The Arabidopsis Mitochondrial Nucleoid–Associated Protein WHIRLY2 Is Required for a Proper Response to Salt Stress

Abstract In the last years, plant organelles have emerged as central coordinators of responses to internal and external stimuli, which can induce stress. Mitochondria play a fundamental role as stress sensors being part of a complex communication network between the organelles and the nucleus. Among the different environmental stresses, salt stress poses a significant challenge and requires efficient signaling and protective mechanisms. By using the why2 T-DNA insertion mutant and a novel knock-out mutant prepared by CRISPR/Cas9–mediated genome editing, this study revealed that WHIRLY2 is crucial for protecting mitochondrial DNA (mtDNA) integrity during salt stress. Loss-of-function mutants show an enhanced sensitivity to salt stress. The disruption of WHIRLY2 causes the impairment of mtDNA repair that results in the accumulation of aberrant recombination products, coinciding with severe alterations in nucleoid integrity and overall mitochondria morphology besides a compromised redox-dependent response and misregulation of antioxidant enzymes. The results of this study revealed that WHIRLY2-mediated structural features in mitochondria (nucleoid compactness and cristae) are important for an effective response to salt stress.


Introduction
Rapid climate change increasingly exposes plants to novel environmental conditions outside their physiological limits and beyond the range to which they are adapted.In this context, mitochondria, the primary powerhouse of the plant cell and essential hub for many metabolic processes, play an important role.Several metabolic pathways either originate from or converge on mitochondria.Maintaining organelle integrity in terms of functionality, morphology and dynamics is crucial for cellular homeostasis and proper responses to environmental challenges (Liberatore et al. 2016).Properly functioning mitochondria are required for efficient responses to any change in developmental or stress-response processes.
Mitochondria can form dynamic, interconnected networks regulated by an equilibrium of fusion and fission events that, in turn, determine the organelle number, size, shape, dynamics and, most importantly, functionality.The regulation of the mitochondrial network dynamics is critical for energy homeostasis, allowing the plant to respond rapidly and directly to acute metabolic perturbations by balancing energy demand and production (Yu and Pekkurnaz 2018).Alterations of the shape and dynamics of mitochondria are strongly linked to genome instability and can harm the organism (Hoppins 2014, Prevost et al. 2018).
The integrity of mitochondria highly depends on the integrity of their genome, which many factors can compromise.Thus, an active DNA recombination system is crucial, allowing the cell to correct detrimental mutations induced by environmental stress.In particular, high-frequency homologous recombination (HR) and base excision repair systems are required for efficient DNA repair (Marechal andBrisson 2010, Gualberto et al. 2014).Defects in any of these systems can lead to failures in maintaining the stability of the mitochondrial genome, resulting in the accumulation of mutations and in genomic rearrangements that, if not correctly repaired or removed, can become deleterious for the whole organism (Gualberto and Kühn 2014).
One of the most abundant proteins in the mtDNA repair system is WHIRLY2, a member of the WHIRLY DNA-binding protein family that is found in angiosperms (Krupinska et al. 2022).In Arabidopsis thaliana, the WHIRLY family is represented by three members: AtWHIRLY1, AtWHIRLY2 and AtWHIRLY3.
These proteins are localized in the DNA-containing organelles: WHIRLY1 and WHIRLY3 have been characterized as plastid DNA-binding proteins (Pfalz et al. 2006), while WHIRLY2 is present in mitochondria (Krause et al. 2005).WHIRLY2 is the most abundant single-stranded DNA (ssDNA) binding protein in Arabidopsis mitochondria (Fuchs et al. 2020).WHIRLIES could bind ssDNA stretches in a sequence-independent manner (Desveaux et al. 2002).This binding may occur in many parts of the mitochondrial genome, including replication forks (Wyman and Kanaar 2006), promoter regions and, of course, during mtDNA repair processes, where WHIRLY2 binds promoting HR and accurate DNA repair (Cappadocia et al. 2010).WHIRLY2 inhibit end-joining mechanisms by blocking the access of ssDNA 30-OH ends to SSBs and DNA polymerases (Gualberto and Kühn, 2014).By binding to resected DNA ends, WHIRLY2 prevents the annealing of ssDNA overhangs to micro-homologous sequences present in stretches of ssDNA (Marechal and Brisson 2010).This restricts the doublestrand break (DSB) repair by error-prone microhomologymediated break-induced replication, which would generate aberrant mtDNA recombination products negatively impacting plant physiology and reproduction efficiency (Cappadocia et al. 2010).Loss of WHIRLY2 inhibits HR repair of genotoxic-induced breaks, leading to significant rearrangements of the mtDNA, by increased ectopic recombination involving repeats of intermediate size and microhomologies (Janicka et al. 2012, Golin et al. 2020).
Recently, we have demonstrated that WHIRLY2, besides being crucial for mtDNA integrity, is also a nucleoid architectural protein determining nucleoid morphology.The reduced compactness of mitochondrial nucleoids in the why2-1 mutant was shown to negatively affect the biogenesis of mitochondria as well as their functionality, morphology and dynamics (Golin et al. 2020).Those results show, on the one hand, a link between the mitochondrial nucleoid structure and mitochondria morphology, dynamics and functionality and, on the other hand, that this link exists in plants, as was already shown in human cells (Li et al. 2016).
Recent research showed that the overexpression of SlWHIRLY2 in tobacco and tomato plants was associated with high resistance to drought stress (Meng et al. 2020) and pathogen attack (Zhao et al. 2018).During drought stress, the overexpression of SlWHIRLY2 is associated with an enhanced capacity of the reactive oxygen species (ROS) scavenging system and reduced ROS accumulation.The authors suggested that SlWHIRLY2 overexpression does enhance the tolerance to drought stress via transcriptional regulation of mitochondrial genes and stabilization of mitochondrial functionality (Zhao et al. 2018).
The present work shows that in A. thaliana, a loss of WHIRLY2 functionality increased the sensitivity of plants to salt stress compared with wild-type (WT) plants.Experimental evidence revealed that WHIRLY2 intervenes in orchestrating the cellular response to salt stress, both by its role in mtDNA repair and by its architectural impact on nucleoids that might function as a recruitment platform for other proteins involved in the maintenance of mitochondria integrity and the regulation of stress signaling pathways.

