Journal of Experimental Botany. May/31/2010; 61(6): 1827-1838

Published online Mar/14/2010

PMID: 20231328

PMC: 2852672

doi: 10.1093/jxb/erq051

Identification and functional analysis of the genes encoding Δ⁶-desaturase from Ribes nigrum†

Abstract

Gamma-linolenic acid (γ-linolenic acid, GLA; C18:3 Δ^{6, 9, 12}) belongs to the omega-6 family and exists primarily in several plant oils, such as evening primrose oil, blackcurrant oil, and borage oil. Δ⁶-desaturase is a key enzyme involved in the synthesis of GLA. There have been no previous reports on the genes encoding Δ⁶-desaturase in blackcurrant (Ribes nigrum L.). In this research, five nearly identical copies of Δ⁶-desaturase gene-like sequences, named RnD8A, RnD8B, RnD6C, RnD6D, and RnD6E, were isolated from blackcurrant. Heterologous expression in Saccharomyces cerevisiae and/or Arabidopsis thaliana confirmed that RnD6C/D/E were Δ⁶-desaturases that could use both α-linolenic acids (ALA; C18:3 Δ^9,12,15) and linoleic acid (LA; C18:2 Δ^9,12) precursors in vivo, whereas RnD8A/B were Δ⁸-sphlingolipid desaturases. Expression of GFP tagged with RnD6C/D/E showed that blackcurrant Δ⁶-desaturases were located in the mitochondrion (MIT) in yeast and the endoplasmic reticulum (ER) in tobacco. GC-MS results showed that blackcurrant accumulated GLA and octadecatetraenoic acids (OTA; C18:4 Δ^6,9,12,15) mainly in seeds and a little in other organs and tissues. RT-PCR results showed that RnD6C and RnD6E were expressed in all the tissues at a low level, whereas RnD6D was expressed at a high level only in seeds, leading to the accumulation of GLA and OTA in seeds. This research provides new insights to our understanding of GLA synthesis and accumulation in plants and the evolutionary relationship of this class of desaturases, and new clues as to the amino acid determinants which define precise enzyme activity.

Introduction

Gamma-linolenic acid (γ-linolenic acid, GLA; C18:3, Δ^6,9,12) is an essential fatty acid (EFA) of the omega-6 family and exists primarily in a few plant-based oils, such as evening primrose oil, blackcurrant oil, and borage oil (Sayanova et al., 1997; Das, 2006). GLA was also found in considerable quantities in Spirulina, a cyanobacterium (Subudhi et al., 2008), some species of Primula (Sayanova et al., 1999a), and fungi (Sperling et al., 2000; Na-Ranong et al., 2005). However, there is no GLA in the major oil crops (Reddy and Thomas, 1996; Sayanova et al., 1999b), such as oilseed rape, soybean, peanut, and palm. GLA may be helpful in the prevention and/or treatment of diabetes, eye disease, osteoporosis, premenstrual syndrome (PMS), eczema, allergies, rheumatoid arthritis, high blood pressure, and heart disease (Horrobin, 1992).

Δ⁶-desaturase is classified as an acyl-lipid desaturase that introduces a double bond between the pre-existing double bond and the carboxyl (front) end of the fatty acid (Qi et al., 2004). It is critical for the metabolic conversion of linoleic acid (LA; C18:2 Δ^9,12) and α-linolenic acids (ALA; C18:3 Δ^9,12,15) into GLA and octadecatetraenoic acids (OTA; C18:4 Δ^6,9,12,15), respectively (Horrobin, 1992; Reddy and Thomas, 1996). The cytochrome b5 (Cyt b5) domain is essential for the borage Δ⁶-desaturase (Sayanova et al., 1999b), and three conserved His boxes of the enzyme are also proved to be essential for its function (Sayanova et al., 2001).

In the last 20 years, a number of Δ⁶-desaturase genes, from cyanobacterium (Reddy et al., 1993), fungi (Sakuradani and Shimizu, 2003), plants (Sayanova et al., 1997, 1999a; Sperling et al., 2000; Whitney et al., 2003), and animals (Napier et al., 1998; Seiliez et al., 2001), were isolated and characterized. Accumulation, function, and expression of Δ⁶-desaturase genes had been investigated in only a few higher plants synthesizing GLA. In Borago officinalis, GLA is accumulated up to 23.1% of the total fatty acids in mature seeds, and more than 8.0% in other tissues. The Δ⁶-desaturase gene is expressed but at different levels in all tissues: higher in seeds, young leaves, and roots, and lower in mature leaves and flowers (Sayanova et al., 1999c). In Anemone leveillei, GLA and OTA are absent in seeds, but are readily detectable in leaves. The Δ⁶-desaturase gene is expressed at the highest level in leaves, a moderate level in seeds, and is completely absent in roots (Whitney et al., 2003). In Primula, different species accumulated varying levels of GLA and OTA. In the species with a high level of Δ⁶-fatty acids, GLA and OTA accumulated mainly in the leaves and seeds, and relatively little in roots (Sayanova et al., 1999a).

It was recently reported that there were two nearly identical copies of the Δ⁶-desaturase gene, which differ by nine silent mutations, in the marine microalga Glossomastix chrysoplasta (Hsiao et al., 2007). While, in the higher plants that had been investigated, there were no instances of several different active Δ⁶-desaturase genes being found in the same organism.

Δ⁶-desaturases from some organisms exhibited substrate selectivity. The Δ⁶-desaturase of moss (Ceratodon purpureus) preferred C16:1 Δ⁹ over C18:1 Δ⁹ (Sperling et al., 2000), whereas the Δ⁶-desaturase from Mucor rouxii had a strong preference for the C18:1 Δ⁹ (Na-Ranong et al., 2005). In Boraginaceae, the Δ⁶-desaturase from Macaronesian Echium gentianoides preferred the LA substrate over the ALA substrate (García-Maroto et al., 2006). Δ⁶-desaturases from different Primula species had different substrate preferences; for example, the Δ⁶-desaturase from Primula vialii had a very strong preference for ALA compared with LA (Sayanova et al., 2003), whereas that from Primula cortusoides mainly desaturated LA (Sayanova et al., 2006).

