Signaling and the Modulation of Pollen Tube Growth

Apical Ca²⁺ Gradients in Growing Pollen Tubes

Several independent studies using ratiometric Ca²⁺ imaging of pollen have established that normally growing pollen tubes possess a steep tip-focused [Ca²⁺]_i apical gradient (Obermeyer and Weisenseel, 1991; Rathore et al., 1991; Miller et al., 1992; Malhó et al., 1994; Pierson et al., 1994, 1996; Franklin-Tong et al., 1997; Holdaway-Clarke et al., 1997; Messerli and Robinson, 1997). As illustrated in Figure 2A, a high apical concentration of [Ca²⁺]_i occurs in growing pollen tubes, but it is not detected in pollen tubes that are not growing. Moreover, treatments that eliminate the apical [Ca²⁺]_i gradients result in the reversible inhibition of pollen tube growth (Rathore et al., 1991; Pierson et al., 1994; Li et al., 1996; Malhó and Trewavas, 1996). It is now generally accepted that apical [Ca²⁺]_i gradients are likely to be a fundamental phenomenon that is common to all growing pollen tubes.

Considerable evidence from a variety of experimental approaches supports the idea that the apical [Ca²⁺]_i gradient results from localized Ca²⁺ influx through active Ca²⁺ channels at the pollen tube tip (Weisenseel and Jaffe, 1976; Kühtreiber and Jaffe, 1990; Malhó et al., 1994, 1995; Pierson et al., 1994; Holdaway-Clarke et al., 1997; Messerli and Robinson, 1997). Both ratiometric imaging and measurements of extracellular Ca²⁺ influx indicate that Ca²⁺ entry is restricted to a small region associated with Ca²⁺ channel activity at the extreme apex of the pollen tube (Pierson et al., 1996). It is generally thought that when growth ceases, the apical gradient is dissipated as these channels close. The nature of these pollen tube Ca²⁺ channels is currently not known. There is speculation that they may be stretch activated (Pierson et al., 1994; Malhó et al., 1995), but further work is required to determine whether or not this is the case.

The apical [Ca²⁺]_i gradient is very steep and is accepted to be ∼2 to 10 μM at the tip, dropping to basal levels of ∼200 nM within ∼20 μm behind the apical region (Obermeyer and Weisenseel, 1991; Rathore et al., 1991; Miller et al., 1992; Malhó et al., 1994; Pierson et al., 1994, 1996; Franklin-Tong et al., 1997). The dissipation of [Ca²⁺]_i behind the apex is thought to be regulated by Ca ²⁺-ATPases, which are likely to be located on the tubular ER close to the apical region of pollen tubes (Figure 1; Obermeyer and Weisenseel, 1991; Lancelle and Hepler, 1992; see Sze et al., 1999, in this issue). This organelle represents a large sink, and it is capable of rapidly sequestering Ca²⁺. Although there are low-level fluctuations in [Ca²⁺]_i in the “shank” region of the pollen tube (Figure 1), they are generally within basal levels.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Imaging of Changes in Cytosolic Calcium ([Ca²⁺]_i) in Pollen Tubes.

(A) Growing and nongrowing pollen tubes. Ratio imaging of [Ca²⁺]_i in a growing pollen tube of Papaver rhoeas microinjected with fura-2 dextran. The top image shows a growing pollen tube, with a high apical [Ca²⁺]_i gradient. The lower image shows a nongrowing tube displaying no apical [Ca²⁺]_i gradient. Note that the basal Ca²⁺_i levels for both of these pollen tubes are very similar. Reproduced from Franklin-Tong et al. (1997), with the permission of Blackwell Science.

(B) Growing pollen tubes exhibit oscillations in [Ca²⁺]_i at the tip. Ratio imaging of [Ca²⁺]_i in a growing pollen tube of Lilium longiflorum microinjected with fura-2 dextran. Time, in seconds, is given on each image. Note that the apical [Ca²⁺]_i gradient is not constant but oscillates over time (Holdaway-Clarke et al., 1997).

