JXB Advance Access originally published online on May 23, 2006
Journal of Experimental Botany 2006 57(9):2001-2013; doi:10.1093/jxb/erj147
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
RESEARCH PAPER |
Style-by-style analysis of two sporadic self-compatible Solanum chacoense lines supports a primary role for S-RNases in determining pollen rejection thresholds
IRBV, Biology Department, University of Montreal, 4101 rue Sherbrooke est, Montreal, Canada H1X 2B2
*To whom correspondence should be addressed. E-mail: mario.cappadocia{at}umontreal.ca
Received 25 October 2005; Accepted 1 February 2006
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
A method for the quantification of S-RNase levels in single styles of self-incompatible Solanum chacoense was developed and applied toward an experimental determination of the S-RNase threshold required for pollen rejection. It was found that, when single style values are averaged, accumulated levels of the S11- and S12-RNases can differ up to 10-fold within a genotype, while accumulated levels of the S12-RNase can differ by over 3-fold when different genotypes are compared. Surprisingly, the amount of S12-RNase accumulated in different styles of the same plant can differ by over 20-fold. A low level of 160 ng S-RNase in individual styles of fully incompatible plants, and a high value of 68 ng in a sporadic self-compatible (SSC) line during a bout of complete compatibility was measured, suggesting that these values bracket the threshold level of S-RNase needed for pollen rejection. Remarkably, correlations of S-RNase values to average fruit sets in different plant lines displaying sporadic self-compatibility (SSC) to different extents as well as to fruit set in immature flowers, are all consistent with a threshold value of 80 ng S12-RNase. Taken together, these results suggest that S-RNase levels alone are the principal determinant of the incompatibility phenotype. Interestingly, while the S-RNase threshold required for rejection of S12-pollen from a given genetic background is the same in styles of different genetic backgrounds, it is different when pollen donors of different genetic backgrounds are used. These results reveal a previously unsuspected level of complexity in the incompatibility reaction.
Key words: Gametophytic self-incompatibility, single style analysis, S-RNase, Solanum chacoense, threshold
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Self-incompatibility (SI) is a widespread genetic mechanism used by many species of flowering plants to prevent inbreeding by promoting outcrossing. This prezygotic barrier is based on recognition of the gene products expressed in specialized cells of the pistil by those expressed in the pollen, which results in rejection of self- but acceptance of non-self pollen (de Nettancourt, 1977, 2001). The Solanaceae, Rosaceae, and Scrophulariaceae are characterized by gametophytic SI, or GSI, where the incompatibility phenotype of the haploid pollen is determined by its own genotype. In these families the male and female determinants to SI are both encoded at a highly complex and multiallelic S-locus. Pollen rejection occurs when the S-haplotype of the haploid pollen matches either of the two S-haplotypes of the diploid pistil, and it takes place inside the upper part of the style. The pistillar gene product to SI is a highly polymorphic ribonuclease termed S-RNase (McClure et al., 1989) that is synthesized by the cells of the transmitting tissue of the style and secreted into the surrounding extracellular matrix where the pollen tubes grow. The pollen determinant to SI (pollen-S gene product) has recently been identified as a polymorphic F-box protein, termed either SLF (S-locus F-box) or SFB (for S-haplotype-specific F-box) by the various authors (for details see Kao and Tsukamoto, 2004; McClure, 2004).
The cytotoxic action of the S-RNases mediates rejection of incompatible pollen by degrading pollen tube RNA in an S-haplotype-specific manner, although the minimal amount of S-RNase required for pollen rejection has not been determined. S-RNases have been shown to enter and accumulate inside the pollen tubes in a haplotype-independent manner (Luu et al., 2000), suggesting that the pollen contains proteins able to inhibit or destroy S-RNases. The mechanism whereby S-RNases penetrate inside the pollen tubes, however, is unknown. It has been suggested that this may occur either by endocytosis, via inclusion into a membrane-bound compartment (McClure, 2004) or through a receptor (or a receptor complex) that recognizes a conserved domain of the S-RNase (Kao and Tsukamoto, 2004). In this regard, the involvement of one of the most attractive possibilities (the conserved C4 region in S-RNases) in uptake has recently been ruled out (Qin et al., 2005).
Permanent self-compatibility (SC) has been reported several times among SI species (for a review, see de Nettancourt, 1977). In most cases, it can be attributed to mutations directly affecting either the pistillar or the pollen determinants to SI (de Nettancourt, 2001). Examples of the former include mutations at the S-RNase gene causing loss of the RNase activity as reported in Lycopersicon (Kowyama et al., 1994; Royo et al., 1994) and Petunia (Huang et al., 1994; McCubbin et al., 1997), or deletion of the S-RNase gene itself (Sassa et al., 1997). With regard to self-compatibility resulting from pollen-part mutations, it is most often associated with the so-called competition effect that takes place when two distinct pollen-S genes are expressed in the same pollen grain. An extra pollen-S gene introduced into a host plant by transgenesis (Qiao et al., 2004; Sijacic et al., 2004) produces the same effect. In some instances, pollen compatibility has been shown to result from the loss of pollen function (Tsukamoto et al., 2003), or from mutations affecting the pollen S-gene (Ushijima et al., 2004; Sonneveld et al., 2005), or from deletion of the pollen S-gene itself (Sonneveld et al., 2005). This last case is particularly important as it suggests S-RNases are inactive in pollen tubes without their cognate pollen-S, as predicted by the two-component inhibitor model (Luu et al., 2000, 2001). Lastly, other cases of SC have been shown to depend on so-called modifier genes, located outside the S-locus, that appear to be required for proper manifestation of the SI response, such as HT-B (O'Brien et al., 2002) or a stylar 120 kDa glycoprotein in Nicotiana (Hancock et al., 2005) (for a further discussion see Kao and Tsukamoto, 2004).
