INTRODUCTION

Plants integrate dynamic environmental cues with growth and stress responses through precisely coordinated intraorganellar communication. Key metabolites not only drive essential biochemical pathways but also function as signaling molecules. A prime example is 2-C-methyl-D-erythritol-2,4-cyclopyrophosphate (MEcPP), a central intermediate in the plastidial methylerythritol phosphate (MEP) pathway, a pathway conserved in all plastid-bearing organisms, most bacteria, and apicomplexans. Beyond serving as an isoprenoid precursor, MEcPP acts as a stress-specific retrograde signal in plants (Xiao et al., 2012).

Elevated MEcPP levels are induced by a variety of abiotic stresses, such as radiation, supraoptimal temperatures, wounding, and high light, as well as biotic challenges including aphid infestation and CMV-m2b infection (Onkokesung et al., 2019;Xiao et al., 2012;Zeng, Wang, et al., 2022). The reason for this accumulation is because hydroxymethyl butenyl diphosphate synthase (HDS), the enzyme responsible for conversion of MEcPP is an [4Fe-4S]-protein susceptible to reactive oxygen species, that is inhibited by reactive oxygen species leading to accumulation of MEcPP, that in turn protects MEP-pathway activity by restricting oxidative stress (Ostrovsky et al., 1998;Rivasseau et al., 2009;Wang et al., 2020;Zeng & Dehesh, 2021). Moreover, in both plants and microbes, excess MEcPP is exported from the chloroplast or cell, although the precise mechanisms governing its export remain largely elusive (Volke et al., 2019;Xiao et al., 2012;Zhou et al., 2012). Notably, elevated MEcPP levels initiate a cascade of events that ultimately activate the calcium-dependent calmodulin-binding transcription activator 3 (CAMTA3), linking metabolic status directly to gene regulation during stress responses (Benn et al., 2016). Subsequently, CAMTA3 activates the rapid stress response element (RSRE; CGCGTT), a regulatory motif prevalent in about 30% of stress-responsive genes and notably enriched in MEcPP-induced genes (Benn et al., 2016). The RSRE was originally identified as a core cis-regulatory motif prevalent in the promoters of genes rapidly induced by a wide range of environmental stresses, including wounding, osmotic stress, UV-B, and pathogen-associated molecular patterns, making it a defining feature of the plant general stress response (GSR) (Benn et al., 2014). The GSR is a highly conserved adaptive program observed across kingdoms-from microbes to plants and animalsthat enables organisms to mount broad, transient transcriptional reprogramming in response to diverse stressors (K€ ultz, 2005;L opez-Maury et al., 2008). In plants, RSRE-containing genes account for a substantial proportion of early stress-responsive genes and are thought to form the transcriptional foundation of GSR (Benn et al., 2014). Through the activation of RSRE-driven transcription, CAMTA3 emerges not only as a key effector of plastidial MEcPP signaling but also as a central regulator of GSR gene expression, bridging organellar metabolic cues and transcriptional stress programs. This led to the notion of MEcPP function as a calcium rheostat, ultimately activating stress-responsive genes and linking plastidial stress perception to nuclear gene expression (Benn et al., 2016). It is noteworthy that CAMTA3, despite its reported role as an activator in cold-response, GSR, and glucosinolate metabolism regulation (Benn et al., 2016;Doherty et al., 2009;Laluk et al., 2012), has also been identified as a transcriptional repressor for immunity-related genes (Jiang et al., 2020). This delicate equilibrium in transcriptional regulation, toggling between gene repression and activation, mirrors observations in the homeodomainleucine zipper (HD-Zip) protein family, encompassing HD-Zip I-IV groups pivotal for plant responses to environmental cues (Ariel et al., 2007;Harris et al., 2011). Notably, HAT1, also known as JAIBA, a member of the class II HD-Zip proteins, showcases a multifaceted nature by acting as an activator of transcriptional pathways in plant development (Z uñiga-Mayo et al., 2012). HAT1 has been shown to modulate auxin signaling and regulate hypocotyl elongation in a light-dependent manner (Zhao et al., 2025). Beyond its role in developmental, HAT1 also negatively regulates abscisic acid (ABA) biosynthesis and signaling under drought conditions serving as a direct substrate of the SnRK2.3 kinase (Tan et al., 2018). Loss-of-function mutants of HAT1 (hat1, hat1hat3) exhibit enhanced drought tolerance, whereas overexpression HAT1 displays hypersensitivity to drought, as evidenced by accelerated water loss and reduced biomass under water-limiting conditions. These results underscore HAT1's function as a negative regulator in drought stress responses and highlight its broader role in integrating developmental and stress signaling pathways. Simultaneously, it functions as a repressor, inhibiting anthocyanin, auxin, brassinosteroid, ABA, and gibberellin pathways (Tan et al., 2018(Tan et al., , 2021;;Turchi et al., 2015;Zhang et al., 2014;Zheng et al., 2019), partly through the recruitment of transcriptional co-repressors such as the TOPLESS (TPL) protein, facilitated by its N-terminal ERF-associated amphiphilic repression (EAR) motif (Zheng et al., 2019).

In Arabidopsis, the TPL/TPR family is categorized into two distinct groups based on sequence similarities, reflecting their evolutionary divergence and specific roles. These proteins are instrumental in governing developmental pathways, mediating hormone signaling, and orchestrating responses to environmental stresses (Causier et al., 2012;Plant et al., 2021). By engaging with diverse transcription factors at gene promoter sites, the TPL/TPR proteins intricately regulate plant growth and enhance adaptability to environmental cues (Ke et al., 2015;Martin-Arevalillo et al., 2017).

