Developmental, genetic and environmental variations of global DNA methylation in the first leaves emerging from the shoot apical meristem in poplar trees

ABSTRACT

In the context of climate changes, clarifying the causes underlying tree phenotypic plasticity and adaptation is crucial. Studies suggest a role of epigenetic mechanisms in response to external stimuli, raising the question whether such processes can promote acclimation of trees exposed to adverse climate conditions. Recently, we revealed an environmental epigenetic footprint in the shoot apical meristem (SAM) which could partially be transmitted mitotically, for several months, up until the winter-dormant bud in field conditions. Here, we extended our previous analysis to the leaves of the same P. deltoides×P. nigra clones. We aimed at estimating the range of developmentally, genetically, and environmentally induced variations on DNA methylation. We showed that only the first leaves emerging from the SAM displayed variations of DNA methylation under changing water conditions. We also found that these variations are genotype- and pedoclimatic site-dependent. Altogether, our data raised questions and perspectives on the direct acquisition, the maintenance of environmentally induced DNA methylation changes, and their mitotic transmission from the SAM to the first emerging leaves.

During their long lifespan, trees must continually respond to environmental pressures, through physical and developmental modulations. Models predict dramatic changes of forest composition as a consequence of climate change.^1,2 An open question is to which extent the phenotypic plasticity of trees will allow them to cope with such rapid climate changes, facilitating their long-term adaptation to such new environmental conditions. In this context, epigenetics was proposed as a potential source of phenotypic plasticity and even adaptation.^3,4 Indeed, epigenetics seems of first importance in plants with the modifications of chromatin marks such as DNA methylation and histone marks altering gene expression and transposable element mobility.^5,6 However, the ecological significance in terms of acclimation and adaptation still needs evidence particularly using field studies and epigenomics population.⁷

Poplar (Populus spp.) is a model tree with sequenced genomes, genetic diversity, fast juvenile growth, and large water requirements^8,9 relevant for investigating the ecophysiological and molecular determinants of tolerance to water deficit.^10–15 Accordingly, poplar has been used to investigate the role of epigenetics in the response of trees to environmental variations.^3,4,16–20 Especially, several studies tested whether signs of epigenetic reprogramming in response to environmental variations could be observed during very early stages of shoot morphogenesis, namely in the shoot apical meristem (SAM). Among these studies, the SAM-specific epigenetic component of geographic site-dependent growth performances,^21,22 environment-induced bud break,²³ and response to water availability^16,17,24,25 have been described. Changes in the epigenome and transcriptome of the poplar SAM in response to water availability were shown to affect preferentially hormone pathways and could participate in the developmental plasticity.¹⁷ In addition, an investigation on field-grown poplar clones or natural populations revealed that winter-dormant SAM shows environmental epigenetic memory of changes occurred during the vegetative period.^16,24 These data suggested a possible mitotic transmission in the SAM from the induction time (vegetative period, active SAM) to the winter (dormant SAM).^16,24

In the present study, we focused on the environment- and genotypic-dependent epigenetic changes in leaves, thus monitoring later stages of shoot morphogenesis and phenotypic plasticity as compared to previous studies on the SAM. We evaluated global DNA methylation that is a widely used epigenetic indicator in plants.^16,24–26 Global DNA methylation was assessed by high-performance liquid chromatography procedure^{16,17,24,27,28} in the leaves of the P. deltoides×P. nigra hybrids recently reported for DNA methylation variations in SAM.^16,17 The corresponding trees had grown in (a) well-watered or moderate water-deficit conditions in greenhouse for one genotype as previously reported^14,17,25,29 or (b) pedoclimatic conditions among two contrasted sites in France for 31 genotypes.^16,30 This way, instead of a direct comparison between experimental designs, the present work allows to detect unifying trends regarding the epigenetic response to limiting growth conditions.

We first investigated whether the DNA methylation of leaves emerged during a suboptimal water system could be affected, suggesting either a de novo change or a mitotic transmission from the SAM to the emerging leaves. Global DNA methylation was assessed in leaves that experience a water deficit at different physiological ages (measured by leaf rank) (Figure 1). A significant change in methylation was noted in response to drought, exclusively in emerging leaves (ranks 0 and 1; rank 0 being the first emerging leaf over 20 mm length) (Figure 1). These emerging leaves were potentially formed during the drought period, but the precise measure of the plastochron index would allow a definitive conclusion. This result suggests a developmental control of environmentally induced variations of DNA methylation. The biological significance of these variations will need further studies particularly since leaves formed during water deficit in Populus balsamifera genotypes have already been characterized by a different phenotype from leaves formed in well-watered conditions, with a lower stomatal density and changes in the expression of ontogenic genes.¹³

Figure 1.

Global DNA methylation percentage in leaves according to leaf rank on P. deltoides × P. nigra ‘Carpaccio’ hybrid. White dots correspond to well water condition (WW) and black dots correspond to severe stress (WD, 10% <REW<20%, during 10 d, Cohen et al., 2010). Values are leaf rank means (±SE, n = 3). Leaf rank starts from ‘0ʹ which corresponds to the first emerging leaf >20 mm to ‘9ʹ which corresponds to the more mature developed leaf we have collected. Levels of significance of t-test for a given leaf rank between WW and WD conditions are *P < 0.05, **P < 0.01, and nothing indicated if non-significant. Leaf ranks potentially formed during stress are indicated by the red vertical dashed arrow.