CRISPR/Cas9 mutant line characterization
To evaluate the role of WHIRLY2 in stress response, two lossof-function mutants were employed: the T-DNA insertional lines why2-1 available at 'The Arabidopsis Information Resource' (SALKseq_118900.0)and the why2-3 mutant lines generated by CRISPR/Cas9 mutagenesis.The why2-1 line has been already characterized (Cappadocia et al. 2010, Golin et al. 2020), while the why2-3 line was characterized in this work.In this latter line, a T insertion mutation is introduced, leading to a shift in the reading frame that results in the generation of a stop codon in the first exon upstream of the WHIRLY domain (Fig. 1).The characterization of the new mutant line compared with WT and why2-1 mutant is reported in Supplementary Fig. S1.The why2-3 phenotype resembles the one of why2-1.No obvious phenotype was observed during the vegetative growth under standard conditions, as revealed by the primary root length (Supplementary Fig. S1A), the primary inflorescence and rosette development (Supplementary Fig. S1D, E).Still, a reduction of about 25% of seed germination at 72 h was observed (Supplementary Fig. S1B), similar to the why2-1 (Golin et al. 2020).In the absence of WHIRLY2, mitochondria of why2-3 appeared elongated (Supplementary Fig. S1C).It has been confirmed that the disruption of WHIRLY2 increased the sensitivity to the genotoxic agents.Arabidopsis plants from the three genotypes (WT, why2-1 and why2-3) were grown in a solid medium supplemented with ciprofloxacin at different concentrations, and the amount of mtDNA was analyzed by using opposite direction primers as previously described (Cappadocia et al. 2010).As shown in Supplementary Fig. S1F, in both mutant lines, aberrant recombination products accumulated more than in WT, where aberrant products were observed only in the presence of the highest concentration of ciprofloxacin.Based on these results, the two mutant lines were considered equivalent.

cis-Regulatory element analysis
To gain information on the possible role of WHIRLY2 in stress responses, an in silico analysis of the gene's cis-regulatory elements was performed.For the analysis, the 344-bp region upstream of the gene and the first exon and first intron were considered.The inclusion of the first exon and first intron was done since they could also contain important regulatory sequences (Schmitz et al. 2022).
The analysis revealed the presence of 11 cis-regulatory elements, some of which are associated with abiotic stress responses (Fig. 2, Table 1).In particular, MYB1AT, MYB2CON-SENSUSAT and MYCCONSENSUSAT are binding sites of MYB and MYC transcription factors that participate in the regulation of many different stress responses (Baldoni et al. 2015).
Fig. 1 Generation of why2-3 knock-out mutant with the CRISPR/Cas system.Overview of the WHIRLY2 locus and target sites for the sgRNA-Cas9 complex.Rectangles and lines depict exons and introns in WHIRLY2 gene, respectively.A spacer sequence was designed in silico to target the sgRNA-Cas9 complex to the corresponding targets site (underlined) in Exon 1. Protospacer adjacent motif sequences are reported (PAM).Expected protein is interrupted at the 15 th amino acid.T-DNA insertion site of why2-1 mutant is indicated by a triangle.
They can be found in the promoter of the RD22, a dehydrationresponsive gene, associated with drought and water stress (Abe et al. 2003).Similarly, MYBCORE is also a binding site for two MYB transcription factors, AtMYB1 and AtMYB2, that have been shown to be involved in dehydration stress response (Urao et al. 1993) as well as AtMYB49 that is associated with an increased salt resistance (Zhang et al. 2020).The RAV1AAT sequence is the binding site of RAV transcription factors and has been found to participate in the regulation of drought and salt stress response and in the response to pathogen attacks (Sohn et al. 2006).The analysis also highlighted the presence of the LTRECOREATCOR15, which is a regulatory element associated with cold and drought response (Baker et al. 1994), and TCA1MOTIF, which is related to biotic and abiotic stresses response and salicylic acid-regulated expression (Li et al. 2020).Finally, several binding sites for light-induced transcription factors (EBOXBNNAPA, GT1CONSENSUS, IBOXCORE and SOR-LIP3AT) were also found, suggesting a possible involvement of WHIRLY2 in different cellular processes.The presence of these cis-regulatory elements in the WHIRLY2 gene supports the likelihood that WHIRLY2 is involved in abiotic stress responses in Arabidopsis.Particularly, the presence of several elements associated with drought, water and salt stress aligns with the findings in tomato by Zhao et al. (2018), hinting at the involvement and potential role of WHIRLY2 in salt stress response in Arabidopsis.