The fatty acid-synthesizing and desaturation system had been widely known to exist in chloroplasts and the endoplasmic reticulum (ER) in higher plants. It had been confirmed by using the [¹⁴C]linoleoyl phosphatidylcholine that the microsomal membrane, prepared from the developing cotyledons, contained an active Δ⁶-desaturase, which catalysed the conversion of linoleate into γ-linolenate (Stymne and Stobart, 1986). Since then, front-end desaturases were generally thought to locate in the ER (Napier et al., 1999). To date, however, there has been no direct evidence for subcellular localization to locate plant Δ⁶-desaturases.

Blackcurrant seed oil, typically containing 15–20% GLA, is known to be one of the available natural sources of GLA (Traifler et al., 1988). However, until now, the Δ⁶-desaturase blackcurrant gene has not been reported.

In this research, GLA was detected mainly in seeds and rarely in other organs of blackcurrant. Three genes were cloned (RnD6C, RnD6D, and RnD6E) encoding Δ⁶-desaturases and two genes (RnD8A, RnD8B) encoding Δ⁸-sphingolipid desaturases in blackcurrant. RnD6C, RnD6D, and RnD6E could use both LA and ALA as substrate. RnD6C/D/E was located in the mitochondrion (MIT) in yeast and the ER in tobacco. RnD6C and RnD6E were expressed in all the tissues at a low level, whereas RnD6D was expressed at a high level but only in seeds. This research should provide a useful addition to the previously identified orthologues from other plants and may well prove to be useful for biotechnological applications. Furthermore, this research may also provide not only new insights into the evolutionary relationship of this class of desaturases but also some new clues as to the amino acid determinants which define the precise enzyme activity (i.e. either Δ⁶-desaturase or Δ⁸-sphingolipid desaturase).

Materials and methods

Plant materials

Blackcurrant (Ribes nigrum L. var. BRÖDTORP), a Finnish variety (Wassenaar and Hofman, 1966) was used in this research.

Blackcurrant genomic DNA extraction and full-length Δ⁶-desaturase gene cloning

Genomic DNA was extracted by a DNA extraction kit (Biomed-tech, Beijing, China) from young leaves of blackcurrant. The DNA fragments of putative cytochrome b5 fusion desaturase genes were amplified from the genomic DNA by using degenerated primers designed according to the sequence of conserved histidine boxes of known Cyt b5 fusion desaturases (Reddy and Thomas, 1996; Whitney et al., 2003; Sayanova et al., 2003). The primer sequences were: PL, 5′-TTGGGTGAAAGACCATCCA(T)GGT(A)GG-3′ PR, 5′-GGGAAACAAA(G)TGA(G)TGCTCAACCTG-3′. Five amplified products (about 1032 bp) were cloned into the pEASY-Blunt vector (TransGen Biotech, Beijing, China) and sequenced. The 5′ and 3′ extension of these fragments were obtained by the genomic walking method using the BD GenomeWalker™ Universal Kit (BD Biosciences, Palo Alto, CA, USA). Conserved genomic walking primers were: 5GSP1 (480), GACATTGTCTG(A/G)TAATGCCCAGAATC; 5GSP2 (270), CCTATAATCTTTGGATACATCTG(A/g)GA(C/t); 3GSP1(702), AGGAA(A/G)TTGGAGTTTGATGCTTTTGCTAGGTTC; 3GSP2(1078), CG(T/C)CTTGGATGGATTGGTTTCATGGTGG. The PCR products were cloned into the pEASY-Blunt vector (TransGen Biotech) and sequenced. Five sequences similar to ‘front-end’ Δ⁶-desaturase or Δ⁸-sphingolipid desaturase genes were obtained and named RnD8A, RnD8B, RnD6C, RnD6D, and RnD6E.

Yeast expression plasmid construction and transformation

The five sequences similar to ‘front-end’ Δ⁶-desaturase or Δ⁸-sphingolipid desaturase genes were purified from 1% agarose gel, digested with the restriction enzymes HindIII and SacI (the restriction sites of the enzymes had been designed in the primers), cloned into the corresponding sites of the yeast expression vector pYES2 (Invitrogen, UK), and sequenced. The five construct plasmids (pRnD8A, pRnD8B, pRnD6C, pRnD6D, and pRnD6E) were transformed into S. cerevisiae strain INV Sc 1 (Invitrogen) by using the lithium acetate method.

Characterization of genes with a Δ⁸-sphingolipid desaturase function in S. cerevisiae

The RnD8A/B-transformed yeasts were grown in SC-U, containing 1% raffinose and 1% (w/v) tergitol NP-40 (Sigma, Germany), and induced with 2% (w/v) galactose. The yeast cells were cultivated for an additional 72 h at 20 °C, and then harvested and dried at 50 °C. 100 mg dried cells were used to prepare the long chain bases (LCBs) from glycerolipids for subsequent analysis by HPLC as previously described (Markham et al., 2006) and phytosphinganine (t18:0) (Sigma) was used as the standard. Briefly, induced yeast cells (100 mg, dried weight) were grounded into a fine powder, and subjected to strong alkaline hydrolysis in 2 ml of 10% (w/v) aqueous Ba(OH)₂ and 2 ml of dioxane for 16 h at 110 °C. After hydrolysis, 2 ml of 2% (w/v) ammonium sulphate was added, and the liberated sphingolipid long chain bases were extracted with 2 ml of diethylether. The upper phase was removed to a second tube, dried under nitrogen, and derivatized with o-phthaldialdehyde (OPA) (Invitrogen) as previously described (Merrill et al., 2000). Individual LCBs were separated by reversed phase HPLC (Waters 600E, USA) using Penomenex C18 column (250×4.6 mm, 5 μm). Elution was carried out at 1.5 ml min⁻¹ with 20% solvent RA (5 mM potassium phosphate, pH 7), 80% solvent RB (100% methanol) for 7 min, increasing to 90% solvent RB by 15 min, returning to 80% solvent RB and re-equilibrating for 2 min (Markham et al., 2006). Fluorescence was excited at 340 nm and detected at 455 nm.

