Overlapping and differential roles of plasma membrane calcium ATPases in Arabidopsis growth and environmental responses

Four plasma membrane calcium ATPases in Arabidopsis, ACA8, ACA10, ACA12, and ACA13, have overlapping and differential roles in vegetative growth, reproductive development, stomatal movement control, and disease resistance.

Abstract

Plant cells have multiple plasma membrane (PM)-localized calcium ATPases (ACAs) pumping calcium ions out of the cytosol. Although the involvement of some of these ACAs in plant growth and immunity has been reported, their individual and combined functions have not been fully examined. Here, we analysed the effects of single and combined mutations of four ACA genes, ACA8, ACA10, ACA12, and ACA13, in a number of processes. We found that these four genes had both overlapping and differential involvements in vegetative growth, inflorescence growth, seeds setting, disease resistance and stomatal movement. Disruption of any of these four genes reduces seed setting, indicating their contribution to the overall fitness of the plants. While ACA10 and ACA8 play major roles in vegetative growth and immunity, ACA13 and ACA12 are also involved in these processes especially when the function of ACA10 and/or ACA8 is compromised. The loss of ACA13 and ACA10 function in combination with a reduction in function of ACA8 leads to seedling death at bolting, revealing the essential role of their collective function in plant growth. Taken together, this study indicates a highly tuned calcium system involving these PM-localized calcium pumps in plant growth and environmental responses.

Introduction

Calcium (Ca²⁺) is an essential second messenger for cellular signal transduction and has a wide range of physiological roles in response to various environmental stimuli such as light, temperature, and pathogens (Sanders et al., 2002; McAinsh and Pittman, 2009; Dodd et al., 2010; Kudla et al., 2010). It has been hypothesized that calcium carries information on the stimulus specifically through the amplitude, frequency, and duration of calcium spikes (Dodd et al., 2010; Bonza and De Michelis, 2011). This Ca²⁺ signature is shaped by the combination of activities of membrane transport proteins in Ca²⁺ influx and Ca²⁺ efflux systems. Calcium influx is mediated by a number of ion channels such as Ca²⁺-permeable cyclic nucleotide-gated channels and voltage-gated channels (Kudla et al., 2010). These channels reside at the plasma membrane (PM), endoplasmic reticulum (ER), vacuole, or mitochondria and are thought to collectively contribute to the dynamics of calcium signals (Geisler et al., 2000a; Sze et al., 2000; Boursiac and Harper, 2007). Ca²⁺ efflux requires cotransport systems and energy-dependent Ca²⁺ pumps such as the ER-type Ca²⁺-ATPases (ECAs) and the autoinhibited Ca²⁺-ATPases (ACAs) (Steinhorst and Kudla, 2013). Fourteen Ca²⁺ pumps have been found in Arabidopsis, ten of which are ACAs and four are ECAs. Based on the sequence similarity and intron positions of their genes, ACA proteins can be classified into four clusters that appear to be conserved in flowering plants (Baxter et al., 2003; Boursiac and Harper, 2007; Bonza and De Michelis, 2011). Cluster 1 consists of ER-localized ACA1 and ACA2 as well as PM-localized ACA7 (Hong et al., 1999; Dunkley et al., 2006; Lucca and León, 2012). Cluster 2 consists of vacuole-localized ACA4 and ACA11 (Geisler et al., 2000b; Boursiac et al., 2010). Cluster 3 consists of PM-localized ACA12 and ACA13, both of which are encoded by intron-less genes (Iwano et al., 2014; Limonta et al., 2014). Cluster 4 consists of PM-localized ACA8, ACA9, and ACA10, which are characterized by a slightly larger molecular mass mainly due to a longer N-terminal domain (Bonza et al., 2000; Schiøtt et al., 2004; George et al., 2008).

Ca²⁺ pumps are considered to be crucial in regulating the shape of the calcium transient, after cytosolic Ca²⁺ concentration is elevated by stimuli, during the recovery of basal cytosolic level (Dodd et al., 2010; Kudla et al., 2010; Steinhorst and Kudla, 2013; Costa et al., 2017). A number of plant ACA genes have been characterized as Ca²⁺ pumps by their expressions in the yeast mutant strain K616, which lacks endogenous Ca²⁺-ATPases (Curran et al., 2000; Bonza et al., 2004; Baekgaard et al., 2006). The N-terminus of several ACAs contains an auto-inhibitory domain that inhibits the activity of the ATPase domain to keep the pump activity low (Giacometti et al., 2011; Costa et al., 2017). This auto-inhibitory domain overlaps with the calmodulin (CaM)-binding motifs, and calmodulin binding is thought to release auto-inhibition and therefore activate calcium ATPase (Tidow et al., 2012; Costa et al., 2017). When the auto-inhibitory domain is deleted, ACA10 becames ‘deregulated’ and complements the calcium pump deficiency of K616 while the full-length ACA10 does not (Schiøtt and Palmgren, 2005). Similar pump activities have been found in yeasts for the deregulated (N-terminus deleted) versions of ACA8, ACA9, and ACA13 (Bonza et al., 2004; Schiøtt and Palmgren, 2005; Iwano et al., 2014). However, ACA12 appears to be a deregulated pump (Limonta et al., 2014); unlike other ACAs, it is not stimulated by CaM although it could bind to CaM, and a full-length ACA12 rescues the defect of K616. This is likely due to the lack of two amino acidic residues that are conserved in other subgroups of ACAs (Limonta et al., 2014).

