PLASTID ENVELOPE ION CHANNELS (PEC1/2) link Ca²⁺ and jasmonic acid signaling in plant cells

Significance

Plants possess rapid defense responses that help them survive environmental stress. Calcium ions (Ca²⁺) act as universal messengers in these responses. However, their role in plastids, a cell organelle with central function in stress perception, hormone biosynthesis, and adaptation, was unclear. We identified PLASTID ENVELOPE ION CHANNELS (PECs) as mediators of plastid Ca²⁺ signals. PEC expression is induced by jasmonic acid (JA) to augment plastid Ca²⁺ signals during reoccurring stress, which sets the plant into a primed state. Loss of PECs disrupts this process, weakening stress-induced JA responses and defense. Conversely, PEC overexpression exacerbates plastid Ca²⁺ signals and improves pathogen tolerance. These findings link plastid Ca²⁺ signals with JA-controlled stress responses, and position PECs as mediators of plant defense.

Abstract

Plants constantly encounter adverse environmental interactions. One organelle is particularly specialized in stress signaling and phytohormone synthesis: the plastid. Calcium (Ca²⁺), a key second messenger, is known to intersect with cellular phytohormone signaling networks. While cytosolic Ca²⁺ dynamics have been studied extensively, the physiological relevance of stromal Ca²⁺ transients and the identity of channels mediating rapid Ca²⁺ flux into plastids remain largely unexplored. In this study, we provide evidence for PLASTID ENVELOPE ION CHANNELS (PECs) as long-sought mediators of fast-activating cation channel-like currents. We show that PEC expression is jasmonic acid (JA)-induced to augment stromal Ca²⁺ transients under stress. Loss of PECs results in decreased JA priming and failure to elicit full defense responses after wounding. In turn, PEC1 overexpression improves Botrytis cinerea tolerance. Our findings link stromal Ca²⁺ signaling with JA-dependent stress responses and position PECs as modulators of plant defense.

Challenged by abiotic and biotic stress, plants must constantly choose between growth and defense (1–3). This tradeoff is in part regulated by rapid signals such as reactive oxygen species (ROS) and Ca²⁺ transients, which elicit within seconds of stress perception (4, 5). Ca²⁺ acts as both a local and systemic signal, transmitting information within and between leaves (6–8). For instance, herbivory triggers cytosolic Ca²⁺ transients that rapidly activate defense programs through biosynthesis of phytohormones (6, 9–11).

The oxylipin phytohormone jasmonic acid (JA) regulates responses to abiotic stress, nutrient starvation, and development (12, 13). However, it is best known for its role in wound-induced defense (12). JA biosynthesis comprises a genetic feed-forward loop initiated in the plastid, where α-linolenic acid, released from membrane lipids by unresolved lipase(s), is oxygenated by 13-lipoxygenases (LOXs) (14, 15). Subsequent enzymatic processing yields 12-cis-oxo-phytodienoic acid (OPDA), which, upon transport into peroxisomes, is converted into JA. Finally, cytosolic JASMONATE RESISTANT 1 (JAR1) conjugates isoleucine to JA to yield bioactive JA-Ile that diffuses into the nucleus. Here, JA-Ile is sensed by CORONATINE-INSENSITIVE 1 (COI1), which degrades JA ZIM domain (JAZ) repressors. This releases MYC2/3/4 transcription factors (TFs) to upregulate JA biosynthesis genes, completing the feed-forward loop (16).

Links between Ca²⁺ and JA signaling have long been proposed: a) Ca²⁺ can promote JA biosynthesis in a MYC-independent fashion via Ca²⁺-dependent disintegration of the nuclear JAV1-JAZ8-WRKY51 suppressor complex (11) and b) some plant lipases and plastid LOXs reportedly require Ca²⁺ for activity (17–20).

Stress-induced cytosolic Ca²⁺ needs to be rapidly redirected into organelles to prevent toxicity (21, 22). Since several compartments possess their own Ca²⁺ signaling pathways, this redistribution can have profound physiological consequences, e.g., modulating Calvin–Benson–Bassham cycle enzymes (23). Chloroplasts, additionally, are capable of actively generating distinct stromal Ca²⁺ transients themselves in response to various stimuli (24–26). However, their physiological relevance remains poorly understood.

The study of mutants defective in cation transport across the plastid inner envelope (IE) may finally provide insights (27). Initially, CHLOROPLAST MITOCHONDRIAL CALCIUM UNIPORTER (cMCU) and BIVALENT CATION TRANSPORTER (BICAT) 2 were identified. However, both are unlikely to mediate rapid Ca²⁺ flux into the stroma of mesophyll chloroplasts, where the initial steps of JA biosynthesis take place, because a) localization of cMCU to chloroplasts is ambiguous and may require further clarification (28, 29), and b) bicat2 mutants are specifically affected in slow Ca²⁺ kinetics (30). A promising candidate is the FAST-ACTIVATING CHLOROPLAST CATION CHANNEL (FACC), first discovered in pea, which conducts K⁺ > Na⁺ > Ca²⁺ ≥ Mg²⁺ (31, 32). However, the molecular identity of FACC remained unknown.

Recently, two homologs coined PLASTID ENVELOPE ION CHANNELS (PEC1/2) were discovered in pea and Arabidopsis envelopes. Heterologous assays done in yeast suggested permeability for K⁺, and pec1pec2 null mutants exhibit dampened stress-induced stromal Ca²⁺ transients in response to all known stress triggers (33). Because of this highly specific molecular phenotype, we set out to a) test if PECs encode FACC and b) employ pec1pec2 mutants to gain an understanding of the physiological relevance of rapid stromal Ca²⁺ signals and the role of PECs for plants.

Results

PECs Mediate FACC-Like Currents and Facilitate Cytosol-to-Plastid Ca²⁺ Signaling in Arabidopsis thaliana.

To determine whether PECs mediate FACC-like currents, we established patch-clamp recordings on isolated Arabidopsis chloroplasts. As a reference, we first reproduced previously published electrophysiological measurements from isolated pea chloroplasts (Fig. 1A) (31). Intact chloroplasts were identified by the characteristic “halo” [due to the refractive index of undamaged envelope membranes (34)]. Excised inside-out envelope patches in symmetric 250 mM KCl showed current-voltage (I/V) relations closely resembling previous data, with minimal activity at ±50 mV (Fig. 1B) (31). Upon stepping the membrane potential to ±150 mV, currents increased significantly.

