Abstract

External Ca2+ supplementation helps plants to recover from stress. This paper considers genetic methods for increasing Ca2+ to augment stress tolerance in plants and to increase their nutritional value. The transport of Ca2+ must be carefully controlled to minimize fluctuations in the cytosol while providing both structural support to new cell walls and membranes, and intracellular stores of Ca2+ for signaling. It is not clear how this is accomplished in meristems, which are remote from active transpiration—the driving force for Ca2+ movement into shoots. Meristems have high levels of calreticulin (CRT), which bind a 50-fold excess of Ca2+ and may facilitate Ca2+ transport between cells across plasmodesmatal ER. Transgenes based on the high-capacity Ca2+-binding C-domain of CRT1 have increased the total plant Ca2+ by 15%–25% and also increased the abiotic stress tolerance. These results are compared to the overexpression of sCAX1, which not only increased total Ca2+ up to 3-fold but also caused Ca2+ deficiency symptoms. Coexpression of sCAX1 and CRT1 resolved the symptoms and led to high levels of Ca2+ without Ca2+ supplementation. These results imply an important role for ER Ca2+ in stress tolerance and signaling and demonstrate the feasibility of using Ca2+-modulating proteins to enhance both agronomic and nutritional properties.

1. Introduction

Plants sense and respond to environmental stimuli using networks of sensors, second messengers, kinases, and transcription factors to regulate gene expression and adapt to the new conditions. Ca2+ is perhaps the best-known second messenger but is also required for proper cell wall structure and membrane integrity [1]. Although Ca2+ is present at relatively high concentrations (0.1–80 mM) in cell walls and organelles, cytoplasmic levels of Ca2+ are maintained at ~100 nM [24]. Signal transduction in plants requires the ability to mobilize and sequester Ca2+ from both internal and external Ca2+ stores. Because both deficiency and high concentrations of Ca2+ cause localized cell death, the transport of Ca2+ throughout the plant must be tightly regulated [5, 6].

Plants grown under Ca2+ deficient conditions are more susceptible to plant pathogens and show reduced growth of apical meristems, chlorotic leaves, and cell wall breakdown leading to softening of tissues [2]. But adding Ca2+ does more than just alleviate these symptoms, it bolsters plant growth by increasing root length and helps them to withstand or recover from stress [714]. Supplemental Ca2+ is also used to improve fruit characteristics and can function to delay ethylene-induced senescence [15]. This information is not new, a report in Science published over 40 years ago described the effect of 1 mM Ca2+ in preventing severe NaCl toxicity in beans [16].

The precise effects of extracellular Ca2+ on a plant system is likely to be complex, because Ca2+ has multiple roles, and because different plants show different responses to supplemental Ca2+. For example, in most plants extracellular Ca2+ reduces Na+ accumulation, which alleviates salt stress. But in some plants (such as maize), Na+ levels remain constant, but a beneficial effect on plant growth is still apparent [17]. Supplemental Ca2+ in a few plants, such as rice, has no apparent effect on salt tolerance [18] (see [19]). Supplementation with K+ has either no effect or is detrimental [20].

In a recent report, a solution of Ca2+ was found to be beneficial when sprayed directly onto the leaves of drought-stressed tea plants [21]. How does simply spraying Ca2+ onto leaves benefit plants? Why have not plants figured out how to increase their own stores, since Ca2+ is readily available in most environments? Alternatively, is this a part of what makes some plants “weedy”? Is there a barrier to the effective long-distance transport of Ca2+? Can we engineer a “work-around”, or alternative mechanism for Ca2+ transport, to help them recover from stress or even to prevent damage in the first place?

To begin to understand how extracellular Ca2+ benefits plants, it is necessary to understand more about the function and mobility of Ca2+ at both the cellular and the whole plant level. Once we understand the different roles of Ca2+, how Ca2+ is sequestered and transported within the plant, released for cellular signaling—and then rapidly sequestered away from detrimental interactions—then we can begin to think about revising or tailoring some of its pathways. This paper will provide a brief, whole-plant overview of Ca2+-regulated pathways and functions with the goal of identifying potential strategies for engineering additional Ca2+ ions into soluble plant reserves, so that they are readily available for signaling and growth. It is hoped that this approach can be part of a strategy to design more nutritional crop plants that are also more resilient to stress.

2. Ca2+ Stores and Signaling

It is commonly believed that Ca2+, one of the most abundant minerals in the earth, evolved as a signaling molecule because of the dual needs of the cell for soluble phosphate and Ca2+ and the propensity for the two to precipitate out as an insoluble salt [22]. Phosphate also plays a critical role in signal transduction, but its role as an energy intermediate requires a presence in the cytoplasm [23]. In plants, metabolic pathways that use ATP are found largely in the cytoplasm and are kept separate from Ca2+ stores, which are found primarily in the apoplast, vacuole, and endoplasmic reticulum (ER) and to a lesser extent in mitochondria, chloroplasts, and the nucleus [4]. In animal cells and early in the plant lineage, the ER was the major source of Ca2+, and its release was controlled by another second messenger, inositol (1,4,5) triphosphate (IP3), through activation of ER-localized IP3 receptors [24, 25]. Similar IP3 receptors have not been found in plants; however, the phosphoinositide pathway is conserved in plants [2629]. Members of the phosphoinositide signaling pathway show transcriptional regulation by environmental and developmental stimuli in Arabidopsis [30], and Ca2+ release by IP3 is conserved [31, 32].

In addition to the apoplast, the ER and vacuoles are the major and metabolically relevant sources of cellular Ca2+ [3335]. Cytosolic Ca2+ levels fluctuate and are controlled by a system of membrane-localized Ca2+ pumps and Ca2+ channels located in the plasmalemma, vacuole, and ER [4, 5, 36]. The electrochemical potential for Ca2+ to enter the cytoplasm, across the plasma membrane, was calculated by Spalding and Harper to be about −52 kJ/mol [22]. Therefore, Ca2+ can enter cells passively through ion channels but requires energy to be pumped out of the cytoplasm. Although energetically unfavorable, removal of Ca2+ is rapid and efficient, resulting in 1000-fold and higher [Ca2+] differences between the cytosol and surrounding organelles and apoplast [37].

Unlike animal systems, mutations in Ca2+ transport proteins often do not produce dramatic phenotypes [22], suggesting that plants are more tolerant of cytosolic Ca2+ or that they have overlapping and redundant systems. This has made it difficult to correlate electrophysiological experiments with genetics to identify exactly which Ca2+ channels function in signaling (or storage) and when. Ca2+ was shown to be released from the ER and possibly other membranes by cADP-ribose, an NAD+ metabolite, similar to what happens in animal cells, over a decade ago [34], however, it now seems clear that cyclic nucleotide-gated channels (CNGC) are found in the plasmalemma [48]. One of the few proteins that do have a phenotype, the phenotype of cngc2, is similar to cax1/cax3 (see Section 3) suggesting that it plays a major role in allowing nonsignaling Ca2+ entry into leaf cells [49]. There are 20 CNGC genes in Arabidopsis and an additional 20 genes that encode glutamate receptor-like channels (GLR), another type of Ca2+ channel found in the plasmalemma [48]. A third type of channel, the two-pore Ca2+ channel (TPC1), was first identified as a plasmalemma protein but is now known to be localized to the tonoplast membrane.

There are two major groups of proteins that function in Ca2+ removal from the cytoplasm [50]. Autoinhibitory Ca2+ ATPase (ACA) uses the energy of ATP to pump Ca2+ out of the cytoplasm and into organelles such as the vacuole and ER. The second group of proteins function as antiporters and are called CAtion eXchange proteins (CAX), found on the tonoplast membrane. CAX exchanges two protons for one Ca2+, using the energy of the proton gradient to dampen cytoplasmic Ca2+ signals [51].

2.1. Calcium Signatures

Cytoplasmic increases in Ca2+ in response to high concentrations of salt were noted at least 25 years ago in plants [52], but the specificity of Ca2+ signaling is still not well understood. There are two nonexclusive models for how Ca2+ functions as a second messenger. The Ca2+ signature model posits that information is encoded in the shape, duration, and frequency of Ca2+ transients and the diversity of cellular Ca2+ stores, all of which may facilitate the formation of microdomains that support and respond to localized Ca2+ changes [4, 53]. These localized changes are specific to the inducing stimulus and result in specific changes to Ca2+-modulated proteins and their targets [5, 39, 5456]. A second model suggests that Ca2+ transients function as a simple binary switch, either on or off, and it is the Ca2+ sensor (a Ca2+-modulated protein) that links different stimuli to the adaptive response [22].