ACA proteins localized on the PM are implicated in growth and development regulation in plants. Several PM-localized ACAs, ACA7, ACA9, and ACA13, have been shown to be important for pollen development and/or pollen function. The ACA9 gene is expressed specifically in pollen, and its loss-of-function (LOF) mutant displays reduced growth of pollen tubes and a high frequency of aborted fertilization (Schiøtt et al., 2004). The LOF mutant of ACA7 has dead pollen grains in mature flowers (Lucca and León, 2012). The ACA13 protein localizes to the PM and vesicles and accumulates at the pollen tube penetration site after pollination; its LOF mutant has a pollination defect (Iwano et al., 2014). These results indicate a critical role of Ca²⁺ efflux at the PM in reproductive growth. ACA10 and ACA8 are implicated in promoting growth of rosettes and inflorescences (Bonza et al., 2000; George et al., 2008; Yang et al., 2017). The ACA10 LOF mutant in the No-0 background has a compact inflorescence stem (George et al., 2008) and the aca10 aca8 double mutant has much reduced growth of the rosette and inflorescence (Yang et al., 2017).

PM-localized ACAs are also involved in plant stress response regulation. ACA8 was reported to be associated with FLS2, a receptor for the pathogen pattern flg22 (dit Frey et al., 2012). The aca10 LOF mutant in No-0 has higher resistance to the bacterial pathogen Pseudomonas syringae pv. tomato (Pst) DC3000, and so does the aca8 aca10 double mutant in Col-0 (Yang et al., 2017). Defense genes are up-regulated in these mutants even under non-infection conditions, and reducing the defense responses alleviates the growth defect in these mutants (Yang et al., 2017). Interestingly, ACAs on the vacuole are also implicated in plant immunity (Boursiac et al., 2010). The double LOF mutant of vacuole-localized ACA4 and ACA11 displays a high frequency of hypersensitive response-like lesions associated with up-regulation of the salicylic acid pathway and enhanced disease resistance (Geisler et al., 2000b; Boursiac et al., 2010). Stomatal movement in response to pathogen has also been shown to be modulated by PM-localized ACAs. The aca8 aca10 double mutant does not close its stomata in response to bacterial pathogen as the wild type does (Yang et al., 2017). In Arabidopsis leaves and roots, ACA8 is also involved in the response to wounding-related signals (Costa et al., 2017).

The role of ACAs in growth and immunity is likely due to their impact on the calcium signal or calcium level in cellular compartments. ACA8 and ACA10 are implicated in calcium signature generation. External calcium application induces cytosolic calcium oscillations in the wild type guard cells, and the loss of either ACA8 or ACA10 function abolishes the oscillations in the cytosol (Yang et al., 2017). In addition, the aca8 aca10 double mutant inoculated with the bacterial elicitor flagellin exhibits a lower cytosolic Ca²⁺ transient increase than the wild type (dit Frey et al., 2012).

Despite of these studies, we still do not fully understand the roles of PM-localized ACAs as a whole or individually even in the model plant Arabidopsis. These proteins might differ in their biochemical activities and regulation. ACA8, ACA10, and ACA13 are thought to have the auto-inhibitory domain while ACA12 likely does not (Iwano et al., 2014; Limonta et al., 2014). Potential genetic redundancy, full or partial, might mask the role of ACAs expressed in the same cell and the same compartment in single mutant studies. In addition, loss of one member’s function might be compensated by up-regulation of another member with similar expression pattern and subcellular localization. ACA12 and ACA13 were both found to have a higher expression in the aca8 aca10 mutant than in the wild type under flagellin treatment (dit Frey et al., 2012). Here we investigated the role of four PM-localized ACAs, namely ACA12 and ACA13 along with ACA8 and ACA10, in a number of processes. Analysis of mutant combinations of the four ACA genes reveals their overlapping and differential involvement in development and defense responses. These PM-localized calcium pumps may contribute to a finely tuned calcium signalling system in plant growth and immunity.

Materials and methods

Plants and growth conditions

Mutants of ACA8 (GK-688H09), ACA10 (GK-044H01), ACA12 (SALK_098383), and ACA13 (SAIL_878_B06) were as previously described (dit Frey et al., 2012; Iwano et al., 2014; Limonta et al., 2014). For generating mutant combinations, aca12 was crossed with aca13; aca12 and aca13 were crossed with aca8 aca10. Mutants aca12 aca13, aca8 aca12, aca8 aca13, aca10 aca12, aca10 aca13, aca8 aca10 aca12, aca8 aca10 aca13/+, and aca10 aca13 aca8/+ were isolated in the F2 populations by PCR (all primers used are summarized in Supplementary Table S1 at JXB online).

The Arabidopsis plants were grown in soil with light intensity at 100 μmol m⁻² s⁻¹ and relative humidity at 50–70%. Constant light conditions were used for growth phenotype analysis and gene expression analysis unless specified otherwise. Plants were grown under 12 h light/12 h dark for pathogen resistance assay and stomatal assay unless specified otherwise.

Hydroponic culture was performed as previously described with slight modification (Boursiac et al., 2010). Arabidopsis seeds were sown on plates on half-strength Murashige and Skoog medium complemented by 1% Suc and 7 g l⁻¹ agar. After 2 d in the dark at 4°C, seeds were germinated under 24 h light at 22°C. Five-day-old seedlings were transferred onto a floating foam support in 300 ml boxes covered by foil and filled with a standard hydroponic solution of 1.25 mM KNO₃, 0.75 mM MgSO₄, 1.5 mM Ca(NO₃)₂, 0.5 mM KH₂PO₄, 50 mM FeEDTA, 50 mM H₃BO₃, 12 mM MnSO₄, 0.7 mM CuSO₄, 1 mM ZnSO₄, 0.24 mM MoO₄Na₂, and 100 mM Na₂SiO₃. Hydroponic solutions were replaced every week. For suppression conditions, the standard hydroponic solution above was supplemented with 15 mM NH₄NO₃ to a final concentration of NO₃⁻ at 19.25 mM.

