L E T T E R S
`
`Stepwise engineering to produce high yields of very
`long-chain polyunsaturated fatty acids in plants
`
`Guohai Wu1, Martin Truksa1, Nagamani Datla1, Patricia Vrinten1, Joerg Bauer2, Thorsten Zank3,4,
`Petra Cirpus2, Ernst Heinz3 & Xiao Qiu1
`
`Very long chain polyunsaturated fatty acids (VLCPUFAs) such
`as arachidonic acid (AA), eicosapentaenoic acid (EPA) and
`docosahexaenoic acid (DHA) are valuable commodities that
`provide important human health benefits1–5. We report the
`transgenic production of significant amounts of AA and EPA in
`Brassica juncea seeds via a stepwise metabolic engineering
`strategy. Using a series of transformations with increasing
`numbers of transgenes, we demonstrate the incremental
`production of VLCPUFAs, achieving AA levels of up to 25% and
`EPA levels of up to 15% of total seed fatty acids. Both fatty
`acids were almost exclusively found in triacylglycerols, with
`AA located preferentially at sn-2 and sn-3 positions and EPA
`distributed almost equally at all three positions. Moreover, we
`reconstituted the DHA biosynthetic pathway in plant seeds,
`demonstrating the practical feasibility of large-scale production
`of this important x-3 fatty acid in oilseed crops.
`
`Plants have the capacity to serve as green factories for the production
`of novel industrial materials, nutritionally enhanced foods or phar-
`maceuticals, via metabolic engineering6–8. One goal of plant metabolic
`engineering is the production of high levels of VLCPUFAs in oilseed
`plants9,10, which would provide a novel and cost-effective source of
`these fatty acids. Several pathways for the biosynthesis of VLCPUFAs
`exist in nature11. To produce VLCPUFAs in seeds, we followed
`the alternating desaturation/elongation pathways of n-6 and n-3
`fatty acids. The two routes commence with linoleic acid (18:2n-6,
`LA) and a-linolenic acid (18:3n-3, ALA), respectively, followed by
`sequential D6 desaturation, D6 elongation and D5 desaturation,
`leading to the synthesis of arachidonic acid (20:4n-6, AA) in the n-6
`and eicosapentaenoic (20:5n-3, EPA) in the n-3 pathway. The two
`pathways can be interconnected by a o3 desaturase that converts AA
`into EPA. Further D5 elongation and D4 desaturation reactions lead to
`the synthesis of docosapentaenoic (22:5n-3, DPA) and finally doc-
`osahexaenoic acid (22:6n-3, DHA). The B. juncea breeding line 1424
`was chosen as a host plant for biosynthesis of VLCPUFAs because of
`its high LA content and lack of erucic acid. The constructs for
`VLCPUFA production in seeds carried three to nine structural
`genes, with each gene under the control of the seed-specific napin
`promoter (Fig. 1).
`
`The first construct (BJ3) introduced into B. juncea contained a D6
`desaturase from Pythium irregulare12, a D5 desaturase from Thraus-
`tochytrium sp.13 and a D6 fatty acid elongase from Physcomitrella
`patens14. This represents the minimal set of transgenes required for the
`synthesis of AA and EPA from endogenous LA and ALA. RT-PCR
`indicated that all three genes were highly expressed in the developing
`seeds (data not shown). Several new fatty acids were detected in BJ3
`seeds (Fig. 2). The most abundant was g-linolenic acid (GLA), the D6
`desaturation product of LA, with an average value of 27.7% of total
`seed fatty acids. AA, the D5 desaturation product of dihomo-g-
`linolenic acid (20:3n-6, DGLA), ranged from 5.0% to 8.5% (average
`7.3%), whereas stearidonic acid (SDA), the D6 desaturation product
`of ALA, averaged 3.1%; several other minor new fatty acids, such as
`18:2n-9 (1.7%), were also present (Table 1). Consequently, LA content
`dropped dramatically from 45.2% in the untransformed control to
`13.7% in transgenic seeds. Thus, the D6 and D5 desaturases func-
`tioned well, with conversion rates of 68.3% and 94.2%, respectively.
`The D6 elongase performed less efficiently, with a conversion rate of
`only 23.6%. The n-6 pathway appeared to be much more effective in
`VLCPUFA biosynthesis, perhaps not surprisingly, given that B. juncea
`oil is characterized by high LA (45.2%) and low ALA (9.7%).
`To increase LA and concurrently reduce the side-product 18:2n-9,
`we added a D12 desaturase gene from Calendula officinalis15 to the
`triple construct, producing the construct BJ4. The addition of this
`desaturase resulted in only a slight decrease (0.5%) of 18:2n-9.
