Petunia Phospholipase C1 Is Involved in Pollen Tube Growth

A PI-PLC Is Present in Actively Growing Petunia Pollen Tubes and Supports Tube Growth

We first isolated candidate cDNAs for PI-PLCs from a pollen/pollen tube cDNA library of P. inflata using as a probe a PCR fragment containing the coding regions for the highly conserved X and Y catalytic domains of At PLC2 of Arabidopsis thaliana (Rebecchi and Pentyala, 2000; Wang, 2001). Only one class of clones was isolated in these screens, and the clone containing the longest insert encoded a full-length PLC, which we named Pet PLC1. As with all plant PLCs described to date, compared with the sequences of the animal PLC family members, Pet PLC1 is most similar to the mammalian δ isoform. Figure 1A shows that Pet PLC1, like other plant PLCs, contains a C2 Ca²⁺/phospholipid binding domain (Kopka et al., 1998), the conserved X and Y catalytic domains, and a region homologous with the second loop of the Ca²⁺ binding EF hand of PLC δ, the so-called EF-loop (Otterhag et al., 2001). However, like other plant PLCs, Pet PLC1 also lacks the PtdInsP₂ binding (PH) domain involved in membrane targeting and processive catalysis found in nearly all animal PLCs (Paterson et al., 1995; Essen et al., 1996; Katan and Williams, 1997). An alignment of the deduced amino acid sequences of Pet PLC1 and its most closely related Arabidopsis homolog, At PLC2, is shown in Supplemental Figure 1 online.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Pet PLC1 Encodes a PLC Predominantly Expressed in Pollen/Pollen Tubes.

(A) Scheme of the domain structures of Pet PLC1 and human PLC δ1. An alignment of the complete sequences of Pet PLC1 and homologous Arabidopsis PLCs, At PLC2 and At PLC4, is shown in Supplemental Figure 1 online. The EF-loop and the C2 domain contain motifs likely involved in the regulation of enzyme activity by Ca²⁺ and/or lipids.

(B) RNA gel blot analysis of the expression of Pet PLC1. Total Petunia RNA (20 μg) was isolated from each of the tissues indicated and from pollen and pollen tubes and loaded in each lane, and the blot was hybridized with radiolabeled Pet PLC1 cDNA. Ethidium bromide staining of the gel used for blotting shows rRNA bands as a loading control. Stages of anther development correspond to developing microspores in the tetrad configuration (1), free uninucleate microspores (2), binucleate microspores (3), microspores that have completed mitotic division (4), and open flowers containing mature pollen (5).

RNA gel blot analysis indicated that Pet PLC1 was expressed predominantly in anther/pollen and was expressed at very low levels, if at all, in all other tissues examined (Figure 1B). During anther development, the Pet PLC1 transcript was first detected at stage 4, approximately the time when microspores are undergoing mitosis (ss., 1994), peaked with pollen maturation, and was maintained during pollen tube growth (Figure 1B).

Pet PLC1 Is Targeted to the Apical Plasma Membrane

There appear to be two major classes of PI-PLCs in plants (Drøbak, 1992). One class is principally cytosolic, requiring millimolar Ca²⁺ concentrations for activity and preferentially using PtdIns as a substrate, whereas the other is membrane-bound, requires micromolar Ca²⁺ concentrations, and preferentially uses PtdInsP₂ and phosphatidylinositol 4-phosphate as its substrates. Mung bean (Vigna radiata) PLC3 has been tentatively localized to the plasma membrane when heterologously expressed in Arabidopsis suspension culture cells, providing a molecular candidate for one membrane-bound enzyme (Kim et al., 2004). Several putative regulators of the tip growth machinery in pollen tubes (e.g., the Rop monomeric G-proteins and their interacting proteins [Wu et al., 2000; Yang, 2002]) have also been shown to localize to the apical membrane of the tube. Therefore, we sought to characterize the localization and PI-PLC activity of Pet PLC1 to ascertain whether we could place it in one of these broad subcellular classes.

We fused green fluorescent protein (GFP) to the N terminus of Pet PLC1 and used biolistic transformation to transiently express this fusion protein in pollen and pollen tubes under the control of the LAT52 promoter. Expression of this construct caused no detectable growth-related phenotype in >150 pollen tubes examined. As shown in Figure 2A , a fraction of GFP:Pet PLC1 was associated with the apical cytoplasm, with a fraction targeted to the apical plasma membrane, a localization not seen with soluble GFP (Figure 2B). This membrane targeting led us to speculate that the localization of the enzyme might be involved in regulating tip growth, as proposed for Rops (Gu et al., 2005). Therefore, we characterized the relationship of the GFP:Pet PLC1 localization to phases of rapid elongation and slowing of growth characteristic of pollen tube elongation. Although pollen tubes from species such as lily (Lilium daviddi) show regular oscillatory growth when grown in vitro (Holdaway-Clarke and Hepler, 2003), pollen tubes of Petunia and Nicotiana do not exhibit this characteristic of clear, regular pulsatile growth (Pierson et al., 1995; Geitmann et al., 1996; Geitmann 1997; Geitmann and Cresti, 1998; P.E. Dowd, T.-h. Kao, and S. Gilroy, unpublished data). Petunia pollen tubes, however, do show phases of rapid elongation interspersed with periods of slowing and cessation of growth, although these occur with less predictable periodicity (see Supplemental Movie 1 online). Therefore, we followed GFP:Pet PLC1 localization as the pollen tubes went through these phases of elongation. Figure 2C shows frames taken from a typical movie of such growth (see Supplemental Movie 1 online). During periods of elongation (0 to 5:00 in Figure 2C), GFP:Pet PLC1 was localized to the lateral plasma membrane just behind the growing tip, with depletion of plasma membrane localization in regions showing rapid expansion. However, during periods of reduced elongation/growth arrest, the apical localization became more prominent around the entire apex of the tube (7:12 to 8:12 in Figure 2C). When growth resumed (12:48 in Figure 2C), this clear apical localization was reduced at the region where the tip started to expand.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Localization of Pet PLC1 in Growing Pollen Tubes.