Quantitative real-time quantitative PCR and gene expression

Total RNA was extracted from soil-grown 15-day-old plants with TRIzol reagent (Invitrogen) according to the manufacturer’s protocol. The cDNAs were synthesized from total RNA using AffinityScript QPCR cDNA Synthesis Kit (Agilent Technologies). Real-time quantitative PCR was performed with the Bio-Rad PCR System using iQSYBR GREEN SuperMix (Bio-Rad). Quantitative real time PCR (qRT-PCR) was performed using primers PR1-QRT-F and PR1-QRT-R to test the expression of PR1, and primers ACT2 F and ACT2 R to amply ACTIN2. All primers are listed in Supplementary Table S1. The relative expression level was calculated by the ${\text{2}}_{}^{-\Delta \Delta C_{\text{T}}}$ method (Livak and Schmittgen, 2001) with three biological replicates. For each biological replicate, there were three technical replicates. ACTIN2 was used as a reference gene for qRT-PCR.

Pathogen resistance assay

The bacteria strain Pst DC3000 was grown for 2 d on King’s B medium and resuspended at 5 × 10⁶ colony forming units ml⁻¹ (OD₆₀₀=0.05) in a solution of 10 mM MgCl₂ and 0.02% (v/v) Silwet L-77. Two-week-old seedlings grown at 22°C or 12-day-old seedlings grown at 28°C were dipping-inoculated with bacterial solution and kept covered for 1 h. The amount of bacteria in the plants was analysed at 1 h and 3 d after dipping (0 and 3 d post-innoculation, respectively). The aerial parts of three inoculated seedlings were pooled as one sample, and three samples were collected for each genotype and time point. Seedlings were ground in 1 ml of 10 mM MgCl₂, and serial dilutions of the ground tissue were used to determine the number of colony forming units (log₁₀) per mg of fresh leaf tissues. For spray-inoculation, the resuspended bacteria were sprayed on plants until all leaves were wet.

Measurement of leaf physiological parameters

Rosettes from 35-day-old plants were weighed as biomass. Rosettes were also photographed from the top, and the diameter of the rosette was defined as the diameter of the smallest circle that covers the whole rosette using ImageJ software. Thirty plants were measured for each genotype/condition.

Stomatal closure assay

Stomatal closure assays were performed as previously described (Zeng et al., 2010; Gou et al., 2015) with slight modifications. Plants were grown under a 12/12 h photoperiod at 22°C for ABA- or Ca²⁺-induced stomatal closure assay. Leaves were collected at 4 weeks after germination and placed in MES buffer (10 mM MES–Tris, pH 6.15, 50 mM KCl) or MES buffer with 20 μM abscisic acid (ABA) or MES buffer with 100 μM CaCl₂ in closed Petri dishes for 1.5 h. Epidermises were peeled and imaged with a Lecia ICC50HD microscope. At least 30 stomatal apertures were measured for each sample from about five expanded leaves using ImageJ software. Each experiment was repeated at least three times.

Statistical analysis

Rosette morphology (weight and diameter), number of seeds in siliques, pathogen resistance assay and stomatal assay data were subjected to a one-way analysis of variance (ANOVA) followed by Duncan’s new multiple range test to assess differences between different genotypes. P was calculated at a 5% significance level to allow easy comparison of differences.

Results

Expression patterns of ACA8, ACA10, ACA12, and ACA13

ACA12 and ACA13 are the only two members in cluster 3, which is most closely related with cluster 4 that consists of PM-localized ACA8, ACA9, and ACA10 (George et al., 2008; Iwano et al., 2014) (Fig. 1A). We examined the tissue expression pattern of these PM ACAs through the public RNA-seq and microarray data (Winter et al., 2007). ACA10 and ACA8 have a higher expression than ACA12 and ACA13 in most tissues and developmental stages, including leaves, roots, young flowers (stages 9–11), and seeds (see Supplementary Table S2A). However, in the stamen at flower stage 15, ACA13 has a much higher expression than the other three ACAs (Supplementary Table S2A). In mature pollens, ACA8 has a slightly higher expression than the other three ACAs (Supplementary Table S2A). In guard cells, ACA12 has the highest expression and ACA8 has the lowest expression (Supplementary Table S2A). These genes also showed differential expression in response to biotic and abiotic stresses (Supplementary Table S2B, C). ACA10 was induced to a greater extent than ACA8 in response to Botrytis cinerea, Pseudomonas syringae, Phytophthora infestans, Erysiphe orontii, and pathogen pattern flg22 (Supplementary Table S2B). ACA12 was also highly induced by these pathogens or pathogen signals except for Erysiphe orontii, with the highest fold induction among the four genes, while no induction was seen for ACA13 except in response; to Pseudomonas syringae (Supplementary Table S2B). The expression of these four genes did not seem to be responsive to many abiotic stresses, except that ACA8 was induced by cold, ACA10 and ACA12 were induced by UV-B, and ACA13 was induced by osmotic stress (Supplementary Table S2C). Therefore, the tissue specificity pattern and environmental response of their RNA expressions are different among the four ACAs.

Fig. 1.

Expression patterns of ACA12 and ACA13. (A) Phylogenetic tree of ACAs modified from Iwano et al. (2014). ER, endoplasmic reticulum; PM, plasma membrane. (B, C) Expression of ACA13 (B) and ACA12 (C) analysed by qRT-PCR in leaves of wild-type Col-0, aca8, aca10, and aca8 aca10 at 2 weeks old grown at 22°C with 12 h light. Data are presented as means ±SD for three independent biological replicates. **Significant differences between Col-0, aca8, aca10 and aca8 aca10 at P<0.01 based on one-way ANOVA followed by Duncan’s new multiple range test.

