Strigolactones modulate jasmonate-dependent transcriptional reprogramming during wound signalling in Arabidopsis

To select wound-induced genes regulated by SLs, DEGs unique to WT were identified (Fig. 1, Table S2). This subset contained 2,186 genes (1,122 up- and 1,064 down-regulated). A comparable number was exclusive to max3 (2,102; 1,115 up- and 987 down-regulated). The largest group, 3,340 DEGs (1,976 up- and 1,364 down-regulated), was shared by both genotypes, indicating that only part of the wound response relies on SLs, with substantial SL-independent pathways acting in parallel. Opposing regulation was rare: 29 genes were up- in WT and down-regulated in max3, while 14 showed the inverse pattern (Fig. 1, Table S2).

Among genes up-regulated exclusively in WT following mechanical wounding, several were previously associated with JA signalling and wound-induced responses. This group includes AOC2 (AT3G25770, log₂FC = 2.45), OPR3 (AT2G06050, 1.15), DGL (AT1G05800, 5.55), and JAZ1 (AT1G19180, 2.60) (Table S2b). These genes encode enzymes or regulators directly involved in JA biosynthesis and signalling, or transcriptional modulators of wound-related pathways. In total, 35 genes related to JA biosynthesis or signalling, and 42 genes related to wounding response, were up-regulated exclusively in WT (Table S2b). However, transcript levels do not necessarily reflect hormone abundance, and no direct measurements of JA or JA-Ile were performed in this study. Future work involving targeted quantification of JA will be necessary to validate these inferences. In contrast, the list of genes down-regulated exclusively in WT included key photosynthesis-related genes such as LHCB3 (AT1G29910, log₂FC = –1.42), PSAK (AT1G30380, –1.29), CP29.1 (AT5G01530, –1.23), and RPE (AT3G01850, –1.16), encoding components of the photosystem I/II antenna and Calvin cycle (Table S2e).

However, among the 1,976 DEGs identified in WT and max3 after wounding (Table S2a), several encode key components of the canonical JA pathway. These include the lipoxygenases LOX3 (AT1G17420; log₂FC = 4.10 in WT, 1.84 in max3) and LOX4 (AT1G72520; 2.92, 1.79), the chloroplast enzymes AOC3 (AT3G25780; 4.21, 3.08) and AOS (AT5G42650; 2.28, 1.64), and the JA-responsive repressor JAZ10 (AT5G13220; 3.50, 2.41). In total, 52 genes annotated with "JA biosynthetic process" or "JA signalling" and 67 genes annotated with "response to wounding" were co-induced in both genotypes. The vast majority of these genes showed up-regulation in both genotypes following wounding (Table S2a and S2e).

GO enrichment highlighted a strong stress-related signature (Fig. S1). "Response to wounding" showed the highest fold enrichment (3.7; FDR = 2.4 × 10⁻⁸), followed by processes linked to jasmonate signalling ("response to fatty acid", "response to jasmonic acid") and detoxification/hypoxia. Secondary-metabolism terms, including sulphur-compound and organic-acid metabolism, were likewise over-represented, consistent with activation of specialised defence pathways. In contrast, GO terms over-represented among repressed genes were almost exclusively photosynthetic (Fig. S1). "Photosynthesis light reaction" (fold = 6.8; FDR = 1.1 × 10⁻¹⁶), "photosystem II assembly" (9.6), and "chloroplast organisation" (4.0) dominated the list, indicating a coordinated shutdown of light harvesting, electron transport and chlorophyll metabolism in WT that is absent in max3. Genes associated with the enriched GO categories are listed in Supplementary Table S2, which includes GO annotations for biological processes for all DEGs.

