Calcium at the Cell Wall-Cytoplast Interface

Abstract

Attention is given to the role of Ca²⁺ at the interface between the cell wall and the cytoplast, especially as seen in pollen tubes. While the cytoplasm directs the synthesis and deposition of the wall, it is less well appreciated that the wall exerts considerable self control and influences activities of the cytoplasm. Ca²⁺ participates as a crucial factor in this two way communication. In the cytoplasm, a [Ca²⁺] above 0.1 μM, regulates myriad processes, including secretion of cell wall components. In the cell wall Ca²⁺, at 10 μM to 10 mM, binds negative charges on pectins and imparts structural rigidity to the wall. The plasma membrane occupies a pivotal position between these two compartments, where selective channels regulate influx of Ca²⁺, and specific carriers pump the ion back into the wall. In addition we draw attention to different factors, which either respond to the wall or are present in the wall, and usually generate elevated [Ca²⁺] in the cytoplasm. These factors include: (i) stretch activated channels; (ii) calmodulin; (iii) annexins; (iv) wall associated kinases; (v) oligogalacturonides; and (vi) extracellular adenosine 5′-triphosphate. Together they provide evidence for a rich and multifaceted system of communication between the cytoplast and cell wall, with Ca²⁺ as a carrier of information.

Peter K. Hepler (Corresponding author)

Introduction

Historically, the plasma membrane (PM) has been considered to be the interface between the living protoplasm and the non-living extracellular matrix. While this membrane marks the boundary between the cell wall to the outside and the cytoplasm to the inside, implying that the cell wall is “non-living” is misleading. Even though it is deposited outside the PM, the cell wall is formed by the activity of the cytoplasm, and maintains a close relationship with it as revealed by the presence of structural attachments to the PM. In addition, the cell wall contains enzymes, and other factors that greatly influence and modify its structure and composition (Carpita and Gibeaut 1993). The cell wall thus possesses active mechanisms for self regulation. Increasingly we are confronted with evidence that the cell wall, due to its structural properties and composition, influences the behavior of the “living” protoplast (Brownlee 2002; Monshausen and Gilroy 2009). For example, during the process of cell growth itself, the yielding of the wall dictates the extension of the cell, and for polarized cells, local yielding determines plant cell shape (Schopfer 2006). It seems clear therefore that much communication takes place between the cell wall and the cytoplasm, with the PM occupying a pivotal position in responding to and transmitting information.

Calcium ions (Ca²⁺) are an important component of the information system that operates between the cytoplasm and the cell wall, participating in crucial ways in both compartments. In the pollen tube cell wall for example, the activity of free Ca²⁺ is in dynamic steady state, influenced by inward diffusion from the external medium, where [Ca²⁺] typically ranges from 10 μM to 10 mM (Picton and Steer 1983), and by ion exchange and coordination with wall polysaccharides. Ca²⁺ participates in cross-linking negative charges, especially on the carboxylic residues of pectins, imparting significant structural rigidity to the wall. In the cytoplasm the [Ca²⁺] is tightly regulated in the vicinity of 0.1 μM, with some spatially and temporally restricted elevated domains at 1 μM–10 μM (Rathore et al. 1991; Miller et al. 1992; Pierson et al. 1996). As is increasingly well appreciated, Ca²⁺ in the cytoplasm participates in myriad events where it acts as a second messenger in a host of signaling pathways.

Given the substantial differences in [Ca²⁺] between the cell wall and cytoplasm, together with the realization that the ions are serving rather different functions in these two compartments, it becomes an interesting question how Ca²⁺ operates at the interface. Of course it is well appreciated that channels and pumps on the PM act as gates to control Ca²⁺ influx and efflux. However, it is often not acknowledged that the Ca²⁺ that crosses the PM and enters the cytosol must first pass through the cell wall. The Ca²⁺ binding reactions, rather than the concentration of the ion in the medium, will determine the activity of Ca²⁺ at the mouth of a selective channel and thus the magnitude of its influx into the cell. Similarly, the Ca²⁺ that is pumped out of the cell, and extruded into the cell wall, will immediately influence wall structure in those local domains near the pumps, being governed by the local [Ca²⁺] and not that of the medium. As a consequence there will be local domains with associated fluxes that become crucial to the growth and development of the plant as a whole.

It will be the purpose of this essay to consider the activity of Ca²⁺ at the cell wall/cytoplast interface. While we draw broadly on work that has been gained from studies of different plant cells, we give emphasis to results from research on the growth of pollen tubes. Although the pollen tube is a specialized cell, it possesses a cell wall composed of polymers well recognized for their presence in other plant cells, and obeys the biophysical laws governing cell wall expansion that apply to plant cells in general (Hepler et al. 2006; Chebli and Geitmann 2007). The pollen tube PM also contains different pumps and channels that participate in regulation of Ca²⁺ import and export (Sze et al. 2006). Although some of the specific proteins are uniquely expressed in pollen tubes, they are members of larger families with conserved functions fundamental to all plant cells. But beyond this, the growing pollen tube demonstrates clear and specific requirements for Ca²⁺ and in addition demonstrates intracellular profiles, such as the tip-focused Ca²⁺ gradient, which are recognized as essential features (Holdaway-Clarke and Hepler 2003). The pollen tube is thus a favorable system in which to explore the role of Ca²⁺ at the cell wall/cytoplast interface.

Ca²⁺ and the Cell Wall

It has long been appreciated that Ca²⁺ plays a key role in determining the structure and function of the cell wall (Hepler 2005). Through cross-linking negatively charged regions in pectins, found primarily in short stretches of de-methoxylated homogalacturonans, Ca²⁺ imparts rigidity to pollen tube cell walls. When the [Ca²⁺] is too low, the wall is weakened and may even break, whereas at high [Ca²⁺] the pectin chains will be cross-linked and aggregated, and the wall maximally rigidified. The experimental control of [Ca²⁺] was pivotal in early studies aimed at separating single living cells. Over 50 years ago, the chelator, ethylenediaminetetraacetic acid (EDTA), was used to remove Ca²⁺ from pectins in the middle lamella, and thus to break down tissues into single cells (Letham 1958).