ACA12 and ACA13 were postulated to compensate for the loss of ACA8 and ACA10 as their expressions after flg22 treatment were higher in the aca8 aca10 double mutant than the wild type (dit Frey et al., 2012). To examine this further, we analysed the expression of ACA12 and ACA13 in the single and double mutants of aca8 and aca10 under normal growth conditions without the treatment of the pathogen signal flg22. qRT-PCR revealed that ACA13 was up-regulated in leaves of the single aca8 and aca10 mutants compared with the wild type by ~2.2- and ~1.4-fold, respectively (Fig. 1B). The transcript abundance of ACA13 in leaves was increased further in the aca8 aca10 double mutant by ~3.0-fold in comparison to the wild type (Fig. 1B). The expression of ACA12 was slightly increased in the single and double aca8 and aca10 mutants by ~0.5-fold compared with the wild type (Fig. 1C), but this increase was not significant statisticallly. This expression pattern suggests that ACA13 and perhaps ACA12 might play compensatory roles for ACA8 and ACA10 even under non-pathogen-invasion conditions.

The role of ACAs in vegetative growth

To identify the biological roles of ACA12 and ACA13, we analysed mutants of ACA12 (SALK_098383) and ACA13 (SAIL_878_B06) from the T-DNA insertion line collections (Alonso et al., 2003; Iwano et al., 2014). These mutants are LOF because T-DNAs were inserted in the only exon and RT-PCR study revealed no expression of ACA12 and ACA13 genes in their respective mutants (Iwano et al., 2014).

To test the potential genetic redundancy among the four ACA genes, we generated all six double mutants among aca8, aca10, aca12, and aca13, namely aca8 aca10, aca12 aca13, aca8 aca12, aca8 aca13, aca10 aca12, and aca10 aca13. At the vegetative stage, none of the single mutants grew drastically differently from the wild type at 22 and 28°C (Fig. 2A, D; Supplementary Fig. S1A, B). Subtle differences were observed when the mutants were quantified by their size and weight at the 3-week-old stage grown under 22°C (Fig. 2B, C). For diameter of rosettes, none of the single mutants showed a significant difference compared with the wild type (Fig. 2B). For weight of whole plant, the aca10 mutant was lighter than the wild type at the vegetative stage (Fig. 2C). Therefore, the loss of ACA8, ACA12, or ACA13 alone did not cause obvious growth defects, suggesting a minor role or overlapping role of these two genes with others in the development of rosettes (see Supplementary Fig. S2A).

Fig. 2.

Growth phenotypes of the aca mutants in the vegetative stage. Shown are growth phenotypes under constant light at 22°C (A, B, C) and 28°C (D). (A) Rosette leaves of 18-day-old plant after germination at 22°C. (B, C) Quantification by diameter (mm) (B) and weight (C) of rosettes at 21 d after germination at 22°C. Data shown are means ±SD (n=24). Each colored line links a single aca mutant with double mutants containing that single mutation. Letters indicate statistically significant differences for different genotypes determined by one-way ANOVA (P<0.05) followed by Duncan’s new multiple range test. **Significant difference (P<0.01) between a double mutant and that single mutant, calculated from one-way ANOVA followed by Duncan’s new multiple range test . (D) Rosette leaves of 15-day seedlings at 28°C.

We then analysed the growth phenotypes of all these six double mutant combinations. The aca8 aca10 double mutant showed the largest visible growth defects compared with the wild type among the six double mutants. Young leaves of aca8 aca10 were more compact and the mature leaves were narrower with leaf blades twisted downward compared with the wild type at 22°C (Fig. 2A; Supplementary Fig. S1A). When quantified by the diameter of the rosette size at vegetative stage, aca10 aca12 and aca8 aca10 were significantly smaller than their respective single mutants (Fig. 2B). Therefore, double mutants containing aca10 exhibited the strongest growth defect, indicating a major role of ACA10 in rosette growth. When biomass was measured, four out of the six double mutants, aca8 aca10, aca12 aca13, aca8 aca12, and aca10 aca12, were all lighter than their respective single mutants (Fig. 2B, C). Only the double mutant aca8 aca13 did not exhibit significant differences from the wild type or single mutants (Fig. 2C). When plants grew at 28°C, all growth defects in the double mutants were abolished, except for aca8 aca10, which still showed a more compact phenotype at seedling stage (Fig. 2D; Supplementary Fig. S1B). Comparison of strength of defects in the double mutants indicates that ACA10 has a major role in vegetative growth followed by ACA8, and then ACA12 and ACA13 (see Supplementary Fig. S2A, B).

Because the aca8 aca10 double mutant exhibited the most severe defect among double mutant combinations, we examined the role of ACA12 and ACA13 in the aca8 aca10 double mutant background. By crossing aca12 and aca13 single mutants with the aca8 aca10 double mutant, we obtained the triple mutant aca8 aca10 aca12 but not aca8 aca10 aca13 due to seed abortion (see below). All the aca8 aca10 aca12, aca8 aca10 aca13/+, and aca8/+ aca10 aca13 triple mutants showed more severe growth defects than the aca8 aca10 double mutant at 22°C (Fig. 2A; Supplementary Fig. S1A). They had narrower leaves, much smaller rosette size, and less biomass (Fig. 2A–C). The aca8 aca10 aca12 and aca8 aca10 aca13/+ mutants exhibited a less severe growth defect at 28°C than at 22°C, and they had a similar morphology to the aca8 aca10 double mutant at 28°C (Fig. 2A, D; Supplementary Fig. S1). Therefore, the function of ACA12 and ACA13 was revealed in the aca8 aca10 double mutant background.