`Enhanced conversion of oleic acid (OA) to LA was evident from the
`decrease in OA content. Although the GLA content remained similar
`to that in BJ3 plants, the average level of AA increased from 7.3% to
`12.0%, with the highest level reaching 17.7%. EPA also increased from
`0.8% to 1.3% (Table 1).
`In view of the results from the BJ3 and BJ4 plants, where poor
`elongation from 18- to 20-carbon fatty acids limited the meta-
`bolic flux, we attempted to enhance elongation by adding a second
`D6 elongase from Thraustochytrium sp. When expressed in yeast, this
`elongase showed activity with both 18-carbon and 20-carbon fatty
`acids, but elongated GLA and SDA much more efficiently than AA and
`EPA (data not shown). In transgenic plants carrying this construct,
`named BJ5, a slight, but still significant increase in GLA elongation
`occurred. This in turn resulted in an increase in AA from an average of
`
`1Bioriginal Food & Science Corporation, 110 Gymnasium Place, Saskatoon, Saskatchewan, Canada S7N 0W9. 2BASF Plant Science GmbH, 67117 Limburgerhof,
`Germany. 3Biozentrum Klein Flottbek, Universita¨t Hamburg, 22609 Hamburg, Germany. 4Present address: BASF Plant Science GmbH, 67117 Limburgerhof, Germany.
`Correspondence should be addressed to X.Q. (xqiu@bioriginal.com).
`
`Published online 12 June 2005; doi:10.1038/nbt1107
`
`©2005 Nature Publishing Group http://www.nature.com/naturebiotechnology
`
`NATURE BIOTECHNOLOGY VOLUME 23 NUMBER 8 AUGUST 2005
`
`CSIRO Exhibit 1008
`
`1 0 1 3
`
`

`

`L E T T E R S
`
`BJ3
`
`NP
`
`Pi∆6
`
`T
`
`NP
`
`Tc∆5
`
`T
`
`NP
`
`PSE1
`
`T
`
`BJ4
`
`BJ5
`
`BJ6
`
`NP
`
`Pi∆6
`
`T NP
`
`Tc∆5
`
`T
`
`NP
`
`PSE1
`
`T
`
`NP
`
`Co∆12
`
`T
`
`NP
`
`Pi∆6
`
`T NP
`
`Tc∆5
`
`T
`
`NP
`
`PSE1
`
`T
`
`NP
`
`Co∆12
`
`T
`
`NP
`
`TcElo
`
`T
`
`NP
`
`Pi∆6
`
`T NP
`
`Tc∆5
`
`T
`
`NP
`
`PSE1
`
`T
`
`NP
`
`Co∆12
`
`T
`
`NP
`
`TcElo
`
`T
`
`NP
`
`Piω3
`
`T
`
`Pi∆6
`
`Tc∆5
`
`Figure 1 Simplified maps of the binary vector constructs used for plant
`transformation. BJ3, BJ4, BJ5, BJ6 and BJ9 represent the three-, four-,
`five-, six- and nine-gene constructs used for B. juncea transformation.
`PiD6, TcD5, PSE1, CoD12, TcElo, Pio3, TcD4, TcAt and OmElo represent a
`D6 desaturase from P. irregulare, a D5 desaturase from Thraustochytrium sp.
`26185, an elongase from P. patens, a D12 desaturase from
`C. officinalis, an elongase from Thraustochytrium sp. 26185, an o3
`desaturase from P. irregulare, a D4 desaturase from Thraustochytrium sp.
`26185, a lysophosphatidyl acyltransferase from Thraustochytrium sp.
`26185 and an elongase from O. mykiss, respectively. NP, napin promoter;
`T, terminator OCS.
`
`instance, in transgenic B. juncea GLA accumulated to a relatively stable
`level (27.1 to 29.4%) regardless of the construct used. These GLA
`molecules could be incorporated into certain lipid classes, such as
`triacylglycerols, where they become unavailable for further modifica-
`tion. Nonelongated GLA seems to remain at a constant level, with
`amounts exceeding this threshold apparently becoming available for
`elongation. This is best illustrated in BJ4 plants, where the addition of a
`D12 desaturase led to a 7.8% decrease in OA, and the consequent
`increase in substrate flux led to a 5.9% increase of 20-carbon VLCPU-
`FAs, rather than in the further accumulation of 18-carbon fatty acids.
