Cytosolic and Nucleosolic Calcium-Regulated Molecular Networks in Response to Long-Term Treatment with Abscisic Acid and Methyl Jasmonate in Arabidopsis thaliana

Calcium acts as a universal secondary messenger that transfers developmental cues and stress signals for gene expression and adaptive growth. A prior study showed that abiotic stresses induce mutually independent cytosolic Ca2+ ([Ca2+]cyt) and nucleosolic Ca2+ ([Ca2+]nuc) increases in Arabidopsis thaliana root cells. However, gene expression networks deciphering [Ca2+]cyt and [Ca2+]nuc signalling pathways remain elusive. Here, using transgenic A. thaliana to selectively impair abscisic acid (ABA)- or methyl jasmonate (MeJA)-induced [Ca2+]cyt and [Ca2+]nuc increases, we identified [Ca2+]cyt- and [Ca2+]nuc-regulated ABA- or MeJA-responsive genes with a genome oligo-array. Gene co-expression network analysis revealed four Ca2+ signal-decoding genes, CAM1, CIPK8, GAD1, and CPN20, as hub genes co-expressed with Ca2+-regulated hormone-responsive genes and hormone signalling genes. Luciferase complementation imaging assays showed interactions among CAM1, CIPK8, and GAD1; they also showed interactions with several proteins encoded by Ca2+-regulated hormone-responsive genes. Furthermore, CAM1 and CIPK8 were required for MeJA-induced stomatal closure; they were associated with ABA-inhibited seed germination. Quantitative reverse transcription polymerase chain reaction analysis showed the unique expression pattern of [Ca2+]-regulated hormone-responsive genes in cam1, cipk8, and gad1. This comprehensive understanding of distinct Ca2+ and hormonal signalling will allow the application of approaches to uncover novel molecular foundations for responses to developmental and stress signals in plants.


Introduction
Sessile plants respond to developmental and environmental cues through various secondary messengers and phytohormones. Calcium ion (Ca 2+ ) is a universal secondary messenger involved in fertilisation, growth, and development; it responds to both biotic and abiotic stresses in plants [1][2][3]. In addition, plants produce hormones (e.g., abscisic acid (ABA) and jasmonic acid (JA)) to help them grow, develop, and adapt to their changing environments [4,5]. Both ABA and methyl jasmonate (MeJA) can induce increases in cytosolic Ca 2+ ([Ca 2+ ] cyt ) in plant guard cells and roots [6][7][8]. Using tobacco suspension cells transiently transformed with cytosolic-or nucleus-targeted apoaequorin, Walter et al. [9] found that both 12-oxophytodienoic acid, the precursor of JA, and JA itself could separately induce a transient increase in nuclear Ca 2+ ([Ca 2+ ] nuc ) and [Ca 2+ ] cyt in a concentration-dependent on genes transcriptionally regulated by Ca 2+ and hormones, we defined a highly reliable gene co-expression network among [Ca 2+ ] cyt -and [Ca 2+ ] nuc -regulated genes, Ca 2+ signal-decoding proteins, and ABA/JA signalling pathway proteins to uncover dynamic subnetwork structures in response to hormone treatment in A. thaliana.
A. thaliana seeds were surface-sterilised with 75% ethanol and plated on half-strength Murashige and Skoog (MS) salts, 2% sucrose, and 0.8% (w/v) agar at pH 5.8. After stratification at 4 • C in the dark for 3 days, the plates were transferred to a growth chamber under 16 h light (120 µmol m −2 s −1 )/8 h dark at 22 • C for another 3 days. Seedlings were then transferred to half-strength MS plates containing 10 µM ABA or 50 µM MeJA, cultured for 5 additional days, and collected for RNA extraction. For the germination assay, seeds were plated on half-strength MS plates containing 0.5 µM ABA or 50 µM MeJA at 4 • C in the dark for 3 days. The plates were then transferred to a growth chamber for 3 days, and the seed germination rate was calculated.