The role of ACAs in inflorescence growth

After bolting, these mutants exhibited growth defects of a similar extent to those of the early seedling stages. At 22°C, the aca10 mutant had the shortest inflorescence among the four single mutants and more axillary stems than the wild type, while single mutants of ACA8, ACA12, and ACA13 also exhibited obvious difference compared with the wild type (Fig. 3A, B). Overlapping functions were found for ACA10 with ACA8, ACA12, and ACA13 in inflorescence development as in vegetative rosette development (see Supplementary Fig. S2C). Each of the aca8 and aca12 single mutations enhanced the inflorescence growth defect of aca10, and the aca8 aca10 and aca12 aca10 double mutants had shorter inflorescence stems than aca10 (Fig. 3A, B). The aca13 mutation had the largest effect when combined with aca10, and the aca10 aca13 mutant had the most reduced inflorescence stem among all double mutants with aca10 (Fig. 3A, B), indicating the dominant role of ACA10 in inflorescence growth. All the double mutants had reduced inflorescence height compared with the wild type (Fig. 3A, B). The severity of the double mutants indicates that the major role of ACA10 is followed by that of ACA13, ACA8, and lastly ACA12 in this process (Supplementary Fig. S2C). The overlapping function was also evident in the triple mutants. The aca8 aca10 aca13/+ triple mutant showed a more severe phenotype than aca8 aca10, and the aca8/+ aca10 aca13 triple mutant never bolted and died at the wild type bolting stage (Fig. 3A, B).

Fig. 3.

Growth phenotypes of the aca mutants at the flowering stage. Shown are growth phenotypes under constant light at 22°C (A, B, D) and 28°C (C). (A) Inflorescence and rosette phenotypes of the wild type Col-0 and mutant plants at 30 d after germination at 22°C. (B) Inflorescent height at 30 d after germination. Data shown are means ±SD (n=6). Each colored line links a single aca mutant with double mutants containing that single mutation. Letters indicate statistically significant differences for different genotypes determined by one-way ANOVA (P<0.05) followed by Duncan’s new multiple range test. **Significant difference (P<0.01) between a double mutant and that single mutant, calculated from one-way ANOVA followed by Duncan’s new multiple range test. (C) Inflorescence and rosette phenotypes of the wild type Col-0 and mutant plants at 30 d after germination at 28°C. (D) Number of seeds per silique after pollination. Data shown are means ±SD (n=25). All annotations are the same as in (B).

The aca10 single mutant also showed reduced inflorescence height compared with the wild type when grown at 28°C, and a more severe defect was observed in the double mutants aca8 aca10, aca10 aca12 and aca10 aca13 (Fig. 3C). Especially, aca10 aca13 showed inflorescence internodes much shorter than those of the aca10 single mutant at 28°C, similar to that at 22°C (Fig. 3A, C). In addition, the aca8 aca12 and aca8 aca13 double mutants had a short inflorescence, while the aca8 aca10 aca12 and aca8 aca10 aca13/+ triple mutants had almost no noticeable inflorescence elongation (Fig. 3C). Together, these data indicate that ACA10 plays an important role in inflorescence development (see Supplementary Fig. S2C), and the function of ACA12 and ACA13 can be revealed in the aca8 aca10 double mutant background.

The role of ACAs in seed setting

Because ACA13 has a critical function for successful pollination (Iwano et al., 2014), we examined the roles of these four ACAs in reproductive growth by examining the number of seeds per silique of aca mutant combinations. For all single mutants, the number of seeds per silique was slightly but significantly less than that in the wild type (Fig. 3D), indicating that all four ACAs could impact seed setting. Among the double mutants, aca10 aca13 exhibited the most severe reduction of seed number, and had almost no seeds set in its siliques (Fig. 3D). This suggests that ACA13 and ACA10 play major roles in seed setting. ACA13 appears to play a larger role than ACA10, as the double mutant aca8 aca13 had less seed setting than aca8 aca10 (Fig. 3D). ACA12 had a minor role comapred with ACA8 as the aca8 double mutants had less seed setting than the respective aca12 doubles with aca13 or aca10 (Fig. 3D). These results indicate that all four ACAs have functions in reproduction. ACA13 has a major role and ACA10 and ACA8 have overlapping functions with ACA13 (see Supplementary Fig. S2D). Interestingly, the aca8 aca10 aca13/+ triple mutant had less seed production than aca8 aca10 (Fig. 3D), suggesting that seed setting is also very sensitive to the level of ACA activity.

The role of ACAs in disease resistance

To examine the possible roles of ACA12 and ACA13 in plant immunity, we analysed pathogen growth in the aca mutant plants. Inoculated by the dipping method, the aca10 mutant, but not other single mutants, supported significantly less growth of the virulent pathogen Pseudomonas syringe pv tomato (Pst) DC3000 compared with the wild type at 22°C (Fig. 4A), which is consistent with the previous report (Yang et al., 2017). We further analysed resistance of double mutants to the virulent pathogen Pst DC3000. Double mutants combined with aca10, including aca8 aca10 and aca10 aca13, exhibited an enhanced resistance similarly to the aca10 single mutants, and aca10aca12 had a slightly reduced resistance than the aca10 mutant (Fig. 4A). None of the other double mutants without aca10 had enhanced resistance compared with their respective single mutants, indicating a major role of ACA10 in plant immunity (Fig. 4A; Supplementary Fig. S2E). The role of ACA12 and ACA13 in immunity was revealed in the absence of both ACA10 and ACA8 as the aca8 aca10 aca12 and aca8 aca10 aca13/+ triple mutants showed even higher resistance to Pst DC3000 than the aca10 or aca8 aca10 mutants (Fig. 4A). This observation was corroborated by resistance assay using a different method, spray inoculation. After being sprayed with Pst DC3000, the wild type displayed yellow diseased areas on most of the true leaves besides yellowing of the cotyledons (Fig. 4B). In contrast, aca10 and aca8 aca10 displayed less yellowing in true leaves and aca8 aca10 aca12 had no yellowing in true leaves (Fig. 4B). Therefore, ACA12 and ACA13 could have overlapping functions with ACA8 in the absence of ACA10 in resistance against Pst DC3000 (see Supplementary Fig. S2E).

Fig. 4.