A special category of partial incompatibility is represented by pseudo-self-compatibility (i.e. formation of fruits containing variable amounts of seeds observed after crosses expected to be incompatible) (Clark et al., 1990) and sporadic self-compatibility (i.e. occasional fruit formation after crosses expected to be incompatible) (de Nettancourt et al., 1971; Qin et al., 2001). In particular, sporadic self-compatibility has been observed in some but not all S12-containing genotypes of Solanum chacoense, and is characterized by occasional bouts of self-compatibility with S12 pollen that can affect from 10% to 60% of the styles on a given plant. Expression of the S12 allele has been analysed in several plant lines and genotype-specific differences were found in the amount of S12-RNase and S12-mRNA. As sporadic self-compatibility occurred only in those genotypes with the lowest average S12-RNase levels (Qin et al., 2001), it is proposed that there may be a variation between flowers that could result in levels of S-RNase in some individual styles too low to reject otherwise incompatible pollen. Indeed, style-to-style variations in S-RNase levels could explain both pseudo and sporadic compatibility but has not previously been demonstrated.
Weakening of the SI response, associated with a reduced level of S-RNases present in the pistil, has led to the hypothesis that a threshold level of the RNase is required for full expression of the SI phenotype (Clark et al., 1990). Support for this idea has been provided by studies on partially compatible Japanese pear cultivars displaying low levels of S-RNase expression (Hiratsuka et al., 1999, 2001; Zhang and Hiratsuka, 2000; Hiratsuka and Zhang, 2002), and by the partial incompatibility of plants expressing an S-RNase transgene at levels significantly below those produced by the endogenous alleles (Lee et al., 1994; Murfett et al., 1994; Matton et al., 1997, 1999; Qin et al., 2005). Finally, accumulation of the S-RNases in the style during flower development is temporally regulated and the increase in S-RNase levels correlates with the acquisition of the incompatibility phenotype (Xu et al., 1990; Clark et al., 1990; Zhang and Hiratsuka, 2000). All these examples are consistent with the hypothesis that a threshold level of S-RNase is required to inhibit the growth of incompatible pollen tubes. However, the threshold itself has never been measured, and the factors that potentially influence it (S-RNase haplotype, pollen genotype, environmental conditions) have not been assessed.
The aim of the present study was the experimental determination of an S-RNase threshold and a preliminary evaluation of the factors that may influence it. To do so, advantage was taken of the sporadic self-compatible phenotype of some of our S12-RNase containing plant lines (Qin et al., 2001). A technique was developed for measurement of the S-RNase levels in single styles, and a definite S-RNase threshold for a particular pollen haplotype in a particular genetic background was found. However, it was also found that the S-RNase threshold required for rejection of this pollen haplotype can vary depending on the genetic background of the pollen donor. It was also noted that the S-RNase threshold differs when incompatibility is defined either by fruit formation or by the lack of pollen tubes entering the ovarian region.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Plant genotypes
The plant material used in these experiments includes the fully self-incompatible G4 (S12 S14), VF60 (S12S12), 582 (S13S14) genetic lines, as well as the two sporadically self-compatible L25 (S11S12) and 314 (S11S12) lines of Solanum chacoense (2n=2x=24) described previously (Qin et al., 2001). In addition, an individual called TP48 (S12S12) issued from the selfed 314 line (Qin et al., 2001), and a plant named 2548 (S12S12) produced by crossing L25 as pistillate parent with TP48 as staminate parent, and selected for its high vigour, pollen fertility, and high in vitro regenerability, were also used.
Genetic crosses
Genetic crosses on recently open flowers were always made with fresh pollen collected from plants of known S-haplotype constitution grown in the Montreal Botanical Garden greenhouses at 23±2 °C under natural light conditions. Pollen viability was estimated by staining with aceto-carmine. Bud pollinations were performed on flower buds at 3, 2, and 1 d before anthesis (DBA). Crosses were classified as fully incompatible if there was no fruit formation after pollination, and compatible when fruits were formed after almost every pollination. Where appropriate, pollen tube growth was monitored by staining the styles with aniline blue about 48 h after pollination, followed by observations by fluorescence microscopy as previously described (Matton et al., 1997). In some cases, the styles were observed 72 h after pollination. In other cases, flowers were gently shaken (touched) 4–5 d post-pollination, and the styles of fallen flowers were observed by fluorescence microscopy to determine if the lack of fruit set could be attributed to self-incompatibility.
Progeny analysis
Seeds obtained from bud pollination of the L25 line selfed at 3 DBA were germinated in vitro as described previously (Van Sint Jan et al., 1996), and the S-constitution of the resulting plantlets assessed by PCR. For each genotype, five leaf discs of 2 mm diameter were crushed with a plastic mortar in 20 µl 0.25 N NaOH, and incubated for 5 min at 95 °C. The mixture was then neutralized with 20 µl 0.25 N HCl, 20 µl TRIS-HCl pH 8.0 and 0.5% w/v Igepal CA-630 (Sigma). The tubes were incubated for five additional minutes at 95 °C, centrifuged for 1 min at 5000 rpm, the supernatant collected, and immediately used for the PCR reactions (40 cycles of 94 °C 30 s, 55 °C 30 s, and 70 °C 1 min) using a commercial PCR buffer (Promega) and Taq polymerase (Promega). The primers used for analysis of the S11 allele were 5'-CTATTTCAGTGTAAGCAGC-3' and 5'-ATTTCTAGAGGACGAAAAAATATTTTC-3', while primers 5'-TAACTTGACCACCACCG-3' and 5'-GTCATGGAAATGTAACCC-3' were used for the S12 allele.
Expression of S11- and S12-RNases in E. coli
The cDNA clones encoding S11- and S12-RNases were first mutated at the active site to avoid possible RNase activity toxic to the E. coli host cells, then cloned into an expression vector pQE30 (Qiagen, Valencia, CA). For the S11-RNase, the histidine encoded by CAC in the conserved region C2 was substituted with arginine (CGT). The three primers used for site-directed mutagenesis were S11HisSmaI (at the C-terminal end of the coding sequence), 5'-CTCTCTCTCT
CAAGGACGAAAAAATATTTCC-3'; S11HisSacI (corresponding to the N-terminal end of the mature coding sequence), 5'-GAGAGAGAGA
AAATTGCAACTGGTATTA-3'; and S11cmc2 (containing the substituted codon), 5'-CCTTATCCGGCCAAAGACCACGAATCGTAAAGTTTTTTG-3'. For the S12-RNase, the histidine encoded by CAT in the C2 conserved region was substituted with arginine (CGT). Two pairs of primers, S12 HindIII (at the C-terminal end of the coding sequence), 5'-CTCTCTCTCT
GGAAATGTAACCCCGGTA-3'; MuS12-HisA, 5'-AACTTTACAATCCGTGGGCTTTGGCCC-3' and S12 BamHI (corresponding to the N-terminal end of the mature coding sequence), 5'-GAGAGAGAGA
GAGCAGTTGCAACTGGT-3', MuS12-HisB, 5'-GGGCCAAAGCCCACGGATTGTAAAGTT-3' were used to amplify the mutant clone. The mutated cDNA fragments were cloned into pQE30 between the SmaI and SacI sites for S11 and between the BamHI and HindIII sites for S12. The sequences of both mutated clones were confirmed by sequencing and named pQE30-S11-his
-C2 and pQE30-S12-his-C2.