`
`WT
`
`BJ3
`
`BJ4
`
`BJ5
`
`BJ6
`
`18:3n-3
`
`18:2n-6
`
`18:1n-9
`
`GLA
`
`SDA
`
`AA
`
`EPA
`
`Response
`
`DPAn-3 + 24:0
`
`DHA
`
`BJ9
`
`24:1n-9
`
`DPAn-6
`
`8
`
`10
`
`12
`
`14
`Time (min)
`
`16
`
`18
`
`20
`
`NP
`
`T NP
`
`T
`
`NP
`
`PSE1
`
`T
`
`NP
`
`Co∆12
`
`T
`
`NP
`
`TcElo
`
`T NP
`
`Piω3
`
`T
`
`BJ9
`
`NP
`
`Tc∆4
`
`T
`
`NP
`
`TcAt
`
`T
`
`NP
`
`OmElo
`
`T
`
`12.0% in BJ4 to 13.7% in BJ5 seeds, with the highest value observed
`being 25.8% (Table 1). Elongation of n-3 fatty acids also increased
`slightly, such that the overall elongation rate of both pathways
`increased from 34.0% in BJ4 to 38.3% in BJ5 seeds.
`The high metabolic flux via the n-6 pathway in B. juncea resulted in
`the accumulation of considerable amounts of n-6 fatty acids such as
`GLA and AA. To use these n-6 fatty acids for the production of n-3
`fatty acids, we included a o3 desaturase from Phytophtora infestans in
`the construct BJ6. In yeast, this desaturase introduced an n-3 double
`bond specifically into AA (data not shown). The o3 desaturase also
`effectively converted AA into EPA in transgenic seeds. As a result, the
`EPA content increased significantly, from an average of 1.4% in BJ5 to
`8.1% in BJ6 plants, with a concurrent decrease in AA (Table 1).
`After achieving substantial production of AA and EPA in plant
`seeds, the next logical step was to produce DHA. Therefore, three
`more genes were added to BJ6, creating the nine-gene construct BJ9.
`One of these genes encodes an elongase from Oncorhynchus mykiss
`that can elongate both 18- and 20-carbon fatty acids in yeast16,
`whereas the second gene encodes a D4 desaturase from Thraustochy-
`trium sp.13. The third gene, also from Thraustochytrium sp., represents
`a putative lysophosphatidic acid acyltransferase. We reasoned that this
`enzyme from a VLCPUFA-rich organism might improve trafficking of
`very long chain fatty acyls among lipid pools. Transcripts from all nine
`genes were detected in transgenic plants (data not shown). A fatty acid
`with a retention time identical to that of DHA was present in BJ9 seeds
`(Fig. 2), and gas chromatography/mass spectrometry (GC/MS) ana-
`lysis confirmed that this fatty acid was indeed DHA (data not shown).
`The average yield of DHA was 0.2% of total fatty acids, whereas the
`highest observed value was 1.5%. BJ9 transgenic plants also produced
`slightly higher levels of EPA, with the highest observed level reaching
`15.0% of total fatty acids. Whether the lysophosphatidyl acyltrans-
`ferase or the third elongase contributes to the improvement in EPA
`production remains to be determined.
`Elongation of EPA appears to be a serious bottleneck in DHA
`synthesis. The elongation rate from EPA to DPA was only 4%;
`consequently, only a low level of DHA was produced in BJ9 seeds.
`This might be due to limitations in the host plant’s ability to release
`EPA into the acyl-CoA pool. Alternatively, the heterologous elongase
`may not cooperate efficiently with the endogenous elongation complex.
`It should be noted that the fatty acid profiles and the derived
`conversion rates reflect only the situation in the whole cell, and
`disregard the possible existence of unavailable fatty acid pools. For
`
`©2005 Nature Publishing Group http://www.nature.com/naturebiotechnology
`
`Figure 2 GC analysis of seed fatty acid methyl esters from wild-type and
`transgenic B. juncea plants. The constructs used for transformation are
`described in Figure 1. GLA, g-linolenic acid; SDA, stearidonic acid;
`AA, arachidonic acid; EPA, eicosapentaenoic acid; DHA, docosahexaenoic
`acid; DPA, docosapentaenoic.