Imaging of [Ca 2+ ] cyt and [Ca 2+ ] nuc in A. thaliana Roots
[Ca 2+ ] measurements were performed in mature root sections of 1-week-old A. thaliana seedlings, as previously described [11]. Transgenic A. thaliana lines containing cytosoliclocalised yellow Cameleon YC 3.6 (NES-YC3.6) or nuclear-localised yellow Cameleon YC 3.6 (NLS-YC3.6) were used to monitor the effects of ABA and MeJA on changes in [Ca 2+ ] cyt and [Ca 2+ ] nuc in root cells. After germination, A. thaliana seedlings were grown vertically on half-strength MS medium for 5-7 days; the roots were immobilised by overlaying 1% (w/v) low-melting-point agarose (AMRESCO, Dallas, TX, USA) in an Attofluor ® Cell Chamber (Invitrogen, Waltham, MA, USA). After a small tunnel in the agarose had been made to expose the root, 200 µL of bathing solution buffer (0.5× MS salt, 1% sucrose, 10 mM MES-KOH, pH 5.8) were applied to the chamber. ABA (10 µM) and MeJA (50 µM) in the same bathing solution were separately perfused as the stimulus into the chamber. The mean fluorescence resonance energy transfer values in response to different stimuli represented measurement of 20-30 root cells from at least nine independent seedlings, each of which included three to six root cells. Analyses of statistical significance were performed using the unpaired Student's t-test with GraphPad Prism 5.0 (GraphPad Software, San Diego, CA, USA) and the results are presented as means ± standard deviations (SDs).

Stomatal Aperture Bioassay
Rosette leaves were detached from 3-to 4-week-old A. thaliana plants and floated in a solution containing 10 mM KCl, 2 mM CaCl 2 , and 10 mM MES-Tris at pH 6.15 for 2 h in a growth chamber under light (120 µmol m −2 s −1 ) at 22 • C. ABA (10 µM) and MeJA (50 µM) were separately added to the solutions for 2 h to assess stomatal closure, as previously described [29]. Stomatal apertures were imaged with a digital camera (DP72; Olympus Corp., Tokyo, Japan) attached to a fluorescence microscope (BX51; Olympus Corp.) and measured using DP2-BAW software (Olympus Corp.).

Total RNA Isolation and Analysis of Microarray Data
Two wild-type (WT) plants and three previously described independent PV-NES lines (1, 7, and 11) and PV-NLS transgenic lines (4, 5, and 8) [11] were used for microarray analysis. First, total RNA was extracted from seedlings that had been treated with 10 µM ABA or 50 µM MeJA for 5 days using the Eastep ® Super Total RNA Extraction Kit (Promega, Madison, WI, USA), in accordance with the manufacturer's instructions. Approximately 2 µg of total RNA were reverse-transcribed into first-strand cDNA using the First-Strand cDNA Synthesis SuperMix (TransScript, Beijing, China). The cDNA was hybridised on an Affymetrix A. thaliana ATH1-121501 genome array using a GPL198 platform (CapitalBio Corporation, Beijing, China). All generated datasets are publicly available in the Gene Expression Omnibus database under the accession number GSE109611.
Files in .cel format were read into R and normalised using the RMA procedure. The cor function was used to calculate the Pearson correlation coefficient, represented by R. The R-value of duplicate samples ranged from 0.98 to 1, indicating high relevance and repeatability. Volcano plot analysis, with a fold-change >2.0 and a false discovery rate-corrected p-value < 0.05, according to Student's t-test, was performed to identify differentially expressed genes; such genes were visualised using Venn diagrams and heatmaps.

Construction of the A. thaliana Gene Co-Expression Network
Seven open or published A. thaliana chip datasets regarding ABA and JA hormone processing were collected from the Affymetrix GPL198 platform and downloaded from the Gene Expression Omnibus database on the National Center for Biotechnology Information website: GSE12715, GSE45662 [30], GSE84446, GSE39384 [31], GSE7432 [32], GSE5620 [33], and GSE109611. In accordance with a previously published method [34], R software was used to pre-process the data of each chip, mainly in terms of removing disproportionately low mean expression genes or genes with poor correlations. Finally, the dataset contained 13,083 genes in 165 samples. Then, the weighted gene co-expression network analysis algorithm in R software was used to construct the A. thaliana co-expression network. This algorithm can generate functional modules in various situations [35,36]. To construct the co-expression network, the weight value power selected in this study was 12; the corresponding Pearson correlation coefficient was approximately 0.9. Cytoscape_V3.2.1 was used to display the co-expression network results.

