Involvement of extracellular calcium influx in the self-incompatibility response of Papaver rhoeas

Authors

  • Vernonica E. Franklin-Tong,

    Corresponding author
    1. 1 School of Biosciences, University of Birmingham, Edgbaston, Birmingham, B15 2 TT, UK, and2Department of Biology, University of Massachusetts, Amherst, Massachusetts 01003, USA
      For correspondence (fax +44 121 414 5925; e-mail
    Search for more papers by this author
  • 1, Terena L. Holdaway-Clarke,

    1. 1 School of Biosciences, University of Birmingham, Edgbaston, Birmingham, B15 2 TT, UK, and2Department of Biology, University of Massachusetts, Amherst, Massachusetts 01003, USA
    Search for more papers by this author
  • 2 Kornelis R. Straatman,

    1. 1 School of Biosciences, University of Birmingham, Edgbaston, Birmingham, B15 2 TT, UK, and2Department of Biology, University of Massachusetts, Amherst, Massachusetts 01003, USA
    Search for more papers by this author
      Present address: Department of Genetics, University of Leicester, University Road, Leicester LE1 7RH, UK.
  • 3 Joseph G. Kunkel,

    1. 1 School of Biosciences, University of Birmingham, Edgbaston, Birmingham, B15 2 TT, UK, and2Department of Biology, University of Massachusetts, Amherst, Massachusetts 01003, USA
    Search for more papers by this author
  • and 2 Peter K. Hepler 2

    1. 1 School of Biosciences, University of Birmingham, Edgbaston, Birmingham, B15 2 TT, UK, and2Department of Biology, University of Massachusetts, Amherst, Massachusetts 01003, USA
    Search for more papers by this author

For correspondence (fax +44 121 414 5925; e-mail V.E.Franklin-Tong@bham.ac.uk).
Present address: Department of Genetics, University of Leicester, University Road, Leicester LE1 7RH, UK.

Summary

We have previously demonstrated that increases in cytosolic free Ca2+ are triggered by the self-incompatibility (SI) response in incompatible Papaver rhoeas (the field poppy) pollen. However, one key question that has not been answered is whether extracellular Ca2+ may be involved. To address this question, we have used an ion-selective vibrating probe to measure changes in extracellular Ca2+ fluxes around poppy pollen tubes. Our data reveal several findings. First, we confirm that there is an oscillating Ca2+ influx directed at the apex of the pollen tube; we also provide evidence that Ca2+ influx also occurs at the shanks of pollen tubes. Second, upon challenge with self-incompatibility (S) proteins, there is a stimulation of Ca2+ influx along the shank of incompatible pollen tubes, approximately 50 µm behind the pollen tube tip. This demonstration of SI-induced Ca2+ influx suggests a role for influx of extracellular Ca2+ in the SI response.

Introduction

Self-incompatibility (SI) is one of the most important devices whereby higher plants prevent inbreeding. It is a genetically controlled mechanism that regulates the acceptance or rejection of pollen that lands on the stigma. In Papaver rhoeas (the field poppy), SI is controlled by a single, multiallelic, gametophytically controlled S locus (Lawrence et al., 1978). Pollen grains carrying S alleles that are genetically identical to those carried by the pistil (incompatible) are discriminated from genetically different (compatible) pollen and selectively inhibited, thereby preventing self-fertilization.

Characterization and cloning of S genes and components involved in the SI reaction in P. rhoeas have begun to elucidate the molecular basis of SI. Several alleles of the stigmatic S gene have now been cloned (Foote et al., 1994; Kurup et al., 1998; Walker et al., 1996). They encode small, basic, highly polymorphic S proteins (approximately 14-kDa) that are developmentally expressed and secreted by the stigmatic papillar cells. Tip growth of incompatible pollen of Papaver rhoeas is inhibited within a few minutes of encountering the S proteins. Both stigmatic extracts and recombinant S proteins have been shown to have S-specific biological activity (Foote et al., 1994; Franklin-Tong et al., 1988). Although the nature of the allelic specificity remains to be elucidated, site-directed mutagenesis has established that certain residues are crucial for the recognition and inhibition of S1 pollen (Jordan et al., 1999; Kakeda et al., 1998). Further, the available data demonstrate that the stigmatic S-proteins act as signal molecules that trigger a signalling cascade in incompatible pollen tubes that results in their inhibition and includes increased phosphorylation of two soluble proteins, p26 and p68 (Rudd et al., 1996, 1997), alterations in the F-actin cytoskeleton (Geitmann et al., 2000; Snowman et al., 2000) and triggering of nuclear DNA fragmentation (Jordan et al., 2000).

Of special interest is the involvement of Ca2+-dependent signal transduction cascade(s) in pollen tube growth, inhibition and the SI response in incompatible P. rhoeas pollen. The importance of cytosolic free calcium (Ca2+i) for the regulation of pollen tube growth is well known (Franklin-Tong, 1999; Malhóet al., 1994, 1995; Miller et al., 1992; Obermeyer and Weisenseel, 1991; Pierson et al., 1994, 1996; Rathore et al., 1991; Rudd and Franklin-Tong, 1999). Inositol trisphosphate (Ins(1,4,5)P3)-induced Ca2+ release has also been shown to play a role in the regulation of pollen tube growth (Franklin-Tong et al., 1996; Malhó, 1998). Large, transient increases in [Ca2+]i have been observed in the ‘shank’ of incompatible pollen tubes (Franklin-Tong et al., 1993, 1995, 1997), implicating a Ca2+-mediated signalling pathway in the SI response. Corresponding decreases in intracellular calcium concentration ([Ca2+]i) in the tip occur concomitantly with SI-stimulated inhibition of tip growth (Franklin-Tong et al., 1997). The nature of the S-specific phosphorylation of p26 suggests that a Ca2+-dependent protein kinase requiring CaM-like domains is activated during the SI response, and acts as an intracellular signal mediating the SI response in P. rhoeas pollen (Rudd et al., 1996).