Characterization of genes with Δ⁶-desaturase function in S. cerevisiae

Heterologous expression of RnD6C/D/E was induced under transcriptional control of the yeast GAL1 promoter. Yeast cultures were grown to logarithmic phase at 30 °C in SC-U containing 1% (w/v) raffinose, 0.67% (w/v) yeast nitrogen and 0.1% (w/v) tergitol NP-40 (sigma), supplemented with 0.5 mM LA (Sigma). The cells were induced by the addition of 2% (w/v) galactose and cultivated for an additional 72 h at 20 °C. Subsequently, cells were harvested by centrifugation, and washed three times with sterile distilled water. The cells were dried and ground into a fine powder. Then the fatty acid analysis was performed using the following method.

Plant plasmid construction and transformation of Arabidopsis thaliana

RnD6C/D/E cDNA under the control of the duplicated CaMV 35S (D35S) promoter and nos terminator were constructed in the binary vector pGI0029 (John Innes Centre, UK). The inserts were sequenced to ensure that undesired mutations were not introduced during the PCR reactions. Agrobacterium tumefaciens strain LBA4404 containing the helper plasmid pSoup (John Innes Centre, UK) was used as the host of the pGI0029 vectors containing the Δ⁶-desaturase gene. Arabidopsis thaliana (ecotype Columbia) was used for plant transformation with the floral dip method (Clough and Bent, 1998). T₁ seeds were plated out on selective media containing kanamycin (50 mg l⁻¹), and then the selected transformed plants were transferred to soil. The harvested seeds were selected for another two generations, and the fixed transformants were selected.

Fatty acid analysis

Cellular fatty acid was extracted by incubating 50 mg yeast powder or 200 mg fresh leaves of transformed A. thaliana or different blackcurrant tissues (young root, young stem, leaves, flowers, ripening exocarp, and ripening seeds) in 3 ml of 7.5% (w/v) KOH in methanol for saponification at 70 °C for 3 h. After the pH was adjusted to 2.0 with HCl, the fatty acid was subjected to methyl-esterification with 2 ml 14% (w/v) boron trifluoride in methanol at 70 °C for 1.5 h. Then 1 ml of 0.9% (w/v) NaCl was added and mixed well. Subsequently, fatty acid methyl esters (FAME) were extracted with 2 ml hexane. The upper phase was removed to a second tube, dried under nitrogen and dissolved in acetic ether. Qualitative analysis of FAME was performed by GC-MS (gas chromatography-mass spectrometry, TurboMass, PerkinElmer, USA) with the capillary column (BPX-70, 30 m×0.25 mm×0.25 μm). FAME were analysed and identified by the comparison of their peaks with those of standards: LA, ALA, and GLA (Sigma).

Subcellular localization of Δ⁶-desaturase in yeast

To determine its subcelluar location, RnD6C/D/E had been cloned in-frame at the 5′ end and 3′ end with the GFP gene in the pYES2 vector and expressed in the yeast strain INV Sc 1. Live yeast cells were incubated with MitoTracker^Red and/or ER-tracker^Blue as described in experimental protocol of ‘MitoTracker^® and MitoFluor™ Mitochondrion-Selective Probes’ and ‘ER-Tracker™ Dyes for Live-Cell Endoplasmic Reticulum Labeling’ (Invitrogen). The localization of RnD6C/D/E-GFP was determined by using a fluorescence microscope (Olympus, Tokyo, Japan). The desaturation function of the fused proteins was identified by GC-MS.

Subcellular localization of Δ⁶-desaturase in tobacco

Wild-type Nicotiana benthamiana was grown as previously described (Ruiz et al., 1998; Voinnet et al., 1998). A. tumefaciens strain LBA4404 with plastids containing RnD6C/D/E-GFP or GFP-RnD6C/D/E and the strain carrying the p19 construct (Voinnet et al.., 2003) were grown at 28 °C in YEB supplemented with 50 mg l⁻¹ kanamycin and 5 mg l⁻¹ tetracycline to the stationary phase. Bacterial cells were precipitated by centrifugation at 5000 g for 15 min at room temperature and resuspended in 10 mM MgCl₂ and 150 mg l⁻¹ acetosyringone. Cells were left in this medium for 3 h and then infiltrated into the abaxial air spaces of 2–4-week-old N. benthamiana plants. The culture of Agrobacterium was brought to an optical density of 1.0 (OD₆₀₀) to avoid toxicity (Voinnet et al., 2000). Transient co-expression of the RnD6C/D/E-GFP or GFP-RnD6C/D/E and the mCherry-tagged ER marker (Nelson et al., 2007) and p19 constructs was at OD600 1.0 (Voinnet et al., 2003). The localization of RnD6C/D/E tagged with GFP was determined under a confocal microscope (Leica TCS SP5, Leica Microsystems, Bensheim, Germany).

Gene expression analysis

Total RNA from different tissues was extracted using a RNA extraction kit (Biomed-tech, Beijing, China). cDNA was synthesized from total RNA using a reverse transcriptional kit (Toyobo, Osaka, Japan). The RT-PCR primers for detecting our cloned genes from blackcurrant were listed in Table 1. The housekeeping actin gene was used as the control in the RT-PCR reactions. The reaction was denatured at 95 °C for 3 min, followed by 30 cycles of 30 s at 95 °C, 30 s at 60 °C, 1 min at 72 °C, and then 7 min at 72 °C.

