The Pollen Receptor Kinase LePRK2 Mediates Growth-Promoting Signals and Positively Regulates Pollen Germination and Tube Growth

In flowering plants, the process of pollen germination and tube growth is required for successful fertilization. A pollen receptor kinase from tomato, LePRK2, has been implicated in signaling during pollen germination and tube growth as well as in mediating pollen (tube)-pistil communication. Here we show that reduced expression of LePRK2 affects four aspects of pollen germination and tube growth. First, the percentage of pollen that germinates is reduced, and the time window for competence to germinate is also shorter. Second, the pollen tube growth rate is reduced both in vitro and in the pistil. Third, tip-localized superoxide production by pollen tubes can not be increased by exogenous calcium ions. Fourth, pollen tubes have defects in responses to STIL, an extracellular growth-promoting signal from the pistil. Pollen tubes transiently over-expressing LePRK2-fluorescent protein fusions had slightly wider tips, whereas pollen tubes co-expressing LePRK2 and its cytoplasmic partner protein KPP (a Rop-GEF) had much wider tips. Together these results show that LePRK2 positively regulates pollen germination and tube growth and is involved in transducing responses to extracellular growth-promoting signals. temperature with occasional gentle agitation. After incubation, the pollen suspension was centrifuged for 5 min at 3000 x g and resuspended to a final concentration of 1 mg pollen/ml in complete PGM (“no STIL” treatment) or supplemented with 0.0003 Abs units (280 nm) of STIL/ μ l of PGM (“+STIL” treatment). In every experiment, each line included 3 replicates for each treatment. Pollen germination was carried out on a rotating shaker (50 rpm) for 3 hours at 28°C, in 24-well microplates and each well contained 400 μ l of the pollen suspension. After germination, the pollen suspension was transferred to 1.5 ml microtubes and 10X fixing solution (5.6% formaldehyde, 0.5% glutaraldehyde and 25% PEG 3350) was added to a final concentration of 1X. Samples were incubated with gentle agitation at 50 rpm for 30 min at 4°C. Fixed pollen tubes were observed using an Axiovert microscope (Zeiss) and 50 pictures were taken for each replicate with a digital camera (Diagnostic Instruments). Fifteen pictures were randomly selected and the lengths of all the pollen tubes in each picture were determined using AxioVision software (Zeiss), and averaged. Pollen tube lengths for each replicate were calculated as the average from all 15 values previously obtained. Control and STIL treatments were analyzed with the Student t-test using Prism version 4.03 for Windows (GraphPad). Germination experiments were repeated twice and since variances did not differ, data were pooled.