Fig. 1.

PECs mediate K⁺ and Ca²⁺ flux across the inner chloroplast envelope. (A) Schematic representation of the experimental setup for inside-out patch-clamp recordings from isolated chloroplasts. (B and C) Representative current traces recorded from chloroplasts isolated from Pisum sativum (B) and A. thaliana (C) in symmetrical 250 mM KCl solution. (D) Representative currents recorded in Arabidopsis WT chloroplasts after application of 100 µM GdCl₃ to the bath solution. (E–G) Quantification of current traces at −150 mV (E) and +150 mV (G) from chloroplasts of WT and two independent pec1pec2 alleles under symmetrical 250 mM KCl conditions. Bars represent mean ± SEM (n ≥ 5). Letters indicate statistically distinct groups (one-way ANOVA followed by Tukey’s post hoc test, P < 0.05). (F) Representative current traces for WT and pec1pec2 chloroplasts. (H and I) Representative traces after application of 100 µM GdCl₃ to the bath solution of pec1pec2-1 (H) and pec1pec2-2 (I) chloroplasts. (J) Cytosolic Ca²⁺ transients measured in WT and pec1pec2 seedlings following elicitation with 200 mM NaCl. Lines indicate mean, shaded areas indicate SEM (n = 12). (K) Quantification of peak cytosolic Ca²⁺ concentrations shown in (J). Bars represent mean ± SEM (n = 12). Letters indicate statistical significance (one-way ANOVA with Tukey’s post hoc test, P < 0.05). (L) Schematic representation showing proposed PEC function: In WT, PECs mediate plastid Ca²⁺ uptake from the cytosol during stress signaling. Loss of PECs traps Ca²⁺ in the cytosol and thus impairs stromal transients.

We next extended this approach to A. thaliana, which allowed for the analysis of mutants. FACC-like currents were detected in isolated WT chloroplasts, albeit I/V relations suggested that channel openings in Arabidopsis are less clearly voltage-dependent than in pea (Fig. 1C). Application of the cation channel blocker Gd³⁺ to the stromal side of the excised patch reduced the open-channel current. Similar to pea FACC, only prolonged incubation (4 to 5 min) led to a complete loss of channel activity (Fig. 1D and SI Appendix, Fig. S1 A and B).

To test whether PECs mediate these FACC-like currents in Arabidopsis, we performed patch-clamp recordings on isolated chloroplasts from two independent loss-of-function mutants, pec1-1pec2-1 and pec1-2pec2-2 (hereafter referred to as pec1pec2-1 and pec1pec2-2, respectively). Recordings were carried out in a double-blind manner. Remarkably, all patches formed with envelope membranes of pec1pec2 loss-of-function mutants displayed severely reduced currents (Fig. 1 E and G). Gd³⁺ further reduced these residual currents upon addition, albeit not significantly (Fig. 1 H and I and SI Appendix, Fig. S1 C and D). To confirm that the FACC-like currents are formed by cations rather than chloride, we next substituted KCl for potassium gluconate in both bath and pipette solutions (SI Appendix, Fig. S1E). This did not abolish the currents recorded in WT (SI Appendix, Fig. S1F), indicating that PECs are permeable for cations rather than chloride. Altogether, this strongly suggests that PECs mediate FACC-like currents in Arabidopsis chloroplasts.

The pea FACC has been shown to conduct K⁺, Na⁺, Mg²⁺, and Ca²⁺ (31). Concomitantly, pec1pec2 mutants exhibit strongly diminished stromal Ca²⁺ transients. This suggests that PEC-dependent FACC-like conductance may contribute to Ca²⁺ flux across the IE into chloroplasts, thus linking cytosolic and plastid Ca²⁺ signaling. However, it remained unclear whether these elicitor-induced (NaCl, Mannitol, etc.) stromal Ca²⁺ transients emerge from the cytosol. To investigate this in planta, we introgressed a cytosolic aequorin-based Ca²⁺ sensor [pMAQ2 (35)] into the pec1pec2 background and quantified cytosolic Ca²⁺ levels. In WT seedlings, we observed large Ca²⁺ transients exceeding 1,000 nM [Ca²⁺]_cyt in response to NaCl. This response was stronger in both pec1pec2 lines, i.e., cytosolic Ca²⁺ transients were significantly elevated compared to WT (Fig. 1 J and K), which indicates that loss of PECs partially impairs Ca²⁺ flux from the cytosol into plastids, trapping Ca²⁺ in the cytosol. Importantly, this increase in cytosolic Ca²⁺ was mirrored by the reduction of stromal Ca²⁺ transients in pec1pec2 (SI Appendix, Fig. S2 A–C). This suggests that cytosolic Ca²⁺ transients precede and drive the stromal Ca²⁺ signals during osmotic stress. In WT, PEC-dependent ion flux appears to directly or indirectly connect cytosolic and plastid Ca²⁺ signals by enabling rapid Ca²⁺ entry into the stroma. In pec1pec2, this connection is disrupted, which causes cytosolic Ca²⁺ to accumulate, while stromal transients are diminished (Fig. 1L).

PEC Expression Is Controlled by JA-Inducible Transcription Factors.

To gain a deeper understanding of the physiological relevance of PEC function, we investigated the regulatory mechanisms governing PEC expression. We screened PEC1/2 promoter regions (2 kb upstream of the start codon) for conserved TF binding sites using in vitro DNA Affinity Purification Sequencing (DAP-Seq) data (36, 37). This analysis identified binding events involving 74 different TFs in the PEC1 promoter, and 53 TFs in the PEC2 promoter (SI Appendix, Fig. S3 A and B and Dataset S1). Gene ontology (GO) enrichment revealed that TFs binding the PEC1 promoter are linked to stimulus-related hormone responses, whereas those associated with PEC2 seem more generally involved in development (SI Appendix, Fig. S3C and Dataset S2). Notably, both promoters contain two conserved G-box motifs (5’-CACGTG-3’), known binding sites for the MYC TF family (38). DAP-Seq confirmed that MYC2, MYC3, and MYC4 bind both G-box motifs in the PEC1 promoter in vitro, whereas the PEC2 promoter was only targeted by MYC3 and MYC4 (Fig. 2A and SI Appendix, Fig. S3D). MYC2 is partially redundant with MYC3 and MYC4, and acts as the master coordinator of JA-mediated stress responses (39).