Plant immunity of the aca mutants. (A) Growth of Pst DC3000 in the wild type Col-0 and mutant plants at 0 and 3 d post-inoculation (dpi) at 22°C. Values represent means ±SD for three independent experiments (n=3). Letters indicate statistically significant differences for different genotypes determined by one-way ANOVA (P<0.05) followed by Duncan’s new multiple range test. FW, fresh weight. Each colored line links a single aca mutant with double mutants containing that single mutation. **Significant difference (P<0.01) between a double mutant and that single mutant, calculated from one-way ANOVA followed by Duncan’s new multiple range test. (B) Disease symptoms in the wild type and aca mutant plants at 22°C after spray inoculation with Pst DC3000. (C) Growth of Pst DC3000 in the wild type Col-0 and mutant plants at 0 and 3 d post-inoculation at 28°C. All annotations are the same as in (A). (D) qRT-PCR analysis of PR1 expression in the wild type Col-0 and mutant plants. Data are means ±SD for three independent replicates. Shown are data from one representative experiment. **Significant differences (P<0.01) between Col-0 and aca mutants based on Student’s t-test.

We subsequently analysed the pathogen growth at 28°C because disease resistance is often suppressed by a high growth temperature (Hua, 2013). Indeed, the increased resistance exhibited at 22°C was not observed for aca10 or its double mutants at 28°C, except for aca10 aca13 (Fig. 4C). Interestingly, the aca8 aca13 double mutant exhibited an enhanced resistance at 28°C compared with the wild type or the single mutants while it only exhibited a slightly enhanced resistance at 22°C compared with the single mutants (Fig. 4A, C). This suggests that ACA13 may have overlapping functions with ACA10 and ACA8 in repressing defense even at 28°C (see Supplementary Fig. S2F). This notion is supported by the enhanced resistance exhibited in the aca8 aca10 aca13/+ triple mutants compared with the aca8 aca10 double mutant (Fig. 4C).

We further analyzed the expression of a defense response gene, PR1, in the wild type and all mutant plants by qRT-PCR. Among the single mutants, PR1 was up-regulated only in the aca10 mutant that exhibited enhanced resistance (Fig. 4D). PR1 was also up-regulated in the aca8 aca10 and aca10aca13 double mutants, but not other double mutants (Fig. 4D). The highest PR1 expression was observed in aca8 aca10 aca12 and aca8 aca10 aca13/+ triple mutant plants (Fig. 4D). These results revealed a correlation of PR1 expression level in the absence of pathogen with the resistance level to the pathogen, indicating a constitutive defense response in some aca mutants that likely leads to the enhanced disease resistance.

The role of ACAs in stomatal movement

A previous study indicates that ACA8 and ACA10 are positive regulators of stomatal closure in response to Ca²⁺ and a coronatine-deficient (COR⁻) Pst DC3000 strain (Yang et al., 2017). We determined whether ACA12 or ACA13 is also involved in stomatal control. Incubation with 20 μM ABA or 100 μM Ca²⁺ induced stomatal closure in the wild type at 22°C (Fig. 5A), which was consistent with previous reports (Zeng et al., 2010; Zou et al., 2010). All single mutants were sensitive to ABA or Ca²⁺ in their stomatal closure response, similarly to the wild type (Fig. 5A). In contrast, ABA- and Ca²⁺-induced stomatal closure at 22°C was impaired in all aca8 double mutants, including aca8 aca10, aca8 aca12, and aca8 aca13 (Fig. 5A). These results suggest that ACA8 has a major role in stomatal response and has an overlapping function with ACA10, ACA12, and ACA13 (see Supplementary Fig. S2G). Not surprisingly, the aca8 aca10 aca12 and aca8 aca10 aca13/+ triple mutants also showed insensitivity to ABA and Ca²⁺ in the stomatal closure response (Fig. 5A). It was also noted that all mutants that were insensitive to ABA or Ca²⁺ in stomatal closure had smaller stomatal aperture compared with the wild type after being incubated with the opening buffer (Fig. 5A).

Fig. 5.

Stomatal closure responses in the aca mutants. (A, B) Stomatal apertures after ABA- and treatment at 22°C (A) and 28°C (B). Epidermal peels of the wild type Col-0 and mutant plants were pre-treated with opening buffer and stomatal apertures were measured before and after treatment with ABA and Ca²⁺. Each data point represents the mean ±SD (n=30). The experiments were repeated three times. Shown are data from one representative experiment. Each colored line links a single aca mutant with double mutants containing that single mutation. **Significant difference (P<0.01) between a double mutant and that single mutant, calculated from one-way ANOVA followed by Duncan’s new multiple range test.

When plants were grown at 28°C, the stomatal closure in response to ABA or Ca²⁺ was significantly disrupted only in aca8 aca10 aca12 and aca8 aca10 aca13/+ triple mutants, much fewer than at 22°C (Fig. 5B). The aca8 aca10 double mutant showed significant insensitivity to Ca²⁺ but not ABA (Fig. 5B). These three mutants, aca8 aca10, aca8 aca10 aca12, and aca8 aca10 aca13/+, exhibited smaller aperture after the opening buffer treatment at 28°C, similarly to 22°C (Fig. 5). Therefore, ABA or calcium insensitivity in the stomatal closure response at 22°C can be rescued by a temperature of 28°C in some mutants but not the more severe mutants such as aca8 aca10 aca12 and aca8 aca10 aca13/+ (see Supplementary Fig. S2H).

Effect of anion supplements on the phenotypes of the aca8 aca10 mutant

In addition to temperature that was shown to modulate the aca mutant phenotype, an anion supplement was shown to suppress hypersensitive response-like necrotic lesions exhibited by the double mutant of vacuole ACAs aca4 aca11 (Boursiac et al., 2010). We therefore tested whether the mutant phenotypes of PM-localized ACAs might be suppressed by a similar anion addition. We used a standard hydroponic solution with NH₄NO₃ at 4 mM or supplemented to 15 mM NH₄NO₃ to grow the aca8 aca10 double mutant. As reported previously (Boursiac et al., 2010), the aca4 aca11 mutant exhibited lesions in standard solution but not with 15 mM NH₄NO₃ (Fig. 6). Interestingly, the narrow leaf phenotype was suppressed in aca8 aca10 double mutants by the addition of NH₄NO₃ (Fig. 6). In addition, the rosette size of aca8 aca10 and aca4 aca11 was comparable to the wild type when grown with the supplemental NH₄NO₃ while they were much smaller in the standard solution.