Eukaryotic cells rely on the nuclear membrane system to meticulously manage transcription and translation processes, ensuring precise regulation. Central to this system is the nuclear import machinery, which facilitates the transportation of proteins, including transcriptional regulators, into the nucleus. These regulators play pivotal roles in governing various aspects of plant growth, development, and responses to environmental stimuli (Allen & Strader, 2021;Castel & Chae, 2021;Gu, 2018). Within this intricate network, nuclear importins, comprising importin a and b, assume key functions in orchestrating responses to diverse stress conditions (Wirthmueller et al., 2013). Recently, we identified IMPa-9 as a critical component of the MEcPP-mediated plastidial retrograde signaling pathway (Zeng et al., 2024). Stress-induced accumulation of MEcPP triggers the proteasome system to degrade IMPa-9 and its associated proteins, thereby alleviating their suppression of stress response genes. This suppression mechanism is part of a broader regulatory network, but direct evidence elucidating how IMPa-9 and its interactors suppress stress response genes remains elusive.

In this study, we demonstrate that HAT1 directly binds to the RSRE motif and interacts with the activator CAMTA3. We also confirm HAT1's role as a suppressor of RSRE, functioning alongside the co-repressors TPL and IMPa-9 to maintain transcriptional repression under normal conditions. Under stress, MEcPP disrupts this repression by triggering the degradation of TPL and IMPa-9 and reducing HAT1 levels, thereby relieving repression on RSRE. Simultaneously, elevated intracellular calcium levels activate CAMTA3, enabling it to bind RSRE and drive the expression of stress response genes. Our findings highlight the critical role of plastidial retrograde signaling in dynamically shifting from gene repression to activation, ensuring precise regulation of gene expression under stress conditions.

RESULTS

HAT1 binds to and suppresses RSRE induction

Our previous research identified RSRE in 30% of stress-induced genes, many of which are activated by elevated MEcPP levels during stress (Benn et al., 2016). To gain deeper insight into the RSRE-regulated transcriptional network, we conducted a yeast one-hybrid (Y1H) assay (Zeng, Chen, et al., 2022) and identified HAT1 as a primary candidate transcriptional regulator binding to RSRE. This finding was supported by both lacZ and luciferase reporter systems (Figure 1A,B).

To further investigate HAT1's biological role in in-planta transcriptional regulation of RSRE, we manipulated HAT1 expression levels genetically, generating both overexpression and knockdown lines. In the overexpression lines, we introduced 35S:HAT1 into the background of the 4xRSRE:Luciferase (LUC ) reporter plants (Walley et al., 2007) via Agrobacterium-mediated transformation (referred to as 35S:HAT1). For functional knockdown, we utilized the CRISPR-Cas9 system to generate HAT1 knockout lines, designated as CRISPR-hat1-1 and CRISPR-hat1-2. In both lines, targeted deletions disrupt the helix-turn-helix (HTH) DNA-binding motif, effectively abolishing HAT1's transcriptional regulatory function (Figure S1). Our analysis of RSRE-driven LUC activity in the control 4xRSRE::LUC/-Col-0 (referred to as WT), 35S:HAT1, and CRISPR-hat1 lines revealed diminished bioluminescence signals in the 35S: HAT1 line compared to the two CRISPR-hat1 lines, and conversely, the CRISPR-hat1 lines exhibited heightened bioluminescence signals under both standard and wounded conditions compared to the control (Figure 1C,D). These findings solidify HAT1's role as an in-planta suppressor of RSRE activation. Utilizing AlphaFold 3 (Abramson et al., 2024) we predicted the 3D molecular architecture of the HAT1-RSRE complex with high confidence (Figure 1E, F). The structure of HAT1 reveals distinct domains, including an N-terminal homeodomain and a HD-Zip region. Notably, the HD-Zip region is characterized by a HTH motif, enriched with basic residues. These electrostatic properties, conserved across land plants (Figure S2), contribute to the stable formation of the HTH-RSRE complex (Figure 1E). These findings prompted us to investigate HAT1's ability to suppress the previously reported MEcPP-inducible RSRE activity (Benn et al., 2016). Thus, we introduced the 35S:HAT1 vector into the ceh1 mutants harboring the 4xRSRE::Luciferase reporter (hereafter referred to as ceh1), which constitutively accumulated high levels of MEcPP (Zeng, Chen, et al., 2022;Zeng, Wang, et al., 2022), thereby generating the ceh1/35S:HAT1 transgenic lines. Subsequent luciferase activity assays revealed a substantial reduction in the constitutive luciferase activity in the ceh1/35S:HAT1 line compared to the ceh1 background (Figure 1G,H). These results collectively demonstrated that HAT1 acts as a suppressor of MEcPP-induced RSRE activation.

In summary, our findings reveal HAT1's ability to bind to the RSRE cis-element and attenuate MEcPP-induced RSRE activation.

MEcPP regulation of HAT1 expression is auxin-dependent

Given the MEcPP-induced RSRE response, we explored a potential feedback loop in which elevated MEcPP levels interfere with HAT1 function through various mechanisms, including transcriptional regulation. This proposition is supported by MEcPP's ability to reshape the transcriptional landscape. To validate this hypothesis, we initially examined HAT1 expression levels in the ceh1 mutant compared to WT plants (Figure 2A). The reduced expression of HAT1 in ceh1 compared to WT plants prompted us to investigate its expression levels in response to wounding, a stress known to induce MEcPP accumulation (Xiao et al., 2012), a phenomenon reaffirmed in this study (Figure 2B). The results demonstrate a decrease in HAT1 transcript levels following wounding (Figure 2C). These collective findings lend support to the concept of MEcPP-mediated transcriptional suppression of HAT1.