The best-studied examples of Ca2+-mediated signal transduction include guard cell opening, nodulation, and tip growth of polarized structures such as pollen tubes [5766]. Specific Ca2+ signatures have also been reported, for example, in response to different chemicals in the root (aluminum, glutamic acid, and ATP [67]) and, at the whole plant level, in response to ozone [68]. Examples of other stimuli that cause transient increases in cytosolic Ca2+ concentrations include touch, cold shock, heat shock, oxidative stress, anoxia, hypo-osmotic shock, salinity, wounding, gravity, and pathogen infection [37, 56, 6981]. Developmental signals including fertilization, senescence, abscission, and ripening also involve Ca2+-regulated proteins [8288].

There is evidence for tissue-specific differences in Ca2+ flux in response to the same stimulus, for example, salt stress. Salt tolerance is a complex trait involving responses to cellular osmotic and ionic stresses and their consequent secondary stresses (e.g., oxidative stress) [89, 90]. Roots show a biphasic transient increase in cytosolic Ca2+ following exposure to acute salt stress [73]. In contrast to cold shock, which is restricted to areas near the root meristem, salt shock increases cytosolic Ca2+ along the entire root [91]. To distinguish tissue-specific differences in Ca2+ flux, different transgenic plants transformed with a gene encoding aequorin (a reporter gene for Ca2+) targeted to the cytoplasm of the epidermis, endodermis, or pericycle of Arabidopsis roots were used [73]. Prolonged oscillations in aequorin luminescence in the endodermis and pericycle occurred that were distinct from the epidermis [73]. This demonstrated that the same stimulus was transduced differently depending on the cell type, which could be due in part to the evolution of multiple family members in genes that transport Ca2+ (Section 2).

2.2. Calcium Sensors

Understanding the transduction of Ca2+ signatures has increased in the past decade due to rapid progress in deciphering the cellular network of Ca2+-responsive proteins. There are several families of Ca2+-binding proteins in plants [9295]. Proteins such as calmodulin, calcineurin B-like proteins (CBL), and Ca2+-dependent protein kinases (CDPK) “sense” Ca2+, having one or more EF-hand domains that bind Ca2+ with high affinity. The Arabidopsis genome encodes ~250 EF-hand containing proteins [96], although it should be noted that the presence of an EF-hand domain does not necessarily mean that a protein is activated by Ca2+ [97]. Calmodulins can interact with transcription factors, directly transducing Ca2+ signals into changes in gene expression [98103]. There is also evidence of Ca2+ signals within the nucleus, where CDPKs can phosphorylate and activate transcription factors [104, 105], and in the chloroplast [4, 106]. It is becoming clear that the cellular location of all parts of the signal transduction pathway plays an important role in proper signal transduction [105]. Sensors “relay” information from Ca2+ signatures (or the binary switch) into downstream events that include phosphorylation, changes in gene expression and protein-protein interactions [107]. The variety of Ca2+ binding proteins in plants suggests that intracellular Ca2+ levels, transport, release, and uptake are interdependent and tightly regulated [92].

2.3. CIPK/CBL Network

Batistic and Kudla [23] argue that a new system of Ca2+-regulated proteins has evolved to replace the IP3 receptor network as plants adapted to life on land. In Arabidopsis this system comprises 10 calcineurin B-like proteins (CBLs), which function as Ca2+ sensors, and 26 CBL-interacting protein kinases (CIPKs) [23]. Elegant experiments combining microscopy and biochemistry have been used to decipher the logistics of this pathway [108]. In addition to Ca2+ sensing, variations in both the cellular distribution and the interaction partners of members in this pathway contribute to an elaborate system capable of interpreting information from a variety of different stimuli [109]. To date, CBL/CIPK complexes have been shown to participate in the transduction of signals caused by the abiotic stress response, abscisic acid, potassium and nitrate uptake mechanisms, anaerobic response, cold, salt, sugar, cytokinin, and light [44, 47, 74, 110122].

Kudla’s group has demonstrated that CIPK6/CBL4 interactions can lead to relocation of the K+ channel, AKT2, from the ER membrane to the plasmalemma [113]. Two lipid modifications of CBL4, myristoylation and palmitoylation, are required for it to associate with the ER to begin the relocalization. CIPK6 serves as a scaffold in this process as phosphorylation is not required [113]. Lipid modifications are also required for CBL1 association with the plasmalemma, where it interacts with CIPK23 to activate a second K+ channel, AKT1 [124]. This interaction results in K+ uptake under low K+ conditions [124] while the CIPK6/CBL4 interaction is needed for normal growth [113].

There is indirect evidence for the role of the CBL/CIPK network in biotic stress as members of this family respond to salicylic acid [125]. CIPK6L was induced by Ca2+ in apples, and exogenous Ca2+ also induced both CIPK and CBL from pea [45, 125] and a CIPK from rice [122].

The overexpression of different CIPK/CBL proteins involved in abiotic stress has been shown to confer increased drought tolerance (Table 1). In addition to nutrient deprivation and abiotic stress, some CIPK/CBL members target particular developmental pathways during abiotic stress including root growth, pollination, and germination [47, 112, 126]. The impact of ectopic CIPK6 expression on root growth was shown to be mediated through auxin [44, 126]. Although CIPK6 expression was shown to confer tolerance to salt, the positive impact of its overexpression in Arabidopsis and tobacco on root growth suggests that those plants may also do well under water-limiting conditions. This is discussed in more details in Section 6.

2.4. Ca2+ Binding Proteins and Modulation of Ca2+ Stores

Suberization of the cell walls in the endodermis might prevent apoplastic Ca2+ from participating in cytosolic signaling events, because the deposition of the wax onto the cell walls would inhibit Ca2+ mobility. White and Knight used this insight to demonstrate that different stimuli do result in the cell accessing different stores of Ca2+ [91]. Transgenic plants that expressed apoaequorin only in the endodermis were used, and the root tips, which had different levels of suberization, were examined for luminescence in the presence of luciferin, which is directly proportional to the concentration of Ca2+. While salt stress resulted in the production of a continuous luminescent Ca2+ signal along the endodermis, cooling the roots produced a signal that was confined to a terminal 4-mm region of the root tip, where suberization was incomplete or lacking [91]. This was an elegant demonstration that signal propagation from salt and cooling require access to different Ca2+ stores. Moore et al. concluded that cytoplasmic signaling in response to salt stress utilized intracellular stores of Ca2+, although it is still not clear what part of the cell contained the store [91].

2.4.1. The Vacuole as a Ca2+ Store

Although a considerable amount of Ca2+ is present in the apoplast, the vacuole is the main storage organelle for Ca2+ within the plant cell. However, there is little direct evidence for the vacuole as a source of Ca2+ for signaling [4, 127, 128], although the identification of Ca2+ channels in the tonoplast membrane is not complete either. Furthermore, most of the Ca2+ in the vacuole is complexed with chelators such as malate, isocitrate, and citrate and is, therefore, not readily available for signaling [4].

There is evidence for an important role for the vacuole in depleting cytosolic Ca2+, which is critical for preventing association with phosphate and for shaping putative Ca2+ signatures. Using mathematical modeling, Bose et al. suggest that the activity of the known major Ca2+ efflux proteins (two members, each of the ACA and CAX gene families) is sufficient to describe a wide variety of Ca2+ signatures, including all of the current experimental results, without having to take into consideration how Ca2+ enters the cytosol [50]. Figure 1 shows a diagram of the major Ca2+ efflux proteins in a leaf cell.

Two vacuolar Ca2+ ATPases, ACA4 and ACA11, have been shown experimentally to be important for removing excess cytoplasmic Ca2+ [129]. When genes for both of these pumps were mutated, groups of cells in the mesophyll began undergoing programmed cell death (PCD). This phenotype requires salicylic acid, suggesting that the increased cytoplasmic Ca2+ by itself was not toxic [129]. It could be that PCD has the lowest threshold for sensing an activating cytoplasmic Ca2+ signal. While many stimuli could activate the release of Ca2+ into the cytoplasm (light, gravity, etc.), without appropriate dampening by ACAs the signal could spread to other parts of the cell to trigger unintended responses. It will be interesting to know if the propensity for cell death is an indirect effect of altered cytosolic Ca2+ on a PCD-related Ca2+ sensor, or if ACA4 and ACA11 are specifically involved in PCD.

2.4.2. The ER as a Ca2+ Store

The ER also contains high levels of Ca2+ and is an attractive candidate for storing signaling Ca2+ [130]. Calreticulin (CRT) is an ER luminal chaperone that has two Ca2+ binding domains. The P-domain contains a high affinity, EF hand-like structure that binds 1-2 moles of Ca2+ per mole protein [131]. The C-domain is the least conserved among organisms but contains a disproportionately high number of acidic amino acid residues that function to bind large amounts of Ca2+ with weak affinity. The C-domain has been estimated to bind 30–50 moles of Ca2+ per mole of protein [131]. Because of its low affinity, C-domain binding requires a relatively high concentration of Ca2+, such as in the ER. Although estimates are scarce, the concentration in the ER of pollen tubes has been estimated to be ~100–500 µM; about 1000-fold higher than in the cytoplasm [130]. The ER of animal cells contains ~1 mM Ca2+ but the concentration is nonuniform [132]. This is also likely to be true in plants due to the conservation of ER pumps and Ca2+ binding proteins such as CRT and Calnexin (CXN) [133]. CXN is a membrane-bound ER protein that functions with CRT and BiP (another chaperone) in glycosylation and quality control of ER proteins [133, 134].