These two plasmids were transformed separately into competent M15 cells, and protein expression was induced by adding IPTG to a final concentration of 1 mM. The cultures were typically grown for 4–5 h before the cells were harvested. The target proteins were purified with Ni-NTA resin (Qiagen, Maryland, USA), electrophoresed on SDS-PAGE, then eluted from gel slices using an Electro-Elutor (Bio-Rad, California, USA) following the manufacturer's protocol. The purity of both proteins was confirmed by Coomassie blue staining after SDS-PAGE and western blot analysis.
Quantification of the standard S11 and S12 proteins
The concentration of purified S11 and S12 proteins was determined both by OD280 and by the Micro BCA Protein Assay Kit (Pierce Inc., Illinois, USA) following the manufacture's protocol. The two measures gave similar results and were thus used to calibrate S-RNases measurements in styles of S. chacoense. One batch of each purified protein was used as the standard for all gels.
Western blots and quantitative analysis of S11 and S12 RNases in individual styles
For western blots, plant styles were collected and frozen immediately in liquid nitrogen. The term ‘style’ as used here comprises both the style per se as well as the stigmatic region. Proteins were extracted from individual styles using 50 µl extraction buffer (0.05M TRIS pH 8.5, 1 mM DTT, 1 mM EDTA, 0.05 M CaCl2, 1 mM PMSF). Typically, 5x SDS sample buffer was added to 25 µl crude extract from each individual style and electrophoresed on SDS-PAGE. Immunological detection of the S-RNases is linear over the range used, so only one aliquot containing 400 ng of each purified S-RNase was run on each gel for standardization. The proteins were transferred to nitrocellulose membranes and stained with 2 (w/v) Ponceau red as a control for a uniform protein load; S-RNase measurements from samples with visibly different protein loads were excluded from these analyses. The membranes were blocked by an overnight incubation with TBS-T (TRIS-buffered saline containing 1.5% w/v BSA fraction V (Sigma) and 0 w/v Tween 80), incubated for 2 h at room temperature with a 1:1000 dilution of either polyclonal anti-S11 (Matton et al., 1999) or anti-S12 (Qin et al., 2001) antibodies, rinsed with TBS-T three times, and incubated with 4 µl of 0.5 µCi mmol–1 I125-protein A (Perkin-Elmer) in 5 ml TBS-T for 2 h. After washing three times with TBS-T, the membranes were exposed with a Phosphor screen for 12–76 h at room temperature and the screen imaged using a PhosphorImager scanner (Amersham Bioscience). The data on the scanned images were quantified using the software supplied by the manufacturer. Both antibodies used are specific for their respective substrates.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
S-RNase levels can be reproducibly assayed in single styles
The threshold hypothesis for S-RNase-mediated pollen rejection posits that an incompatible phenotype requires a minimum level of S-RNase within the styles. To determine this level experimentally, it was first necessary to develop a technique that would permit the absolute levels of S-RNase to be accurately and reproducibly measured in extracts from single styles. This assay thus requires two elements, a sensitive detection system and a calibration method to calculate the amount of S-RNase at ng level. For the latter, S-RNase protein standards were prepared by expressing an inactive S-RNase as a His-tagged construct in bacteria. Two tagged constructs, an S11-RNase and an S12-RNase, were purified by Ni-affinity chromatography and SDS-PAGE elution, and were homogenous by the criteria of Coomassie blue staining (Fig. 1).
|
For development of a sensitive and easily quantifiable assay, western blotting was coupled using an I125-labelled protein A with detection using a PhosphorImager. To characterize the response to the antibodies, standard curves using different amounts of both S-RNases were prepared. The immunological response is linear under these conditions (Fig. 1C). To characterize the precision of the measurements, S11-RNase levels in each of four equal aliquots of an individual style extract were compared (Fig. 2). These S11-RNase measurements have a coefficient of variation of only 5% and show that the method can faithfully assess the RNase levels in individual styles.
|
Since there was interest in measuring levels of both S11- and S12-RNases in single style extracts, tests were carried out to see if single transfers could be analysed using both anti-S11- and anti-S12-RNases. In one experiment, two membranes containing half the extracts from 18 styles of plant 314 (S11S12 genotype) were prepared. One membrane was treated first with the anti-S11-RNase and the S-RNase levels quantitated using the PhosphorImager. This membrane was then stripped and treated with the anti-S12-RNase (Fig. 3A). A second membrane, containing the other half of the same samples was similarly treated except that the order of the two antibodies on the membrane was reversed (Fig. 3B). An additional 18 styles from the plant L25 (also S11S12) were analysed in parallel using the same protocol (Fig. 3C, D). All samples show a substantial decrease in calculated S11-RNase (Fig. 3E) and S12-RNase (Fig. 3F) levels when measured on previously used membranes. It was concluded that this technique does not accommodate multiple analyses from the same membrane. However, as reliable measurements were obtained from one-half of the stylar extracts, both S-RNases can be measured in a single style extract by simply preparing two membranes from each sample. It was also noted from this preliminary study that a substantial difference in the average levels of S11- and S12-RNases can be observed in the styles of both plant lines.