(A) and (B) Pollen was biolistically transformed with constructs encoding a GFP:Pet PLC1 fusion protein (A) or soluble GFP as a control (B). Note the localization of GFP:Pet PLC1 to the apical plasma membrane, which is magnified in the boxed inset. Images represent an extended depth of focus of 20 confocal optical sections 0.5 μm apart. The images shown are representatives of those obtained from >100 individual pollen tubes.

(C) Time course of GFP:Pet PLC1 localization. Single midplane optical sections were taken at the indicated times (minutes). The series of images shown is representative of >30 individual movies.

(D) Colocalization of FM 4-64 and GFP:Pet PLC1. Representative images are of growing (1 and 2) and nongrowing (3) pollen tubes showing the colocalization of FM 4-64 and GFP:Pet PLC1 signals and the overlay of these images. Colocalizing pixels are coded yellow in the overlay images. Note the colocalization in the inverted cone of FM 4-64–stained vesicles (arrows). Controls for bleed through of FM 4-64 or GFP:Pet PLC1 labeling alone indicated no contamination of FM 4-64 signal in the GFP:Pet PLC1 image and vice versa.

(E) FRAP analysis of pollen tubes expressing GFP:Pet PLC1. Numbers represent times (seconds) after photobleaching, which was performed in the areas indicated in red. The series of images shown are representative of more than nine separate experiments. Bars in (A) to (E) = 10 μm.

(F) Quantitative analysis of photobleaching recovery. Fluorescence intensity was monitored in the indicated regions (dashed boxes in the insets) before and after photobleaching of the regions defined by the red boxes. Intensities were normalized to the measurement made directly before bleaching. Note that all regions of the lateral membrane recover with similar kinetics, suggesting that the flow of unbleached GFP:Pet PLC1 occurs not only from insertion at the tip of the pollen tube.

To further characterize the localization of GFP:Pet PLC1, we imaged the expression of GFP:Pet PLC1 and labeled the growing pollen tubes with the fluorescent dye FM 4-64 (Figure 2D). FM 4-64 has been used extensively in pollen tubes to visualize both the plasma membrane and a pool of vesicles that form an inverted cone centered on the growing tip (Parton et al., 2001). These vesicles are thought to represent the endocytotic pool associated with membrane recovery during the massive exocytosis associated with tip growth (Parton et al., 2001; de Graaf et al., 2005; Monteiro et al., 2005). We observed that, although GFP:Pet PLC1 did not colocalize with FM 4-64 staining in the shank or more subapical regions of the growing pollen tube, there was substantial overlap in localization in the apex, with apical plasma membrane and the vesicle cone both being costained (arrows in Figure 2D and Supplemental Movie 2 online). In nongrowing tubes, there was no evidence of the vesicle cone by either GFP:Pet PLC1 or FM 4-64 staining, which uniformly stained the plasma membrane (Figure 2D; see Supplemental Movie 3 online), consistent with the proposed role of the FM 4-64 staining vesicle pool in active tip growth.

We next used fluorescence recovery after photobleaching (FRAP) analysis to define whether this membrane localization of GFP:Pet PLC1 was dynamic. We photobleached the GFP:Pet PLC1 associated with the lateral plasma membrane toward the tube apex and the region of the cytoplasm where the vesicle cone was observed (Figures 2E and 2F). Recovery from photobleaching of the apical cytoplasm occurred within 6 s (Figures 2E and 2F; see Supplemental Movie 4 online), likely by replenishment of unbleached molecules through the vigorous cytoplasmic streaming within the shank of the pollen tube. Fluorescence recovery into bleached areas of the membrane was slower, with full recovery taking ∼50 s (Figures 2E and 2F; see Supplemental Movie 5 online). Analysis of fluorescence intensities across the bleached area showed equal recovery rates in the center and periphery of the bleached zone. This observation indicated that unbleached GFP:Pet PLC1 was not moving just into the bleached area from the end closest to the growing tip, suggesting that Pet PLC1 is not only inserted into the membrane at the very apex of the tube (i.e., in the region associated with the FM 4-64 staining vesicle cone). As the cytoplasmic pool of GFP:Pet PLC1 is rapidly redistributed throughout the pollen tube by cytoplasmic streaming, we were unable to robustly determine whether the recovery of Pet PLC1 in the membrane was limited to lateral diffusion of unbleached GFP:Pet PLC1 within the plane of the membrane or via insertion from this cytoplasmic pool. As cytoplasmic streaming does not extend completely to the tip of the tube (Holdaway-Clarke and Hepler, 2003), recovery of the very apical cytoplasmic GFP:Pet PLC1 signal in our FRAP experiments cannot be entirely explained by cytoplasmic streaming.

We next determined that recombinant Pet PLC1 and GFP:Pet PLC1 had associated PtdInsP₂-cleaving PLC activity (Figure 3A ), whose micromolar Ca²⁺–dependent activity profile was similar to that detectable in crude microsomal and soluble extracts from pollen tubes (Figure 3B). As noted above, the membrane-bound form of PLC is thought to require micromolar Ca²⁺ concentrations and preferentially use PtdInsP₂ and phosphatidylinositol 4-phosphate as its substrates. Thus, the biochemical profile of Pet PLC1 is consistent with the deduction from the GFP imaging described above that this enzyme is likely a membrane-associated PtdInsP₂-cleaving enzyme.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Effect of Pet PLC1-H126A, an Inactive Form of Pet PLC1, on Pollen Tube Growth.

(A) Biochemical assay of recombinant Pet PLC1, GFP:Pet PLC1, and Pet PLC1-H126A. Note that Pet PLC1-H126A has no detectable PI-PLC activity, whereas Pet PLC1 and GFP:Pet PLC1 have nearly identical enzymatic actions (results represent means ± se; P > 0.05, t test; n = 3).

