When fluo-4/AM was loaded into ‘Housui’ pollen tubes at room temperature, the normally growing pollen tubes preserved the Ca²⁺ gradient at the tube apex (Figures 1A,B). It had no effect on the apical Ca²⁺ gradient by adding 1.0 μg μl^-1 non-self-S-RNase (‘Imamuraaki’) (Figures 1C,D). However, after adding 1.0 μg μl^-1 self-S-RNase (‘Housui’), it decreased [Ca²⁺]_i at the pollen tube tips and sub-tips (Figures 1E,F). These results indicate that self-S-RNase can decrease Ca²⁺ concentrations in pollen tube apices and disrupt Ca²⁺ gradients.

We determined Ca²⁺ currents in ‘Housui’ pollen tube apex protoplasts with whole-cell recordings 30 min after adding 1.0 μg μl^-1 (final concentration) ‘Imamuraaki’ non-self-S-RNase into the bath solution, and replacement with 0.4 or 1.0 μg μl^-1 ‘Housui’ self-S-RNase. Compared to controls with 10 mM Ca²⁺, non-self-S-RNase had no effect on Ca²⁺ currents. However, in the presence of ‘Housui’ self-S-RNase, the calcium ion current presented very significant decrease (Figure 2A). The current amplitudes of 1.0 μg μl^-1 non-self-S-RNase were >4-fold bigger than the amplitudes of 1.0 μg μl^-1 self-S-RNase at a reference voltage of -200 mV (Figure 2B). These results suggest that self-S-RNase was selective in its regulation of Ca²⁺ channel activity in pollen tube protoplast membranes.

Phosphatidylinositol-specific phospholipase C controls the levels of PIP₂ (i.e., substrate; phosphatidylinositol 4,5-bisphosphate) and IP₃ (i.e., cleavage product), both of which are strong tip growth regulator candidates (Helling et al., 2006). We recorded the Ca²⁺ currents in pollen tube apices with our established method (Qu et al., 2007). When U-73122, an inhibitor of PLC activity (Mogami et al., 1997; Evdonin et al., 2004), was added to protoplasts through a pipette at a final concentration of 10 μM, the Ca²⁺ currents significantly decreased (p < 0.01). The current amplitudes, which were assessed at a standard reference voltage of -200 mV, were approximately sixfold smaller than those of the controls (Figure 3A). After adding U-73343-a biologically inactive analog of U-73122, it did not show significant difference from the control (p > 0.01) (Figure 3A). We observed that compared to controls, IP₃ increased Ca²⁺ currents when the final concentration was 10 or 100 pM in the pipette solution, with a greater effect at 100 pM (Figure 3B). Moreover, U-73122, with a final concentration of 10 μM in culture medium, also disrupted tip-focused Ca²⁺ gradients of pollen tubes, as revealed by fluo-4/AM staining (Figures 3C,D). These results suggest that hyperpolarization-activated Ca²⁺ channels at the tips of pollen tubes were regulated by IP₃, and that PLC regulated the tip-focused Ca²⁺ gradient of pollen tubes.

We collected plasma membrane fractions (containing cell-membrane Phospholipase C) from ‘Housui’ pollen tubes (Figure 4A), and analyzed the influence of ‘Housui’ self-S-RNase and ‘Imamuraaki’ non-self-S-RNase on its activity. PLC activity was monitored in vitro by incubating plasma membrane fractions with ³H-PIP₂ and measuring the labeled product in the soluble fraction (presumably IP₃) (Figure 4B). It remarkably decreased PLC activity by adding 10 μM (final concentration) U-73122. However, 1 μg μl^-1 (final concentration) self-S-RNase also severely inhibited PLC activity, while non-self-S-RNase of the same final concentration only had minimal effect on the PLC activity (Figure 4B). These results suggest that self-S-RNase inhibited PLC activity in self-pollen tubes.

When the basal pipette solution contained 100 pM IP₃, Ca²⁺ currents in ‘Housui’ protoplasts (whole-cell configuration) were fully recovered from the suppression caused by the presence of 1 μg μl^-1 ‘Housui’ self-S-RNase in the bath solution. Furthermore, compared to the control, no significant differences were found in current when the pipette solution contained a mixture of 10 μM U-73122 and 100 pM IP₃. These results indicate that IP₃ enabled the recovery of Ca²⁺ currents previously suppressed by U-73122 (Figures 4C,D). The data suggest that self-S-RNase suppressed the activity of Ca²⁺ channels at pollen tube tips by inhibiting PLC activity and decreasing IP₃ levels in protoplasts.