Next, the promoter sequences (1500 bp) of each WT-specific wound-induced DEG were screened to identify the TF binding sequences. Those analyses allowed for the identification of the 19 overrepresented TFs that regulate the transcription of WT DEGs (Table S4). Among them, several have been previously characterised as wound-induced and functionally relevant to tissue damage responses (Table 1). The AP2/ERF-domain TF WIND1 (AT1G78080) was known to act as a master regulator of wound-induced cellular reprogramming. WIND1 promotes callus formation and de novo shoot regeneration at wound sites, and its transcription is rapidly induced following injury (Iwase et al. 2017). Two members of the calmodulin-binding transcription activator family, CAMTA3 (AT2G22300) and CAMTA5 (AT4G16150), were also identified. CAMTA3 is required for wound- and herbivore-induced JA biosynthesis and defence activation (Qiu et al. 2012) (Qiu et al. 2012), while CAMTA5 binds the CCGCGT element in the promoters of ethylene-responsive genes and contributes to transcriptional activation under mechanical stress (Moore et al. 2022). Another AP2/ERF-domain TF, ORA47 (AT1G74930), previously shown to function as a positive regulator of the JA pathway in a COI1-dependent manner, was also predicted as a key regulator. Its expression is induced in the wound-related transcriptome dataset (ArrayExpress E-GEOD-41779). DDF1 (AT1G12610), another AP2-like factor, was up-regulated in light-dependent wound-response assays (Morker and Roberts 2011) and showed consistent induction in both root and leaf wounding datasets (ArrayExpress E-GEOD-41779, E-MTAB-7609). Additional TFs with transcriptomic evidence of wound-responsiveness include HSFB3 (AT2G41690), HSFA6B (AT3G22830), and AT5G59780. All were identified as up-regulated in independent microarray or RNA-seq datasets profiling Arabidopsis tissues at early timepoints post-wounding (ArrayExpress E-GEOD-41779, E-MTAB-7609). Importantly, since these TFs were wound-induced exclusively in WT but not in max3, this indicates that their activation depends on a functional strigolactone pathway.

Among 867 DEGs under control conditions, 12 encoded transcription factors (Table S5e). Three of them, AT3G53600, ZAT10, and ZAT12, have previously been implicated in the plant wound response, supporting the hypothesis that some key regulators of the SL-dependent transcriptional programme are constitutively dysregulated in the mutant. Promoter sequence analysis of the 867 misregulated genes revealed enrichment of binding motifs for 135 transcription factors (Table S5g). Among these, 9 TFs were previously predicted to regulate wound-induced genes specific to WT (Fig. 3) (Table S5h). Notably, five of these nine TFs have documented roles in the wound response (Table 1), further supporting their potential contribution to the impaired transcriptional activation observed in max3.

Together, these basal and interaction analyses show that (i) max3 seedlings are transcriptionally reprogrammed even before wounding, and (ii) a discrete cohort of genes requires SL signalling for full induction after injury, beyond constitutive differences between the genotypes.