In pollen tubes the growing apical wall consists almost entirely of pectin, with cellulose and callose being located several microns from the apex (Ferguson et al. 1998). As a consequence, when the [Ca²⁺] is lowered sufficiently the pollen tube wall completely loses its structural integrity and bursts. The permissive [Ca²⁺] for pollen tubes extends between 10 μM to 10 mM (Picton and Steer 1983). Below 10 μM, the pollen tube bursts because the wall is too weak, and above 10 mM, pollen tube growth frequently stops, which is likely due to excessive cross-linking. Because a [Ca²⁺] of 10 μM or less should be sufficient to satisfy the number of binding sites in the wall it may be more pertinent to speak about the availability of Ca²⁺ at the cell wall/cytoplast interface as new wall material is being deposited. Because 20%–30% of the newly deposited pectin will be de-esterified (Bosch and Hepler 2005), there will be always an immediate need for Ca²⁺ by the growing pollen tube. Therefore a [Ca²⁺] of 10 μM is the amount needed in the medium to ensure that sufficient ions are available at the cell wall/cytoplast interface in order to prevent the tube from bursting. But even in other cell types, the Ca²⁺/pectate bonds play a major role in regulating the ability of the cell wall to extend. For example, mature cucumber hypocotyls cells, which become insensitive to the growth promotive effects of expansin, can regain their sensitivity simply through chelation of the wall Ca²⁺ with ethyleneglycoltetraacetic acid (EGTA) (Zhao et al. 2008).

Pectins are secreted largely as methoxy-esters (70%–80%), and only later de-esterified in the apoplast by the enzyme pectin methyl esterase (PME), which has also been secreted (Willats et al. 2001; Bosch and Hepler 2005). The methoxylation pattern of newly secreted pectins is not known, but it is known that plant PMEs carry out block-wise de-methoxylation, creating contiguous stretches of galacturonic acid residues (Bosch et al. 2005). The extent and strength of Ca²⁺ cross-linking will depend on the pattern of de-methoxylation as well as on the number and availability of the acidic residues. Recent evidence from in vitro systems under conditions of a high degree of methoxylation, and a low degree of block-wise de-methoxylation, or simply degree of blockiness, indicate that pectins associate ionically, with carboxyl moieties participating in labile binding with free Ca²⁺ forming plastic gels with low shear strength (Fang et al. 2008). As pectins are de-methoxylated block-wise by PME, dimers begin to form in a cooperative fashion, so that binding strength increases rapidly as the ratio of Ca²⁺ to available binding sites increases. The number of consecutive de-methoxylated galacturonic acid residues required to form stable chains in a modified, or shifted “egg-box” configuration (Braccini and Perez 2001; Braccini et al. 2005) has been estimated in various systems to range from 6 to 20 (Fraeye et al. 2009).

Ultimately, at high [Ca²⁺], with a low degree of methoxylation and high degree of blockiness, pectins reach maximum strength (Fraeye et al. 2009). Thus at a given, but permissive [Ca²⁺], pollen tube wall expansion and growth can be stopped simply by adding exogenous PME (Bosch et al. 2005; Parre and Geitmann 2005). Enzymatic de-methoxylation creates an abnormal number of blocky, de-methoxylated regions over the entire pollen tube tip, concomitantly increasing wall stiffness due to the extent of Ca²⁺ cross-linking. From these brief comments it can be appreciated that pollen tube growth is dependent on a balance between the number of available acidic residues and the [Ca²⁺]. If there are too few acidic groups or too little Ca²⁺ the pollen tube bursts, but if there are too many acidic groups and/or too much Ca²⁺, the wall is overly cross-linked and unable to extend.

Another factor that influences Ca²⁺ binding is the stress that turgor pressure imposes on the wall binding sites (Boyer 2009). When Ca²⁺ cross-linked pectates are placed under physical tension, as imposed by turgor pressure, for example, 0.2 MPa, the bonds may lengthen and thus weaken and decrease their affinity for Ca²⁺ (Proseus and Boyer 2007; Boyer 2009). Dissociation may occur, allowing turgor dependent expansion. This scenario has been developed from studies on single internode cells of the green alga, Chara/Nitella, but could also apply to cells of higher plants including pollen tubes. In this scheme it is also important to recognize the ability of turgor pressure to force polymeric building blocks, notably pectates, into isolated Characean cell walls, and to achieve normal rates of growth (Proseus and Boyer 2006). Here the effects are two-fold: first, the insertion of new material within the matrix weakens the existing bonds and facilitates growth by intussusception, and second, the new pectates, which would exhibit a low [Ca²⁺], locally act as a chelator and remove the ion from load bearing bonds, thus weakening ionic linkages, and permitting turgor-dependent elongation (Boyer 2009).

It is now well established that the rate of pollen tube growth oscillates, with changes of up to sixfold (100 to 600 nm/s) occurring during periods of 20–50 s (Cárdenas et al. 2008). Simply from a consideration of the above factors, including notably the availability and activity of PME, there will be ongoing changes in the extensibility of the cell wall. Particularly because the secretion of PME itself oscillates at the apical cell surface (McKenna et al. 2009), it follows that there will be an associated oscillation in the appearance of acidic groups, which will lead to changes in the cell wall yielding properties and ultimately to changes in growth rate. A direct confirmation comes from studies using the extracellular Ca²⁺ selective electrode to monitor fluxes associated with growth. The results reveal an influx of Ca²⁺ focused at the tip of the tube (Kühtreiber and Jaffe 1990), which oscillates (Holdaway-Clarke et al. 1997; Messerli et al. 1999). Cross-correlation analysis further shows that the increase in influx follows the increase in growth rate by over one third of a growth rate cycle (Holdaway-Clarke et al. 1997; Messerli et al. 1999).