(C) Reorientation of an Agapanthus pollen tube. Calcium Green-1 was microinjected into a pollen tube of A. umbellatus together with caged Ca²⁺ (Nitr-5). [Ca²⁺]_i was imaged using laser confocal scanning microscopy (LCSM) before treatment (first two images) and after release of caged Ca²⁺ in a region to the left of the pollen tube tip (indicated by the arrow). Increases in [Ca²⁺]_i in the tip region are detected coincident with the reorientation of pollen tube growth over time (given in seconds toward the top of each image). Based on a figure originally published in Malhó and Trewavas (1996). Bar = 10 μm.

(D) Ca²⁺ wave propagated in a P. rhoeas pollen tube after release of caged Ins(1,4,5)P₃. Calcium Green-1 was microinjected together with caged Ins(1,4,5)P₃, and Ca²⁺_i was imaged in a P. rhoeas pollen tube by using LSCM. The top image shows the pollen tube before treatment, after which caged Ins(1,4,5)P₃ was released throughout the imaged region by UV photolysis (indicated by the arrow). Note the slow increases in [Ca²⁺]_i, which “flood” toward the pollen tube tip in the form of a Ca²⁺ wave following Ins(1,4,5)P₃ release (Franklin-Tong et al., 1996). Elapsed time is given in seconds (s) to the right of each image.

(E) Ratio imaging of P. rhoeas pollen tubes during the SI response. Fura-2 dextran was microinjected into two pollen tubes to visualize alterations in [Ca²⁺]_i in response to the addition of incompatible stigmatic S proteins over time (indicated in seconds [s] after the addition). (i) Alterations in [Ca²⁺]_i in the apical and subapical region of a pollen tube. The oscillating tip-focused gradient is lost within 60 sec. In contrast, [Ca²⁺]_i in the subapical region continues to increase. (ii) Changes in [Ca²⁺]_i appear to originate in the shank of the pollen tube, in a region ∼100 μm behind the tip. Within 1 to 2 sec (indicated as 0 s), [Ca²⁺]_i increases from basal levels to ∼300 nM, and localized regions rapidly reach 1 μM Ca²⁺. [Ca²⁺]_i redistribution occurs and is especially notable at 80 and 100 sec; these changes appear to take the form of a Ca²⁺ wave. Reproduced from Franklin-Tong et al. (1997), with the permission of Blackwell Science.

Apical Ca²⁺ Oscillates in Growing Pollen Tubes

Recent observations, made by using both ratiometric imaging and photon counting of the photoprotein aequorin, have provided good evidence that apical Ca²⁺ in growing pollen tubes oscillates over time, as illustrated in Figure 2B (Pierson et al., 1996; Calder et al., 1997; Franklin-Tong et al., 1997; Holdaway-Clarke et al., 1997; Messerli and Robinson, 1997). This oscillatory behavior is closely associated with growth, and the oscillations in [Ca²⁺]_i and Ca²⁺ influx at the pollen tube tip have the same periodicity as the pollen tube growth rate (20 to 40 μm sec^-1), which also pulses (Pierson et al., 1996; Holdaway-Clarke et al., 1997; Messerli and Robinson, 1997). This observation has resulted in the suggestion that there is a direct causal link between the two kinds of pulsation, although there remains debate as to whether the changes in [Ca²⁺]_i drive the oscillations in growth or whether they are caused by them. Data that have been taken to support the former possibility come from experiments demonstrating that increases in Ca²⁺ can trigger the onset of Ca²⁺ oscillations in growth-arrested pollen tubes and that these events can stimulate tip swelling and the resumption of growth (Calder et al., 1997). On the other hand, Holdaway-Clarke et al. (1997) have provided evidence that pollen tube growth rate and the [Ca²⁺]_i gradient oscillate in phase, whereas pulses in Ca²⁺ influx appear to lag by ∼11 sec, implying that Ca²⁺ influx follows growth.