(B) Production of InsP₃ from PtdInsP₂ by recombinant Pet PLC1 and by soluble and microsomal fractions of P. inflata pollen tubes. Results represent means ± se (n = 3 separate experiments) of PLC activity from the indicated fractions normalized to the activity measured at 10 μM Ca²⁺.

(C) Morphology of pollen tubes producing both Pet PLC1-H126A and GFP or soluble GFP alone. Left, Pet PLC1-H126A in a bright-field image; middle, GFP fluorescence image showing cotransformed soluble GFP; right, soluble GFP alone used as a control. Note that Pet PLC1-H126A causes arrest of tip growth and apical swelling of the growing tube. pg, pollen grain; st, swollen pollen tube tip. Bar = 20 μm.

(D) Ratio imaging of cytosolic Ca²⁺ in a pollen tube producing GFP alone (control) and two pollen tubes each transiently expressing Pet PLC1-H126A and soluble GFP. Pollen tubes were microinjected with the fluorescent Ca²⁺ imaging dye Indo-1-dextran, and Ca²⁺ distribution was calculated by ratio imaging. Ca²⁺ levels have been pseudo-color-coded according to the scale at bottom right. Bar = 20 μm.

(E) Localization of GFP:Pet PLC1-H126A in pollen tubes before induction of bulging. Note the localization to the apical plasma membrane, as also seen for GFP:Pet PLC1. Bar = 10 μm.

Expression of Pet PLC1-H126A Disrupts Tip Growth

As Pet PLC1 showed a localization consistent with a possible role in pollen tube growth, we sought a more direct indication that this enzyme was indeed a candidate for a PLC involved in regulating pollen tube function. Therefore, we used site-directed mutagenesis to change a catalytically required His-126 (in the X domain; see Figure 1A and Supplemental Figure 1 online) to Ala-126 to produce a catalytically inactive protein, named Pet PLC1-H126A. A similar mutation in the human PLC δ1 causes a 20,000-fold decrease in phospholipase activity (Ellis et al., 1998). We first examined whether Pet PLC1-H126A was indeed catalytically inactive by comparing its turnover rate of PtdInsP₂ with that of the recombinant wild-type protein. Pet PLC1-H126A showed no detectable PtdInsP₂ cleavage under conditions in which both the wild-type enzyme and the GFP:Pet PLC1 were active (Figure 3A).

We reasoned that as Pet PLC1-H126A was catalytically inactive, if produced in pollen tubes it might compete with the endogenous Pet PLC1 for its regulators, signaling intermediates, and/or targeting machinery and so might have a dominant negative–like effect on the endogenous enzyme. Thus, we used biolistic transformation to transiently express Pet PLC1-H126A (under the control of the pollen-expressed LAT52 promoter of tomato) (Twell et al., 1990), along with soluble GFP, in pollen. GFP expression was used to identify transformants and was driven by a separate LAT52 promoter present in the same plasmid as the test construct. This dual construct/single vector approach allowed us to use the expression of GFP in different pollen grains/tubes to more reliably assess the likely expression of a particular test construct. Expression of Pet PLC1-H126A led to the arrest of tube growth and swelling of the pollen tube tip (Figure 3C). We also used biolistic transformation to express wild-type Pet PLC1 under the control of the LAT52 promoter; however, no alteration of pollen tube growth was observed (n > 50, at least three separate experiments; data not shown). We hypothesize that the overexpressed wild-type protein was being regulated similarly to the endogenous enzyme and so did not alter growth patterns.

Pet PLC1-H126A Expression Disrupts the Apical Ca²⁺ Gradient and Actin Dynamics

Tip growth in pollen tubes is known to be associated with a tip-focused Ca²⁺ gradient that is required to drive and orient apical growth (reviewed in Holdaway-Clarke and Hepler, 2003). Therefore, we monitored Ca²⁺ gradients in the pollen tubes expressing wild-type Pet PLC1, Pet PLC1-H126A, or GFP alone to determine their effects on the tip-focused Ca²⁺ gradient (Figure 3D). We observed a Ca²⁺ gradient in untransformed and GFP-expressing pollen tubes, and this gradient was unaltered in tubes that overexpressed wild-type Pet PLC1 (data not shown). In pollen tubes expressing Pet PLC1-H126A, the Ca²⁺ gradient was disrupted, with the region of high Ca²⁺ concentration being much more extensive and filling the entire bulging tip (Figure 3D). This is unlikely to be a feature of high Ca²⁺ always being associated with a bulge at the end of the tube, rather than a more selective effect of Pet PLC1-H126A, as other researchers have reported treatments leading to bulging pollen tube tips that are not associated with an increased Ca²⁺ level within the bulge (Cardenas et al., 2005; Gu et al., 2005).

We hypothesized that these effects of Pet PLC1-H126A on normal growth were possibly attributable to competition with the endogenous PLC. Because we had characterized this wild-type enzyme to be localized to the apical plasma membrane (Figure 2A), we tested whether the inactivated enzyme was likewise localized, as expected if it were truly competing with the endogenous PLC for sites on the plasma membrane. Images of GFP:Pet PLC1-H126A taken before it had induced obvious tip swelling showed that this protein was indeed localized to the apical plasma membrane in a pattern very similar to that of the active enzyme, consistent with it acting in a dominant negative manner on the endogenous enzyme (Figure 3E).

Actin is also thought to play a major role in directing normal localized tip growth in pollen tubes (Hepler et al., 2001). Therefore, we coexpressed Pet PLC1-H126A with an actin-decorating GFP fusion protein, GFP:mTalin, to assess the effects on actin dynamics in the swollen tube (Figure 4 ). Similar to the findings reported by others (reviewed in Vidali and Hepler, 2001; Wasteneys and Yang, 2003), in growing pollen tubes used as controls, GFP:mTalin decorated the prominent actin cables running along the shank of the tube, but these cables did not protrude to the very tip of the growing apex (Figure 4A). However, in pollen tubes in which Pet PLC1-H126A had caused an apical swelling, GFP:mTalin decorated a prominent subapical actin band. Such swelling and disruption of actin distribution were suppressed by germinating the pollen grains expressing Pet PLC1-H126A in medium containing 5 nM Lat B (Figure 4A), suggesting that Pet PLC1-H126A may exert its effects on tip growth through disruption of actin dynamics.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Effect of Pet PLC1-H126A, an Inactive Form of Pet PLC1, on Actin and Golgi Dynamics.