WHIRLY2 loss of function increased salt stress sensitivity in Arabidopsis
To functionally characterize WHIRLY2 and to test its role in stress response, the T-DNA insertion line (why2-1) and the why2-3 mutant were exposed to salt stress to evaluate their sensitivity in terms of germination efficiency and primary root length.It was observed that both parameters were significantly reduced in stress conditions.Still, there was a significant difference between WT and mutants: the reduction in germination efficiency at 96 h was about 44-49% in the two mutants compared to 24% in WT (Fig. 3A), and the primary root length was about 18-26% lower in the two mutants than in WT (Fig. 3B).
Based on these results, the two mutants showed an equivalent response to salt treatment, and thus the following experiments were conducted using the why2-1 mutant.
To evaluate the expression profile of WHIRLY2 during salt stress, WT plants were grown for 14 d in standard solid medium and subsequently exposed to salt (150 mM NaCl) for 8 h before collecting them for quantitative reverse transcription (qRT)-PCR analysis.The chosen salt concentration and the duration of the treatment were previously shown to substantially inhibit root growth without causing seedling death (Geng et al. 2013).After 8 h of exposure to salt, WHIRLY2 expression was significantly induced (Fig. 4A) and, at the same time, a stress signal was generated.This was evidenced by the upregulation of WRKY15 (Fig. 4B), an early hydrogen peroxide (H 2 O 2 )-responsive transcription factor playing a key role in salt stress response (Vanderauwera et al. 2012), and RD29A (Fig. 4B), a marker gene for drought/salt stress (Lee et al. 2016).
To determine spatial and stress-associated expression patterns of pWHIRLY2, the WHIRLY2 promoter:-glucuronidase (GUS) reporter gene fusion construct was examined.In 14-dayold seedlings treated as described earlier, a solid blue staining was observed in the perivascular tissue of the primary root (Fig. 4C), indicating the specific induction of the WHIRLY2.
It is worth noting that, within the WHIRLY gene family, only WHIRLY2 was induced by salt stress.At the same time, the expression of other members, WHIRLY1 and WHIRLY3, did not show any significant variation (Fig. 4A).In addition, there was no significative induction of genes coding for other proteins involved in mtDNA repair as ODB1, MSH1, OSB1-4 and SSB1 and SSB2.A slight induction has been observed in the gene coding for the mitochondria/chloroplast-localized OSB3 (Fig. 4A).

WHIRLY2 ensured mtDNA integrity during stress
To evaluate the impact of WHIRLY2 disruption on mitochondrial morphology in salt stress conditions, WT and mutant plants were grown for 23 d on a standard solid medium  The list was obtained by analyzing the sequence using the PLACE database.The position of the elements is reported from the translation start site (−upstream, +downstream) containing different concentrations of NaCl, and the root mitochondria were analyzed using transmission electron microscopy (TEM).As reported in Fig. 5, salt treatment induced alterations of mitochondrial morphology in both WT and mutant plants.
In the WT plants, under salt treatment, some mitochondria exhibited fewer cristae than in control conditions along with the presence of a translucent area, associated with nucleoid disassembly (Golin et al. 2020).In mutant plants, it has been confirmed that mitochondria contained fewer cristae than WT already under control conditions (Golin et al. 2020).Upon exposure to salt stress, mitochondria morphology was severely compromised.Due to an almost complete absence of cristae, the organelles appeared swollen and low electron-dense, while the translucent area, indicative of a disorganization of nucleoids, appeared to be enlarged.To evaluate the impact of WHIRLY2 on mtDNA integrity in salt stress conditions, PCR analyses were performed on specific mtDNA regions using opposite direction primers.In WT mitochondrial genome, these primers are expected to yield no product.However, in a rearranged genome with circular or head-tail concatemers, new fragments containing both extremities of the amplified regions would be present (Marechal and Brisson, 2010).As reported in Fig. 6A, many amplicons correspond to aberrant rearrangements accumulated in the mitochondria of mutant plants upon salt stress treatment.WT plants showed a low level of aberrant products even after exposure to the highest salt concentration.After 10 d of recovery in standard medium without salt, aberrant recombination products were no more visible in WT but still evident in the mutants (Supplementary Fig. S2).It has been observed that salt treatment induced an increase in mtDNA content (Fig. 6B) associated with an augmented expression of PolIB (Fig. 6C, D), likely due to an increase in the energy demand to cope with the stress.Interestingly, in the mutant, this increase was higher than that in WT plants.This result is consistent