Table 1.

Primers (from 5′ to 3′) used in RT-PCR

Primer name	Sequences
RTDL94	GGTAAGGTTTACAATGTCACCC
RTDR830	CGCTTTGAGAGCAACATG
RTA&BL420	ATG(C/T)(T/G)TTGTGGAGGCTTA
RTA&BR1077	AGCATGAAATATCTATAGTCCCTT
RTC&EL322	GAAAGGGCATGGTGTTTTC
RTC&ER864	CAA(A/G)AGTCTGTTGGGTACTTTTTT
Actin L	CATCAGGAAGGACTTGTACGG
Actin R	GATGGACCTGACTCGTCATAC

Cladistic analysis

Nucleotide sequences of the desaturases were aligned by using the program ClustalX 2.0.10 (Thompson et al., 1994) and manual adjustment. The phylogenetic tree was generated using the alignment output based on the Minimum evolution method (Rzhetsky and Nei, 1992), as implemented in the Mega 4.1 (Kumar et al., 2001). The complete deletion option and the Jukes and Cantor metric were used, and confidence of the tree branches was checked by bootstrap generated from 1000 replicates.

Results

Fatty acid composition in different blackcurrant tissues

The fatty acid compositions from young roots and stems, flowers, mature leaves, ripening exocarp, and seeds of blackcurrant were determined by GC-MS. The results showed that all the detected tissues except ripening exocarp contained GLA and OTA. There was very little GLA (<1.1% of total fatty acid) and OTA (<0.4% of total fatty acid) in young roots and stems, flowers and mature leaves, but abundant GLA (>12% of total fatty acid) and OTA (>2.6% of total fatty acid) in ripening seeds (Table 2).

Table 2.

Fatty acid composition of different blackcurrant tissues

	Young root	Young stem	Leaf	Flower	Ripening exocarp	Ripening seeds
16:0	32.3±0.7	30.4±0.7	17.2±0.6	34.7±1.5	35.0±0.2	10.1±0.0
16:1	23.4±0.9	–	1.7±0.0	–	–	–
16:2	–	–	1.2±0.0	–	–	–
16:3	–	–	5.0±0.2	–	–	–
18:0	14.3±0.6	11.2±0.8	2.5±0.3	11.1±0.6	19.5±1.0	3.9±0.0
18:1	6.5±0.2	2.9±0.3	1.7±0.1	9.9±0.5	4.6±0.3	12.6±0.2
LA	16.7±0.3	30.8±1.2	10.7±0.4	24.6±1.3	22.2±0.8	42.7±0.1
GLA	0.8±0.0	1.1±0.0	0.4±0.0	0.3±0.1	–	12.7±0.2
ALA	3.6±0.0	20.7±0.5	57.1±1.3	16.9±1.2	16.7±0.7	13.4±0.1
OTA	0.4±0.0	0.9±0.0	0.4±0.0	0.3±0.0	–	2.6±0.1

The levels of individual fatty acids were determined by GC-MS and expressed as mole% of the total. Values given represent means ±SE obtained from three independent experiments.

Isolation of putative Δ⁶-desaturase genes

Using the genomic DNA walking method, five DNA sequences (1347 bp in length) were cloned and sequenced. They were very similar in DNA sequences (>85% identity) and encoded five distinct deduced polypeptides of 448 amino acids. All of them had an N-terminal Cyt b5 domain as the heme-binding region and three conserved histidine boxes which exist in all membrane-bound desaturases (Fig. 1) (Shanklin et al., 1994). These five DNA sequences could be grouped into two classes. One included two sequences that showed high identities (>69%) to the DNA sequence of Δ⁸-sphingolipid desaturase from borage (Sperling et al., 1998). They were designated as RnD8A and RnD8B and registered in GenBank as GU198924 and GU198925, respectively. The other contained three sequences showing slightly lower identities (>60%) to the DNA sequence of Δ⁶-desaturase from borage. They were designated as RnD6C, RnD6D, and RnD6E, which were registered in GenBank as GU198926, GU198927, and GU198928, respectively. The deduced amino acid sequences of the RnD6C, RnD6D, and RnD6E proteins shared a 93% identity to each other and a 70% identity to the sequence of Δ⁶-desaturase from borage. And the predicted amino acid sequences of RnD8A and RnD8B shared a 92% identity to each other and 61% to the sequence of Δ⁸-sphingolipid desaturase from borage (Sperling et al., 2001).

Fig. 1.

Alignment of deduced amino acid sequences of blackcurrant desaturases (RnD6A, RnD6B, RnD6C, RnD6D, and RnD6E), borage Δ⁶-desaturase (acc. no. AAC49700), and Δ⁸-sphingolipid desaturase (AAG43277) by using DNAMAN 6.0. Numbers on the right denote the number of amino acid residues. Residues identical for all sequences in a given position are in white text on a black background, and those identical in two of the three or similar in the three sequences are on a grey background. Characteristic regions of the N-terminal cyt b5 motif are underlined and the conserved His boxes are indicated by black dots.

Functional identification of RnD8A, RnD8B, RnD6C, RnD6D, and RnD6E in Saccharomyces cerevisiae

To validate the putative Δ⁶-desaturase activity of RnD6C, RnD6D, and RnD6E, they were expressed in S. cerevisiae. The fatty acid compositions of transformants and the control were analysed by GC-MS, respectively. In the control, no conversion into non-native products was found by feeding potential fatty acid precursors LA (Fig. 2) or ALA (data not shown) to the yeast cells transformed with empty vector pYES2. By contrast, yeast cells transformed with RnD6C, RnD6D or RnD6E could utilize two precursor forms LA (Fig. 2) and ALA (data not shown), to produce a new product GLA (Fig. 2) or OTA (data not shown), respectively. There were no significant difference in Δ⁶-desaturase activity among RnD6C, RnD6D, and RnD6E.