(A) Actin distribution visualized by GFP:mTalin in control pollen tubes and pollen tubes transiently expressing Pet PLC1-H126A. Arrowheads point to the subapical actin-rich band that forms in the swollen tip induced by Pet PLC1-H126A. Note that swelling and subapical actin band formation are suppressed by treatment with 5 nM Lat B.

(B) Golgi dynamics in control pollen tubes and those producing Pet PLC1-H126A, as visualized by expression of the Golgi-retained NAG:CFP. Linear tracking of Golgi in the shank of the pollen tube (white arrowheads) is present in both control tubes and tubes producing Pet PLC1-H126A. Note that Golgi are excluded from the extreme tip in the control tube but are accumulated in this region of the Pet PLC1-H126A–producing tube (black arrowheads) from the streaming of Golgi through the swollen tip. pg, pollen grain; sa, subapical region representing maximal travel of Golgi. Numbers represent time in seconds. Images are representatives of those obtained from >20 separate pollen tubes.

Bars = 10 μm.

Although the subapical actin band was never observed in the GFP:mTalin-expressing control tubes and GFP:mTalin has been used successfully by several groups to monitor the dynamics of actin structure in pollen tubes (Fu et al., 2001; Cardenas et al., 2005), we were concerned that expressing an actin binding protein such as talin could alter the actin dynamics itself and may artifactually implicate Pet PLC1 in modulating the activity of the actin cytoskeleton. Therefore, we also imaged Golgi dynamics to independently confirm likely alterations in actin function induced by Pet PLC1-H126A. Golgi motility is dependent on actin dynamics (Boevink et al., 1998), so monitoring Golgi movements in the pollen tube provided another, albeit indirect, assessment of actin and pollen tube function. We coexpressed a Golgi marker, NAG:CFP, made from fusion of the first 79 amino acids of an A. thaliana N-acetylglucosaminyl transferase I to cyan fluorescent protein (CFP) (Dixit and Cyr, 2002), along with Pet PLC1-H126A. Importantly, NAG:CFP does not bind to actin and so is unlikely to generate artifacts as a result of potential direct actions on actin dynamics. Golgi are known to be distributed throughout the pollen tube but to be excluded from the secretory vesicle-rich tip (Hepler et al., 2001). In control pollen tubes, NAG:CFP labeled punctate structures characteristic of the Golgi apparatus (Figure 4B), as observed previously by Dixit and Cyr (2002). These structures moved at between 2 and 6 μm/s in the shank of the tube in a linear manner (Figure 4B), consistent with the Golgi tracking along actin microfilaments reported in a variety of cell types (Boevink et al., 1998). This motility was arrested by the addition of 5 μM Lat B, an actin-disrupting drug, again consistent with an actin-based motility system (data not shown). The labeled Golgi were excluded from the very apex of the tube, and once delivered to the cytoplasm just below the tip of the tube, they flowed around this subapical region before returning to linear streaming, back along the length of the tubes' shank (Figure 4B; see Supplemental Movie 6 online). In pollen tubes expressing Pet PLC1-H126A, the Golgi dynamics in the tube shank were unaltered relative to controls (Figure 4B; moving in linear streams at 2 to 7 μm/s and qualitatively following similar linear paths as the control). By contrast, at the tube apex, the Golgi now streamed through the end of the tube, protruding to the apex of the swollen tip (see Supplemental Movie 7 online). These observations are consistent with Pet PLC1-H126A expression disrupting the actin cytoskeletal organization that prevents Golgi from entering the tube tip. However, the continued Golgi motility in the shank of the tube and through the apex under these circumstances suggests that expression of Pet PLC1-H126A did not disrupt all actin-dependent functions within the cell.

Expression of the C2 Domain Alone in Pollen Tubes Phenocopied the Expression of Pet PLC1-H126A

The effect of Pet PLC1-H126A on pollen tube growth led us to ask whether the entire protein or a specific domain of Pet PLC1 was responsible for altering tip growth. Therefore, we produced the full-length Pet PLC1 and four different truncated proteins (Figure 5A ) in pollen tubes and assessed their effects on growth, Ca²⁺ gradient, actin structure, and Golgi dynamics (Figures 5B to 5E). In addition, we expressed the truncated constructs that retained the X+Y catalytic core in Escherichia coli and isolated the recombinant proteins to determine whether they still possessed PI-PLC activity. The truncated protein lacking the EF-loop domain (X+Y+C2), but not the truncated protein lacking the C2 domain (EF+X+Y), retained the catalytic activity (Pet PLC1 = 3.34 ± 0.29, X+Y+C2 = 4.14 ± 1.06, EF+X+Y= 0 μmol·min⁻¹·mg⁻¹). Expressing either the X+Y+C2 or the EF+X+Y construct in pollen tubes did not alter tube morphology or affect the growth-related tip-focused Ca²⁺ gradient (Figure 5B; data not shown). Similarly, expression of the EF-loop construct alone did not alter pollen tube growth or the tip-focused Ca²⁺ gradient (Figure 5B). By contrast, expressing the C2 domain construct alone was able to phenocopy the effects of Pet PLC1-H126A at the level of tube growth. These included swelling of the apex of the tube, alteration of the apical Ca²⁺ gradient (Figure 5C), formation of a subapical actin band (Figure 5D), and protrusion of Golgi to the very tip of the swollen pollen tube (Figure 5E). In a similar manner to Pet PLC1-H126A (Figure 4A), when tubes expressing C2 were treated with 5 nM Lat B, no bulging was apparent (Figure 5F).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Effects of the Expression of Various Truncated Forms of Pet PLC1 on Pollen Tube Growth.