While these studies indicate that the arrest of pollen tube growth due to the SI reaction involves increases in [Ca2+]i, the nature and localization of the source(s) of Ca2+ involved in generating the SI-stimulated increases in [Ca2+]i has yet to be determined. The ion selective vibrating electrode (Kühtreiber and Jaffe, 1990; Smith et al., 1994) has previously been used effectively to demonstrate localized influx and efflux of Ca2+ in both animal and plant systems, including pollen tubes and root hairs (Cárdenas et al., 1999; Holdaway-Clarke et al., 1997; Kühtreiber and Jaffe, 1990; Messerli and Robinson, 1998; Messerli et al., 1999; Pierson et al., 1994, 1996). This method has several advantages over the more traditional technique of patch clamping for investigation of transmembrane ion fluxes. Most importantly, it is non-invasive, not requiring removal of the cell wall, and so may be used to study tip growth in living cells. Furthermore, in contrast to patch clamping, no unnatural changes in membrane potential are imposed and the natural conditions on both sides of the membrane are maintained during measurement by the extracellular ion selective vibrating probe. Using this approach we have measured extracellular Ca2+ fluxes around pollen tubes of P. rhoeas, and show here that poppy pollen tubes exhibit novel extracellular Ca2+ influxes along their shank that are stimulated during the SI response. This implicates a role for Ca2+ influx in the increases in [Ca2+]i previously reported as mediating the SI response in Papaver.

Results

Calcium fluxes in normally growing P. rhoeas pollen tubes

We have measured Ca2+ fluxes at the extreme apex of growing pollen tubes of P. rhoeas using the vibrating probe technique and show, as with other studies, that influx occurs and that it oscillates over time (n = 5), for example see Figure 1. Depending on the growth conditions of the pollen tubes, the fluxes varied considerably, though oscillations were still detectable when fluxes were small. The mean maximal Ca2+ influx at the peak of the oscillations was 4.64 pmol cm−2 sec−1 (SEM ± 1.83, n = 5).

Figure 1.

Oscillations of Ca2+ influx at the pollen tube tip.

Measurements of oscillating net influx of Ca2+ at the apex of a growing pollen tube of P. rhoeas were measured using a Ca2+–selective vibrating probe and are displayed graphically here. The plot shows net Ca2+ flux (pmol cm−2 sec−1), with net influx as a negative reading, and net efflux as a positive reading. The latter part of the plot shows a short background measurement 10 µm away from the pollen tube, in order to demonstrate that the fluxes measured are attributed to the pollen tube, and not noise.

In order to obtain a more thorough description of the overall pattern of Ca2+ current activity in growing pollen tubes, we measured the Ca2+ fluxes at various points along the shanks of growing pollen tubes of P. rhoeas using 2D mapping (n = 5). This approach enabled the mapping of both the magnitude and direction of the Ca2+ fluxes. Because we had previously observed increases in [Ca2+]i in the shank region approximately 50–100 µm behind the tip region during the SI response, we were particularly interested to ascertain if there was any evidence for Ca2+ fluxes in this region of the pollen tube. Figure 2 shows 2-D mapping of Ca2+ influxes along part of the length of a typical growing pollen tube shank, concentrating on this area of interest. Three of the average vectors were taken when adjacent to a nucleus. The SEs of these vectors were small, even at a 95% confidence interval (data not shown). We have ascertained that the probe was not measuring markedly different fluxes, by calculating the means and SEs for the fluxes at these points (data not shown). Although some of the fluxes measured seem to be parallel to the tube, there is also clear evidence for influx localized in the region of the nucleus, since the bounding flux vectors, left and right, in Figure 2 are small. There is therefore no strong evidence that the tube-parallel component observed in the longest vector has its source or sink in another region (e.g. the pollen grain, which is to the left in Figure 2). Notably there are relatively large extracellular Ca2+ fluxes in this ‘shank’, particularly in the region occupied by the male germ unit. An inwardly directed Ca2+ flux of 4.1 pmol cm−2 sec−1 was measured directly adjacent to the generative cell in Figure 2. In another pollen tube, a large inwardly directed Ca2+ flux of 5.02 pmol cm−2 sec−1 was measured in the region of the male germ unit. The standard errors on pollen tubes measured in this way were small; for the pollen tube in Figure 2, the average coefficient of variation for the vectors was 11%.

Figure 2.

Image of part of a pollen tube of Papaver rhoeas with vector plots of Ca2+ fluxes.

Vector plots (indicated by white lines) displaying Ca2+ fluxes along the shank of a typical growing pollen tube, measured using the Ca2+–selective vibrating probe. The site of each measurement is indicated by a small box with the vector plot emanating from this; the line indicates the magnitude and direction of the Ca2+ fluxes. The unlabelled calibration bar indicated is 10 pmol cm−2 sec−1. These Ca2+ flux measurements are superimposed upon an image of the pollen tube measured, obtained from the concomitant video recording; the labelled bar indicates scale (10 µm). The vibrating probe tip (to the top left) is approximately 50 µm back from the pollen tube tip, which is to the left of the image. GC indicates the generative cell and VN the vegetative nucleus. Note that the largest Ca2+ influx in this region is towards the generative cell.