|
16) S11-RNase levels calculated on a per-style basis after PhosphorImager detection in styles of L25 and 314 plants when the anti- S11-RNase was used first (lightly stippled bars) or second (darkly stippled bars). (F) The average ±SD (n It has previously been shown that average S12-RNase levels in pooled stylar extracts were dependent on the plant genotype used (Qin et al., 2001). These observations are confirmed here using the single style measurements of S12-RNase in sporadically self-compatible S11S12 lines L25 and 314 (Fig. 4A, B), the strictly incompatible S12 homozygote line VF60 (Fig. 4C), and the S12S14 G4 line (Fig. 4D). As expected, average S12-RNase levels are low (86±55 ng) in styles from L25 plants, intermediate (136±86 ng) in those from 314 plants, and high in VF60 (269±55 ng) and G4 (301±93 ng) plant styles (Fig. 4E). More important, however, are the style-to-style variations observed within each plant genotype which here can vary up to 3-fold (Fig. 4F). Interestingly, as S-RNase levels in the fully incompatible lines VF60 and G4 can be as low as 160 ng (Fig. 4F), this level of S-RNase must lie above the minimum required for pollen rejection. The S12-RNase appears as a doublet in L25 and 314 but not in G4 or VF60 due to genotype-dependent glycosylation differences (Qin et al., 2001). Samples with markedly different Ponceau staining (for example, Fig. 4B, lane 5) were excluded from further analysis.
|
Experimental determination of S12-RNase thresholds for pollen rejection
In contrast to the fully incompatible VF60 and G4 lines, L25 and 314 plant lines have previously been demonstrated to display a sporadic self-compatibility (SSC) phenotype (Qin et al., 2001). Although the SSC phenotype has not yet been traced to a specific environmental or physiological cause, the fact that only a fraction of the L25 or 314 styles pollinated set fruit (Qin et al., 2001), coupled with a large standard deviation in measured S12-RNase levels, suggested that a style-by-style comparison of S12-RNase levels and fruit set might allow the threshold levels required for pollen rejection to be determined.
During the course of these experiments, 14 styles were collected from L25 plants during a bout of almost complete compatibility with S12 pollen (31 fruits set from 33 flowers pollinated with S12 pollen from line 2548). The maximum amount of S12-RNase in the styles sampled was 68 ng (Fig. 5A, B), and the complete compatibility phenotype during this period suggests that this amount of S-RNase must lie below the minimum level required for pollen rejection.
|
The S-RNase levels in individual styles described above appear to define the upper limit for pollen acceptance and the lower limit for pollen rejection, yet two problems are associated with this conclusion. First, in these cases the phenotype is complete (i.e. no pollinations produce fruits), and to determine more precisely the threshold, additional data must of necessity use incomplete phenotypes (i.e. where some pollinations are observed to set fruits). Since S-RNase levels and the incompatibility phenotype cannot both be measured in the same style, correlative techniques are thus necessary. Second, it is possible that other factors might contribute to pollen rejection. It is therefore necessary to repeat these correlative experiments under several different conditions to ensure that the effect observed can indeed be ascribed to the S-RNase itself. To address these issues, S-RNase levels were examined in the SSC plant line L25 at several developmental stages. In one series of experiments, S-RNase levels were measured in individual styles of L25 plants at the time of anthesis (Fig. 5C). The average fruit set in these plants with S12 pollen from line 2548 is 55%, and if the S-RNase level was the principal contributor to the SI phenotype, then a threshold value of 80 ng S12-RNase would leave this proportion of individual data points below the line. It was also noted that the levels of S11-RNase are far above this in agreement with a full S11 pollen rejection phenotype if the threshold was similar (Fig. 5H).
In a second series of experiments, styles were taken between 1 d and 3 d before anthesis (DBA), a timing based on the size and morphology of the flower buds (Fig. 5D–F). Again, the predicted threshold value of 80 ng S-RNase is able to account for the incompatibility phenotype observed with S12 pollen from line 2548 (Table 1). It must be noted that complete compatibility with flowers pollinated at 3 DBA is not obtained, even with completely compatible pollen. This is due to the fragility of the buds, the tendency of the stylar tissues to dry after opening of the buds, and the fact that the stigma is apparently only partially receptive, as assessed by microscopic observations of fewer pollen grains that adhered and germinated.
|
Next, a similar series of experiments was performed with the styles taken from 314 plants. Once again, styles were analysed at anthesis (Fig. 6A) as well as between 1 d and 3 d prior to flower opening (Fig. 6B–D). The general pattern of S11- and S12-RNase values is similar to that observed in L25, with S11-RNase high at anthesis and at 2 d prior to flower opening (Fig. 6E) while S12-RNase levels were generally lower at anthesis and substantially lower 2 d prior (Fig. 6F). A predicted pollen rejection phenotype based on the number of individual styles with S12-RNase levels of less than 80 ng (32%) agrees well with the observed results using S12 pollen from line 2548 (25%) (Table 1) suggesting that the threshold S12-RNase level is similar in styles from 314 and L25 plants. Taken together, therefore, these data support the idea that a threshold of 80 ng S12-RNase is sufficient to block fruit set after pollination with S12-pollen.
|
Progeny analysis
Bud pollination of plant L25, selfed at 3 DBA, resulted in fruit formation. Plants were raised in vitro from the seeds and analysed by PCR using S-allele-specific oligos to assess whether the three possible genotypes (S11S11, S11S12, and S12S12) were present in the expected ratios. Of the 202 F1 progeny, 90 were S11S12, 112 were S12S12, and none were S11S11 (x2=59.17, P <0.001). It is concluded that S11 pollen is fully rejected even at 3 DBA, despite the observation of low (53 ng) S11-RNase levels measured in one out of six styles analysed (Fig. 5). It is possible that the threshold for S11-RNase is lower that the 80 ng threshold estimated for S12-RNase. Alternatively, the steady increases in S11-RNase observed between 3 and 2 DBA may be sufficient to block S11 pollen tubes before they reach the ovary.
