Sequence Characterization and Spatiotemporal Expression Patterns of PbS 26 -RNase Gene in Chinese White Pear (Pyrus bretschneideri)

Many flowering plants exhibit an important intraspecific reproductive barrier phenomenon, that is, self-incompatibility (SI), in which S-RNase genes play a significant role. To clarify the specific function of S-RNase genes in Chinese pears, the full length cDNA of PbS 26 -RNase was isolated by rapid amplification of cDNA ends (RACE) technology from Chinese white pear (Pyrus bretschneideri) cultivar “Hongpisu.” The cDNA sequence for PbS 26 -RNase was deposited in GenBank under accession number EU081888. At the amino acid level, the PbS 26 -RNase displayed the highest similarity (96.9%) with PcSa-RNase of P. communis, and only seven amino acid differences were present in the two S-RNases. Phylogenetic analysis of rosaceous S-RNases indicated that the PbS 26 -RNase clustered with maloideous S-RNases, forming a subfamily-specific not a species-specific group. The PbS 26 -RNase gene was specifically expressed in the style but not other tissues/organs. The expression level of the PbS 26 -RNase gene rapidly increased at bell balloon stage (BBS), and then it dropped after pollination. However, the abundance of the PbS 26 -RNase gene transcript in the style was greater after cross-pollination than after self-pollination. In addition, a method for rapidly detecting the PbS 26 -RNase gene was developed via allele-specific primers design. The present study could provide a scientific basis for fully clarifying the mechanism of pear SI at the molecular level.


Introduction
In flowering plants (also called angiosperms), an intraspecific reproductive mechanism, self-incompatibility (SI), has so far been widely known. This mechanism can prevent inbreeding and promote outcrossing [1]. One type of the SI, gametophytic SI (GSI), mainly exists in three plant families including Solanaceae, Scrophulariaceae, and Rosaceae, and it is controlled by a multiallelic S-locus [1][2][3]. The S-locus harbors a series of S-alleles that are divided into two types, that is, female S-genes and male S-genes. The expression products of the S-allele in the female organ (style) are S-RNases, a class of polymorphic proteins with ribonuclease activity.
Many fruit trees of Rosaceae, including apple, pear, and almond, belong to GSI-type plants. Among rosaceous members, Japanese pear (Pyrus pyrifolia) is the first species in which pistil S-allele specific proteins were successfully identified as S-RNases [3]. Afterwards, cDNAs encoding seven S-RNases (S 1 to S 7 ) of Japanese pear were cloned with their primary structure being determined [4,5]. A hypervariable (HV) region, possibly associated with S-allele specific recognition, was identified [5]. Based on sequence characterization of Japanese pear S-RNase genes, a PCRrestriction fragment length polymorphism (PCR-RFLP) system was established for identifying S-genotypes of cultivars and searching new S-alleles [5]. Using this system, four additional alleles S 8 , S 9 , S 10 , and S were discovered and cloned, and accordingly the system was modified [6][7][8].
Pears native to China include four main cultivated species, that is, Chinese sand pear (P. pyrifolia), Chinese Signal peptide T P C K D P P D K L F T V H  white pear (P. bretschneideri), Ussurian pear (P. ussuriensis), and Xinjiang pear (P. sinkiangensis). Chinese pears were considered to be the major origin of Asian pears [9]. Knowledge of S-genotype is essential for pear production and breeding programs. The conventional pollination-based method for cultivars S-genotyping is time consuming and labor intensive. Moreover, the obtained results are often ambiguous due to environmental and physiological factors. PCR-based molecular techniques, such as PCR-RFLP and direct sequencing of amplified fragments, provide a rapid, reliable, and easily manipulated approach for pear cultivars S-genotyping. To date, a series of S-alleles have been found in Chinese pears and approximately 100 cultivars have been Sgenotyped [10][11][12][13][14][15]. Of these S-alleles, only three full length cDNAs encoding S 13 -, S 16 -, and S 34 -RNases were reported and analyzed for their possible function in SI reaction [16,17]. More recently, several pollen specific polymorphic F-box genes were isolated and considered to be the candidates of male S-genes in pears [18][19][20][21][22][23][24]. To fully clarify the mechanism of pear SI, it is necessary to investigate the interaction of the S-RNase genes with the male S-genes. In this study we cloned and characterized the full length of cDNA of the PbS 26 -RNase gene in P. bretschneideri. We studied the spatiotemporal expression patterns of this gene and compared its transcript abundance between incompatible pollination and compatible pollination. In addition, we developed a method for rapidly and accurately detecting the PbS 26 -RNase gene.