Gene expression reprogramming upon salt stress is altered in WHIRLY2 loss-of-function mutants
The morphology alterations of mitochondria are associated with general stress occurring at the level of mitochondria as indicated by the increased expression of the mitochondrial chaperon mtHSC70-1 (Fig. 7).To acclimate to stress, plants must reprogram their gene expression and cellular metabolism to divert energy from growth and developmental processes toward stress responses.It is known that alteration in DNA metabolism of both chloroplasts and mitochondria can affect gene expression and genetic reprogramming in the nucleus (Cupp and Nielsen 2014).The organelles' activity is closely coordinated with transcriptional activity in the nucleus through retrograde regulation, a feedback mechanism modifying nuclear gene expression to maintain or restore organellar functions.This intricate interplay ensures efficient cellular function and overall homeostasis.To investigate whether the lack of WHIRLY2 could compromise the signaling of stress from mitochondria to the nucleus, the expression profile of the marker genes under retrograde regulation has been investigated.One of the primary targets of retrograde regulation is the alternative oxidase (AltOx) which contributes to a reduction of ROS produced under conditions of mitochondrial dysfunction (Umbach et al. 2005).UGT74E2, UPOX1 and HRG1 are also identified as marker genes for mitochondrial dysfunction under retrograde control (Khan et al. 2023, Kacprzak et al. 2020).The gene expression data revealed that all the considered genes were induced upon salt stress in WT and the why2-1 mutant.However, in the mutant line, their induction was significantly lower than that in WT (Fig. 7), indicating an apparent impairment in mitochondria-nucleus communication.

Salt-induced Ca 2+ dynamics
To assess whether WHIRLY2 disruption could impact stress perception, WT and why2 mutant were compared in order to evaluate the calcium signature, which represents one of the earliest signaling pathways triggered by salt stress (Othman et al. 2017).According to Yang and Guo (2018), a rapid calcium uptake induced by salt activates the CBL/CIPK complex, which triggers sodium extrusion through specific transporters.
A why2-1 transgenic line harboring the Cameleon YC3.6 calcium probe (Behera et al. 2013) was generated, and in vivo cytosolic Ca 2+ dynamics was analyzed in 5-day-old seedlings and in three different regions of the root (ROI 1, ROI 2 and ROI 3).Both WT and why2 roots showed rapid Ca 2+ transients upon salt addition and removal (Supplementary Fig. S3).The Förster resonance energy transfer (FRET) efficiency ratio calculated upon NaCl addition (IN) and removal (OUT) did not significantly differ between WT and mutant.These results suggest that salt sensitivity variance between WT and mutant is not linked to initial calcium signaling impairment triggered by salt perception.

WHIRLY2 loss of function compromised the redox-dependent response triggered by salt stress
It is known that salt stress, through the activation of NADPH oxidases, can trigger ROS accumulation, activating a signaling cascade and downstream responses (Rejeb et al. 2015, Yang andGuo 2018).Thus, H 2 O 2 accumulation was analyzed by dehydroascorbate (DAB) staining in the leaves of WT and why2-1 mutant, grown under standard conditions or treated for 8 h with 150 mM NaCl.Under standard conditions, H 2 O 2 accumulation was higher in the mutant line than in the WT (Fig. 8A).However, upon salt stress exposure, the level of H 2 O 2 increased in WT, while it did not show substantial differences in the mutant line.H 2 O 2 accumulation in the roots, measured by roGFP2-Orp1, showed the same behavior observed in the leaves, being higher in the why2-1 mutant than in WT under control conditions and increasing only in WT in response to salt stress (Fig. 8B).Accordingly, lipid peroxidation, evaluated in the whole seedlings, did not vary significantly in response to salt stress in the why2-1 mutant, whereas it increased in the WT (Fig. 8C).To characterize the redox-dependent defense response triggered by salt, the total content of the hydrophilic antioxidants, ascorbate (ASC) and glutathione (GSH), as well as the behavior of the two major enzymes involved in H 2 O 2 scavenging, namely, catalase (CAT) and ASC peroxidase (APX), were analyzed in the whole seedlings (Fig. 8, Supplementary Fig. S4).Under control conditions, the ASC content was significantly higher in the why2-1 mutant than in WT (Fig. 8D, Supplementary Fig. S4A); however, the total content of this antioxidant did not change after salt stress in both genotypes, whereby it remained higher in the mutant than in WT plants treated with NaCl.Total GSH content did not show any significant difference among the genotypes and in response to salt stress (Fig. 8E, Supplementary Fig. S4B).Total APX and CAT activities increased in the WT line after NaCl treatment.At the same time, they did not change significantly in the knock-out line (Fig. 8F, G, Supplementary Fig. S4C, D), showing a failure in activating ROS detoxification.To investigate whether the lack in the activity increase of H 2 O 2 scavenging enzymes was due to an impairment of signaling, the transcript levels of the two genes CAT2 and APX1 were analyzed (Fig. 8G,  H).After salt stress, CAT2 transcript levels rose in both WT and mutant lines, whereas cytosolic APX1 expression remained unaltered in both genotypes.(Fig. 8H, I).These results indicate that the lack of response to stress in terms of ROS scavenging system activation in mutant for WHIRLY2 is not due to an alteration of the transcriptional regulation of detoxification enzymes.