In exploring the regulatory elements governing HAT1 transcription, we analyzed its promoter sequence and identified three auxin response cis-elements (AuxRE) (Figure 2D). This discovery prompted us to assess HAT1 transcript levels in both mock-treated and indole-3-acetic acid (IAA)-treated WT plants. We observed a significant increase in HAT1 transcript levels 30 min post IAA treatment, compared to mock-treated plants, underscoring the auxin-dependent regulation of HAT1 expression (Figure 2E). Moreover, considering MEcPP's established role in diminishing auxin levels (Jiang et al., 2018) and influencing the transcript levels of ARF7 and ARF19, key members of the transcription factor family responsible for activating AuxREs (Hagen & Guilfoyle, 2002), we proceeded to assess HAT1 transcript levels in arf7/arf19 double mutants under both mock-and IAA-treated conditions, contrasting them with those of WT plants. The marked reduction in HAT1 transcript levels observed in the arf7/arf19 mutants relative to WT in both treatment scenarios (Figure 2E) underscores the participation of ARF7 and ARF19 in the transcriptional regulation of HAT1 in response to MEcPP.

These findings reveal the involvement of MEcPP and auxin in regulating HAT1 expression levels, compelling us to examine RSRE activities in WT and ceh1 plants following IAA treatment. As expected, we observed a significant reduction in RSRE signals in IAA-treated plants compared to representative mock-treated plants (Figure 2F,G).

In summary, these data elucidate the molecular mechanisms underlying the MEcPP-mediated decrease in HAT1 expression levels and the resulting increase in RSRE activity.

MEcPP-mediated reduction of HAT1-interacting protein, TPL

Building on the known interaction between HAT1 and TPL, a transcriptional co-suppressor (Zheng et al., 2019), and Ó 2025 Society for Experimental Biology and John Wiley & Sons Ltd., The Plant Journal, (2025), 123, e70478

MEcPP's negative regulation of HAT1 (Figure 2), we hypothesized that MEcPP might similarly impact TPL. To explore this, we initially reaffirmed the TPL/HAT1 interaction through structural modeling analyses (Figure 3A). We found that TPL contains an N-terminal TPR1-like fold, followed by two WD40/YVTN repeats that form a seven-blade propeller structure (Figure S3A), domains known to mediate protein-protein interactions (Collins et al., 2019). Further modeling revealed that the stability of the HAT1-TPL complex is primarily maintained by salt bridges formed between polar residues, a feature highly conserved across land plants (Figure 3A; Figure S3b,c).

Next, we investigated whether TPL, as a known transcriptional co-repressor, plays a role in suppressing RSRE activity. To test this, we introduced the 35S:TPL-Myc overexpression construct into ceh1 (ceh1/4xRSRE::Luciferase) plants via Agrobacterium-mediated transformation to generate ceh1/4xRSRE::Luciferase/35S:TPL-Myc (hereafter refer as ceh1/35S:TPL-Myc) transgenic lines. Homozygous lines were selected and used for all subsequent assays. In the high MEcPP-accumulating ceh1 mutant, RSRE:Luciferase activity is constitutively activated under non-stressed conditions compared to that in WT plants. However, RSRE: Luciferase activity is significantly decreased in ceh1/TPLoverexpressing plants compared to that in ceh1 plants under normal conditions (Figure 3B,C). These results provide strong genetic evidence that TPL functions as a negative regulator of RSRE motif-containing genes.

Our genetic data raised the question of why native TPL levels fails to reduce RSRE activity in the high MEcPPcontaining ceh1 mutant background. To explore this, we first examined TPL transcript levels in 2-week-old WT and ceh1 plants and found no significant differences between these two lines (Figure S4), suggesting that MEcPP does not regulate TPL transcriptionally. Next, we investigated whether MEcPP regulates TPL's protein level. However, in the absence of a TPL specific antibody, we introduced the 35S:TPL-cLuc vector into previously generated DEX-inducible HDS RNAi (HDSi) plants. These plants accumulate MEcPP at levels similar to those observed in ceh1 mutant after 72 h of DEX induction (Jiang et al., 2018). We subjected 2-week-old 35S:TPL-cLuc/HDSi seedlings to DEX or control mock, followed by immunoblot analyses conducted 72 h later to evaluate total TPL protein levels. The data showed a significant reduction in TPL total protein levels as evident in the DEX-induced 35S: TPL-cLuc/HDSi seedlings compared to the mock-treated counterparts (Figure 3D). This result clearly establishes a link between MEcPP accumulation and the decreased TPL protein levels in DEX-induced HDSi seedlings.