(A) Schemes of the full-length and four truncated constructs of Pet PLC1, quantifying the effect of overproducing the full-length protein or producing each truncated protein on pollen tube morphology. The morphology of at least 80 pollen tubes that produced GFP from at least three independent experiments was examined for each construct.

(B) Effects of producing the EF domain alone or an EF+X+Y–truncated version of Pet PLC1 on tip growth and Ca²⁺ levels. Each construct was cobombarded with pLAT52:GFP, and the pollen tubes shown were imaged for GFP fluorescence (left two panels). The same pollen tubes were also imaged for cytosolic Ca²⁺ after they had been microinjected with the fluorescent Ca²⁺ imaging dye Indo-1-dextran (right two panels). The Ca²⁺ distribution was calculated by ratio imaging, and Ca²⁺ levels have been pseudo-color-coded according to the scale at right. Note that neither construct causes tip bulging or detectable alteration in the tip-focused Ca²⁺ gradient.

(C) Effect of expressing the C2 domain alone. Two separate pollen tubes were imaged for GFP fluorescence (left two panels) and Ca²⁺ (right two panels), as described for (B). Note that this construct causes apical swelling of the growing tube and expansion of the tip-focused Ca²⁺ gradient to fill the swollen apex.

(D) Actin distribution visualized by GFP:mTalin in a pollen tube producing the C2 domain alone. The arrowhead points to the subapical actin-rich band that forms in the swollen tip (st). pg, pollen grain.

(E) Golgi dynamics of a pollen tube producing the C2 domain alone as visualized by expression of the Golgi-retained NAG:CFP. Note that Golgi move to the extreme tip of the tube.

(F) Actin distribution visualized by GFP:mTalin in pollen tubes producing the C2 domain with or without 5 nM Lat B added at pollen tube germination.

The images shown are representatives of those obtained from >10 separate pollen tubes ([C], [E], and [F]), >20 separate pollen tubes (D), or >50 separate pollen tubes (B). Bars = 10 μm.

The similarity in the phenotypes of pollen tubes producing Pet PLC1-H126A and the pollen tubes producing the C2 domain suggested to us that both Pet PLC1-H126A and C2 might affect tube growth by a similar mechanism. In addition, our experiments localizing GFP:Pet PLC1-H126A (Figure 3E) were consistent with this inactivated enzyme binding to similar plasma membrane sites on the wild-type enzyme, raising the possibility that it exerted its effects on growth by displacing the endogenous PLC. Therefore, we coexpressed each of the Pet PLC1-H126A and C2 domain constructs with different molar ratios of GFP:Pet PLC1 to determine (1) whether the wild-type protein was mislocalized by expression of these other constructs and (2) whether excess wild-type protein could rescue the effect on pollen tube growth, as would be expected if Pet PLC1-H126A and the C2 domain acted in a dominant negative manner and competed with the endogenous PLC. Indeed, for both Pet PLC1-H126A and C2, GFP:Pet PLC1 was displaced from the apical plasma membrane as the tubes bulged. However, increasing the proportion of competing wild-type protein suppressed both the bulging phenotype and mislocalization (Figure 6 ). Together, these observations are consistent with both Pet PLC1-H126A and the C2 domain acting in a similar, competitive manner with the endogenous phospholipase.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Effects of Increasing the Relative Amounts of GFP:Pet PLC1 to Pet PLC1-H126A or the C2 Domain on Pollen Tube Morphology and GFP:Pet PLC1 Localization.

Numbers represent amounts of the construct DNA used in each bombardment relative to the 2 μg per bombardment of either the Pet PLC1-H126A or the C2 domain construct. Note that as the relative amount of wild-type GFP:Pet PLC1 is increased, the swelling of the tip induced by either the C2 domain or Pet PLC1-H126A is reduced. Similarly, as wild-type GFP:Pet PLC1 levels are increased, the localization of Pet PLC1 to the apical plasma membrane becomes more pronounced. Images are representative of >30 tubes per treatment from three independent experiments. Bar = 10 μm.

We reasoned that disruption of PLC localization/function by Pet PLC1-H126A or the C2 domain might alter the levels of PLC's substrate (PtdInsP₂) or its products, such as InsP₃. Such a change in PtdInsP₂ levels could lead to the disruption of the actin cytoskeleton, whereas altering the dynamics of InsP₃ might lead to changes in the apical Ca²⁺ gradient, both phenotypes we observed when expressing Pet PLC1-H126A (Figures 3D and 4A) or the C2 domain (Figures 5C and 5F). In the Rop-based signaling network of the pollen tube, phenotypes associated with the alteration of actin dynamics are suppressed by low levels (5 nM) of Lat B (Fu et al., 2001), whereas those related to an altered tip-focused Ca²⁺ gradient can change the Ca²⁺ requirement for pollen germination and growth (Gu et al., 2005). We observed that expression of Pet PLC1-H126A or the C2 domain did not alter the requirement for millimolar Ca²⁺ concentrations in the medium for pollen to germinate and undergo tip growth (e.g., 0% germination at 100 μM Ca²⁺, 70% for Pet PLC1-H126A, C2, or control at 3 mM [n > 100 tubes, three separate experiments]). However, germination in 5 nM Lat B did suppress the bulging of the tube tip and the formation of the subapical actin ring in pollen expressing either Pet PLC1-H126A (Figure 4A) or the C2 domain (Figure 5F). Addition of 5 nM Lat B to untransformed pollen had no observable effect on germination or growth, whereas addition of ≥1 μM Lat B led to arrested tip growth in both the control and transformed pollen tubes, irrespective of the construct being expressed (data not shown).