Discussion
The results of this study indicate that WHIRLY2 plays a crucial role in maintaining mtDNA integrity under stress conditions and that its disruption increases salt stress sensitivity in Arabidopsis.
Under standard conditions, the dramatic alterations at the mitochondrial level caused by the absence of WHIRLY2 do not have a noticeable impact on plant growth (Golin et al. 2020).However, under salt stress, mutants show more severe impairment in the germination rate and primary root growth compared to WT. Stress conditions, besides inducing widespread damage to cellular structures, including membranes and proteins, strongly impact the nucleoid structure and mtDNA integrity, which, in turn, activated mtDNA repair primarily involving WHIRLY2.Gene expression analyses in WT plants revealed indeed that expression of WHIRLY2, but not of other genes related to DNA repair in mitochondria, was significantly enhanced under salt stress conditions.Maintaining mtDNA integrity becomes imperative during stress response as it preserves mitochondrial functionality and, consequently, ensures efficient energy production.This agrees with the presence of stress-related regulatory sequences in the promoter of WHIRLY2.Under stress, the loss-of-function mutant did not repair damage in mtDNA, properly leading to the accumulation of aberrant recombination products that were not removed even after the stress subsided.The accumulation of aberrant mtDNA recombination fragments and the increased mtDNA synthesis were associated with a severe disturbance of nucleoid architecture and a profound alteration of the overall mitochondrial morphology.In the absence of WHIRLY2, the nucleoid architecture losing compactness may disrupt protein recruitment into complexes, causing improper replication and non-HR events.The data of this study revealed that WHIRLY2 is required to maintain mtDNA integrity because of its role in mtDNA repair (Marechal and Brisson 2010).In addition, the mtDNA integrity is necessary for nucleoid stability and, consequently, the maintenance of mitochondrial morphology and dynamics.
Mitochondria, besides being the primary powerhouse of the plant cell, play a role as environmental sensors being central to perceiving stress and orchestrating the cellular responses that allow the plant to cope with the stress (Liberatore et al. 2016).The absence of WHIRLY2 has a consequence not only on the alteration of mtDNA and nucleoid integrity but also on the efficiency of the communication between mitochondria and nucleus, as evidenced by the reduced stress-triggered transcriptional activation of nuclear genes that are under retrograde control.This interpretation follows the finding that the mutant plants are not impaired in stress perception and calcium-mediated signaling in the cytosol, while the redoxdependent response to stress was altered.First, it should be considered that under standard conditions, why2-1 mutant shows a higher H 2 O 2 level than WT.As previously shown in Golin et al. (2020), the lack of WHIRLY2 is associated with the disruption of mitochondrial internal structures with why2-1 mutants showing nucleoid and cristae disorganization (Golin et al. 2020).The alterations can lead to disorganization of the mitochondrial electron transport (mtETC), resulting in altered electron flow and loss of cytochrome c, leading to an increase in ROS accumulation.Moreover, why2-1 mutant compared to WT showed a higher ASC content.It is known that the last ASC biosynthetic step, catalyzed by l-galactono-1,4-lactone dehydrogenase (GALDH), is associated with the mtETC.The reaction provides an electron to cytochrome c (Bartoli et al. 2000), contributing to an over-reduction of the QH 2 pool and hence an increased level of mitochondrial ROS, which indeed was observed in the mutant, under standard conditions.It is important to note that GALDH is considered a moonlighting protein that also plays a role in the structural regulation of the mtETC being a key assembly factor of complex I (Schimmeyer et al. 2016).WHIRLY2 is known to be associated with complex I too (Senkler et al. 2017), so it can be proposed that the absence of WHIRLY2 impairs the interaction between complex I and GALDH, impacting the architecture/stability of mtETC.This, in turn, might shift the role of moonlighting GALDH to its function in ASC synthesis (Phua et al. 2021).Thus, the high ASC level under standard conditions appears to be sufficient to prevent oxidative damage and preserve plant vitality.After salt exposure, H 2 O 2 increase in WT plants leads to an enhancement of APX and CAT activities.In why2-1 mutant, the high ASC content effectively prevents ROS accumulation and damage over a short period of time, as it is in plant exposed to salt for 8 h.However, in the absence of H 2 O 2 accumulation, the mutant line failed to induce CAT and APX activities.Thus, the consequent lower ROS scavenging activity can impair the tolerance of plants to prolonged salt stress and could explain, at least in part, the elevated level of damage to the mtDNA under oxidative stress conditions.It is known that salt stress-derived injuries can be mitigated by a proper antioxidant response that is triggered by a reaction cascade that begins with an early H 2 O 2 generation by NADPH oxidase.Indeed, mutants lacking the NADPH oxidase are more sensitive to salt stress (Rejeb et al. 2015).However, after salt stress, no difference was observed among the two genotypes in the transcriptional profile of CAT2 and APX1.For these reasons, WHIRLY2 is likely involved in the regulation of the two enzymes responsible for H 2 O 2 scavenging at the posttranscriptional level, thus providing protection from long salt stress-derived damages.
All these data taken together show that disruption of WHIRLY2 in the mutant line causes a pleiotropic phenotype, which might be primarily due to a compromised nucleusmitochondria communication.Subsequently, the salt stress response, which includes the activation of those pathways that help the cell to buffer the damage and recover from the stress, is impaired.Our studies underline the importance of understanding the role of mitochondrial nucleoid architecture and integrity in stress response.Also, in humans/mammals, alterations of the mitochondrial nucleoid structure affect mitochondria function and mitochondrial-nuclear crosstalk, which is necessary to prevent severe diseases (Sharma et al. 2021).In plants, similar as in humans, mitochondrial nucleoids likely serve as a recruitment platform not only for oxidative phosphorylation complexes (Kim et al. 2021) but also for other proteins involved in the maintenance of mitochondria integrity and the regulation of stress signaling pathways.