To further investigate how MEcPP influences TPL protein abundance, we treated 2-week-old 35S:TPL-cLuc/HDSi seedlings with bortezomib, a 26S proteasome inhibitor, or a mock solution for 18 h following 72 h of DEX induction. Immunoblot analyses showed a significant increase in TPL protein levels in bortezomib-treated seedlings compared to mock-treated controls (Figure 3E; Figure S5). This bortezomib-induced stabilization of TPL protein strongly suggests that proteasomal degradation is a key mechanism by which MEcPP regulates TPL protein abundance. TPL interacts with nuclear importin IMPa-9

In our prior investigation using genetic mutagenesis analysis, we identified nuclear importin IMPa-9 as a key suppressor of the MEcPP-mediated retrograde signal transduction pathway. Additionally, we found that MEcPP negatively affects IMPa-9 protein levels (Zeng et al., 2024). Through immunoprecipitation-mass spectrometry and yeast two-hybrid library screening assays, we discovered various members of the TPL family of transcriptional co-repressors in the interactome of IMPa-9, including TPL (Zeng et al., 2024). We subsequently utilized additional methods to confirm the protein-protein interaction between TPL and IMPa-9. First, we generated pIMP-a9: IMPa-9-GFP/35S:TPL-cLuc (referred to as IMPa-9-GFP/35S: TPL-cLuc) and control pIMP-a9:IMPa-9-GFP/35S:cLuc (referred to as IMPa-9-GFP/35S:cLuc) transgenic plants to verify the interaction through targeted coimmunoprecipitation (Co-IP) experiments. We used GFP-Trap and empty magnetic beads for IP of IMPa-9-GFP in both lines, followed by immunoblot analyses using GFP and luciferase antibodies. The clear presence of the TPL-cLuc band in the IP fraction of the IMPa-9-GFP/35S: TPL-cLuc line, but not in the control empty beads or the IMPa-9-GFP/35S:cLuc line, verified the in vivo interaction between TPL and IMPa-9 proteins (Figure 3F). We further confirmed the interaction by luciferase reconstitution assay. We constructed vectors that fused the IMPa-9 coding sequence (CDS) with the amino-terminal fragment of the luciferase gene under the control of the 35S promoter (35S:nLuc-IMPa-9) and the TPL CDS with the carboxyl-terminal fragment of the luciferase gene under the 35S promoter (35S:TPL-cLuc). Infiltrating Nicotiana (F) The in vivo interaction of IMPa-9 with TPL determined by co-immunoprecipitation assay. Protein samples obtained from IMPa-9-GFP/35S:TPL-cLuc and IMPa-9-GFP/35S:cLuc were immunoprecipitated (IP) using GFP (+) and empty (À) magnetic beads followed by immunoblots performed using a-Luc. Each blot shows protein inputs before (input, right panels) and after (IP, left panels) immunoprecipitation. (G) Split luciferase complementation assays in Nicotiana benthamiana leaves expressing 35S:IMPa-9 (N-terminal Luc fragment fused with IMPa-9) and 35S:TPL-cLuc (TPL fused with C-terminal Luc fragment), demonstrating the interaction between IMPa-9 and TPL. Negative controls include 35S:IMPa-9 and 35S:cLuc.

benthamiana leaves with both 35S:nLuc-IMPa-9 and 35S: TPL-cLuc constructs resulted in the reconstitution of luciferase activity, while controls showed no activity (Figure 3G).

In summary, these results confirm the protein-protein interaction between TPL and IMPa-9. Given IMPa-9's role as a key suppressor in the MEcPP-mediated retrograde signaling pathway (Zeng et al., 2024), this suggests that HAT1, TPL, and IMPa-9 function as interconnected components of the RSRE suppressor network.

HAT1 interacts with CAMTA3, the calcium/calmodulinbinding activator driving MEcPP-mediated RSRE induction

It is widely acknowledged that CAMTA3, activated by stress-induced MEcPP, is the primary activator of RSRE (Benn et al., 2016;Prasad et al., 2023). The predicted structure of CAMTA3 reveals conserved functional domains such as the N-terminal DNA binding domain, tandem ankyrin repeats that are involved in protein-protein interaction, and IQ motifs which are implicated in the association of Ca 2+ loaded CaM and CaM-like proteins (CML) to CAMTA (Prasad et al., 2023) (Figure S6). Given MEcPP's established role in CAMTA3 activation, we explored whether the RSRE suppressor, HAT1, interacts with inactive CAMTA3 under normal conditions or if they function independently. To test these possibilities, we initially utilized a luciferase reconstitution assay and constructed vectors expressing fusion proteins: the HAT1 CDS with the amino-terminal fragment of the luciferase gene (-nLuc-HAT1) under the 35S promoter, and the CAMTA3 with the carboxyl-terminal fragment of luciferase (cLuc-CAMTA3) under the 35S promoter. Co-infiltration of N. benthamiana leaves with these constructs showed the reconstitution of luciferase activity exclusively in leaves infiltrated with both nLuc-HAT1 and cLuc-CAMTA3, supporting their interaction (Figure 4A).

To further confirm this interaction, we generated transgenic Arabidopsis plants expressing HAT1-GFP under its native promoter (pHAT1:HAT1-GFP) and conducted Co-IP experiments. Utilizing GFP-Trap magnetic beads, we isolated immunoprecipitation fractions from the transgenic and wild-type control lines, followed by immunoblot analyses with antibodies against GFP and CAMTA3. The detection of CAMTA3 in the immunoprecipitate from pHAT1: HAT1-GFP plants, but not from wild-type plants, provided additional evidence for the in vivo interaction between HAT1 and CAMTA3 (Figure 4B; Figure S7). Lastly, our structural modeling analyses revealed the interaction between HAT1 and CAMTA3, mediated by CAMTA3's protein-protein interaction domains formed by tandem ankyrin repeats (Figure 4C). Notably, these interaction sites are highly conserved across land plants (Figure S8a,b).

In conclusion, the Co-IP experiments, combined with structural modeling, demonstrate the interaction between HAT1 and CAMTA3 under normal condition, mediated by conserved protein-protein interaction domains in CAMTA3.