Isolation of Nucleic Acids.
Genomic DNA was isolated from young leaves with a CTAB method as described in our previous study [22]. Total RNA was extracted using Micro-to-Midi Total RNA Purification System (Invitrogen) and following the manufacturer's protocols. The quality of DNA and RNA was checked by electrophoresis in 1.0% TAE agarose gel.

Controlled Pollinations.
In order to elucidate the expression profiles after pollination of the PbS 26 -RNase gene, selfpollination ("Hongpisu" × "Hongpisu") and cross-pollination ("Hongpisu" × "Jinhua") tests were performed using 30 flowers with three replicates for each pollination combination. Anthers were collected from flowers at the BBS and incubated for 30 h to collect the pollens. For each pollination combination, flowers were emasculated, pollinated, and covered with paper bags to avoid contamination. In order to monitor pollen tubes growth in the style, samples were taken at 2, 8, 12, and 24 h after pollination and fixed in FAA solution.
After fixation, the flowers were maintained in 70% ethanol and stored at 4 ∘ C until the time of pollen tubes growth was observed using a fluorescence microscope. According to the pollen tube growth, styles at 12 h and 24 h after pollination were collected for quantitative RT-PCR (qPCR) analysis. In addition, the percentages of fruit set and the number of full seeds were recorded four months after pollination. The fruit set percentage of below 30% was considered as incompatible [7]. 2.6. Quantitative RT-PCR. Quantitative RT-PCR (qPCR) was carried out with three replicates in each reaction using the BIO-RAD CFX system (Bio-Rad). PbS 26 -RNase gene-specific primers, reference gene, and its corresponding primers were the same as described in semi-qPCR. PCR was performed in a 20 L volume containing 2 L diluted cDNA, 250 nM each primer, and 1 × SYBR Premix Ex Taq II (TaKaRa) using the same conditions with semiquantitative RT-PCR. The results from PbS 26 -RNase gene-specific amplification were analyzed using the comparative Cq method which uses an arithmetic formula, 2 −ΔΔ , to obtain results for relative quantification [26]. Sequence data were analyzed with DNAMAN (version 6.0.40). Prediction of isoelectric point (pI), molecular weight (Mw), and Prosite was performed by using online ExPASy proteomics tools (http://expasy.org/tools/pi tool.html). The BLAST algorithm was used to search the NCBI GenBank (http://www.ncbi.nlm.nih.gov/) databases for homologous sequences. Motif analysis of the target gene was conducted by Motif Scan (http://myhits.isb-sib.ch/cgi-bin/motif scan) and sequence comparisons. Multiple amino acid sequences The Scientific World Journal  alignment was performed using the DNAMAN software. Based on this alignment, a phylogenetic tree was generated. 26 -RNase. Previously, the PbS 26 -RNase allele was identified in P. bretschneideri cultivar "Hongpisu" (S 12 S 26 ) [13]. Its partial genomic sequence was isolated, but the functional completed cDNA has not been gained. The most common way to isolate full length cDNA includes the processes of 5 RACE and 3 RACE. The process of 5 RACE is a more complicated and challenging procedure than 3 RACE, with many more steps. The PF2 primer anneals to a conserved region in the 5 untranslated region (UTR) of Japanese pear S-RNases [6]; accordingly the produced target gene contains initiation codon ATG in theory. Following RACE-PCR with primer PF2 and AUAP in this study, an approximately 900 bp product was yielded. Since the PF2 does not match the S 12 -RNase [27], the 900 bp product was considered to correspond to the PbS 26 -RNase gene. The target clone was sequenced and confirmed as a full length cDNA encoding S 26 -RNase by detailed sequence analysis. The cDNA sequence for PbS 26 -RNase was 906 bp long with a complete open reading frame (ORF) of 684 bp encoding 228 amino acids (GenBank accession number EU081888). The putative initiation codon ATG was located at poisons 29-31. The termination codon TAA was present at positions 713-715, followed by 3 UTR. The deduced amino acid sequences for the PbS 26 -RNase had similar structural characteristics of pear S-RNases, for example, a signal peptide, five conserved regions, and a HV region. The molecular weight (Mw) and the isoelectric point (pI) of the PbS 26 -RNase were predicted to be 26.2 kDa and 9.14, respectively, which agreed with the basic properties of pear S-RNases [7]. The PbS 26 -RNase showed the conserved motifs with two histidine residues, His-60 and His-116, which are essential for T2/S type RNase activity [28]. It also harbored eight cysteine residues (Cys-42, Cys-49, Cys-75, Cys-119, Cys-183, Cys-198,Cys-209, and Cys-221) that were mostly conserved in S-RNases. These eight cysteine residues could form four disulfide bridges, which plays a significant role in the formation or stabilization of their tertiary structure [7]. The PbS 26 -RNase also presented six potential N-glycosylation sites, of which only one, that is, Asn-, has glycans which were considered to be important for the folding and the stabilization of the core structure ( Figure 1) [7].