We confirmed the presence of the LePRK2 antisense construct in T1 plants of all six lines by genomic DNA PCR (data not shown). For each line, most of the T1 plants had about 50% GFP-expressing pollen, as expected for heterozygotes, but in at least one plant in each line, all pollen expressed GFP, consistent with these plants being homozygotes (Supplemental Table 1).
We self-pollinated heterozygous T1 plants of 4 lines, and counted the Kanamycin resistance to sensitive ratio of their progenies. Lines 2, 4 and 6 again gave a 3:1 ratio, and line 1 again gave a 1:1 ratio (Supplemental Table 1). We also pollinated wild-type tomato pistils with pollen from heterozygous lines 2 and 4, but didn't see any male transmission defects (Supplemental Table 1).
We pollinated wild-type tomato pistils with pollen from the putative homozygous antisense LePRK2 lines 1-6 and germinated the resulting seeds on medium containing kanamycin. All seedlings were resistant to kanamycin, confirming that these plants were homozygous for the construct. We obtained homozygous transgenic plants for all six lines and all subsequent experiments used homozygous plants.
To determine whether LePRK2 expression was reduced in these plants, we performed quantitative RT-PCR with LePRK2-specific primers, using total RNA of mature pollen as templates. Figure 1A shows that the LePRK2 expression level was significantly reduced in all 6 lines, to 20-30% of wild-type levels in lines 1 and 6, and to less than 5% in lines 2-5. LePRK1 is the closest homolog for LePRK2 among known tomato genes, with 54% identity in overall protein sequence, and 64% identity in overall nucleotide sequence (Kim et al., 2002). The longest identical DNA fragment between LePRK1 and LePRK2 is 26 nucleotides. To address the specificity of the antisense construct, we checked whether the LePRK1 expression level was also altered in LePRK2 antisense plants. Figure 1A shows that the LePRK1 mRNA level varied from 80% to 105% of the levels in wild-type pollen. We therefore concluded that LePRK1 expression was not affected significantly, and that LePRK2 expression was specifically reduced in mature pollen of lines 1-6.
To test whether the reduction of LePRK2 expression is maintained after pollen germination, we also examined the expression level of LePRK2 in pollen that had been germinated in vitro, and found that LePRK2 mRNA levels were reduced to about 10% of wild-type levels in line 1 and line double-blinded assays. After 10 hours the difference was readily apparent with the naked eye.
Wild-type pollen formed long tubes and overnight growth resulted in interlocked pollen tubes resembling a mat of fungal hyphae, whereas the antisense LePRK2 pollen tubes formed only small clumps (Fig. 3C). Figure 3B shows that pollen tubes of lines 1 and 6, which express ~10% of wild-type levels of LePRK2 (Fig. 1B), were slightly longer than pollen tubes of lines 2-4, which express less than 5% of wild-type levels of LePRK2 (Fig. 1B). The correlation between LePRK2 expression level and pollen tube length supports a positive role of LePRK2 in regulating pollen tube growth. These results were confirmed in the T2 and T3 generations.
To determine whether the reduced tube length of antisense LePRK2 pollen was due to a slower growth rate, to early termination of tube growth, or both, we measured pollen tube lengths after 2-4 hours, and calculated an average growth rate. The growth rate of line 4 pollen tubes was ~0.11mm/hour, half of the rate (~0.22mm/hour) for wild-type pollen tubes (Supplemental Fig. 1).
Although many tubes of antisense LePRK2 pollen continued to grow after 10 hours, many stopped growth much earlier. The reduction in growth rate is at least one of the causes for the reduction of tube length, but earlier termination of tube growth might also contribute.
To determine whether antisense LePRK2 pollen tubes also grew slower in pistils, we pollinated wild-type pistils with pollen from lines 2 and 4 as well as with wild-type pollen, then checked pollen germination status on the stigmas and recorded the time pollen tubes arrived at the ovary. On the stigmas, the wild type pollen and antisense LePRK2 pollen germinated 3-5 hours after pollination ( Fig. 4A-D). Wild-type or GFP pollen tubes arrived at ovaries 7-9 hours after pollination (growth rate estimated at 1.2 mm/hour), while pollen tubes of lines 2 and 4 did not reach the ovaries until 10-12 hours after pollination, with growth rates estimated at 0.9 mm/hour (Table 1 and Fig. 4). Thus antisense LePRK2 pollen tubes grew slower than wild-type tubes both in vitro and in vivo, although the difference was smaller in vivo.

Antisense LePRK2 pollen tubes have vacuoles near the tip and more frequent callose plugs
To determine what downstream processes might account for the slower growth of antisense LePRK2 pollen tubes, we observed the sub-cellular morphology of growing pollen tubes. Vacuoles perform multiple functions in plant cells, including the storage and degradation of cellular components, osmoregulation, and modulation of turgor (Bassham and Raikhel, 2000;Lew, 2004;MacRobbie 2006). Growing wild-type pollen tubes usually have large vacuoles (>5 μ m in diameter) at the very rear, while only thin tubular vacuoles are seen near the tip (Lovy-Wheeler et al., 2007). Many of the antisense LePRK2 pollen tubes had large vacuoles near the tip (Fig. 5A) and sometimes the front edge of these vacuoles were as near as 10 μ m from the tip. These large vacuoles were not present in early stages of tube growth, but were seen in about 50% of the tubes, starting as early as 5 hrs after germination. Upon further observation of such pollen tubes, larger and larger vacuoles moving closer to the tip were noted. If they reached the tip they should disrupt the normal organelle distribution of the tip and the tube should stop growth. However, continued observations of these pollen tubes showed that the large vacuoles sometimes moved away from the tip -this backward and forward movement kept the vacuoles near the front but away from the clear zone, allowing such pollen tubes to grow for quite a while (Supplemental movies 1-3).
Pollen tubes form periodic callose plugs to keep the cytosol and the sperm towards the front (Nishikawa et al., 2005). Figures 5B and 5C show that wild-type pollen grown in vitro had callose plugs that were spaced at regular intervals of ~350 µm, whereas callose plug placement in the antisense LePRK2 pollen tubes was more frequent and the intervals were more variable, occurring from 220-250 µm.