MEcPP potentiates extracellular Ca 2+ influx

We have established the pivotal role of MEcPP in activating RSRE, which was subsequently nullified by the presence of the Ca 2+ chelator EGTA (Benn et al., 2016). This highlights the indispensable function of Ca 2+ for CAMTA3 activity, leading to the notion of MEcPP as a regulator for Ca 2+ release, thereby facilitating CAMTA3 activation (Benn et al., 2014(Benn et al., , 2016)). To investigate this notion, we utilized our previously reported multi-compartmental Ca 2+ sensor construct (CamelliA_NucG/PmG/CytR) to simultaneously image Ca 2+ dynamics in the nucleus, sub-plasma membrane, and cytosol (Guo et al., 2022). We expressed this triple Ca 2+ sensor in both WT plants and the high MEcPP-accumulating ceh1 mutant to monitor Ca 2+ dynamics. The results revealed significantly higher Ca 2+ levels in the nuclei of epidermal cells in ceh1 compared to WT, as reflected by increased CamelliA_NucG fluorescence intensity (Figure 5A,B; Figure S9; Movies S1 and S2). The functionality of the CamelliA_NucG sensor in WT plants was validated by its ability to report Ca 2+ influxes triggered by a high dosage of 405 nm laser wounding as previously described (Guo et al., 2022) (Figure 5B; Figure S9). These findings indicate elevated nuclear Ca 2+ levels in ceh1, potentially driven by increased MEcPP accumulation, suggesting a role for MEcPP in modulating nuclear Ca 2+ signaling. To investigate the specificity of MEcPP in potentiating these Ca 2+ signatures, we examined the Ca 2+ signature in WT plants treated with either a mock or exogenous MEcPP. The results demonstrated that exogenously applied MEcPP, but not the mock treatment, induced both nuclear and cytosolic Ca 2+ elevation in epidermal cells of true leaves (Figure 5C,D; Movies S3 and S4).

Furthermore, we employed the commonly used calcium chelator, BAPTA (1,2-bis(o-aminophenoxy) ethane-N, N,N 0 ,N 0 -tetraacetic acid), which sequesters Ca 2+ by forming stable complexes through its multiple oxygen atoms, effectively reducing the concentration of free Ca 2+ in the cellular environment. Our data unequivocally demonstrate the abolition of Ca 2+ influx in the presence of BAPTA (Figure 5E; Movie S5).

Given the constitutive expression of RSRE in the ceh1 mutant background and the significant increase in Calmodulin-like protein 24 (CML24) expression levels in ceh1 (Benn et al., 2016;Xiao et al., 2012), along with the role of CML24 as a potential Ca 2+ sensor and the function of Ca 2+ -Calmodulin in activating CAMTA3, we were prompted to further analyze the structure of Ca 2+ -loaded CML24 in complex with CAMTA3 (Figure 5F; Figure S10a). These analyses revealed the four conserved EF-hand motifs of CML24 responsible for Ca 2+ binding (Nie et al., 2012), primarily stabilized by aspartic/glutamic acid and asparagine residues (Figure S10b). Furthermore, the predicted structure of CML24 features a dumbbell-shaped architecture with N-and C-terminal domains connected by a flexible linker that may wrap around CAMTA3, positioning CAMTA3's IQ-motif peptide at the interface of these domains. This potential interaction is suggested to be stabilized by polar salt bridge interactions (Figure S10c). However, we acknowledge that further experimental validation is required to confirm the predicted protein-protein interaction between CML24 and CAMTA3. The potential binding of CML24 could facilitate CAMTA3 activation, potentially enhancing its ability to access the RSRE-binding site (Figure 5G), and thereby triggering the expression of stress-induced genes. Notably, evolutionary analyses show that the CAMTA3-RSRE interaction sites, some confirmed through point mutations (Prasad et al., 2023), are highly conserved across land plants (Figure 5G; Figure S11).

Collectively, these findings are consistent with our previous notion that MEcPP potentiates extracellular Ca 2+ influx, leading to elevated cytosolic and nuclear Ca 2+ levels. This, in turn, could activate Ca 2+ -CML-dependent activation of CAMTA3, thereby facilitating the transcription of stress response genes, although further experimental evidence is needed to identify the specific Calmodulin-like protein.

DISCUSSION

Understanding the regulatory components of the GSR, also known as the cellular stress response or core stress response, is essential for deciphering the fundamental principles governing early stress response mechanisms impacting cellular homeostasis, providing cross-protection, where exposure to one stressor can confer resistance to another, and enabling tailored adaptive responses to specific stressors, among other functions. Extensive research across bacteria, fungi, animals, and plants has confirmed the evolutionary conservation of GSR molecular components across all domains of life, highlighting their critical role in survival under stressful conditions (K€ ultz, 2005;L o pez-Maury et al., 2008;Walley et al., 2007).

The rapid activation of the GSR, often occurring within minutes of exposure to stimuli, demands efficient perception and integration of signal transduction, alteration of pre-existing transcription factors, and epigenetic modifications. While specific regulatory elements may vary among organisms, the underlying principles of stress detection, signal transduction, and adaptive responses remain evolutionarily conserved. Moreover, within the same organism, shared regulatory infrastructure allows for effective management and response to diverse environmental stressors. This conservation has spurred extensive inquiries into the mechanisms involved, providing deeper insights into how organisms maintain cellular homeostasis and adapt to environmental challenges.