Firefly Luciferase Complementation Imaging (LCI) Assay
The LCI assay was conducted as previously described [37]. Coding sequences of glutamate decarboxylase 1 (GAD1), calmodulin 1 (CAM1), and calcineurin B-like proteininteracting protein kinase 8 (CIPK8) were inserted into the pCAMBIA-nLUC or pCAMBIA-cLUC vector. The Ca 2+ -regulated hormone response genes were cloned into the pCAMBIA-cLUC vector. The constructed plasmids and conjugative P19 plasmid were separately transformed into Agrobacterium GV3101 strains. A single colony was inoculated in the corresponding resistant YEB liquid medium and cultured overnight at 28 • C and 200 rpm. After centrifugation for 2 min at 12,000 rpm, the medium was discarded. The pellet was washed five times with tobacco transformation buffer (10 mM MES, 10 mM MgCl 2 , pH 5.6), resuspended with tobacco transformation buffer containing 0.1 mM acetosyringone, and infiltrated into the leaves of 3-week-old Nicotiana benthamiana (16 h day/8 h night, 25 • C) with pairs of nLUC and cLUC vectors. After 2-3 days, the luciferase assay substrate (Promega) was infiltrated into the leaf and reaction imaging was immediately captured by a low-light cooled charge-coupled device imaging system (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The experiments were repeated at least three times.

qRT-PCR
Four-day-old seedlings were transferred to half-strength MS medium containing 10 µM ABA or 50 µM MeJA for 6 h, followed by RNA extraction and first-strand cDNA as described in Section 2.4. qRT-PCR was performed with Power SYBR ® Green PCR Master Mix (TransStart, Beijing, China) on a 7500 Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). Value changes of more than two-fold (>2 or <0.5) were considered to indicate the induction or repression of gene expression. All primers used in this study are listed in Table S1. The A. thaliana actin2 gene served as an internal control. At least three independent biological replicates were performed.

Statistical Analysis
Data analyses were carried out using the Data Processing System [38]. Data are shown as means ± SDs of at least three independent experiments. Statistical comparisons were performed using one-way or two-way analysis of variance with Student's t-test to identify significant differences among group means; p < 0.05 was considered to indicate statistical significance.

Both ABA and MeJA Triggered [Ca 2+ ] cyt and [Ca 2+ ] nuc Increases in A. thaliana Roots
Transgenic A. thaliana lines containing cytosolic-localised yellow Cameleon YC 3.6 (NES-YC3.6) or nuclear-localised yellow Cameleon YC 3.6 (NLS-YC3.6) were used to monitor the effects of ABA and MeJA on changes in [Ca 2+ ] cyt and [Ca 2+ ] nuc in root cells. [Ca 2+ ] measurements were performed in mature root sections of 1-week-old A. thaliana seedlings, as previously described [11]. Consistent with previous results from plant stomata [6,39], we demonstrated that both ABA and MeJA triggered [Ca 2+ ] cyt elevations in A. thaliana roots ( Figure 1A Figure S3). These results suggest that both cytosolic and nucleosolic Ca 2+ signals are involved in hormone-regulated gene expression and network crosstalk in A. thaliana.

