henhaobio
Sign in

bioRxivEvolutionary BiologyPreprint

Repeated drought induces a reproducible DNA methylation response associated with gene expression in Quercus lobata

Lily D. PeckiD, Victoria L. SorkiD

PDFDOICode
In-depth full-text digest available · Jump to digest ↓

Abstract

Long-lived trees must continually adjust to environmental change and face sustained climatic shifts over their lifetimes. One increasingly important challenge is the rising frequency of drought caused by climate change. Environmentally responsive DNA methylation is widespread in plants, but whether it contributes to gene expression during environmental stress remains unclear, particularly in long-lived trees. Here, we integrated long read methylomes and transcriptomes from valley oak (Quercus lobata) seedlings exposed to repeated drought and well-watered treatments. Repeated drought induced a reproducible DNA methylation response that repeatedly targeted the same genomic regions despite turnover of individual methylated sites. These repeatedly targeted regions were transposable elements (TEs) located near genes. Genes adjacent to CHH-methylated TEs were enriched for core drought-response pathways, including abscisic acid signaling, osmotic adjustment and cell-wall remodeling, and remained transcriptionally activated under drought. However, higher CHH methylation levels were associated with progressively smaller transcriptional responses, suggesting that environmentally responsive DNA methylation influences how strongly drought-response genes are activated rather than simply switching them on or off. At the same time, greater CHH methylation was associated with continued repression of nearby TEs, suggesting that this response may simultaneously regulate gene activity while maintaining genome stability. Together, these findings identify a reproducible genome-regulatory response associated with repeated environmental stress in a long-lived tree. By repeatedly targeting the same genomic regions despite turnover of individual sites, this response provides a framework for how long-lived trees repeatedly adjust gene expression while maintaining genome stability during environmental change.

In-depth summary

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 2 (7 Sept 2026) · CC BY 4.0

Repeated drought in valley oak seedlings induced CHH hypermethylation recurring at the same gene-proximal transposable elements, associated with elevated yet methylation-attenuated expression of nearby genes.

Long-lived trees must adjust gene regulation within a lifetime as droughts intensify, yet methylation and transcription are usually analyzed separately, so direct tests of whether stress-induced CHH methylation at transposable elements (TEs) shapes nearby gene expression are rare. Pairing long-read methylomes with transcriptomes across repeated drought in Quercus lobata, it links reproducible TE-targeted CHH gains to the magnitude of neighboring transcriptional responses.

Key results

  1. ShownDrought left genome-wide methylation landscapes broadly stable but produced replicate-concordant differential methylation dominated by CHH gains, which emerged mainly after the second drought exposure; CG and CHG changes were rare and mostly losses.

    Differentially methylated positions (DMPs) per million cytosines, drought vs well-watered: CHH 64 vs 59 at T1-T2 and 1,834 vs 549 at T1-T3; CG 42 vs 18 and CHG 13 vs 9 at T1-T3. CHH DMPs were almost all hypermethylation (about 99.8% in sensitivity catalogues). Based on 24 PromethION methylomes from eight seedlings in two families. · Drought induces context-specific changes in DNA…; Sensitivity analyses

  2. ShownRepeated drought re-targeted the same transposable elements rather than the same cytosines: exact-site recurrence of CHH DMPs was limited, but TE-level recurrence far exceeded a permutation null, with no recurrence for CG or CHG DMPs.

    About 25% of first-exposure (T1-T2) CHH DMPs recurred within ±10 bp at T1-T3, with overlap rising as the matching window widened; roughly half of TEs with T1-T2 CHH DMPs again had CHH DMPs at T1-T3, a 5-6x enrichment over a genomic CHH permutation null (1000 permutations, P < 0.001), holding for each of the four drought-treated seedlings. · Drought induces context-specific changes in DNA…; Differentially methylated positions

  3. ShownDrought-responsive CHH DMPs concentrated inside TEs and in 1.5 kb gene-flanking regions, while CG DMPs avoided TEs; enrichment was structured by TE lineage, and copies of enriched families sat preferentially upstream of genes.

    TE overlap: CHH DMPs 95%, background CHH sites 68% (χ² = 78,331; enrichment 1.4-fold; adjusted P < 2.2 x 10^-16); CG DMPs 43%, background 83% (0.52-fold). TcMar-Stowaway, MuDR and tRNA-derived SINE superfamilies strongly enriched; 312 TE families significant (adjusted P < 0.05). Enriched-family copies: upstream odds ratio = 1.54, downstream 1.16. · Drought-associated CHH methylation…; Genomic annotation and enrichment analyses

  4. SuggestedDrought raised expression of genes carrying CHH-hypermethylated TEs in their 1.5 kb upstream region versus controls, but not for other TE positions or CG/CHG DMPs; gene set enrichment favored cell wall polysaccharide, pectin and wax biosynthesis plus xylan metabolism, and disfavored photosynthesis.