Plant materials
The experiments were performed on A. thaliana plants, ecotype Columbia-0.Two different whirly2 mutants of A. thaliana (ecotype Columbia-0) were used.
WT and the why2-1 mutant line SALKseq_118900.0were obtained from the Nottingham Arabidopsis Stock Centre (NASC) and had been used in previous investigations (Maréchal et al. 2008, Golin et al. 2020).PCR analysis was performed to screen the plant collection and to check the insertion integrity.The why2-3 mutant was obtained by CRISPR/Cas9-mediated genome editing (Li et al. 2021).Annealed oligos were integrated via BbsI restriction enzyme sites in plasmid pSI57 yielding pGH511.The gRNA/Cas9 expression cassettes were introduced into p6i-d35S-TE9 (DNA-Cloning-Service, Hamburg, Germany) via SfiI, generating plasmid pGH483.This construct was used for the Agrobacterium-mediated transformation of ecotype Col-0.Seeds were germinated on a hygromycin B-containing medium and screened for the presence of the T-DNA by specific primers as described (Li et al. 2021).Mutations in WHIRLY2 were confirmed by Sanger sequencing.The pWHY2::GUS line was kindly provided by Qiang Cai from Peking University, China.Arabidopsis thaliana Col-0 lines harboring the Yellow Cameleon 3.6 NES biosensor cytosolic line (Loro et al. 2012) were crossed with the why2-1 line to produce why2 mutant lines harboring the different biosensors.
Treatments were performed on plants grown in 1/2MS medium either supplemented with 100 or 150 mM NaCl or with ciprofloxacin (0.25 and 0.75 μM) or treated for indicated times with the chemicals under different conditions.

β-GUS analysis
Histochemical staining to detect GUS activity in plants harboring the construct pWHIRLY2::GUS (Cai et al. 2015) was performed by immersion of 12-day-old plants, treated for 24 h with 150 mM NaCl, in GUS staining solution [2 mM X-Gluc; 0.05% of Triton-X-100; 0.5 mM of K 3 (Fe(CN) 6 ) × 3H 2 O; 0.5 mM of K 4 (Fe(CN) 6 ) × 3H 2 O; 10 mM EDTA; 50 mM buffer phosphate, pH 7] overnight (O/N), at 37 ∘ C in the dark.Root samples were incubated at 37 ∘ C O/N and the day after washing in sterile distilled water.After the incubation period and the washing, tissue samples were collected and placed on microscope slides with 70 μl of imaging buffer (10 mM MES Tris base; 1 mM CaCl 2 ; 5 mM KCl pH 5.8).GUS-stained plants were imaged using a stereomicroscope DMI 4000 and Leica DM6 B.

Phenotype analyses
For root length analysis, photographs of vertically grown seedlings were taken at indicated time using a Bio-Rad ChemiDoc Touch Imaging System (Flamingo setting) and each image was processed using Fiji-ImageJ bundle software.The experiments were performed at least three times with three biological replicates, and each replicate involved 20 seedlings.For germination analysis, sterile seeds were sown on solid medium and grown horizontally.The images were acquired using a stereomicroscope (Leica MZ16F).Each image was processed using Fiji-ImageJ bundle software.The experiments were performed at least three times with three biological replicates, and each replicate included 36 seeds.

Detection of DNA rearrangement products
Total DNA was extracted using the High-Quality DNA extraction kit following the protocol (Qiagen, Hilden, Germany).The samples were then analyzed by PCR using the GoTaq Polymerase (Merck KGaA, Darmstadt, Germany).For each reaction, 100 ng/μl of total DNA was used.DNA rearrangement events were detected using both outward-and inward-facing PCR primers spaced by ∼5-30 kb as previously described (Cappadocia et al. 2010).The sequence of the primer used in the analysis is reported in Supplementary Table S1.DNA was separated on 1.5% (w/v) agarose gel supplemented with GelRed.Cox1 (ATMG01360) was used as a reference gene to normalize the DNA samples.

mtDNA copy number assay
Total DNA was isolated with a High-Quality DNA extraction kit.To measure the amount of mtDNA, quantitative PCR (qPCR) was performed by using a mitochondrial gene (Orf170mito; AtMG00820) as a target and the nuclear-encoded RPOTP gene (GenBank accession number Y08463) as an internal standard (Preuten et al. 2010) for nuclear genes.The mtDNA amount was calculated using the ΔCt method (Livak and Schmittgen 2001).