Fig. 2.

GC analysis of FAME of total lipids of Saccharomyces cerevisiae. (A) The total fatty acid in yeast INV Sc 1; (B) the total fatty acid in yeast INV Sc 1 with LA and GLA standard (Sigma); (C) the total fatty acid in induced yeast transformed with pYES2; (D, E, F) the total fatty acid in induced yeast transformed with RnD6C/D/E (arrows indicated the novel peak of GLA). All the GC-MS an alyses were repeated three times and representative results are shown.

To identify the functions of RnD8A and RnD8B, they were expressed in S. cerevisiae under the control of the GAL1 promoter. Besides the phytosphinganine, cis- and trans-desaturated sphingoid bases were found in the transformed yeast cells with RnD8A or RnD8B by detecting their LCBs with HPLC (Fig. 3) using the method described in the Materials and methods (Markham et al., 2006). The results indicated that RnD8A and RnD8B were Δ⁸-sphingolipid desaturases.

Fig. 3.

Formation of phytosphingenines in yeast cells by heterologous expression of a Δ⁸-sphingolipid desaturase from blackcurrant. (A) The phytosphinganine (t18:0) (Sigma) as the standard. (B) The predominating LCBs from Saccharomyces cerevisiae cells (INV Sc 1) as the control harbouring the empty vector pYES2 is phytosphinganine (t18:0). (C, D) Formation of cis-(Z) and trans-(E) of phytosphingenine (t18:1) in yeast cells expressing pRnD8A/B. LCBs from yeast cells were converted into their OPA derivatives and analysed by reversed-phase HPLC.

Functional identification of RnD6C, RnD6D, and RnD6E in Arabidopsis thaliana

The cDNAs of RnD6C, RnD6D, and RnD6E were individually expressed in Arabidopsis thaliana. The transgenic plants were verified by PCR using two primers targeting the internal sequence of the genes. The fatty acid profile of lipids from leaves of transgenic plants was determined. Figure 4 showed the fatty acid compositions in leaves of 3-week-old seedling of T₃ RnD6C transgenic plants. Compared with the wild type, the leaves of the transgenic plants had an altered fatty acid profile containing two additional peaks corresponding to GLA and OTA. There were no significant differences in the production efficiency of GLA and OTA among transgenic RnD6C, RnD6D, and RnD6E lines (data not shown).

Fig. 4.

GC analysis of FAME of total lipids of Arabidopsis thaliana leaves transformed with RnD6C. (A) The total fatty acids of wild-type A. thaliana leaves; (B) the total fatty acids of A. thaliana leaves transformed with RnD6C (arrows indicate the novel peak of GLA and OTA). All the GC-MS analyses were repeated three times and representative results are shown.

Subcellular localization of RnD6C/D/E in S. cerevisiae and tobacco

The Δ⁶-desaturase RnD6C/D/E was tagged with GFP in both the N-terminal and C-terminal, and expressed in S. cerevisiae and N. benthamiana. RnD6C/D/E led to similar results and the representative results are shown. The results showed that RnD6C/D/E was localized in mitochondria when introduced into yeast cells (Fig. 5) and ER in tobacco epidermal cells (Fig. 6).

Fig. 5.

The localization of RnD6C/D/E tagged with GFP at the C-terminal end (A–C) and N-terminal end (D–F) in yeast cells and yeast cells INV Sc 1 with pYES2 (G–H) as the negative control. (A, D, G) Fluorescent images of yeast cells that expressed GFP and the negative control under an exciter filter of 485 nm. (B, E, H) Fluorescent images of yeast cells that were stained with MItotracker^Red (Invitrogen) under an exciter filter of 510 nm. (C, F, I) The merged image. (This figure is available in colour at JXB online.)

Fig. 6.

The localization of RnD6C/D/E tagged with GFP in tobacco epidermal cells. The results shown are from transient transformed Nicotiana benthamiana epidermal cells. (A–C) RnD6C/D/E fused with GFP at the C-terminal end; (D–F) RnD6C/D/E fused with GFP at N-terminal end. (A, D) Fluorescent images of epidermal cells that expressed GFP under an exciter filter of 485 nm. (B, E) Fluorescent images of epidermal cells that expressed the endoplasmic reticulum marker ER-mCherry under an exciter filter of 510 nm. (C, F) The merged image. (This figure is available in colour at JXB online.)

The fused protein also showed Δ⁶-desaturase activity in yeast. However, RnD6C/D/E-GFP had a relatively lower level Δ⁶-desaturase activity than RnD6C/D/E and GFP-RnD6C/D/E and the representative results are shown in Fig. 7.

Fig. 7.

GC analysis of FAME of total lipids of RnD6C/D/E tagged with GFP in Saccharomyces cerevisiae. The total fatty acid in induced yeast transformed with (A) RnD6C/D/E, (B) GFP-RnD6C/D/E, and (C) RnD6C/D/E-GFP (arrows indicated the novel peak of GLA).

Expression profiles of RnD8A, RnD8B, RnD6C, RnD6D, and RnD6E

RT-PCR was used to investigate the expression profiles of RnD8A, RnD8B, RnD6C, RnD6D, and RnD6E. The sequence identity between RnD8A and RnD8B is 91.6%, and that between RnD6C and RnD6E is 97.1%. Thus, the expression of RnD8A and RnD8B, and that of RnD6C and RnD6E, were investigated together. The results showed that RnD8A&RnD8B and RnD6C&RnD6E were expressed in all the tissues at a low level, while RnD6D was expressed only in seeds at a high level (Fig. 8). Using genomic walking, a seed specific expression promoter upstream of RnD6D has been cloned and sequenced (data not shown).