These results suggested to us that Pet PLC1-H126A might affect tip growth by altering PtdInsP₂ levels, thus affecting actin dynamics. Therefore, we expressed a GFP fusion to the PH domain of human PLC δ1 to investigate whether there were changes in PtdInsP₂ levels in the pollen tubes expressing Pet PLC1-H126A. This construct has been used to infer PtdInsP₂ localization in animal and plant cells, including tip-growing cells such as root hairs (Vincent et al., 2005) and pollen tubes (Kost et al., 1999). We observed that, as reported by Kost et al. (1999), high expression of this construct led to growth arrest in pollen tubes (data not shown). However, at lower expression levels (as inferred from the brightness of GFP), growth was sustained, and we observed that in growing pollen tubes the GFP:PH signal appeared higher toward the growing apex, whereas in nongrowing tubes this labeling was uniform throughout the pollen tube (Figures 7A and 7B ). This membrane localization suggested that GFP:PH selectively labeled the membrane lipid (i.e., PtdInsP₂) as opposed to its other potential reported ligand, soluble InsP₃ (Lemmon et al., 1995). When the intensity and distribution of GFP:PH fluorescence were compared with those of colabeling with FM 4-64, the two signals did not mirror each other, indicating that GFP:PH did not merely reflect the distribution of membranous material at the apex that is associated with growth. Thus, GFP:PH signal appeared locally enhanced in the regions where tip growth occurred (Figures 7A and 7B; see Supplemental Movies 8 and 9 online). In tubes induced to bulge by the expression of Pet PLC1-H126A or the C2 domain alone, GFP:PH decorated the entire plasma membrane of the bulge.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Localization of FM 4-64 and GFP:PH in Growing and Nongrowing Pollen Tubes.

(A) Pollen tubes expressing the GFP:PH construct were stained with 5 μM FM 4-64, and midplane confocal images were taken from pollen tubes that were actively elongating (growing), had arrested in elongation (nongrowing), or were expressing Pet PLC1-H126A or the C2 domain construct. Results are representative of >30 tubes from at least three separate experiments. Bar = 10 μm.

(B) Intensity profiles along the plasma membrane of GFP:PH signal normalized to the highest intensity pixel in each transect. One-pixel-wide transects were traced along the plasma membrane according to the image at right, and pixel intensities were recorded using Zeiss LSM 510 image-analysis software. Transects are representative of at least five separate measurements.

(C) Model of the role of Pet PLC1 and PtdInsP₂ in tip growth. When growth is slow, Pet PLC1 accumulates on the apical plasma membrane, cleaving PtdInsP₂; consequently, levels of apical PtdInsP₂ are low. Upon growth, Pet PLC1 leaves the plasma membrane, likely accumulating on the FM 4-64–stained endocytotic/exocytotic vesicle pool at the tip. PtdInsP₂ increases in the tube apex, sustaining growth through regulation of membrane dynamics and/or alterations in cytoskeletal structure. Expression of high levels of the inactive Pet PLC-H126A displaces endogenous PLC, leading to increased PtdInsP₂ throughout the tip and unrestricted, depolarized growth.

Assay of PI-PLC Activity

Pollen from 200 Petunia inflata flowers was collected in 25 mL of germination medium containing 0.07% (w/v) Ca(NO₃)₂·4H₂O, 0.02% (w/v) MgSO₄·7H₂O, 0.01% (w/v) KNO, 0.01% (w/v) H₃BO₃, 0.2% (w/v) sucrose, 15% (w/v) polyethylene glycol 4000, and 20 mM MES, pH 6.0. Pollen was grown for 5 h at 30°C with shaking at 200 rpm. Soluble and membrane fractions were prepared from these pollen tubes essentially as described previously (Drøbak et al., 1994). Briefly, pollen tubes were homogenized in 4 mL of extraction medium (50 mM Bis-Tris propane, MOPS, pH 7.5, 0.33 M sucrose, 5 mM EDTA, 1 mM DTT, 1 mM phenylmethylsulfonyl fluoride, and Complete protease inhibitor cocktail without EDTA [10 μg/mL]; Roche Applied Science) and centrifuged at 10,000g for 10 min. The resulting supernatants were centrifuged at 100,000g for 30 min. The microsomal pellet was resuspended in 200 μL of 10 mM HEPES/KOH, pH 7.5, 0.33 mM sucrose, 1 mM EDTA, 1 mM DTT, 1 mM phenylmethylsulfonyl fluoride, and Complete protease inhibitor cocktail without EDTA (10 μg/mL). Soluble and membrane fractions were then frozen with liquid N₂ and stored at −80°C. PI-PLC activity was assayed according to Coursol et al. (2000).

Screening of a P. inflata Pollen cDNA Library, and RNA Gel Blot Analysis

Floral cDNA was synthesized from total RNA isolated from 10 Arabidopsis thaliana flowers by SuperScript II RNase H⁻ reverse transcriptase (Invitrogen) as described by Skirpan et al. (2001). A 789-bp fragment of At PLC2, which included the coding sequences for both the X and Y catalytic domains, was obtained from total floral cDNA by PCR using 5′-GACTGGTAATCAGCTGAGCAG-3′ as the forward primer and 5′-ATTTGAGCACCGTGGCTC-3′ as the reverse primer. The PCR product was cloned into pGEM-T Easy (Promega), radiolabled with ³²P using the RTS RadPrime DNA labeling kit (Life Technologies), and used to screen a previously constructed P. inflata pollen cDNA library as described by Mu et al. (1994). RNA gel blot analysis was performed as described by Skirpan et al. (2001), except that 20 μg of total RNA was used. The blot was hybridized with a ³²P-labeled full-length Pet PLC1 cDNA clone isolated from the library screen.