qRT-PCR
One hundred milligrams of plant material was collected and ground in liquid nitrogen with mortar and pestle.Total RNA was extracted using the RNeasy® Plant Mini Kit (Qiagen) with an extra-protocol passage of 15 min of RNase-Free DNase (Qiagen) and resuspended in 30 μl of sterile nuclease-free water.RNA concentration was measured using a Nanodrop ND-1000 spectrophotometer (Nano Drop Technologies Llc, Wilmington, Delaware, USA).First-strand cDNA synthesis was performed with a SuperScript-IV Reverse Transcriptase kit (Thermo Fisher Scientific, Waltham, Massachusetts, USA), using 2 μg of RNA and 1 μl of random primers (Sigma).qRT-PCR was performed in plate using Taq® qPCR Master Mix (Promega) with SYBR Green technology in either QuantStudio 12K Flex or QuantStudio 5 (Thermo Fisher Scientific, Waltham, Massachusetts, USA) instrument.The total volume of each reaction was 10 μl using 0.25 μl of primer mix (10 μmol).Sequences of the used primers are reported in Supplementary Table S1.Arabidopsis Actin-2 (ACT2; At3g18780) was used as internal control.The relative expression was calculated using the ΔCt and ΔΔCt method (Livak and Schmittgen 2001).

Confocal imaging
For mitochondrial nucleoid analyses, seedlings were incubated for 4 min with 200 nM tetramethylrhodamine methyl-ester (TMRM) in imaging buffer (10 mM MES Tris base; 10 mM CaCl 2 ; 5 mM KCl, pH 5.8) and washed for 10 min in imaging buffer.Stained samples were analyzed under the Confocal Laser Scanning Microscope Zeiss LSM700.For TMRM detection, samples were excited at 535 nm, and fluorescence was collected at 600 nm.

Analyses of Ca 2+ dynamics
Arabidopsis seedlings were grown vertically in 1/2MS solid medium for 7 d.Samples were then gently mounted on a perfusion chamber and stabilized with cotton wool soaked in 200 μl of imaging buffer (10 mM MES Tris base; 1 mM CaCl 2 ; 5 mM KCl pH 5.8); an ISMATEC pump was used to administrate stimuli in continuous, setting the flow rate at 3 ml/min; each stimulus was added into the imaging buffer.The seedlings were kept under continuous perfusion.For Cameleon (Krebs and Schumacher 2013) analysis, the FRET CyanoFP/Yel-lowFP (CFP/YFP) optical block A11400-03 (Emission 1, 483/32 nm for CFP and Emission 2 542/27 nm for the FRET) with a dichroic 510 nm mirror (Hamamatsu Photonics, Shizuoka, Giappone) was used for the simultaneous CFP and FRET acquisitions (cpVenus for YC3.6 and YC4.6).Images were acquired in 5-s intervals with an exposure time of 500 ms.Filters and dichroic mirrors were purchased from Chroma Technology, Olching, Germany.The NIS-Element (Nikon) was used to control the microscope, illuminator, camera, and post-acquisition analyses, as previously reported (Giovanna Loro et al. 2016).FRET efficiency values were calculated as previously described (Loro et al. 2013).

Analyses of H 2 O 2 root content
For H 2 O 2 detection in the roots, we exploit the genetically encoded roGFP2-Orp1-probe (Nietzel et al. 2019).Arabidopsis seedling harboring the probe was crossed with why2-1 mutant lines to generate mutant lines harboring the probe itself.Arabidopsis seedlings grown vertically, as previously described for 14 d, were incubated for 8 h in liquid medium [1/2MS medium including vitamins (Duchefa), with 1% (w/v) sucrose, 0.5 g/l MES-OH, pH 5.8] supplemented with 0 or 150 mM NaCl.Samples were then gently mounted on a chamber and stabilized with cotton wool soaked in 200 μl of the same incubation liquid medium, and the signal was acquired for 3 min.RoGFP2-Orp1 was excited sequentially at 405 and 488 nm, and emission was recorded at 505-535 nm.Images were acquired at 5-s intervals with an exposure time of 300 ms.The NIS-Element (Nikon) was used to control the microscope, illuminator, camera and post-acquisition analyses.

TEM
TEM analyses were performed on plants collected at 23 d after sowing (DAS), which were grown in vitro on 1/2MS solid medium supplemented with 0 or 100 mM NaCl.Entire plants were fixed by incubating them O/N at 4 ∘ C in 2.5% (v/v) glutaraldehyde plus 2% paraformaldehyde (v/v) in 0.1M sodium cacodylate buffer, pH 7.4.The samples were post-fixed with 1% (w/v) osmium tetroxide for 2 h at 4 ∘ C.After three washes in water, the samples were dehydrated in ethanol and embedded in Epon resin (Sigma).Ultrafine sections (60-80 nm) of roots were obtained with a Leica Ultracut EM UC7 ultramicrotome, subsequently contrasted with 1% (w/v) uranyl acetate and 1% (w/v) lead citrate and visualized with a Tecnai G2 (FEI) transmission electron microscope operating at 100 kV.Images were captured with Velta (Olympus Soft Imaging System, Tokyo, Japan) digital camera.