Integrative Co-Expression Analysis to Identify Hub Genes within Calcium-Regulated Transcriptional Modules
To further characterise the roles of calcium in phytohormone-regulated gene expression, we generated integrative co-expression networks of [Ca 2+ ]-regulated hormoneresponsive genes with Ca 2+ signal-decoding genes and ABA or JA pathway proteins. First, we collected A. thaliana microarray datasets that were used in the Affymetrix platform GPL198 analysis, filtered outliers from the data using a published method [34], and compiled microarray datasets containing 165 samples and 13,083 genes (Table S4); these covered approximately half of the coding genes in the A. thaliana genome. Subsequently, we constructed a co-expression network map by calculating the Pearson correlation coefficient among the expression values of different genes in accordance with the standard WGCNA procedure; we identified modules with biologically correlated genes. The power value was set to 12 and the Pearson correlation coefficient was set to >0.9, indicating a high correlation; the most highly connected genes within each module, designated as 'hubs', could be key regulators in response to ABA and/or MeJA in A. thaliana.  Tables S2 and S3. Next, we placed Ca 2+ -regulated ABA-responsive genes, calcium signal-decoding genes (Table S5), and ABA pathway member genes (Table S6) into the constructed A. thaliana gene co-expression database to obtain a Ca 2+ -regulated ABA signalling co-expression network ( Figure 3A,B, Table S7). Similarly, we placed Ca 2+ -regulated MeJA-responsive genes, calcium decoding protein genes, and JA pathway member genes into the constructed gene co-expression database to obtain a Ca 2+ -regulated JA signalling co-expression network ( Figure 3C,D, Table S7). We found that three calcium signal-decoding genes-calmodulin 1 (CAM1), calcineurin B-like protein-interacting protein kinase 8 (CIPK8), and glutamate decarboxylase 1 (GAD1)-were hub genes in the main submodules of Ca 2+ -regulated ABA and JA signalling co-expression networks ( Figure 3A,C). In these two modules, CAM1, CIPK8, and GAD1 were co-expressed with [Ca 2+ ] cyt -regulated ABA-responsive genes such as  Figure 3A,C). In two other submodules, the calcium decoding protein gene CPN20 co-expressed with PSBZ, a [Ca 2+ ] cyt -regulated ABA-responsive gene, and CLA1, an ABA pathway member, constituted the hub genes of a Ca 2+ -regulated ABA signalling co-expression network ( Figure 3B). Similarly, CPN20 co-expressed with AT3G26440, a [Ca 2+ ] cyt -regulated JA-responsive gene, was identified as a hub component in a Ca 2+regulated JA signalling co-expression network ( Figure 3D). These results indicated that these hub genes, CAM1, CIPK8, GAD1, and CPN20, have important regulatory roles in cytosolic and nucleosolic calcium-mediated hormone-responsive gene expression.

Interactions among the Hub Proteins CAM1, CIPK8, and GAD1 and with Proteins Encoded by Ca 2+ -Regulated Hormone-Responsive Genes
We next performed the LCI assay in the leaves of 3-week-old N. benthamiana to detect interactions among the three hub proteins, as well as interactions with proteins encoded by Ca 2+ -regulated hormone-responsive genes. First, we showed an interaction of CIPK8 with CAM1, GAD1, and itself; we also showed an interaction of the protein encoded by AT1G28400, a [Ca 2+ ] cyt -regulated ABA-responsive and [Ca 2+ ] cyt -and [Ca 2+ ] nuc -coregulated JA-responsive gene, with CAM1, but not with GAD1 and CIPK8 ( Figure 4). Interactions between the encoding proteins of several Ca 2+ -regulated hormone-responsive genes and each of the three hub genes are summarised in Table 1; the interaction between CAM1 and GAD1 was used as a positive control. In the [Ca 2+ ] cyt -regulated ABA-responsive genes AT1G64330, NDHA, HIR2, CAPE2, PGSIP2, and PSBI, we observed interactions of AT1G64330 with GAD1 or CAM1, PSBI with CAM1 or CIPK8, and PGSIP2 with CAM1 ( Figures S4 and S5). The encoded protein of AT3G45160, a [Ca 2+ ] cyt -regulated JA-responsive gene, interacted with GAD1 and CAM1, but not CIPK8 ( Figure S5). In the [Ca 2+ ] cytand [Ca 2+ ] nuc -co-regulated ABA-responsive genes NTMC2T6.1 and RCI2B, we observed interactions of RCI2B with GAD1 or CAM1, and NTMC2T6.1 with CAM1 ( Figure S6). In the [Ca 2+ ] cyt -and [Ca 2+ ] nuc -co-regulated JA-responsive genes AT3G19370 and AT1G54410, we observed interactions of AT3G19370 with GAD1 or CAM1; there were no interactions between AT1G54410 and any of the three hub proteins ( Figure S7). These results provide clues for dissecting the mechanism underlying hub protein-mediated regulation of the expression of ABA and JA responsive genes.  Empty vectors were used as negative controls. Luminance intensity indicates an interaction between any two proteins. The experiments were repeated three times and a representative read-out is shown.  (Figures S4-S7).