Fig. 2.

PEC expression is controlled by the MYC branch of the canonical JA signaling pathway. (A) DAP-Seq profiles showing MYC2/3/4 binding to the PEC1 promoter. Binding events are colored from orange to purple by normalized peak height. (B) ChIP-Seq showing increased MYC2 binding to the PEC1 promoter region after water spray treatment compared to the untreated control. Enrichment over input was calculated (bin size = 1 bp), normalized by read count, and plotted over the PEC1 gene model (Chr5:683630-691192). (C) Luciferase activity in pPEC1::LUC2 reporter lines following treatment with 200 µM MeJA, 10 µM COR, or mechanical wounding. Photon counts were integrated over 3.5 h and images were taken at indicated time points; asterisks mark wounded leaves. Scale bar equals 1 cm. (D) Immunoblot analysis of PEC1 protein accumulation in 2-wk-old WT and pec1pec2-2 at indicated time points following treatment with 20 µM MeJA, 1 µM COR, or wounding. ACTIN or Coomassie Brilliant Blue (CBB) staining of RuBisCO LSU serves as loading control, respectively. (E) Immunoblot displaying PEC1 levels in 2-wk-old myc234 loss-of-function mutants compared to WT and pec1pec2-2 before and after 24 h of MeJA treatment. CBB staining of RuBisCO functions as loading control. (F) Basal PEC1 protein levels in 3-wk-old JA signaling mutants (lox2-1, jar1-1, coi1-16, myc234) compared to WT. ACTIN serves as loading control. (G) Schematic representation showing that PEC expression is induced MYC2/3/4-dependently upon activation of JA signaling.

To affirm MYC2 binding to PEC1 in planta, we further analyzed Chromatin Immunoprecipitation Sequencing (ChIP-Seq) datasets (40, 41). In untreated plants, MYC2 showed little association with the PEC1 or PEC2 promoter. However, MYC2 binding to the PEC1 promoter was strongly enriched in stress-treated plants (Fig. 2B), which was also observed for PEC2 (SI Appendix, Fig. S3E), albeit to a lower degree, in line with DAP-Seq analysis (SI Appendix, Fig. S3B). Given that MYC2 is a prominent regulator of JA-mediated responses, this suggests that PEC1 may be the more JA-responsive isoform.

To confirm that JA induces PEC1 expression, we treated Arabidopsis pPEC1::LUC2 reporter lines with methyl jasmonate (MeJA) or coronatine (COR), a JA-Ile mimic (42), or by wounding and monitored PEC1-driven LUC2 expression (Fig. 2C). After 3.5 h, luminescence was considerably higher in all treated plants compared to controls, confirming PEC1 promoter activation.

This raised the question how JA-dependent PEC1 promoter regulation translates into transcriptional dynamics during JA signaling. Thus, we reanalyzed transcriptomic data from a 16-h time series following MeJA treatment (43). Similar to established JA marker genes, PEC1 and PEC2 transcript abundance increased significantly within 1 to 2 h and gradually declined over the next 14 h. However, PEC1 expression remained ~2.5 to 20 times higher than PEC2 at all times (SI Appendix, Fig. S4). Broader analysis of JA-responsive genes revealed at least three distinct temporal expression patterns: 1) very early responders, peaking and subsiding within 1 to 2 h (e.g., OPR3, JAR1) (SI Appendix, Fig. S4A), 2) rapid and transient responders, peaking within 1 to 4 h and then returning to baseline (e.g., AOS, AOC1) (SI Appendix, Fig. S4B), and 3) rapid and sustained responders, which reach peak expression within 1 to 4 h and remain at least 30% elevated throughout the 16 h period (e.g., LOX2, MYC2) (SI Appendix, Fig. S4C). Based on this classification, PEC2 acts as a rapid, transient JA responder, with expression peaking ~2 h posttreatment and returning to baseline by 12 h. PEC1, by contrast, shows a rapid and sustained JA-controlled expression pattern, with a broad peak expression between 1 to 4 h posttreatment and no return to baseline. Notably, PEC1 and LOX2 share similar expression patterns, with a rapid MeJA-dependent increase followed by sustained expression. This sustained induction pattern, together with its MYC2-dependency and overall higher expression levels, again suggests PEC1 as the more prominent isoform during stress signaling.

To address whether the coinduction of PEC1 and PEC2 with JA marker genes reflects potential cofunction, we verified PECs’ temporal- and cell-type-specific expression in Arabidopsis (44). This analysis revealed that PECs and selected JA marker genes (e.g., LOX2, AOC2) have similar diurnal expression patterns (SI Appendix, Fig. S4D) and are expressed mainly in mesophyll and phloem tissues, confirming that their diurnal and JA-dependent coexpression occurs within the same cell types.

Next up, we monitored PEC1 accumulation over 24 h after treatment with MeJA, COR, or wounding to determine whether the JA-induced expression of PECs translates to changes at the protein level. Across all treatments, PEC1 protein levels remained largely unchanged 1.5 h after treatment, increased after 3 h, and accumulated substantially at 12 and 24 h posttreatment (Fig. 2D and SI Appendix, Fig. S5 A–C).

Subsequently, we tested the extent to which JA-induced PEC1 accumulation is dependent on the MYC TFs. For this, WT and myc234 null mutants (39) were treated with MeJA for 24 h. While WT showed a strong accumulation, PEC1 levels in myc234 remained low (Fig. 2E and SI Appendix, Fig. S5D), which confirms that stress-triggered PEC1 expression is largely MYC2/3/4-dependent.

Further analysis of basal PEC1 levels in untreated lox2-1 (45), jar1-1 (46), myc234, and coi1-16 (47) mutants, all of which harbor defects in key components of the JA signaling pathway, revealed a clear pattern: Basal PEC1 protein accumulation progressively declined with increasing severity of JA signaling disruption (i.e., myc234 and coi1-16 exhibit lowest PEC1 levels) (Fig. 2F and SI Appendix, Fig. S5E). This shows that even basal PEC1 expression, in part, relies on intact JA biosynthesis and signaling. However, residual PEC1 detected in these mutants suggests that basal expression is maintained by additional, JA-independent factors (Fig. 2G).

Stress-Induced PEC Accumulation Augments Stromal Ca²⁺ Signaling.