Second, our data from greenhouse were extended in a field study to assess the genetic and environmental variations of global DNA methylation on the first emerging leaf (rank 0) in 31 poplar genotypes grown in favorable or unfavorable field conditions (two contrasted pedoclimatic sites, Echigey (ECH), favorable, versus Saint-Cyr-En-Val (SCV), unfavorable growth conditions). ECH for Echigey site was characterized by fertile, silty-clay soil, and high soil water content during the growing season (>40% soil volumetric water content at 20 cm).³⁰ The SCV for Saint Cyr en Val site was characterized by a poor, sandy-loam soil type with low soil water content (<20%).³⁰ A significant genotype×treatment interaction was detected as already observed for the dormant SAMs (Figure 2). A similar range of variation of global DNA methylation was observed (16–50%; Figure 2) than those reported in the dormant SAM (20–54% in the SAM¹⁶) across the 31 unrelated genotypes and the two sites (Figure 2). Significant differences across sites were also detected for 17 out of 31 genotypes, while previously for the dormant SAM only 8 out of 31 genotypes were identified (Figure 2). These environmentally induced variations of DNA methylation in the first emerging leaves in field conditions are in agreement with the data in greenhouse (Figure 1). However, no correlation between DNA methylation percentages in dormant SAM and first emerging leaves could be detected (r = 0.7 at P = 0.71 for favorable and r = −0.03 at P = 0.85 for unfavorable growth sites) as well as between DNA methylation in first emerging leaf and biomass production (r = −0.07 at P = −0.05 for favorable and r = 0.18 at P = 0.08 for unfavorable growth sites). This is in contrast with the previously reported correlations between DNA methylation in SAM (active or dormant) and biomass production in poplar.^16,25 However, it could be explained by the recent work showing that negative genetic correlations between growth and SAM global DNA methylation in black poplar populations subjected to drought are compensated by positive environmental correlations resulting in some case in no significant phenotypic correlations.²⁴

Figure 2.

Genotypic variation and environmental effect for first emerging leaf global DNA methylation on different Populus deltoides × P. nigra hybrids grown in experiment 2. White and black bars correspond, respectively, to favorable (ECH) and unfavorable (SCV) growth conditions for experiment 2 (Toillon et al., 2013). Values are genotype means ± SE (n = 4). Global genotype variations and the effect of growing conditions were evaluated by ANOVA. G: genotype effect; T: treatment effect for growth conditions; G × T: interaction between genotype and treatment. *P < 0.05, **P < 0.01, ***P < 0.001; nothing is indicated if non-significant.

Altogether, these data and our recent reports^16,17,24 illustrate that global DNA methylation levels are developmentally, genetically, and environmentally determined in the SAM and its emerging leaves. We previously proposed that mitotic transmission occurs during the transition from the vegetative growth period to dormancy in the SAMs, constituting an epigenetic phenomenon and a potential stress memory.¹⁶ While this needs to be further tested, the possibility of a similar mitotic transmission from the SAM to the newly formed leaves, proposed here as a potential developmental memory, is an exciting perspective to dissect how developmental programs and phenotypic plasticity work as a whole, especially in long-living organisms such as trees. However, it is also possible that the sensitivity to environmentally induced variations of DNA methylation is directly realized on leaf primordia and not being transmitted from the SAM. As only emerging leaves (ranks 0 and 1) showed variations, this could suggest that there is a need for a sufficient duration of stress to get epigenetic modification as in the vernalization³¹ process or that these variations are not stable during leaf development and restricted to ranks 0 and 1 before being erased.

In this context, it will be of first importance to study the leaf formation in the SAM by measuring the number of leaf primordia under rank 0, estimating the leaf production rate in distinct environmental conditions, and determining a leaf plastochron index to reduce variations among individuals and focus on developmental effects. Then, it would be relevant to perform kinetics analysis along a repeated drought stress experiment (with different intensities and durations) in SAM and emerging leaves on poplar clones (no genetic variation among individuals), identify the loci affected by variations of DNA methylation in the SAM that are transmitted to and persist in the daughter leaves, and evaluate how this impacts the drought-induced leaf phenotype. Finally, it will be important to test the causal role of these epigenetic signatures in trees through a reverse genetic approach with hyper- or hypomethylated RNAi lines^23,28 (A-L. Le Gac and S. Maury, unpublished results), through CRISPR/Cas9 strategy,^32,33or using epigenetic quantitative, population, or integrative approaches.^21,34,35

Funding Statement

This work was supported by the Agence Nationale de la Recherche [ANR-08-BIOE-0006].

Acknowledgments

CLP and ALLG received Ph.D. grants from the Region Centre et le Ministère de la Recherche et Enseignement Supérieur. This work was funded by the ANR France, within the project BIOENERGIE SYLVABIOM (ANR-08-BIOE-0006). We acknowledge Prof. Franck Brignolas, Dr. Régis Fichot, and Dr. Sophie Jacquot for discussion and all colleagues that participate in the previously associated publications: Le Gac et al. 2018 (¹⁶) and Lafon Placette et al. 2018 (¹⁷).

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.