CAM1 and CIPK8 Are Required for MeJA-Induced Stomatal Closure and Are Associated with ABA-Inhibited Seed Germination
We monitored the roles of [Ca 2+ ] cyt and [Ca 2+ ] nuc in hormone-induced stomatal closure in A. thaliana. Both ABA-and MeJA-induced stomatal closure were severely impaired in PV-NES plants, but not in PV-NLS plants ( Figure 5A,B), indicating that ABA-and MeJA-induced stomatal closure requires [Ca 2+ ] cyt , but not [Ca 2+ ] nuc . Furthermore, ABAinduced stomatal closure was impaired in cam1, but not gad1 and cipk8 ( Figure 5C); MeJAinduced stomatal closure was impaired in cam1 and cipk8, but not gad1 ( Figure 5D). These results suggest that CAM1 is required for ABA-and MeJA-induced stomatal closure, while CIPK8 is essential for MeJA-induced stomatal closure. In addition, ABA-inhibited seed germination was more severe in both PV-NES and PV-NLS plants than in WT plants ( Figure 5E). Similarly, ABA-inhibited seed germination was more severe in cam1 and cipk8 than in gad1 ( Figure 5F). However, MeJA had no effect on seed germination in any of the tested plants ( Figure 5E,F). These results indicate that CAM1 and CIPK8 regulate ABA-inhibited seed germination. Here, we found that NCED3 was a [Ca 2+ ] cyt -regulated ABA-responsive gene, ERF104 was a [Ca 2+ ] cyt -regulated ABA-and MeJA-responsive gene, PR1 was a [Ca 2+ ] nuc -regulated ABA-and MeJA-responsive gene, and AGL21 was a [Ca 2+ ] cyt -and [Ca 2+ ] nuc -regulated MeJA-responsive gene (Tables S2 and S3). Therefore, we monitored the expression patterns of these [Ca 2+ ]-regulated genes in different A. thaliana plants treated with ABA or MeJA for 6 h by qRT-PCR. First, we found that ERF104 was induced by ABA only in PV-NES plants and by MeJA only in WT plants ( Figure 6A,B), consistent with our microarray data. In addition, ERF104 was induced by ABA in cam1, but not in cipk8 and gad1; it was not induced by MeJA in these three mutant lines ( Figure 6C,D). Thus, CAM1 is required for ABA-induced expression of ERF104; CAM1, CIPK8, and GAD1 positively regulate MeJAinduced expression of ERF104. Second, we found that PR1 was downregulated by ABA and MeJA in WT plants, but strongly upregulated by MeJA in PV-NLS plants ( Figure 6A,B). Furthermore, PR1 expression was activated by ABA and MeJA in cipk8, but the effects were less robust in cam1 and gad1 ( Figure 6C,D). Third, NCED3 was activated by ABA in WT plants; however, this type of activation was decreased in PV-NES, PV-NLS, and cam1 lines ( Figure 6A,C). Finally, AGL21 was induced by MeJA in PV-NLS and cipk8 plants, but not in WT, PV-NES, or gad1 plants ( Figure 6B,D). These results suggested that CAM1, CIPK8, and GAD1 act as key regulators that modify the expression patterns of calcium-regulated genes in response to ABA and MeJA in A. thaliana seedlings.