In this study, we investigated the intricate mechanisms governing the transmission of stress-induced plastidial retrograde metabolite MEcPP signaling to the nucleus, with a particular focus on its crucial role in initiating the expression of GSR genes, notably through activating the GSR transcriptional hub RSRE. Our findings highlight MEcPP's dynamic roles in coordinating key regulators of adaptive responses through two distinct pathways. First, MEcPP disrupts the suppressor network comprising HAT1, TPL, and IMPa-9 through a dual mechanism: it reduces levels of the hormone IAA (Jiang et al., 2018;Zeng, Chen, et al., 2022;Zeng, Wang, et al., 2022), lowering IAA-dependent HAT1 expression, and it promotes the degradation of TPL and IMPa-9 via the 26S proteasome. Zhao et al. (2025) identified HAT1 as a key regulator in auxin-mediated light signaling. While their work emphasized light-auxin interactions in developmental contexts, our study revealed HAT1's role in repressing RSRE-mediated stress responses and showed that MEcPP negatively regulates HAT1 expression via suppression of auxin levels. Thus, we highlight a distinct plastid-tonucleus signaling mechanism that integrates stressderived MEcPP signals with auxin-responsive transcriptional regulation. Additionally, our recent discovery of MEcPP-induced ASK1 accumulation, a crucial component in the SKP1-Cullin1-F-box (SCF) E3 ubiquitin ligase complex, reveals another facet of MEcPP's regulatory role (Zeng et al., 2024). ASK1 facilitates IMPa-9 degradation, further weakening the suppression of RSRE-regulated stress genes. Together, these mechanisms dismantle the repression network, allowing RSRE to transition to an active state under stress conditions. Importantly, our protein-protein interaction assays and structural modeling analyses (Figures 3 and 4) support the formation of a repressive complex comprising HAT1, CAMTA3, TPL, and IMPa-9. Upon MEcPP accumulation, these interactions are disrupted, coinciding with reduced HAT1 expression and proteasomal degradation of TPL and IMPa-9. Structural modeling further supports the feasibility of these interactions, reinforcing the proposed multi-component regulatory framework depicted in Figure 6.

Remarkably, this mechanism shares similarities with established signaling pathways such as jasmonate (JA) and ABA pathways. In JA signaling, JAZ (Jasmonate ZIM-Domain) proteins act as repressors inhibiting transcription factors like MYC2 responsible for defense gene expression. However, during stress, increased JA levels trigger JAZ degradation through the Skp/Cullin/F-box complex (SCF COI1 )-dependent 26S proteasome pathway, freeing MYC2 to activate defense-related genes (Chini et al., 2007;Thines et al., 2007). Similarly, in the ABA signaling pathway, protein phosphatase 2C (PP2C) proteins dephosphorylate and repress SnRK2 kinases, but under stress ABA inhibits PP2C, allowing SnRK2 kinases to activate downstream targets and initiate ABA-responsive gene expression (Park et al., 2009).

These signaling pathways collectively illustrate a coordinated regulation amounting to relieving repression and activating key components to mount an effective response to stress.

Besides its role in releasing transcription factors and alleviating repression, MEcPP also enhances the influx of Ca 2+ , a crucial step for activating CAMTA3. This adds complexity to the regulatory network of retrograde signaling in adaptive responses. This discovery raises an intriguing question about the nature of calcium channels and the mechanism by which MEcPP activates them. Unraveling these mechanisms provides insights into how plants integrate various signals to orchestrate effective responses to environmental stresses.

In summary, this report outlines the multicomponent regulatory framework that balances growth and stress responses, as depicted in the simplified schematics (Figure 6). Under non-stress conditions, HAT1 binds to the RSRE motif, present in 30% of stressresponsive genes, suppressing their expression. However, stress-induced accumulation of MEcPP initiates a coordinated dual-channel mechanism that dismantles the repressor network and activates CAMTA3 by enhancing nuclear calcium influx. This interplay between retrograde signaling, nuclear transport, transcriptional regulation, and ion signaling highlights the pivotal and multifaceted role of MEcPP. Our findings offer significant insights into retrograde metabolite signaling and its role in plastid-tonucleus communication, reinforcing the concept of plastids as central regulatory hubs orchestrating plant stress responses.