Previous studies indicate that SLs modulate JA signalling at both transcriptional and metabolic levels (Nagata et al. 2015; Song et al. 2020; Lahari et al. 2024; Han et al. 2025). In this study, the SL biosynthesis mutant max3 exhibited impaired induction of canonical JA biosynthetic genes (DGL, AOC2, OPR3) (Hyun et al. 2008; Chini et al. 2018; Yang et al. 2023) and JA-responsive repressors (JAZ1) (Thines et al. 2007) following wounding, suggesting that SLs are required for the full activation of the early JA-dependent response to injury. It is important to note that these conclusions are based solely on transcriptomic data alone; no direct quantification of JA, JA-Ile, or related metabolites was conducted. Thus, the presented model should be viewed as a transcriptional hypothesis pending metabolic validation. Previous reports show that SL mutants display attenuated JA responses under stress conditions, including drought and pathogen infection (Piisilä et al. 2015; Lahari et al. 2024). In rice, SLs were shown to act as negative regulators of pathogen defence in rice by antagonising jasmonate (JA) signalling. SL mutants (biosynthesis—d10 and signalling—d14) and SL-depleted plants (treated with SL biosynthetic inhibitor TIS108) showed enhanced resistance to Pyricularia oryzae, accompanied by elevated levels of JA, sugars, and flavonoid phytoalexins. This enhanced defence required an intact JA pathway, as JA-insensitive mutants (jar1, coi1-18) did not benefit from SL depletion (Lahari et al. 2024). Recent multi-omics studies in barley support and extend these findings. In the SL-insensitive mutant hvd14, 2-week-old seedlings accumulate ~ 2.6-fold more JA than WT, whereas 4-week-old mutants show a ~ 40% reduction in JA content (Korek et al. 2025). This developmental inversion of JA levels is paralleled by transcriptional upregulation of lipoxygenase (LOX) genes, key enzymes in the JA biosynthetic pathway. Suggesting that SL signalling affects not only gene expression but also downstream metabolic flux or stability of JA. In line with this, proteomic data from hvd14 mutants revealed elevated levels of enzymes involved in early JA biosynthesis but no corresponding increase in late-pathway components or JA conjugates (Korek et al. 2025). These findings suggest that SL signalling influences JA-related processes at multiple regulatory levels, and that the effects may vary depending on tissue, developmental stage, or type of stress encountered. While SL mutants show impaired wound-induced JA responses in Arabidopsis, they may activate alternative defence mechanisms or accumulate JA differently under pathogen attack, as observed in rice and barley.

Promoter enrichment and protein–protein interaction analyses identified a transcriptional module regulated by SLs and centred around C2H2-type zinc-finger proteins (ZAT10, ZAT12), AP2/ERF-domain transcription factors (ERF5, ORA47, WIND1), and WRKY proteins. Several of these transcription factors are known integrators of jasmonate (JA) and reactive oxygen species (ROS) signalling (Davletova et al. 2005; Moffat et al. 2012). This supports a model in which SLs modulate early wound responses by enhancing the activity of transcriptional regulators that coordinate JA production and signalling (Fig. 4B).

SLs also appear to play an important role in maintaining ROS homeostasis under stress conditions. In barley, the SL-insensitive mutant hvd14 exhibits increased susceptibility to drought or heavy metal stress, as reflected by severe chlorophyll loss, reduced PSII efficiency, accumulation of anthocyanins, and elevated hydrogen peroxide levels, compared to WT plants (Daszkowska-Golec et al. 2023). Application of synthetic SL analogues (GR24) or the upregulation of antioxidant enzymes (e.g., SOD, CAT) ameliorates these effects in various species (Qiu et al. 2021; Wani et al. 2023; Shah et al. 2023). Given the well-established role of ROS as upstream signals in JA biosynthesis and wound responses, impaired ROS detoxification in SL-deficient genotypes may contribute to the reduced JA accumulation observed following wounding.

In addition to defence-related signalling, several TFs in the SL-dependent module are involved in post-injury tissue regeneration. Notably, WIND1 (Iwase et al. 2015, 2017) and WOX11 (Wan et al. 2023), transcriptional regulators required for callus formation and vascular reconnection, are induced specifically in WT plants and are predicted targets of JA/ROS-associated TFs (Iwase et al. 2017). These findings indicate that SLs contribute to reprogramming processes necessary for tissue repair and developmental re-patterning following mechanical damage.

Collectively, these results indicate that SLs regulate a transcriptional programme linking hormone signalling, oxidative stress responses, and regenerative processes. This programme is activated both prior to and in response to injury, and its disruption in SL-deficient plants compromises the efficiency of JA-mediated signalling, ROS homeostasis, and tissue regeneration.

The data support a model in which SLs act upstream of a transcriptional hub that integrates JA biosynthesis, reactive-oxygen signalling and tissue repair. In WT plants, wounding rapidly induces a set of C2H2 zinc-finger and AP2/ERF transcription factors, together with regulators of cellular reprogramming (Table 1). Promoter enrichment analysis suggests that these TFs may regulate a substantial portion of the SL-dependent wound transcriptome, including genes encoding core JA biosynthetic enzymes. Thus, it is proposed that SL signalling maintains these TFs at low basal levels while preserving their inducibility. Upon injury, SL perception relieves repression of the hub, allowing a rapid JA burst, ROS detoxification and activation of regeneration pathways.