Fig. 6.

Effects of nutritional supplements on growth of aca8 aca10 and aca4 aca11. (A, B) Fifteen-day-old seedlings grown in a hydroponic solution without or with supplemental 15 mM NH₄NO₃ after transfer from plate at 5 d of age; 4 mM NH₄NO₃: standard solution; 15 mM NH₄NO₃: standard solution supplemented with 15 mM NH₄NO₃.

We tested whether the anion supplement could reduce defense response defects in the aca8 aca10 mutant as observed in aca4 aca11. The aca8 aca10 mutant exhibited a higher PR1 expression, although not as high as aca4 aca11, when compared with the wild type (Fig. 7A). The NH₄NO₃ supplement significantly reduced the PR1 transcript level in all plant lines including the wild type, and both aca8 aca10 and aca4 aca11 had significantly more PR1 expression compared with the wild type under NH₄NO₃ supplement (Fig. 7A).

Fig. 7.

Effects of nutritional supplements on the immunity of Col-0, aca8 aca10, and aca4 aca11. (A) qRT-PCR analysis of PR1 expression in Col-0, aca8 aca10, and aca4 aca11 grown in two types of hydroponic solutions (4 mM NH₄NO₃: standard solution; 15 mM NH₄NO₃: standard solution supplemented with 15 mM NH₄NO₃). (B) Growth of Pst DC3000 in Col-0 and aca8 aca10 at 0 and 3 dpi (+NH₄NO₃: standard solution supplemented with 15 mM NH₄NO₃). Asterisks indicate significant growth difference between two solutions or different genotypes (*P<0.05; **P<0.01), calculated from one-way ANOVA followed by Duncan’s new multiple range test. Data are means ±SD for three independent replicates. Shown are data from one representative experiment of two biologically independent experiments.

We further examined the effect of the anion supplement on disease resistance in the aca8 aca10 mutant. Plants were grown in hydroponic solution with or without the NH₄NO₃ supplement for 4 weeks and dipping-inoculated with Pst DC3000. After 3 d of growth, pathogen growth was quantified in the wild type and the mutant. The aca8 aca10 mutant was slightly more resistant than the wild type to Pst DC3000 when grown in standard solution (Fig. 7B). With the anion supplement, resistance was greatly reduced in both the wild type and the aca8 aca10 mutant, and the aca8 aca10 mutant was more resistant than the wild type even under anion-supplemented condition (Fig. 7B).

Discussion

In this study, we analysed the contribution of four PM-localized Ca²⁺ ATPases, ACA8, ACA10, ACA12 and ACA13, to plant growth and immunity. Mutant combinations of their genes revealed overlapping functions between multiple ACA pairs, including ACA8–ACA10, ACA12–ACA13, ACA8–ACA12, ACA8–ACA13, ACA10–ACA12, and ACA10–ACA13 in plant growth, stomatal response, and resistance to bacterial pathogens (see Supplementary Fig. S2). They indicate not only the importance of each ACA in development and environmental responses but also a differential contribution of these ACAs in rosette growth, inflorescence stem elongation, seed setting, stomatal closure response, and disease resistance. We demonstrate for the first time that ACA12 and ACA13 play broader roles in development and environmental responses than previously thought. We found that ACA10 plays a major role in plant immunity, and ACA8, ACA12, and ACA13 also play roles in immunity in the absence of ACA10 (Supplementary Fig. S2E). In addition, ACA10 takes the largest role in vegetative growth, followed by ACA8 and then ACA12 and ACA13 (Supplementary Fig. S2A, B).

The developmental function of ACAs, especially ACA9 and ACA7, was demonstrated in pollen germination (Schiøtt et al., 2004; Lucca and León, 2012), and the stigmatic function of ACA13 was also demonstrated (Iwano et al., 2014). Here we find that the aca10 aca13 double mutant, but not any of the other double mutants, led to no seed production (Fig. 3D). This could be due to a failure in gamete formation, pollination, fertilization, and/or embryogenesis. Detailed anatomical and physiological analysis will reveal which processes and which cells these ACAs could play critical roles in. In any case, these data indicate an overlapping role of ACAs in one or several processes and an essential function of calcium signaling in reproductive growth.

ACA10 and ACA8 have, in general, a larger role in leaf growth and immunity than ACA12 and ACA13. This appears to largely result from differential expression pattern of these four genes. ACA8 and ACA10 expression is much higher than ACA12 and ACA13 in the vegetative rosette, which likely accounts for the severe growth defect of the aca8 aca10 double mutant compared with the aca12 aca13 double mutant (see Supplementary Table S2A). It is also interesting that ACA13 expression was up-regulated in the rosette when ACA10 and ACA8 are knocked out (Fig. 1B), suggesting that there could be compensation among the members. Indeed, the function of ACA12 and ACA13 became apparent when the ACA10 and ACA8 functions were knocked out. In pollen, ACA13 has a much higher expression than in leaves and roots, and the level is comparable to that of ACA10 (Supplementary Table S2A). This is correlated with the major role it plays among the four ACAs in seed production. In addition, expression of ACA12 has a higher induction in response to pathogens than the other three ACA genes, while ACA10 has a much higher expression overall than the others (Supplementary Table S2B), suggesting a possible fine tuning of biotic responses through regulation of ACA gene expression. Interestingly, none of the four genes has a drastic change in the expression in response to abiotic stress. The roles of these genes in abiotic stress responses are largely uncharacterized. When they are characterized, the dynamics of calcium signaling might be found to be different in biotic and abiotic responses.

