the initial osmotic stress (Shabala and Lew 2002). However, this water contains very high Na+ and Cl− concentrations leading to the second major physiological constraint on growth: ionic stress. Na+ and K+ ions possess similar physio-chemical properties: Na+ can compete with K+ for important binding sites within the cell, impairing enzyme activity (Flowers and Colmer 2008; Shabala and Lew 2002). In addition, Na+ and Cl− ions can accumulate in the cell wall causing cell dehydration (Munns and Passioura 1984). The inability to compartmentalize/exclude harmful ions inside the cell (e.g. in the vacuole) inhibits regular cell function and leads to cell death from either toxicity or dehydration (Munns 2002). This internal injury inhibits new leaf growth, reduces overall plant photosynthesis and thus reduces supply of carbohydrates to new cells (Shabala 2017). Plants have evolved different ionic stress tolerance mechanisms, which can vary across species, and may depend on local environmental conditions as well as the length of salinity exposure. These are (i) the ability of roots to recognize Na+ ions and exclude them from accumulating within the plant (Alberico and Cramer 1993; Byrt, et al. 2007; Fortmeir and Schubert 1995; Matsushita and Matoh 1991), and (ii) tissue tolerance of Na+ and Cl− ions through compartmentalization (Apse, et al. 1999; Flowers and Colmer 2008; Flowers, et al. 2010; Mühling and Läuchli 2002; Munns and Tester 2008).
Arabidopsis thaliana accessions display variation in tolerance to low concentrations of NaCl, but not high NaCl
The A. thaliana accessions used in this experiment originate from across the northern hemisphere (Supplementary File 1). While many A. thaliana accessions show only small differences in nucleotide sequence they can display large genetic variation for phenotypic characteristics, e.g. flowering time (Kowalski, et al. 1994). At low concentrations of NaCl there were notable variations between genetically different accessions in this study. Four accessions (Bur-0, Col-0, T910, Wilna) display normal or above-normal above-ground biomass accumulation at 50mM NaCl, while others (Ler-0, Sorbo, TAL07) display ~50% reduction in above-ground biomass accumulation at this concentration (Figure 1).
Previous work has demonstrated large differences between A. thaliana accessions for NaCl tolerance, as measured using days-to-death (Katori, et al. 2010) and leaf rosette area coupled with electrolyte leakage (Julkowska, et al. 2016). Our experiments did not identify a large variation in NaCl tolerance between accessions at stressful NaCl concentrations (Figure 1), even though there is some overlap with the accessions used in our study and the previously studies. It is noteworthy that previous experiments used much higher concentrations of NaCl than our study, both in soil and in artificial growth media in vitro experiments.
Different accessions of Arabidopsis thaliana display parent-of-origin independent and dependent genome dosage effects on salt stress tolerance
We have previously demonstrated that A. thaliana tetraploid plants accumulate more above-ground biomass than their diploid equivalents (Fort, et al. 2016), indicating a genome dosage effect. Likewise, paternal-excess F1 triploid plants can also accumulate more above-ground biomass than their diploid equivalent, as well as maternal-excess F1 triploid plants. In this study, we identify genome dosage effects (and also parent-of-origin dependent genome dosage effects) on salt stress tolerance (Figure 3). We reveal that in some genetic backgrounds (accessions) there are parent-of-origin independent genome dosage effects on salt stress tolerance, that are not evident in other genetic backgrounds (Figure 3). Indeed, five of the ten accessions tested (i.e. Col-0, TAL07, Wilna, Zurich and Ler-0) displayed significant parent-of-origin independent genome dosage effects on salt stress tolerance where diploids were more stress-tolerant than both of the reciprocal triploids (Figure 3). In addition, in three out of ten accessions (i.e. Col-0, Ler-0 and TAL07) the tetraploid plants were were more stress-tolerant than both of the reciprocal triploids (Figure 3). If the genome dosage effects on salt tolerance we have identified were linear, we should expect that the salt tolerance value will either be 2x > 3x > 4x or 2x < 3x < 4x. However, there are no cases where such trends are evident. Instead, our results indicate that in five genetic backgrounds, the salt tolerance of both isogenic reciprocal triploid lines is lower than both the diploid and tetraploid parental lines. For Col-0, TAL07 and Ler-0, the finding that diploid and tetraploid plants perform better under salt stress tolerance than triploid plants is strongly suggestive of a parental-genome dosage balance effect on salt stress tolerance, where plants with an equal parental contribution of chromosome sets (i.e. 2x and 4x) perform better than those with unequal parental genome dosage.
In addition, we demonstrate that there are also parent-of-origin dependent genome dosage effects on salt stress that are accession-specific. The reciprocal triploid lines are genetically identical at the nuclear level differing only in whether the two genome copies of the genome in the triploid plant have been inherited paternally (via pollen) or maternally (via ovule). Any differences in salt stress tolerance between the maternal-excess triploid and the paternal-excess triploid are likely to have an epigenetic basis, either due to parent-of-origin specific epigenetic marks and/or parent-of-origin specific dosage dependent factors. We have identified such parent-of-origin dependent genome dosage effects on salt tolerance in three genetic backgrounds (Figure 3). Such parent-of-origin dependent genome dosage effects on salt tolerance between the reciprocal triploid plants could be due to epigenetic marks at the nuclear genome level (in the embryo or endosperm), cytoplasmic differences between diploid and tetraploid maternal parents, or dosage dependent factors in the maternal seed coat that differ between diploid and tetraploid maternal parents.
Overall, the genome dosage effects between diploids and tetraploids, and the reciprocal F1 triploid pairs, are occurring in plants that are genetically identical (isogenic) apart from their differential genome dosage or whether their genome copies are maternally derived (madumnal) or paternally derived (padumnal). Hence, such accession-specific genome dosage effects are likely epigenetic in nature as they do not involve any changes in DNA sequence.