Most plants have two forms of CRT [135, 136]. In Arabidopsis, CRT1a and CRT1b (also called CRT2) have the highest homology and form the first group, while CRT3, which is specifically needed for viral cell-to-cell movement [137], is in the second group. All three CRTs function as chaperones and play an important role in protein folding and glycosylation [136, 138141]. The C-domain of CRT3 is reduced in size compared to CRT1a and CRT1b, but was specifically required for proper folding of the brassinosteroid receptor, BRI1 [142]. All three CRTs have been implicated in innate immunity for proper folding of different receptor proteins [143148], and CRT1 appears to participate in signaling [149]. CRT has also been associated with increased tolerance to abiotic stress [147, 150].

CRT is highly expressed in meristematic and reproductive tissues. It shows lower expression associated with vascular tissue. CRT1 and CRT2 are largely coexpressed, except that CRT2 is high in senescing leaves, perhaps as a mechanism for retrieving Ca2+. CRT2 also shows guard cell -specific expression.

The ER also contains at least one Ca2+ ATPase, ACA2, that is activated by calmodulin and inhibited by a CDPK [151, 152]. Inhibition of an ER-type Ca2+ ATPase (ECA1) in pollen tubes decreased ER Ca2+ and inhibited pollen tube growth suggesting that the ER serves as a Ca2+ store for signaling [130]. In addition, mutants with 4-fold lower ECA1 activity showed poor growth on medium with low Ca2+ (0.2 mM versus 1.5 mM, normal) [153]. It is not clear why ECA1 is needed to pump Ca2+ into the ER under low Ca2+ conditions.

In animal cells, Ca2+ is constantly leaking out of the ER and constantly being pumped back in by SERCAs, membrane pumps that are similar to ECA [132]. But the major mechanism for ensuring adequate ER levels of Ca2+ is a specialized plasma membrane pump that responds only to low ER Ca2+. In a mechanism called store-operated Ca2+ entry [154], the pump (Orai1) forms a structure adjacent to an ER protein (STIM1) that contains an EF hand to sense ER Ca2+ levels. Together, they allow ER Ca2+ levels to be refilled [132]. It is not known if a similar mechanism could function in plants.

2.5. Ca2+ Transduction and Regulation of Gene Expression

How many genes and proteins are associated with Ca2+ regulation? In addition to the ~250 EF-hand containing proteins, ~700 are thought to be involved with Ca2+ signaling for Arabidopsis, according to proteomic data [155]. These proteins generate Ca2+ signatures and transduce the signal into changes in protein phosphorylation, protein localization, protein-protein interactions, and changes in gene expression. It is the latter that is most difficult to identify due to the difficulty in testing Ca2+ without other secondary effects that result when stimuli such as NaCl are used that also cause chemical and ionic perturbations of the system. Knight’s group addressed this by using an applied voltage to alter membrane permeability in combination with transgenic aequorin to monitor changes in cytoplasmic Ca2+ levels [156]. Conditions for a transient increase in cytoplasmic Ca2+ from less than 100 nM to almost 600 nM were established, and microarrays were used to profile genetic changes. A combination of transient and oscillating Ca2+ fluxes produced the greatest number of genes (269) with increased expression levels, while a single long increase in Ca2+ to 200 nM produced only 10 genes with increased expression.

Analysis of the promoter regions of the Ca2+-upregulated genes revealed a surprising bias for genes that respond to abiotic stress. Three out of the four Ca2+-regulated promoter motifs were previously identified as being important for abiotic stress responses and included the ABA-response element and the drought-responsive element [156]. This bias could be due to the nature of the Ca2+ flux, which may have resembled signatures produced from an apoplastic source of Ca2+, or could be a feature of Ca2+ regulation.

In addition to the cytoplasm, transient Ca2+ fluctuations have also been reported in the nucleus, chloroplast, mitochondrion, and peroxisome [4]. Ca2+ oscillations in the cytosol and chloroplast have been linked to circadian rhythms [32, 157]. It is not known whether these fluctuations also lead to changes in gene expression.

We used a genetic method to specifically increase Ca2+ in the ER by taking advantage of the high capacity, low affinity Ca2+ binding activity of the C-domain from CRT. A green fluorescent protein-calcium binding peptide (GFP-CBP) fusion protein consisting of the C-domain from Zea mays CRT1 was fused to the C-terminal region of GFP [158]. The GFP-CBP construct included a signal protein for ER-targeting and the C-terminal region of CRT1, which contains an HDEL sequence for ER retention. Total Ca2+ in seedling shoots was increased by ~25%, when GFP-CBP was expressed in Arabidopsis using a constitutive promoter. Microarray analysis of seedlings expressing GFP-CBP compared to seedlings expressing GFP showed that 31 genes were upregulated by >3.5-fold. As expected, none of these genes included the cytosolic Ca2+-regulated genes identified by Whalley et al. Only one of the genes was involved in Ca2+ regulation—CIPK6 [158]. Whalley et al. also identified a single CIPK, CIPK9 [156]. The other genes we found were enriched for microsome-associated proteins and glycine-rich proteins, which are often targeted to the cell wall [158]. One of the proteins encoded a subunit of the anaphase-promoting complex [158]. This expression pattern could indicate a regulatory role for ER Ca2+ levels in mitosis. We will come back to this in Section 5.2.

Of course steady-state modulation of Ca2+ levels in an organelle is quite different from generating a cytosolic Ca2+ signal. According to the eFP browser [159], CIPK6 is induced by salt, drought, and abscisic acid and is expressed at a low level in guard cells, leaves, flowers, and developing fruit and seed. Although some of the genes coexpressed with CIPK6 in the GFP-CBP plants showed similar expression profiles to CIPK6, there is nothing to suggest a connection with ER Ca2+.

2.6. Summary of Cellular Ca2+ Dynamics

Cells contain stores of Ca2+ in the apoplast and in various compartments within the cell. Cytoplasmic Ca2+ is kept low to prevent interference with phosphate-containing pathways. Signal transduction uses discrete Ca2+ fluxes to connect stimuli with adaptive responses. Different stores of Ca2+ are used in the generation of these fluxes and the location, magnitude, and duration of the fluxes appear to contain information for the appropriate response. Vacuolar pumps and antiporters participate in removing Ca2+ from the cytoplasm before deleterious interactions occur. It has been difficult to determine which intracellular stores participate in different kinds of signaling, but the ER is an attractive candidate because of its distribution throughout the cell, and the ability of CRT to bind large quantities of Ca2+ with low affinity.

We still need more information on the plant’s ability to generate stimulus-specific Ca2+ signatures. What is the source of the Ca2+ used for different signals? What dampens the signature? How is information about the signal (magnitude, oscillations, and duration) transduced into specific responses? With respect to the original question—what, exactly, could the presence of supplemental Ca2+ contribute to increase stress tolerance? Are certain Ca2+ stores normally limited, or does spraying Ca2+ onto a plant trigger oscillations as Ca2+ is assimilated? Understanding Ca2+-regulated networks is plagued by the ubiquity of the molecule, and dissecting pathways in different cells and tissues is still tedious and difficult. However, the combination of biochemistry, Ca2+ reporter genes, and genetics is providing tremendous information that is building a solid foundation for understanding Ca2+ regulation.

The next section begins to discuss tissue-specific differences in Ca2+ levels to better understand how exogenous Ca2+ is assimilated.

3. Calcium Distribution within the Leaf

Eating roots and leaves is the best way for vegans (people who do not eat meat, fish, or dairy products) to increase Ca2+ intake [160]. This makes sense because Ca2+ is transported from roots to shoots through transpiration, and leaves carry out the bulk of transpiration. But not all cells within a leaf have equivalent Ca2+ levels. In grasses, Ca2+ is found mainly in the upper epidermis [161]. In dicots, Ca2+ levels are low in both upper and lower epidermis, but are higher in mesophyll, a distribution that facilitates Ca2+ control over stomatal aperture [161, 162].

A landmark study looked at the distribution of Ca2+ in different cell types of the leaf and found that mesophyll cells have ~6-fold more Ca2+ than epidermal cells, due largely to the differential expression of CAX1 in those cells [162]. CAX1 is located on the tonoplast membrane and couples proton export with Ca2+ transport into the vacuole. cax1/3 double mutants not only had reduced growth, reduced photosynthesis, and thicker cell walls, but also had higher apoplastic levels of Ca2+ [162]. This resulted in reduced stomatal apertures, which led to reduced growth due to a lack of carbon assimilation compared to nonmutant lines [162]. Although the cell walls were thicker, they were also more brittle and contained more pectin. Supplementation with low Ca2+ media reduced free apoplastic Ca2+ levels and suppressed the phenotype, while returning the plants to normal Ca2+ caused the phenotype to return. Free Ca2+ (sorbitol-exchangeable) was ~3-fold higher in the apoplast of cax1/3 double mutants compared to the nonmutant line. In fact, CAX1, CAX3, CAX4, and ACA4 (encoding a Ca2+ ATPase) and ACA11 are coregulated to make sure total Ca2+ levels are constant [162].