There has been disagreement about what this influx represents, with some investigators thinking that it is due to movement of Ca²⁺ directly into the cytoplasm, while other supporting wall binding. Kwack (1967), applying autoradiography to pollen tubes that had been administered with ⁴⁵Ca²⁺, showed extensive labeling of the tube apex, which he interpreted as Ca²⁺ binding by the acidic pectic residues. A similar study by Jaffe et al. (1975) on lily pollen tubes yielded similar results, but the authors argued in favor of uptake into the cytoplasm. Their reason against wall binding was mainly that it would be insufficient to account for the magnitude of signal observed. Also the ions were found to be tightly bound and resistant to washing. However, these conclusions do not fit with data on the wall binding of Ca²⁺. In studies of mung bean hypocotyls the [Ca²⁺] has been shown to increase during maturation from 80 μmol/g dry weight of wall to 122.5 μmol/g dry weight (Goldberg et al. 1986), which when extrapolated to the pollen tube accounts for the bulk of influx observed. Thus, Holdaway-Clarke et al. (1997) calculated that one could predict an influx of 35 pmol/cm² per s, which is reasonably close to the 15–20 pmol/cm² per s that has been measured, especially when you take into account the fact the electrode is only 50% efficient. By contrast, the calculation of the influx needed to sustain the intracellular gradient yields values of 0.5–2.0 pmol/cm² per s, which is a small fraction of the amount measured. Indeed the small amount of influx needed to support the intracellular gradient is within the noise of the extracellular flux measurements, explaining why an influx directly corresponding to the change in the intracellular gradient has not been observed (Holdaway-Clarke et al. 1997). While there is no question that some Ca²⁺ does enter the cytoplasm, it must be recognized that the cell wall, including that of the pollen tube, has enormous binding capacity and in addition can bind Ca²⁺ tightly. Recent work on the dimensions and dynamics of pollen tube walls (McKenna et al. 2009) suggest that the Ca²⁺ flux required may be even greater than that estimated in Holdaway-Clarke et al. (1997) since the maximum wall thickness is 0.5 μm, compared with the 0.2 μm used earlier. In addition, extensive work on in vitro pectin systems has demonstrated three main regions in pectin/Ca²⁺ state diagrams (Vincent and Williams 2009). If the most energetically stable configurations found in vitro also occur in pollen tube walls, it would require 320 to 640 μmol Ca²⁺/g of pectin compared with the 80 to 120 μmol/g used in earlier calculations. With improved methods for measuring Ca²⁺ fluxes we anticipate values on the order of 250 to 500 pmol/cm² per s.

A second argument against the idea that the bulk of the Ca²⁺ detected from the influx data directly crosses the PM and enters the cytoplast stems from the lack of agreement in the phase relationships. Thus the increase in the intracellular tip-focused gradient increases +10–40° behind the increase in growth rate (Messerli et al. 2000; Cárdenas et al. 2008), whereas the extracellular influx increases +130° after the increase in growth (Holdaway-Clarke et al. 1997). If Ca²⁺ were directly entering the cytoplasm there should be a close agreement between the intracellular and extracellular signals. To explain the difference it has been argued that Ca²⁺ enters the cytoplasm but is immediately sequestered within the endoplasmic reticulum (ER) or other compartments, and thus not detected by cytosolic Ca²⁺ indicators (Trewavas and Malhó 1997; Malhó et al. 2006). This process, which has been described in animal cells, is referred to as “capacitatively coupled uptake” or “store operated calcium entry” (Putney et al. 2001; Parekh and Putney 2005). However, studies with manganese, an ion that enters cells through Ca²⁺ channels and quenches indicator dyes such as indo-1, indicate that the apical signal in pollen tubes is rapidly reduced (Malhó et al. 1995), providing support for influx from the extracellular medium directly into the cytosol, and not via capacitative processes.

We can explain the lag of Ca²⁺ influx relative to growth if we consider both the spatial and temporal binding patterns. Ca²⁺ demand in the wall is a function of the amount of wall material, the degree of methoxylation and the degree of blockiness. The peak in the amount of wall material at the pollen tube tip (Region I, Figure 1) occurs at −100° (McKenna et al. 2009) relative to the peak in growth rate. This material has a degree of methoxylation of 70%–80%, with random dispersed carboxyl groups, and a low degree of blockiness. Ca²⁺ demand in this region will be relatively low, and binding will be primarily ionic and labile.

Figure 1. Ca²⁺ flux at the pollen tube tip. Ca²⁺ diffuses into the pollen tube wall (large green arrows) at different rates across the expanding tip, driven by gradients created by different Ca²⁺ binding mechanisms in the wall. In zone I, a moderate flux of Ca²⁺ (green circles) binds ionically to carboxyl groups on newly secreted, loosely associated, highly methoxylated pectin chains (red lines). More pectin arrives continuously in this zone by exocytosis of secretory vesicles (red circles). Pectin methylesterase (PME) is also secreted in inactive form (yellow circles). As the pollen tube wall expands, stretch-activated channels (blue boxes) open creating a small Ca²⁺ influx (narrow green arrows) that sustains a dynamic Ca²⁺ gradient in the cytoplasm, created by Ca²⁺ uptake by the endoplasmic reticulum. In zone II PME becomes active (yellow stars) and de-methoxylates pectins in block-wise fashion, leading to tight association of pectin dimers through coordination with Ca²⁺. Ca²⁺ demand for this process is very high. In zone III pectin dimers form aggregates, also bound with Ca²⁺, creating a mature, inextensible pollen tube wall.

Wall expansion begins to accelerate as the highly methoxylated pectin intercalates and the wall softens. Continued wall expansion displaces the methoxylated pectin away from the primary site of deposition. At the same time, PME becomes more active, either as the pro region is cleaved or as inhibitor diffuses away and begins block-wise de-methoxylation, dramatically increasing both the Ca²⁺ demand for dimerization of pectin chains (Region II, Figure 1) and wall stiffness. Shear forces may orient pectin dimers, possibly leading to anisotropy in wall stiffness. Those Ca²⁺ that are bound between pairs of pectin chains by multiple coordination bonds are likely to be less labile than those that are ionically bound to pairs of carboxyl residues on the exterior of the pectin dimers.