Discussion
Ca 2+ serves as a vital secondary messenger in the mediation and integration of multiple hormone signalling pathways that specify cell-signalling information during plant growth and development. Therefore, it is challenging to elucidate the complexity of hormone and Ca 2+ crosstalk; such analysis requires a combined approach that involves both experimental measurements and systems-level modelling. A prior study revealed 269 [Ca 2+ ] cyt -upregulated genes in A. thaliana seedlings that responded to three specific types of [Ca 2+ ] cyt elevations elicited by artificial electrical stimulation [28]. However, the identification of [Ca 2+ ]-responsive genes under physiological conditions in plants remains challenging. Our recent study demonstrated that both osmotic-and salt-stress-induced [Ca 2+ ] cyt and [Ca 2+ ] nuc increases were impaired in the roots of transgenic A. thaliana lines containing PV-NES or PV-NLS, respectively [11], thus providing a powerful tool for the establishment of an interaction landscape of hormone and Ca 2+ signalling in A. thaliana. Here, we identified 244 [Ca 2+ ] cyt -and [Ca 2+ ] nuc -regulated ABA-responsive and 144 [Ca 2+ ] cytand [Ca 2+ ] nuc -regulated MeJA-responsive genes in A. thaliana seedlings using full genome microarray analysis; we found that 22 genes overlapped. Unlike other Ca 2+ -based transcriptome studies, we confined our conditions to a long duration of exposure to ABA or MeJA; this facilitated enrichment of late transcriptional events. Through co-expression network analysis, we unveiled two common modules among [Ca 2+ ] cyt -and [Ca 2+ ] nuc -regulated hormone-responsive genes, Ca 2+ signal-decoding genes, and ABA/JA signalling pathway genes; our results provide valuable clues for use in exploring novel functions of known genes or potential functions of unknown genes, while enabling the dissection of crosslinks among Ca 2+ signalling and hormonal pathways via co-expression networks.
Previous studies showed that cytosolic and nucleosolic calcium increases in response to various stimuli are mutually independent in animal [40] and plant cells [11]. Moreover, gene expression patterns regulated by nuclear calcium are also independent of cytosolic calcium in animal cells. For example, signalling pathways activated by [Ca 2+ ] cyt target the serum-response element, whereas [Ca 2+ ] nuc increases are critical for cyclic AMP response element-dependent calcium-activated transcription in hippocampal neurons [41]. However, Thompson et al. [42] found that [Ca 2+ ] nuc , but not [Ca 2+ ] cyt , negatively regulates the activity of transcription enhancer factor in Chinese hamster ovary (CHO) cells. In plants, [Ca 2+ ] nuc oscillations in response to Nod factor treatment are mediated by three nuclear-localised cyclic nucleotide-gated channels (CNGC15a/b/c); such oscillations are required for the establishment of symbiosis by nitrogen-fixing rhizobial bacteria in Medicago truncatula [43]. Here, we characterised [Ca 2+ ] cyt -activated/repressed, [Ca 2+ ] nuc -activated/repressed, and [Ca 2+ ] cyt -and [Ca 2+ ] nuc -co-regulated genes in response to ABA and MeJA treatment in A. thaliana seedlings. Furthermore, some genes were shared among the [Ca 2+ ]-regulated ABAresponsive and [Ca 2+ ]-regulated MeJA-responsive genes, indicating that both cytosolic and nucleosolic calcium are involved in the transcriptional regulation triggered by ABA or MeJA treatment in A. thaliana. Notably, we found that four calcium signal-decoding genes, CAM1, GAD1, CIPK8, and CPN20, were hub genes in two modules of Ca 2+ -regulated ABA and JA signalling co-expression networks. Additionally, we found the direct interactions among hub genes and Ca 2+ -regulated hormone-responsive genes with different combinations in LCI assays. These results provide important clues concerning hub genes CAM1, GAD1, CIPK8, and CPN20 that mediate crosstalk between Ca signalling and hormone pathways in A. thaliana, suggesting the implications for other plants or crops under biotic and abiotic stress [44,45].