Why was high apoplastic Ca2+ a problem? Guard cells use Ca2+ to signal downstream components to close or open stomata. In the presence of excess Ca2+, stomata remain closed even under conditions favorable for gas exchange and carbon fixation. The exact mechanism for how extracellular Ca2+ interferes with guard cell signaling is not known. As mentioned in Section 2, the electrochemical gradient for Ca2+ across the cell membrane strongly favors passive Ca2+ entry—it is the removal of Ca2+ from the cytoplasm that requires energy. Thus, the presence of high levels of free Ca2+ on the other side of the plasmalemma may either make it difficult to remove Ca2+ from the cytoplasm or make it too easy for Ca2+ to enter it. Extracellular Ca2+ has been shown to cause guard cells to close by generating H2O2 and NO, which generate an intracellular Ca2+ spike, leading to stomatal closure [63].

Thus, keeping free Ca2+ out of the apoplast enables proper guard cell function and allows normal plant growth. CAX1 keeps apoplastic Ca2+ low by storing it in the vacuole [162]. Rather than viewing the apoplast as a separate entity that protected plant cells from extracellular threats, it now seems important to acknowledge that unbound extracellular Ca2+ must be maintained in equilibrium across the apoplast/symplast boundary. At least in leaves, it is the vacuole, a membrane-bound organelle on the symplastic side of the divide, not the cell wall, that serves as the reservoir for excess accumulation of Ca2+.

Where does Ca2+ come from? In leaf cells, Ca2+ is transported through the xylem by transpiration [2]. Ca2+ is one of the most immobile ions in the plant, with Mg2+ and Mn2+ not far behind [2]. In the leaf, Ca2+ is thought to diffuse through the apoplast up to about 15 cells away from the xylem. Transpiration would seem to direct Ca2+ to guard cells, which are mostly on the lower side of leaves, but the pattern of veins, anatomy of the leaf, and presence of air spaces all help to dissipate the pattern of water flow [163].

The pattern of Ca2+ transport is thought to vary with the developmental stage of the leaf, the species, and environmental conditions [163]. In eudicots, Ca2+ is trapped in the vacuoles of mesophyll cells by CAX1 [163], while in monocots higher relative levels of Ca2+ are found in the epidermis [2, 123]. Root pressure can contribute to the transport of Ca2+, especially when humidity is high and transpiration low [164]. Ca2+ deficiency is first noticed as tip burn, and diseases such as blossom end rot in tomato are a visual demonstration of the limited mobility of Ca2+. Since leaves develop acropetally, the apex is the last to differentiate. This suggests that dividing cells may be particularly vulnerable to Ca2+ depletion. We will come back to this in Section 5.

4. Ca2+ Is Transported from the Roots to the Shoot by Transpiration through the Xylem

There could be three points of control for transpiration—uptake in the root apoplast, entry into the xylem across the endodermis, and exit through guard cells. The apoplast shows very little electrical resistance and allows the free exchange of most ions. Ca2+ is absorbed from the soil by the apoplast and by cation channels in the root epidermis [165]. The extent of symplastic transport of Ca2+ between cells is not known, although a cadmium resistant channel was recently identified that facilitates radial movement of Ca2+ in roots [166].

Two pathways for Ca2+ transport to the shoot can be experimentally tested, a symplastic or cell-to-cell pathway and an apoplastic pathway. The symplastic pathway involves passage through at least one membrane. The Casparian strip of the endodermis, which contains suberin, restricts solute passage through the apoplast, and promotes passage through the symplastic pathway. Studies with radio-labeled Ca2+ suggest that this pathway predominates in onion [6]. Identification of enhanced suberin (esb) mutants in Arabidopsis allowed the role of the endodermis to be directly tested [167]. Shoot Ca2+ levels decreased ~50% compared to wild type. If there was no change, it could be concluded that transport was entirely apoplastic or entirely symplastic. So the reduction in Ca2+ transport suggests that restriction by the Casparian strip of the endodermis is incomplete—some apoplastic flow is permitted through the Casparian strip in its wild type state. There was no change in Mg2+ in the esb mutants, which is also transported through the phloem, but Zn2+ and Mn2+ also decreased [167]. Surprisingly, accumulation of the monovalent ions Na+, S+, and K+ increased. Transpiration was also decreased and the plants were less susceptible to wilting.

The existence of the apoplastic pathway was demonstrated from experiments that showed that the ratios of Ca2+, Br2+, and Sr2+ do not change after they are applied to roots, although channels and pumps have a clear preference for Ca2+ [168]. In many plants, the amount of Ca2+ transported depends on the rate of transpiration, which is consistent with solvent drag, not symplastic processes [168]. In some plants under certain conditions, Ca2+ transport may be almost entirely apoplastic with channels at the destination cell controlling cellular Ca2+ entry, followed by rapid assimilation into different organelles by pumps and antiporters. Ca2+ transport through the endodermal cytosol in the symplastic pathway is thought to be achieved using Ca2+ channels and pumps, but must be carefully regulated to avoid interfering with signaling pathways. According to White, apoplastic transport may be necessary to meet the demand for adequate Ca2+ in the shoot [168]. Breaks in the endodermis, for example where lateral roots emerge, allow Ca2+ transport without an intervening symplastic step.

Transpiration is considered to be the driving force for Ca2+ transport into shoots and leaves, and Ca2+ travels with the bulk water flow [2, 6, 163, 169, 170]. The pattern of Ca2+ deficiency symptoms can be explained by a combination between demand for Ca2+ and variation in transpiration. Tip burn, which affects the leaf margin and the undeveloped distal region of the leaf, is thought to result from a lack of well-developed veins in the undifferentiated part of the leaf and high rates of cell wall deposition.

Recent experiments actually compared the shoot accumulation of several minerals in members of 7 different plant families grown together under different fertilizer regimes [123]. The correlation with phylogeny (versus fertilizer treatment or residual) was the strongest for Ca2+ (70%) and total Ca2+ varied over 5-fold (Table 2). In contrast, Mg (with a 32.8% correlation with phylogeny) showed little more than a 2-fold variation. Dicotyledonous plants are known to accumulate more Ca2+ than monocots, partly as a function of the structure of their cell walls, and there only was ~3-fold variation in Ca2+ in different dicot families (Table 2). To put this in perspective, there was a ~2-fold variation in Ca2+ among Arabidopsis ecotypes, which are all members of the same species [171]. The molecular basis for the difference in Ca2+ levels between different families is not known, but the data suggest that factors are at play that ultimately limit the amount of Ca2+ absorbed from the soil.

The endodermis clearly has a role in regulating water transport, and likely helps the plant to conserve water by preventing unrestricted transpiration. Gilliham et al. argue that Ca2+ transport and transpiration are linked—Ca2+ regulates both stomatal activity in leaves and aquaporin (water channel) density and function in roots [163]. Thus, Ca2+ could increase its own transport by affecting aquaporin function [14]. Global mechanisms such as this may also play a role in limiting the amount of Ca2+ that ultimately reaches the shoot. In support of this, the overexpression of an aquaporin in Arabidopsis increased Ca2+ levels by ~33% under normal conditions and almost doubled Ca2+ under 100 mM NaCl [172]. The regulation of hydraulic conductivity (aquaporin function) under stress is reviewed by Aroca et al. [173].

A second mechanism for Ca2+ regulation of Ca2+ leaf concentration has been proposed [174]. A plasma membrane-localized CAlcium Sensing receptor, CAS, is upregulated in guard cells. High levels of apoplastic Ca2+ cause stomata to close, a process that requires CAS. When transpiration levels are high, Ca2+ has the potential to be too high. CAS mutants grown in soil had ~40% more Ca2+ than wild type plants [174]. Together with the aquaporin overexpression [172], this suggests that global regulation of Ca2+ levels occurs primarily through mechanisms found in the shoot, not through the endodermis in the root.

5. An ER Ca2+ Network for Meristems

Meristems are critically important for plant growth and reproduction. Meristems require high amounts of Ca2+ because of cell wall deposition and organelle biogenesis, but it is not clear how Ca2+ moves from areas with high rates of transpiration (leaves) into the protected region of the meristem (Figure 2). An alternative mechanism for Ca2+ transport is through the endoplasmic reticulum (ER). The ER is contiguous with the nuclear envelope and forms a symplastic continuum throughout the plant by spanning cell walls through plasmodesmata. Consistent with the idea of CRT as a Ca2+ transporter/regulator, high levels of CRT are found in plasmodesmata [175, 176] and in meristems [177]. This may be especially important in meristems, where the need for Ca2+ is high due to the formation of new cell walls, but the ability to transpire Ca2+ is limited by the lack of differentiated xylem. Transport through the ER would avoid the problem of cytoplasmic transit disrupting signaling pathways and could either augment apoplastic transport to ensure the protection of developing areas of the plant or bypass it, depending on where Ca2+ enters the ER.