Fig. 8.

Expression profiles of Δ⁶-desaturases and Δ⁸-sphingolipid desaturases by RT-PCR in different tissues. (A) Expression levels in root, stem, leaf, flower, exocarp, and seed and relative transcript levels (normalized to actin mRNA) were determined by reverse transcriptase (RT)-PCR. (B) Seed RNA from different ripening stages: green (G), green and black (G&B), black and green (B&G), black (B) and relative transcript levels (normalized to actin mRNA) determined by reverse transcriptase (RT)-PCR. All the RT-PCR experiments were repeated three times, and representative results are shown. The actin gene for actin RNA was amplified as a control. The PCR products were separated by 2% (w/v) agarose gel electrophoresis and visualized by ethidium bromide staining.

Discussion

In a few previously studied higher plants, the profiles of the GLA and OTA accumulation were different. GLA is accumulated in mature seeds and in other tissues in Borago officinalis at a high level (Sayanova et al., 1999c), but was absent in seeds and was at a high level in leaves in Anemone leveillei (Whitney et al., 2003). In some primula species, GLA and OTA were accumulated mainly in leaves and seeds, and were relatively low in roots (Sayanova et al., 1999a). In blackcurrant, mature seeds (Traifler et al., 1988) and ripening seeds accumulate GLA and OTA at a high level, while other tissues contained very little GLA and OTA.

In Borago officinalis, Anemone leveillei, and some species of primula with high GLA, a single gene encoding Δ⁶-desaturase had been identified and characterized. In this research on blackcurrant, three genes encoding Δ⁶-desaturase (namely RnD6C, RnD6D, and RnD6E) had been cloned. Their desaturase activities were similar in yeast and Arabidopsis. However, their expression profiles were different. RnD6C&RnD6E was expressed in all the tissues at a low level, while RnD6D expressed only in seeds. RnD6D may mainly contribute to the accumulation of GLA and OTA in the seeds of blackcurrant.

It is widely believed that the Δ⁶-desaturases, which were only found in some plants, have evolved from Δ⁸-sphingolipid desaturases that existed in nearly all plants (Libisch et al., 2000; Sperling et al., 2003). Cladistic analysis of Δ⁶-desaturases and Δ⁸-sphingolipid desaturases genes (Fig. 9) revealed that Δ⁶-desaturases and Δ⁸-sphingolipid desaturase genes from the same plants clustered together, rather than being separated into two groups, for example, Ribes nigrum and Anemone lendsquerelli. The Δ⁶-desaturases and Δ⁸-sphingolipid desaturases genes of Primula vialli (Pv), and Primula farinose (Pf) clustered in two groups as the two species belong to the same genus. The only exception is Borago officinalis in which the genes were not clustered together due to the greater difference between the two desaturases than that in other plant species. The sequence identity of Δ⁶-desaturase and Δ⁸-sphingolipid desaturase in Borago officinalis is 58%, suggesting that the Δ⁶-desaturase and Δ⁸-sphingolipid desaturase separated much earlier in Borago officinalis than those in other species. The identity of Δ⁶-desaturases and Δ⁸-sphingolipid desaturases is more than 80% in blackcurrant and the distance between them is much shorter than that in other species. Furthermore, among the plant species with reported Δ⁶-desaturase genes, blackcurrant is the only one in which several functional genes or copies of Δ⁶-desaturase have been found. These genes have identities higher than 92% based on the amino acid sequences. The existence of three such genes may have resulted from a recent gene duplication event, as the delta desaturases were thought to be conserved in evolution until now (Libisch et al., 2000). It is speculated that Δ⁶-desaturase genes in blackcurrant may have evolved later than those in other species.

Fig. 9.

Phylogenetic analysis of Δ⁶-desaturases and Δ⁸-sphingolipid desaturase genes (ORF region) based on the multiple nucleotide sequence alignment program ClustalX and the Minimum evolution method. Organism abbreviations stand for: Borago officinalis (Bo), Anemone lendsquerelli (Al), Camellia sinensis (Cs),Nicotiana tabacum (Nt), Echium gentianoides (Eg), Arabidopsis thaliana (At), Brassica napus (Bn), Primula vialli (Pv), Primula farinosa (Pf). GenBank accession numbers are: BoD6/8 (U79010/AF133728), AlD6/8 (AF536525/AF536526), CsD6 (AY169402), NtD8 (EF110692), EgD6 (AY055117), AtD8 (NM_116023), BnD8 (AJ224160), PvD6/8 (AY234127/ AY234126), PfD6/8 (AAp23034/AY234124), RnD8A/B (GU198924/GU198925), and RnD6C/D/E (GU198926/GU198927/GU198928). Bootstrap values at nodes represent the percentages of times a clade appeared in 1000 replicates.

Δ⁶-desaturases from previously reported plants always exhibited different substrate selectivity. Some prefer LA (García-Maroto et al., 2006; Sayanova et al., 2006) and some prefer ALA as the substrate (Sayanova et al., 2003). In the present study, the three Δ⁶-desaturases, RnD6C, RnD6E, and RnD6D from blackcurrant can use both LA and ALA as the substrates, but their substrate selectivity needs to be confirmed by future studies.

Front-end desaturases are generally thought to locate in the ER, although the actual location is unknown (Hsiao et al., 2007). Expression of GFP-tagged RnD6C/D/E in yeast and tobacco could facilitate the visualization of its exact location. The present study provides the unequivocal evidence that RnD6C/D/E is located in the ER in tobacco, but it is located in mitochondria when introduced into yeast.