Determination of hydrogen peroxide and lipid peroxidation
In situ H 2 O 2 accumulation in leaves was detected with DAB as described in Fortunato et al. (2022).'ImageJ' software digitally acquired and quantified the staining intensity (https://imagej.nih.gov/ij/).The relative H 2 O 2 levels were calculated as the percentage of DAB-stained area of leaves.The level of lipid peroxidation was evaluated in terms of malondialdehyde (MDA) content determined by thiobarbituric acid (TBA) reaction, as described by Paradiso et al. (2008).The amount of MDA-TBA complex was calculated using an extinction coefficient of 155 mM −1 cm −1 .

Analysis of enzymatic and non-enzymatic antioxidants
For ASC and GSH analyses, 0.3 g of samples was homogenized at 4 ∘ C with 1.8 ml of 5% (v/v) trichloroacetic acid.After centrifugation at 18,000×g for 20 min, the supernatants were collected and ASC and GSH levels were determined through the colorimetric assay described in de Pinto et al. (1999).For the determination of the activities of antioxidants enzymes, seedlings were ground in liquid nitrogen and homogenized at 4 ∘ C in a 1:8 (w/v) ratio with the extraction buffer (50 mM Tris-HCl pH 7.5, 0.05% (w/v) cysteine, 0.1% (w/v) bovine serum albumin, 1 mM phenylmethanesulfonyl fluoride).To determine APX activity, 1 mM ASC was added to the extraction buffer.Homogenates were centrifuged at 20,000×g for 15 min, and the supernatants were used for spectrophotometric and electrophoretic analyses.The protein concentration was determined according to Bradford (1976), using bovine serum albumin as a standard.CAT (EC 1.11.1.6)activity was spectrophotometrically measured according to Paradiso et al. (2020), and APX (APX,EC 1.11.1.11)activity was determined according to de Pinto et al. (1999).

cis-Regulatory elements analysis
A sequence of 464 bp composed of the region upstream (344 bp), the first exon and the first intron (120 bp) of the gene WHIRLY2, was selected.The sequence was analyzed for cis-element prediction by using the Plant cis-acting regulatory DNA elements (PLACE) database (Higo et al. 1999).

Fig. 2
Fig. 2 Putative cis-regulatory elements of WHIRLY2.Position of the different transcription factor binding sites found by the use of the PLACE database.The gray line represents the 344-bp region upstream of the start codon, the black box the first exon and the black line the first intron of the gene.Each colored box represents a different cis-element, with the width corresponding to the sequence length and its position on the sense (+) or antisense (+) strand.

Fig. 5
Fig. 5 WHIRLY2 disruption affects mitochondrial and nucleoid organization.Transmission electron microscope images of WT and why2-1 root epidermal cells of seedlings grown for 23 d on solid medium with or without 100 mM NaCl.Bar 500 nm.

Fig. 6
Fig. 6 Why2-1 mutants accumulate aberrant recombination products during salt stress.(A) Electrophoretic analysis of representative PCR performed with two inward-facing mitochondria genome-directed primers.The Cox1 mitochondrial gene was used as loading controls.The oligonucleotides used for each PCR are indicated.Plants were grown for 23 DAS in normal solid medium under different concentrations of salt (0, 100 or 150 mM NaCl).The experiment was repeated three times, and 36 plants per biological replicate were used.(B) mtDNA copy number calculated on the level of orf170mito (mitochondrial RNA-dependent DNA polymerase).RPOTP (nuclear plastid-RNA polymerase) has been used as housekeeping gene.(C) Relative expression of PolIB.ACT2 has been used as housekeeping gene.The results were analyzed using the 2 (−ΔCt) method.Statistical significance (P-value from Student's t-test and Tukey's multiple comparisons) is indicated by asterisks (*P < 0.05; **P < 0.005; ***P < 0.001; ****P < 0.0001).Error bars: SE.

Fig. 8
Fig. 8 WHIRLY2 is necessary for proper redox-dependent response under salt stress.(A) Representative image of H 2 O 2 accumulation, visualized by diaminobenzidine (DAB staining) in WT and why2-1 mutant seedlings grown for 14 d and treated for 8 h with or without 150 mM NaCl in liquid medium.The analyses were repeated three times showing reproducible results.The percentage area (±SE) of 60 leaves (20 for each experiment) stained with DAB is reported inside the images.(B) H 2 O 2 level detected in root in roGFP2-Orp1 sensor line WT and why2-1 in response to 150 mM NaCl.Normalized average sensor ratios are plotted ± SE value (n = 9).(C) Lipid peroxidation, measured as MDA content, (D) total content of ASC and (E) total GSH, (F) total APX and (G) CAT activities in WT and why2-1 mutant seedlings grown for 14 d and treated for 8 h with 150 mM NaCl.The values are plotted as ±SE (n = 5-7).(H and I) Expression profile of Ascorbate peroxidases 1 and Catalase2 on 14 DAS seedlings treated for 8 h with 150 mM NaCl.ACT2 has been used as housekeeping gene.The results were analyzed using the 2 (−ΔCt) method.Statistical significance (P-value from the t-test) is indicated by asterisks (*P < 0.05; **P < 0.005; ***P < 0.001).Error bars: SE.

Table 1
Putative cis-regulatory elements present in the 344-bp region upstream of WHIRLY2, first exon and first intron