`
`1 0 1 4
`
`VOLUME 23 NUMBER 8 AUGUST 2005 NATURE BIOTECHNOLOGY
`
`CSIRO Exhibit 1008
`
`

`

`Table 1 Total fatty acid composition of oilseeds from the wild-type and transgenic B. juncea plants (wt%)
`Wild type (n ¼ 14) BJ3 (N ¼ 10; n ¼ 19)
`BJ4 (N ¼ 7; n ¼ 20)
`BJ5 (N ¼ 4; n ¼ 28)
`BJ6 (N ¼ 3; n ¼ 12)
`
`Fatty acid
`
`16:0
`18:0
`18:1n-9 (OA)
`18:1n-7
`18:2n-9
`18:2n-6 (LA)
`18:3n-6 (GLA)
`18:3n-3 (ALA)
`18:4n-3 (SDA)
`20:3n-6 (DGLA)
`20:4n-6 (AA)
`20:4n-3 (ETA)
`20:5n-3 (EPA)
`22:5n-3 (DPA)
`22:6n-3 (DHA)
`Other
`
`5.6 7 0.2
`1.7 7 0.1
`33.2 7 0.7
`3.4 7 0.1
`
`45.2 7 0.5
`
`9.7 7 0.2
`
`5.7 7 0.1
`5.8 7 0.1
`5.7 7 0.2
`2.5 7 0.1
`2.2 7 0.1
`2.1 7 0.1
`15.8 7 0.9
`18.7 7 0.5
`26.5 7 0.4
`3.3 7 0.1
`3.2 7 0.1
`3.5 7 0.1
`0.5 7 0.1
`1.2 7 0.2
`1.7 7 0.2
`15.4 7 0.7
`14.2 7 0.4
`13.7 7 0.3
`28.6 7 0.8
`29.4 7 0.9
`27.7 7 0.5
`3.1 7 0.2
`4.1 7 0.1
`4.6 7 0.3
`2.2 7 0.1
`2.7 7 0.1
`3.1 7 0.1
`2.1 7 0.1
`1.2 7 0.1
`0.5 7 0.1
`7.3 7 0.4 (5.0–8.5) 12.0 7 0.3 (8.2–17.7) 13.7 7 0.7 (8.8–25.8)
`0.5 7 0.1
`1.4 7 0.1 (0.9–2.7)
`
`0.8 7 0.1 (0.1–1.1)
`
`1.3 7 0.2 (0.9–1.7)
`
`5.6 7 0.1
`2.4 7 0.1
`18.4 7 1.0
`2.8 7 0.10
`0.2 7 0.04
`16.9 7 0.5
`27.1 7 0.8
`4.2 7 0.5
`2.4 7 0.1
`2.0 7 0.1
`5.4 7 0.3 (4.0–7.4)
`0.8 7 0.1
`8.1 7 0.4 (6.4–11.0)
`
`1.2 7 0.2
`
`2.8 7 0.3
`
`4.0 7 0.2
`
`5.2 7 0.4
`
`3.7 7 0.4
`
`L E T T E R S
`
`BJ9 (N ¼ 8; n ¼ 30)
`
`5.1 7 0.2
`3.0 7 0.2
`18.9 7 1.0
`2.4 7 0.1
`0.2 7 0.1
`16.0 7 0.5
`27.3 7 0.7
`3.0 7 0.1
`2.2 7 0.1
`1.9 7 0.1
`4.0 7 0.2 (2.0–7.3)
`1.1 7 0.1
`8.1 7 0.4 (2.8–15.0)
`0.1 7 0.02
`0.2 7 0.03 (0–1.5)
`6.5 7 0.4
`
`BJ3, BJ4, BJ5, BJ6 and BJ9 represent the transgenic plants generated from the three-, four-, five-, six- and nine-gene constructs. N and n are the number of independent transgenic plants and the
`total number of positive seeds analyzed, respectively. Each value represents the mean 7 s.e.m. The values in brackets indicate the fatty acid ranges for AA, EPA and DHA. Other fatty acids include
`14:0, 16:1n-7, 20:0, 20:1n-9, 20:2n-6, 22:0, 24:0 and others.
`
`Transgenic BJ9 plants were phenotypically normal, as were plants
`carrying other constructs, and seeds containing substantial amounts of
`VLCPUFAs showed no obvious germination problems (data not
`shown). Thus, transgenic B. juncea seeds might be able to use the
`newly synthesized VLCPUFAs to support initial growth.
`Although overall elongation rates in transgenic B. juncea appear
`low, they are actually much higher than those observed in flax and
`tobacco10. Accordingly,
`transgenic B.
`juncea accumulated higher
`amounts of VLCPUFAs in seeds. This could be due to the contribution
`of endogenous enzymes in B. juncea that are directly involved in the
`elongation process. The elongation complex includes four enzymes17,
`and although the condensing enzyme (elongase) is critical in deter-
`mining the substrate specificity of the elongation process, the remain-
`ing three enzymes may also play important roles in the overall
`elongation efficiency. Whereas erucic acid represents approximately
`45% of total seed fatty acids in traditional B. juncea lines, the line used
`here contains only a trace amount of erucic acid. This might be due to
`a mutation in a particular condensing enzyme, as was observed in low
`erucic acid Brassica napus18. The remaining three enzymes in the
`elongation complex may thus be free to co-act with the transgenic
`elongase, resulting in higher elongation rates than in flax or tobacco.
`Alternatively, in B. juncea the shuffling of fatty acyls between the
`phospholid and acyl-CoA pools during the biosynthesis of VLCPUFAs
`may be more efficient. Indeed, in transgenic flax and tobacco, the low
`rate of elongation of D6-desaturated 18-carbon fatty acids to their
`20-carbon counterparts was
`associated with a
`low level of
`D6-desaturated 18-carbon fatty acids in the acyl-CoA pool10.