As a result of the elapsed time, the maximum in de-methoxylated pectin occurs well after the peak in growth, which is thought to be about one third of a growth cycle based on extracellular Ca²⁺ influx data (Holdaway-Clarke et al. 1997). Spatially this would take place at the “shoulder” of the pollen tube tip, where expansion must eventually slow as the wall becomes less extensible. For this reason, the flux of Ca²⁺ into the tube tip is maximal well after the peak in growth. Final wall maturation occurs in Region III (Figure 1), where dimers further aggregate with small amounts of additional Ca²⁺ to form the final rigid wall of the pollen tube shank.

Although the focus herein is on Ca²⁺ we must also recognize that cell wall structure and activity will be influenced by other ions, notably protons. De-methoxylation of the methyl-esters by PME produces methanol and a carboxyl group. Given the pH of the wall (5–6), and the pK of pectic acid (about 3.7), it is likely that all of the acidic residues are fully dissociated, and available for Ca²⁺ binding. However, there could be local domains, governed in part by the influx of protons at the tip of the pollen tube and efflux along the sides of the clear zone, which could influence the local degree of dissociation.

Ca²⁺ and the Cytoplasm

Many studies richly document the extensive role for Ca²⁺ as a second messenger in signal transduction. A full review of this topic is beyond the scope of this article, but can be gained from several recent and insightful reviews (Harper et al. 2004; Bothwell and Ng 2005; Harper and Harmon 2005; Wheeler and Brownlee 2008; Kim et al. 2009; Mazars et al. 2009; McAinsh and Pittman 2009). We will, however, focus on a few aspects that seem especially important in the formation and structural organization of the cell wall. Most notable is the process of secretion itself, which delivers many of the building blocks for cell wall formation, including notably the non-cellulosic components (Nebenführ and Staehelin 2001). Ca²⁺ would be expected to occupy a major role because it has been shown to stimulate secretion in both plant and animal cells (Zorec and Tester 1992; Battey et al. 1999). In plants a good example can be found in studies of maize coleoptile protoplasts, where measurements by patch-clamp analysis of whole-cells show that increasing the [Ca²⁺] causes a marked increase in the total surface area of the cell, as detected by an increase in membrane capacitance (Sutter et al. 2000). While noting the agreement between these results and those obtained from animal cells, Sutter et al. (2000) allow that plant cells appear to be more responsive to small changes in the [Ca²⁺] than typical animal cells. Thus half maximal saturation for the maize coleoptile protoplasts is achieved at 0.9 μM and full saturation at 1.5 μM. These data and several other examples establish a role for Ca²⁺ in secretion in plants.

Secretion in the pollen tube appears to fit into the Ca²⁺-stimulus paradigm. First, pollen tubes of all species, which have been investigated, possess a “tip-focused” gradient of free Ca²⁺ that is tightly appressed to the apical PM (Holdaway-Clarke and Hepler 2003). Ca²⁺ gradients have even been observed in Arabidopsis grown in vivo (Iwano et al. 2004, 2009). In lily pollen tubes this gradient reaches from 3–10 μM (Messerli et al. 2000) and declines to a basal level of 0.1–0.2 μM within 20 μm of the tip (Pierson et al. 1996). Second, the region of high [Ca²⁺] also contains numerous secretory vesicles (Figure 1), which have been brought forth by the actin cytoskeleton (Lancelle and Hepler 1992). Third, experimental perturbations aimed at affecting the Ca²⁺ gradient indicate a close coupling between cell elongation and the presence of the tip-focused gradient; growth correlates with a high apical [Ca²⁺] (Pierson et al. 1994, 1996). One contrary example involves inhibition of growth with Yariv reagent, which blocks arabinogalactan protein activity; however, the pollen tubes exhibit a high apical [Ca²⁺] (Roy et al. 1999). In this instance, although growth is blocked, secretion continues but is markedly delocalized, consonant with a similar delocalization in the elevated Ca²⁺ domains. Fourth, experiments in which the position of the gradient has been modified, either by blocking its formation with local application of inhibitors such as lanthanides, or stimulating its formation with local application of Ca²⁺ ionophores, reveal that the location of high [Ca²⁺] defines the region of cell elongation (Malhó and Trewavas 1996). A general conclusion has been that the tip-focused gradient, which is sufficiently high to activate Ca²⁺ dependent proteins, facilitates the process of secretion in local domains, and thus not only stimulates growth, but establishes the polarity of the cell.

Additional insight regarding the connection between Ca²⁺ and secretion comes from studies of oscillatory pollen tube growth. Because the underlying processes also oscillate with the same period but not usually the same phase, it has been possible using cross-correlation analysis to establish whether a process differs in phase from growth, if so by how much and perhaps most importantly, does it precede or follow the increase in growth rate (Holdaway-Clarke and Hepler 2003; Hepler et al. 2006). It has been an assumption that processes that increase in anticipation of an increase in growth rate, might play a role in initiating the growth process, whereas processes that increase after an increase in growth rate, either serve to inhibit the growth event or by some means prepare the cell for the next stimulatory period.

The phase analysis of intracellular Ca²⁺ versus growth rate yields a conclusion that comes as a surprise. Although the intracellular Ca²⁺ signal clearly oscillates with changes from 0.75 μM to above 3.0 μM, the cross-correlation analysis unambiguously indicates that its increase follows by +10–40° the increase in growth rate (Messerli et al. 2000; Cárdenas et al. 2008), an observation that also holds for oscillatory growth of root hairs (Monshausen et al. 2008). Additionally perplexing have been the data showing that exocytosis itself oscillates, but that it anticipates the increase in growth rate by −100° in lily pollen tubes (McKenna et al. 2009). Ca²⁺ and secretion thus appear to differ with one another by as much as 140°. A somewhat similar lack of correspondence between the [Ca²⁺] and secretion comes from the work of Camacho and Malhó (2003), who showed that application of the non-hydrolyzable guanosine triphosphate (GTP) analog, GTPγS, which activates G-proteins, causes an increase in exocyotsis. However, these treatments led to a small but detectable decrease in the intracellular [Ca²⁺].

