Omega-3 long-chain polyunsaturated fatty acids (ω3 LC-PUFAs) such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are important for human health. Suboptimal levels of ω3 LC-PUFAs are associated with increased risk of several diseases (Ghasemi Fard et al., 2019). Docosapentaenoic acid (ω3-DPA, C22:5) is a rare LC-PUFA but of special interest because of its unique properties (Drouin and Legrand, 2019; Kaur et al., 2013). Multiple studies showed direct effects of DPA on inflammation, improved plasma lipid profile and cognitive function (Ghasemi Fard and Cameron-Smith, 2021). The principal sources of DPA are wild oceanic fish species. However, DPA is not currently available in sufficient quantities for commercial production. An inexpensive and sustainable supply of this important ω3 fatty acid is highly desirable to conduct large-scale human intervention studies to examine the role of ω3-DPA in relation to optimal health.

Significant efforts to engineer the production of ω3 LC-PUFAs in oilseed crops have been attempted recently. Two distinctive approaches have been used to produce ARA, EPA and DHA in seed oil crops comparable to the levels of wild fish oil (Petrie et al., 2020; Usher et al., 2017; Walsh et al., 2016). These include the anaerobic polyketide synthase system and the aerobic desaturase pathway of LC-PUFA biosynthesis. The aerobic pathway involves sequential desaturation and elongation steps (Robert et al., 2005, Figure 1). Introduction of additional Δ12-desaturase (Δ12-Des) and ω3-desaturase (ω3-Des) in oil crop enhanced the DHA level (Petrie et al., 2020). However, there has been no attempt to produce DPA in higher plants. We introduced the aerobic LC-PUFA biosynthesis pathway into Brassica juncea and produced high levels of both DHA and DPA, the first successful production of DPA in a crop. The level of DPA was two to three times higher than the highest found in fish oil, providing a scalable platform for efficient DPA production. We also report the first successful production of DHA to DPA in 1:1 ratio in B. juncea seed oil.

Brassica juncea was transformed with the binary vector GA7_ModB (Figure 1a) used previously to develop DHA canola (Petrie et al., 2020). Full T-DNA insertion from this vector produces DHA, while an incomplete T-DNA insertion could lead to the accumulation of intermediates, including DPA (Figure 1b). Seed fatty acid composition was analysed by gas chromatography (Zhou and Singh, 2013). DPA positional distribution on triacylglycerol was determined as previously described (Petrie et al., 2014). Seed oil content was verified by NMR using an MQC benchtop analyser (Oxford Instruments) following the manufacture's instruction.

Among 21 independent transgenic B. juncea lines, DHA levels in pooled T1 seeds ranged from 0% to 6.6% of total fatty acids. Interestingly, Line 4 had a substantial level of DPA (3.7%), and 6.6% DHA. Single T₁ seed analysis of Line 4 showed a DPA content from 0.3% to 16.1% and DHA from 0 to 17.9%. Eight of 30 single T₁ seeds contained 2.5–16.1% DPA but no DHA. Line 4 was then further analysed for fatty acid composition in half cotyledons of 48 germinating seeds. A range of 3.8–18.1% DPA was observed in half cotyledons of 11 T₁ seeds, without any DHA, while others contained various levels of DHA. This suggested there was a segregation of multiple T-DNA insertions in T₁ seeds leading to either DHA or DPA accumulation. Nineteen plants with either high DHA or high DPA without DHA were established. Fourteen of these plants had 3.6–17.2% DHA in T₂ seeds. One progeny with 17.2% DHA was designated as BjDHA-4-17 and advanced to T₄ seeds by selfing. The DHA level in T₄ seeds remained 17% (Figure 1c). The other five T₁ plants contained 4.2–12.5% DPA with no DHA in T₂ seeds. These were designated BjDPA-4-13, BjDPA-4-19, BjDPA-4-25, BjDPA-4-34 and BjDPA-4-39, potentially containing truncated inserts without a functional Δ4-desaturase gene (Δ4-Des). BjDPA-4-34 was advanced to T₆ seeds which contained 12 ± 1.3% DPA.

Line BjDHA-4-17 (17.2% DHA) was crossed with line BjDPA-4-19 (11.6% DPA), resulting in 68 F₁ seeds. Half cotyledon analysis revealed that 11 F₁ seeds contained both 4.1–6.0% DPA and 14.7–20.3% DHA. Pooled seed analysis of F₂ seeds from these 11 F₁ plants showed DPA levels ranging from 1.3% to 7.2% and DHA from 4.0% to 10.2%. Progeny 52, which had 7.2% DPA and 10.1% DHA in F₂ seeds, was advanced to F₂ plant. Pooled seed analysis of F₃ seeds from 14 F₂ plants showed variation in the amount of DHA, DPA and the sum of DHA+DPA, including six plants with an almost 1:1 DPA:DHA ratio (Figure 1c).