Amino Acid Sequences Comparison and Phylogenetic
Analysis. The newly cloned PbS 26 -RNase gene allows us to better understand the evolutionary relationship of rosaceous S-RNases. The multiple amino acid sequences alignment was made using clustal X program with DNAMAN. A total of 39 complete amino acid sequences of S-RNases of eight species, including sweet cherry (P. avium), apricot (P. armeniaca), almond (P. duslics), plum (P. salicina), Japanese pear (P. pyrifolia), Chinese white pear (P. bretschneideri), European pear (P. communis), and apple (M. domestica), were used in the alignment. Among the P. bretschneideri species, the PbS 26 -RNase showed the lowest similarity (64.0%) with the PbS 13 -RNase and the highest (90.4%) with the PbS 16 -RNase. However, the PbS 26 -RNase showed extremely high similarities to P. communis PcSa-RNase (96.9%), which were remarkably higher than to P. bretschneideri S-RNases. Amino acid comparison between the PbS 26 -RNase and the PcSa-RNase showed that only seven amino acids differences were presented in the two S-RNases ( Figure 2). The existence of such two S-alleles, in one case, agreed with the hypothesis that new S-alleles are generated by an accumulation of point mutations of earlier-formed alleles, which may or may not lead to differences in amino acid sequence [29]. In another case, the two genes also may generate from the same ancestor after separate mutations.
Based on the multiple sequence alignment, a phylogenetic tree was generated ( Figure 3). As indicated by the tree, the 39 S-RNases were divided into two subfamily-specific groups, that is, amygdaloideous S-RNases and maloideousS-RNases. The amygdaloideous group includes S-RNases of pears (P. pyrifolia, P. bretschneideri, and P. communis) and apple (M. domestica). The maloideous group includes those of almond (P. duslics), apricot (P. armeniaca), sweet cherry (P. avium), and plum (P. salicina). In the maloideous group, pear and apple S-RNases were closely related that they did not form species-specific subgroup. In the case of amygdaloideous group, the almond, apricot, plum, and sweet cherry S-RNases were grouped with one another and did not form speciesspecific subgroup either. The phylogenetic analysis of the study agreed with the proposal of Ushijima et al. [30]; namely, the divergence of S-RNase alleles predated speciation of the Rosaceae family, but it occurred shortly after the divergence into subfamilies.

The PbS 26 -RNase Gene Was Specifically Expressed in the Style.
To investigate tissues/organs expression of PbS 26 -RNase gene, semi-qPCR was performed on total RNA of styles prior to pollination, stamens, roots, stems, and leaves of "Hongpisu" pear. Following RT-PCR with PbS 26 -RNase genespecific primers, one fragment of expected sized (130 bp) was produced from styles, but not from the other tissues/organs ( Figure 4). The results suggested that the PbS 26 -RNase gene was specific in pears. The expression patterns coincided with those of other S-RNase genes in Rosaceae plants such as apple [31], Japanese pear [7], sweet cherry [32], and Japanese apricot [33].