The differential role of these ACAs might also result from their different biochemical properties (Bonza and De Michelis, 2011). Relative expression levels are not always correlated with the relative roles of these ACAs. ACA12 and ACA13 have much higher expression in guard cells than other tissues and their expression levels are comparable to or higher than ACA8 and ACA10 (Supplementary Table 1A), but ACA8 and ACA10 play a more critical role in guard cell response to calcium and ABA signals (Fig. 5). The regulation of ACA10 and ACA8 proteins is thought to be different from that of ACA12 and ACA13 proteins. The activities of ACA10 and ACA8 are calcium regulated through the auto-inhibitory domain that overlaps with the calmodulin binding motif (Tidow et al., 2012). These features confer regulation on ACA10 and ACA8 activities by calcium, CaM and the interacting protein BON1 (Yang et al., 2017). In contrast, the activity of ACA12 is deregulated, that is, it does not contain the auto-inhibitory domain (Limonta et al., 2014). Sequence alignment suggests that ACA13 is also likely to be deregulated (Limonta et al., 2014). The lower expression in the vegetative tissues of ACA12 and ACA13 might provide a low constitutive calcium pump activity irrespective of calcium status in the cell.

The aca10 aca13 aca8/+ triple mutants exhibited lethality at bolting (Fig. 2A). Two scenarios, alone or in combination, could contribute to the lethality. One is that ACAs have essential functions in cell physiology such as calcium homeostasis, and shoot apical meristem cells cannot transit from vegetative growth to reproductive growth (Bonza et al., 2016; Zhu, 2016). The other scenario is that the shoot apical meristem cells have heightened defense responses leading to cell death. This would be similar to the mutant of BON1 which interacts with and potentially activates ACA8 and ACA10 (Yang et al., 2017). The progressive loss of the BON1 family members results in progressively increased disease resistance and ultimately lethality that can be suppressed by inhibition of defense response up-regulation (Yang et al., 2006; Li et al., 2009).

Calcium signaling is critical for the closing and opening of stomata (Kudla et al., 2010; Zou et al., 2010). ACA8 and ACA10 have been shown to modulate stomatal closure in response to calcium, ABA, and pathogens (Yang et al., 2017). Here we revealed a role of both ACA12 and ACA13 in stomatal closure when the ACA8 function is abolished. Because ACA8 and ACA10 are important for calcium signature generation and calcium homeostasis (Yang et al., 2017), it is likely ACA12 and ACA13 also participate in generating the calcium signature and impact the steady level of calcium (Iwano et al., 2014; Limonta et al., 2014). Intriguingly, stomata of aca8 aca10, aca8 aca10 aca12, and aca8 aca10 aca13/+ mutants had a smaller aperture after treatment with opening buffer (Fig. 5), suggesting that they are defective in stomatal opening as well. These data suggest that the ACAs are important for both opening and closure responses in guard cells.

The growth defects observed in the ACA mutant combinations are likely partially due to up-regulation of defense responses. The reduced rosette size and inflorescence stem observed in multiple aca double mutants at 22°C are greatly reduced at 28°C (Figs 2, 3). Elevated temperature could inhibit disease resistance gene-mediated defense responses and the associated growth inhibition (Hua, 2013). It is possible that up-regulation of defense responses in the aca mutant combinations contributed greatly to the growth defects. This was demonstrated for the aca10 and aca8 aca10 mutants where inhibition of defense responses rescued their growth defect (Yang et al., 2017). The lesion phenotype of the aca4 aca11 double mutant could be suppressed by high concentrations of various anions, such as NO₃⁻ (Boursiac et al., 2010). Here we found that the narrow leaf phenotype of aca8 aca10 could also be largely suppressed by anion supplements (Fig. 6). Anion supplements reduced the accumulation of PR1 in all plants including Col-0, aca8 aca10, and aca4 aca11 (Fig. 7A). Despite the greatly reduced PR1 expression in general, the aca8 aca10 mutant still exhibited a higher resistance to Pst DC3000 compared with the wild type under standard hydroponic conditions or anion supplemented conditions (Fig. 7B). This suggests that anion suppression works downstream of the early signaling pathway and upstream of the transcriptional response of the PR1 gene. Therefore, perturbation of ACA8/10 function, similarly to ACA4/11 function, induces heightened disease resistance as well as up-regulation of PR1, which can be differentially influenced by the anion supplement.

Interestingly, we found that both the wild type and the mutants had higher PR1 expression under standard solution than solution supplemented with NH₄NO₃ (Fig. 7A). The NH₄NO₃ supplement also alleviated the growth defects but not the enhanced resistance in the aca mutants. Previous study showed that the anion supplements did not suppress the accumulation of reactive oxygen species and salicylic acid but prevented the accumulation of PR1 in the aca4 aca11 mutant (Boursiac et al., 2010). Therefore, anion balance likely has a large impact on the expression of PR1 even in the absence of pathogen invasion. It may also affect expression of other genes directly related to plant growth independent of disease resistance.

While we have evaluated the roles of four PM-localized ACA genes, ACA8, ACA10, ACA12, and ACA13, in a number of processes through genetic characterizations, how they exert their regulation on these processes is not yet determined. Future studies should reveal how they individually and combinatorically affect the basal levels of calcium in specific cell types and tissues as well as the kinetics of calcium transients in response to environmental changes. That knowledge will contribute to the ultimate understanding of the regulation of calcium signals to achieve a fine-tuned and highly adaptive calcium signaling system in plant growth and environmental responses.

Supplementary data

Supplementary data are available at JXB online.

Fig. S1. Growth phenotypes of the full set of aca mutants in the seedling stage.

Fig. S2. Diagram of genetic interaction of the four ACA genes.

Table S1. List of all oligonucleotides used in this study.

Table S2. Expression level of four PM ACA genes.