Anatomy Of Floral Development And Cloning Of Key Genes Related To Floral Organs In Feijoa Sellowiana Berg

Feijoa sellowiana Berg., a new kind of fruit tree species of Myrtaceae family, has edible and ornamental value. It has a relatively long juvenile phase, and the seedlings blossom and bear fruits after four to five years. Therefore, elucidation of flowering mechanism of F. sellowiana is of great significance to shortening and adjusting the juvenile phase, genetical modification, and new cultivars breeding. In this paper, systematic studies on the development of F. sellowiana floral organs and related expressed genes via comparative anatomy experiment, cDNA and EST libraries construction, cloning and expression pattern detection of floral organs-related genes, and bioinformatics analysis etc. The main results are as follows:1. Studies on comparative anatomy of F. sellowiana floral organs. The development of F. sellowiana's flower bud differentiation was divided into six stages: predifferentiation, sepal formation, petal formation, stamen and pistil formation, ovary and anther formation and stamen and pistil maturation. F. sellowiana had four ventricles in each anther; the development of the anther wall was basic type; glandular tapetum; the meiosis in the microspore mother cells were belonged to a simultaneous type and the microspores were arranged in a tetrahedron shape in the tetrads; the pollen was matured in two or three days before blooming; The ripened pollen was of2-cell type. Megaspore mother cell meiosis to the formation of four spores were linear arrangement, which the chalazal megaspore had function. After three mitosis, the formation of mature eight-nucleate embryo sac was in the blossoming day. The embryo sac was a Polygonum type. Between male and female gametophyte and flower morphology had stably relationship in the developmental process.2. Construction of cDNA library of F. sellowiana floral organs. Total RNA was isolated from mixed floral buds of'Coolidge'in six different developmental stages. Then message RNA (mRNA) was purified from total RNA and was reverse transcribed to cDNA with SuperSciptTM II RnaseH-Reverse Transcriptase. The cDNA was ligated to pBluescript Ⅱ SK(+)XR vector and co-transformed into competent Escherichia coli cells DH10B. Eventually, a high-quality cDNA library of developing floral organs of F. sellowiana was successfully constructed. After the examination, the cDNA library aggregate the capacity of2.02×106clones, the recombination rate is92.32%, and the size of insertion segment range from600bp to2000bp. The cDNA library provieds a basis for EST library construction, isolation and identification of floral organs-related genes, gene miccroarry construction, and gene expression detection for F. sellowiana.3. Construction of EST library of F. sellowiana floral organs. Using the cDNA library of F. sellowiana floral organs as the material,3500positive clones was selected at random for DNA sequencing at the5'end, accordingly the EST library of F. sellowiana floral organs was constructed. After analysis and classification,2856original sequences were obtained with the sequencing seccess rate of81.6%.2840high-quality ESTs with the size of more than100bp were obtained, of which numbers of ESTs ranging from500bp to700bp were2520, accounting for88.73%. Assembling of the total of2840ESTs resulted in1682Unigenes that include438contigs and1244Singletons with the average size of698.76bp.1379ORFs (open reading frame) were predicted from1682Unigenes, of which1120ORFs were from1244Singletons accounting for81.22%, and259were from Contigs accounting for18.78%.4. Informatics analysis of EST sequences of F. sellowiana floral organs. Using EST sequences of1682Unigenes as queries for Blast analysis in Nucleotide Database,637could search out homologous sequences accounting for49.53%, of which367represented putative functional genes. Blast analysis in Non-redundant Database revealed that851Unigenes could search out homologous sequences accounting for66.17%, of which612represented putative functional genes. Blast analysis in SWISSPROT revealed that547Unigenes could search out homologous sequences accounting for42.53%. Blast analysis in COG database showed that236annotated sequences were obtained and were classified into18groups. Most of annotated sequences were found in four groups including Translation, ribosomal structure and biogenesis, Carbohydrate transport and metabolism, Posttranslational modification, protein turnover, and chaperones, and Lipid transport and metabolism. The numbers of the annotated sequences were50,38,34, and24, accounting for21.19%,16.1%,14.41%, and10.17%, respectively.5. Cloning and informatics analysis of FsLFY. Based on known588bp EST