An alternate pathway to long-chain polyunsaturates: the FADS2 gene product Delta8-desaturates 20:2n-6 and 20:3n-3.

The mammalian #6-desaturase coded by fatty acid desaturase 2 (FADS2; HSA11q12-q13.1) catalyzes the first and rate-limiting step for the biosynthesis of long-chain polyunsaturated fatty acids. FADS2 is known to act on at least five substrates, and we hypothesized that the FADS2 gene product would have #8-desaturase activity. Saccharomyces cerevisiae transformed with a FADS2 construct from baboon neonate liver cDNA gained the function to desaturate 11,14-eicosadienoic acid (20:2n-6) and 11,14,17-eicosatrienoic acid (20:3n-3) to yield 20:3n-6 and 20:4n-3, respectively. Competition experiments indicate that #8-desaturation favors activity toward 20:3n-3 over 20:2n-6 by 3-fold. Similar experiments show that #6-desaturase activity is favored over #8desaturase activity by 7-fold and 23-fold for n-6 (18:2n-6 vs 20:2n-6) and n-3 (18:3n-3 vs 20:3n-3), respectively. In mammals, 20:3n-6 is the immediate precursor of prostaglandin E1 and thromboxane B1. 20:3n-6 and 20:4n-3 are also immediate precursors of long-chain polyunsaturated fatty acids arachidonic acid and eicosapentaenoic acid, respectively. These findings provide unequivocal molecular evidence for a novel alternative biosynthetic route to long-chain polyunsaturated fatty acids in mammals from substrates previously considered to be dead-end products.—Park, W. J., K. S. D. Kothapalli, P. Lawrence, C. Tyburczy, and J. T. Brenna. An alternate pathway to long-chain polyunsaturates: the FADS2 gene product#8-desaturates 20:2n-6 and 20:3n-3. J. Lipid Res. 2009. 50: 1195–1202. Supplementary key words polyunsaturated fatty acid biosynthesis • dihomo-g-linolenic acid • eicosanoid precursor biosynthesis Long-chain polyunsaturated fatty acids (LCPUFAs) are ubiquitous in mammalian tissue, achieving highest concentrations in the membranes of neural and other excitable tissue (1). LCPUFA of the n-3 and n-6 families, especially eicosapentaenoic acid (EPA; 20:5n-3), docosahexaenoic acid (22:6n-3), and arachidonic acid (20:4n-6), are bioactive components of membrane phospholipids and serve as substrates for signalingmolecules (2). Thedegreeof unsaturation of the membranes is determined by the action of enzymes involved in fatty acid biosynthesis and metabolism (3). Most organisms synthesize unsaturated fatty acids, but the pathways are specific to cell types and species. Fatty acid desaturases are enzymes that catalyze the introduction of cis double bonds at specific positions in a fatty acid chain (4). Desaturases in plants and lower animal species can introduce double bonds near the methyl end. Eukaryotic cells of higher animals, fungi, and dinoflagellates express membrane-bound acyl-CoA front-end desaturases (5, 6) catalyzing double bond introduction into the D6, D5, D8, and D4 positions. Mammalian front-end desaturases operate on diet-derived PUFA to synthesize LCPUFA, which can also be derived from the diet but possibly not in sufficient quantities to optimize health (7). The front-end desaturases are remarkable for their structural similarity and functional diversity. They all contain theN-terminal cytochrome b5 domain (HPGG) as electron donor and three histidine motifs, HXXXH, HXXHH, and QXXHH, conserved from human to microalgae (8). Molecular cloning and isolation of a D5-desaturase from Caenorhabditis elegans (9) and Mortierella alpina (10) and a D6-desaturase from C. elegans (11), M. alpina (12), rat (13), and mouse (14) have all been reported. The human fatty acid desaturase (FADS) gene cluster at 11q12-q13.1 encodes two desaturases with known function,D5-desaturase (FADS1) andD6-desaturase (FADS2) (14, 15), as well as a third putative desaturase gene (FADS3) (16), which thus far has no known substrate despite high homology to FADS1 and FADS2. Figure 1 shows the common n-3 and n-6 LCPUFA pathwaysmediated byD6 andD5 desaturases. TheD6-desaturase (FADS2) is known to operate on both 18:3n-3 and 18:2n-6, resulting in the synthesis of 6,9,12,15-18:4 and 6,9,12-18:3 (g-linolenic acid), respectively. This step is rate limiting This work was supported by National Institutes of Health Grant GM071534. Manuscript received 5 December 2008 and in revised form 