An intriguing finding was that the chloroplast genome encoded several [Ca 2+ ] cytactivated ABA-responsive genes, including PSBI, YCF9, PSBJ, RPOA, RPL14, RPS3, RPL22, RPS19, YCF1.1, PSAC, ndhE, and ndhA, as well as one [Ca 2+ ] cyt -and [Ca 2+ ] nuc -activated ABA-responsive gene, PSBT, and one [Ca 2+ ] nuc -repressed MeJA-responsive gene, PSBD. A previous study showed that the chloroplast can function as a sensor for environmental stimuli, such as drought stress, and initiate signals that regulate nuclear gene expression [46]. Plastid-derived signals that target the regulation of nuclear gene expression are considered retrograde signals [47]; nucleus to plastid signalling is considered an anterograde pathway [48]. Wang et al. [49] showed that many genes related to drought stress responses, ABA metabolism, chloroplast biogenesis, and chlorophyll degradation are strongly expressed at early time points, followed by gradual decreases in induction or even suppression at later time points, during long-term ABA treatment in A. thaliana. Here, we found that PSB1 and ndhA were co-expressed with hub genes CAM1, CIPK8, and GAD1, while RPOA was co-expressed with CAM1 and GAD1 in the Ca 2+ -regulated ABA signalling co-expression network. In addition, LCI assays showed that PSB1 interacts directly with CAM1 and CIPK8. These results indicated that the ABA-induced increase in [Ca 2+ ] cyt , but not in [Ca 2+ ] nuc , has a vital role in combining and coordinating the expression patterns of nuclear and plastid genes for bi-directional communication between chloroplasts and the nucleus [50].
Crosstalk between JA and ABA signalling through interactions of ABA receptor PYRABACTIN RESISTANCE1-Like proteins (PYLs) and JASMONATE ZIM DOMAIN (JAZ) proteins, transcription inhibitors of JA signalling, is known to coordinate the balance between plant growth and defence resistance [51]. Here, we revealed a new crosstalk mechanism between JA and ABA signalling via [Ca 2+ ] cyt and [Ca 2+ ] nuc . We first showed that both ABA and MeJA induce [Ca 2+ ] cyt and [Ca 2+ ] nuc increases in A. thaliana roots; ABAand MeJA-induced stomatal closure is impaired in PV-NES, but not PV-NLS, plants. In addition, ABA-and MeJA-induced stomatal closure is impaired in cam1 plants. These results indicated that CAM1 perceives ABA-and MeJA-induced increases in [Ca 2+ ] cyt to regulate stomatal closure in A. thaliana. Hossain et al. [52] showed that MeJA-induced, but not ABA-induced, [Ca 2+ ] cyt elevation and stomatal closure are disrupted in the ABAdeficient mutant aba2-2 or by treatment with the ABA synthetic inhibitor fluridon in WT plants, consistent with our results in cipk8 plants; thus, CIPK8 presumably regulates ABA biosynthesis in A. thaliana. Furthermore, we found that ABA-inhibited seed germination is more sensitive in PV-NES, PV-NLS, cam1, and cipk8 plants than in WT plants. These results indicated that both CAM1 and CIPK8 participate in cross-talk during ABA-and JA-regulated biological processes in A. thaliana. Using LCI assays, we showed that CIPK8 interacts with CAM1; many Ca 2+ -regulated ABA-and JA-responsive genes, such as ERF104, PR1, and AGL21, exhibit various expression patterns in cam1 and cipk8 plants. These results suggested that the CAM1-CIPK8 complex could be a connection between the ABA and JA signalling pathways. Further genetic and molecular studies are needed to explore this potential connection.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/genes13030524/s1, Figure S1: Characterisation of A. thaliana T-DNA insertion mutants cam1, cipk8, and gad1, Figure S2: Volcano plots of the total gene expression profiles of WT, PV-NES, and PV-NLS transgenic lines after treatment with ABA (A-C) or MeJA (D-F), Figure S3: Venn diagram identifying the unique and shared genes among PV-NES and PV-NLS transgenic lines after treatment with ABA or MeJA, Figure S4: Interactions of several [Ca 2+ ] cyt -regulated ABA-responsive genes with four hub genes in tobacco leaves, as determined by LCI assays, Figure S5: Interactions of two [Ca 2+ ] cyt -regulated JA-responsive genes with three hub genes in tobacco leaves, as determined by LCI assays, Figure S6: Interactions of two [Ca 2+ ] cyt -and [Ca 2+ ] nuc -regulated ABA-responsive genes with three hub genes in tobacco leaves, as determined by LCI assays, Figure S7: Interactions of two [Ca 2+ ] cyt -and [Ca 2+ ] nuc -regulated JA-responsive genes with three hub genes in tobacco leaves, as determined by LCI assays, Table S1: Primers used in this study, Table S2: [Ca 2+ ]-regulated ABAresponsive genes, Table S3: [Ca 2+ ]-regulated JA-responsive genes, Table S4: Microarray datasets used for the construction of co-expression networks, Table S5: Decoding Ca 2+ signalling proteins, Table S6: Hormone signalling pathway proteins, Table S7: Genes in the co-expression network, Table S8: Genes co-expressed with CAM1, CIPK8, GAD1, and GUN5.