Sequencing of genomic DNA from the seedlings of BjDPA-4-34-2-8-7 (T₄) showed there were three partial inserts containing the functional gene cassettes from GA7_ModB except for the ω3-Des and Δ4-Des, leading to no conversion of DPA to DHA (Figure 1b). The function of the missing ω3D, converting C18:2 to C18:3, was complemented by the endogenous Δ15-desaturase.

Seed oil content remained same in T₅ (34.3 ± 2.1%) and T₆ (33.5 ± 2.0%) seeds derived from BjDPA-4 compared to the wild type (35.1 ± 2.1%) grown at the same time with no statistical difference. DPA was preferentially located at the sn-1/3 positions (91.6%) of the triacylglycerol molecules. Similar preferential distribution of DHA was previously reported (Petrie et al., 2020).

In this study, we explored the introduction of LC-PUFA biosynthesis pathway into B. juncea to produce DPA or DHA. Their levels were stable over four generations in BjDHA-4-17 and six generations in BjDPA-4-34. Although T-DNA truncations were observed and integration occurred at three different loci in the BjDPA-4 event, DPA levels were stable in both the glasshouse and field over several years and up to the T₆ generation. The oil from BjDPA lines has several unique features. It is relatively high in DPA, a highly beneficial ω3 LC-PUFA for dietary supplementation with increasing interest from the medical community (Kaur et al., 2013). In addition, the oil contained a high level of α-linolenic ALA (ca. ~>20%, compared to 15% in WT B. juncea oil), contributing to an increased ω3:ω6 ratio, with concomitant health benefits. An almost 1:1 ratio of DHA to DPA was produced in F₃ pooled seeds from the BjDHA × BjDPA crosses, with a total of ~17% DPA+DHA in the seed oil (Figure 1c). An oil with DHA and DPA in a 1:1 ratio may be an excellent source for promoting cardiovascular health.

This study demonstrates the production of 12% of DPA (two to three times higher than any other natural source) or 18% DHA in transgenic B. juncea, and the production of equal amounts of DHA and DPA in seed oil. Production of DPA in B. juncea is also more sustainable, removing the need to exploit ocean resources.

二十碳五烯酸 (EPA) 和二十二碳六烯酸 (DHA) 等 Omega-3 长链多不饱和脂肪酸 (ω3 LC-PUFA) 对人类健康很重要。ω3 LC-PUFA 的次优水平与几种疾病的风险增加有关（Ghasemi Fard等人，2019 年）。二十二碳五烯酸 (ω3-DPA, C22:5) 是一种罕见的 LC-PUFA，但因其独特的性质而备受关注（Drouin 和 Legrand，2019 年；Kaur等人，2013 年）。多项研究表明 DPA 对炎症、改善血浆脂质谱和认知功能有直接影响（Ghasemi Fard 和 Cameron-Smith，2021）。DPA 的主要来源是野生海洋鱼类。然而，目前 DPA 的数量不足以用于商业生产。这种重要的 ω3 脂肪酸的廉价且可持续的供应对于进行大规模的人类干预研究以检查 ω3-DPA 在最佳健康方面的作用是非常可取的。

最近尝试了在油料作物中设计生产 ω3 LC-PUFA 的重大努力。两种独特的方法已被用于在种子油料作物中生产与野生鱼油水平相当的 ARA、EPA 和 DHA（Petrie等人，2020 年；Usher等人，2017 年；Walsh等人，2016 年）。这些包括厌氧聚酮化合物合成酶系统和 LC-PUFA 生物合成的需氧去饱和酶途径。需氧途径涉及连续的去饱和和延伸步骤（Robert et al ., 2005， 图1）。在油料作物中引入额外的 Δ12-去饱和酶 (Δ12-Des) 和 ω3-去饱和酶 (ω3-Des) 可提高 DHA 水平（Petrie等人，2020 年）。然而，没有尝试在高等植物中生产DPA。我们将好氧 LC-PUFA 生物合成途径引入芥菜中，并产生高水平的 DHA 和 DPA，这是作物中首次成功生产 DPA。DPA 的含量是鱼油中最高含量的两到三倍，为高效生产 DPA 提供了一个可扩展的平台。我们还报告了在芥菜籽油中以 1:1 的比例首次成功生产 DHA 与 DPA 。

芥菜用之前用于开发 DHA 油菜的二元载体 GA7_ModB（图 1a）进行转化（Petrie等人，2020 年）。来自该载体的完整 T-DNA 插入产生 DHA，而不完整的 T-DNA 插入可能导致中间体的积累，包括 DPA（图 1b）。通过气相色谱分析种子脂肪酸组成（Zhou and Singh, 2013）。如前所述确定 DPA 在三酰基甘油上的位置分布 (Petrie et al ., 2014 )。根据制造商的说明，使用 MQC 台式分析仪（Oxford Instruments）通过 NMR 验证种子油含量。

