Enhanced Production of Docosahexaenoic Acid in Mammalian Cells

Docosahexaenoic acid (DHA), one of the important polyunsaturated fatty acids (PUFA) with pharmaceutical and nutraceutical effects, may be obtained through diet or synthesized in vivo from dietary a-linolenic acid (ALA). However, the acumulation of DHA in human body or other mammals relies on the intake of high dose of DHA for a certain period of time, and the bioconversion of dietary ALA to DHA is very limited. Therefore the mammalian cells are not rich in DHA. Here, we report a new technology for increased prodution of DHA in mammalian cells. By using transient transfection method, Siganus canaliculatus Δ4 desaturase was heterologously expressed in chinese hamster ovary (CHO) cells, and simultaneously, mouse Δ6-desaturase and Δ5-desaturase were overexpressed. The results demonstrated that the overexpression of Δ6/Δ5-desaturases significantly enhanced the ability of transfected cells to convert the added ALA to docosapentaenoic acid (DPA) which in turn get converted into DHA directly and efficiently by the heterologously expressed Δ4 desaturase. This technology provides the basis for potential utility of these gene constructs in the creation of transgenic livestock for increased production of DHA/related products to meet the growing demand of this important PUFA.


Introduction
Docosahexaenoic acid (DHA; 22:6n-3) is an n-3 polyunsaturated fatty acid (PUFA) that is particularly enriched in mammalian tissues such as brain, testis, and retina. Clinical studies have demonstrated the crucial role of DHA in the development and functions of these tissues [1,2]. DHA enhances membrane elasticity and molecular motion and thus promotes signal transduction via enhanced protein/receptor interactions [3][4][5][6]. DHA is also the activating ligand for multiple transcriptional factors that control the expression of enzymes involved in fatty acid synthesis and b-oxidation [7]. Depletion of brain DHA in some mammals leads to distinct impairments in brain and neural function [8][9][10][11][12][13]. In adult humans, low levels of DHA in blood have been correlated with psychological disturbances such as alcoholism, major depression (nonpsychotic), postpartum depression, and senile dementia [19][20][21]. The essentiality of DHA for human infant nutrition in support of neuronal function has been shown by DHA supplementation, enhancing visual acuity and cognition-related test scores in human infants [14][15][16][17][18].
Although dietary ALA can be converted to DHA in vivo, the efficiency of conversion is limited [24]. It was shown that the mature rat brain has little capacity to synthesize its own DHA, even when stimulated to do so, under dietary conditions of feeding ALA alone or during n-3 PUFA dietary deprivation [25,26]. In contrast, the adult rat liver does convert ALA to DHA, but at a rate sufficient to convert ,2% of the total ALA that enters the liver per unit time, with the majority (.70%) of the entering ALA being lost to b-oxidation. Therefore, mammals are unable to produce high levels of DHA in their bodies. Algae are the primary producers of DHA and EPA in the ecosystem, and several refined algal oils are rich sources of DHA. Fish consume micro algae (phyto plankton) and therefore are rich in DHA and EPA, thus fish become the main dietary source of these PUFA. Nevertheless, marine products such as encapsulated fish oil are not suitable for daily use. Furthermore, the decline in marine fish stocks and the potential contamination of marine products with mercury and other chemicals call for land-based dietary sources of DHA. One strategy is to create genetically modified land animals by making use of the desaturases and elongases involved in DHA biosynthesis. In this paper, we report the successful production of DHA in chinese hamster ovary (CHO) cells by heterologous expression of Siganus canaliculatus D4 desaturases after overexpression of mice D6/D5-desaturase. The results provide a basis for potential applications of this technology in creating of transgenic livestock for enhanced production of DHA in the related products.

Construction of Recombinant Plasmids
The D4 desaturase gene from Siganus canaliculatus, sScD4, was artificially synthesized (with genetic codons optimized) from its original sequence (GenBank accession number: GU594278). Then, the synthetic gene (1352 bp) was inserted into the pcDNA3.1 expression vector with EcoRI/HindIII double digestion. The expression vectors for fads2 and fads1genes (cDNAs), coding for mouse D6/-desaturase and D5-desaturase respectively, were constructed previously in our laboratory, with the cDNAs cloned into pcDNA3.1. The double-gene expression vector pcDNA3.1-F2F1 was constructed as follows: based on the enzyme sites BglII/MluIcarried on pcDNA3.1-fad1, PCR primers were designed to amplify the pCMV-fad2-polyA DNA fragment on pcDNA3.1-fad2, and then the amplicon was inserted into pcDNA3.1-fad1 by BglII/MluIdouble digestion and ligation.