Holdaway-Clarke et al. (1997) have proposed two models that could account for the observed delay between the peak of the Ca²⁺_i gradient and the peak influx of extracellular Ca²⁺ in the apical region. In the first model, they propose that a brief opening of stretch-activated Ca²⁺ channels allows the entry of a small amount of Ca²⁺. This triggers calcium-induced calcium release from internal stores (e.g., the ER), which increases the [Ca²⁺]_i gradient. However, there are difficulties with this model, particularly because the ER is not especially concentrated in the tip region and is not closely associated with the plasma membrane (Figure 1; Lancelle and Hepler, 1992). In the second model, which is depicted in Figure 3 and is discussed below, it is suggested that the cell wall acts as a Ca²⁺ store and that periodic changes in the ion binding properties of the pectin in the cell wall could regulate the rate of Ca²⁺ influx, resulting in the observed oscillations (Holdaway-Clarke et al., 1997). Although these models are speculative at present, they do provide hypotheses to test.

What Is the Function of the Apical [Ca²⁺]_i Gradient?

The precise function of the apical [Ca²⁺]_i gradient remains unclear, although it seems likely that it is involved in the regulation of vesicle secretion in the apical dome of the pollen tube. It is known that high levels of Ca²⁺ are associated with secretion (Blackbourn and Battey, 1993) and that annexins, which are Ca²⁺ and phospholipid binding proteins that play an important function in secretion, are localized in the apical dome of the pollen tube tip (Blackbourn et al., 1992; Battey and Blackbourn, 1993; Battey et al., 1999, in this issue). Elevated [Ca²⁺]_i at the tip is predicted to promote vesicle exocytosis and hence cell elongation. It is therefore tempting to speculate that the colocalization of high Ca²⁺ and high annexin at the pollen tube apex may represent a functional and possibly direct relationship among annexins, Ca²⁺ influx, and vesicle fusion, all of which are required for pollen tube tip extension. Similarly, because microfilament assembly and disassembly are known to be regulated by Ca²⁺ and Ca²⁺-dependent enzymes, it is also considered likely that the pollen tube cytoskeleton is directly or indirectly influenced by fluctuations in the apical Ca²⁺ gradient.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

The External Stores Model for Oscillatory Growth in Pollen Tubes.

This model seeks to explain the oscillatory nature of pollen tube growth, which is out of phase with Ca²⁺ influx. See text for details. Ca²⁺, extracellular Ca²⁺_o; Max, maximum; Min, minimum; PM, plasma membrane; PME, pectin methylesterase; X-links, cross-links; up and down arrows indicate increase and decrease, respectively. Reproduced from Holdaway-Clarke et al., 1997.

The Ca²⁺–pectin model mentioned above (Figure 3; Holdaway-Clarke et al., 1997) has the attractive feature that it attempts to explain oscillatory growth directly in terms of the consequences of the action of high [Ca²⁺]_i. These authors hypothesize that pectin methylesterases (PMEs), which are responsible for cross-linking unesterified pectins, control the rate of production of cell wall binding sites for Ca²⁺. As shown in Figure 3, when incoming Ca²⁺ binds to the wall, the binding of Ca²⁺ also releases H⁺. The resulting acidification could deactivate PME activity, and assuming that exocytosis continues, unesterified methylpectin would be pushed into the tip zone and would stretch the plasma membrane here, leading to tip extension. This could also result in the opening of stretch-activated Ca²⁺ channels and the entry of Ca²⁺ ions at the tip. Such a system would be self-regulating, because incoming Ca²⁺ would be sequestered into the cell wall, the [Ca²⁺]_i gradient would decrease, and the free Ca²⁺ in the cell wall would result in the cross-linking of pectins, which would slow down exocytosis and growth.

Support for this model comes from studies demonstrating that the pulsing growth of pollen tubes can be accompanied by the deposition of distinct cell wall bands. Wall material deposited during periods of slow growth are enriched with acidic pectins and arabinogalactan proteins and contain relatively less esterified pectin (Li et al., 1992, 1994; Geitmann et al., 1995). Although there is now general acceptance that tip-focused [Ca²⁺]_i oscillations in pollen tubes are likely to be a general phenomenon, the relationship between Ca²⁺ influx and growth needs further clarification, as does the exact function of the Ca²⁺ gradient.