Given the role of PECs in mediating stromal stress-induced Ca²⁺ transients (33), we investigated whether JA-induced changes in PEC levels alter stromal Ca²⁺ dynamics. Thus, we pretreated Arabidopsis WT and pec1pec2 seedlings expressing plastid-targeted YFP-Aequorin (sAEQ) (33) with or without MeJA (24 h) and subsequently elicited stromal Ca²⁺ signals with NaCl (Fig. 3A). WT seedlings pretreated with MeJA exhibited significantly higher stromal Ca²⁺ transients compared to the untreated controls. By contrast, pec1pec2 mutants displayed consistently dampened Ca²⁺ transients and did not exhibit increased Ca²⁺ levels upon MeJA treatment (Fig. 3 B and C and SI Appendix, Fig. S5 F–H). This shows that plants can modulate stromal Ca²⁺ levels through JA-controlled PEC expression (Fig. 3D).

Fig. 3.

Accumulation of PEC1 augments stromal Ca²⁺ transients JA-dependently. (A) Schematic workflow for MeJA treatment and Ca²⁺ quantification. 5-d-old seedlings carrying stromal Aequorin were incubated with 200 µM MeJA or a control solution for 24 h prior to Ca²⁺ measurements. (B) Stromal Ca²⁺ transients measured in WT and pec1pec2-2 seedlings with or without MeJA treatment following elicitation with 200 mM NaCl. Lines indicate mean, shaded areas indicate SEM (n = 24). (C) Quantification of peak stromal Ca²⁺ concentrations shown in (B). Bars represent mean ± SEM (n = 24). Letters denote statistically distinct groups (one-way ANOVA with Tukey’s post hoc test, P < 0.05). (D) Schematic model of the proposed mechanism: MeJA treatment induces PEC1 expression, leading to enhanced plastid Ca²⁺ influx and thus elevated stromal Ca²⁺ transients. (E) Immunoblot showing basal PEC1 protein levels in WT, fou2, fou2pec1pec2, and pec1pec2 mutants. RuBisCO LSU was stained with CBB as loading control. (F) Stromal Ca²⁺ transients in response to NaCl elicitation in WT, fou2, pec1pec2-1, and fou2pec1pec2-1 mutants. Lines represent mean, shaded areas indicate SEM (n ≥ 8). (G) Quantification of peak stromal Ca²⁺ concentrations shown in (F). Bars represent mean ± SEM (n ≥ 8). Letters indicate statistically distinct groups (one-way ANOVA with Tukey’s post hoc test, P < 0.05). All data are representative of at least three independent biological replicates.

To further verify our hypothesis, we employed fou2, a mutant with constitutively elevated JA levels (48). Since fou2 accumulated more PEC1 protein already at basal levels (Fig. 3E and SI Appendix, Fig. S5I), we outfitted fou2 with plastid-targeted YFP-Aequorin and additionally isolated fou2pec1pec2 triple mutants (SI Appendix, Fig. S5J). Stromal Ca²⁺ transients in fou2 were significantly higher compared to untreated WT and largely resembled those of MeJA-treated WT. By contrast, fou2pec1pec2 triple mutants, like pec1pec2, exhibited strongly diminished stromal Ca²⁺ transients in response to NaCl (Fig. 3 F and G). This confirms that JA signaling tunes stress-induced stromal Ca²⁺ signals mainly by controlling PEC expression (Fig. 3D).

Loss of PECs Impairs Defense Gene Expression and JA Accumulation.

For an overview of the downstream effects of altered stromal Ca²⁺ dynamics, we performed RNA-Seq on pec1pec2-1 and WT plants under control conditions and 90 min after wounding (Fig. 4A). Initially, gene expression patterns were explored by k-means clustering (k = 6) of the top 2,000 most variable genes 90 min after wounding (SI Appendix, Fig. S6A). Cluster_2 (409 genes, Dataset S3) displayed the most uniform expression changes, and heatmap visualization revealed a consistent trend toward lower expression of cluster_2 genes in pec1pec2-1 (Fig. 4 B and C). Interestingly, GO term enrichment of genes in cluster_2 indicated significant overrepresentation of genes involved in stress-related processes i.e., response to water deprivation, wounding, cold, and JA (Fig. 4D and Dataset S4). Differentially expressed genes (DEGs) were identified by a statistical analysis across the entire dataset. This allowed us to identify differences in steady state transcriptional abundances between untreated and wounded WT and pec1pec2-1 plants and to assess how transcriptional wounding responses differ between genotypes. A heatmap of the most diverse genes and Principal Component Analysis (PCA) revealed that the overall transcriptional wounding response remains largely intact in pec1pec2-1, and, in line with the 16-h time series, revealed ~6 to 7 times higher expression of PEC1 compared to PEC2 in WT (SI Appendix, Figs. S6B and S7 A and B). However, we identified 73 DEGs in pec1pec2-1 compared to WT in response to wounding, out of which only 20 were shared with cluster_2 (Fig. 4B). 60 DEGs were downregulated and only 13 upregulated in pec1pec2-1 vs. WT (Fig. 4C and Dataset S5). These genes were enriched for two major functional groups: “metabolism” (e.g., JA and fatty acid metabolism) and “response to stimulus” (e.g., wounding, oxygen-containing compounds, hormone, and JA) (Fig. 4E and Dataset S6). Intriguingly, the downregulated genes included JA pathway components such as LOX3 and JAR1 (SI Appendix, Fig. S7C). This indicates that while the global response to wounding in pec1pec2 remains largely intact, loss of PECs attenuates expression of a subset of JA-related, wound-induced genes, which leaves the plant unable to elicit the full defense response.

Fig. 4.