sequence for FsLFY gene, specific primers were designed for3'RACE (Rapid amplification of cDNA ends) and5'RACE PCRs. Accoringly, the full-length cDNA were obtained by assembling3'and5'teminal sequences. Genomic DNA sequences for FsLFY were also successfully amplified by genomic PCR and DNA sequencing. The FsLFY sequences were subbmited to GenBank and the Accession No. was JN562738. The full-length cDNA of FsLFY was1336bp with5'non-transcription region of91bp,3'non-transcription region of174bp and complete CDS of1071bp, encoding356amino acids. FsLFY gene contained two introns, the size of which was63bp and109bp, respectively. The amino acid sequence of FsLFY contained two viable regions and two conserved regions. At DNA sequence level, the FsLFY showed the lowest similarity with ClLFY (59.4%), the highest with EglLFY (94.4%). At amino acid sequence level, the FsLFY showed the lowest similarity with OiLFY (60.3%), the highest with EglLFY (95.2%). The predicted molecular weight and isoelectric points of FsLFY were40631.8Da and8.59, respectively, and belonged to unstable protein. Hydrophobicity index of the FsLFY varied from-3.933to1.456, exhibiting hydrophilicity in general. The FsLFY did not contain signal peptide. Secondary structure prediction showed that the FsLFY contained31Helixs,30Strands, and34Coils.6. Cloning and informatics analysis of FsPI. Based on known626bp EST sequence for FsPI gene, specific primers were designed for3'RACE and5'RACE PCRs. Accoringly, the full-length cDNA of FsPI were obtained by assembling3'and5'teminal sequences. The FsPI sequences were subbmited to GenBank and the Accession No. was JN562739. The full-length cDNA of FsPI was891bp with5' non-transcription region of6bp,3'non-transcription region of258bp and complete CDS of627bp, encoding208amino acids. FsPI exhibited characteristic structual of B type, MADS-box, ie., contained MADS domain, K domain, and PI domain. At DNA sequence level, the FsPI showed the lowest similarity with CrGLO1(66.7%), the highest with MqPI (93.0%). At amino acid sequence level, the FsPI showed the lowest similarity with CaPI (62.0%), the highest with MqPI (91.8%). The predicted molecular weight and isoelectric points of FsPI were24286.4Da and8.83, respectively, and belonged to unstable protein. Hydrophobicity index of the FsPI varied from-2.956to1.933, exhibiting hydrophilicity in general. The FsPI did not contain signal peptide. Secondary structure prediction showed that the FsPI contained25Helixs,21Strands, and20Coils.7. Cloning and informatics analysis of FsSEP1. Based on known338bp EST sequence for FsSEP1gene, specific primers were designed for3'RACE and5'RACE PCRs. Accoringly, the full-length cDNA of FsSEP1were obtained by assembling3' and5'teminal sequences. The FsSEP1sequences were subbmited to GenBank and the Accession No. was JN562740. The full-length cDNA of FsSEP1was1120bp with5' non-transcription region of83bp,3'non-transcription region of299bp and complete CDS of738bp, encoding245amino acids. FsSEP1exhibited characteristic structual of E type, MADS-box, ie., contained MADS domain, K domain, and SEP domain. At DNA sequence level,the FsSEP1showed the lowest similarity with AthSEPl (69.1%), the highest with EgM3(92.4%). At amino acid sequence level, the FsSEP1also showed the lowest similarity with AthSEPl (66.7%), the highest with EgM3(91.4%). The predicted molecular weight and isoelectric points of FsSEP1were28009.8Da and8.80, respectively, and belonged to unstable protein. Hydrophobicity index of the FsSEP1varied from-2.411to2.089, exhibiting hydrophilicity in general. The FsSEP1did not contain signal peptide. Secondary structure prediction showed that the FsSEP1contained23Helixs,25Strands, and26Coils.8. Expression pattern analysis of FsLFY. Fscalm (F. sellowiana Calmodulin) gene was isolated by RT-PCR and used as a reference gene for expression analysis of FsLFY gene. Semiquantitative RT-PCR analysis revealed that FsLFY gene was expressed in both reproductive organs and vegetative organs. In general, the FsLFY gene was strongly expressed in reproductive organs with significantly higher expression level was than in vegetative organs including fruit-bearing shoots, vegetative shoots, stems, tender leaves, and axillary buds. A stable high expression level was detected from April10to April22, whereas the expression level began to decrease with floral buds approaching to maturity after April26. In addition, FsLFY mRNA accumulation was slightly higher in fruit-bearing shoots and tender leaves than in the other three organs. The expession pattern of FsLFY suggests that it should play a role in both vegetative and reproductive development in F. sellowiana, but appears to be more closely related to reproductive development.