21 January 2009. Published, JLR Papers in Press, February 6, 2009. DOI 10.1194/jlr.M800630-JLR200 Abbreviations: EPA, eicosapentaenoic acid; FADS, fatty acid desaturase; FAME, fatty acid methyl ester; GC-CACI-MS/MS, gas chromatography-covalent adduct chemical ionization tandem mass spectrometry; LCPUFA, long-chain polyunsaturated fatty acids; wt, wild type. 1 To whom correspondence should be addressed. e-mail: jtb4@cornell.edu Copyright © 2009 by the American Society for Biochemistry and Molecular Biology, Inc. This article is available online at http://www.jlr.org Journal of Lipid Research Volume 50, 2009 1195 and is followed by elongation to 8,11,14,17-20:4 and 8,11,1420:3 (dihomo-g-linolenic acid). A rapid D5-desaturation (FADS1) on these PUFA produces EPA and arachidonic acid. EPA can be further elongated and desaturated to yield docosahexaenoic acid by the pathway shown, which is accepted as mammalian pathway, or via a D4-desaturase as demonstrated in Thraustochytrium (17). The operation of an alternative pathway via C20 fatty acids using a D8-desaturase reported in unicellular organisms (18–21) has been verified by molecular cloning and functional characterization studies in Euglena gracilis (22), Acanthamoeba castellanii (23), and Perkinsus marinus (24). There are many reports of D8-desaturation activity in mammalian cells (25, 26), in rat and human testes (27, 28), and in mouse liver (29), though it has not been verified by molecular cloning, and the existence of D8-desaturation in rat microsomes has been questioned (30). The putative substrate of the D8-desaturase, 11,14-eicosadienoic acid (20:2 n-6), is found in human plasma and red cells as well as other tissues, and its concentration has recently been associated with human genetic variation in the FADS gene cluster (31, 32). The mammalian D6-desaturase coded by FADS2 uses at least five substrates, 18:2n-6, 18:3n-3, 24:6n-3, 24:5n-3 (33, 34), and 16:0. D6-desaturase in the sebaceous glands catalyzes desaturation of 16:0 to 16:1n-10 (sapienate), the most abundant fatty acid in human sebum, showing that substrate specificity is influenced by the cellular environment in which it is expressed (35). We hypothesized that the primate FADS2 gene product would have D8-desaturase activity and cloned baboon FADS2 into Saccharomyces cerevisiae, an organism with no native PUFA biosynthetic capability, to test for gain of D8-desaturation activity. Here, we report unambiguous evidence of the existence of D8desaturation in primates, suggesting alternative pathway for LCPUFA biosynthesis. MATERIALS AND METHODS RNA isolation and cDNA synthesis Total RNA from 30mg neonate baboon liver tissue homogenate was extracted using the RNeasy Mini kit (Qiagen, Valencia, CA). The yield of total RNA was assessed by 260 nm UV absorption. The quality of RNA was analyzed by 260/280 nm ratios of the samples and by agarose gel electrophoresis to verify RNA integrity. One microgram total RNA was reverse transcribed into first-strand cDNA using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA). The resulting cDNA was used as template for RT-PCR reactions. Cloning of baboon FADS2 and sequence analysis To identify baboon FADS2 cDNA sequence, primers were generated using human cDNA sequences for FADS2 (GenBank accession number NM_004265). PCR primers, FADS2 forward (5′ATGGGGAAGGGAGGGAACCAGGGCGA-3′) and FADS2 reverse (5′-TCATTTGTGAAGGTAGGCGTCCAGCCA-3′) were ordered from Integrated DNATechnologies (Coralville, IA) and were amplified with baboon liver cDNA as template and high-fidelity Taq polymerase (Roche Diagnostics) using Eppendorf gradient thermal cycler. Cycling conditions were as follows: initial denaturation at 95°C for 5 min followed by 40 cycles of denaturation at 95°C for 30 s, annealing at 72°C for 45 s, and extension at 72°C for 1 min, with a final extension at 72°C for 5 min. PCR product was separated by electrophoresis on 2% agarose gel stained with ethidium bromide and band of appropriate size was obtained. The PCR product was gel purified and cloned in pGEM T-Easy vector (Promega) and sequenced using T7 forward and SP6 reverse universal primers at Cornell University Life Sciences Core Laboratories Center using the Applied Biosystems automated 3730 DNA analyzer. We