‘Dangshansuli’ (Pyrus x bretschneideri Rehd.) is the most important commercial Asian pear cultivar, with a production volume of 4 million tons per year. The draft genome of the ‘Dangshansuli’ was the first comprehensive fine pear genome (Wu et al., 2012b). Its S-genotype is S₇S₃₄. To further identify possible PLC and S-RNase interactions, we performed yeast two-hybrid experiments. The PLC coding region was subcloned into the pGBKT7 vector, and it was tested against S₇-RNase and S₃₄-RNase pGADT7 fusion constructs. Full-length PLC interacted with full-length S₇-RNase, but not with S₃₄-RNase (Figure 5 and Supplementary Figure S1). To further gain insights into the relationship between PLC and S-RNase, we transiently expressed GFP-PLC and RFP-S-RNase in tobacco leaves. Based on the fluorescence microscopy, we found that when fused to GFP, the same PLC fragment used in the yeast-two-hybrid experiment (without the membrane-associated C2 region) mainly localized near the nucleus, and the signal overlapped with the DsRed fluorescence emitted by the RFP-S7-RNase (Figure 6A). However, it did not result in fluorescence superposition by expressing RFP-S34-RNase and GFP-PLC in tobacco cells (Figure 6B). These results suggest that PLC interacted with the products of multiple alleles of one S-RNase, rather than the products of two S-RNases.

Self-specific direct interactions between PLC and self-S-RNase require PLC polymorphisms at a level matching the S-RNase polymorphisms. To clone the PLC cDNA from Pyrus pyrifolia pollen tubes, one degenerate oligonucleotide, designed according to a conserved amino acid sequence, was used for PCR analyses with cDNA from ‘Housui’ pollen tubes as templates. Two DNA fragments (approximately 750 bp) were amplified. A BLAST analysis indicated that these two fragments were homologous to plant PI-PLCs. Sequence alignments of the two DNA fragments revealed the encoded amino acid sequences (according to Open Reading Frame Finder¹) with a polymorphism frequency of 22% (Figure 7A). The two partial PLC genes were determined to be PI-PLC 2-like (LOC103936790) according to Protein Blast (P. x bretschneideri), with identities of 96 and 83%.

For fruit trees, including pear trees, it is very difficult to construct a stable transgenic system. However, we obtained ‘Jinzhuili’ as a naturally occurring self-compatible mutant of ‘Yali.’ Field pollination tests showed that >76% of flowers set fruits, and many pollen tubes grew down to the base of the style during ‘Jinzhuili’ self-pollination and cross-pollination of ‘Yali’ (female) × ‘Jinzhuili’ (male). In contrast, during ‘Yali’ self-pollination and cross-pollination of ‘Jinzhuili’ (female) × ‘Yali’ (male), fruit set did not occur, and most pollen tubes were arrested at the upper part of the style. The results indicate that ‘Yali’ and ‘Jinzhuili’ had a normal SI response in the style, which rejected competent ‘Yali’ pollen, but accepted ‘Jinzhuili’ pollen. Further molecular analyses revealed that both ‘Yali’ – and ‘Jinzhuili’ – labeled S₂₁-RNase and S₃₄-RNase genes were transcribed normally with identical mRNA sequences. Therefore, we suggest that it was most likely ‘Jinzhuili’ had lost pollen SI activity, resulting in a style that behaved normally during the SI response, similar to that of the wild-type ‘Yali’ (Supplementary Figure S2).

We used a previously described method to clone PLC genes from ‘Jinzhuili’ and ‘Yali.’ The amino acid sequence alignments (Figure 7B) showed that the PLC of ‘Jinzhuili’ had a 26-amino acid insertion. We inserted this nucleotide fragment to encode the 26-amino acid peptide segment into the PLC gene of ‘Dangshansuli’ (initially used in the yeast-two-hybrid experiment as shown in Figure 5). This mutated PLC protein did not interact with either the S7-RNAse or the S34-RNAse in our yeast-two-hybrid experiment (Figure 8 and Supplementary Figure S1). The self-S-RNase could not identify mutant PLC, and was unable to inhibit PLC activity. Therefore, ‘Jinzhuili’ exhibited self-compatibility. These results suggest that the PLC in the pollen tube is one of the male determinants in the self-incompatibility reaction.