    Upstream difference-in-differences BH-adjusted P = 2.0 x 10^-4; Treatment x Time x Position F3,160020 = 5.2, P = 0.0014; per-gene T3-T1 log2 fold change model Treatment x Feature F3,77043 = 8.9, P = 6.7 x 10^-6. Candidate genes include KCS1, COR47, P5CS, PLIP2 and BLH6. Expression from five seedlings (three drought, two well-watered). · CHH-methylated upstream transposable elements…; RNA extraction, sequencing and orthology…; Expression responses associated with CHH…

  5. SuggestedExpression gains shrank as CHH methylation accumulated at associated TEs, across all TE positions; intragenic TEs showed lower expression with greater CHH gain under drought, while the TE transcriptome overall did not shift.

    Treatment x DMP number F1,76835 = 17.9, P = 2.3 x 10^-5 (Position interaction P = 0.23); Treatment x ΔCHH F1,77233 = 8.8, P = 0.0029. Intragenic TE slope (drought minus well-watered) -0.16 log2 CPM per unit ΔCHH, P = 0.045; no drought effect at average methylation (P = 0.58); three-way TE interaction P = 0.037. · Increasing CHH methylation is associated with…; Transposable element expression analyses

Methods, data and reproducibility

Oak seedlings were sampled before drought, after a 10-day drought, and during a second drought; Nanopore PromethION methylomes yielded replicate-concordant, control-masked differentially methylated positions (DMPs) via a difference-in-differences filter, related to TE annotations and 3' Tag-Seq expression using linear mixed-effects models.

  • Greenhouse seedlings grown from acorns of seven maternal trees at Sedgwick Reserve (CA), assigned to drought or well-watered treatment; transpiration tracked by porometer · Quercus lobata (valley oak) · n = 80 seedlings from seven maternal families
  • Nanopore PromethION leaf methylomes at T1, T2 and T3 from half-sib seedlings of maternal families 667 and 786 under drought and well-watered treatment · Quercus lobata (valley oak) · n = 24 methylomes from eight seedlings
  • 3' Tag-Seq leaf transcriptomes (3-6 million single-end reads per sample) with paired T1 and T3 samples from the same two families · Quercus lobata (valley oak) · n = five seedlings (three drought, two well-watered)
  • Valley oak 3.2 reference genome, gene models and RepeatMasker transposable element annotation · Quercus lobata (valley oak) · GCF 001633185.2
  • Arabidopsis orthologs assigned via best SwissProt hit (Trinotate) and UniProt-TAIR mappings for GO enrichment and methylation-machinery gene analysis · Arabidopsis thaliana
Methods & tools
Oxford Nanopore PromethION sequencing, Dorado 0.5.0 (basecalling and modified-base calling), modkit v0.2.8 (pileup, dmr pair), RepeatMasker / RepeatModeler TE annotation, bedtools (closest, map), 3' Tag-Seq, Trimmomatic v0.39, FASTQC v0.11.2, BBTools bbmap (SILVA rRNA filtering), STAR v2.7.11b, deepTools bamCoverage, lmerTest, emmeans, emtrends (R 4.4.0), clusterProfiler (GO gene set enrichment analysis), Licor Li-600 porometer
Software & environment
R 4.4.0 with lmerTest, emmeans, emtrends, clusterProfiler; Dorado 0.5.0; modkit v0.2.8; STAR v2.7.11b; Trimmomatic v0.39; FASTQC v0.11.2
Assay conditions & compute
Basecalling and all-context methylation calling on an A5000 GPU box; PromethION flowcell sequenced for 48 h.
Reproducibility gaps
Raw reads are said to be deposited at INSDC, but the BioProject accession is printed only as a placeholder (XXXXXXX); no link to processed methylation or expression data is given.
Limitations (5)
  • Several RNA-seq libraries failed quality control, leaving five seedlings (three drought, two well-watered) with paired T1 and T3 transcriptomes, and these only partially overlap the seedlings profiled for methylation. · stated by the authors
  • 3' Tag-Seq relies on oligo-dT priming, so expression covers only polyadenylated transcripts and cannot quantify the small RNAs that guide RNA-directed DNA methylation (RdDM). · stated by the authors
  • The molecular basis of the dose-dependent attenuation is unresolved, and whether weaker gene induction under CHH methylation benefits the plant or is an unavoidable cost of TE repression remains unknown. · stated by the authors
  • DMP concordance is described inconsistently: Results and the Figure 2 legend define DMPs as shared by all four drought seedlings, whereas Methods and sensitivity analyses use a three-of-four threshold for CHH, which recovered roughly 16-fold more CHH DMPs than the strict catalogue. · noted in this summary
  • Methylation-expression links are correlational and rest on few seedlings from two maternal families, with no experimental manipulation of methylation (such as RdDM mutants or demethylating treatment), so whether CHH gains causally tune transcription is untested. · noted in this summary

Read this if you study drought epigenomics or transposable element regulation in long-lived trees and want paired long-read methylome and transcriptome evidence linking CHH methylation to gene expression.

Citations
1

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISSow '20Peña‐Ponton '24Duplan '25Sánchez‐Luc '26Reynolds '25Steele '25
71% semantic matchPlant Biology
M. D. Sow, A.-L. Le Gac, R. Fichot+18 more·bioRxiv·Apr 16, 2020