| Limonoids, as a class of triterpenoid, have excellent biological activities and ecological effects, which can effectively eliminate Plutella xylostella, Oxya chinesis, Pieris rapae and other agricultural pests. Thus limonoids have been considered as a kind of ideal botanical pesticide. Melia azedarach and Azadirachta indica, which are two members of Meliaceae plants, have closely genetic relationship in evolution. Chemical composition analysis shows many different kinds of limonoids involved, among which no azadirachtin was detected in M. azedarach. Based on these differences in M. azedarach and A. indica, the transcripts of both leaf samples were sequenced using the Illumina high-throughput sequencing technique in this study, acquiring huge numbers of gene sequences information. Then the sequenced data were further analyzed by series of bioinformatic methods, such as alter splicing site forecast, single nucleotide polymorphisms(SNP) and insertion-deletion(In Del) site prediction, new genes detection, gene structure optimization and gene expression quantification for A. indica; coding sequences(CDS) prediction, simple sequence repeat(SSR) and functional annotation for M. azedarach; the orthology genes comparatively analysis for both. Finally, the genes involved in the biosynthesis of limonoids were screened from the bioinformatic data mentioned above. Cloning and expression of these target genes were performed, and the main results were as follows:1. Through total RNA extraction, strand-specific cDNA library construction and high throughput sequencing for three A. indica leaf samples(named as Y1, Y2 and Y3), 23 M, 22 M and 28 M pair-end reads were obtained, respectively. Fast QC analysis showed that clean reads accounted for more than 98.2% of the total raw reads, and the sequenced base quality value Q30 reached more than 91.44%. These results suggested that the sequenced data were reliable. Based on the referenced neem genome information, more than 81.26% clean reads were found to map to the genome. The mapped reads were joint using Cufflink software and then compared with the reference genome. About 83,000 alternative splicing events were detected, among which the alternative 5’ first exon and 3’ last exon accounted for over 80% of the total events. Meanwhile, 5037 gene structures were optimized and 1179 new genes were found by comparative analysis to neem genome, of which 1055 new genes were functionally annotated by alignment with known database. Analysis of gene expression levels revealed that the FPKM value were between 0.1 and 100 for most genes, and the coefficient of correlation showed good repeatability in three A. indica samples. Additionally, 72,644 potential SNP sites and 10,368 In Del sites were predicted by GATK software. The original sequenced data were submitted to Sequence Read Archive(SRA) Database of NCBI and the accession number were obtained: SRR3180937 for Y1, SRR3181105 for Y2 and SRR3181166 for Y3.2. Similarly, 26 M, 23 M and 21 M pair-end reads were separately obtained from three M. azedarach leaf samples(called K1. K2 and K3) by strand-specific RNA-seq. By Fast QC analysis, clean reads were found to account for more than 98% of the raw reads and the value of Q30 reached more than 91.27%, which revealed high sequencing quality. Then de novo assembly of the clean reads were conducted by Trinity, acquiring 6,014,722 contigs, 225,972 transcripts and 91,607 Unigenes, and the value of N50 were 48, 2628 and 1321 respectively. The results showed good assembly integrity. By using Bowtie software, the clean reads were mapped to the unigenes and transcripts pool and high mapped ratio was observed, up to 93%, suggesting that perfect assembly was occurred. The potential CDS region prediction by Trans Decoder software displayed 68,225 possible CDS, of which 25,574 CDS had the complete open reading frame. SSR analysis of 18,099 unigenes with more than 1 kb length by MISA software displayed 7338 possible SSR sites located in 5521 unigenes. A total of 3732 unigenes were functionally annotated by comparing with the known database. Gene expression level analysis showed that the FPKM values of most unigenes were between 0.1 and 100, and the coefficient of correlation also exhibited good reproducibility in three M. azedarach samples. In the same way, the raw data were deposited in NCBI SRA Database and the accession number were obtained: SRR3183379 for K1, SRR3183380 for K2 and SRR3183381 for K3.3. Through homologous alignment of the protein sequences from two Meliaceae plants, 3867 orthologous genes were screened. Then gene expression levels were normalized on account of different species background, and the genes with more than a twofold difference in the expression quantity were defined as the differentially expressed genes(DEGs). About 2478 DEGs were selected and 1388 genes of them showed up-regulation while 1090 genes showed down-regulation. Alignment of the DEGs with known database suggested that 2459 DEGs were functionally annotated. Top GO and KEGG enrichment analysis demonstrated that in A. indica, 71 genes were significantly up-regulated, involving ATP binding, sequence-specific DNA binding, arginine/serine-rich splicing factor, protein tyrosine phosphatase, deformylase, isomerase and methyltransferase, while 281 down-regulated DEGs were related to photolyase, helicase, N-acetyltransferase, succinyltransferase, prenyltransferase and so on. After KEGG annotation, genes involved in terpenoid biosynthesis were screened from A. indica and M. azedarach. The results showed that 107, 24, 74 and 30 genes were separately involved in terpenoid backbone, monoterpene, diterpene, sesquiterpene and triterpene biosynthesis in A. indica, while 135, 29, 50 and 27 unigenes in M. azedarach were referred to participate in the above-mentioned pathways respectively, and 6, 2, 3 and 1 corresponding DEGs were also observed. There were 6 of 12 DEGs involved in terpenoid backbone biosynthesis appeared significantly up-regulated in A. indica in further analysis. The result suggested that the total terpene biosynthesis in neem was more active than that in M. azedarach, which was consistent with the chemical component analysis data. Furthermore, 5 of 6 DEGs involved in monoterpene, diterpene, sesquiterpene and triterpene biosynthesis were down-regulated in neem, including β-amyrin synthase,(+)-neomenthol dehydrogenase, ent-kaurenoic acid hydroxylase and gibberellin 2-oxidase. Only ent-kaurene oxidase was found to be up-regulated in M. azedarach.4. By using molecular biology technique, 26 genes related to limonoids biosynthesis were successfully cloned, including 6 from A. indica named as Ai DXR, Ai FPPS, Ai HMGS, Ai HMGR, Ai IPI and Ai SS; and 20 from M. azedarach called Ma DXR, Ma HMGS, Ma IDS, Ma HMGR, Ma ATCC, Ma MK, Ma CMK, Ma MDC, Ma SS, Ma DXS1, Ma DXS2, Ma HDS, Ma MECPS, Ma IPI, Ma GGPPS1, Ma GGPPS2, Ma GGPPS3, Ma SQLE1, Ma SQLE2 and Ma GPPS. Ai SS, which was the most proximal known node to the end-product azadirachtin biosynthesis, was selected to construct the recombinational prokaryotic expression plasmid p ET32a-Ai SS and then transformed into E. coli BL21(DE3) competent cell for protein expression. SDS-PAGE analysis showed that the recombinant protein was expressed as inclusion body, existing in precipitation component. Therefore, another recombinant eukaryotic expression plasmid p PICZαA-Ai SS was constructed and expressed in Pichia Pastoris GS115. Although positive clones were selected, SDS-PAGE results showed no detectable protein of interest in the extracellular solution. This may attribute to the N-terminal signal peptide of Ai SS, which could lead recombinant protein to be trapped into the membrane when secreted to extracellular space. The successful clone of these 26 genes can not only help us to further understand limonoids biosynthetic pathway in Meliaceae plants, but also transform engineering bacteria to harvest high-yield squalene, laying the foundation for large-scale fermentation of limonoids in the future. |
PDF Full Text Request