Construction of Plasmids for Recombinant Protein Production in Escherichia coli

The full-length Pet PLC1 cDNA clone was used to generate by PCR all of the expression constructs of Pet PLC1 and its various truncated or mutant versions. Full-length Pet PLC1 was obtained using 5′-CCATGGCGTCGAAACAGACATA-3′ (primer 1) as the forward primer and 5′-GCGGCCGCTAGATAAATTCAAAACGATA-3′ (primer 2) as the reverse primer. Pet PLC1-H126A was amplified by overlapping PCR, using primer 1 (forward primer) and 5′-TAAGAATTGGCGCCCGTGTA-3′ (reverse primer) in one reaction and primer 2 (forward primer) and 5′-TACACGGGCGCCAATTCTTA-3′ (reverse primer) in the other. The DNA fragments generated by the first two reactions were used as a template for a third reaction in which only primer 1 (forward primer) and primer 2 (reverse primer) were used. A truncated Pet PLC1 containing the coding sequence for the X+Y+C2 domains was made using 5′-CCATGGCGCCTGTCAATCCTAAACGTAAG-3′ as the forward primer and primer 2 as the reverse primer. Another truncated Pet PLC1 containing the coding sequence for the EF+X+Y domains was amplified using primer 1 as the forward primer and 5′-GCGGCCGCTAATTGTTAGGACCAGCTTTCA-3′ as the reverse primer. The resulting DNA fragment contained at its 3′ end the codon for amino acid residue 446, followed by a stop codon. Each PCR product (flanked by NcoI and NotI sites included in the primers) was first cloned into pGEM-T Easy. After the sequence had been confirmed, the DNA fragment was released from pGEM-T and cloned into the NcoI and NotI sites of pRSET B (Invitrogen), which generates 6xHis-tagged recombinant proteins. GFP:Pet PLC1 was generated by first amplifying soluble GFP (sGFP) (Marc et al., 1998) by PCR using the forward primer 5′-CCATGGTGAGCAAGGGCG-3′ and the reverse primer 5′-CCATGGCACCTCCGCCTCCTCCGCCACCTCCGCCTCCCTTGTACAGCTCGTC-3′. This resulted in the deletion of the stop codon of sGFP and the insertion of a linker encoding 10 Gly and 1 Ala residues. The PCR product was then cloned into the NcoI site of the full-length Pet PLC1 in pRSET B.

Bacterial Expression and Protein Purification

Each expression construct was transformed into E. coli (BL-21 pLysS). An overnight culture in 1 mL of Luria-Bertani medium was transferred to 250 mL of Luria-Bertani medium, and the culture was grown at 37°C until OD₆₀₀ reached 0.6. Isopropylthio-β-galactoside was then added to a final concentration of 0.5 mM, and the culture was grown at 22°C for 16 h. Bacterial cells were pelleted and frozen at −80°C. All subsequent steps were performed on ice or at 4°C. The pellets were resuspended in 20 mL of sonication buffer (10% [v/v] glycerol, 20 mM Tris-HCl, pH 8, 50 mM NaCl, Complete protease inhibitor cocktail without EDTA, trypsin inhibitor [2 μg/mL], and 0.1% [w/v] lysozyme). The suspension was sonicated, and the lysates were centrifuged for 20 min at 10,000g. The supernatant was incubated with 0.5 mL of Talon resin (Clontech) for 2 h with gentle mixing. The resin was poured into a column and washed with 3 volumes of sonication buffer and then with 3 more volumes of sonication buffer containing 10 mM imidazole. Protein was eluted in 300-μL fractions with sonication buffer containing 100 mM imidazole. The purity of each fraction was assessed using 12% (w/v) SDS–polyacrylamide gels.

Construction of Single Expression Plasmids for Transient Expression in Pollen Tubes

All single expression plasmids used for biolistic experiments were constructed in a vector derived from the pBGD vector (James et al., 1996). The 2.8-kb SalI-EcoRI fragment, containing the LAT52 promoter, GUS coding sequence, and NOS terminator, was released from pLAT52-7 (Twell et al., 1990) and inserted into the SalI and EcoRI sites of pBGD to generate pLAT-GUS. The GUS sequence and the NOS terminator were then removed by digestion with NcoI and EcoRI and replaced with the NcoI–EcoRI fragment from a pUC18 derivative that contains the sGFP coding sequence and the NOS terminator to generate pLAT-GFP. Pet PLC1 (full length), Pet PLC1-H126A, X+Y+C2 (truncated Pet PLC1 lacking the EF-loop), and EF+X+Y (truncated Pet PLC1 lacking the C2 domain) from the E. coli expression constructs described above were released from their respective pRSET B constructs and ligated separately into the NcoI and NotI sites of pLAT-GFP (the GFP fragment was released in the cloning). The EF-loop domain was amplified by PCR using primer 1 (forward primer) and 5′-GCGGCCGCTAATGGTGATCTTACGTTTAGGA-3′ (reverse primer) to give a fragment starting at the initiation (start) codon within an NcoI site and ending at amino acid 102, which is followed by a stop codon and a NotI site. The C2 domain was amplified by PCR using forward primer 5′-CCATGGCGAAAGCTGGTCCTAACAATC-3′ and primer 2; this resulted in a fragment with an in-frame ATG within an NcoI site and ending at the stop codon and a NotI site. Both the EF-loop and C2 fragments were then cloned into the NcoI and NotI sites of pLAT-GFP as described above.

Construction of GFP:PH

GFP:PH was constructed by PCR of the coding sequence for the PH domain of human PLC δ1 essentially according to Kost et al. (1999), except that the sequence for a seven–amino acid linker (Gly-Gly-Ala-Gly-Ala-Ala-Gly) was inserted between the sequences for GFP and the PH domain. Aequorea coerulescens GFP (AcGFP) (Clontech) was amplified using 5′-TTAAGCTTCCATGGTGAGCAAGGGCGC-3′ (primer 1) as the forward primer and 5′-GTCTCCTGCAGCTCCAGCTCCACCCTTGTACAGCTCATCCATGCC-3′ as the reverse primer, and the PH domain from a cDNA clone of human PLC δ1 (Open Biosystems) was amplified from the coding sequence corresponding to amino acid residues 2 to 174 using 5′-GGTGGAGCTGGAGCTGCAGGAGACTCGGGCCGGGACTTCCTG-3′ as the forward primer and 5′-GCGGCCGCTAGATGTTGAGCTCCTTCAGGAAGTTCTGC-3′ as the reverse primer (primer 2). AcGFP:PH was amplified by primer 1 and primer 2 using both AcGFP and PH domain fragments as templates. This PCR product (flanked by HindIII and NotI sites included in the primers) was first cloned into pGEM-T Easy. After the sequence had been confirmed, the DNA fragment was released from pGEM-T and cloned into the HindIII and NotI sites of pLAT-GFP.