EXPERIMENTAL PROCEDURES

Plant material and growth condition

All Arabidopsis thaliana mutants and transgenic lines are in the Columbia (Col-0) ecotype. Plants were grown in 16-h light/8-h dark cycles at ~22°C in the growth chamber with light intensity ~100-150 lmol m À2 sec À1 . Two-week-old seedlings grown on ½ MS plates (pH 5.7) were used for all the experiments. The ceh1, HDSi lines were developed previously (Jiang et al., 2018;Xiao et al., 2012). OX-ARF seeds were obtained from Professor Lucia C. Strader at Washington University. Transgenic pIMPa-9:IMPa-9-GFP/35S:TPL-cLuc, pIMPa-9:IMPa-9-GFP/35S:cLuc, 35S:HAT1, CRISPR-hat1-1, CRISPR-hat1-2, 35S:TPL-cLuc/HDSi, and pHAT1: HAT1-GFP lines were generated in this study. Under normal conditions, the plant size of the aforementioned transgenic lines is comparable to that of wild-type plants. The egg cell-specific promoter-controlled CRISPR/Cas9 system with the Golden Gate cloning method was used to obtain CRISPR-hat1 mutants as described (Wang et al., 2015). The sequences of the two CRISPR guide RNAs are provided in Table S1. The 35S:HAT1 and pHAT1: HAT1-GFP constructs were introduced into the RSRE::Luciferase background via Agrobacterium-mediated transformation to generate transgenic lines. Similarly, the 35S:TPL-cLuc construct was introduced into the HDSi background by Agrobacterium-mediated transformation. To generate the pIMPa-9:IMPa-9-GFP/35S:TPL-cLuc and pIMPa-9:IMPa-9-GFP/35S:cLuc double transgenic lines, WT plants were first transformed with the pIMPa-9:IMPa-9-GFP construct to obtain the pIMPa-9:IMPa-9-GFP line and screened for the homozygous lines. This pIMPa-9:IMPa-9-GFP homozygous line was then used as the recipient for a second round of transformation with either the 35S:TPL-cLuc or 35S:cLuc construct, respectively. For overexpression lines, those exhibiting a moderate increase in transcript levels (two-to threefold relative to control plants) were selected based on qPCR analysis, in order to minimize potential artifacts associated with excessive ectopic expression. Forceps were used to mechanically wound leaves. Two-week-old seedlings were treated with IAA (10 lM) for 30 min to extract RNA, or 1 h for luciferase signals measurements. Luciferin was dissolved in sterile water to a final concentration of 1 mM and uniformly sprayed on all plants used for RSRE::LUC imaging, across all genotypes and treatments. Auxin (IAA) was initially dissolved in 100% ethanol to make a stock solution, then diluted in water to the final working concentration for spraying. For mock treatments, control plants were sprayed with the corresponding solvent used for dissolving the IAA to ensure consistency. Luciferase imaging was initiated only after droplets had dried (%10 min). The RSRE::LUC signals were checked by using the CCD camera as previously described (Zeng, Chen, et al., 2022).

Yeast one-hybridization assay

The Gold Yeast One-Hybrid Library Screening System (#630491; Takara, Shiga, Japan) was used for screening RSRE motif binding Ó 2025 Society for Experimental Biology and John Wiley & Sons Ltd., The Plant Journal, (2025), 123, e70478

proteins. The sequence of RSRE motif (the 'DNA bait') was cloned to create the 4xRSRE::LacZ and 4xRSRE::Luciferase reporter constructs. X-gal was used to detect LacZ activity. The Andor Technology CCD camera was used to detect luciferase activity signals.

MEcPP extraction and analyses

MEcPP was extracted and analyzed as previously described (Zeng, Wang, et al., 2022).

Quantification of gene expression

Total RNA was extracted from 2-week-old plants using the Aurum Total RNA Mini Kit (Bio-Rad). Reverse transcription of 1 lg of total RNA was carried out using the iScript reverse transcriptasepolymerase chain reaction (RT-PCR) according to the manufacturer's instructions (Bio-Rad, Carlsbad, CA, USA). The resulting complementary DNA (cDNA) was used as the template for subsequent analysis. Real-time PCR was performed using the SsoAdvanced Universal SYBR Green Supermix on a CFX96 real-time PCR detection system (Bio-Rad). The reference gene AT4G26410 was used as an internal control.

Statistical analysis

Two-tailed Student's t-test and ANOVA test with Tukey's post hoc test were used for the statistical analyses. The asterisks and different letters on figures denote statistical significance (P ≤ 0.05).

Bortezomib treatment

Two-week-old 35S:TPL-cLuc/HDSi seedlings 72 h post DEX treatment were used for these analyses. Leaves were cut as small pieces and submerged in bortezomib (40 lM, dissolved in 0.1% DMSO) or in mock (0.1% DMSO) buffer for 18 h on a shaker at room temperature.

Agro-infiltration-based transient assays in N. benthamiana

The N. benthamiana transient assay was employed to investigate the RSRE activity and protein-protein interactions. Vectors pENTR/D-TOPO (Invitrogen) and Gateway systems were used for vector construction. The vectors carrying C/N-terminal luciferase fused with IMPa-9, TPL, HAT1, and CAMTA3, and luciferase fused with the RSRE motif were introduced into Agrobacterium GV3101, which were then used for leaf infiltration in N. benthamiana. After 48-72 h of infiltration, N. benthamiana leaves were sprayed with 1 mM luciferin solution. The Andor Technology CCD camera was used to detect luciferase activity signals. Images were acquired every 5 min for 3 h for N. benthamiana plants. The Andor Technology software was used to quantify the luciferase activity.

Protein extraction and immunoblot analyses

Two-week-old seedlings were collected and ground in liquid nitrogen and suspended in 29 SDS lysis buffer and boiled (100°C, 10 min) for total protein extraction. Proteins were then separated on a 10% SDS-PAGE gel and transferred onto the nitrocellulose membrane. The monoclonal a-GFP (1:1000, Roche), the monoclonal a-Luciferase (1:1000, Sigma), and a-CAMTA3 (1:500) antibodies (Sun et al., 2020) were used to detect the, respectively, tagged proteins. And the monoclonal a-Actin (1:10 000, Sigma) antibody was used as the loading control. The goat anti-mouse (1:3000) IgG (H+L)-HRP (1:3000) and goat antirabbit IgG (H+L)-HRP (1:3000) conjugated secondary antibodies (Bio-Rad) were used.

Co-immunoprecipitation (Co-IP)

Two-week-old plants were ground and suspended in extraction buffer (50 mM Tris-HCl at pH 7.5, 150 mM NaCl, 10% glycerol, 0.1% NP-40 and protease inhibitor cocktail) at 4°C for 30 min. The protein suspensions were then centrifuged at 12 000 g for 10 min at 4°C, and the precipitation were then filtered out using a 100 lm Nylon Mesh. The supernatant was incubated with GFP-Trap or empty magnetic (ChromoTek) beads for 2 h at 4°C. Beads were then collected by using a magnetic column and washed five times with 29 extraction buffer.