We also performed 1-D vibrating probe measurements of Ca2+ fluxes along the shanks of growing pollen tubes of P. rhoeas that were several hundred µm long (n = 18), as shown in Figure 3. Typical measurements made along an individual pollen tube are shown in Figure 3(a). There is Ca2+ influx at several points along the pollen tube shank. From many measurements made on different tubes (n = 34), as well as along the length of the same pollen tube, it was evident that these Ca2+ fluxes were variable in their magnitude. The reason for this is not known, but it may be related to the length of the pollen tube and/or the rate of growth. Figure 3(b) shows that despite the substantial variation in the extent and amplitude of the Ca2+ fluxes along individual pollen tube shanks (which explains the large error bars shown), there is clear evidence for influx in the region around 50 µm behind the pollen tube tip. The Ca2+ influxes detected along the shank of untreated tubes were generally in the range 1–9 pmol cm−2 sec−1 (Figure 4). These data, together with the 2D vector data, provide evidence for the presence of Ca2+ influx in regions of pollen tube other than the pollen tube tip. In contrast, at 130–170 µm behind the tip there was consistently scarcely any evidence for Ca2+ influx (Figure 3b).

Figure 3.

Measurements of Ca2+ influx along the pollen tube shank.

The plots show net Ca2+ flux, with net influx as a negative reading along the shank of tubes several hundred µm long.

(a) A representative plot of Ca2+ fluxes along an individual growing pollen tube of P. rhoeas. Measurements of Ca2+ flux were made at various points along the pollen tube shank (indicated by the grey bars and the secondary y-axis on the right-hand side). These were made by recording sequentially at a point along the pollen tube, then moving the probe to another location; the times at which recordings were made are indicated by the x-axis. The primary y-axis shows net Ca2+ flux (pmol cm−2 sec−1). Note that there is a high level of variability in the level and location of the Ca2+ fluxes between the measurements made at different points. However, there is clearly evidence for net Ca2+ influx at some points along the pollen tube shank.

(b ) Mean Ca2+ fluxes measured along the shanks of pollen tubes of P. rhoeas. The means in this figure were obtained from sets of data (n = 7) where measurements were made at various points along the pollen tube length. Despite the high variability, there is a region of the pollen tube around 50 µm behind the tip, that consistently exhibits net Ca2+ influx (mean influx 4.17 pmol cm−2 sec−1, SEM ± 1.92, n = 7).

Figure 4.

Average Ca2+ fluxes before and after treatment with S proteins and Ca2+ channel blockers.

Average Ca2+ fluxes measured 50 µm behind the tip of Papaver pollen tubes before and after 500 µl La3+, 500 µl Ga3+, compatible, incompatible challenge and incompatible challenge of pollen tubes pre-treated with Gd3+. Comparisons were made between before and after measurements of individual pollen tubes treated similarly, and not with the overall mean in untreated pollen tubes.

Significance (T-test) levels are indicated as follows: *P < 0.05 one-sided; **P < 0.005 one sided; ‡P < 0.05 two-sided; ‡‡P < 0.01 two-sided; ‡‡‡P < 0.001 two sided.

Use of Gd3+ and La3+ to block calcium influx and growth in pollen tubes

We used the lanthanides Gd3+ and La3+ (Malhóet al., 1995) in an attempt to provide further evidence that the fluxes measured are due to Ca2+ movement across the plasma membrane. Addition of 75 µm Gd3+ resulted in considerable reduction in apical Ca2+ influxes (n = 11), since the tip oscillations were effectively damped out, leveling out at approximately 0–0.5 pmol cm−2 sec−1; 100 µm La3+ also markedly inhibited influx at the apex (data not shown). We also tested the effect of these agents on Ca2+ influx measured 50 µm behind the tip (n = 13). 100 µm La3+ and Gd3+ markedly reduced Ca2+ influx; however, in several examples, between 100 and 500 µm La3+ did not completely block the Ca2+ influx in this region. As shown in Figure 4, the mean maximal influx at the shank (50 µm behind the tip) before addition of lanthanides was reduced by 50% after treatment with 500 µm La3+ and to less than half with Gd3+.

We also examined the concentrations of Gd3+ and La3+ required for the inhibition of pollen tube growth in an attempt to corroborate the findings described above, with respect to inhibition of Ca2+ flux activity. Figure 5 demonstrates that although approximately 40 µm La3+ or Gd3+ is required for half-maximal inhibition of pollen tube growth, considerably higher concentrations are required for complete inhibition.

Figure 5.

Inhibition of pollen tube growth by La3+ and Gd3+.

These plots contain data from seven independent experiments, with n = 105 pollen tubes measured for each data point.

(a) Dose–response curve showing inhibition of pollen tube growth by GdCl3. Half-maximal inhibition was obtained at approximately 40 µm.

(b) Dose–response curve showing inhibition of pollen tube growth by LaCl3. Half-maximal inhibition was obtained at approximately 40 µm.