Construction of Dual Expression Constructs

Dual expression constructs were made to express both GFP and full-length Pet PLC1, or both GFP and Pet PLC1-H126A, from the same plasmid. The pLAT-GFP plasmid was used as a template for PCR with the forward primer 5′-AGGCTTCTCGACTCAGAAGGTATTGAG-3′ and the reverse primer 5′-AAGCTTCATGTTTGACAGCTTATCATC-3′. The resulting DNA fragment contained the 0.6-kb LAT52 promoter, GFP coding sequence, and NOS terminator. This fragment was cloned into pGEM-T Easy, released by digestion with EcoRI, and cloned into the EcoRI site of the Pet PLC1 and Pet PLC1-H126A single expression constructs, respectively.

Biolistic Transformation

Pollen grains collected from 10 flowers of greenhouse-grown P. inflata plants were suspended in 200 μL of germination medium and bombarded with DNA-coated gold particles (1.0 μm diameter) according to Fu et al. (2001). Gold particles (0.5 mg) were coated with 2 μg of plasmid DNA, except for GFP:mTalin, for which only 0.8 μg was used. For experiments in which GFP:Pet PLC1 was titrated with different amounts of the C2 domain construct or Pet PLC1-H126A, 2 μg of GFP:Pet PLC1 was used and C2 or Pet PLC1-H126A was added at between 0.5 and 4 μg. The bombarded pollen grains were incubated at room temperature for 30 min, washed into culture tubes with 1 mL of germination medium, and shaken at 200 rpm at 30°C for 5 h before observation using either a Diaphot 300 epifluorescence microscope (Nikon) or a Zeiss LSM 410 confocal microscope. GFP was imaged using 480 ± 20 nm excitation, 500 nm dichroic mirror, and >510 nm emission (Diaphot) or 488 nm excitation, 488 nm dichroic mirror, and 520 to 560 nm emission (LSM 410). CFP was monitored using 440 ± 20 nm excitation, 460 nm dichroic mirror, and 480 ± 20 nm emission (Diaphot).

Lat B Treatment

To determine the effect of Lat B on the actin organization in P. inflata pollen tubes, pollen grains were bombarded with GFP:mTalin alone, C2 domain construct + GFP:mTalin, or Pet PLC1-H126A + GFP:mTalin. Lat B (Calbiochem) in a 1 mM stock in DMSO was added to each bombarded sample to a final concentration of 5 nM, with the final DMSO concentration at <0.001% (v/v). DMSO was added to the germination medium of the controls at an equivalent concentration. Pollen tubes were imaged after 4.5 to 5 h as described above.

FM 4-64 Staining, Calcium Measurement, and NAG:CFP Imaging

Pollen tubes were stained according to Parton et al. (2001). FM 4-64 (Molecular Probes) was added to a final concentration of 10 μM into pollen germination medium and visualized after 15 min using a Zeiss LSM 510 Meta microscope (543 nm excitation from a HeNe laser, 488/514/543 nm dichroic mirror, and 560 nm long-pass emission filtration). For measurement of cytosolic Ca²⁺, pollen tubes of ∼50 μm in length were impaled with borosilicate glass micropipettes and pressure-microinjected with Indo-1 conjugated to 10 kD of dextran (Molecular Probes), and Ca²⁺ levels were measured as described previously (Bibikova et al., 1997). To image Golgi distribution, pollen tubes were biolistically transformed with a construct to express NAG:CFP, a protein made from a fusion of the first 79 amino acids of A. thaliana N-acetylglucosaminyl transferase I to CFP (Dixit and Cyr, 2002). Transformed pollen tubes were imaged using a Diaphot 300 inverted epifluorescence microscope, a CFP filter set (Chroma), and a SenSys cooled CCD camera (Roper Scientific). Individual Golgi were tracked through frames of time-lapse movies (3 s between frames) using IPLabs image-analysis software (Signal Analytics).

Accession Numbers

Sequence data from this article can be found in the GenBank/EMBL databases under the following accession numbers: Pet PLC1, DQ322461; At PLC2, AC074395; At PLC4, NM125257; Ld PLC1, AAW81961; Ld PLC2, AAW81960.

Supplemental Data

The following materials are available in the online version of this article.

• Supplemental Figure 1. Amino Acid Sequence Alignment of Pet PLC1 and Two Arabidopsis PLCs, At PLC2 and At PLC4.

• Supplemental Movie 1. Time Lapse of a Growing Pollen Tube Expressing GFP:Pet PLC1.

• Supplemental Movie 2. Time Lapse of a Growing Pollen Tube Expressing GFP:Pet PLC1 and Stained with 5 μM FM 4-64.

• Supplemental Movie 3. Time Lapse of a Nongrowing Pollen Tube Expressing GFP:Pet PLC1 and Stained with 5 μM FM 4-64.

• Supplemental Movie 4. Fluorescence Recovery after Photobleaching the Cytoplasm in the Tip Region of a Pollen Tube Expressing GFP:Pet PLC1.

• Supplemental Movie 5. Fluorescence Recovery after Photobleaching the Membrane along the Shank of a Pollen Tube Expressing GFP:Pet PLC1.

• Supplemental Movie 6. Golgi Dynamics in Wild-Type Growing Pollen Tubes.

• Supplemental Movie 7. Golgi Dynamics in Pollen Tubes Expressing Pet PLC1-H126A.

• Supplemental Movie 8. Time Lapse of a Growing Pollen Tube Expressing GFP:PH to Visualize PtdInsP₂ Localization.

• Supplemental Movie 9. Time Lapse of a Nongrowing Pollen Tube Expressing GFP:PH to Visualize PtdInsP₂ Localization.