Loss of PECs attenuates wounding-induced transcriptional and hormonal responses. (A) Schematic overview of the experimental workflow for short-term transcriptomic analysis. 3-wk-old WT and pec1pec2-1 mutants were wounded and harvested after 90 min. (B) Venn diagram showing the overlap between genes in k-means cluster_2 (SI Appendix, Fig. S6A) and differentially expressed genes (FDR < 0.05) in pec1pec2-1 and WT, and after wounding (pec1pec2 DEGs). (C) Heatmaps of expression profiles (z-scored transcript abundance) for genes in k-means cluster_2 (Left) and pec1pec2 DEGs (Middle). For pec1pec2 DEGs, corresponding log₂FCs are shown on the Right. (D) Selected enriched GO-terms of genes in k-means cluster_2 (Benjamini–Hochberg FDP < 0.05). Circle size corresponds to the number of cluster_2 genes in the GO term, and coloring depicts fold enrichment. (E) Selected enriched GO terms of pec1pec2 DEGs. Circle size corresponds to the number of DEGs in the GO term, and coloring depicts fold enrichment. (F) Experimental workflow of hormone quantifications. WT and pec1pec2 seedlings were primed twice by wounding, then subjected to a final wounding or left untreated (Control). Samples were harvested 60 min after the final treatment. (G–I) Quantification of cis-OPDA (G), JA (H), and JA-Ile (I) in primed WT and pec1pec2 after wounding or at control conditions (n = 4). Bars represent mean ± SEM. Statistical significance was determined within each treatment by two-way ANOVA with Šidák correction, *P < 0.05, **P < 0.01.

As PEC1 expression is rapidly induced by JA, our transcriptomic analysis mostly focused on its effects on the early wound response. However, PEC1 expression, like that of LOX2, remains elevated long after JA induction (SI Appendix, Fig. S4C). Thus, we decided to further explore its role in the sustained phase of the defense response by examining PEC1 and LOX2 protein levels over 96 h after wounding in WT and pec1pec2. In WT, LOX2 and PEC1 levels strongly increased between 24 to 48 h and remained elevated over 96 h, consistent with their prolonged expression after MeJA treatment. Strikingly, in pec1pec2, LOX2 protein levels did not increase notably at any point during the 96-h period beyond basal level (SI Appendix, Fig. S7 D and E). Thus, PECs seem crucial for sustained LOX2 protein accumulation after stress.

Since this implies that loss of PECs is most disruptive days after wounding, we examined hormone levels during sustained stress. To this end, we primed WT and pec1pec2 mutants by wounding twice during development followed by a final wounding or mock treatment (=unwounded) 60 min before sampling, and quantified cis-OPDA, JA, and JA-Ile (Fig. 4F). In primed but unwounded plants, OPDA was reduced to ~30% of WT levels in both pec1pec2 mutants (Fig. 4G). This trend persisted after wounding but was no longer significant. Reduced OPDA accumulation during priming also affected JA levels: While JA was equally low in primed but unwounded plants, pec1pec2 accumulated only ~66% of WT JA levels after a final wounding (Fig. 4H). JA-Ile followed JA levels by trend, reaching only ~75% of WT levels in pec1pec2 after a final wounding (Fig. 4I).

In unprimed plants, hormone levels showed a similar pattern that was generally less pronounced (SI Appendix, Fig. S7 F–I). In conjunction with prolonged PEC1 expression, this indicates that PECs may play a more prominent role in sustained defense rather than the immediate wounding response.

PEC1 Overexpression Exacerbates Stress-Induced Stromal Ca²⁺ Transients, Augments the JA-Dependent Priming Response, and Increases Resistance Against Botrytis cinerea.

Since loss of PECs attenuates stromal Ca²⁺ transients and overall stress response, we wondered whether PEC1 overexpression would similarly alter plastid Ca²⁺ signals and potentially benefit defense. For this, we generated pec1pec2 mutants stably expressing the PEC1 genomic locus under its native promoter but fused to the HSP18.2 terminator (49), known to stabilize mRNA (50). Three independent overexpression lines designated PEC1ox-3, PEC1ox-6, and PEC1ox-11 were isolated, all accumulating more PEC1 protein than WT to varying degrees. PEC1ox-3 showed moderately increased PEC1 levels, while PEC1ox-6 and PEC1ox-11 were strong overexpressors (Fig. 5A). We verified that the cloned PEC1 promoter region remained JA-responsive, as MeJA treatment further increased PEC1 protein levels substantially in all overexpression lines to a similar extent as in WT (Fig. 5B and SI Appendix, Fig. S8A).

Fig. 5.

PEC1 overexpression enhances stromal Ca²⁺ transients, the JA-dependent priming response, and resistance to B. cinerea. (A) Immunoblot showing PEC1 protein levels in WT, pec1pec2, and three independent PEC1 overexpression lines (PEC1ox-3/-6/-11). Coomassie Brilliant Blue (CBB) staining of RuBisCO LSU serves as loading control. (B) Immunoblot confirming JA-responsiveness of the pPEC1::PEC1 construct. Two-wk-old plants were treated with 20 µM MeJA for 24 h. CBB staining of RuBisCO indicates equal loading. (C) Stromal Ca²⁺ transients in response to 200 mM NaCl in WT, pec1pec2, and PEC1ox-11. Shaded areas indicate SEM (n ≥ 16). (D) Quantification of peak Ca²⁺ levels from (C) and pec1pec2-2 and PEC1ox-6 (SI Appendix, Fig. S9 B and C). Data represent mean ± SEM (n ≥ 16). Letters indicate statistical significance (one-way ANOVA with Tukey’s test, P < 0.05). (E) Schematic overview of the experimental workflow for long-term transcriptomic analysis. Plants were either left untreated as time-matched controls or subjected to priming. WT, PEC1ox-6, and PEC1ox-11 mutants were primed by wounding and harvested 48 h or 96 h after priming. For 96 h samples, plants were subjected to a second priming at 48 h. In parallel, primed plants were wounded at the respective time points and harvested 90 min later to assess the acute wounding responses. (F) Hierarchically clustered heatmaps of the z-scored transcript abundances (z-scores of normalized read counts) of genes differentially expressed in WT upon wounding (Wound vs. Primed) at 48 and 96 h (SI Appendix, Fig. S10A) at control, Primed, and Wound conditions. (G and H) Violin plots showing the z-scored transcript abundances of genes in DEG Cluster 1 (G) and Cluster 2 (H) identified from (F) during control conditions, priming and acute wounding. Letters indicate statistically distinct groups (Kruskal-Wallis test followed by Dunn’s post hoc test, P < 0.05). Dashed lines indicate the median, while the dotted lines denote the upper and lower quartiles. (I) Line plots showing z-scored transcript abundances in PEC1ox-6 relative to WT at 48 h. The selected DEGs are unique in PEC1ox at 48 h, either in Primed (Primed vs. Control) or Wound (Wound vs. Primed) contrasts (SI Appendix, Fig. S11 A and B). Displayed PEC1ox unique DEGs are color-coded by functional category: JA signaling, pathogen response, redox/detoxification, and hormone crosstalk. Bubble plots below show the top four significantly enriched GO terms (g:SCS-corrected, P < 0.05) among all PEC1ox unique DEGs at 48 h Primed and Wound contrasts. Bubble size corresponds to the intersection size (number of DEGs annotated to the GO term), and coloring depicts fold enrichment. (J) Line plots showing z-scored transcript abundances in PEC1ox-6 relative to WT at 96 h. The selected DEGs are unique in PEC1ox at 96 h, either in Primed (Primed vs. Control) or Wound (Wound vs. Primed) contrasts (SI Appendix, Fig. S11 C and D). Displayed PEC1ox unique DEGs are color-coded by functional category: JA signaling, defense associated molecules, cell wall remodeling, and hormone crosstalk. Bubble plots below show the top four significantly enriched GO terms (g:SCS-corrected, P < 0.05) among all PEC1ox unique DEGs at 96 h Primed and Wound contrasts. Bubble size corresponds to the intersection size (number of DEGs annotated to the GO term), and coloring depicts fold enrichment. (K) RGB and false color panel showing F_v/F_m of representative lesions in adult leaves of 5-wk-old WT, pec1pec2, and PEC1ox-3/-6/-11 3 d after infection with B. cinerea. (L) Quantification of the lesion size from (K). Bars represent mean ± SEM (n ≥ 16). Statistical significance was assessed by unpaired two-tailed t test against WT (*P < 0.05; **P < 0.01; ***P < 0.001; P < 0.0001). (M) Trypan Blue staining of infected leaf discs 2 d postinfection to analyze primary lesion formation. Images are representative of six leaves per genotype.