In the max3 mutant, this regulation is uncoupled. Nearly one-fifth of the genes mis-expressed under control conditions belong to the WT wound-responsive set, indicating partial pre-activation of the hub. However, the same TFs and biosynthetic genes fail to reach WT expression after wounding, leading to insufficient JA accumulation and an attenuated defence switch. Constitutive up-regulation of negative regulators and impaired ROS homeostasis likely feed back to restrict further induction. Thus, SL deficiency creates a "primed-but-hyposensitive" state in which defence costs are incurred without full protective benefit. This working model may explain how SLs calibrate the balance between preparedness and responsiveness in the JA pathway and clarifies why SL-deficient mutants are handicapped in acute wound signalling despite exhibiting basal stress signatures.

This study presents a preliminary, transcriptome-based analysis of wound-induced responses in a single SL-deficient mutant (max3-9) at one timepoint (3 h post-wounding). As such, the conclusions drawn here are restricted to transcriptional dynamics and cannot directly infer changes at the metabolic, hormonal, or physiological level. In particular, no quantification of endogenous SL or JA levels was performed, and functional validation via phenotypic assays (e.g. pathogen resistance) was not included. These omissions reflect the exploratory nature of the study as well as current technical constraints. Future work should include time-resolved transcriptomic and hormonal profiling, targeted JA and SL quantification, and the use of additional SL mutants to validate the generality of the observed transcriptional patterns. Combining transcriptomic data with phenotypic or biochemical outputs would also allow for the identification of functionally relevant SL-JA interactions during tissue regeneration or defence. Despite its limitations, the dataset generated here provides a useful starting point and defines a testable framework for SL-mediated modulation of wound signalling.

The WT Arabidopsis thaliana plants used in this study belonged to the Columbia-0 (Col-0) ecotype. The SL-depleted mutant max3-9, lacking CCD7 (Carotenoid Cleavage Dioxygenase 7) function (Booker et al. 2004), was obtained from Prof. Christine Beveridge. This line was generated in the Col-0 background. WT and mutant seeds were propagated simultaneously and grown under identical conditions prior to the experiment. Arabidopsis seeds were sterilised by immersion in 70% ethanol containing 0.5% Tween 20 for 10 min, followed by three washes in absolute ethanol and subsequent air drying. Seeds were then sown on half-strength Murashige and Skoog (1/2 MS) medium (Murashige and Skoog 1962) and grown under controlled environmental conditions (22 °C, 16 h light/8 h dark photoperiod, 200 µmol m⁻² s⁻¹ light intensity). The wounding assay was performed on 4-week-old seedlings. Each Petri dish contained a single set of plants, either wounded or unwounded, representing a specific genotype. Mechanical wounding was carried out using a 0.5 mm diameter laboratory needle, puncturing each leaf at four distinct sites. Control plates were briefly opened (2 min) to match the exposure time during wounding. All samples were subsequently returned to the growth chamber for a 3-h recovery period.

Plant material (shoots) was harvested three hours after the onset of stress treatment and immediately snap-frozen in liquid nitrogen. For each biological replicate, tissue samples from four individual plants were pooled. A total of four biological replicates per genotype were prepared, each originating from a separate plate. Frozen samples were ground in liquid nitrogen using a mortar and pestle. Total RNA was extracted using the mirVana™ miRNA Isolation Kit (ThermoFisher Scientific, Cat. No. AM1560), following the manufacturer's protocol.