We explored the difference in gene expression at the early stage after pollination, and compared gene expression levels in the styles of ‘Yali’ and ‘Jinzhuili’ 0.5 h after self-pollination. Although the differences in gene expression were compared after pollination, the GO terms associated with “pollen tube growth,” “cell tip growth,” “pollen tube development,” and “pollination” identified cell activities consistent with pollen growth activity (data unpublished). We searched for Pyrus PLC genes in the GenBank database (Supplementary Table S1). The differences in gene expression levels were determined based on high-throughput sequencing results (Supplementary Table S1). Importantly, 0.5 h after pollination, the pollen grain had just germinated on the stigma, and there was no difference in pollen tube growth between plants that underwent compatible and incompatible pollinations (Supplementary Figure S3). Therefore, at this time point, differences in gene expression between stigmas derived from compatible and incompatible pollinations likely reflect differences related to the compatibility of the interaction rather than differences in pollen tube size or viability.

In theory, if the PLC gene is expressed specifically in pollen tubes, its expression during compatible pollination should be higher than that of the control or during incompatible pollination. This is because pollen tubes can grow during compatible pollination. The values of Log₂(Jinzhuili _{0.5 h}/ Jinzhuili _CK) and Log₂(Jinzhuili _{0.5 h}/ Yali _{0.5 h}) were positive for only LOC103958605, LOC103936790, and LOC103966250, although the p-value for LOC103966250 was higher than 0.05 (i.e., 0.252186). Therefore, we selected these three genes for quantitative PCR analysis, and detected the expression levels at 0.5 h after pollination. The expression levels in the pollen tube after compatible pollination were significantly higher (p < 0.01) than those after incompatible pollination and the expression level of non-pollinated styles. During compatible pollination, the expression levels were 7, 10, and 14 times higher than those during incompatible pollination (Figures 9A–C). Moreover, based on Supplementary Table S1, we selected LOC103932386, which exhibited no differential expression, for further RT-PCR analysis. We confirmed that the expression level did not change following pollination (Figure 9D). These results suggest that the three genes of LOC103958605, LOC103936790, and LOC103966250, were expressed specifically in the pollen tube, and they were related to pollen tube growth in the style. The expression levels of these three genes in the styles were higher (p < 0.01) after incompatible pollination than in non-pollinated styles.

By using the iTRAQ technique, we measured the style proteome of the ‘Jinzhuili’ and ‘Yali’ after 0.5 h of self-pollination respectively. The expression of PI-PLC related proteins was different under ‘Jinzhuili’ (0.5 h) – vs. – ‘Yali’ (0.5 h), ‘Yali’ (0.5 h) – vs. – ‘Jinzhuili’ (CK), ‘Jinzhuili’ (0.5 h) – vs. – ‘Jinzhuili’ (CK) (Table 1). In the case of self-compatibility comparing to self-incompatibility, PI-PLC-related protein up-regulated expression was significant (Protein ID: XP_009373787.1, XP_009343292.1, XP_009344939.1). The proteins, XP_009343292.1 and XP_009344939.1, were up-regulated only in the case of self-compatibility. In particular, the changes in XP-009344939.1 protein expression were consistent with the results of high-throughput sequencing and RT-PCR. Although expression of the XP-009346269.1 protein was down-regulated in styles derived from compatible pollination, and up-regulated in styles derived from incompatible pollination, the differences were not significant. These results indicate that by using the ITRAQ technology to analyze PLC protein expression differences, it could be useful to distinguish between compatibility and incompatibility.

S-RNase inhibited PLC activity and arrested pollen tube growth. Furthermore, we cloned a gene fragment from PI-PLC 2-like gene (LOC103936790), and observed that the expression levels of the other two PLC genes (LOC103958605 and LOC103966250) were relatively higher after compatible pollination. According to proteomic analysis, there are other PLC genes involved in self-incompatibility. Therefore, it requires further research on the role of these PLC genes.

We used two pear varieties of ‘Housui’ and ‘Imamuraaki’ to investigate GSI in P. pyrifolia. The S-genotypes of ‘Housui’ and ‘Imamuraaki’ are S₃S₅ and S₁S₆, respectively (Ishimizu et al., 1998). The two varieties are completely self-incompatible after self-pollination, but show strong compatibility when cross-pollinated.