Antisense LePRK2 pollen tubes have impaired responses to Ca 2+
In tobacco, ROS production was detected at the pollen tube tip, and the ROS level was increased with exogenous Ca 2+ (Potocký et al., 2007). To determine whether there was a difference in ROS levels between wild-type and antisense LePRK2 pollen tubes, we stained growing pollen tubes with nitroblue tetrazolium (NBT), which reacts with O 2 − and forms a blue precipitate. NBT staining of wild-type tomato pollen tubes was darker in a medium with 1mM CaCl 2 than in a medium without CaCl 2 , indicating an increase in the ROS level with exogenous Ca 2+ (Fig. 6 A  Wild-type tomato pollen grew the longest tubes within a range of 1-3 mM exogenous Ca 2+ (data not shown). They had much shorter tubes without additional Ca 2+ in the medium, but still grew, probably due to the presence of endogenous Ca 2+ in pollen grain. In contrast, the pollen tube lengths of antisense LePRK2 lines 2 and 4 were not significantly different in medium with or without Ca 2+ (Fig. 6C), indicating that antisense LePRK2 pollen tubes have impaired responses to exogenous Ca 2+ . Boric acid can also stimulate pollen tube growth (Johri and Vasil 1961). Both wild-type tomato pollen and antisense LePRK2 pollen grew longer tubes in medium with 1.6 mM boric acid than in medium with 0.16mM boric acid (Supplemental Fig. 2), indicating that antisense LePRK2 pollen still responds to boric acid.

LePRK2 pollen
LePRK2 is specifically dephosphorylated by a component of style extract and this component can also cause dissociation of a complex that includes LePRK1 and LePRK2 (Wengier et al., 2003).
We speculated that the style component triggers LePRK2 signaling during pollen tube growth through dephosphorylation and complex dissociation, but the nature and the biological function of this style component remained unknown. Recently STIL, the factor in the style component that is responsible for LePRK2 dephosphorylation, was purified and shown to stimulate pollen tube growth (Wengier et al., submitted). Figure 7 confirms that adding STIL to pollen germination medium increased the length of wild-type and GFP-expressing pollen tubes. However, the lengths of antisense LePRK2 pollen tubes were not increased in the presence of STIL, suggesting that this stimulatory effect depends on LePRK2 expression.
Pollen tubes transiently over-expressing LePRK2 have slightly swollen tips, but pollen tubes

co-expressing LePRK2 and full-length KPP have much wider tips
Overexpression phenotypes are sometimes informative (Li et al., 1999;Kaothien et al., 2005). We attempted to obtain transgenic plants that over-expressed LePRK2, but were unsuccessful.
Therefore we used microprojectile bombardment of pollen (Twell et al., 1989a) to transiently overexpress a LePRK2-GFP fusion protein. The green pollen tubes were not shorter than untransformed pollen tubes, but their tips were 20% wider than tips of wild-type or GFP-expressing pollen tubes ( Fig. 8 A

and B). A similar swollen tip phenotype was seen with
LePRK2-RFP-expressing pollen tubes ( and full-length KPP had tips that were much wider than those on tubes expressing either LePRK2 or KPP fusion proteins alone ( Fig. 8G and Supplemental Fig. 3).