These results, however, cannot be construed as evidence against a role for Ca²⁺ in secretion. It must be emphasized that the growing pollen tube retains a marked tip-focused gradient, which even at its low point (0.75 μM) (Pierson et al. 1996) is still quite substantial. If the pollen tube has properties similar to the maize coleoptile protoplast (Sutter et al. 2000), and reaches maximal rates of secretion at 1.5 μM, then the oscillating gradient even at its low point may be sufficient to saturate the system. We must keep in mind that the low values of 0.75 μM derive from studies using fluorescent dyes, and are subject to signal averaging within the microscope system. At the mouth of a channel, where the ion first enters the cytosol, the [Ca²⁺] would be expected to be much higher. However, these postulated high values would go undetected because they will be averaged with much lower numbers within the resolution limit of 0.2 μm. Because secretion will be regulated by the [Ca²⁺] at the surface of the PM, it may be that those values always exceed saturation (1.5 μM). Taken together these considerations support the idea that Ca²⁺ participates in exocytosis, but that the observed oscillations in [Ca²⁺] do not cause oscillations in exocytosis because the ion concentration is already above threshold.

If Ca²⁺ participates in exocytosis in the pollen tube, what proteins are involved in responding to the ion and facilitating the process? Given the general role of Ca²⁺ in both plants and animals, it has been attractive to imagine that proteins similar to those identified in animal cells are also active in plants. One candidate includes the annexins, which constitute a large family of proteins found in plants and animals (Bushart and Roux 2007; Mortimer et al. 2008; Laohavisit et al. 2009). A feature that implicates these proteins in secretion consists of a 70 amino acid domain that contains a Ca²⁺ binding region through which the protein cross-bridges with the negatively charged phospholipids on membranes. Further pertinent information includes the observation that these proteins are abundant and close to the PM in cells undergoing active secretion. For example, annexins are abundant at the apex of tip growing cells such as pollen tubes (Blackbourn et al. 1992), root hairs (Clark et al. 2005), and fern rhizoids (Clark et al. 1995). Particularly persuasive has been the observation that annexin stimulates Ca²⁺ dependent vesicle fusion to the PM of maize root cap protoplasts (Carroll et al. 1998). A second family of proteins that are likely involved in transmitting the Ca²⁺ stimulus during exocytosis are the synaptotagmins, which participate in exocytosis in animal cells (Martens et al. 2007). Earlier studies had identified these proteins in plants (Kiyosue and Ryan 1997). Recent work indicates that synaptotagmin 1 in Arabidopsis plays a major role in controlling membrane stability, where it facilitates the ability of the PM to reseal following lesions induced by osmotic or freezing stress (Schapire et al. 2008; Yamazaki et al. 2008). These properties make synaptotagmin 1 a good candidate for a role in exocytosis where membrane breakage and resealing occurs during vesicle fusion, and release of material into the extracellular matrix.

Within the framework of pollen tube growth there are likely several processes, in addition to exocytosis, that are affected by the tip-focused Ca²⁺ gradient. Indeed some of these processes may respond to the oscillations in the [Ca²⁺]. Obvious candidate proteins are calmodulin (CaM) and CDPKs, which participate in many different processes. Although bulk CaM is relatively uniformly distributed throughout the length of the pollen tube, the activated form, as revealed through the use of fluorescent analog (TA-CaM) that binds to the Ca²⁺ bound form, localizes to the apex of the pollen tube in a manner similar to Ca²⁺ (Rato et al. 2004). In addition the results reveal that activated CaM oscillates similarly to Ca²⁺ itself, and thus is able to modulate downstream effectors. However, given the breadth and extent of the CaM literature, we will not attempt to comprehensively review it, but instead refer the reader to several recent reviews (Yang and Poovaiah 2003; Zhang and Lu 2003; Du and Poovaiah 2005; Harper and Harmon 2005; McCormack et al. 2005; Boursiac and Harper 2007; Finkler et al. 2007; Ma et al. 2008; Kim et al. 2009). We will, however, mention two CaM target proteins that are particularly pertinent to the control of Ca²⁺ itself, and thus likely central in the control of pollen tube growth.

These two Ca²⁺/CaM targets are: (i) the cyclic nucleotide gated channel CNGC-18 (Frietsch et al. 2007), and (ii) the autoinhibited Ca²⁺ pump, ACA9 (Schiøtt et al. 2004). CNGC-18 localizes to the apical PM, and is presumably activated by either cyclic adenosine monophosphate (AMP) or cyclic guanosine monophosphate (GMP), and inactivated by Ca²⁺/CaM (Talke et al. 2003; Kaplan et al. 2007; Ma and Berkowitz 2007). Like other plant CNGC channels, it is thought to be a non-selective cation channel allowing the influx of both divalent and monovalent cations including Ca²⁺ and potassium (Talke et al. 2003; Kaplan et al. 2007; Ma and Berkowitz 2007). The presence of CNGC-18 fits well with an earlier study showing that transient elevations of cAMP modulate the pollen tube growth axis (Moutinho et al. 2001). ACA9 also localizes to the PM, but is distributed uniformly over the length of the pollen tube, and is predicted to be activated Ca²⁺/CaM. Their different localization and complementary functions suggest that CNGC-18 and ACA9 closely interact to control the intracellular [Ca²⁺]. Thus the opening of the CNGC-18 allows Ca²⁺ to enter the cell. Because this channel is located to the apical PM, Ca²⁺ influx would be restricted to the tip region, consistent with numerous observations. However, the elevated [Ca²⁺] would bind CaM, and close the channel, blocking further entry of Ca²⁺ (Hua et al. 2003; Kaplan et al. 2007). In this sense CNGC-18 is likely self-regulatory, and potentially able to restrict a massive build up of Ca²⁺. Nevertheless, elevated domains are generated, and these, if allowed to persist and spread, could cause serious physiological problems; for example, Ca²⁺ could react with phosphates and block energy (ATP) production. The ACA9 pump thus becomes an essential factor in reducing the intracellular [Ca²⁺] to basal levels. While we acknowledge the presence of Ca²⁺-pumps on the internal membranes (ER, mitochondria, vacuole), ACA9 serves a key role in extruding Ca²⁺ into the cell wall. Although ACA9 is localized over the entire pollen tube, it will only be active where Ca²⁺/CaM is high. Normally this would be in the apical domain, where ACA9 would return Ca²⁺ specifically to the apical apoplasm (Schiøtt et al. 2004). However, if the pollen tube suffers a wound anywhere along its length, the inevitable rise in [Ca²⁺] would locally bind CaM, and activate ACA9, thus lowering the [Ca²⁺] and restricting damage.

