An RNA quality-control pathway has opposing roles in plant heat recovery

An early advantage that does not last

Plants can retain a molecular memory of heat exposure that helps them cope with a later episode. But preparing for stress and returning to growth require different kinds of regulation. A new bioRxiv preprint by Syed Hussam Abbas and colleagues suggests that nonsense-mediated mRNA decay, or NMD, helps coordinate these competing demands. The study, posted November 2, 2026, has not been certified by peer review.

NMD is an RNA surveillance pathway that removes certain messenger RNAs, including faulty transcripts and some normal transcripts with features that mark them for degradation. Because messenger RNAs carry instructions used to make proteins, controlling their persistence can influence both stress responses and development.

The researchers compared wild-type Arabidopsis thaliana with two mutants, upf1-5 and upf3-1, whose NMD function is impaired. Seedlings received a heat-priming treatment followed by a later, challenging heat exposure at 44°C. Recovery was assessed four and fourteen days after that challenge, with untreated plants providing controls.

At four days, both mutants appeared less affected than wild-type seedlings. Their leaves were larger and less shrunken, although they developed brownish-green pigmentation. By fourteen days, the pattern had reversed: wild-type plants recovered more successfully, while the mutants showed greater damage.

Physiological measurements in upf3-1 supported this reversal. During early recovery, the mutant retained more chlorophyll, had less electrolyte leakage—a measure of membrane damage—and maintained a higher effective yield of photosystem II than wild-type plants. During late recovery, those comparisons favored the wild type instead.

A head start outside the main memory pathways

To investigate the early advantage, the team examined gene activity before heat treatment, immediately after the priming-and-challenge sequence, and one day into recovery.

Even without treatment, upf3-1 showed modestly elevated expression across a set of heat-responsive genes. The authors interpret this as a partially stress-primed state that might help explain its initial tolerance. However, the mechanism connecting those RNA changes to protection remains unresolved.

Genetic experiments provided another clue. The researchers combined upf3-1 with mutations affecting either HsfA2 or HsfA3, transcription factors involved in heat-stress memory. Each double mutant was less sensitive during early recovery than the corresponding hsfa single mutant, although neither matched upf3-1 alone.

These results support an early protective contribution from UPF3 loss that does not require either factor individually. The benefit did not persist: during late recovery, the double mutants were not significantly different from the corresponding hsfa single mutants.

A changed balance of RNA variants

The longer-term disadvantage was associated with weaker expression of a branch of the heat response linked to HsfA2. By contrast, the delayed, HsfA3-correlated gene group showed little difference between genotypes. These expression relationships support the authors’ model, but correlation alone does not establish which factor directly controls every gene.

Separate measurements after priming revealed changes among RNA variants produced by the HSFA2 gene. Compared with wild type, upf3-1 had approximately 35% less of the functional HsfA2-I transcript, about six times as much HsfA2-II, and roughly two to two-and-a-half times as much HsfA2-III. The latter two variants are NMD targets; HsfA2-I is not.

An unspliced HSFA2 RNA measurement, used as a proxy for transcriptional activity, showed no significant genotype difference. The authors therefore propose an indirect change in RNA processing or handling, potentially shifting production toward the NMD-sensitive variants. The responsible mechanism was not established.

Recovery involves more than heat protection

Heat-treated upf3-1 plants also showed increased expression of immune-response genes and reduced expression of genes associated with growth and cell walls. The authors suggest that this altered growth–defense balance, together with weakened HsfA2-associated memory responses, may contribute to poor late recovery.

That explanation remains a model rather than a demonstrated causal chain. The study did not test whether suppressing the immune response would restore recovery, and possible consequences for pathogen resistance after heat remain untested.

The central finding is nevertheless clear within this Arabidopsis seedling assay: an early advantage can conceal a later recovery cost. The results argue against treating reduced NMD activity as a straightforward route to lasting heat tolerance.