The effect of the fusion of GFP with Δ⁶-desaturases on their enzyme activity has not been reported. Results from the present study indicate that the yeast cells containing RnD6C/D/E-GFP accumulated a lower level of GLA than those containing GFP-RnD6C/D/E or RnD6C/D/E and there were no significant differences between GFP-RnD6C/D/E and RnD6C/D/E (Fig. 7). It is shown that the GFP tagged with Δ⁶-desaturases in the N-terminal (GFP-RnD6C/D/E) had no effect on the activity of Δ⁶-desaturases, but tagged with C-terminal (RnD6C/D/E-GFP) could decrease the activity of Δ⁶-desaturases. This reduced activity may result from the alteration of the three-dimensional structure of the C-terminal when the protein is tagged with GFP at that end. The results suggested that the three-dimensional structure of the C-terminal of Δ⁶-desaturases should be important for their enzyme activity. However, the effect of the three-dimensional structure of the C-terminal of Δ⁶-desaturases on their enzyme activity is worthy of further investigation.

In conclusion, three genes encoding Δ⁶-desaturases (RnD6C/D/E) and two encoding Δ⁸-sphingolipid desaturases (RnD8A/B) from blackcurrant that were very similar in their sequences have been isolated and characterized. RnD6C and RnD6E was expressed in all the tissues at a low level, while RnD6D was expressed only in seeds. RnD6D could mainly contribute to the accumulation of GLA and OTA in blackcurrant seeds. To the authors’ knowledge, this is the first report of the identification of several Δ⁶-desaturases in a single plant species and of finding a Δ⁶-desaturase gene expressed specifically in seeds. Blackcurrant Δ⁶-desaturases were confirmed to localize in the ER of plant cells and in the mitochondria of yeast cells. It is speculated that the evolution of Δ⁶-desaturase genes from Δ⁸-sphingolipid desaturase genes in blackcurrant occurred more recently than that in other species.

Acknowledgments

We thank Dr Chengcai Chu and Dr Yiqin Wang, Institute of Genetics and Developmental Biology, CAS, for their help with the HPLC analyses. This research was supported by project (No. 2008ZX08009-003) from the Ministry of Agriculture of China for transgenic research, project KSCX-YW-N-009 from the Chinese Academy of Sciences, and project 2006AA10A113 from the Ministry of Science and Technology of China.