The Ca²⁺ dependent protein kinases or the calmodulin domain protein kinases (CDPKs) constitute another large family of Ca²⁺ binding proteins uniquely found in plants and some protists (Harper et al. 2004; Harper and Harmon 2005). They too are being widely and extensively investigated, and thus will not be considered in detail here. Briefly we note that several of these kinases are membrane associated through myristoylation or palmitoylation linkages. In early studies Moutinho et al. (1998) identified a Ca²⁺-dependent protein kinase activity, possibly a CDPK, associated with the pollen tube apex, which appeared to participate in polarized growth. Recent studies in Arabidopsis have focused on CDPK 17 and 34, which are similar to one another, and localized to the PM being biased towards the apical domain (Myers et al. 2009). These appear to have a profound effect on pollen tube growth as evidenced by observations that double mutants have slower growth rates, and are nearly sterile. Petunia CDPK 1, which is similar to Arabidopsis CDPK 17/34 and also PM localized, likewise participates in pollen tube growth (Yoon et al. 2006). Although these kinases are crucial target proteins that respond to changes in the [Ca²⁺], we know much less about the downstream components. With CDPK 34 alone there may be more than 50 proteins that are phosphorylated (Myers et al. 2009); determining the specific targets and processes associated with them constitute important goals in future research.

Ca²⁺ at the PM/Cell Wall Interface

It has been appreciated for years that extracellular Ca²⁺ affects the stability and permeability of the plant cell PM, with concentrations between 0.1 and 1.0 mM being required to retain the integrity and selective ion transport properties (Hanson 1984; Hepler 2005). Diverse processes such as retardation of leaf and tissues senescence (Hanson 1984), and closure of stomatal pores are stimulated by extracellular Ca²⁺ in the range of 0.1 to 1.0 mM (McAinsh and Pittman 2009). The likely basis for the effects in guard cells may derive from the activity of a Ca²⁺ sensor found on its PM (Han et al. 2003). Plant cells thus have mechanisms for responding to Ca²⁺ in the cell wall. These responses may involve changes in the permeability of the PM itself to Ca²⁺ and other ions, but they may involve changes in the interaction between the PM and cell wall that affect many aspects of cell growth and development. In the paragraphs below we provide specific examples where extracellular Ca²⁺ and extracellular factors act as intermediaries between the PM and cell wall, and in most instances induce a rise in intracellular [Ca²⁺].

Mechanical activation: a role for stretch activated channels

It is well known that plant cells respond to a variety of mechanical stimuli, some of which may be generated endogenously (Monshausen et al. 2009; Monshausen and Gilroy 2009). A prime example of the close interplay between the cell wall and the cytoplasm revolves around the activity of stretch activated channels (SACs) on the PM. Although we have mentioned channels earlier, that is, CNGC-18, we discuss SACs here because they respond to deformations of the PM, which are dependent on preceding changes in the extensibility of the cell wall. In pollen tubes a candidate SAC has been identified that appears to localize specifically to membrane patches derived from the growing tip of lily pollen tubes, and that selectively conducts Ca²⁺ (Dutta and Robinson 2004). Within the context of oscillatory pollen tube growth this channel emerges as the best candidate to account for the corresponding oscillations in the tip-focused Ca²⁺ gradient. It seems plausible, therefore, that as the apical cell wall relaxes and the pollen tube extends, the PM in this specific location would deform due to the force of turgor pressure. The SAC would be activated and Ca²⁺ entry would occur, being tightly localized to those regions of the PM experiencing deformation. Not only does this scenario explain the spatial positioning of the tip-focused Ca²⁺ gradient, but it accounts for its temporal properties in relation to changes in growth rate. Thus as the cell extends, and the PM deforms, a Ca²⁺ rise occurs, which slightly follows the increase in growth rate, as observed. Similarly as growth slows, PM deformation lessens, the SAC closes and the tip-focused gradient declines (Dutta and Robinson 2004). Although CNGC-18 may contribute to the formation and oscillation of the tip focused Ca²⁺ gradient (Frietsch et al. 2007), we think that the SAC, described by Dutta and Robinson (2004) is a much better candidate. For example, while CNGC-18 localizes to the apical PM, its distribution is rather broad, extending across the dome and partly down the flank of the pollen tube apex. There is no evidence that it would give rise to the focused high [Ca²⁺] that is observed. Also the rise and fall of the tip-focused Ca²⁺ gradient, which is tightly coupled to changes in the growth rate, is elegantly explained by an apically positioned SAC that is activated by wall loosening and turgor-dependent extension of the wall and associated PM. While the studies of Dutta and Robinson (2004) provide physiological evidence for the presence of a stretch activated channel in the PM of pollen tubes, more recent work on Arabidopsis provides molecular evidence for the existence of a SAC. Nakagawa et al. (2007) have isolated and characterized an Arabidopsis cDNA, MCA1, which complements a lethal phenotype in the yeast Saccharomyces for a stretch activated Ca²⁺ channel. In Arabidopsis this channel increases [Ca²⁺] when the plasma membrane is deformed.