Ca2+ fluxes at the shanks of pollen tubes during the SI response

Having established the patterns of Ca2+ fluxes in normally growing pollen tubes, we challenged growing pollen tubes with incompatible S proteins in order to investigate whether there is a role for extracellular Ca2+ in the SI response. Using the Ca2+-selective vibrating probe, we chose a region of the pollen tube 50 µm behind the pollen tube tip, because we had previously observed alterations in Ca2+i using ratio imaging that appeared to be generated in this region of the pollen tube (Franklin-Tong et al., 1993, 1995, 1997). Moreover we had previously noted that this region also appeared to correlate with the general localization of the male germ unit, and we had obtained evidence, presented earlier in this paper, for Ca2+ influxes in this region. Figure 6 demonstrates that large Ca2+ fluxes that are S-specific (n = 13) are triggered along the pollen tube shank. It can be clearly seen in Figure 6(a,b,c) that S proteins may trigger influxes of Ca2+ as large as 10–18 pmol cm−2 s−1 in incompatible pollen tubes 50 µm back from the tip. Challenge of pollen tubes with incompatible S proteins induced an average 13.6-fold increase in Ca2+ influx at 50 µm from the tip (Figure 4) (highly significant; P < 0.001, T-test).

Figure 6.

Ca2+ fluxes are triggered at the pollen tube shank in the SI response.

The plots show net Ca2+ flux, with net influx shown as a negative reading. a, b and c illustrate typical examples of the response detected by the vibrating probe in an incompatible SI response, while d, e and f are typical compatible SI reactions. These comparisons demonstrate the S-allele specificity of the Ca2+ influxes measured, as the same S proteins were used, and the only differences are due to different S-alleles carried by the pollen samples. All measurements were made in the shank region of the pollen tube, at 50 µm behind the pollen tube tip. Measurements before treatment and after addition of S proteins (indicated by arrow) are shown. Some plots indicate readings taken afterwards, at 10 µm away from the pollen tube, in order to give a background reading.

Temporally, an increase in Ca2+ influx was detected very soon after addition of S proteins. Although increases were not detected immediately in all cases, in several experiments the first point measured was approximately 2-fold higher than before treatment. Even where increases appeared to be delayed (e.g. Figure 6a), < 10 sec after treatment Ca2+ influx was increased almost 4-fold. Ca2+ influx increased gradually over the subsequent approximately 4 min (Figure 6a–c). However, detailed analysis revealed that within this overall trend of an apparently slow increase in influx, Ca2+ influx oscillated, sometimes as rapidly as 10–15 sec, usually altering several-fold between oscillations, throughout this period. Ca2+ influx subsequently gradually decreased over a further 5–8 min, returning to the original pre-stimulation levels.

Figure 6(d–f) illustrate typical Ca2+ fluxes in a fully compatible SI situation. Under these conditions fluxes of only 2–3 pmol cm−2 sec−1 were observed after challenge, which is far below those detected in response to an incompatible situation. Indeed, this figure is more in line with the mean Ca2+ fluxes detected in normally growing, unchallenged pollen tubes. The mean Ca2+ influxes for the compatible pollen tubes before challenge with compatible S proteins was significantly different from those measured after the challenge (P < 0.01, T-test) but at a significance level 10 times lower than observed in the incompatible situation. As Figure 4 shows, the average magnitude of fluxes induced by incompatible S proteins is clearly much larger than the average flux induced by compatible S proteins. Although some low-level Ca2+ influx was detected in most cases (as expected for this region of the pollen tube shank), no Ca2+ fluxes comparable with the levels measured in the incompatible SI response were detected. This demonstrates the S-allele specificity of the alterations in Ca2+ fluxes measured at pollen tube shanks in response to S protein challenge.

Ca2+ fluxes at the tip of incompatible pollen tubes during the SI response

We also examined what effect the SI response had on Ca2+ fluxes at the pollen tube tip. This region of the pollen tube usually has a high level of Ca2+ influx activity, which is inhibited when pollen tube growth is blocked (Pierson et al., 1994). We have previously shown that the intracellular apical Ca2+ gradient is rapidly dissipated in the SI response (Franklin-Tong et al., 1997). Using the Ca2+-selective vibrating probe, we found that upon addition of incompatible S proteins, there was a gradual reduction of Ca2+ influx at the tip, with Ca2+ fluxes continuing to oscillate, but decreasing with each oscillation as shown in Figure 7. A few minutes after challenge the mean maximal Ca2+ fluxes at the tips of these treated pollen tubes were reduced to around 5% of the mean pre-treatment flux (n = 3).

Figure 7.

Ca2+ influx at the pollen tube tip is inhibited in the SI response.

Addition of incompatible S proteins results in Ca2+ influx being rapidly inhibited in the apical region of the pollen tube. Measurements at the tip, before and after addition of incompatible S proteins.

Effect of pre-treatment with La3+ and Gd3+ on SI-induced Ca2+ influx at the shank

Our data demonstrate that incompatible S proteins stimulate Ca2+ influx at the shank of the pollen tube. Since approximately 500 µm La3+ and Gd3+ reduced or blocked Ca2+ influx at the shank, we examined if the SI-induced influx is sensitive to lanthanides. Pollen tubes were treated with La3+ or Gd3+ prior to addition of incompatible S proteins. As shown in Figures 4 and 8, tubes pre-treated with 500 µm Gd3+ or La3+ had substantially reduced Ca2+ influx triggered by the SI response. Although exposure to lanthanides reduced the Ca2+ flux prior to challenge, the proportional increase in Ca2+ influx induced by an incompatible challenge was similar to that observed in tubes not pre-treated with Gd3+ or La3+ (11-fold compared with 13-fold). However, as shown in Figure 4, the Ca2+ influx, though increased, was far below that detected in the incompatible SI response, and was within the levels of fluxes detected in unchallenged pollen tubes. These data suggest that the SI-response might be mediated by two different kinds of Ca2+-conducting channels, one sensitive to inhibition by lanthanides and the other insensitive.