We collected ‘Housui’ (P. pyrifolia Nakai) pollen from blooming flowers in a fruit experimental garden in Jiangsu, China. The collected pollen was stored at -20°C. S-RNase was extracted from ‘Housui’ and ‘Imamuraaki’ styles according to previously described methods (Hiratsuka et al., 2001). The final S-RNase concentration was 1.0 μg μl^-1 with 0.15 units of activity in the medium. The medium composed of 0.55 mM Ca(NO₃)₂, 1.6 mM H₃BO₃, 1.6 mM MgSO₄, 1 mM KNO₃, 440 mM sucrose and 5 mM 5, 2-(N-Morpholino) ethane sulfonic acid hydrate (MES), and pH was adjusted to pH 6.0 with Tris. Protoplasts were isolated from pollen tube apices according to our established method (Qu et al., 2007).

Pollen tubes were cultured for 2.0 h, and then stained with 1 μM fluo-4/AM dye (Dôjindo Laboratories) for 15 min at room temperature in the dark according to our methods (Qu et al., 2016a). We used a TCS SP-2 MP AOBS Confocal System (Leica) to observe the tubes. The fluorescence intensity was measured with Image Pro Plus software (Qu et al., 2016a).

Pipettes were pulled from borosilicate glass blanks and coated with Sylgard (184 Silicone Elastomer kit; Dow Corning). Pipettes with solution had resistance values ranging from 15 to 35 MΩ in 10 mM CaCl₂. Whole-cell currents across the plasma membrane of protoplasts isolated from pollen tubes were measured with an Axon 200B amplifier (Axon Instruments). Whole-cell preparations were obtained by forming a giga seal in the cell-attached mode, and the membrane was ruptured with a short burst of extra suction. A substantial increase in capacitance indicated that the whole-cell configuration was achieved, and the series resistance and capacitance were adjusted accordingly. Voltage-pulses of 2.5 s were used to elicit voltage-dependent currents. Data were sampled at 2 kHz and filtered at 0.5 kHz. Records were stored and analyzed by using pClamp 9.0 (Axon Instruments). The junction potential was corrected according to a published report (Amtmann and Sanders, 1997). All experiments were conducted at room temperature (20–22°C). The current-voltage curves were constructed by using the current values measured at the final voltage.

The basal solution consisted of 10 mM CaCl₂, 0.2 mM glucose, and 0.05 mM MES. The bath solutions were adjusted to 800 mOsM and pH 5.8 with mannitol and Tris. The internal basal pipette solution contained 1 mM MgCl₂, 0.1 mM CaCl₂, 4 mM Ca(OH)₂, 10 mM EGTA, 2 mM MgATP, 10 mM HEPES, 100 mM CsCl, and 0.1 mM GTP, with pH 7.3. The solution was adjusted to 1,100 mOsM with mannitol and Tris. The free Ca²⁺ concentration of the pipette solution was approximately 10 nM. Details regarding the bath and pipette solutions are provided in the figure legends.

Mature pollen grains (0.1 g, approximately 1 × 10⁷ cells ml^-1) were suspended in 10 ml medium and cultured for 3 h at 25°C. Subsequently, pollen tubes were washed with 0.8 M sorbitol and quickly frozen with liquid nitrogen. The tubes were then homogenized in 20 mM Hepes-Tris (pH 7.8), 1 mM EGTA (pH 8.0), 0.25 mM sorbitol, 0.5% (w/v) BSA, 1 mM PMSF, 5 mM DTT, and 0.7% PVPP for membrane isolation. Tissue debris was removed by centrifugation at 10,000 × g for 20 min at 4°C. The supernatant was further centrifuged at 100,000 × g for 60 min at 4°C for microsomal membrane preparation. Highly purified plasma membranes were prepared from the pellet through two-phase partitioning (Tate and Gruner, 1989). All steps were performed at 4°C. Protein concentrations were measured by using a bicinchoninic acid protein assay reagent with BSA as a standard (Smith et al., 1985). The cell membrane extracted from the experimental procedure can be used to evaluate the activity of phospholipase C (Liu et al., 2006).