Calmodulin

A persuasive case is building for a role for extracellular CaM in the control of plant cell growth and development. Ma and Sun (1997) showed that an anti-CaM serum, and the CaM antagonist, W-7 strongly inhibited pollen germination and tube growth, and that the inhibitory effects could be reversed by the addition, up to 1.0 μM, of purified cauliflower CaM. In further studies Ma et al. (1999), show that heterotrimeric G-proteins may be involved in the signal transduction scheme for CaM. Using microinjection, they show that the addition of GDP-β-S or an anti-G_α antibody, both of which block heterotrimeric G-proteins, also slows pollen tube growth. In addition they report that anti-CaM inhibition can be reversed by cholera toxin, a G-protein agonist, or that CaM stimulation can be inhibited by pertussis toxin, a G-protein antagonist. Taken together they suggest that heterotrimeric G-proteins participate in the signal transduction pathway that involves extracellular CaM as a regulator of pollen tube growth. A possible downstream target has been identified in suspension-culture cells of Angelica; this is a 21 kDa CaM binding protein that is localized in the cell wall (Mao et al. 2005).

More recent studies indicate that CaM, although present in the apoplasm, specifically binds to the outer surface of the PM, and that it stimulates the influx of Ca²⁺. Binding at the PM has been shown most convincingly with a quantum dot-CaM probe, using both fluorescence and transmission electron microscopy. With ion imaging, Shang et al. (2005) first showed that extracellular addition of purified cauliflower CaM stimulates Ca²⁺ entry into pollen grains. In an extension of that study, Wang et al. (2009) report that the quantum dot-CaM probe also stimulates Ca²⁺ influx. In parallel with results noted above, Shang et al. (2005) further show that application of a CaM anti-serum, or the CaM antagonist W-7, inhibits the increase in intracellular Ca²⁺. Patch clamp studies further indicate that Ca²⁺ influx occurs through a voltage gated channel, which based on its characteristics, appears to belong to the category of hyperpolarization-activated Ca²⁺-permeable channels that participate in growth regulations at potentials more negative than −100 to −150 mV (Shang et al. 2005).

Because CaM is synthesized in the cytosol it is a puzzle how it is delivered to the apoplasmic space, or how its concentration might change to cause the reported changes in intracellular Ca²⁺. It is conceivable that during secretion CaM leaks out as a result of microlesions in the PM. Because exocytosis oscillates, so too would the transport of CaM to the wall. For the pollen tube this would translate into an oscillatory profile for extracellular CaM in the apical apoplasm. In turn this could generate a positive feedback loop wherein the process of exocytosis delivers more CaM to the wall, which then causes further Ca²⁺ influx into the apical cytosol.

Annexins

The annexins are soluble, multifunctional proteins usually found in the cytosol, but also associated with membranes and sometimes integral components of the membrane (Bushart and Roux 2007; Mortimer et al. 2008; Laohavisit et al. 2009). They bind Ca²⁺, hydrolyze nucleotides, interact with membrane phospholipids, exhibit peroxidase activity, bind to the actin cytoskeleton, and in general seem capable of myriad functions. We previously made reference to their role in facilitating secretion; here we draw attention to the observations showing that these proteins are themselves secreted and that they can affect the activity of the PM from the apoplasm (Laohavisit et al. 2009). Somewhat similarly to CaM, annexins can facilitate the influx of Ca²⁺ into the cytosol. However in this instance they appear to do so by inserting in the membrane and constructing conductance channels through which Ca²⁺ can pass (Laohavisit et al. 2009). The annexins, together with CaM, appear able to establish a positive feedback loop of reinforcing signals. Indeed the annexins may facilitate their own secretion and through their subsequent activity as a stimulator of Ca²⁺ influx, they may be able to promote further secretion or at least bias secretion to the regions where the annexins are active.

Wall associated kinases

The cell wall associated kinases (WAKs) are potentially key elements in the continuum between the cell wall and the cytoplasm (Kohorn 2000; Kohorn et al. 2006a, 2006b). These proteins, which occur widely in plants including pollen tubes, consist of a single transmembrane domain with a serine/threonine kinase on the cytoplasmic side and an extracellular fragment that associates with the wall polymers in the apoplasm. In the wall the WAKs bind to pectins, in a manner that is dependent on the Ca²⁺ cross-linked or “egg-box” configuration of de-esterified galacturonic acid residues (Decreux and Messiaen 2005). Reducing the [Ca²⁺] in the wall, or more precisely the Ca²⁺/monovalent ion ratio that retards egg-box formation, specifically inhibits the binding of WAK1. The association between pectin and WAKs may occur very early in development as studies show that these components appear together in the Golgi apparatus and are secreted as a complex (Kohorn et al. 2006a). Although we lack a great deal of information about how these kinases operate, some recent evidence suggests that they may affect sugar metabolism (Kohorn et al. 2006b). Specifically it has been reported that mutants in WAK2, which exhibit a dwarf phenotype, are defective in vacuolar invertase activity. Because cells experience a slight decrease in turgor pressure as a loosened wall extends, Kohorn et al. (2006b) suggest that the reduced invertase activity may prevent mutant cells from restoring normal turgor pressure in association with the cell elongation process. While much remains to be discovered the WAKs emerge as extremely interesting candidates for transducing a signal from the cell wall to the cytoplasm that is influenced by the Ca²⁺ poise within the wall.

Oligogalacturonides

An important development has been the realization that certain cell wall fragments, notably the α 1, 4 oligogalacturonides (OGs), derived from pectins, serve as signaling agents in response to plant pathogens and also for normal development (Lecourieux et al. 2006). Although we are not aware of specific information that applies directly to pollen tubes, given the pectin rich cell wall and the likely presence of the appropriate OGs, their role in the pollen tube growth seems entirely plausible. Very briefly these OGs, with a degree of polymerization of 10 to 15 and exhibiting the Ca²⁺ cross-linked egg box configuration, are able to stimulate an intracellular increase in the [Ca²⁺]. For example, in studies in aequorin transformed Arabidopsis cells, the application of OGs induces a sharp rise in Ca²⁺. This rise can be blocked by lanthanum, which is a Ca²⁺ channel blocker or by tetrabromobenzotriazole, which is a protein kinase inhibitor, indicating that you can prevent the cellular response either by preventing Ca²⁺ from entering the cell at the channel or by blocking a kinase activity inside the cell (Moscatiello et al. 2006). Moscatiello et al. (2006) further performed a transcript analysis following OG application and their results draw attention to changes in cell wall modifying factors as well as to factors involved in the biosynthesis of jasmonate. To some degree there appears to be an intersection between the WAK requirement, mentioned in the previous section, and OGs discussed here; both involve a role for Ca²⁺ cross-linked pectin fragments exhibiting the egg-box configuration and a degree of polymerization of 10 to 20. Additional work on these pectic wall fragments could provide important insight about the ability of cell wall to transmit signals to the cytoplasm.