Comparison of Expression Patterns of the PbS 26 -RNase Gene between Self-Pollination and Cross-Pollination.
Although many data show that the S-RNase genes are specific in the style, little is known about the expression profiles of the pear S-alleles during the style development, including before and after pollination. In this study, qPCR indicated that the PbS 26 -RNase gene showed an increased expression pattern in the developing style before pollination in "Hongpisu. " A slight transcript level of the Pb 26 -RNase allele was detected at 96 h BA, and then it rapidly increased at 48 h BA and at BBS (Figure 7).
To further investigate the expression patterns after compatible pollination and incompatible pollination and compare their difference, controlled pollinations, that is, "Hongpisu" × "Hongpisu" and "Hongpisu" × "Jinhua, " were conducted. In self-pollination, pollen tubes were inhibited and twisted 12 h after pollination ( Figure 5, a4). The selfpollination led to the low percentages of fruit set (5.8%) and seed set (on the average 3.7 seeds per fruit) (Figure 6), showing incompatible behaviors. In cross-pollination, in contrast, pollen tubes could grow to the cut end of the style, that is, the stylar base 24 h after pollination ( Figure 5, b4). The cross-pollination resulted in remarkably higher percentages of fruit set (84.6%) and seed set (on the average 8.7 seeds per fruit) (Figure 6), showing compatible behaviors. After pollination, the expression level of PbS 26 -RNase allele gradually decreased in the style of both different pollinations ( Figure 7). However, the expression level of the PbS 26 -RNase gene was higher after compatible pollination than after incompatible pollination, with the abundance of transcript being around 9-fold and 7-fold higher at 12 h AP and 24 AP, respectively, in cross-pollination than in self-pollination. This is the first report of the expression patterns of pear S-RNase genes in the developing style after pollination. The difference of S-RNase gene expression level between selfpollination and cross-pollination may be associated with the behaviors of different pollen S-genes. The pollen S-gene is identified as an F-box protein (SFB/SLF) which appears to act as a component of SCF complex. The SCF complex is associated with 26S proteasome degradation pathway in which non-self-S-RNases are targeted for degradation, and then the self-S-RNases are retained to specifically inhibit the growth of self-pollen tubes [34]. More recently, Xu et al. reported that two pollen specific SLF-interacting Skp1-like (SSK) proteins, PbSSK1 and PbSSK2, in P. bretschneideri could connect PbSLFs to PbCUL1 to form a putative canonical SCF SLF (SSK/CUL1/SLF) complex [24]. Based on the studies of Qiao et al. [34] and Xu et al. [24], we presume that pollen specific SFB 26 /SLF 26 may be involved in the formation of 26S proteasome pathway and specifically degrade the non-self-S-RNases after pollination, resulting in the fact that the PbS 26 -RNase is reserved. Therefore, the abundance of the PbS 26 -RNase gene transcript in the style could be greater after crosspollination than after self-pollination.

Development of the Rapid Method for
Identifying the PbS 26 -RNase Gene. In Japanese pear, a PCR-RFLP system was established for S-genotype assignments based on cloning of S 1 -to S 9 -RNase gene [5,7]. Herein, we reported a different method for rapidly identifying S-RNase genes in Chinese pears. For the PbS 26 -RNase gene, a novel primer pair S26PF2 and S26PR2 was designed based on DNA sequence comparisons. Genomic DNA of "Hongxiangsu" was amplified with S26PF2 and S26PR2, and one specific fragment with expected size (254 bp) was generated. Sequencing of this fragment revealed that it corresponded to the PbS 26 -RNase allele. Based on the PbS 26 -RNase allele, the other one allele was easily identified as S 21 by further PCR amplification and DNA sequencing. The S26PF2 and S26PR2 primer 8 The Scientific World Journal combination was also applied in 46 progenies of "Hongpisu" (S 12 S 26 ) × "Jinhua" (S 13 S 18 ), and the 254 bp sized fragment was produced in 13 progenies (Figure 8), suggesting that the 13 individuals harbor the PbS 26 -RNase allele. These results showed that the newly designed PbS 26 -specific primers could be used for rapid identification of the PbS 26 -RNase allele and will be useful in pear breeding programs.