在 21 个独立的转基因芥菜品系中，混合的 T1 种子中的 DHA 水平占总脂肪酸的 0% 至 6.6%。有趣的是，第 4 行的 DPA（3.7%）和 6.6% DHA 的含量很高。品系 4 的单 T ₁种子分析显示 DPA 含量为 0.3% 至 16.1%，DHA 含量为 0 至 17.9%。30 个单 T ₁种子中有 8 个含有 2.5-16.1% DPA 但不含 DHA。然后进一步分析第 4 行的 48 个发芽种子的半子叶中的脂肪酸组成。_{在 11 个 T 1}种子的半子叶中观察到 3.8-18.1% 的 DPA 范围，不含任何 DHA，而其他种子则含有不同水平的 DHA。_{这表明在 T 1}中存在多个 T-DNA 插入的分离导致 DHA 或 DPA 积累的种子。建立了 19 株具有高 DHA 或高 DPA 而没有 DHA 的植物。_{这些植物中的 14 株在 T 2}种子中含有 3.6-17.2% 的 DHA 。一个具有 17.2% DHA 的后代被命名为 BjDHA-4-17，并通过自交获得 T ₄种子。_{T 4}种子中的 DHA 水平保持在 17%（图 1c）。其他五株 T _{1植物在 T 2}种子中含有 4.2-12.5% DPA，不含 DHA 。这些被命名为 BjDPA-4-13、BjDPA-4-19、BjDPA-4-25、BjDPA-4-34 和 BjDPA-4-39，可能包含没有功能性 Δ4-去饱和酶基因 ( Δ4-Des ) 的截短插入片段。BjDPA-4-34 被推进到 含有 12 ± 1.3% DPA的 T _6种子。

品系BjDHA-4-17 (17.2% DHA)与品系BjDPA-4-19 (11.6% DPA)杂交，产生68个F ₁种子。半子叶分析表明，11 颗 F ₁种子同时含有 4.1-6.0% DPA 和 14.7-20.3% DHA。来自这 11 株 F _{1植物的 F 2}种子的混合种子分析显示 DPA 水平在 1.3% 至 7.2% 之间，DHA 水平在 4.0% 至 10.2% 之间。_{F 2}种子中含有7.2% DPA和10.1% DHA的后代52被推进到F ₂植物。对来自 14 株 F _{2植物的 F 3}种子进行的混合种子分析显示 DHA、DPA 和 DHA+DPA 总和的变化，包括 DPA:DHA 比率几乎为 1:1 的六株植物（图 1c）。

来自 BjDPA-4-34-2-8-7 (T ₄ ) 幼苗的基因组 DNA 测序显示，除了 ω3-Des 和 Δ4-Des 之外，存在三个包含来自 GA7_ModB 的功能基因盒的部分插入片段，导致没有DPA 到 DHA 的转化（图 1b）。缺失的 ω3D 的功能，将 C18:2 转换为 C18:3，由内源性 Δ15-去饱和酶补充。

与同时生长的野生型 (35.1 ± 2.1%) 相比，源自 BjDPA-4的 T ₅ (34.3 ± 2.1%) 和 T ₆ (33.5 ± 2.0%) 种子的种子油含量保持不变，无统计学差异。DPA优先位于三酰基甘油分子的sn -1/3 位置（91.6%）。先前报道了 DHA 的类似优先分布（Petrie等人，2020 年）。

在本研究中，我们探索了将 LC-PUFA 生物合成途径引入B中。 芥菜生产 DPA 或 DHA。在 BjDHA-4-17 和 BjDPA-4-34 中，它们的水平在四代和六代中保持稳定。尽管在 BjDPA-4 事件中观察到 T-DNA 截断并且在三个不同的位点发生了整合，但 DPA 水平在温室和田间均保持稳定多年，直至 T ₆代。BjDPA 生产线的油具有几个独特的特性。它的 DPA 含量相对较高，这是一种非常有益的 ω3 LC-PUFA，可用于膳食补充剂，越来越受到医学界的关注（Kaur等，2013）。此外，该油含有高水平的 α-亚麻酸 ALA（约 >20%，而 WT B.juncea油中 为 15% ），有助于提高 ω3:ω6 比率，并带来健康益处。_{在来自 BjDHA × BjDPA 杂交的 F 3}混合种子中，DHA 与 DPA 的比例几乎为 1:1 ，种子油中的 DPA+DHA 总量约为 17%（图 1c）。DHA 和 DPA 比例为 1:1 的油可能是促进心血管健康的极好来源。

该研究表明转基因芥菜中产生了 12% 的 DPA（比任何其他天然来源高 2 至 3 倍）或 18% 的 DHA ，并且在种子油中产生了等量的 DHA 和 DPA。芥菜中 DPA 的生产也更具可持续性，无需开发海洋资源。