If the ER functions in Ca2+ transport, why has not this been detected in leaves? A key aspect of the proposed Ca2+ network in meristems is the presence of CRT, whose gene shows high expression in meristematic tissues [177]. As described in Section 2.4.2, CRT has three conserved domains, one of which binds 30–50 Ca2+ ions with low affinity (the C-domain). CRT may function in intercellular Ca2+ distribution by acting as a buffer, partly neutralizing the charge. CRT is further proposed here to act as a sort of matrix to facilitate Ca2+ absorption and movement by the cell and to provide a gradient for additional Ca2+ to be transported cell-to-cell from mature tissues. But because CRT is not expressed at high levels in mature leaves, Ca2+ transport appears to follow a bulk flow pattern of distribution with the rate of transpiration dictating where it accumulates.

5.1. Desmotubules Allow Movement through the Plasmodesmata

Cytoplasmic Ca2+ transients have been demonstrated to result in rapid closure of plasmodesmata [178]. The biggest obstacle to Ca2+ transport through an ER network is the plasmodesmata. Plasmodesmata consist of a central desmotubule (see Figure 1), which is derived from the compaction of the two sides of the ER tubule that traverses the cell wall. A thin cytoplasmic sleeve that lies between the desmotubule and the plasma membrane serves as the conduit for cytoplasmic proteins and solutes that show intercellular trafficking [175].

CRT has been localized to plasmodesmata [175, 176] and could serve as a Ca2+ donor to maintain an internal network of stored Ca2+. High concentrations of CRT on either side of the plasmodesmata may result in a Ca2+ gradient, which could facilitate the distribution of Ca2+ to adjacent cells. Any cytoplasmic Ca2+ transients would occur independently of luminal concentrations [178].

Despite the narrow aperture of the desmotubule, transit of fluorescent molecules across the desmotubule appeared to be rapid. Microinjection studies were used to study the spread of the small molecular weight fluorescent tracers carboxyfluorescein and FITC-conjugated triglutamic acid in epidermal cells of tobacco and Torenia [179]. About 10% of the injections resulted in a punctate pattern of label that corresponded to the pattern obtained with DiOC6, a fluorescent dye that labels ER. This was explained by insertion of the needle into the lumen of the ER. In each case, the fluorescent molecules rapidly spread into adjacent cells through the desmotubule of the plasmodesmata. Spread of the fluorescence was more rapid through the desmotubule than through the cytoplasmic sleeve of the plasmodesmata and occurred more readily (100% of the cases versus ~88% for injections into the cytoplasm) [179]. Fluorescent dextrans corresponding to 10 kDa showed luminal transport in Torenia in 3 out of 3 injections. This demonstrates that sufficient space exists within the desmotubule for cell-to-cell Ca2+ transport. Although movement of ER-targeted GFP through the desmotubule was not demonstrated, Martens et al. discuss the possibility of the desmotubule functioning both as a conduit for cell-to-cell Ca2+ transport and as a mechanism for whole-plant signaling [180].

GFP fusions have also been used to study intercellular trafficking in leaf epidermal cells following microinjection. A CRT-GFP fusion protein in the ER lumen did not traffic into adjacent cells, but calnexin-GFP, an ER membrane-localized protein, did spread cell to cell [181]. CXN also binds Ca2+ and functions with CRT as a protein chaperone. It contains an N-terminal Ca2+-binding domain on the luminal side and an acidic tail of ~90 amino acids. These characteristics could enable it to transport Ca2+ across the plasmodesmata.

Why would transport through the desmotubule be needed? Plasmodesmata are regulated by Ca2+. When a cold shock was used to increase cytoplasmic Ca2+ from 100 to 200 mM, there was a 4-fold increase in resistance, but the resistance returned to normal within 10 sec [182]. Thus, cytoplasmic Ca2+ transients would be expected to close plasmodesmata. By compartmentalizing Ca2+ away from the cytoplasm, it could equilibrate between cells at levels that would interfere with plasmodesmata function if it were on the cytosolic side of the plasmodesmata.

5.2. Ca2+ and Cell Division

Vascular tissue forms de-novo and differentiates acropetally (phloem) and basipetally (xylem) in developing leaves after they have begun to expand and differentiate. The leaf midvein does not connect to the stem until after xylem and phloem have differentiated, and the leaves have begun to actively photosynthesize. The high rates of cell division in developing leaf primordia require significant amounts of Ca2+ to bind to cell wall pectin, stabilize the plasma membrane, and ensure completion of mitosis.

Ca2+ plays a major role in mitosis at anaphase, where it concentrates at the spindle poles at levels that cause microtubule depolymerization [183]. Interestingly, two proteins, one of them a CRT-like protein, have been identified in plants that could facilitate this process. Tonsoku (TSK) localized to the nucleoplasm while tonsoku-associated protein (TSA) has a signal peptide and was found in cytoplasmic vesicles derived from the ER [184]. During anaphase, the two proteins colocalized and appeared to interact. TSA has 10 repeats of an EFE motif consisting of acidic amino acids and was shown to bind Ca2+ in vitro. Although there was no homology with CRT, it may have a very similar function—to provide a matrix for storing Ca2+ until it is needed. Although a function has not been reported for these proteins, other than to bind Ca2+ and colocalize, it seems possible that they would be needed, along with kinesins [185], for depolymerization of microtubules during anaphase.

When plant cells enter prophase, the nuclear envelope (which is contiguous with the ER) disintegrates into vesicles. Following anaphase, a new cell wall is deposited, which requires vesicle secretion and membrane fusion. The ER is well positioned to provide Ca2+ during this process, which would be needed for stabilizing the developing cell wall by binding to pectate. The dynamic nature of cell-to-cell movement through desmotubules could ensure that the ER has a ready supply of Ca2+ available for the new cell wall that could be delivered through the vesicles.

CRT is known to be expressed at high levels in dividing cells, but the reason for this has not been obvious. Clearly, there is a higher need for glycosylated proteins as new cells are formed, but it appears to play more than a structural role, as overexpression of CRT has been shown to increase regeneration [186]. One possibility is that cell division, and possibly regeneration, may have become linked to the expression of CRT, such that if ER Ca2+ levels were not adequate to support new growth, the process of cell division would arrest. Our microarray results provide some tantalizing evidence in favor of this hypothesis. We found that GFP-CBP caused a 3.7-fold increase in the expression of At5g26635, which encodes one subunit of a putative anaphase-promoting complex. However, this is probably too late in the cell cycle to arrest development. A more likely explanation is that ER Ca2+ levels need to be high enough to facilitate the depolymerization of microtubules.

Nevertheless, the relationship between ER Ca2+, CRT expression, and mitotic activity would be interesting to study—to determine why “meristem burn” is not a problem, for example. It would also be interesting to examine CRT expression in tomato, since it suffers from the occurrence of blossom end rot, discussed in Section 6. Blossom end rot is a Ca2+-related disorder that results in tissue softening and necrosis at the distal end of the tomato fruit, which contains the highest proportion of dividing cells. Tomato is not the only fruit to undergo extensive cell division during fruit development (papayas, watermelons, and jack fruit are also quite large); what makes it more susceptible?

5.3. Summary

In summary, a gradient of Ca2+ ions in the ER is proposed to help the plant guard its vulnerable meristem from fluctuations in the transpiration of Ca2+. CRT networks in the ER could provide a conduit for Ca2+ transport to ensure that adequate levels of Ca2+ reach the meristem to support growth. CRT could serve as a buffer to help neutralize charge and to draw Ca2+ towards the meristem. Cell division may be coupled to CRT expression to ensure that adequate levels of Ca2+ are present when the cell divides. Transport through the ER would avoid competition with the vacuole and protect the cytoplasm while ensuring that enough Ca2+ is transported to meet the demands of the cell wall and organelles.

6. Genetic Manipulation of Ca2+ Stores

Many postmenopausal women take supplemental Ca2+ to help prevent osteoporosis, a crippling disease related to aging. With the demographics of most developed countries showing a rise in the aging population, the impact of nutrient deficiencies on human health is likely to increase. Many people do not like to take Ca2+ in the form of a pill because of its large size, which is needed due to the relatively poor absorption of chemical Ca2+. The best way to obtain more nutrients is to consume more fruits and vegetables, especially roots and leaves for Ca2+, Mg2+, and K+ [160, 187]. Unfortunately, almost 10% of the adult population of the USA and UK are deficient for those three elements [160], due, in part, to consumption of cereal grains rather than vegetables (although breakfast cereals are often sprayed with supplemental Ca2+). Although other countries who rely on rice as a major staple face a similar problem, they are much more likely to combine it with vegetables, if they have the money. Thus, Ca2+ deficiencies are more of a problem in developed countries, which are also more likely to have an aging population. Since fortifying plants with supplemental Ca2+ increases their tolerance to stress, it would be prudent to consider the genetic alteration of Ca2+ stores with transgenes that benefit consumers as well as farmers.