Phospholipase C activity was assayed according to a published method (Helling et al., 2006). The standard reaction mixture contained 50 mM Tris-maleate (pH 6.0), 0.8 mM Ca²⁺/EGTA, 0.08% (w/v) sodium deoxycholate, 10 μl self-S-RNase (‘Housui’) or non-self-S-RNase (‘Imamuraaki’), 0.74 kBq [³H]PIP₂ (251.6 GBq mmol^-1, head group labeled; Perkin-Elmer), and 10 μg membrane protein in a final volume of 50 μl. The reaction was initiated by adding micellar substrate solutions, and stopped after 20 min at 25°C by adding 1 ml chloroform: methanol (2: 1, v/v) and 250 μl 1 M HC1. After vortexing and a brief centrifugation, the radioactive reaction products were recovered in the upper phase, and quantified with a liquid scintillation counter (LS 6500 Multi-Purpose Scintillation Counter; Beckman Coulter). Background radioactivity detected in control samples incubated without protein was subtracted from all data.

The diploid P. x bretschneideri Rehd. is the first pear species with its genome comprehensively sequenced (Wu et al., 2012b). We compared the sequence of our cloned PLC cDNA fragments isolated from pollen tubes with that of the ‘DangshanSuli’ (P. x bretshneideri Rehd.) genome to determine the full-length PLC sequence. The full-length cDNA sequences of S₇-RNase and S₃₄-RNase (DQ414813) were based on the P. x bretschneideri Rehd. genome and GenBank sequences, respectively. The S₇-RNase and S₃₄-RNase cDNA sequences were separately cloned into the pGADT7 vector (Clontech). The PLC cDNA sequence was cloned into the pGBKT7 vector (Clontech). Comparing with the PLC gene mutation of the ‘Jinzhuili,’ we inserted 78 nucleotide residues, which encoded 26 amino acid residues, into the same mutant position in the ‘Dangshansuli’ PLC gene. The mutation of the ‘Dangshansuli’ PLC gene was also cloned into the pGBKT7 vector (Clontech). Each bait/prey pair was introduced to the AH109 yeast strain (Clontech). As a control for auto-activation false-positives, each bait was also co-transformed into the yeast strain with an empty AD vector, and each prey was co-transformed with an empty BD vector. The bait/prey pair colonies that grew on all selective media (Trp-Leu and Trp-Leu-Ade-His) were considered positive for interaction.

For Tobacco leaves expressing PLC-GFP and S-RNase-RFP, fluorescence localization and image processing were performed as described in Francinallami et al. (2011) and Zhong et al. (2012).

Total RNA was extracted from ‘Housui’ and ‘Imamuraaki’ pollen grains as previously described (Tao et al., 1999), and treated with DNase I (Invitrogen). The integrity of the RNA was checked by electrophoresis, and concentrations were determined by spectrophotometry. Total RNA (1 μg) was used for first-strand cDNA synthesis with the Micro-FastTrack 2.0 mRNA Isolation Kit (Invitrogen) by following the manufacturer’s instructions. The cDNA then served as templates for subsequent polymerase chain reaction (PCR) amplifications.

We reverse-transcribed mRNA to synthesize first-stand cDNA by using the adapter primer NotI-(dT)18. The 3′-region of cDNA encoding PI-PLC was amplified with NotI-(dT)18 and a forward primer (5′-CAGCTKAGYAGYGAYTGYAGTG-3′) designed based on the X-domain region. The 5′-ends of the cDNA were amplified by using the 5′-RACE System 2.0 (Invitrogen) with reverse primers (5′-TGAACATWCCDTGCATRAACCAAAG-3′ and 5′-TNAGDATYTTCCGCTGAGTYAACCTG-3′) specific for PI-PLC. The partial-length cDNA encoding PI-PLC was obtained through 3′ and 5′ RACE cloning.

‘Jinzhuili’ is a spontaneous self-compatible mutant of ‘Yali’ (P. x bretschneideri Rehd., S₂₁S₃₄). ‘Yali’ is a major Chinese pear cultivar, which displays a typical S-RNase-based GSI (Wu et al., 2013). We collected the styles from 100 flowers of each cultivar 0.5 h after self-pollination. We used non-pollinated ‘Jinzhuili’ styles as controls. Total RNA was extracted from each sample with Takara reagent (Takara Bio) by following the manufacturer’s protocol. RNA quantity, purity, and quality were analyzed with a spectrophotometer (SMA4000 UV-VIS spectrophotometer; Varian, Inc.). Double-stranded cDNA libraries were then constructed by using the TruSeq RNA Sample Preparation kit v2 (Illumina). Each sample pool was sequenced on the MiSeq (Illumina) platform with sequence runs of 2 × 250 paired-end reads at The Central Facility of the Institute for Molecular Biology and Biotechnology, McMaster University, Canada.