Figure 8.

Figure 8.

Pre-treatment of pollen tubes with Gd3+ and La3+ prior to the SI response blocks Ca2+ fluxes.

Effects of (a) 500 µm La3+ and (b) 500 µm Gd3+ on the calcium influxes in pollen tubes in response to incompatible S-proteins. The plots show net Ca2+ flux, with net influx as a negative reading.

(c) Ratiometric Ca2+-imaging of pollen tubes response to incompatible S proteins after pre-treatment with La3+. The pale grey area indicates the pre-treatment period. The dark grey area indicates treatment with 500 µm La3+. The arrow and line indicate addition of S proteins. Open symbols indicate ratiometric values of [Ca2+]i after La3+ treatment. The plot with black symbols shows a plot without La3+ pre-treatment for comparison.

Figure 8.

Figure 8.

Pre-treatment of pollen tubes with Gd3+ and La3+ prior to the SI response blocks Ca2+ fluxes.

Effects of (a) 500 µm La3+ and (b) 500 µm Gd3+ on the calcium influxes in pollen tubes in response to incompatible S-proteins. The plots show net Ca2+ flux, with net influx as a negative reading.

(c) Ratiometric Ca2+-imaging of pollen tubes response to incompatible S proteins after pre-treatment with La3+. The pale grey area indicates the pre-treatment period. The dark grey area indicates treatment with 500 µm La3+. The arrow and line indicate addition of S proteins. Open symbols indicate ratiometric values of [Ca2+]i after La3+ treatment. The plot with black symbols shows a plot without La3+ pre-treatment for comparison.

We obtained data using ratio imaging of pollen tubes that were pre-treated with 500 µm La3+ prior to challenge with incompatible S proteins (n = 8). Figure 8(c) shows that the increases in intracellular cytosolic free Ca2+ ([Ca2+]i) normally stimulated by the SI response (Franklin-Tong et al., 1993, 1997) are virtually completely inhibited by pre-treatment with La3+. Only one pollen tube showed increases in [Ca2+]i and this was quite abnormal, being delayed (initiating 80 sec versus 1–2 sec after challenge) and short-lived (sustained for approximately 50 sec, rather than 5–10 min). We suspect that this single event was due to an incomplete blocking of channels through which Ca2+ enters. These results demonstrate that the Ca2+ influx stimulated in the SI response is required for the increases in [Ca2+]i previously reported, and indicates that Ca2+ influx is required for the SI response.

Discussion

The ion-selective vibrating probe provides a powerful method for measuring the magnitude and direction of extracellular ion fluxes in regions adjacent to living cells; it has been especially effective in studies of pollen tubes (Feijóet al., 1995, 1999; Holdaway-Clarke et al., 1997; Kühtreiber and Jaffe, 1990; Messerli et al., 1999; Pierson et al., 1994). Here we have used the ion-selective vibrating probe to investigate Ca2+ fluxes associated with growing poppy pollen tubes, with particular reference to the SI response.

Calcium influxes are detected at the pollen tube shank

Our observations show, in agreement with previous studies on lily pollen tubes (Holdaway-Clarke et al., 1997; Messerli and Robinson, 1997), that there is an oscillating Ca2+ influx directed at the tip of the poppy pollen tube. We also demonstrate significant influxes of Ca2+ at the shank of normally growing pollen tubes. There is clear evidence for localized influx, despite some of the fluxes measured appearing parallel to the tube. Although it is not possible to rationalize the parallel component outside the cell by any diffusion process, it is known that large electrical fields surround the growing pollen tube (Weisenseel et al., 1975). These could potentially interact with the concentration gradient of Ca2+ to produce local distortions of the resultant vectors parallel to the pollen tube shaft. Until now, it has been generally considered that there is negligible Ca2+ influx except that occurring at the tip (Malhóet al., 1995). Although Weisenseel et al. (1975) reported net ion influx at the pollen tube shank, the contribution of Ca2+ was not determined, and Kühtreiber and Jaffe (1990) demonstrated only small Ca2+ influxes of approximately 1 pmole cm−2 sec−1 at the shank of a tobacco pollen tube.

Extracellular Ca2+ is involved in the SI response

Although increases in [Ca2+]i are known to be stimulated by the SI response in incompatible poppy pollen (Franklin-Tong et al., 1993, 1995, 1997), one key question remaining has been the source(s) of Ca2+. Here we demonstrate that the SI response stimulates an influx of extracellular Ca2+. Our data indicate two alterations in the normal pattern of Ca2+ fluxes in the pollen tube: increased influx in a region of the shank, and decreased influx at the tip, as illustrated in Figure 9(a). We used Gd3+ and La3+, which are reported to be inorganic Ca2+ channel blockers (Marshall et al., 1994), in an attempt to further decipher the observed influxes, and found that both inhibitors blocked Ca2+ influx in pollen tubes. Furthermore, our ratio-imaging data, showing that La3+ blocked SI-induced increases in [Ca2+]i, clearly demonstrates that Ca2+ influx is required for these increases in [Ca2+]i.