successfully cloned the baboon FADS2 protein coding region (GenBank accession number EU780003). The pGEM TEasy vector with FADS2 was named pTFADS2. Transformation into yeast (S. cerevisiae) The entire coding regions of baboon FADS2 was amplified from pTFADS2 with primers FADS2 -KOZAK forward (5′ CCCAAGCTTACCATGGGGAAGGGAGGGAACCAGGGCGA-3′) including the HindIII site and FADS2-KOZAK reverse (5′CCGCTCGAGTCATTTGTGAAGGTAGGCGTCCAGCCA-3′) including XhoI site. The high-fidelity Taq polymerase (Roche Diagnostics) was used to minimize potential PCR errors. The amplified PCR product containing baboon FADS2 was gel purified, restriction digested, and inserted into HindIII and XhoI sites behind the GAL1 promoter of pYES2 vector (Invitrogen) to yield the plasmid pYFADS2. The constructed plasmid of pYFADS2 was transformed into S. cerevisiase (strain INVSc1 from Invitrogen) using S. c. Easy CompTM Transformation Kit (Invitrogen), and the transformants were verified by DNA sequencing. Expression of baboon FADS2 For functional expression characterization, transformed yeast strains with pYES2 (empty vector) as a negative control and pYFADS2 were grown for 24 h in S. cerevisiae minimal media without uracil. As another negative control, wild S. cerevisiae (INVSc1) was cultured in S. cerevisiae minimal medium with uracil. Expression of the transgene was induced whenOD600 reached 0.4. At that time, appropriate fatty acids, 1mM linoleic acid (18:2n-6),a-linolenic acid (18:3n-3), eicosadienoic acid (20:2n-6), and eicosatrienoic acid (20:3n-3), were added in the presence of 1% tergitol-Nonidet P-40 (Sigma-Aldrich) to the cultures and were grown at 30°C with constant shaking. The samples were collected after 48 h for fatty acid analysis. All treatments were performed in duplicate. Fig. 1. Pathways for LCPUFA biosynthesis. The conventional pathway consists of alternating desaturation and elongation leading to LCPUFA. D8-Desaturation of 20:2n-6 and 20:3n-3 would yield 20:3n-6 and 20:4n-3, intermediates in the conventional pathway to 20:4n-6 and 20:5n-3, as well as immediate eicosanoid precursors. 1196 Journal of Lipid Research Volume 50, 2009

Long-chain polyunsaturated fatty acids (LCPUFAs) are ubiquitous in mammalian tissue, achieving highest concentrations in the membranes of neural and other excitable tissue (1). LCPUFA of the n-3 and n-6 families, especially eicosapentaenoic acid (EPA; 20:5n-3), docosahexaenoic acid (22:6n-3), and arachidonic acid (20:4n-6), are bioactive components of membrane phospholipids and serve as sub-strates for signaling molecules (2). The degree of unsaturation of the membranes is determined by the action of enzymes involved in fatty acid biosynthesis and metabolism (3). Most organisms synthesize unsaturated fatty acids, but the pathways are specific to cell types and species.
Fatty acid desaturases are enzymes that catalyze the introduction of cis double bonds at specific positions in a fatty acid chain (4). Desaturases in plants and lower animal species can introduce double bonds near the methyl end. Eukaryotic cells of higher animals, fungi, and dinoflagellates express membrane-bound acyl-CoA front-end desaturases (5,6) catalyzing double bond introduction into the D6, D5, D8, and D4 positions. Mammalian front-end desaturases operate on diet-derived PUFA to synthesize LCPUFA, which can also be derived from the diet but possibly not in sufficient quantities to optimize health (7).
The operation of an alternative pathway via C20 fatty acids using a D8-desaturase reported in unicellular organisms (18)(19)(20)(21) has been verified by molecular cloning and functional characterization studies in Euglena gracilis (22), Acanthamoeba castellanii (23), and Perkinsus marinus (24). There are many reports of D8-desaturation activity in mammalian cells (25,26), in rat and human testes (27,28), and in mouse liver (29), though it has not been verified by molecular cloning, and the existence of D8-desaturation in rat microsomes has been questioned (30). The putative substrate of the D8-desaturase, 11,14-eicosadienoic acid (20:2 n-6), is found in human plasma and red cells as well as other tissues, and its concentration has recently been associated with human genetic variation in the FADS gene cluster (31,32).