`The VLCPUFAs produced in B. juncea seeds are almost exclusively
`present as
`triacylglycerols. For example in BJ4 and BJ9 seed,
`93.9–98.6% of total AA and 96.0–98.1% of total EPA was found in
`triacylglycerols. Other lipid classes contain only very small amounts of
`these fatty acids. This is not unexpected, given the high levels to which
`AA and EPA accumulated in the seeds.
`Positional analysis of BJ4 and BJ9 phospholipids (Fig. 3a) showed
`that GLA and AA were mainly located at the sn-2 position, which
`appears to be the main site of D5 and D6 desaturation19. However,
`EPA, which is also a D5-desaturation product, was not predominantly
`located at the sn-2 position. For most fatty acids, distribution at the
`
`sn-1 and sn-2 positions of triacylglycerol reflected the distribution
`pattern in phospholipids (Fig. 3b). The majority of AA was located at
`the sn-2 and sn-3 positions and EPA was almost equally distributed at
`all three positions of triacylglycerol. Preliminary analyses of developing
`BJ4 seeds (30 days after flowering) indicated that the fatty acid
`compositions of phospholipids and triacylglycerols were similar to
`those in mature seeds (data not shown). This suggests that the relative
`AA content in phospholipids is not dramatically reduced during seed
`desiccation, in contrast to what has been observed for medium-chain
`fatty acids20,21.
`A recent study reconstituted the conventional D6 desaturase/elon-
`gase pathway in seeds of transgenic flax and tobacco10. While the
`transgenic seeds accumulated high levels of D6 desaturated fatty acid,
`amounts of AA and EPA were only in the range of 1–2%. In work with
`Arabidopsis thaliana9, somewhat higher levels of AA and EPA accu-
`mulated in leaves of plants carrying genes from the alternative D9
`elongase/D8 desaturase pathway. This suggested that using the alte-
`rnative pathway might overcome the problems associated with rela-
`tively poor elongation rates in the conventional pathway9,22,23. In fact,
`our preliminary experiments showed that the constitutive expression
`of the D6 elongase pathway in B. juncea resulted in even higher EPA
`levels in leaves than those observed in seeds (data not shown),
`implying that the particular host plant and targeted tissue have
`major effects on the efficiency of individual systems.
`The synthesis of VLCPUFAs in plant seeds is an intricate biochem-
`ical process, requiring the sequential activity of multiple transgenic
`enzymes. The use of stepwise metabolic engineering provides an
`opportunity to observe the effects of individual genes in the biosyn-
`thetic pathway and offers insights into the intermediate steps of this
`complex process. Using this stepwise engineering, we increased the AA
`from 8.5% in BJ3 plants to 25.8% in BJ5 plants (maximum observed
`levels). The highest EPA level observed in BJ3 seeds was 1.1%, which
`increased to 15.0% in BJ9 plants. Since these measurements were
`taken from segregating populations,
`the highest observed value
`demonstrates the potential of lines to produce specific fatty acids.
`The cloning of the first D4 desaturase from Thraustochytrium sp.13
`implied the existence of a simple pathway for DHA biosynthesis and
`suggested the possibility of producing this important fatty acid in
`
`NATURE BIOTECHNOLOGY VOLUME 23 NUMBER 8 AUGUST 2005
`
`CSIRO Exhibit 1008
`
`1 0 1 5
`
`©2005 Nature Publishing Group http://www.nature.com/naturebiotechnology
`
`

`

`TAG
`
`sn-1
`
`sn-2
`
`sn-3
`
`BJ4
`
`BJ9
`
`50
`
`40
`
`30
`
`20
`
`10
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`Mol (%)
`
`Mol (%)
`
`b
`
`PC sn-1
`PC sn-2
`
`PE mol (%)
`
`0 1
`
`0
`
`20
`
`30
`
`40
`
`PE sn-1
`PE sn-2
`
`BJ4
`
`BJ9
`
`L E T T E R S
`
`a
`
`30
`
`20
`
`10
`
`PC mol (%)
`
`0
`
`22:6n-3
`
`22:5n-3
`
`20:5n-3
`
`20:4n-3
`
`20:3n-3
`
`20:4n-6
`
`20:3n-6
`
`18:4n-3
`
`18:3n-3
`
`18:3n-6
`
`18:2n-6
`
`18:1n-9
`
`18:0
`
`16:0
`
`Fatty acid
`
`20:6n-3
`20:5n-3
`20:4n-3
`20:3n-6
`20:4n-6
`20:3n-6
`18:4n-3
`18:3n-3
`18:3n-6
`18:2n-6
`18:1n-9
`18:0
`16:0
`20:5n-3
`20:4n-3
`20:4n-6
`20:3n-6
`18:4n-3
`18:3n-3
`18:3n-6
`18:2n-6
`18:2n-9
`18:1n-9
`18:0
`16:0
`
`Fatty acid
`
`Figure 3 Stereospecific analysis of phospholipids and triacylglycerols. (a) The positional distribution of phosphatidylcholine (upper panels) and of
`phosphatidylethanolamine (lower panels). The distribution of selected fatty acids between the sn-1 and sn-2 positions are shown for BJ4 (left) and BJ9
`(right) seeds. Fractionated phospholipids from mature seeds were digested with phospholipase A2 and the products were resolved by TLC, transmethylated
`and analyzed by GC. (b) The positional distribution of fatty acids in triacylglycerols from mature BJ4 and BJ9 seeds. Triacylglycerols were partially deacylated
`using ethyl magnesium bromide and the purified a,b-diacylglycerols were used in the synthesis of phosphatidylcholine. The distribution of fatty acids at the
`sn-1 and sn-2 positions was analyzed by digesting the resulting PC with phospholipase A2 and the sn-3 position was calculated as described in the text.