Polyadenylated transcripts were enriched from total RNA using oligo(dT)-attached magnetic beads. The isolated mRNA was then fragmented at elevated temperatures to yield appropriately sized RNA fragments. First-strand cDNA synthesis was performed using random hexamer primers and reverse transcriptase, followed by second-strand synthesis, during which dUTP was incorporated instead of dTTP to maintain strand orientation. Subsequently, the double-stranded cDNA underwent end repair, 3′-adenylation (A-tailing), and adapter ligation. Fragments of the desired size were selected, and strand specificity was ensured by digestion with the USER enzyme. Final libraries were generated through PCR amplification and purification steps. Library quality was assessed by Qubit fluorometry for concentration, real-time PCR for quantification accuracy, and fragment analysis with an Agilent Bioanalyzer. Only libraries meeting the quality criteria were pooled based on concentration and sequencing requirements and subjected to high-throughput sequencing on an Illumina platform, following Novogene's standard workflow.

Raw reads (FASTQ format) were subjected to quality filtering using in-house Perl scripts. Sequences containing adapter contaminants, poly-N stretches, or excessively low-quality bases were removed. The remaining high-quality reads (referred to as clean reads) were used for all subsequent analyses. Key quality metrics, including the percentage of bases with Phred scores ≥ Q20 and Q30 and overall GC content, were calculated. All bioinformatic procedures were carried out using only the filtered reads to ensure data reliability and reproducibility.

The Arabidopsis thaliana reference genome and annotation files (TAIR10 version) were downloaded from Ensembl Plants (https://plants.ensembl.org↗). Genome indexing and read alignment were performed using the splice-aware aligner HISAT2 (v2.0.5). Paired-end reads were mapped to the indexed reference genome with default settings.

HISAT2 was chosen due to its capability of integrating splice junction data from the provided annotation file, which enhances alignment accuracy and facilitates the identification of exon–intron boundaries and transcript variants.

Gene expression levels were estimated using featureCounts (v1.5.0-p3), which calculated the number of reads mapped uniquely to each annotated gene model. Raw counts were converted into FPKM (Fragments Per Kilobase of transcript per Million mapped reads), a normalisation metric that accounts for both gene length and sequencing depth, allowing for comparison across genes and samples.

While FPKM is not suitable for statistical tests of differential expression, it remains useful for descriptive analysis and visualisation of gene expression profiles.

Differential gene expression between experimental groups was assessed using the DESeq2 package (v1.20.0) in R. This method models count data using a negative binomial distribution, allowing for precise estimation of fold changes and statistical significance.

Raw counts from featureCounts served as input, and p-values were adjusted using the Benjamini–Hochberg procedure to control the false discovery rate (FDR). Genes with adjusted p-values (FDR) ≤ 0.05 were deemed significantly differentially expressed.

All differential expression analyses were performed within each genotype (WT_wound vs WT_control; max3_wound vs max3_control) to account for potential baseline expression differences.

Protein sequences of significantly regulated genes were retrieved from the TAIR10 dataset using the BioMart interface (https://plants.ensembl.org/biomart↗). These sequences were submitted to the PlantTFDB prediction tool (http://planttfdb.gao-lab.org/prediction.php↗) for transcription factor classification based on domain structure and sequence similarity.

Promoter regions, defined as the 1500 bp upstream of the start codon (ATG), were also retrieved using BioMart. To identify putative transcription factor binding sites (TFBS), these promoter sequences were analysed using the PlantTFDB binding site prediction tool (http://plantregmap.gao-lab.org/binding_site_prediction.php↗), with a significance cutoff of p ≤ 1e − 4.

To annotate the biological functions of selected Arabidopsis thaliana genes, Gene Ontology (GO) terms were retrieved using the Araport portal (https://www.araport.org/↗). A list of gene identifiers (AGI codes) was submitted to the"ThaleMine"search interface via the "Gene" data class. The output table was customised to include the associated GO terms, covering the three main categories: Biological Process, Molecular Function, and Cellular Component. The resulting data were exported in tab-delimited format (TSV) and used for downstream analysis.