Cell Cultures and transfection
Transient transfections in chinese hamster ovary (CHO) cells were carried out using Lipofectamine TM 2000 (Invitrogen Inc., USA) as described in the manufacturer's manual. One day prior to transfection, cells were plated onto 60-mm culture dishes in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 10 mM PUFA substrates (DPA for sScD4 transfections, LA and ALA for F2F1 transfections). Cells were transfected with 8 mg of the plasmids. To monitor transfection efficiencies, cells were transfected with an expression vector containing the enhanced green fluorescent protein (EGFP) cDNA, pcDNA3.1-EGFP. The day following transfection, media were replaced with fresh media containing the same PUFA substrates as above. Two days after transfection, cells were harvested and used for analysis of gene expression or fatty acid composition.

RT-PCR
RT-PCR was performed to analyze transcripts in transgenic cells.Total RNA was extracted from the harvested cells, using TRIzol Reagent (Invitrogen) followed by DNase treatment to remove genomic DNA contamination. cDNA was prepared using the TaKaRa RNA PCR Kit 3.0(TaKaRa). The following primer pairs were used to detect the target genes-sScD4: 59-TCGGGCATCTGAGCGTGTTC-39 and 59-ATACTCCACG-GGCTGGGTCTCC-39; F2: 59-TAGTCTAGAGCATGGG-GAAGGGAGGTAAC -39and 59-AGCAAGCTTCATTTATG-GAGGT AAGCATCCAG -39; F1: 59 TACTCTAGAGCTTG-CTATGGCTCCCGACC -39, 59-ATGAAGCTTGCTATTG-GTGAAGGTAAGCGTCC-39. The PCR conditions were as follows: 94uC for 2 min for desaturation, 35 cycles of 94uC for 30 s for desaturation, 60uC for 30 s for annealing, and 72uC for 30 s for extension, followed by a final extension period of 72uC for 10 min. The amplified products were subjected to electrophoresis on a 2% agarose gel. The amplification of beta-actin gene was carried out as control.

Lipid Analysis
The fatty acid composition of total cellular lipids was analyzed as following: an aliquot of cell pellet homogenate in a micro reaction vessel (Supelco) was mixed with 1 ml of 14% BF3/ MeOH reagent containing 0.005% butylated hydroxytoluene (as antioxidant). After blanketed with nitrogen, the mixture was heated at 95-100uC for 1 h, then cooled to room temperature and methyl esters extracted in the hexane phase following addition of 1 ml H 2 O and 1 ml of hexane. The samples were centrifuged for 5 min at 3000 rpm, then the upper hexane layer was removed and concentrated under nitrogen. Fatty acid methyl esters were quantified with GC-MS by using an HP-5890 Series II gas chromatograph equipped with an Omegawax 250 capillary column (Supelco, Bellefonte, PA) attached to an HP-5971 mass spectrometer. The oven program was maintained initially at 150uC for 2 min, then ramped to 200uC at 10uC/min and held for 4 min, ramped again at 5uC/min to 240uC, held for 3 min, and finally, ramped to 270uC at 10uC/min and maintained for 5 min. Peaks were identified by comparison with fatty acid standards (Sigma), then area and percentage for each resolved peak were analyzed.

Statistical Analysis
Statistical analysis was performed with SPSS version 18.0 for Windows, using one-way ANOVA followed by LSD comparison test for posthoc analyses. Results were expressed as the mean 6 SD. A p-value of ,0.05 was considered to be statistically significant.