Phenotypic appearance, photosynthetic performance, and chlorophyll content remained largely unaffected in PEC1 overexpressors. However, all PEC1ox lines were smaller and accumulated less fresh weight than WT or pec1pec2, which may reflect a growth penalty, typical for mutants with elevated JA levels (51) (SI Appendix, Fig. S8 B–I). To ensure that this phenotype does not stem from general alterations in the plastid ionome caused by PEC1 overexpression, we analyzed the total leaf and chloroplast ionome (52). Metal ion content of whole rosettes as well as isolated chloroplasts was unchanged in both pec1pec2 and PEC1ox lines (SI Appendix, Fig. S8 J–M and Dataset S7). This further indicates that PECs are not involved in general plastid cation loading but fulfill a highly specialized role, likely conducting only under specific physiological conditions i.e., during stress-induced Ca²⁺ signaling.

Thus, we hypothesized that PEC1 overexpression may enhance stress-induced stromal Ca²⁺ transients, and with this, plant defense. To test this, we outfitted PEC1ox-6 and PEC1ox-11 with plastid-targeted YFP-Aequorin (SI Appendix, Fig. S9A). Already at baseline, PEC1ox-6 and PEC1ox-11 exhibited significantly enhanced NaCl-induced stromal Ca²⁺ transients comparable to MeJA-treated WT, and MeJA treatment exacerbated stromal Ca²⁺ transients even further (Fig. 5 C and D and SI Appendix, Fig. S9 B and C).

Enhanced stromal Ca²⁺ transients may alter JA responses. Therefore, we assessed PEC1ox sensitivity to MeJA-induced primary root growth inhibition (46). PEC1ox exhibited shorter roots than WT under control conditions (SI Appendix, Fig. S9 D and E), and MeJA further exacerbated this effect, mirroring the fou2 mutant (SI Appendix, Fig. S9 F and G). Conversely, pec1pec2 roots developed similarly to WT under control conditions (SI Appendix, Fig. S7 D and E), but, surprisingly, grew significantly longer on MeJA plates, resembling the MeJA-insensitive myc234 mutants (39) in root length, but not overall appearance (SI Appendix, Fig. S9 F and G). This indicates that changes in PEC expression affect JA sensitivity.

Next, we determined whether PEC1 overexpression alters defense priming i.e., whether sustained wounding shifts the basal and short-term wound-induced transcriptomic profile over a medium (48 h) or prolonged (96 h) time frame, mirroring the JA quantifications (Fig. 4 F–I). To that end, we performed RNA-Seq on unwounded and wounding-primed WT, PEC1ox-6, and PEC1ox-11 plants. Tissue was collected at 48 h after an initial wounding to capture the primed state, or 90 min after a second wounding applied at that 48 h point to capture the acute transcriptional response. For an extended analysis, plants were wounded a second time at 48 h and then wounded again at 96 h, generating a double-primed state prior to sampling (Fig. 5E).

To assess whether transcriptomes are changed during defense priming in WT plants, we selected wound-responsive genes differentially expressed at both 48 and 96 h (SI Appendix, Fig. S10A and Dataset S8). Hierarchical clustering of transcript abundances separated genes into two major clusters: genes mildly yet significantly repressed during priming and strongly downregulated after wounding (Cluster 1), and genes slightly elevated during priming and further induced upon wounding (Cluster 2; Fig. 5 F–H). Notably, both clusters showed a significantly stronger priming response (Primed) at 96 h after a second priming, which indicates that repeated stress progressively reinforces plant defense on a transcriptional level. However, the acute short-term wounding response (Wound) was comparable between plants primed for 48 and 96 h, suggesting that the depth of the primed state does not further amplify the immediate transcriptional response to wounding. GO term enrichment revealed distinct differences between both clusters: Cluster 1 was enriched for genes associated with circadian clock and light-responsive processes (SI Appendix, Fig. S10B and Dataset S9). By contrast, Cluster 2 showed significant overrepresentation of genes involved in stress-related processes, i.e., response to stress, wounding, and JA (SI Appendix, Fig. S10C and Dataset S9), indicating that the upregulated component of the priming response is in part JA-dependent.

To test whether the defense priming relies on PECs, we compared the transcriptomic responses between PEC1ox lines and WT. We focused on genes uniquely responsive in both PEC1ox-6 and PEC1ox-11 relative to WT, applying a moderate fold-change threshold for the primed state (|log₂FC| > 0.5) and a more stringent threshold for acute wounding (|log₂FC| > 1), reflecting the stronger transcriptional changes induced by direct wounding (SI Appendix, Fig. S11 A–D and Dataset S8).