S12-RNase thresholds differ for different pollen types
The S12-RNase threshold estimated from pollen rejection phenotypes using pollen from plant 2548 is similar when styles of L25 and 314 plants are compared (Table 1). However, this experiment does not address the potential influence of the pollen itself on the estimated thresholds. Thus, in another experiment, the pollination efficiency of four types of pollen from different genetic background was compared on L25 styles. All these pollen show similar viability (based on their appearance after staining with acetocarmine) and good germination. Despite this, major differences are observed in pollination efficiency (Table 2), and this suggests that the S-RNase thresholds may be different for the different types of pollen. Indeed, if thresholds are estimated as before by placing an arbitrary threshold line in Fig. 5G at a value where the proportion of styles with lower S-RNase levels corresponds to the percentage fruit set, the estimated S12-RNase threshold differs by up to 5-fold (Table 2). It is concluded from this that the S-RNase threshold must be defined for each particular pollen type.
|
Microscopic observations
To complement the fruit set measurements on genotypes L25 and 314, microscopic observations were also made to assess the behaviour of pollen tubes inside the styles 2–3 d after pollination. Following self-pollination, S12 pollen tubes were typically observed in the middle or lower third of the style, unlike S11 pollen tubes, which all arrested in the upper third of the style. Following pollination with S12 pollen from VF60 or plant 2548, however, a large variation was found in the number of pollen tubes observed at the stylar basis. In some cases, numerous pollen tubes were observed to have entered the ovary, and the appearance of the stylar bases were indistinguishable from compatible pollinations (usually yielding about 120 seeds) (Fig. 7A). This was interpreted as being consistent with a low level of S12-RNase in these styles. In other cases, no tubes were observed at the stylar base, which was interpreted as consistent with a high level of S-RNase (Fig. 7B). In other instances, a reduced number of pollen tubes (from one up to ten, but most often one or two) can be seen to have reached the base of the style and to have entered the ovarian region. Some of these pollen tubes may arrest just after their entrance into the ovarian region, while others can be observed to penetrate inside the ovules (Fig. 7C, D). Curiously, however, fruits containing only one or two seeds were never observed. Indeed, the smallest numbers that were ever observed were four and five seeds in two different fruits and, in both cases, they were accompanied by numerous aborted seeds and swollen ovules.
|
These observations suggested that a single pollen tube reaching the ovary may be insufficient to allow fruit set, and that assessment of SI either by number of pollen tubes at the stylar base or by fruit set may differ. To test this, the simple expedient of collecting, 4–5 d after pollination of L25 or 314 plants with S12 pollen, flowers that would fall when gently shaken was used. These incompatible crosses were then examined microscopically to determine the number of pollen tubes at the stylar base. Although in many cases the pollen tubes were all arrested in the middle or lower third of the style, instances were also found where up to ten tubes could be found entering the ovarian region (Fig. 7E, F). It was deduced from this that fruit formation may require more than 10 pollen tubes entering the ovarian region, although it was noted as a caveat that this value may include some slow growing tubes that reach the ovary after the process leading to flower abscission has already been initiated. As a complement to these studies, 24 styles of L25 plants pollinated during its bout of almost complete self-compatibility were also examined at 72 h post-pollination. While 21 styles had 20 or more (in most cases, uncountable) pollen tubes in the ovarian region, in the remaining cases only 14 (two cases) or 15 (one case) pollen tubes were found in the ovarian region. Since the ovarian regions contained 14 pollen tubes in these compatible crosses, and ten in the incompatible crosses described above, these results suggest that a number of pollen tubes in excess of this 10–14 tubes threshold must enter the ovary to ensure fruit formation.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Style-to-style variations in S-RNase levels can be used to estimate the S-RNase threshold
A considerable natural variability in the amount of S-RNases present in individual styles has been found. The extent of this variation was surprising, as individual styles can differ in S12-RNase levels by over 20-fold. This phenomenon has not previously been reported, and it was observed here only because S-RNase levels had been measured for individual styles. Interestingly, these variations neatly account for the observations reported here in which fruit set for an individual style is an all-or-none phenomenon but where different styles on the same plant may behave differently. It is not believed that these variations represent technical problems with the assay since our standard curves are linear (Fig. 1) and replicates from the same biological samples have a low coefficient of variation (Fig. 2).
These variations in individuals can be exploited to estimate the threshold below which S-RNases are ineffective in pollen rejection. For example, when an S-RNase expressing plant is completely compatible (Fig. 5B), the highest stylar level found (68 ng) defines the highest point still below the threshold. Similarly, when the plant is completely incompatible (Fig. 4F), the lowest value of S-RNase obtained (160 ng) defines the lowest point still fully capable of pollen rejection.
Do these values truly reflect the existence of an S-RNase threshold? This question is not trivial, as other factors such as HT-B (O'Brien et al., 2002) or a stylar 120 kDa glycoprotein in Nicotiana (Hancock et al., 2005) have been shown to be required for pollen rejection. However, if a factor other than the S-RNase were to contribute substantially to pollen rejection, a repeated correlation between incompatibility and S-RNase levels would be unlikely. It is thus significant that the thresholds estimated from two different plants lines (L25 and 314) under all these conditions are so similar. Furthermore, the same threshold accounts for the acquisition of the incompatibility phenotype in developing flowers (Figs 5G, 6E), long cited as support for the threshold hypothesis itself. Taken together, it is proposed that the numerous times the same correlation is observed provides strong support for the idea that S-RNases play the primary role in determining the SI phenotype. If so, this in turn suggests that the S12-RNase threshold required for pollen rejection is indeed around our value of 80 ng per style.
S-RNase concentrations inhibiting pollen tube growth are remarkably similar in different plants
The value proposed here for the threshold of the S12-RNase in S. chacoense should be considered in the context of the average amounts of S-RNase present in the styles of other fully self-incompatible Solanaceae genotypes. Average S-RNase levels have appeared in a number of different reports, and it must first be noted that these values can differ markedly depending on the method used. For example, calculations of the RNase levels in S2S2 homozygous self-incompatible N. alata based on the amount of S-RNases purified from styles yielded estimates of 10 µg (Jahnen et al., 1989a, b) or 
20 µg (Gray et al., 1991) S2-RNase per style. By contrast, measurements based on comparisons of staining intensity of an S2-RNase band after electrophoresis of crude extracts with that of the purified S-protein, yielded estimates of 90 µg S2-RNase per style in the same plants (Harris et al., 1989). The lower levels calculated after protein purification may perhaps be attributable to losses during purification. Alternatively, the higher levels found after electrophoresis may have been due to the presence of contaminating proteins with the same electrophoretic mobility as the S2-RNase. This study's protocol, which compares immunoreactivity in crude extracts with those of a pure standard S-RNase, appears less susceptible to experimental errors of either type. Our values about 1 µg S11-RNase and 0.1 µg S12-RNase in the styles of L25 plants (Fig. 3) appear lower than that found in Nicotiana, but it must be borne in mind that Nicotiana styles (15 mg) weigh roughly ten times more than S. chacoense styles (1.3 mg). Indeed, the calculated concentrations of S-RNases in styles of S. chacoense (0.25–1 mg ml–1) and N. alata (0.5–5 mg ml–1) are quite similar.
