Cloning And Characterization Of Genes Involved In The Biosynthesis Of Very Long Chain Polyunsaturated Fatty Acids And The Reconstitution Of This Pathway In Crop Plants

Very long chain (≥C20) polyunsaturated fatty acids (VLCPUFA), such as arachidonic acid (AA; 20:4ω6), eicosapentaenoic acid (EPA; 20:5ω3) and docosapentaenoic acid (DHA; 22:6ω3) are essential for human health and nutrition. EPA and DHA are the main active ingredients of fish oil. The oily fish, such as salmon and mackerel are good sources for these fatty acids. However, the natural fish resources have declined rapidly in recent years. In addition, the recent findings of toxic chemicals in fish oil have raised the safety concerns on the consumption of fish and fish products, this again reduced the intake of these beneficial fatty acids. Therefore, alternative source of these VLCPUFAs are desirable. To reconstruct EPA/DHA metabolic pathway in crop plants through genetic engineering will be a sustainable and safe alternative to supplement these fatty acids.To achieve this, we first isolated a number of desaturase and elongase genes related to EPA and DHA biosynthesis pathways from different micro-organisms that are rich in these VLCPUFAs, including Isochrysis galbana, Phytophthora infestans and Amphidinium carterae. We then developed a novel multiple gene transfer (MGT) system for the construction of 3 plus genes in a single binary vector for a one- step transformation of higher plant. Using this method, we reconstructed a multigene vector containing five desaturase and elongase genes for the production of EPA. Various crop plants, including maize and cotton were transformed with this vector, and AA and EPA were detected in maize. The main results are listed as follows:(1) Isolation and functional analysis of aΔ5 desaturase gene from I.galbanaWe employed RACE strategy to isolate a cDNA from an I.galbana cDNA library. The full length cDNA consisted of 1329 nucleotides, encoding a protein of 442 amino acids with predicted molecular mass of 49.9 kDa. Bioinformatics analysis showed that it shared homology with other functionally known front-end fatty acid desaturases and the highest homology of 56% was found with aΔ5 desaturase from another microalgae Pavlova salina. As characterized by this family of desaturases, it contained an N-terminal cytochrome b5 domain, and three histidine rich motifs (his-boxes) related to electron transfer. Functional analysis by expression in Saccharomyces cerevisiae revealed that it can convert DGLA (20:3Δ^8,11,14) to AA (20:4Δ^5,8,11,14) by introducing a double bond in the acyl chain at theΔ5 position, indicating that this newly isolated cDNA sequence encoded a protein that specifically catalyzed the conversion of DGLA to the C20-Δ5-polyunsaturated fatty acid, AA, hence it was designated as IgD5.(2) Isolation and characterization of three fatty acid desaturase and elongase genes involved in eicosapentaenoic acid biosynthesis from P. infestansThree novel cDNA sequences putatively encoding aΔ6 (PinDes1), aΔ5 (PinDes2), and aΔ12 (PinDes3) desaturases have been identified from P. infestans by searching through its genome database with known desaturase sequences. These three putative desaturase cDNAs were subsequently cloned by RT-PCR from total RNA isolated from this fungus.. Sequencing results showed that the ORFs for PinDes1, PinDes2 and PinDes3 genes were 1371, 1151 and 1197bp in length, encoding for predicted proteins of 456, 516 and 398 amino acids, respectively. They all contained three histidine-rich domains (his boxes) that are characteristics of all membrane bound desaturases reported so far. Both PinD6 and PinD5 also contained a cytochrome b5 domain with a signature HPGG motif fused at their N-terminii that is common to other so-called front-end desaturases. Their functionalities were verified by expression and feeding experiments in S. cerevisiae. The results demonstrated that PinD6, PinD5 and PinD12 encodedΔ6 desaturase,Δ5 desaturase andΔ12 desaturase, respectively. These, together with the previously characterizedΔ17 desaturase andΔ6 elongase genes enabled us to propose that AA and EPA were synthesized in P. infestans mainly via theω6/Δ6 pathway.(3) Isolation of genes involved in DHA biosynthesis from the microalgae A. carteraeWe sequenced the transcripts of microalgae A. carterae. Bioinformatics analysis enabled us to identify four putative desaturase and five elongase expressed sequence tags (ESTs). 5′and 3′RACE strategy were employed to amplify these sequences from first strand cDNA prepared from total RNA isolated from this microalgae. The full length sequence of one of the putative desturase genes, named here as AcD20, was isolated. Its function is currently being analysed.(4) A novel method for the construction of plant multigene transformation vector was developedA simple and user friendly method that can be applied in most molecular laboratories for the construction of a single vector containing 3 or more transgenes for one step transformation of higher plant was dveloped. The main feature of this system was the usage of a pair of isocaudomers, XbaⅠand AvrⅡin a nested fashion, i.e., XbaⅠ-MCS-AvrⅡ-XbaⅠ. This structure was inserted in the MCS of the cloning vector pCR2.1-TOPO (Invitrogen), replacing its existing MCS to form the auxiliary vector, pAUX1. Multiple genes/constructs can be linked together by the alternate use of XbaⅠand AvrⅡin a reiterative fashion via conventional digestion and ligation processes. This is because the DNA sequence of interest, such as D1, D2, D3 etc, can be individually subcloned into pAUX1, resulting in the construction of pAUX-XbaⅠ-D1-AvrⅡ-XbaⅠ-,pAUX-XbaⅠ-D2-AvrⅡ-XbaⅠ- and pAUX-XbaⅠ-D3-AvrⅡ-XbaⅠ- etc. To link D1 and D2 together, the sequence D2 is first released by XbaⅠfrom the pAUX-XbaⅠ-D2-AvrⅡ-XbaⅠ-, and then ligated into the AvrⅡsite of predigested pAUX-XbaⅠ-D1-AvrⅡ-XbaⅠ-. This results in the generation of the double construct containing -XbaⅠ-D1-(?)-D2-AvrⅡ-(?)-XbaⅠ-. Because the newly formed sequences -AvrⅡ/XbaⅠ- and -XbaⅠ/AvrⅡwere no longer digestible by either XbaⅠor AvrⅡenzymes, yet a new AvrⅡsite was re-introduced following the insertion of sequence D2, the original XbaⅠ--AvrⅡ-XbaⅠin the double construct is reinstated. Therefore, in the same way, the third sequence D3 can be ligated into this double construct to produce a triple construct containing all 3 DNA sequences. The process can be repeated many more times so that multiple genes/DNA molecules can be stacked together with ease. In a proof-of-concept experiment, we constructed a plant transformation vector containing three reporter gene (GUS, BAR and GFP) expression cassettes flanked by two matrix attachment region sequences. The expression of all three genes was confirmed in transgenic Arabidopsis thaliana.(5) Production of EPA in crop plantsWe constructed a plant transformation vector pCamBAR-5EC containing four desaturase and one elongase genes for the production of EPA in higher plant. We transformed maize and cotton via the Agrobaterium mediated transformation method with this vector. Transgenic maize and cotton lines were obtained by selection on Basta herbicide containing agar plates. These transgenics were further tested for the existence of the transgenes by PCR anylysis. The production of AA and EPA were confirmed by GC and GC-MS analysis in transgenic maize.