Extracellular ATP

Finally we draw attention to extracellular ATP (eATP), which is yet another factor that is able to modulate intracellular Ca²⁺ (Demidchik et al. 2003; Tang et al. 2003; Jeter et al. 2004; Kim et al. 2006; Song et al. 2006; Foresi et al. 2007; Roux and Steinebrunner 2007; Demidchik et al. 2009). As with CaM we normally think of ATP as residing entirely in the cytoplasm in the mM level and where it serves as a universal energy currency for the cell. However it is well known in animal cells and now emerging in plants including pollen tubes (Steinebrunner et al. 2003) that ATP occurs in the extracellular or apoplasmic domain, where its concentration (0.3–40.0 μM) is much less than in the cytoplasm (Roux and Steinebrunner 2007). Consequently eATP probably does not energize reactions; rather it appears to act as a signaling agent. Specifically eATP has been shown to increase the intracellular [Ca²⁺].

Because ATP is synthesized in the cytoplasm a question arises on how it moves to the apoplasmic space. Different possibilities emerge including leakage through wound sites, or osmotic and mechanical stress lesions, and through secretion itself, where the ATP will be trapped in vesicles along with other material that is being exocytosed. Different studies show that the highest [eATP] occurs in association with amplified exocytosis, as for example at the tip of root hairs (Kim et al. 2006; Wu et al. 2008). Once in the apoplasm eATP is partly inactivated by the enzyme apyrase. It might seem that these conditions would completely destroy the activity of eATP; however, that is not true because studies on pollen germination reveal that grains with much reduced apyrase activity are prevented from growing (Steinebrunner et al. 2003). From these observations Roux and Steinebrunner (2007) suggest that the [eATP] must be at an intermediary level between low and high extremes to support normal growth. The activity of eATP also depends upon the availability of Ca²⁺ in the extracellular matrix. If Ca²⁺ is chelated, eATP cannot stimulate an increase in the intracellular [Ca²⁺] (Demidchik et al. 2003; Jeter et al. 2004).

With all of the agents that stimulate an increase in [Ca²⁺] much remains to be learned about the identity and function of the downstream targets. A recent study makes a compelling link between the rise in intracellular [Ca²⁺] induced by eATP, and the stimulation of a reduced nicotinamide adenine dinucleotide (NADPH) oxidase-dependent increase in reactive oxygen species. Here the train of events appears to flow from eATP to the activation of a PM NADPH oxidase, with an increase in reactive oxygen species. The latter agent then stimulates the opening of a Ca²⁺ channel, which causes the intracellular elevation of this ion (Demidchik et al. 2009). We acknowledge that research on reactive oxygen species attracts a great deal of interest these days owing to the central role that these toxic factors play as signaling agents in controlling response to pathogens, as well as normal growth and development (Mori and Schroeder 2004; Carol and Dolan 2006). Although we start to gain information on the pathway for eATP activity, we nevertheless lack knowledge of the presumed receptor. Continued research on this exciting topic is clearly warranted.

Conclusion

A considerable amount of information increasingly shows the existence of an extensive interplay between the cytoplasm and the cell wall, which contributes to the control of plant growth and development. We have long known that the cytoplasm gives rise to the cell wall through the secretion of the building blocks. More recently we realize that the cell wall can have a profound effect on the cytoplasm with Ca²⁺ participating in different regulatory pathways. Ca²⁺ has recently held our attention because of its central role as a second messenger to control many different processes. Now it becomes crucial to consider the interaction between the cytoplasmic Ca²⁺ and cell wall Ca²⁺ with the intervening PM occupying a pivotal position in the process of information transfer. The conversation, however, is definitely two ways; the cell wall through its ability to bind and control the [Ca²⁺] at the surface of the PM, influences the entry of the ion into the cytoplasm. But, through the activity of pumps the cytoplasm can extrude Ca²⁺ into the wall and affect local structure and activity of wall components. With the increasing number of factors that exert further control over this process, we come to recognize a complexity that could not be imagined a few years ago.

In the last section of this brief essay we have drawn attention to a few factors, namely SACs, CaM, annexins, WAKs, OGs, and ATP that either occur in the apoplasm, or interact with the cell wall to produce a signal that requires Ca²⁺ or influences its translocation in ways that affect growth and development. With the exception of the WAKs, a common theme appears to emerge that ties these factors together, namely the ability to stimulate an increase in the intracellular [Ca²⁺]. Although we have ideas about the specific signatures of the Ca²⁺ responses for SACs, we know much less about the other factors that modulate the intracellular [Ca²⁺]. For example, are the Ca²⁺ elevations restricted to specific locations in the cell, do they oscillate and if so their amplitude and period, and are they temporally brief or prolonged? This information could provide clues about the downstream targets, and begin to make it possible to sort out different pathways. We also imagine that additional factors will emerge. When considered as a whole it becomes evident that there is an extensive system involving the interaction between the cell wall and the cytoplast, with Ca²⁺ as an intermediary. Resolving the details will provide insight about the control of plant growth and development.

(Co-Editor: Kurt V. Fagerstedt)

Acknowledgements

We thank our colleagues at the University of Massachusetts at Amherst for stimulating discussion on the ideas presented in this article. This work was supported by a grant from the USA National Science Foundation (MCB-0846876).