The calculation of S-RNase thresholds in vivo is difficult, as in general the genotypes studied are fully incompatible and the degree to which their S-RNase levels surpass the threshold can therefore not be ascertained. Even in plants where some degree of compatibility is observed, measurements of S-RNase levels and incompatibility phenotype in an individual style are mutually exclusive. Methods employing correlations in populations of styles are thus required. One such approach has been taken with the Japanese pear (Rosaceae), where some cultivars can be up to 15% compatible (Hiratsuka and Zhang, 2002). Since the S-RNase values vary 2-fold between different plant genotypes, and the fully incompatible S3-RNase is present at roughly 0.2 mg ml–1 in styles of the fully incompatible cultivar ‘Choiuro’ (Matsuura et al., 2001), this suggests that the threshold may lie close to 0.1 mg ml–1 (Hiratsuka and Zhang, 2002). This value is remarkably similar to the 80 ng (0.06 mg ml–1) per style estimated from the studies reported here.
Genotypic differences in expression of style and pollen components of SI
The observation that different S-RNases are expressed to different levels in styles of a given plant (Fig. 3) leaves open the important question of the expression of the same S-RNase in different plant genotypes. In general, an effect of the genotype of the donor plant on the amounts of the S-RNase it produces, i.e. the level of expression of the same S-RNase in different genetic backgrounds, has only rarely been considered. In one study, the expression of SA2-RNase constructs in different Nicotiana species was found to depend on the genetic background of the host (Murfett and McClure, 1998). In another study using Japanese pear, the same S-allele produced different amounts of S-RNase depending upon the cultivars (Zhang and Hiratsuka, 1999). In S. chacoense, a comparison of the levels of S11-RNase in genotypes L25 and 314 suggests that only slight differences may be observed (Fig. 3E). However, in the case of the S12-RNase, up to 3-fold differences were observed among the four genotypes tested (Fig. 4). For the S12-homozygote line VF60, at least part of the increase may be related to the number of S12 genes present. However, this explanation cannot account for the high levels in the S12S14 heterozygote G4 line. In this regard it is interesting to note, in various cultivars of Japanese pear, the systematic (almost exclusive) association of S-RNases (S1, S3, S5) that are both more abundant in the styles and yield strong SI phenotypes, with those (S2, S4, S6 and S7) with weaker SI phenotypes and whose abundance in the styles is reduced (Zhang and Hiratsuka, 1999; Hiratsuka and Zhang, 2002). It was concluded that the strength of the SI system in the various cultivars depends on the total S-RNase content rather than the levels of individual S-RNases (Zhang and Hiratsuka, 1999).
The genetic background of the pollen may influence the S-RNase threshold
In addition to genotype-dependent differences in expression of the stylar component to SI, these analyses also suggest that differences may be found in expression of the pollen component. For example, when pollen of different staminate genotypes was tested on the same pistillate parent, the proportion of fruits set and the calculated S12-RNase thresholds were considerably different (Table 2). Around a 4-fold difference was calculated in the S12-RNase thresholds for pollen from the three S12- homozygous individuals tested, similar to the over 3-fold differences observed for S12-RNase expression in different genotypes (Fig. 4). Although the difference in the calculated threshold of L25 pollen appears even lower, it must be noted that only half the pollen derived from the L25 pollen has the S12 haplotype. It is tempting to speculate that expression levels of the pollen component to SI may also vary according to the genetic background, and that these variations may be responsible for the differential sensitivity of the pollen to its cognate S-RNase. Alternatively, it is possible that the different pollen types tested here differ in their S12-RNase uptake efficiency. These factors can be scrutinized when the pollen component is finally identified in S. chacoense.
Consequences of an S-RNase threshold for conceptualizing SI
The determination of what appears to be a relatively sharp S-RNase threshold has some important implications that have not previously been made explicit in studies on self-incompatibility. The reason for this is that several molecular mechanisms have previously been reported to give rise to threshold phenomena, also termed ultrasensitive or switch-like. For example, ultrasensitive responses can be due to co-operative interactions between subunits, to a requirement for multiple phosphorylation events, or to covalent modifications carried out by modifying enzymes working at saturating substrate levels (zero-order kinetics) (Ferrell, 1998). The mechanism leading to the threshold phenomenon in GSI is presently unknown, but to date the only covalent modification suspected is ubiquitination, and as S-RNase levels must rise in the pollen after import from the style they will not be initially saturating. However, as it has previously been proposed that pollen S may function as a multimer (Luu et al., 2001), one attractive possibility is that co-operative interactions might also exist between S-RNase subunits. Indeed, it would be of interest to examine the kinetics of purified S-RNases in vitro for evidence of co-operative behaviour. A second intriguing possibility is that the threshold might reflect the presence of stoichiometric levels of a factor that binds with high affinity to the RNase. If true, this suggests that biochemical approaches might be successful in isolating the factor responsible.
Fruit formation requires a minimum number of fertilization events
These results suggest that more than 10 pollen tubes entering the ovary and accomplishing fertilization may be necessary to trigger fruit formation and sustain its subsequent development. This facet of the SI phenotype has been revealed by observations by fluorescence microscopy of styles from flowers that drop, after gentle shaking, 4–5 d after pollination. Since in SI studies a cross is generally considered incompatible when pollinated flowers drop, the presence of up to 10 pollen tubes entering the ovary seems surprising. It is known that a minimum number of fertilization events is required for fruit set and its subsequent growth (Gillaspy et al., 1993; Hiratsuka and Zhang, 2002), and this is consistent with the finding that when fruits contained only a few well-developed seeds, numerous aborted seeds or swollen ovules were always present. This is similar to what was observed in other species, for example in Lycopersicon (de Nettancourt et al., 1974; Gradziel and Robinson, 1989). However, it is difficult to assess if this may be more generally true, as reports of fruits with only a single seed (de Nettancourt and Ecochard, 1968; de Nettancourt et al., 1971) made no mention of either the presence or absence of aborted seeds, nor were microscopic observations performed in those studies. In addition, it is important to recognize that the entrance of pollen tubes into the ovarian region represents only an estimation of the outcome of the cross, as the number of pollen tubes actually accomplishing fertilization is the determinant factor for fruit set. It is known that environmental factors such as temperature, heat, humidity etc. can affect self-incompatibility, although it seems unlikely that these factors would selectively affect only some styles on a plant as was observed here.