For Co-IP experiments, proteins were eluted from GFP-trapbeads using 29 SDS lysis buffer (50 mM Tris-HCl at pH 6.8, 2% SDS, 10% glycerol, 0.1% bromophenol blue, 1% 2mercaptoethanol) and boiled at 100°C for 10 min. Proteins were then separated on a 10% SDS-PAGE gel and subsequently transferred (20% methanol, 25 mM Tris, 192 mM glycine, pH 8.3) onto a nitrocellulose membrane for probing with the corresponding antibodies.

Ca 2+ dynamics imaging in epidermal cells

To image the Ca 2+ dynamics in ceh1 mutant plants, we first introduced the triple Ca 2+ sensor line CamelliA_NucG/PmG/CytR expressing nuclear-and PM-targeted Ca 2+ sensor GCaMP6f and cytosolic-targeted Ca 2+ sensor jRGECO1a (Guo et al., 2022) into the ceh1 mutant background by crossing. Seeds of the triple Ca 2+ sensor line in WT and ceh1 background were germinated on 0.5 9 Murashige and Skoog (MS) solid medium with 1% sucrose and buffered with 0.05% MES to pH 5.8. True leaves from 14-day post-stratification seedlings were mounted onto a glass-bottomed chamber slide using Hollister medical adhesive and incubated in leaf imaging buffer (5 mM KCl, 50 lM CaCl 2 , 10 mM MES pH 6.15) in an illuminated Percival growth chamber for 2 h. Subsequently, the leaf samples were imaged under a Zeiss confocal microscope LSM 900 equipped with a Zeiss W Plan-Apochromat 409/1.0 DIC water-dipping objective lens in a 'top-imaging' setup as described previously (Guo et al., 2022). Green-colored Ca 2+ sensors (Camel-liA_NucG/PmG) are excited with a 488 nm laser line, and the emission is collected between 493 and 590 nm; chlorophyll autofluorescence is collected between 656 and 700 nm. Image resolution is set to 512 9 512 pixel and scan speed ~1.4 sec per frame.

To monitor Ca 2+ dynamics in WT plant treated with exogenous applied MEcPP, true leaves of the 14-day post-stratification seedlings of the aforementioned triple Ca 2+ sensor plant were first attached to the glass bottom chamber slide using the medical adhesive and immediately covered with 100 ll of leaf imaging buffer (5 mM KCl, 50 lM MgCl 2 , 10 mM MES pH 6.15). The mounted leaves were rested for at least 2 h in an illuminated Percival growth chamber before imaging under an upright Zeiss confocal microscope LSM 880 equipped with a Zeiss W Plan-Apochromat 409/1.0 DIC water dipping objective lens in a 'topimaging' setup as described previously (Guo et al., 2022). Greencolored Ca 2+ sensors (CamelliA_NucG/PmG) are excited with a 488 nm laser line and the emission is collected between 493 and 550 nm; the red-colored Ca 2+ sensor (CamelliA_CytR) is excited with a 561 nm laser line and its emission is collected between 566 and 635 nm. Image resolution is set to 512 9 512 pixel and scan speed ~0.95 sec per frame. Leaf samples were first imaged for ~120 sec prior to MEcPP treatment, then were imaged for another ~600 sec after adding 100 ll of 200 lM MEcPP (final concentration of 100 lM) or mock (200 lM NH 4 Cl) diluted in leaf imaging buffer (indicated by the pink lines in Figure 5B). To chelate extracellular Ca 2+ , 10 mM BAPTA is included in the imaging buffer during the initial 2 h resting incubation before imaging.

Analysis of Ca 2+ imaging data

Each nucleus in the time-series images was segmented manually in the FluoroSNNAP code (Patel et al., 2015) in MATLAB (R2024a) and the mean fluorescence intensity of each nucleus is analyzed and quantified using the same tool. Baseline fluorescence intensity (F 0 ) is defined as the average intensity of each nucleus during the resting period before any treatments. Relative fluorescence intensity changes over time are then calculated using the equation ΔF/F 0 = (F À F 0 )/F 0 in GraphPad Prism and are plotted as a heatmap using PlotTwist (huygens.science.uva.nl/PlotTwist) (Goedhart, 2020). F 0 represents the baseline fluorescence intensity before stimulation, and DF indicates the change in fluorescence intensity relative to that baseline. The ratio DF/F 0 is a standard method used to quantify relative changes in fluorescence, providing a normalized measure of calcium dynamics. Each line of the heatmap represents the relative fluorescence intensity over time of a single nucleus.

Protein structural modeling and analysis

The predicted 3D models were generated using the AlphaFold 3 module with default setting, based on the Arabidopsis sequences for HAT1 (UniProt P46600), CAMTA3 (UniProt Q8GSA7), TPL (Uni-Prot Q94AI7), CML24 (UniProt P25070), and IMPa-9 (UniProt F4KF65). The illustration and analysis of the models were produced using PyMOL v3.

ACCESSION NUMBERS

The gene sequences mentioned in this article are available in the Arabidopsis Genome Initiative Information Resource (http://www.arabidopsis.org/). The locus numbers corresponding to the genes are as follows: IMPa-9 (AT5G03070), TPL (AT1G15750), HAT1 (AT4G17460), HDS (AT5G60600), and CAMTA3 (AT2G22300).