Figure 9.

Figure 9.

Models for Ca2+ fluxes in a poppy pollen tubes and alterations induced by the SI response.

(a) This model is based upon the Ca2+ influx data presented. In normally growing poppy pollen tubes there is evidence of Ca2+ influx (indicated by black arrows) at the tip and the shank. This implies there are Ca2+-conducting channels in the pollen tube (grey boxes). The effects of an incompatible SI response on Ca2+ influx (indicated by open arrows) are Ca2+ influx at the tip is inhibited and significant influx is detected in the shank of the pollen tube.

(b) Model indicating some of the events identified as being SI-stimulated, how they might interrelate, and how alterations in [Ca2+]i may affect them. Stigmatic S proteins (S1, Sx) interact with assumed S-specific pollen receptor via SBP (all bind SBP, but only S1 protein can interact with S1 receptors thereby generating incompatible S-specificity). Binding the S-receptor stimulates increases in Ca2+ influx (perhaps through a non-specific cation channel), which is postulated to interact (indicated by the double-headed arrow) in some way with the pollen S-receptor to achieve S-specific influx of Ca2+. Evidence suggests that Ca2+ influx stimulates increases in [Ca2+]i, with some contribution from intracellular stores as well as extracellular sources. This activates several protein kinase cascades. Other events known to be triggered downstream of increases in [Ca2+]i are indicated.

Figure 9.

Figure 9.

Models for Ca2+ fluxes in a poppy pollen tubes and alterations induced by the SI response.

(a) This model is based upon the Ca2+ influx data presented. In normally growing poppy pollen tubes there is evidence of Ca2+ influx (indicated by black arrows) at the tip and the shank. This implies there are Ca2+-conducting channels in the pollen tube (grey boxes). The effects of an incompatible SI response on Ca2+ influx (indicated by open arrows) are Ca2+ influx at the tip is inhibited and significant influx is detected in the shank of the pollen tube.

(b) Model indicating some of the events identified as being SI-stimulated, how they might interrelate, and how alterations in [Ca2+]i may affect them. Stigmatic S proteins (S1, Sx) interact with assumed S-specific pollen receptor via SBP (all bind SBP, but only S1 protein can interact with S1 receptors thereby generating incompatible S-specificity). Binding the S-receptor stimulates increases in Ca2+ influx (perhaps through a non-specific cation channel), which is postulated to interact (indicated by the double-headed arrow) in some way with the pollen S-receptor to achieve S-specific influx of Ca2+. Evidence suggests that Ca2+ influx stimulates increases in [Ca2+]i, with some contribution from intracellular stores as well as extracellular sources. This activates several protein kinase cascades. Other events known to be triggered downstream of increases in [Ca2+]i are indicated.

Although the levels of Gd3+ and La3+ (100–500 µm) used are high, they are consistent with the 100 µm La3+ used by Malhóet al. (1995) and Messerli and Robinson (1997) to dissipate Ca2+ influx at the tip. In addition, Marshall et al. (1994) found that although half-maximal inhibition was obtained at 40 µm La3+ or Gd3+, 500 µm La3+ or Gd3+ did not completely inhibit Ca2+ uptake in maize vesicles; in intact plant protoplasts even 1 mm La3+ and 1 mm Gd3+ did not inhibit Ca2+ uptake completely. Thus our observations reflect the findings from other systems. Since more La3+ or Gd3+ is required to inhibit Ca2+ influx in the shank of the pollen tubes than at the tip, this suggests that there might be different types of Ca2+-conducting channels, including non-specific cation channels, in the tips and shanks of pollen tubes. Furthermore, since the proportional increase in Ca2+ flux with incompatible challenge is similar, regardless of whether or not tubes were pre-treated with lanthanides, this suggests that the SI-response might be mediated by two different kinds of Ca2+ channel, one sensitive to lanthanides and the other insensitive. We conclude therefore that there must be active Ca2+-conducting channels along the shank of the pollen tube (though we have no direct evidence for their presence).

The relationship between Ca2+ influx and increases in [Ca2+]i

Although our data show that Ca2+ influx is triggered early in the SI response and in line with the timing of the first detectable increases in [Ca2+]i imaged previously, we did not detect the dramatic and immediate influx that might, perhaps, be expected if the extracellular Ca2+ influxes relate directly to the increases in [Ca2+]i. Nevertheless, the initial Ca2+ influxes measured are potentially large enough to trigger the initial increases in [Ca2+]i, since Holdaway-Clarke et al. (1997) demonstrated that the Ca2+ influx required to fuel the large apical gradient was very small. Thus, it is likely that the relatively modest initial Ca2+ influxes that occur immediately after SI-induction are sufficient to initiate the increases in [Ca2+]i. However, since localized Ca2+ fluxes are reported to remain at their sites of cell entry for many minutes following channel activation (Clapham, 1995; Trewavas, 1999), we favour the view that, although the initial [Ca2+]i increases involve the extracellular influx reported here, at least some of the elevations in [Ca2+]i originate from intracellular Ca2+ stores, since we previously visualized large changes in [Ca2+]i that appear to move as a wave through the pollen tube shank (Franklin-Tong et al., 1997). Increases in [Ca2+]i through influx could be amplified through Ca2+-induced Ca2+ release or IP3-induced Ca2+ release as a secondary process, and propagated as Ca2+ waves previously observed (Franklin-Tong et al., 1996). Since the magnitude of Ca2+ influxes stimulated by SI gradually increased over several minutes, this suggests that this extracellular Ca2+ contribution may serve to further ramp up [Ca2+]i. One possible explanation is that only a small amount of influx is required to trigger Ca2+ release from intracellular stores, but increasing amounts are needed to sustain and propagate this Ca2+i release.