At 48 h priming, GO term enrichment of PEC1ox unique genes identified JA signaling related genes as the most prominent category, which indicates that PEC1 overexpression modulates the JA-dependent primed state (Fig. 5I and Dataset S9). To visualize transcriptional patterns within PEC1ox uniquely induced genes, we plotted z-scored transcript abundance relative to WT for selected genes associated with JA signaling, redox/detoxification, pathogen response, and hormone crosstalk, using PEC1ox-6 as the representative PEC1ox line. In line with the GO term enrichment, at 48 h Primed, PEC1ox showed a stronger induction of core JA signaling genes together with redox/detoxification markers, while selected pathogen response and hormone crosstalk genes were more repressed (Fig. 5I and SI Appendix, Fig. S11E). Upon rewounding (48 h Wound), JA- and redox-related genes in WT reached expression comparable to PEC1ox, whereas pathogen response and hormone crosstalk genes were more strongly induced in the overexpression line. This indicates that PEC1 overexpression shifts specifically the primed state toward JA-dependent processes, while during acute wounding, it mainly results in enhanced pathogen response and hormone crosstalk.

At 96 h Primed, PEC1ox-specific differences in core JA-associated genes became less pronounced (Fig. 5J). Instead, the PEC1ox unique priming response after 96 h was characterized by enhanced expression of genes linked with defense-associated molecules and cell wall remodeling (Fig. 5J and SI Appendix, Fig. S11F). This was also reflected in the GO term enrichment, which shifted toward broader defense and biosynthetic metabolism (Dataset S9). Upon acute wounding (96 h Wound), PEC1ox again showed a stronger induction of some JA and hormone crosstalk related genes, indicating that PEC1 overexpression continues to affect the transcriptional response, and JA dependent gene expression, even after prolonged priming (Fig. 5J). Overall, the transcriptomic responses during defense priming were complex, with aspects of the response varying across time points and wounding regimes. Nevertheless, one feature emerged consistently: JA-related processes were disproportionately represented among PEC1-dependent priming responses, both in terms of GO term enrichment and individual gene expression patterns.

Since JA signaling is a key defense pathway against necrotrophic pathogens (12), we finally tested whether these changes in JA-related processes are reflected in resistance against B. cinerea. When infected with B. cinerea, strong overexpressors PEC1ox-6 and PEC1ox-11 developed significantly smaller lesions and accumulated less fungal biomass than WT, while moderate expressor PEC1ox-3 behaved statistically indistinguishable from WT. By contrast, pec1pec2 displayed in most replicates significantly larger lesions compared to WT, as well as stronger primary lesions (Fig. 5 K–M and SI Appendix, Fig. S12A).

Together, these results establish a functional link between PEC-mediated regulation of stromal Ca²⁺ signaling and the modulation of downstream JA-mediated physiological defense responses.

Discussion

The ability of plastids to integrate environmental signals is central to their role in plant stress adaptation as they host key biosynthesis and regulatory steps for stress-related hormones such as JA, SA, and ABA (53). While Ca²⁺ signaling reportedly intersects with all three (54), most studies have focused on their links to cytosolic Ca²⁺ dynamics. By contrast, the physiological significance of stress-induced stromal Ca²⁺ transients, and plastid Ca²⁺ influx mechanisms remain largely unknown (27). Electrophysiological studies on pea chloroplasts identified the FACC with high Ca²⁺ permeability in the inner envelope (31). However, its molecular identity remained elusive. In Arabidopsis, we observed a FACC-like, highly conductive envelope channel, which was absent in pec1pec2 mutants that exhibited strongly diminished envelope currents and ~80% reduced K⁺ conductance (Fig. 1 E and G). Residual currents indicate the presence of other K⁺ or Cl⁻ permeable channels with lower conductivity than PEC1/2 [e.g., MSL2/3 (55, 56)].

Channel gating for PECs revealed weaker voltage dependence than for pea FACC, which may relate to physiological differences, the presence of distinct isoforms, or species-specific regulation. Importantly, PEC-mediated currents share key characteristics with pea FACC, including permeability for K⁺ (33), high total conductance, rapid activation, and Gd³⁺ sensitivity. These properties position PECs as best candidates for the FACC to date.

FACC has confirmed Ca²⁺ permeability and exhibits high open probability at physiological cytosolic Ca²⁺ concentrations. Concomitantly, loss of PECs diminishes stress-induced stromal Ca²⁺ transients (33) while simultaneously trapping Ca²⁺, normally directed toward plastids, in the cytosol (Fig. 1 J and K). Therefore, PECs provide a mechanism to manipulate stress-induced cellular and especially chloroplast Ca²⁺ signaling and enable investigation of the physiological function of stromal Ca²⁺ transients.

Promoter analysis revealed that PECs represent direct targets of MYC2/3/4 TFs (Fig. 2 A and B), master coordinators of JA signaling (39). This aligns with recent research adding PEC1 as a MYC2 core target (57). Time-resolved transcriptomics of MeJA treatments show that PEC1 is JA-inducible and follows a sustained induction profile, peaking at around 1 to 3 h posttreatment but subsiding slowly thereafter over 16 h (SI Appendix, Fig. S4). This sustained expression pattern contrasts with early canonical JA biosynthesis genes such as JAR1 and OPR3, which peak rapidly and return to baseline within 4 h, and instead parallels that of LOX2 and MYC2.

Like LOX2, which is essential for sustaining JA biosynthesis beyond the initial burst 90 min postwounding (45), MYC2 mediates not only immediate wounding responses but also contributes to long-term immune memory (58). PEC1 protein levels increase robustly following MeJA treatment, wounding, and coronatine exposure (Fig. 2 C and D), and, like LOX2, remain elevated for more than 96 h after wounding (SI Appendix, Fig. S7D). This indicates that PECs may function as part of a sustained plastid amplification circuit. Our data show that JA-dependent control of PEC expression actively augments stromal Ca²⁺ transients: Stress exposure leads to higher PEC protein levels and thus amplified stromal Ca²⁺ flux (Fig. 3 B and C). This mechanism may represent a physiological strategy linking rapid Ca²⁺ transients, which rise seconds after stress perception, to long-distance sustained hormone signaling (59).