`PC, phosphatidylcholine; PE, phosphatidylethanolamine; TAG, triacylglycerols.
`
`plants. Here we described the reconstitution of the entire DHA biosyn-
`thetic pathway in plants. Achieving DHA synthesis in seeds, albeit at
`low levels, is a basis for further optimization to attain commercially
`viable levels, as has been demonstrated here for AA and EPA.
`
`METHODS
`Vector construction and plant transformation. A triple cassette containing
`three napin promoters24, three different multiple cloning site linkers and three
`octopine synthase (OCS) terminators was prepared in the plasmid pUC19. A
`three-gene construct (BJ3) was built by inserting PiD6, a phospholipid-acyl D6
`desaturase gene from P. irregulare12, TcD5, a phospholipid-acyl D5 desaturase
`gene from Thraustochytrium sp.13, and PSE1, an acyl CoA elongase gene from
`P. patens14, into the multiple cloning sites (Fig. 1). For the four-gene construct
`(BJ4), an XhoI/SalI fragment containing a phospholipid-acyl D12 desaturase
`gene from C. officinalis (CoD12)15 linked to a napin promoter and OCS
`terminator was removed from a one-gene construct and subcloned into the
`three-gene construct. The same approach was applied to make the five- and six-
`gene constructs (BJ5 and BJ6) by adding an elongase gene from Thraustochy-
`trium sp. (TcElo) and a phospholipid-acyl o3 desaturase gene from P. infestans
`(Pio3), respectively. Finally these three-, four-, five- and six-gene constructs
`were removed from pUC19 by digestion with AscI, and cloned into the binary
`vectors pGPTV or pSUN2. For the nine-gene construct (BJ9), a three-gene
`construct containing the TcD4 phospholipid-acyl D4 desaturase gene from
`Thraustochytrium sp.13, an elongase, OmElo, from the fish O. mykiss16 and a
`lysophosphatidyl acyltransferase, TcAt, from Thraustochytrium sp., was con-
`structed and transferred into the six-gene binary vector using the Gateway
`system (Invitrogen). All binary vectors used the NPTII gene with the NOS
`promoter as a selection marker. Binary vectors were transferred into Agrobac-
`terium tumefaciens strain GV3101 (pMP90) by electroporation.
`For plant transformation, hypocotyls from 5 to 6 day old seedlings of the
`0% erucic acid B.
`juncea breeding line 1424 were used as explants for
`inoculation with A. tumefaciens containing the binary constructs described
`above. Transformation of B. juncea was performed as described25.
`
`Fatty acid and lipid analyses. Fatty acid analyses of seeds and yeast cultures
`were performed by GC as described previously15. Individual fatty acids were
`identified by comparing the GC peaks with authentic fatty acid standards and/
`or by GC/MS.
`
`If seeds were to be used for more detailed lipid analyses, individual seeds
`were first heated for 10 min at 95 1C in 1 ml of isopropanol, and after
`homogenization, 50-ml aliquots were removed and analyzed by GC to identify
`the segregating transgenic and nontransgenic seeds. The isopropanol extracts of
`transgenic seeds were then pooled (12 seeds per sample), centrifuged, the
`supernatant collected and the pellet reextracted with isopropanol/chloroform
`1:1 (vol/vol). The two extracts from each sample were combined, evaporated
`and dissolved in chloroform. T2 seeds were processed directly without GC
`screening. The resulting lipid extract was prefractionated into neutral lipids,
`glycolipids and phospholipids on a silica PrepSep column (Fisher Scientific)26.