RNA isolation and cDNA synthesis
Total RNA from 30 mg neonate baboon liver tissue homogenate was extracted using the RNeasy Mini kit (Qiagen, Valencia, CA). The yield of total RNA was assessed by 260 nm UV absorption. The quality of RNA was analyzed by 260/280 nm ratios of the samples and by agarose gel electrophoresis to verify RNA integrity. One microgram total RNA was reverse transcribed into first-strand cDNA using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA). The resulting cDNA was used as template for RT-PCR reactions.

Cloning of baboon FADS2 and sequence analysis
To identify baboon FADS2 cDNA sequence, primers were generated using human cDNA sequences for FADS2 (GenBank accession number NM_004265). PCR primers, FADS2 forward (5′-ATGGGGAAGGGAGGGAACCAGGGCGA-3′) and FADS2 reverse (5′-TCATTTGTGAAGGTAGGCGTCCAGCCA-3′) were ordered from Integrated DNA Technologies (Coralville, IA) and were amplified with baboon liver cDNA as template and high-fidelity Taq polymerase (Roche Diagnostics) using Eppendorf gradient thermal cycler. Cycling conditions were as follows: initial denaturation at 95°C for 5 min followed by 40 cycles of denaturation at 95°C for 30 s, annealing at 72°C for 45 s, and extension at 72°C for 1 min, with a final extension at 72°C for 5 min. PCR product was separated by electrophoresis on 2% agarose gel stained with ethidium bromide and band of appropriate size was obtained. The PCR product was gel purified and cloned in pGEM T-Easy vector (Promega) and sequenced using T7 forward and SP6 reverse universal primers at Cornell University Life Sciences Core Laboratories Center using the Applied Biosystems automated 3730 DNA analyzer. We successfully cloned the baboon FADS2 protein coding region (GenBank accession number EU780003). The pGEM T-Easy vector with FADS2 was named pTFADS2.

Transformation into yeast (S. cerevisiae)
The entire coding regions of baboon FADS2 was amplified from pTFADS2 with primers FADS2-KOZAK forward (5′-CCCAAGCTTACCATGGGGAAGGGAGGGAACCAGGGCGA-3′) including the HindIII site and FADS2-KOZAK reverse (5′-CCGCTCGAGTCATTTGTGAAGGTAGGCGTCCAGCCA-3′) including XhoI site. The high-fidelity Taq polymerase (Roche Diagnostics) was used to minimize potential PCR errors. The amplified PCR product containing baboon FADS2 was gel purified, restriction digested, and inserted into HindIII and XhoI sites behind the GAL1 promoter of pYES2 vector (Invitrogen) to yield the plasmid pYFADS2. The constructed plasmid of pYFADS2 was transformed into S. cerevisiase (strain INVSc1 from Invitrogen) using S. c. Easy Comp™ Transformation Kit (Invitrogen), and the transformants were verified by DNA sequencing.

Expression of baboon FADS2
For functional expression characterization, transformed yeast strains with pYES2 (empty vector) as a negative control and pY-FADS2 were grown for 24 h in S. cerevisiae minimal media without uracil. As another negative control, wild S. cerevisiae (INVSc1) was cultured in S. cerevisiae minimal medium with uracil. Expression of the transgene was induced when OD 600 reached 0.4. At that time, appropriate fatty acids, 1 mM linoleic acid (18:2n-6), a-linolenic acid (18:3n-3), eicosadienoic acid (20:2n-6), and eicosatrienoic acid (20:3n-3), were added in the presence of 1% tergitol-Nonidet P-40 (Sigma-Aldrich) to the cultures and were grown at 30°C with constant shaking. The samples were collected after 48 h for fatty acid analysis. All treatments were performed in duplicate.

Fatty acid analysis
The yeast cells were harvested by centrifugation at 4,000 rpm for 5 min. The cell pellets were washed twice with tergitol-Nonidet P-40 and finally twice with distilled water. Fatty acid methyl esters (FAMEs) were prepared using modified one-step lipid extraction method of Garces and Mancha (36). FAMEs were structurally identified by gas chromatography-covalent adduct chemical ionization tandem mass spectrometry (GC-CACI-MS/MS) (37-39) and quantitatively analyzed by GC-flame ionization detection. An equal weight FAME mixture was used to verify response factors on a daily basis (40). For competition experiments, GC analyses were performed in triplicate.