Ca2+ levels show a high degree of heritability but vary from species to species. Ca2+ distribution was shown to vary ~2-fold among Arabidopsis ecotypes and was correlated with Mg2+ in all tissues except seeds, [171]. In general, there is more Ca2+ in shoots than in roots, and the distribution within leaves is nonuniform. In grasses, Ca2+ is found only in the upper epidermis. In dicots, Ca2+ levels are low in both upper and lower epidermis, but are higher in mesophyll, a distribution that facilitates Ca2+ control over stomatal aperture. Much of the variation in Ca2+ levels could be traced to the expression of CAX1 [162]. So far, Ca2+ levels have been altered by mutation or overexpression in the vacuole, ER, and apoplast using two proteins—CAX1 and a derivative of CRT. This section will describe these results and examine their collateral impact on abiotic stress responses and make some recommendations for future experiments.

6.1. Transgenic Expression of CAX Family Members

The protein family with the best potential for increasing bioavailable Ca2+ in plants is CAX, located on the tonoplast membrane [188]. As previously mentioned, CAX1 expression levels are the primary determinant for Ca2+ levels in Arabidopsis [162, 189]. In leaves, CAX1 functions to clear free Ca2+ from the apoplast, so that guard cell signaling, which requires extracellular Ca2+, can be regulated properly (Section 3). In sCAX1, the N-terminal autoinhibitory loop has been removed so that it can transport increased amounts of Ca2+ into the vacuole [190].

Ectopic sCAX1 expression increased Ca2+ in potato tubers by 2-3 folds with no change in morphology or yield when supplemented with 2 mM CaCl2 during the first 3 months [191]. However, CAX1 transports other cations in addition to Ca2+, which are not as beneficial from a nutritional standpoint. Hirschi’s group, therefore, modified the CAX2 gene, which shows a greater specificity for Ca2+, to eliminate its Mn2+ transport function and then showed ~50%–60% increase in Ca2+ in transgenic potatoes [192].

Tomato was transformed with CAX4, which is more specific for Ca2+ than CAX1 [193]. This resulted in a 40% increase in total Ca2+ and was not associated with Ca2+ deficiency symptoms even in the absence of CaCl2 supplementation. CAX4 increased fruit firmness (and, therefore, postharvest life), but did not impact ethylene production or sugar content [193]. In addition, root growth was enhanced [193]. Later experiments in Arabidopsis demonstrated that CAX4 expression, which is uniquely confined to roots, is needed for normal root growth and that cax4 mutants had reduced DR5 : GUS expression [194]. DR5 is a synthetic promoter that responds to auxin. The authors postulate that cytosolic Ca2+ levels may have increased, due to altered CAX4-mediated efflux into the vacuole. This may have affected polar transport of auxin, which is regulated by CDPKs [194]. The impact of CAX genes on root growth is important and deserves further study. Although CAX1 and CAX4 are thought to act primarily by depleting cytosolic Ca2+, roots of CAX1 transformants were less sensitive to inhibition by applied auxin than the wild type [195], while roots of CAX4 transformants were more sensitive [194]. It would be very interesting to know how these alterations in root phenotype affect tolerance to abiotic stress.

There are no deleterious effects of sCAX1 expression under normal conditions if supplemental Ca2+ is added. Otherwise, Ca2+ deficiency symptoms result, which include increased sensitivity to salt and cold stress. The yeast 2-hybrid experiments demonstrated that CAX1 interacts with SOS2, a CIPK that usually requires SOS3 (a CBL) for activity, through its N-terminal domain [196]. This may help to deplete the cytosol of excess Ca2+ following salt stress, which is known to produce a transient increase in Ca2+. Overexpression of sCAX1 increased the plant’s sensitivity to salt, perhaps by being too efficient in the removal of excess Ca2+, leading to store depletion [196]. The impact of drought and osmotic stress on sCAX1 overexpression has not been studied, but would be expected to show similar responses (enhanced sensitivity). In contrast to sCAX1 transformants, mutants of CAX3 show increased salt sensitivity [197]. Both decreased Ca2+ transport into the vacuole during salt stress and decreased H+ ATPase activities at the plasma membrane were associated with the cax3 mutation.

Overexpression of CAX1 also resulted in increased sensitivity to salt while mutations in this gene produce salt and drought tolerant plants [195]. Interestingly, exogenous Ca2+ can reverse salt sensitivity in CAX1 transgenic plants and can also reverse the salt tolerance of cax1 mutants [195]. CAX1 may be involved in sequestering Ca2+ to the vacuole following release into the cytoplasm. If Ca2+ signals cannot be dampened by transport into the vacuole, cytosolic levels may remain high, activating salt tolerance pathways. Conversely, if Ca2+ is sequestered into the vacuole at a faster rate than normal (as in the CAX1 over-expressors), cytosolic levels may never reach the threshold required to activate pathways for salt tolerance. cax1 mutants showed developmental abnormalities including reduced root growth and delayed flowering [195].

Ectopic expression of sCAX1 in tobacco was also associated with increased sensitivity to cold shock [188]. This correlated with the positive impact of mutations in cax1 on cold tolerance [51]. The negative impact of CAX1 on cold tolerance was shown to be due to decreased upregulation (relative to wild type) of DREB1 and a subset of cold-responsive genes induced by DREB1 [51]. These results are interesting because they are the first to demonstrate altered gene expression by CAX1, although the signal transduction pathway has yet to be demonstrated. Although DREB1 was upregulated by cold in the cax1 mutants, there were no changes in gene expression associated with exposure to dehydration or salt [51].

There are also beneficial effects of ectopic CAX expression. Both CAX1 and CAX4 expressions have been associated with enhanced tolerance to heavy metals [194, 198201]. The potential impacts on other traits are difficult to assess. CAX1 and CAX3 have been shown to regulate phosphate homeostasis by repressing phosphate starvation-associated genes [202]. A cax1/3 double mutant resulted in increased shoot phosphorous accumulation [202]. Grafting experiments suggested that CAX1 and CAX3 could be involved in the generation of a shoot to root signal that represses phosphate transport [202], but the impact of sCAX1 over-expressing plants on phosphate transport has not been determined.

Unfortunately, sCAX1 expression has not contributed towards mitigation of Ca2+ deficiency diseases. Massive cell death is associated with Ca2+ deficiency resulting, for example, in fruit that is not suitable for consumption. Tomato fruit development is especially susceptible to cell death (blossom end rot) caused by Ca2+ deficiency [203] a situation aggravated by increased salinity [204]. Blossom end rot in tomato is known to be related to Ca2+ deficiency [205]. Instead of helping to prevent blossom end rot, sCAX1 expression resulted in 100% of the tomato fruits developing the disorder [206]. This may have been due to reduced free Ca2+ in the apoplast, where it likely helps to stabilize membrane structure, among other things. Ca2+ deficiency near the plasma membrane causes destabilization, which could precipitate the disorder [206]. Although sCAX1 expression may make Ca2+ more bioavailable to humans, it does not appear to have the same effect in plants.

In contrast to Arabidopsis, over expression of soybean CAX1 homolog in Arabidopsis increased salt tolerance [207]. GmCAX1 has an N-terminal autoinhibitory loop, also found in AtCAX1, but shows only 65% homology to it and 68% homology to CAX2 [207]. In contrast to Arabidopsis, GmCAX1 was not induced by cold suggesting that the regulation and function of different CAX homologs may show considerable variation across species [207]. It is not clear what this means for predicting the impact of overexpression of sCAX1 in other species. As acknowledged [197], it may be difficult to predict the effects of overexpression of a major transporter on the phenotype of any plant.

6.2. CRT and CBP

CRT is a multifunctional protein that is highly conserved in eukaryotic cells [208210]. It has at least three functional domains: a globular N-domain, a proline rich, high affinity ( = 1.6 μM), low capacity ( = 1 mol/mol of protein) Ca2+-binding domain (the P-domain), and a highly acidic, low affinity ( = 0.3–2 mM), high capacity ( = 20–50 mol/mol of protein) Ca2+-binding domain (the C-domain) [211]. In animals, CRT has been suggested to be involved in Ca2+ signaling [212, 213], chaperone activity [211], cell adhesion [214], gene expression [215], apoptosis [216], and in controlling store-operated fluxes through the plasma membrane [217219]. Overexpression of CRT in both plants [220] and animals [221] increases total ER Ca2+ stores.