`These fractions were further resolved on silica G-25 thin layer chromatography
`(TLC) plates (Macherey-Nagel). Neutral lipids were developed with hexane/
`diethyl ether/acetic acid (70:30:1, vol/vol/vol), glycolipids with chloroform/
`methanol/ammonia (65:25:4, vol/vol/vol) and phospholipids with chloroform/
`methanol/ammonia/water (70:30:4:1, vol/vol/vol/vol). The individual
`lipid
`classes were identified under UV light after a primuline spray (0.05% (wt/
`vol) in acetone/water, 80:20, vol/vol; Sigma), removed from the plate by
`scraping, and used for direct transmethylation or extracted by an appropriate
`solvent for further analysis. The diacylglycerols were extracted and rerun on a
`boric acid–containing silica TLC plate with chloroform/acetone (96:4, vol/vol)
`before GC analysis.
`
`Positional analysis of triacyglycerols and phospholipids. Separated and
`extracted phospholipid classes were dissolved in 0.5 ml of borate buffer
`(0.5M, pH 7.5, containing 0.4 mM CaCl2). After a brief sonication, 5U of
`phospholipase A2 from venom of Naja mossambica (Sigma P-7778) and 2 ml
`diethyl ether were added and samples were vortexed for 2 h at 22 1C. The ether
`phase was evaporated, the digestion was stopped with 0.3 ml 1M HCl, and the
`reaction mixture was extracted with chloroform/methanol (2:1, vol/vol). The
`digested phospholipids were separated by TLC in chloroform/methanol/
`ammonia/water (70:30:4:2, vol/vol/vol/vol) and spots corresponding to released
`free fatty acids and lysophospholipids were removed by scraping and directly
`transmethylated.
`Fatty acid profiles of triacylglycerol stereopositions were determined by
`partial chemical deacylation of 20–30 mg of TLC-purified triacylglycerol, as
`described previously10, with some modifications. a,b-diacylglycerol was pur-
`ified by TLC on boric acid-treated silica plates (chloroform/acetone, 96:4, vol/
`vol), extracted and used for phosphatidylcholine synthesis27. The mixture of
`
`1 0 1 6
`
`VOLUME 23 NUMBER 8 AUGUST 2005 NATURE BIOTECHNOLOGY
`
`CSIRO Exhibit 1008
`
`©2005 Nature Publishing Group http://www.nature.com/naturebiotechnology
`
`

`

`L E T T E R S
`
`phosphatidylcholine molecules with head groups at sn-1 and sn-3 positions was
`extracted from silica
`after TLC (chloroform/methanol/ammonia/water,
`70:30:4:1, vol/vol/vol/vol) and digested with phospholipase A2 as described
`above for phospholipids. The fatty acid profile of the resulting lysophos-
`phatidylcholine represented the triacylglycerol sn-1 position, and the released
`free fatty acids, the sn-2 position. The remaining sn-3 position was calculated
`according to the formula sn-3 ¼ 3  TAG–(sn-1 + sn-2).
`
`ACKNOWLEDGMENTS
`The authors thank Derek Potts for providing B. juncea germplasm, and Darwin
`Reed and Mike Giblin for assistance with GC and GC/MS analysis. We also
`thank Jonathan Page for comments on an earlier version of this manuscript.
`
`COMPETING INTERESTS STATEMENT
`The authors declare competing financial interests (see the Nature Biotechnology
`website for details).
`
`Received 11 February; accepted 10 May 2005
`Published online at http://www.nature.com/naturebiotechnology/
`
`1. Funk, C.D. Prostaglandins and leukotrienes: advances in eicosanoid biology. Science
`294, 1871–1875 (2001).
`2. Hong, S. et al. Docosatrienes and 17S-resolvins generated from docosahexaenoic acid
`in murine brain,human blood, and glial cells. Autacoids in anti-inflammation. J. Biol.
`Chem. 278, 14677–14687 (2003).
`3. Crawford, M.A. et al. Are deficits of arachidonic and docosahexaenoic acids responsible
`for the neural and vascular complications of preterm babies? Am. J. Clin. Nutr. 66,
`1032S–1041S (1997).
`4. Benatti, P., Peluso, G., Nicolai, R. & Calvani, M. Polyunsaturated fatty acids: biochem-
`ical, nutritional and epigenetic properties. J. Am. Coll. Nutr. 23, 281–302 (2004).
`5. Spector, A. Essentiality of fatty acids. Lipids 34, S1–S3 (1999).
`6. Slater, S. et al. Metabolic engineering of Arabidopsis and Brassica for poly(3-
`hydroxybutyrate-co-3-hydroxyvalerate) copolymer production. Nat. Biotechnol. 17,
`1011–1016 (1999).
`7. Ye, X. et al. Engineering the provitamin A (beta-carotene) biosynthetic pathway into
`(carotenoid-free) rice endosperm. Science 287, 303–305 (2000).
`8. Thelen, J.J. & Ohlrogge, J.B. Metabolic engineering of fatty acid biosynthesis in plants.