The alternative synthetic pathways to arachidonic acid from 20:2n-6 are either by sequential action of a D8-desaturase and a D5-desaturase or vice versa. Initial D8-desaturation yields the eicosanoid precursor 20:3n-6, whereas initial D5desaturase activity yields 5,11,14-20:3. There are numerous reports showing that 11,14-20:2 is D5-desaturated to 5,11, 14-20:3 (30, 47-50). However, no clear evidence for the conversion of 5,11,14-20:3 to 20:4n-6 has been found (30,47,49). Sprecher and coworkers have studied the desaturation of 11,14-20:2 with isotope labeling in vitro and in vivo, and consistently find that rat liver does D5-desaturate it to 5,11,14-20:3, but they find no evidence of D8-desaturation activity on this product (47,49). Fourteen day feeding of 5,11,14-20:3 led to the accumulation of this PUFA in liver phosphoglycerides where it decreased 20:4n-6, while feeding of 11,14-20:2 did not alter 20:4n-6 levels (30). The production of 5,11,14-20:3 and 5,11,14,17-20:4 in human leukemia K562 cells has been reported in which the D5-desaturase is the only active desaturase operating because of the lack of D6-desaturase activity in these cells (50). Consistent with this report, we recently found significant amounts of 7,11, 14-20:3, 7,11,14,17-20:4, and 9,13,16,19-22:4 in the liver lipids of chow-fed FADS2 null mice, all of which may be synthesized by action of D5-desaturase, coded by FADS1, on 18:2n-6 or 18:3n-3, followed by prompt elongation (C. Stroud, P. Lawrence, J. T. Brenna, and M. Nakamura, unpublished observations). These data support the hypothesis that the D5-desaturase acts on PUFA only when its preferred substrate is not available, which may well imply that its products found in experimental studies are not relevant in vivo under normal conditions. Reports of D8-desaturase activity in rodent and human testes have appeared (27,28), and the most recent study shows stable isotope labeling best explained by direct conversion of 11,14-20:2 to 20:4n-6 via D8-desaturation, albeit as a minor pathway (30), but there are no existing molecular data to implicate a specific gene responsible for coding for vertebrate D8-desaturase activity. D8-Desaturation has been shown unequivocally in unicellular organisms where the gene has been cloned and is active when expressed in Arabidopsis thaliana (51). A D8-desaturase was first reported in the single cell protist E. gracilis (22) and later, along with a D9PUFA-elongase, in Isochrysis galbana (21) as well as the free living amoeba A. castellanii (23). The present report is the first to show that a vertebrate gene product introduces a double bond at the D8 position, demonstrating an alternative pathway to LCPUFA biosynthesis.
The competition experiments provide insight as to whether D8-desaturase activity of the FADS2 protein product can be important in vivo. The synthesis of 20:4n-3 dominates by 3.1fold over 20:3n-6, consistent with long-established observations for D6-desaturase preference for 18:3n-3 over 18:2n-6 (52-54). These observations are also are consistent with in vitro work showing that the biosynthesis of n-6 PUFA is strongly suppressed by ,2% of calories of 18:3n-3, whereas nearly 10 times as much 18:2n-6 is required to equally suppress n-3 PUFA biosynthesis (55), indicating that the affinity of the biosynthetic apparatus favors n-3 PUFA. Our competition experiments ( Table 2) also establish that the FADS2 gene product exhibits both D6-desaturase and D8-desaturase activities when both substrates are available. As expected, the FADS2 gene product showed higher D6-desaturase activity by acting on the 18:2n-6 substrate to generate 7-fold more product than for the 20:2n-6. The relative action D6/D8 activity toward the n-3 was much greater, at 23-fold, indicating that the conventional pathway would be strongly favored when both substrates are available.
In conclusion, baboon FADS2 gene cloned into S. cerevisiae causes gain of D8-desaturase activity, in addition to coding for D6-desaturase activity. D8-Desaturase activity on 20:2n-6 leads directly to 20:3n-6, the immediate precursor of PGE1 and of 20:4n-6. All available evidence indicates that D8desaturation is a minor pathway, but further study may show that it becomes important when there is high demand for eicosanoid synthesis, such as in inflammation or vasodilation, particularly in situations in which specialized tissues require 20:3n-6 as a precursor to prostaglandins E1 and F1a, hydroxyeicosatrienoic acids, or thromboxane B1. This alternative pathway to the eicosanoid precursors may explain data suggesting that 20:2n-6 levels are related to human health.