Which Way to Go? Ca²⁺ Is Involved in the Reorientation of Growing Pollen Tubes

How the perception of extracellular signals influences the modulation of growth patterns and the directionality of plant cells is currently a key question in the plant signaling field. Although the study of “normal” pollen tube growth has provided important information in recent years about the mechanisms that control pollen tube growth, investigations of the signals that may alter growth are also beginning to provide insights. In particular, by studying the inhibition and reorientation of pollen tubes, some of the signaling pathways that modulate pollen tube growth have been identified. Below, I discuss some of the evidence that supports the involvement of Ca²⁺ signaling in modulating pollen tube growth in response to known stimuli.

Pollen tubes undergo necessary changes in direction during their progress through the pistil before they effect fertilization. Indeed, the nature of the signals, signaling pathways, and the mechanisms involved in the reorientation of pollen tube growth are of central importance to an understanding of this process. A series of studies by Malhó (Malhó et al., 1994, 1995; Malhó and Trewavas, 1996) using Ca²⁺ imaging of Agapanthus pollen tubes has begun to advance our knowledge on this front. These studies have established a role for [Ca²⁺]_i in the reorientation of Agapanthus pollen tubes. For example, Figure 2C shows the effect on pollen tube orientation provoked by releasing caged Ca²⁺. Local increases in [Ca²⁺]_i apparently perturb the polarity of the pollen tube; growth stops temporarily (the apical Ca²⁺ gradient is lost), and the tip swells. Tip swelling coincides with the establishment of a new apical Ca²⁺ gradient and with the resumption of growth, which usually occurs in a new direction (Figure 2C; Malhó et al., 1994, 1995). On the question of how the direction of reorientation may be controlled, Malhó and Trewavas (1996) have shown that the direction and angle of reorientation could be predicted accurately by locally increasing [Ca²⁺]_i within the pollen tube, as shown in Figure 2C, and also by locally altering external Ca²⁺.

These data strongly suggest that Ca²⁺ channel activity in the apical dome plays a critical role in determining pollen tube reorientation. Pierson et al. (1996) have suggested that there is a region of higher Ca²⁺ channel activity within the apical dome that defines the precise point from which elongation will proceed. Malhó and Trewavas (1996) have shown that Ca²⁺ channel activity is localized to the first 20 μm of the pollen tube and that alterations in the exact localization of this activity can influence directional growth in pollen tubes. The nature of the type of Ca²⁺ channels, as mentioned earlier, awaits further investigation. Together, these data suggest that a Ca²⁺-based signal transduction pathway can control the direction of pollen tube growth. However, it should be borne in mind that there are likely to be many other factors controlling directional growth of the pollen through the pistil tissues toward the ovary. Some of these factors are discussed later.

A Role for Phosphatidylinositol-(1,4,5)-Triphosphate (Ins[1,4,5]P₃) in the Regulation of [Ca²⁺]_i in Pollen Tubes

So far I have discussed the central role for [Ca²⁺]_i in modulating pollen tube growth and the potential roles of the cell wall and the ER as Ca²⁺ pools. Nevertheless, the signals involved in Ca²⁺ release remain, for the most part, unidentified. However, there is emerging evidence that a functional phosphoinositide signal transducing system operates in growing pollen tubes and that this pathway may be involved in coordinating and regulating pollen tube growth.

The presence of phosphoinositides and phosphatidylinositol phospholipase C activity in pollen tubes of Lilium longiflorum was demonstrated some years ago (Helsper et al., 1987). This finding was recently confirmed when it was demonstrated that a Ca²⁺-dependent Ins(4,5)P₂-specific phospholipase C activity is present in pollen of Papaver rhoeas (Franklin-Tong et al., 1996).

Further evidence comes from data showing that changes in intracellular concentrations of Ins(1,4,5)P₃, as illustrated in Figure 2D, can stimulate large increases in [Ca²⁺]_i in growing pollen tubes, which initiate in the “nuclear region” and are propagated toward the pollen tube tip (Franklin-Tong et al., 1996). It has been suggested that these waves are mediated, at least in part, by Ins(1,4,5)P₃-induced Ca²⁺ release, and a model describing this possibility proposes that the waves are a result of a series of Ins(1,4,5)P₃-generating and Ca²⁺-mobilizing reactions (Franklin-Tong et al., 1996). Although there is good evidence for the presence of Ins(1,4,5)P₃-sensitive Ca²⁺ stores in the shank of pollen tubes (Franklin-Tong et al., 1996; Malhó, 1998), their nature remains to be established.