Plant lipoxygenases e.g., LOX2, which catalyzes the first committed JA biosynthesis step in the stroma, are Ca²⁺-dependent, a requirement also conserved in animal LOXs, to associate LOX with membrane lipids (17, 18, 60, 61). Ca²⁺-dependent regulation has also been reported for some plant lipases (20). Coexpression and highly similar temporal behavior of PEC1 and LOX2 suggest a functional interplay in which JA-dependent upregulation of PECs ensures that stromal Ca²⁺ supply meets the demand for plastid oxylipin biosynthesis during priming and sustained defense. This may facilitate the activation of Ca²⁺-responsive enzymes, e.g., LOX2, and potentially influence candidate lipases for the release of α-linolenic acid during the initiation of JA signaling, such as DAD1-LIKE LIPASES (DALLs) (19) or other as-yet-unidentified lipases. However, during early defense, PEC-dependent modulation of stromal Ca²⁺ transients appears supportive rather than essential to elicit the full stress response (Fig. 4 B–E).

This hypothesis is supported by a significant reduction of plant hormone levels only in primed pec1pec2 mutant plants (~30% OPDA, ~66% JA) compared to WT (Fig. 4 G and H), coinciding with increased susceptibility to B. cinerea (Fig. 5 K–M). This suggests that PECs support plastid-derived oxylipin precursor biosynthesis during prolonged or repeated stress, ensuring a stronger and faster hormonal response. Modulating PEC protein levels shapes stromal Ca²⁺ signaling, and, by extension, JA-dependent defense. Consequently, PEC1 overexpression exacerbates stromal Ca²⁺ transients, both JA-mediated priming and broader wounding-induced defense responses on a transcriptional level, and thus ultimately enhances resistance to B. cinerea without disturbing chloroplast ion composition or photosynthesis (Fig. 5 and SI Appendix, Figs. S8–S12). Altogether, this suggests that PECs do not mediate general plastid cation loading but are tightly regulated and fulfill highly specialized roles in modulating stress signaling. While PEC1 overexpressors exhibit modest growth reductions (SI Appendix, Fig. S8C), hallmarks of enhanced JA signaling (51), these effects were mild compared to known JA-overproducing mutants such as fou2 and cev1 (48, 62). Thus, tuning PEC1 levels may offer a well-dosed avenue to modulate defense while minimizing growth tradeoffs.

In summary, we propose a model in which PECs and the stromal Ca²⁺ transients they mediate act as modulators within plant defense (Fig. 6). Our data show that PECs are not essential for early Ca²⁺-dependent defense responses, likely due to Ca²⁺ saturation at this stage and/or compensatory mechanisms, e.g., unknown ion channels that maintain sufficient relay of cytosolic and stromal Ca²⁺ signals in pec1pec2. However, prolonged stress and concomitant JA-dependent upregulation of PEC expression set the plant cells into a primed state. In this state, increased PEC-dependent Ca²⁺ flux promotes strong and fast activation of lipases or coexpressed lipoxygenases e.g., LOX2, ensuring an enhanced JA response upon recurring stress (Fig. 6). Importantly, PECs do not only affect stromal Ca²⁺ transients: Through altering compartmentalization of stress-induced Ca²⁺ signals between cytosol and plastids, changes in PEC protein levels may further affect downstream Ca²⁺ decoding in the cytosol. This likely contributes to the phenotypes observed here, e.g., reduced MeJA sensitivity in pec1pec2 mutants. Altogether, this study establishes PECs as modulators that link plastid Ca²⁺ dynamics and cellular compartmentalization to JA-dependent stress defense responses and provides a genetic tool to dissect how compartment-specific Ca²⁺ signatures shape hormonal signaling pathways in plants.

Fig. 6.

PECs are modulators of the plant defense response. Working model: Under basal conditions, PEC1 levels are low, directly or indirectly limiting plastid Ca²⁺ influx. Upon priming or stress, PEC1 expression is induced controlled by JA signaling. ≥24 h after priming, elevated PEC1 protein levels enhance stromal Ca²⁺ transients, which allows for the accumulation of higher levels of cis-OPDA, which translates into higher JA and JA-Ile levels upon trigger. This ultimately likely enables a faster and stronger defense response. Note, the plastid outer envelope was omitted for simplicity.

Materials and Methods

All mutants isolated or designed in this study were donated to the stock centers: PEC1ox-3 (N2112900), PEC1ox-6 (N2112901), PEC1ox-11 (N2112902), PEC1ox-6 sAEQ (N2112905), PEC1ox-11 sAEQ (N2112906), pec1pec2-1 pMAQ2 (N2112907), pec1pec2-2 pMAQ2 (N2112908), fou2pec1pec2-1 (N2112898), fou2pec1pec2-2 (N2112899), fou2 sAEQ (N2112903), and fou2pec1pec2-1 sAEQ (N2112904). Further detailed description of the materials and methods used in this study are provided in the SI Appendix. In brief, Arabidopsis plant material and growth conditions, the generation of transgenic lines, chloroplast isolation, electrophysiological measurements on isolated chloroplasts, measurement of Ca²⁺ transients, various stress treatments, protein extraction, and immunoblotting are described in detail. The SI Appendix further includes information on the reanalysis of publicly available DAP-Seq, ChIP-Seq, RNA-seq, and spatiotemporal single-cell RNA-Seq data, quantification of luciferase luminescence, RNA Sequencing, and data analysis, as well as fresh weight determination, quantification of photosynthetic parameters, and leaf chlorophyll determination. Detailed procedures for whole rosette and chloroplast ionomics, the B. cinerea infection assay and Trypan Blue staining, root length analysis, and phytohormone quantification are also included. The applications used for data processing and analysis are also provided in the SI Appendix. Information on the respective tests applied, P values, and numbers of replicates for each experiment are indicated in the figure captions.

Supplementary Material

Appendix 01 (PDF)

Dataset S01 (XLSX)

Dataset S02 (XLSX)

Dataset S03 (XLSX)

Dataset S04 (XLSX)

Dataset S05 (XLSX)

Dataset S06 (XLSX)

Dataset S07 (XLSX)

Dataset S08 (XLSX)

Dataset S09 (XLSX)

Dataset S10 (XLSX)

Data, Materials, and Software Availability

Study data are included in the article and/or supporting information. Raw sequencing data have been deposited with NCBI SRA under the bioproject accessions PRJNA1268713 (63) for Dataset S1 and PRJNA1464431 (64) for Dataset S2.