How nitric oxide regulation relates to DNA methylation after plant stress

Plants respond to drought, heat, ozone and other environmental challenges in many ways, including changes in gene activity. DNA methylation—chemical marks on DNA—can also change. Some methylation differences may remain after conditions improve, but it is unclear how often this happens or what helps determine which marks persist. This study asks whether GSNOR, which helps regulate nitric oxide signalling, is associated with methylation patterns during repeated exposure and recovery.

Researchers compared wild-type Arabidopsis thaliana with a mutant lacking functional GSNOR. Separate groups experienced drought, elevated carbon dioxide, ozone, heat or a combined treatment during generations G1 to G3. The next two generations, G4 and G5, were grown under control conditions. Researchers sampled leaves in G1, G3 and G5 and used whole-genome bisulfite sequencing to measure DNA methylation. They also compared the results with matched RNA-sequencing data and previously published datasets.

The mutant began with a different methylome

Even under control conditions, the mutant’s methylation profile differed from the wild type. It also had many more differentially methylated regions (DMRs)—stretches of DNA whose methylation differed from the study’s comparison baseline—and an overall bias toward higher methylation. This matters because differences in a treated mutant cannot automatically be attributed to treatment: substantial differences were already associated with the mutation.

To assess persistence, the researchers tracked whether a region remained differentially methylated in the same direction in G1, G3 and G5. For treatment-associated CG-context DMRs, 0.9–4.5% of the mutant’s G1 regions met this definition, compared with 0.1–1.2% in wild type. Persistence was lower in the other sequence contexts. These percentages should not be confused with the much larger set of persistent differences in the mutant’s control comparison with wild type, which primarily reflects the baseline genotype difference rather than inheritance of a treatment effect. Overall, most treatment-associated changes were not detected as persistent across all three sampled generations.

The researchers examined three DNA sequence contexts, called CG, CHG and CHH, because different biological pathways help establish and maintain methylation in each. Persistent CG DMRs were often associated with gene bodies, while CHG and CHH DMRs were more commonly linked to transposable elements—DNA sequences capable of moving within the genome. Regions were also found at promoters and elsewhere. These results map where methylation differences occurred; they do not demonstrate what those differences do.

Methylation and gene activity

When the team compared methylation with RNA-sequencing results, more mutant candidate regions were associated with altered gene expression than wild-type candidates. But there was no universal relationship: higher methylation could coincide with lower, higher, unchanged or statistically nonsignificant expression, depending on the region and conditions. This comparison is correlational and cannot show that methylation caused expression changes. Nor does methylation at a transposable element establish whether it became more active or moved.

The study also compared its results with a published dataset of GSNOR-associated DMRs. Of 292 published regions, 211 overlapping regions had the same direction of methylation change in the current comparison. This supports a reproducible subset of mutant-associated differences, not every region identified here. Comparisons across experiments can be affected by differences in stress conditions, tissue, plant age, sequencing and analysis methods.

What remains uncertain

The findings associate GSNOR-dependent nitric oxide regulation with methylation patterns and their persistence in Arabidopsis. They do not reveal the precise mechanism. The authors cannot determine whether the mutant’s methylation bias results from reduced removal of marks, increased establishment or maintenance, or another indirect effect. The study also does not show that persistent marks benefit descendants, improve stress resilience, or apply to crops or natural settings.

This work is a bioRxiv preprint posted in November 2026; this version had not been certified by peer review. The sequencing included three biological replicates for control samples and two for each treatment, so treatment-specific results should be interpreted with that design in mind. Targeted experiments would be needed to test whether particular methylation changes cause expression changes. For now, the results suggest that GSNOR is linked to how the Arabidopsis methylome changes during repeated environmental exposure and recovery.