That a functional phosphoinositide signal transducing system involving Ins(1,4,5)P₃-stimulated increases in [Ca²⁺]_i plays a role in both reorientation and inhibition of pollen tube growth is suggested by data showing that changes in intracellular concentrations of Ins(1,4,5)P₃ provoke many of the changes in pollen tube tip morphology and growth that were described earlier. One could envisage a two-tier level of control involving phosphoinositide-mediated signals modulating pollen tube growth. In addition to a role for Ins(1,4,5)P₃-stimulated increases in [Ca²⁺]_i in inhibiting pollen tube growth, low-level phosphoinositide turnover appears to be required for normal pollen tube growth (Franklin-Tong et al., 1996).

The SI Response in P. rhoeas Pollen Is Mediated by Ca²⁺

Further evidence for the central role of [Ca²⁺]_i in modulating pollen tube growth also comes from studies of the SI response in P. rhoeas, in which inhibition of “self” pollen in response to a precise, defined signal has been studied in some detail. Several alleles of the stigmatic S gene have been cloned, and both stigmatic extracts and recombinant S proteins have been shown to possess the expected S allele–specific biological activity, that is, they inhibit incompatible pollen soon after germination (Foote et al., 1994; Franklin et al., 1995; Walker et al., 1996). The current model for the initial events in the SI response proposes that the stigmatic S proteins, which act as signal molecules, bind a membrane-bound pollen receptor in an S allele–specific manner, triggering a Ca²⁺-dependent signal transduction pathway.

Calcium imaging of pollen tubes challenged with S proteins has provided good evidence that Ca ²⁺ acts as a second messenger responsible for mediating inhibition of pollen tube growth, usually soon after germination, during the SI response (Franklin-Tong et al., 1993, 1995, 1997). This point is illustrated in Figure 2E. Transient increases in [Ca²⁺]_i, which are triggered virtually immediately and last several minutes, can be detected only when biologically active S proteins are used in combination with incompatible pollen.

What is the effect of these increases in [Ca²⁺]_i? Data from several different experiments suggest that increases in [Ca²⁺]_i can inhibit pollen tube growth and that the S proteins act as signal molecules and achieve their effect by stimulating alterations in [Ca²⁺]_i (Franklin-Tong et al., 1993, 1995, 1997). For example, early imaging studies of pollen tubes responding to the SI reaction suggested the S allele–specific [Ca²⁺]_i increases were localized in the shank of the pollen tube (Franklin-Tong et al., 1993, 1995; Drøbak et al., 1997). More recent studies using ratiometric Ca²⁺ imaging have allowed alterations in [Ca²⁺]_i to be quantified. These studies have provided further evidence that Ca²⁺ waves are elicited by the SI response (Franklin-Tong et al., 1997) and that increases in [Ca²⁺]_i are detected in the subapical and shank regions of the pollen tube, as shown in Figure 2E. Unexpectedly, after some fluctuation, the tip-focused gradient at the pollen tube tip diminishes to basal levels within ∼1 min (Franklin-Tong et al., 1997). This suggests that, whereas alterations in [Ca²⁺]_i in the apical region are crucial for some processes regulating pollen tube growth, alterations in [Ca²⁺]_i in other regions of the pollen tube also play a role. The mechanisms involved, the source of the Ca²⁺, and the nature of the Ca²⁺ waves are, at present, not known and require further investigation.

The rapidity of the SI response suggests that Ca²⁺ is acting as a second messenger and that increases in [Ca²⁺]_i are likely to be one of the first events in the signaling pathway. Ca²⁺-dependent increases in protein phosphorylation triggered by the SI response, which are likely to be downstream of these initial Ca²⁺ signals, have been identified (Rudd et al., 1996, 1997). The increased phosphorylation of a 26-kD cytosolic protein, p26.1, specifically induced by the SI response, appears to be Ca²⁺ and calmodulin dependent (Rudd et al., 1996). This suggests that a Ca²⁺-dependent protein kinase requiring calmodulin-like domains is activated during the SI response and that it acts as an intracellular signal mediating the SI response in P. rhoeas pollen. These observations provide compelling support for the idea that the S proteins profoundly modulate Ca²⁺ homeostasis.