Results
To test whether pcDNA3.1-sScD4 could be expressed in mammalian cells, the vector was introduced into CHO cells, by using liposome transfection method. DPA was added to the cell culture medium for 3 days. The RT-PCR results showed mRNA of this synthetic gene could be greatly transcribed, while the cells transfected with the control plasmid pcDNA3.1-EGFP, which carries the enhanced green fluorescent protein as reporter gene, detected no sScD4 mRNA (Fig.1). The transient transfections of pcDNA3.1-sScD4 in CHO cells confirmed sScD4 can directly and efficiently convert DPA to DHA, as shown in Fig 2. Table 1 shows fatty acids composition of total cellular lipids from the CHO cells transfected with EGFP or sScD4. None of the percentage distribution of a fatty acid varied between the CHO cells transfected with EGFP and sScD4 except DPA and DHA. The CHO cells transfected with pcDNA3.1-F2F1 have the ability to overexpress of fads2 and fads1 genes as shown in Fig.1. The fatty acid profiles were remarkably different between the control cells transfected with pcDNA3.1-EGFP and the cells transfected with pcDNA3.1-F2F1 (Fig.3). In the cells overexpressing the D6/D5 desaturase, the percentage distribution of 18:2n-6 and 18:3n-3 were significantly decreased, while 20:3n-6, 20:4n-6, 20:5n-3, 22:5n-3 were significantly increased, as compared to the control cells. This means the substrates (18:2n-6 and 18:3n-3) added to the transfected cells were efficiently converted into longer PUFA. However, the DHA level increased slightly but was not statistically significant, indicating that the overexpression of the D6/D5 desaturases did not contribute to the elevation of DHA level. An interesting phenomenon is that the percentage composition of 20:3n-3 and 20:4n-3 were also significantly decreased in the cells overexpressing the D6/D5 desaturase when compared to the control cells. Obviously, the cells tend to convert them to longer chain PUFA, such as 20:5n-3, 22:5n-3.

Discussion
Here, we reported a new technology for enhanced production of DHA in mammalian cells. Mammals lack the D12 and D15 fatty acid desaturases responsible for converting oleic acid (18:1n-9) into linoleic acid (18:2n-6) and a-linolenic acid (18:3n-3) and thus, are unable to biosynthesize polyunsaturated fatty acids (PUFA) de novo [27,28]. Therefore, the ingested plant-derived 18:2n-6 and 18:3n-3 are essential dietary nutrients for mammals [27,28]. Nevertheless, most of the ingested plant-derived 18:2n-6 and 18:3n-3 can not be converted to ARA, EPA or DHA but are lost to b-oxidation. Our results demonstrated that, by overexpression of D6 and D5 fatty acid desaturases, the ability of converting 18:2n-6 to 20:4n-6, 18:3n-3 to EPA and DPA in mammalian cells was significantly elevated. This may be attributed to the fact that the overexpression of these desaturases disequilibrated the regulatory system of PUFA biosynthesis. We had studied that the overexpression of D6 and D5 fatty acid elongases (the most two important elongases involved in PUFA biosynthesis) in CHO cells, and found that they contributed little to elevating the level of any longer PUFA biosynthesized from 18:2n-6 or 18:3n-3 (unpublished data). We conclude that, among the fatty acid desaturases and elongases involved in PUFA biosynthesis in mammals, D6 and D5 desaturases are the key ones that contribute to production of longer PUFA.
The conversion of DPA to DHA is the rate-limiting step in the pathway of biosynthesis of DHA from ALA [29]. As shown in our results, even the overexpression of D6 and D5 fatty acid  Production of DHA in land animals may be a good choice for mankind to meet the growing demand for this important LC-PUFA, and our results provide potential solution on this regard. Like humans, most land animals can accumulate DHA through their diet, but usually there is no adequate DHA in animal feeds. Adding DHA to animal feeds for the accumulation of it in animal bodies is expensive and unadvisable. Mammalian cells in culture, to a large extent, can simulate the process of PUFA biosynthesis in living animals. So our technology, by heterologous expression of Siganus canaliculatus D4 desaturases and overexpression of mice D6/ D5-desaturases simultaneously, will probably produce high level of DHA in transgenic livestock carrying these gene constructs as in transfected CHO cells reported here. If such genetically modified land animals are successfully created and widely accepted by the public, DHA can be provided abundantly by these animals, which will probably change our views on the ordinary 'unhealthy' animal fat.
In conclusion, this work has demonstrated that mammalian cells can be genetically modified by transfection of an delta-4 desaturase that is heterologously expressed with concomitant overexpression of delta-5 and delta-6 desaturases to convert ALA to DHA efficiently, and provide a potential for creation of new land animal breeds who can produce abundant DHA in their related products.

Author Contributions
Conceived and designed the experiments: GZ TZ. Performed the experiments: GZ XJ JS TG MW ZW. Analyzed the data: GZ XJ QO GS. Contributed reagents/materials/analysis tools: QO TZ GS. Wrote the paper: GZ.