`Metab. Eng. 4, 12–21 (2002).
`9. Qi, B. et al. Production of very long chain polyunsaturated omega-3 and omega-6 fatty
`acids in plants. Nat. Biotechnol. 22, 739–745 (2004).
`10. Abbadi, A. et al. Biosynthesis of very-long-chain polyunsaturated fatty acids in
`transgenic oilseeds: constraints on their accumulation. Plant Cell 16, 2734–2748
`(2004).
`
`11. Qiu, X. Biosynthesis of docosahexaenoic acid (DHA, 22:6–4,7,10,13,16,19):
`two distinct pathways. Prostaglandins Leukot. Essent. Fatty Acids 68, 181–186
`(2003).
`12. Hong, H. et al. High-level production of gamma-linolenic acid in Brassica juncea
`using a delta 6 desaturase from Pythium irregulare. Plant Physiol. 129, 354–362
`(2002).
`13. Qiu, X., Hong, H. & MacKenzie, S.L. Identification of a delta 4 fatty acid desaturase
`from Thraustochytrium sp. involved in the biosynthesis of docosahexaenoic acid by
`heterologous expression in Saccharomyces cerevisiae and Brassica juncea. J. Biol.
`Chem. 276, 31561–31566 (2001).
`14. Zank, T.K. et al. Cloning and functional characterisation of an enzyme involved in the
`elongation of delta 6 polyunsaturated fatty acids from the moss Physcomitrella patens.
`Plant J. 31, 255–268 (2002).
`15. Qiu, X., Reed, D.W., Hong, H., MacKenzie, S.L. & Covello, P.S. Identification and
`analysis of a gene from Calendula officinalis encoding a fatty acid conjugase. Plant
`Physiol. 125, 847–855 (2001).
`16. Meyer, A. et al. Novel fatty acid elongases and their use for the reconstitution of
`docosahexaenoic acid biosynthesis. J. Lipid Res. 45, 1899–1909 (2004).
`17. Leonard, A.E., Pereira, S.L., Sprecher, H. & Huang, Y.S. Elongation of long-chain fatty
`acids. Prog. Lipid Res. 43, 36–54 (2004).
`18. Han, J. et al. Functional characterization of beta-ketoacyl-CoA synthase genes from
`Brassica napus L. Plant Mol. Biol. 46, 229–239 (2001).
`19. Domergue, F. et al. Acyl carriers used as substrates by the desaturases and elongases
`involved in very long-chain polyunsaturated fatty acids biosynthesis reconstituted in
`yeast. J. Biol. Chem. 278, 35115–35126 (2003).
`20. Wiberg, E., Banas, A. & Stymne, S. Fatty acid distribution and lipid metabolism in
`developing seeds of laurate-producing rape (Brassica napus L.). Planta 203, 341–348
`(1997).
`21. Wiberg, E., Edwards, P., Byrne, J., Stymne, S. & Dehesh, K. The distribution of
`caprylate, caprate and laurate in lipids from developing and mature seeds of transgenic
`Brassica napus L. Planta 212, 33–40 (2000).
`22. Graham, I.A., Cirpus, P., Rein, D. & Napier, J.A. The use of very long chain
`polyunsaturated fatty acids to ameliorate metabolic syndrome: transgenic plants as
`an alternative sustainable source to fish oils. Nutr. Bul. 29, 228–233 (2004).
`23. Napier, J.A., Beaudoin, F., Michaelson, L.V. & Sayanova, O. The production of long
`chain polyunsaturated fatty acids in transgenic plants by reverse-engineering. Biochi-
`mie 86, 785–792 (2004).
`24. Ericson, M.L. et al. Structure of the rapeseed 1.7 S storage protein, napin, and its
`precursor. J. Biol. Chem. 261, 14576–14581 (1986).
`25. Radke, S., Turner, J. & Faccioti, D. Transformation and regeneration of Brassica rapa
`using Agrobacterium tumefaciens. Plant Cell Rep. 11, 499–505 (1992).
`26. Lynch, D. & Steponkus, P. Plasma membrane lipid alterations associated with cold
`acclimation of winter rye seedlings (Secale cereale L. cv. Puma). Plant Physiol. 83,
`761–767 (1987).
`27. Myher, J.J. & Kuksis, A. Stereospecific analysis of triacylglycerols via racemic phos-
`phatidylcholines and phospholipase C. Can. J. Biochem. 57, 117–124 (1979).
`
`©2005 Nature Publishing Group http://www.nature.com/naturebiotechnology
`
`NATURE BIOTECHNOLOGY VOLUME 23 NUMBER 8 AUGUST 2005
`
`
`
`View publication statsView publication stats
`
`CSIRO Exhibit 1008
`
`1 0 1 7
`
`