A model for Ca2+ increases in the SI response

Together, our data indirectly implicate a role for intracellular release, but more importantly, directly demonstrate that Ca2+ influx is a key early step in the SI response; indeed it may represent the first event. The rapidity of the Ca2+i response is in line with the idea that the SI reaction is a receptor-mediated response, with the S-protein acting as a signal molecule. The notion that the Ca2+ influx is likely to represent an early step in the signal transduction cascade is not unprecedented. Transient influx of Ca2+ across the plasma membrane has been shown to play a role in the early steps of the pathogen defence response signal cascade (Romeis et al., 1999). Similarly, one of the earliest measured responses to Nod factors by their host legumes is a stimulation of Ca2+ influx (Felle et al., 1998; Yokoyama et al., 2000).

We consider it unlikely that the S proteins act directly on Ca2+-conducting channels since S-specificity needs to be encoded in some way, as elevations in [Ca2+]i do not occur as a consequence of a compatible SI reaction. Our current model (Figure 9b) therefore proposes that there is a membrane-located pollen S-receptor that binds S proteins in an S-specific manner. Although we have not identified the pollen S-receptor, a pollen-specific integral membrane proteoglycan, SBP, has been shown to bind S proteins in vitro (Hearn et al., 1996). We hypothesize that SBP either acts as an accessory receptor in conjunction with the S-specific receptor in pollen, or that it is the S-receptor, and has both high and low-affinity binding sites (Jordan et al., 1999). Since the present study demonstrates that when incompatible stigmatic S proteins interact with the pollen S-receptor, SI-induced increases in Ca2+ influx are stimulated, it appears likely that the pollen S-receptor is in some way associated, or interacts with, channels that conduct Ca2+. In this way, the S-specific influx of Ca2+ could be achieved. Our data also indicate that Ca2+ influx stimulates increases in [Ca2+]i, which suggests some contribution from intracellular stores as well as extracellular sources. These increases in [Ca2+]i trigger the SI-induced signalling cascade (Figure 9b) described earlier.

In summary, we have demonstrated that normally growing poppy pollen tubes exhibit relatively large influxes of Ca2+ in the shank region. Of particular importance are the findings that these shank influxes are amplified during the SI response. It seems plausible therefore that the SI receptor is in some way associated with Ca2+-conducting channels along the shank of the tube. These results further emphasize the role of extracellular Ca2+ currents as part of the complex signalling cascade leading to the inhibition of pollen tube tip growth.

Experimental methods

Plant material

Plants of P. rhoeas L. (var. Shirley) segregated for known incompatibility genotypes (S1S3 and S2S4) were used for these experiments (Franklin-Tong et al., 1988). Pollen was stored at −20°C, over silica gel, until required.

Expression of recombinant S-proteins and assessment of S-protein activity

Recombinant S proteins for alleles S1 and S3 (S1e and S3e, respectively) were used to challenge pollen. Nucleotide sequences specifying the mature peptide of the S gene for S1 and S3 were cloned into the expression vector pMS119 to produce pPRS100 and pPRS300, respectively (Foote et al., 1994). High-level expression of recombinant S proteins were induced in these E. coli cells containing pPRS100 or pPRS300 with 1 mm isopropyl β-d-thiogalactoside. The resulting inclusion bodies were isolated, purified and refolded as described in Kakeda et al. (1998). Protein concentration was estimated using the Coomassie Brilliant Blue R-250 dye-binding method (Bradford, 1976). The recombinant S proteins S1e and S3e were assayed for their S-specific pollen inhibitory activity prior to each experiment. They were dialyzed against liquid germination medium (see In vitro system for growing pollen) and tested against incompatible (S1S3) pollen and compatible (S4S5 or S2S4) pollen genotypes for S-specific inhibitory activity; see Kakeda et al. (1998) for full details.

In vitro system for growing pollen for vibrating probe work

Pollen was grown using an in vitro system, adapted from Franklin-Tong et al. (1993) so that the vibrating probe technique could be used to make measurements from growing pollen tubes. The concentration of Ca2+ in the germination medium (GM) was adjusted to 50 µm, which was the lowest concentration that would support good growth. This GM comprised 7.5% sucrose, 5% polyethylene glycol 6000, 0.01% KNO3, 0.01% H3BO3, 0.01% Mg(NO3)2⋅6H2O, 50 µm CaCl2⋅2H2O. Slides were prepared as described in Franklin-Tong et al., 1997. Briefly, pollen was pre-hydrated and a pre-warmed chamber-slide was prepared with GM and 1.5% low-temperature gelling agarose (type VII; Sigma-Aldrich Co. Ltd, Poole, UK). Pollen was sown on this thin layer of germination medium, left to cool and solidify for approximately 30 sec. This was then covered with 300 µl liquid germination medium and incubated at room temperature in a moist chamber until pollen tubes of approximately 100 µm had grown.