The number of fertilization events required for fruit formation has a particularly interesting implication for the evolution of the SI system. It has long been recognized that mutations of S-alleles occur in nature, yet attempts to produce new S-alleles by mutagenic treatments have not been successful (de Nettancourt, 1977, 2001), and have only been observed after site-directed mutagenesis (Matton et al., 1997, 1999). These results suggest that, even if pollen mutants were generated experimentally, these mutations could not be transmitted to the progeny upon experimental selfing if a single mutant pollen grain is insufficient to allow fruit formation, unless hormone treatments were applied to the pollinated flowers in order to prevent premature abscission (de Nettancourt et al., 1971; Golz et al., 1999). By contrast, under open pollination conditions such as occur in nature, the concomitant pollination with a mutant pollen and abundant compatible pollen from neighbour plants would allow fruit set to occur normally, in turn allowing the mutant zygote to develop and begin to spread the mutant genotype within the population.
|
Acknowledgements |
|---|
We thank Dr G Laublin for in vitro culture of immature seeds, and G Teodorescu for plant care. This work was supported by grants from Natural Sciences and Engineering Research Council of Canada (MC) and Fonds de Recherche sur la Nature et les Technologies du Québec (DM, MC).
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Clark KR, Okuley JJ, Collins PD, Sims TL. (1990) Sequence variability and developmental expression of S-alleles in self-incompatible and pseudo-self-compatible petunia. The Plant Cell 2:815–826.
de Nettancourt D. (1977) Incompatibility in Angiosperms (Springer Verlag, New York).
de Nettancourt D. (2001) Incompatibility and incongruity in wild and cultivated plants 2nd edn (Springer-Verlag, Berlin).
de Nettancourt D, Devreux M, Laneri U, Cresti M, Pacini E, Sarfatti G. (1974) Genetical and ultrastructural aspects of self and cross incompatibility in interspecific hybrids between self-compatible Lycopersicon esculentum and self-incompatible L. peruvianum. Theoretical and Applied Genetics 44:278–288.
de Nettancourt D and Ecochard R. (1968) Effects of chronic irradiation upon a self-incompatible clone of. Lycopersicon peruvianum. Theoretical and Applied Genetics 38:289–293.[CrossRef]
de Nettancourt D, Ecochard R, Perquin MDG, van der Drift T, Westerhof M. (1971) The generation of new S alleles at the incompatibility locus of Lycopersicon peruvianum Mill. Theoretical and Applied Genetics 41:120–129.[Web of Science]
Ferrell JE Jr. (1998) How regulated protein translocation can produce switch-like responses. Trends in Biochemical Science 23:461–465.
Gillaspy G, Ben-David H, Gruissem W. (1993) Fruits: a developmental perspective. The Plant Cell 5:1439–1451.
Golz JF, Su V, Clarke AE, Newbigin E. (1999) A molecular description of mutations affecting the pollen component of the Nicotiana alata S locus. Genetics 152:1123–1135.
Gradziel T and Robinson R. (1989) Breakdown of self-incompatibility during pistil development in Lycopersicon peruvianum by modified bud pollination. Sexual Plant Reproduction 2:38–42.
Gray J, McClure B, Bönig I, Anderson M, Clarke A. (1991) Action of the style product of the self-incompatibility gene of Nicotiana alata (S-RNase) on in vitro-grown pollen tubes. The Plant Cell 3:271–283.
Hancock CN, Kent L, McClure B. (2005) The stylar 120 kDa glycoprotein is required for S-specific pollen rejection in Nicotiana. The Plant Journal 43:716–723.[Medline]
Harris P, Weinhandl J, Clarke A. (1989) Effect on in vitro pollen growth of an isolated style glycoprotein associated with self-incompatibility in Nicotiana alata. Plant Physiology 89:360–367.
Hiratsuka S, Nakashima M, Kamasaki K, Kubo T, Kawai Y. (1999) Comparison of an S-protein expression between self-compatible and -incompatible Japanese pear cultivars. Sexual Plant Reproduction 12:88–93.[CrossRef]
Hiratsuka S and Zhang S. (2002) Relationship between fruit set, pollen tube growth, and S-RNase concentration in the self-incompatible Japanese pear. Scientia Horticulturae 95:309–318.[CrossRef]
Hiratsuka S, Zhang S, Nakagawa E, Kawai Y. (2001) Selective inhibition of the growth of incompatible pollen tubes by S-protein in the Japanese pear. Sexual Plant Reproduction 13:209–215.[CrossRef]
Huang S, Lee H-S, Karunanandaa B, Kao T-h. (1994) Ribonuclease activity of Petunia inflata S proteins is essential for rejection of self pollen. The Plant Cell 6:1021–1028.[Abstract]
Jahnen W, Batterham MP, Clarke AE, Moritz RL, Simpson RJ. (1989a) Identification, isolation, and N-terminal sequencing of style glycoproteins associated with self-incompatibility in Nicotiana alata. The Plant Cell 1:493–499.
Jahnen W, Lush W, Clarke A. (1989b) Inhibition of in vitro pollen tube growth by isolated S-glycoproteins of Nicotiana alata. The Plant Cell 1:501–510.
Kao TH and Tsukamoto T. (2004) The molecular and genetic bases of S-RNase-based self-incompatibility. The Plant Cell 16:72–83.
Kowyama Y, Kunz C, Lewis I, Newbigin E, Clarke AE, Anderson M. (1994) Self-compatibility in a Lycopersicon peruvianum variant (LA2157) is associated with a lack of style S-RNase activity. Theoretical and Applied Genetics 88:849–864.
Lee HS, Huang S, Kao T. (1994) S proteins control rejection of incompatible pollen in Petunia inflatas. Nature 367:560–563.[CrossRef][Medline]
Luu DT, Qin X, Laublin G, Yang Q, Morse D, Cappadocia M. (2001) Rejection of S-heteroallelic pollen by a dual-specific s-RNase in Solanum chacoense predicts a multimeric SI pollen component. Genetics 159:329–335.