As an indication of the potential complexity of this signaling pathway, p68, another S-specific phosphorylation-responsive pollen protein, has been identified (Rudd et al., 1997). The timing of p68 phosphorylation suggests that its phosphorylation is downstream of p26.1. Surprisingly, however, the kinase(s) responsible for the phosphorylation of p68 is not Ca²⁺ dependent. This suggests that a “second wave” of Ca²⁺-independent signaling may follow the initial Ca²⁺-dependent SI signaling.

Protein Kinases in Pollen

Although the evidence for signaling, and Ca²⁺-dependent signaling in particular, in pollen–pistil interactions is now good, there are remarkably little data relating to the activities of protein kinases in pollen. However, as the above discussion illustrates, it is becoming apparent that Ca²⁺-mediated signaling cascades involving pollen protein kinases are likely to play a major role during pollination. Ca²⁺-dependent protein kinases were first identified in germinated pollen of Nicotiana alata (Polya et al., 1986). Since then, a pollen-specific, Ca²⁺-dependent, calmodulin-independent protein kinase (CDPK), which is thought to be required for pollen germination and growth, has been cloned from maize pollen (Estruch et al., 1994). This CDPK may be involved in altering the cytoskeletal dynamics required for pollen tube growth.

More recently, soluble and microsomal Ca²⁺-dependent protein kinases, which share characteristics of CDPKs, have been identified in N. alata (Kunz et al., 1996). One of these, Nak-1, has been shown to phosphorylate the stylar S-RNases involved in SI in N. alata. Alterations in the phosphorylation status of pollen proteins during incompatible and compatible pollinations have also been reported in rye and in the Brassicas (Wehling et al., 1994; Hiscock et al., 1995). Recent data suggest that Ca²⁺-dependent protein kinase activity in pollen tubes may be localized in the apical region (Moutinho et al., 1998).

A number of receptor-like protein kinases, including PRK1, LePRK1, and LePRK2, have been cloned from pollen (Mu et al., 1994; Muschietti et al., 1998). These are all functional protein kinases, and although their biological function has not been determined, it has been suggested that PRK1 may have a role in pollen development (Mu et al., 1994), whereas LePRK1 and LePRK2 appear more likely to play roles in pollen–pistil interactions (Muschietti et al., 1998). The existence of functional protein kinases with an extracellular receptor domain clearly provides good evidence that signals are transduced into the pollen grain or tube, presumably from the pistil. The next challenge will be to identify and characterize in detail the pistil components with which the pollen interacts, the nature of the signaling pathways triggered by the interaction, and the resulting responses.

Rho and Small GTP Binding Proteins

The Rho family of GTPases function as key molecular switches, controlling a variety of actin-dependent cellular processes in diverse eukaryotic organisms. Recent work provides good evidence that Rho-type GTPases play a key role in the control of polarized growth of pollen tubes (Lin et al., 1996; Lin and Yang, 1997). For example, a Rho-type GTPase, Rop1, is expressed predominantly in pollen and pollen tubes (Lin et al., 1996), and its localization in a tip–base gradient concentrated toward the apical region implies a role in tip growth and movement of the generative cell in pollen tubes. Moreover, because it is also distributed toward the periphery of the generative cell, where it colocalizes with myosin, Rop GTPase may have a role in the modulation of an actinomyosin motor system involved in the movement of the generative cell (Lin et al., 1996).

Further evidence that Rop may be involved in controlling tip growth comes from elegant functional studies in which microinjected anti-Rop antibodies inhibited pollen tube growth (Lin and Yang, 1997). Not only did these studies provide the first direct demonstration that microinjected antibodies can be used as tools to study pollen tube growth, they also provided good evidence that Rop may regulate a Ca²⁺-dependent pathway involved in vesicle docking or fusion, whereas a distinct Rho GTPase may mediate cytoplasmic streaming. Further studies aimed at unraveling the signaling pathways and the structural elements with which they interact should greatly enhance our understanding of the apparently complex processes involved in the modulation of pollen tube tip growth.