Vibrating probe work

Measurements of Ca2+ fluxes were made using Ca2+ selective vibrating probes (Smith et al., 1994). These were made at various points at the extreme apex and along the shank region of growing pollen tubes. For the SI challenges, we chose a region of the pollen tube approximately 50 µm behind the pollen tube tip to position the probe, and monitored Ca2+ fluxes over time.

Some 2-dimensional scans were made of the shanks of growing pollen tubes, which enabled the mapping of both the magnitude and vector direction of the Ca2+ fluxes. The Ix and Iy were made sequentially, not simultaneously or interleaved. As a result, saying the Ix is stable for the time the measurements were made does not necessarily mean that the Iy, measured next, shares the same stability. Matlab scripts were used to place vectors on grey scale images with appropriate scale bars. A sample script is available for download at the following URL: http://marlin.bio.umass.edu/biology/kunkel/pub/matlab/. Data matrices produced by the ion selective probe software, as described in Holdaway-Clarke et al. (1997), were analyzed using an MS Excel (Excel 97, Microsoft) spreadsheet to convert data from the background-mV estimation of concentration and microvolt-difference estimation of the local gradient into specific ion flux (pmole cm−2 sec−1). A model MS Excel spreadsheet is available at http://marlin.bio.umass.edu/biology/kunkel/mage/uv2flux.html

A Nernst slope and intercept was determined for each electrode used. Calibration of the efficiency of the probe in a dynamic state was performed using point sources of Ca2+ as previously described (Smith et al., 1994). The background mV response was measured prior to each series of 1D measurements and prior to each 2D measurement. Background mV values away from and close-by the tube were routinely determined to establish if the local environment of the pollen tubes was being modified by evaporation or local cellular metabolism. Mean data on the Ca2+ fluxes was generated from various data sets. Means of fluxes for data where no detectable changes were elicited (e.g. at a certain point behind the pollen tube tip, or before challenge) were taken from a complete data set where that measurement was made. ‘Mean maximal’ fluxes for comparisons of data where changes were detected (e.g. oscillations at pollen tube tips and fluxes triggered by the SI response) were taken from 10 adjacent data points in time where the largest influx was detected. For those experiments examining challenges, means were calculated only from the pollen tubes subsequently challenged with compatible or incompatible S-proteins (i.e. ‘before’ and ‘after’ challenge comparisons).

Use of Gd3+ and La3+ to block calcium influx and growth in pollen tubes

A range of concentrations between 10 µm and 500 µm of Gd3+ and La3+ was used to establish concentrations required to block Ca2+ influx. These were added to pollen tubes prior to measurement with the ion-selective vibrating probe. Pollen tubes were also treated with addition of La3+ and Gd3+ prior to addition of incompatible S proteins. To establish the concentrations of La3+ and Gd3+ required to inhibit pollen tube growth, pollen tubes were grown on germination medium solidified with 0.5% Agarose MP (Boehringer-Ingelheim, Bracknell, Berkshire, UK). After 30 min addition of LaCl3 or GdCl3 were made by adding liquid germination medium supplemented with 0, 10, 25, 50, 75, 100, 250 or 500 µm LaCl3 or GdCl3. At this point, a control was fixed in 2% glutaraldehyde for t = 0. The pollen tubes were grown for 30 min after treatment, fixed in 2% glutaraldehyde and the length of 15 tubes for every treatment was measured. Independent experiments were repeated (n = 7), and the data analyzed using Minitab software.

Microinjection, treatment and [Ca2+]i imaging of fura-2 dextran-loaded pollen tubes

Fura-2 dextran was used to visualize the [Ca2+]i in individual living pollen tubes in order to determine if La3+ was effective in blocking intracellular [Ca2+]i, as in our vibrating probe studies it was found to block Ca2+ influx stimulated by the SI response. Pollen tubes grown in vitro were microinjected with fura-2 dextran (10 000 Mr) (Molecular Probes Inc., Eugene, OR, USA) as described in detail by Franklin-Tong et al. (1997). [Ca2+]i was measured in pollen tubes pre-treated with 500 µm La3+ using ratiometric imaging before and after SI induction using biologically active S proteins, as described in Franklin-Tong et al. (1997). A series of pairs of images imaged at 10-sec intervals, at the wavelengths 340 nm and 360 nm was captured in a timed sequence. Pollen tube [Ca2+]i was analyzed through a series of images captured by fluorescence ratio-imaging on a Nikon Diaphot microscope (Nikon UK Ltd., Kingston, Surrey, UK) outfitted with a highly regulated mercury vapor lamp and a thermoelectrically cooled CCD (Pierson et al., 1994, 1996). Images were obtained and processed using the PMIS imaging software package (Photometrics, Tucson, AZ, USA) and the ratio data were plotted using excel.

Acknowledgements

We are indebted to Wolfram Nagel, Department of Physiology, University of Munich, for discussions on the physical interpretation of the vectors. VEF-T wishes to thank Chris Franklin for his continued support and encouragement. Thanks are due to John Martin and colleagues for expert horticultural services. VEF-T is funded by Biotechnology and Biological Sciences Research Council (BBSRC). JGK is currently funded by a grant from the Deutsche Forshungsgemeinschaft. P.K.H. is funded by the USA National Science Foundation (MCB-0077599) and the Constantine Gilgut Endowed Professorship at UMass. The UMass Vibrating Probe Facility is supported by Applicable Electronics, Inc.

Ancillary