Luu D, Qin X, Morse D, Cappadocia M. (2000) S-RNase uptake by compatible pollen tubes in gametophytic self-incompatibility. Nature 407:649–651.[CrossRef][Medline]
Matsuura T, Unno M, Sakai H, Tsukihara T, Norioka S. (2001) Purification and crystallization of Japanese pear S-RNase associated with gametophytic self-incompatibility. Acta Crystallographica, Section D, Biological Crystallography 57:172–173.[CrossRef][Medline]
Matton DP, Luu DT, Xike Q, Laublin G, O'Brien M, Maes O, Morse D, Cappadocia M. (1999) Production of an S RNase with dual specificity suggests a novel hypothesis for the generation of new S alleles. The Plant Cell 11:2087–2097.
Matton DP, Maes O, Laublin G, Xike Q, Bertrand C, Morse D, Cappadocia M. (1997) Hypervariable domains of self-incompatibility RNases mediate allele-specific pollen recognition. The Plant Cell 9:1757–1766.[Abstract]
McClure B. (2004) S-RNase and SLF determine S-haplotype-specific pollen recofnition and rejection. The Plant Cell 16:2840–2847.
McClure BA, Haring V, Ebert PR, Anderson MA, Simpson RJ, Sakiyama F, Clarke AE. (1989) Style self-incompatibility gene products of Nicotiana alata are ribonucleases. Nature 342:955–957.[CrossRef][Medline]
McCubbin A, Chung Y, Kao T. (1997) A mutant S3 RNase of Petunia inflata lacking RNase activity has an allele-specific dominant negative effect on self-incompatibility interactions. The Plant Cell 9:85–95.[Abstract]
Murfett J, Atherton TL, Mou B, Gasser CS, McClure BA. (1994) S-RNase expressed in transgenic Nicotiana causes S-allele-specific pollen rejection. Nature 367:563–566.[CrossRef][Medline]
Murfett J and McClure BA. (1998) Expressing foreign genes in the pistil: a comparison of S-RNase constructs in different Nicotiana backgrounds. Plant Molecular Biology 37:561–569.[CrossRef][Web of Science][Medline]
O'Brien M, Kapfer C, Major G, Laurin M, Bertrand C, Kondo K, Kowyama Y, Matton DP. (2002) Molecular analysis of the stylar-expressed Solanum chacoense small asparagine-rich protein family related to the HT modifier of gametophytic self-incompatibility in Nicotiana. The Plant Journal 32:985–996.[CrossRef][Web of Science][Medline]
Qiao H, Wang F, Zhao L, Zhou J, Lai Z, Zhang Y, Robbins TP, Xue Y. (2004) The F-box protein AhSLF-S2 controls the pollen function of S-RNase-based self-incompatibility. The Plant Cell 16:2307–2322.
Qin X, Luu D, Yang Q, Maes O, Matton D, Morse D, Cappadocia M. (2001) Genotype-dependent differences in S12-RNase expression lead to sporadic self-compatibility in Solanum chacoense. Plant Molecular Biology 45:295–305.[CrossRef][Web of Science][Medline]
Qin X, Soulard J, Laublin G, Morse D, Cappadocia M. (2005) Molecular analysis of the conserved C4 region of the S11-RNase of Solanum chacoense. Planta 221:531–537.[CrossRef][Web of Science][Medline]
Royo J, Kunz C, Kowyama Y, Anderson M, Clarke AE, Newbigin E. (1994) Loss of a histidine residue at the active site of S-locus ribonuclease is associated with self-compatibility in Lycopersicon peruvianum. Proceedings of the National Academy of Sciences, USA 91:6511–6514.
Sassa H, Hirano H, Nishio T, Koba T. (1997) Style-specific self-compatible mutation caused by deletion of the S-RNase gene in Japanese pear (Pyrus serotina). The Plant Journal 12:223–227.[CrossRef]
Sijacic P, Wang X, Skirpan AL, Wang Y, Dowd PE, McCubbin AG, Huang S, Kao TH. (2004) Identification of the pollen determinant of S-RNase-mediated self-incompatibility. Nature 429:302–305.[CrossRef][Medline]
Sonneveld T, Tobutt KR, Vaughan SP, Robbins TP. (2005) Loss of pollen-S function in two self-compatible selections of Prunus avium is associated with deletion/mutation of an S haplotype-specific F-box gene. The Plant Cell 17:37–51.
Tsukamoto T, Ando T, Takahashi K, Omori T, Watanabe H, Kokubun H, Marchesi E, Kao TH. (2003) Breakdown of self-incompatibility in a natural population of Petunia axillaris caused by loss of pollen function. Plant Physiology 131:1903–1912.
Ushijima K, Yamane H, Watari A, Kakehi E, Ikeda K, Hauck NR, Iezzoni AF, Tao R. (2004) The S haplotype-specific F-box protein gene, SFB, is defective in self-compatible haplotypes of Prunus avium and P. mume. The Plant Journal 39:573–586.[CrossRef][Web of Science][Medline]
Van Sint Jan V, Laublin G, Birhman R, Cappadocia M. (1996) Genetic analysis of leaf explant regenerability in Solanum chacoense. Plant Cell Tissue Organ Culture 47:9–13.[CrossRef]
Xu BB, Grun P, Kheyr-Pour A, Kao T-h. (1990) Identification of pistil-specific proteins associated with three self-incompatibility alleles in Solanum chacoense. Sexual Plant Reproduction 3:54–60.
Zhang S and Hiratsuka S. (1999) Variations in S-protein levels in styles of Japanese pears and the expression of self-incompatibility. Journal of the Japanese Society of Horticultural Science 68:911–918.
Zhang S and Hiratsuka S. (2000) Cultivar and developmental differences in S-protein concentration and self-incompatibility in the Japanese pear. HortScience 35:917–920.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  O. Kodad, R. Socias i Company, A. Sanchez, and M. M. Oliveira The Expression of Self-compatibility in Almond May Not Only Be Due to the Presence of the Sf Allele J. Amer. Soc. Hort. Sci., March 1, 2009; 134(2): 221 - 227. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z.-H. Hua, A. Fields, and T.-h. Kao Biochemical Models for S-RNase-Based Self-Incompatibility Mol Plant, July 1, 2008; 1(4): 575 - 585. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Liu, D. Morse, and M. Cappadocia Glycosylation of S-RNases may influence pollen rejection thresholds in Solanum chacoense J. Exp. Bot., February 10, 2008; (2008) erm339v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||