We found that ectopic expression of the maize CRT1 or a Ca2+-Binding Peptide (CBP) consisting of only the CRT C-domain can not only increase Ca2+ stores, but also enhance the survival of Arabidopsis plants grown in low Ca2+ medium [222, 223], suggesting that the extra Ca2+ could be used by the plant in times of stress. The hypothesis guiding this research is that the CBP sequesters Ca2+ in the ER in a manner similar to CRT. However, Ca2+ may bind the CBP protein in the ER, but then travel as a complex through the secretory system to the vacuole, cytoplasm, or even the nucleus [224]. It is highly unlikely that Ca2+ will be bound by ER-CBP in the cytoplasm, because of its low affinity. It is, therefore, reasonable to use the ER-CBP as a tool for altering intracellular stores of Ca2+.

Our previous work demonstrated that intracellular Ca2+ levels could be manipulated in Arabidopsis by heat shock induction of an ER-targeted GFP-CBP peptide constructed by translationally fusing the green fluorescent protein gene to a sequence corresponding to 126 amino acids derived from the maize calreticulin C-domain [223]. ER-CBP plants induced on Ca2+ containing medium survived longer than similarly heat-shocked ER-GFP control plants when transferred to Ca2+ depleted medium [223]. This work suggested that the ER capacity for Ca2+ could be directly related to a physiological response, early senescence in the absence of Ca2+. Importantly, ER Ca2+ could be modulated without the addition of external Ca2+ and deleterious effects due to Ca2+ depletion were not apparent. To further examine physiological differences in these plants and to avoid the complications of heat shock induction, we transformed Arabidopsis with the same GFP-CBP construct (or CBP without GFP, for indo-1 experiments) but under the control of the constitutive 35S cauliflower mosaic virus promoter.

Why not over-express CRT to increase Ca2+? Overexpression of ZmCRT1 in tobacco cells increased Ca2+ by 2-fold, and transformation of Arabidopsis with ZmCRT1 reduced the rate of senescence following transfer to low Ca2+ media [222]. There are two potential problems with over-expressing full-length CRT, silencing of the endogenous gene, and deleterious effects under some conditions. Overexpression of CRT2 resulted in the production of dwarfed plants, caused by high levels of salicylic acid [145]. Although overexpression of Chinese cabbage CRT1 enhanced shoot and root regeneration in tobacco, the subsequent growth of tobacco plants was retarded [225]. CRT1 overexpression was also shown to be deleterious in rice [186].

My group initially used a soybean heat shock promoter to drive the expression of a maize CRT1 C-domain, which we called CBP for Ca2+ binding peptide, fused to GFP to stabilize it. This turned out to be unnecessary although it was very useful for detecting gene silencing. Nevertheless, we were able to increase Ca2+ in heat-shocked plants by ~15%. Now we know that total Ca2+ levels can be increased by ~25% using constitutively expressed ER-localized CBP [158]. Arabidopsis plants transformed with 35S : CBP showed better salt and drought tolerance and had longer roots, even in the absence of stress [158]. There were no detectable differences in GFP-CBP plants compared to GFP or control plants under normal conditions except that seed production was slightly higher and seedling root growth was increased [158].

Preliminary experiments using both cytoplasmic aequorin-expressing plants and indole-1 ratio imaging suggested that there were no significant differences in [Ca2+]cyt concentrations between 35S : CBP-expressing Arabidopsis and wild type or 35S : GFP control plants [226]. However, after 4-5 days growth in Ca2+-deficient media, the peak [Ca2+]cyt in control plants was significantly lower than in CBP-expressing plants in response to a 150–300 mM NaCl challenge [226]. This suggested that expression of CBP allowed plants to respond to stimuli over a longer period of time due to the excess ER-localized reserves of Ca2+. This was a very interesting result that could provide a mechanism for how CBP benefits plants with respect to stress tolerance.

Microarray results of 35S : GFP-CBP compared to 35S : GFP plants showed that genes for endomembrane and cell wall-associated proteins were upregulated [158]. One Ca2+-regulated gene was strongly upregulated (greater than 3.5-fold), CIPK6. As described in Section 2.3, CIPK6 is a protein kinase that interacts with a Ca2+ sensor protein, CBL. Mutants in AtCIPK6 are sensitive to salt [44], and overexpression of a constitutively active mutant of AtCIPK6 in tobacco confers salt tolerance and also increases root length [44, 126], which are both found in CBP-expressing plants. We, therefore, asked if the enhanced salt tolerance was due to co-expression of CIPK6. When CBP was crossed with a cipk6 knockdown mutant (50% reduction in mRNA) and then challenged with NaCl, it showed the same response as wild type plants. This was somewhat disappointing, as we believed that CBP would enhance stress tolerance by providing a Ca2+ reserve. Of course the induction of CIPK6 may have been caused by the presence of additional ER Ca2+; but the eradication of the response by a single mutation was surprising. It remains possible that there is an extra advantage of CBP expression in drought tolerance or under different conditions. The cipk6 mutant has been complemented with a CIPK6 transgene (D. Chattopadhyay, pers. Comm.).

CBP-expressing plants also downregulate CIPK23, which is also involved in salt tolerance, by 2-fold [158]. We believe this is why the CBPxcipk6 plants showed a similar response to NaCl as the controls, despite the presence of ~50% CIPK6 in the knockdown mutant.

How does CIPK6 enhance salt tolerance? Recent experiments from Kudla’s group have shown that CIPK6 interacts with AKT2, a K+ channel [113]. Interaction occurs on the ER membrane, although both proteins are translated in the cytosol. CIPK6 interacts specifically with CBL4, which was originally identified as an SOS (salt overly sensitive) mutant [227229]. When CBL4 is modified by both myristoylation and palmitoylation, the AKT2/CIPK6/CBL4 complex moves from the ER membrane to the plasma membrane, where AKT2 participates as a K+ channel. Mutations in CIPK6, AKT2, and CBL4 confer similar phenotypes when grown under short days, reduced leaf number and size and delayed flowering [113]. K+ is needed for phloem transport, and the reduced size of the mutant plants is restored under long day conditions [113]. This phenotype is consistent with a reduction in phloem transport, but does not provide an explanation for the altered response by cipk6 to NaCl.

The role of AKT1, which is modulated by CIPK23/CBL1/CBL9 in a similar manner as AKT2, was recently called into question. Mutants defective in akt1 or cipk23 showed better drought tolerance than wild type plants, suggesting that CIPK23/CBL1/CBL9 regulation of AKT1 may actually decrease abiotic stress tolerance [230]. However, overexpression of CBL1 and CIPK23 has been shown to increase tolerance to abiotic stress [39, 47]. Clearly, more experiments are needed to understand the relationship between K+ and abiotic stress.

In addition to Arabidopsis, CBP has been transformed into potato and rice ([231], S. Y. Lee, R. Qu, and D. Robertson, in preparation). The goal for CBP expression in potato was to prevent internal heat necrosis (INH), a disorder affecting the quality of potato tubers [232]. There is strong but indirect evidence for an involvement of Ca2+ in this disorder. The application of antitranspirants to potato leaves reduced total Ca2+ levels and increased Ca2+ in tubers. This led to a decreased incidence of the disorder. However, when 3 independent transgenic potato lines (cv. Atlantic) expressing a 35S : CBP gene were grown under greenhouse conditions, the incidence of INH correlated positively with expression of 35S : CBP, which also increased potato tuber yield and total Ca2+ in leaves [231]. It was not possible to measure Ca2+ in tubers. There were also increased levels of Mg2+ and Mn2+ in the CBP-expressing plants, and reduced levels of K+ [231]. Although the increased yield was statistically significant, the experiment would need to be repeated. It is not known if it was the increased yield that was responsible for greater incidence of INH, but it is unlikely that it could be separated from the expression of CBP.

It would be interesting to know if CBP expression in other plants (besides Arabidopsis) causes an increase in CIPK6 orthologs, and these experiments are currently in progress for rice (Lee, Qu, and Robertson, unpublished). Does the induction of CIPK6 depend on a flux or an increase of ER Ca2+? Could this result from ACA and ECA activity in removing Ca2+ from the cytosol? Confocal microscopy of the GFP-CBP fusion protein showed ER and, to a lesser extent, nuclear activity [158]. Although CBP would not be expected to bind Ca2+ in the cytosol, it could bind Ca2+ in the nucleus. Acidic domains can act as transcriptional coactivators [233], providing a possible mechanism for CBP action. These results illustrate the difficulty of using genetic methods to modulate specific stores of Ca2+. Although targeting of CBP to the nucleus could be used as a control, the molecular weight of CBP is estimated to be ~5 kDa so it should enter the nucleus without a targeting sequence.