The P. pyrifolia reference genome and annotation data were obtained from GenBank². The transcriptome genome coverage was deduced based on our transcriptome data (4.0 Gb) and the P. x bretschneideri Rehd. genome data (432 Mb) from the Centre of Pear Engineering Technology Research, State Key Laboratory of Crop Genetics and Germplasm Enhancement (Wu et al., 2012b).

In our project, we sequenced three samples on the Illumina HiSeq platform. On average, we generated about 4.46 Gb bases from each sample. After mapping sequenced reads to the reference genome and reconstructing transcripts, we obtained 7,251 novel transcripts from all samples. Among them, 4,997 were previously unknown splicing variants from known genes, 999 were novel coding transcripts without any known features, and the remaining 1,255 were long non-coding RNA. The sequencing reads containing low quality, adaptor polluted and high content of unknown bases (N) were removed before downstream analyses. After filtering, the read-quality metrics were generated (Supplementary Table S3). The filtered reads were mapped to the reference genome by using HISAT (Kim et al., 2015). On average, 59.79% of the reads were mapped, and the uniformity of the mapping results between samples suggested that the results of these analyses were comparable. The mapping details are summarized in Supplementary Table S4.

Gene ontology (GO) analysis was performed to determine the biological implications of the expression of unique genes in significant or representative profiles of genes that were differentially expressed. We downloaded GO annotations from NCBI³, UniProt⁴, and the Gene Ontology Consortium⁵. Fisher’s exact test was conducted to identify significant GO categories, and false discovery rate (FDR) was used to correct the p-values (Dupuy et al., 2007).

Pathway analysis was used to determine the significant pathway(s) of the DEGs based on the KEGG database. Fisher’s exact test was conducted to select significant pathways, and the threshold of significance was defined according to the p-value and FDR (Kanehisa et al., 2004; Yi et al., 2006; Draghici et al., 2007).

To estimate mRNA expression levels of PLC genes in style after compatible and incompatible pollination, qRT-PCR analysis was performed. After 0.5 h of artificial pollination, we collected the styles and immediately placed them into liquid nitrogen for preservation. Total RNA was extracted with TRIzol reagent (Invitrogen), and the first-strand cDNA was synthesized from 2 μg of total RNA with an Omniscript RT kit (Qiagen). The primers used for RT-PCR are shown in Supplementary Table S2. Actin gene was used as the reference gene. All quantitative RT–PCR experiments were performed in biological triplicate and technical duplicate. Relative quantification was normalized to the housekeeping control gene ubiquitin, and FC (fold-change) in gene expression was calculated with the 2^-ΔΔCT method (Livak and Schmittgen, 2001).

The three samples used in high-throughput sequencing were used to perform the proteome analysis. The total protein of each corresponding group was blocked, digested and labeled according to the iTRAQ protocol (BGI, Co., Ltd, China). The labeling was as follows: ‘Yali’ 0.5 h (113), ‘Jinzhuili’ Control (114) and ‘Jinzhuili’ 0.5 h (115). All LC-ESI-MSMS analysis was based on the Triple TOF 5600 system (SCIEX, Framingham, MA, United States) fitted with a Nanospray III source (SCIEX, Framingham, MA, United States), and a pulled quartz tip as the emitter (New Objectives, Woburn, MA, United States). The IQuant software was used to conduct protein identification, tag impurity correction, data normalization, missing value imputation, protein ratio calculation, statistical analysis, and results presentation for iTRAQ (Wen et al., 2014). The three samples went through a batch of machine tests. No difference in the number of proteins was identified between samples, and no abnormal sample was found. The labeling efficiency and label quality after mass spectrometry were subject to quality-control tests. The ratios of the medians of tag intensities were close to 1. The iTRAQ labeling efficiency was 96.44%. Protein identification was based on unique peptide segments with its threshold set at a minimum of one unique peptide segment. The peptide segments retained for quantification were the segments that uniquely identified individual proteins. The search was performed against P. x bretschneideri Rhd. genome data.