Abbreviations

GLAγ-linolenic acid

EFAessential fatty acid

ALAα-linolenic acids

LAlinoleic acid

OTAoctadecatetraenoic acids

MITmitochondrion

ERendoplasmic reticulum

FAMEfatty acid methyl esters

GC-MSgas chromatography-mass spectrometry

Cyt b5

cytochrome b5

PMSpremenstrual syndrome

LCBslong chair bases

OPAo-phthaldialdehyde

GFP-Green fluorescence protein tagged in C-terminal

-GFPGreen fluorescence protein tagged in N-terminal

References

1. CloughSJBentAFFloral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thalianaThe Plant Journal199816735743[PubMed][Google Scholar]
2. DasUNEssential fatty acids: biochemistry, physiology and pathologyBiotechnology Journal20061420439[PubMed][Google Scholar]
3. García-MarotoFManas-FernandezAGarrido-CárdenasJAAlonsoDLSubstrate specificity of acyl-Δ⁶-desaturases from Continental versus Macaronesian Echium speciesPhytochemistry200667540544[PubMed][Google Scholar]
4. HorrobinDFNutritional and medical importance of gamma-linolenic acidProgress in Lipid Research199231163194[PubMed][Google Scholar]
5. HsiaoTYHolmesBBlanchHWIdentification and functional analysis of a delta-6 desaturase from the marine microalga Glossomastix chrysoplastaMarine Biotechnology20079154165[PubMed][Google Scholar]
6. KumarSTamuraKJakobsenIBNeiMMEGA2: molecular evolutionary genetics analysis software2001Arizona State University, Tempe, Arizona, USA
7. LibischBMichaelsonLVLewisMJShewryPRNapierJAChimeras of Δ⁶-fatty acid and Δ⁸-sphingolipid desaturasesBiochemical and Biophysical Research Communications2000279779785[PubMed][Google Scholar]
8. MarkhamJELiJCahoonEBJaworskiJGSeparation and identification of major plant sphingolipid classes from leavesJournal of Biological Chemistry20062812268422694[PubMed][Google Scholar]
9. MerrillAHCaliganTBWangEPetersKOuJAnalysis of sphingoid bases and sphingoid base 1-phosphates by High-Performance Liquid ChromatographyMethods in Enzymology200031239[PubMed][Google Scholar]
10. NapierJAHeySJLaceyDJShewryPRIdentification of a Caenorhabditis elegans Δ⁶-fatty-acid-desaturase by heterologous expression in Saccharomyces cerevisiaeBiochemical Journal1998330611614[PubMed][Google Scholar]
11. NapierJASayanovaOSperlingPHeinzEA growing family of cytochrome b₅-domain fusion proteinsTrends in Plant Science1999424[Google Scholar]
12. Na-RanongSLaotengKKittakoopPTantichareonMCheevadhanarakSSubstrate specificity and preference of Δ⁶-desaturase of Mucor rouxiiFEBS Letters200557927442748[PubMed][Google Scholar]
13. NelsonBKCaiXNebenführAA multicolored set of in vivo organelle markers for co-localization studies in Arabidopsis and other plantsThe Plant Journal20075111261136[PubMed][Google Scholar]
14. QiBFraserTMugfordSDobsonGSayanovaOButlerJNapierJAStobartAKLazarusCMProduction of very long chain polyunsaturated omega-3 and omega-6 fatty acids in plantsNature Biotechnology200422739745[Google Scholar]
15. ReddyASNuccioMLGrossLMThomasTLIsolation of a Δ⁶-desaturase gene from the cyanobacterium Synechocystis sp. strain PCC6803 by gain-of-function expression in Anabaena sp. strain PCC7120Plant Moleclular Biology199322293300[Google Scholar]
16. ReddyASThomasTLExpression of a cyanobacterial Δ⁶-desaturase gene results in γ-linolenic acid production in transgenic plantsNature Biotechnology199614639642[Google Scholar]
17. RuizMTVoinnetOBaulcombeDCInitiation and maintenance of virus-induced gene silencingThe Plant Cell199810937946[PubMed][Google Scholar]
18. RzhetskyANeiMA simple method for estimating and testing minimum evolution treesMolecular Biology and Evolution19929945967[Google Scholar]
19. SakuradaniEShimizuSGene cloning and functional analysis of a second Δ⁶-fatty acid desaturase from an arachidonic acid-producing Mortierella fungusBioscience, Biotechnology, and Biochemistry200367704711[Google Scholar]
20. SayanovaOBeaudoinFLibischBCastelAShewryPRNapierJAMutagenesis and heterologous expression in yeast of a plant Δ⁶-fatty acid desaturaseJournal of Experimental Botany20015215811585[PubMed][Google Scholar]
21. SayanovaOVBeaudoinFMichaelsonLVShewryPRNapierJAIdentification of primula fatty acid Δ⁶-desaturases with n-3 substrate preferencesFEBS Letters2003542100104[PubMed][Google Scholar]
22. SayanovaOHaslamRVenegas-CalerónMNapierJAIdentification of primula ‘front-end’ desaturases with distinct n-6 or n-3 substrate preferencesPlanta200622412691277[PubMed][Google Scholar]
23. SayanovaOShewryPRNapierJAHistidine-41 of the cytochrome b₅ domain of the borage Δ⁶ fatty acid desaturase is essential for enzyme activityPlant Physiology1999b121641646[PubMed][Google Scholar]
24. SayanovaOShewryPRNapierJACharacterization and expression of a fatty acid desaturase from Borago officinalisJournal of Experimental Botany1999c50411412[Google Scholar]
25. SayanovaOSmithMALanpinskasPΔ⁶-Unsaturated fatty acids in species and tissues of the PrimulaceaePhytochemistry1999a52419422[Google Scholar]
26. SayanovaOSmithMALapinskasPStobartAKDobsonGChristieWWShewryPRNapierJAExpression of a borage desaturase cDNA containing an N-terminal cytochrome b₅domain results in the accumulation of high levels of Δ⁶-desaturated fatty acids in transgenic tobaccoProceedings of the National Academy of Sciences, USA19979442114216[Google Scholar]
27. SeiliezIPanseratSKaushikSBergotPCloning, tissue distribution and nutritional regulation of a Δ⁶-desaturase-like enzyme in rainbow troutComparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology20011308393[Google Scholar]
28. ShanklinJWhittleEFoxBGEight histidine residues are catalytically essential in a membrane-associated iron enzyme, stearoyl-CoA desaturase, and are conserved in alkane hydroxylase and xylene monooxygenaseBiochemistry1994331278712794[PubMed][Google Scholar]
29. SperlingPZähringerUZHeinzEA sphingolipid desaturase from higher plantsThe Journal of Biological Chemistry19982732359028596[PubMed][Google Scholar]
30. SperlingPLeeMGirkeTZähringerUStymneSHeinzEA bifunctional Δ⁶-fatty acyl acetylenase/desaturase from the moss Ceratodon purpureusEuropean Journal of Biochemistry200026738013811[PubMed][Google Scholar]
31. SperlingPLibischBZähringerUNapierJAHeinzEFunctional identification of a Δ⁸-sphingolipid desaturase from Borago officinalisArchives of Biochemistry and Biophysics2001388293298[PubMed][Google Scholar]
32. SperlingPTernesPZankTKHeinzEThe evolution of desaturasesProstaglandins, Leukotrienes and Essential Fatty Acids2003687395[Google Scholar]
33. StymneSStobartAKBiosynthesis of γ-linolenic acid in cotyledons and microsomal preparations of the developing seeds of common borage (Borago officinalis)Biochemical Journal1986240385393[PubMed][Google Scholar]
34. SubudhiSKurdridPHongsthongASirijuntarutMCheevadhanarakSTanticharoenMIsolation and functional characterization of Spirulina D6D gene promoter: role of a putative GntR transcription factor in transcriptional regulation of D6D gene expressionBiochemical and Biophysical Research Communications2008365643649[PubMed][Google Scholar]
35. ThompsonJDHigginsDGGibsonTCCLUSTALW: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positions-specific gap penalties and weight matrix choiceNucleic Acids Research19942246734680[PubMed][Google Scholar]
36. TraiflerHWilleHJStuderAFractionation of blackcurrant seed oilJournal of the American Oil Chemists’ Society198865755760[Google Scholar]
37. VoinnetOLedererCBaulcombeDCA viral movement protein prevents systemic spread of the gene silencing signal in Nicotiana benthamianaCell2000103157167[PubMed][Google Scholar]
38. VoinnetORivasSMestrePBaulcombeDCAn enhanced transient expression system in plants based on suppression of gene silencing by the p19 protein of tomato bushy stunt virusThe Plant Journal200333949956[PubMed][Google Scholar]
39. VoinnetOVainPAngellSBaulcombeDCSystemic spread of sequence-specific transgene RNA degradation is initiated by localised introduction of ectopic promoterless DNACell199895177187[PubMed][Google Scholar]
40. WassenaarLMHofmanKDescription and identification of blackcurrant varieties by winter charactersEuphytica196615117[Google Scholar]
41. WhitneyHMMichaelsonLVSayanovaOPickettJANapierJAFunctional characterisation of two cytochrome b₅-fusion desaturases from Anemone leveillei: the unexpected identification of a fatty acid Δ⁶-desaturasePlanta2003217983992[PubMed][Google Scholar]

Identification and functional analysis of the genes encoding Δ6-desaturase from Ribes nigrum†