Discussion
Homozygous plants were readily obtained from all 6 antisense LePRK2 lines. This was surprising, especially from line 1, which had a 1:1 ratio for Kanamycin resistance to sensitivity, and from the lines with greatly reduced levels of LePRK2 mRNA. These results indicate that pollen with less than 5% of the normal level of LePRK2 expression still can deliver sperm for successful Cytosolic turgor has to be maintained for pollen tube growth (Benkert et al., 1997). Among other functions, vacuoles modulate turgor during cell growth (Hicks et al., 2004;Lew, 2004). In normal growing pollen tubes only thin thread-like vacuoles are usually seen in the front, while large vacuoles are usually seen in the rear, close to callose plugs (Hicks et al. 2004;Lovy-Wheeler et al., 2007). As growth proceeded in the antisense LePRK2 pollen tubes larger vacuoles were seen more often closer to the tip region. The phenomenon might be interpreted as compensation for reduced cytosolic turgor in the front. The periodic formation of callose plugs that separate the growing front from the evacuated regions is thought to be useful for maintaining a manageable cytosolic volume for pollen tubes, because pollen tubes need to extend many times the grain diameter to reach ovules (Nishikawa et al. 2005). The antisense LePRK2 pollen tubes had shorter intervals between callose plugs, giving a smaller cytosolic volume for the growing tube cell.
Although the callose plug interval difference can be interpreted in many different ways, it might be a compensation for or a result of the reduced cytosolic turgor, because less turgor would be required to maintain the force needed for a smaller cell volume to grow forward. The presence of large vacuoles in front of the tube and the more frequent callose plugs both suggest that cytosolic turgor in the antisense LePRK2 pollen tubes was reduced, and might explain why antisense LePRK2 pollen tubes can not grow as fast as wild-type tubes.
Consistent with the idea that pollen tubes with reduced LePRK2 expression have less cytosolic turgor, wild-type pollen tubes grew as slowly as antisense LePRK2 pollen tubes when cultured in germination medium containing 32% PEG (data not shown), which increases the external osmotic pressure. Pollen germination also requires the accumulation of cytosolic turgor to a threshold level to start protruding from the pollen grain (Taylor and Hepler 1997). In antisense LePRK2 pollen fewer pollen grains might have enough turgor to start germination, perhaps explaining the reduced germination percentage. Consistent with the idea that LePRK2 might regulate cytosolic turgor, overexpression of LePRK2-GFP caused slightly swollen tips, which could be interpreted as higher turgor. Swollen tips are also seen in depolarized growth, such as when a constitutively active ROP (Li et al., 1999) is overexpressed, but such depolarized growth is usually more extreme (so-called balloon tips), and also causes slower growth or arrest of pollen tube growth. The cell wall at the tip of pollen tubes is thin as it lacks callose or cellulose and contains mainly esterified pectins (Krichevsky et al., 2007). Under slightly higher cytolic turgor, the tip will be the first place to swell. In summary, we interpret the multiple morphological defects caused by altering LePRK2 expression as affects on cytosolic turgor. For the antisense LePRK2 pollen, the average reduction in vitro tube growth (50-70% less than wild type) was very significant, but the delay in arriving at ovaries was small (2-3 hours delay for a journey that normally takes 7-9 hours) and, except for line 1, the transmission ratio of the transgene was not significantly distorted. Considering that there is always more pollen on the stigma than the number of ovaries to be fertilized, only the earliest-arriving pollen tubes will  rpm. We observed pollen tubes four to seven hours after bombardment.

Transgenic tomato development
A GFP expression cassette and an antisense full-length LePRK2, both driven by the LAT52 promoter, were individually inserted into pCAMBIA2300 to obtain the pCAMBIA-antisense LePRK2 plasmid. Agrobacterium tumefaciens strain LBA4404 (Hoekema et al., 1983) carrying this plasmid was used to transform tomato (Solanum lycopersicon cv. VF36) as described (McCormick, 1991).

NBT and DAB staining of pollen tubes
Production of O 2 − was determined by its ability to reduce nitroblue tetrazolium (NBT) (Rossetti and Bonatti, 2001). Pollen tube staining was slightly modified from the method described in  5'-GGCCTGAAGTACAAGCAGTACAACA-3' and 5'-CGAACCAAACACACCGCTA-3'. The primers used to amplify a 123-bp fragment of a tomato actin gene (BG140412) were:

Style component preparation and treatment
STIL was purified as in Wengier et al. (submitted). Briefly, a pistil exudate was obtained by cutting 100 tobacco styles and stigmas transversely in 5 mm segments and incubating overnight in 50 mM ammonium bicarbonate (25 ml) at 4°C with gentle agitation. The pistil exudate was filtered through miracloth and filter paper and then subjected to chloroform-methanol extraction.
The aqueous phase was dried by rotary evaporation and the pellet was dissolved in water. The             WT1 and antisense LePRK2 line 2 were measured in experiment 1. WT2, GFP and antisense LePRK2 lines 3 and 4 were measured in experiment 2. B. Pollen tube length measured after 3 hour in vitro culture. 1-6 are antisense LePRK2 lines 1-6. C. Tomato pollen tubes from wild-type (left) or from antisense LePRK2 line 2 (right), after 10 hour in germination medium. In each plate, pollen was added at 1mg/ml. Arrows point to the "mats" of pollen tubes. The tip regions of a representative GFP-expressing pollen tube (control) and antisense LePRK2 pollen tube of lines 2 and 4. The pollen tubes were imaged after 5 hours in vitro culture. Scale bar=10μm. DIC: differential interference contrast. B. Representative pictures of pollen tubes with callose plugs stained by decolorized aniline blue. Wild-type and antisense LePRK2 pollen tubes after 10 hr and 12 hour in vitro culturing respectively. Insets show enlargements of the callose plugs. Scale bar=200 μ m. G: pollen grain. T: tip of pollen tube. C. The average interval lengths between callose plugs in pollen tubes of wild-type, GFPexpressing or antisense LePRK2 lines after 12 hours in vitro culturing. Error bar=standard error (n=3).