6.3. Coexpression of sCAX1 and CRT1

100% of tomato plants expressing sCAX1 developed blossom end rot, a Ca2+ deficiency related disorder that leads to necrosis in the distal, developing end of the fruit [206]. These plants were grown in a greenhouse under conditions where none of the nontransgenic control plants developed the syndrome. The expression of sCAX1 was shown to reduce apoplastic Ca2+ levels, which increased membrane leakiness [234]. Co-expression of CRT resulted in a significant decrease in Ca2+ deficiency symptoms in both tomato and tobacco without the addition of supplemental Ca2+ [235]. This is very interesting and, if it can be repeated in other species, may suggest several things about the ER and vacuole with respect to signaling. Questions that this observation raises include the following:(1) How is the Ca2+ level in the shoot increased, without an increase in transpiration? (This is relevant to all sCAX-expressing plants.) (2) Does the ER form a symplastic Ca2+ network distinct from the apoplast and vacuole? (3) Is CRT needed to keep a bioavailable pool of Ca2+ inside the ER for signaling? If so, then extra CRT may have successfully competed with CAX1 for the limited pool of free Ca2+ in the apoplast in the dual transgenic plants. (4) Can the vacuole serve as the source of Ca2+ for some stimuli? (5) What would CRT overexpression in a cax1/3 mutant do? Could it help to bind excess apoplastic Ca2+?

6.4. Other Transgenes for Manipulating Ca2+ Stores in Plants

Several Ca2+-related proteins have the potential to serve as a mechanism for altering Ca2+ stores in plants. Theoretically, any part of the cell except the cytoplasm could sustain increased levels of Ca2+ without deleterious consequences, although this needs to be experimentally verified. As a group, plants vary in Ca2+ content and show differential sensitivity to Ca2+ as a nutrient [236]. Since we know there is variation in Ca2+ levels between plants, even between ecotypes of Arabidopsis (and that variation correlates with CAX1 expression [189]), we should be able to genetically manipulate it.

One of the benefits of large-scale scientific experiments (“omics”) is the availability of data for gene expression and ion concentrations for a variety of closely related plants. Arabidopsis ecotypes have been collected from around the world, and there are hundreds of accessions, each of which shows less genetic variation than would be found between two species, but together there is a large pool of variation that can be correlated with a variety of different phenotypes. The leaf ionome of 31 of these accessions has now been completed, and Conn and his colleagues have outlined methods for using this data to identify candidate genes controlling elemental accumulation [237]. This promises to be a very productive avenue of research, especially if some of the candidates can be correlated with positive agronomic properties.

In addition to proteins found in Arabidopsis, there are other Ca2+ binding proteins that have been identified in various species. Examples include a celery vacuole-associated dehydrin-like protein [238] and a radish vacuolar Ca2+ binding protein [239] that is induced by lack of Ca2+. Neither of these proteins has mutants nor has been overexpressed, so it is not clear how much Ca2+ can be increased by using them.

Recently, TPC1, the slow vacuolar channel found in all plants, has been shown to contain a novel Ca2+ binding site that senses Ca2+ and alters its activity. Mutants have been created that are insensitive to feedback inhibition by luminal Ca2+, which leads to an increase in the store of vacuolar Ca2+ [240].

Simply adding Ca2+ to fertilizers can increase leaf Ca2+ levels by up to 3-fold [241], and there is an argument that transgene manipulation may be unnecessary as breeding for increased Ca2+ levels should be sufficient to meet nutritional requirements for Ca2+. There are two arguments against this notion: adding Ca2+ to the right compartment has the potential to boost the resiliency of plants to stress and providing Ca2+ loosely complexed to protein might result in enhanced nutritional absorption. Since overexpression of CRT can be detrimental to plant growth [145, 225], transgenic approaches that separate out the C-domain are the most straightforward approach to boosting ER Ca2+.

6.5. Biofortification Studies

The potential role for CAX in biofortification has been demonstrated in carrots expressing sCAX1 [242]. Human consumption of the genetically engineered carrots resulted in a 41% increase in Ca2+ absorption compared to controls, demonstrating the bioavailability of vacuolar Ca2+ in this system [242]. Lettuce was also transformed with sCAX1 and contained 25%–32% more Ca2+ than controls [243]. The response of a human panel to the engineered lettuce was positive for its sensory characteristics [243]. As long as sCAX1-expressing plants have good agronomic properties and can be grown in the presence of excess Ca2+ or cotransformed with CRT (or, better, CBP), this is a very promising method for biofortification.

The absorption of Ca2+ from vegetables can be complicated by the presence of “antinutrients” such as oxalic acid, which forms insoluble Ca2+ oxalate crystals. As long as the diet is varied, it should not have a significant impact. Antinutrients are more important when choosing a plant for transgenic modification. These requirements are fulfilled in carrots, a good choice for one of the first plants to be transformed for increased Ca2+ absorption [242].

It has never been tested in clinical trials, but the delivery of Ca2+ ions complexed with protein, such as found in the ER in the form of the C-domain of CRT (CBP), could increase the absorption rate of Ca2+. Although the use of sCAX1 to increase vacuolar Ca2+ has achieved remarkable increases in Ca2+ absorption on a per gram basis [242], the overall efficiency of Ca2+ absorption was 10% less than for controls. The reason for this is not clear, unless the level of antinutrients increased (which would be important to know). Comparing the efficiency of absorption between CBP transgenic and sCAX1 transgenic carrots could help to determine if Ca2+ absorption efficiency decreases as its concentration increases, or whether the cellular context of the extra Ca2+ plays a role in absorption. In the long run, it will be important to be able to use Ca2+ as efficiently as possible. Since CBP and the combination of CBP and sCAX1 lead to higher total Ca2+ levels without external supplementation, the added nutritional benefit may not require supplemental Ca2+ to be added.

Because the CRT C-domain is not highly conserved, it should be possible to choose sequences that retain a high number of acidic amino acids, which are known to bind Ca2+, without causing silencing of the endogenous CRT genes. The potential for CBP expression alone to increase Ca2+ absorption from food should be tested, because Ca2+ loosely bound to a protein may be even more bioavailable than Ca2+ salts in the vacuole. CBP has not been associated with Ca2+ deficiency symptoms under normal or stress conditions in the laboratory. It would be interesting to compare Ca2+ uptake from sCAX-expressing plants to those expressing CBP, along with a combination of the two transgenes. The long-term goal for sustainable agriculture should be to maximize the efficiency of Ca2+ supplementation in the human diet, so that the effective use of Ca2+ as a fertilizer can be maximized.

6.6. Summary

Transgenic expression of sCAX1 or CAX4 may be the best way to increase vegetative sources of Ca2+ but this can require supplementation with CaCl2. When coexpressed with CRT1, the need for Ca2+ supplementation appears to be reduced, but more studies are needed to determine the effect of two Ca2+ binding transgenes on agronomic properties, because CRT1 overexpression by itself can have deleterious effects on plant growth under certain conditions.

Transgenic expression of CBP also increases total Ca2+ but not by as much as CAX1. This may be a better transgene to co-express with CAX1 than CRT1 because it retains Ca2+ binding but lacks most of the functions of CRT1. CBP expression by itself increases root growth under nonstress conditions and reduces the effects of drought and salt stress, perhaps in part by increasing root growth but we think also by providing a more extensive store of bioavailable Ca2+.

7. Conclusion

As described in the beginning, many studies show that Ca2+ applied externally can benefit plants by increasing stress tolerance. Even postharvest fruit characteristics are improved following a CaCl2 soak. It is still not known where in the plant this supplemental Ca2+ is absorbed and distributed, or how it is used to benefit the plant. How much is actually necessary for the enhanced growth and stress responses? Is it the change in Ca2+ concentration or the absolute amount of Ca2+ available to the plant that is relevant?

One explanation for the beneficial response is that Ca2+ induces genes involved in abiotic stress tolerance, such as members of the CIPK/CBL family, some of which are known to be induced by exogenous Ca2+ [45, 125]. But rather than overexpressing Ca2+-regulated genes, it may be more beneficial to increase the Ca2+ stores that are used to cause their induction. Finding ways to genetically increase Ca2+ levels in plants may allow us to capture the Ca2+-stimulated enhancement under normal conditions or with minimal Ca2+ supplementation. Additional research on targeting Ca2+ binding proteins to various organelles may, therefore, be useful.

More robust signaling pathways and stress responses would seem to be a good thing in the face of global climate change. By increasing just the second messenger, one could conceivably preserve the ability of the plant to adapt to different stresses. By increasing the degree of stress response, but not the specific pathway, plants may be better able to deploy valuable reserves into tolerating a wide variety of different stresses. This would make the ubiquity of Ca2+ an asset rather than an impediment to research. The more we understand about Ca2+-regulated pathways, the more we can optimize the response to adverse conditions. One thing is clear, more exploratory research on the ectopic expression of Ca2+ binding or exchange proteins could be very promising for plants, agronomists, and consumers.

Acknowledgments

The author would like to thank Dr. Sang Yoon Lee for his dedication and initiative in working on the CBP project and for many interesting discussions. Drs. Pei-Lan Tsou and Sarah Wyatt started this work, and it was Dr. Wendy Boss’s idea to use the CRT C-domain as a transgene for manipulating ER Ca2+. I would also like to thank Dr. George Allen for his critical comments and encouragement and Dr. Steven Nagar for Figure 2. The CBP project was originally funded by NASA.