Biotechnology of α-linolenic acid in oilseed rape (Brassica napus)using FAD2 and FAD3 from chia (Salvia hispanica)

XUE Yu-fei , Inkabanga Tseke ALAIN , , YIN Neng-wen , JIANG Jia-yi , ZHAO Yan-ping , LU Kun , LI Jia-na , DING Yan-song , ZHANG Shi-qing , CHAI You-rong #

1 Integrative Science Center of Germplasm Creation in Western China (Chongqing) Science City and Southwest University/Chongqing Key Laboratory of Crop Quality Improvement/College of Agronomy and Biotechnology, Southwest University,Chongqing 400715, P.R.China

2 Engineering Research Center of South Upland Agriculture, Ministry of Education/Academy of Agricultural Sciences, Southwest University, Chongqing 400715, P.R.China

3 Faculty of Agricultural Sciences, National Pedagogic University (UPN), Kinshasa 8815, D.R.Congo

Abstract α-Linolenic acid (ALA, 18:3Δ9,12,15) is an essential fatty acid for humans since it is the precursor for the biosynthesis of omega-3 long-chain polyunsaturated fatty acids (LC-PUFA).Modern people generally suffer from deficiency of ALA because most staple food oils are low or lack ALA content.Biotechnological enrichment of ALA in staple oil crops is a promising strategy.Chia (Salvia hispanica) has the highest ALA content in its seed oil among known oil crops.In this study, the FAD2 and FAD3 genes from chia were engineered into a staple oil crop, oilseed rape (Brassica napus), via Agrobaterium tumefaciens-mediated transformation of their LP4-2A fusion gene construct driven by the seed-specific promoter PNapA.In seeds of T0, T1, and T2 lines, the average ALA contents were 20.86, 23.54, and 24.92%, respectively,which were 2.21, 2.68, and 3.03 folds of the non-transformed controls (9.42, 8.78, and 8.22%), respectively.The highest seed ALA levels of T0, T1, and T2 plants were 38.41, 35.98, and 39.19% respectively, which were 4.10-4.77 folds of the respective controls.FA-pathway enzyme genes (BnACCD, BnFATA, BnSAD, BnSCD, BnDGAT1, BnDGAT2, and BnDGAT3) and positive regulatory genes (BnWRI1, BnLEC1, BnL1L, BnLEC2, BnABI3, BnbZIP67, and BnMYB96) were all significantly up-regulated.In contrast, BnTT1, BnTT2, BnTT8, BnTT16, BnTTG1, and BnTTG2, encoding negative oil accumulation regulators but positive secondary metabolism regulators, were all significantly down-regulated.This means the foreign ShFAD2-ShFAD3 fusion gene, directly and indirectly, remodeled both positive and negative loci of the whole FA-related network in transgenic B.napus seeds.

Keywords: biotechnology, α-linolenic acid, oilseed rape (Brassica napus), FAD2, FAD3, chia (Salvia hispanica)

1.Introduction

Omega-3 long-chain polyunsaturated fatty acids(LC-PUFA) eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are of great importance in human health, e.g., body growth and development,prevention of cardiovascular diseases, and improvement of inflammatory diseases (Wuet al.2005; Ghasemi Fardet al.2019).α-Linolenic acid (ALA, 18:3Δ9,12,15) is the biosynthesis precursor of EPA and DHA, but it cannot be synthesized in the human body (Bakeret al.2016).Therefore, ALA is an essential fatty acid and must be obtained from daily diets.Modern people generally suffer from deficiency of ALA because most staple food oils are low or lack ALA content.Dietary ALA is rich in some nonstaple oil crops (Ciftciet al.2012; Yuanet al.2022), such as chia (Salviahispanica), perilla, flax, and tree peony.However, they have low yield and poor environmental adaption and thus cannot be widely grown and bountifully produced.Therefore, the enrichment of ALA in staple oil cropsviametabolic engineering is a promising strategy.

Attempts have been made to engineer ALA using seed- or endosperm-specific expression ofFAD3in plants with high-LA precursor, but most studies were performed not on oil crops (Liuet al.2012).For oil crops, there are only four reports on ALA engineering in soybean and cottonseed: over-expressing bifunctional Δ12/ω3 fatty acid desaturase from fungusFusariummoniliforme(Damudeet al.2006),FAD3fromArabidopsisthaliana(Eckertet al.2006),FAD3-1fromLesquerella(Yeomet al.2020), andFAD3fromBrassicanapus(Gaoet al.2020).

Brassicanapus(rapeseed or oilseed rape) is a staple oil crop of world importance, and canola oil has high oleic acid (OA, about 60%) (Xuet al.2020).However, to date,there is no report on the creation of high-ALA rapeseed.FAD2 and FAD3 are key enzymes in ALA biosynthesis from OA precursor (Fig.1-A).Thus,FAD2andFAD3genes are key targets for ALA engineering.The mysterious ancient Mesoamerican Indian crop chia is revived and expanding worldwide due to its richness of valuable nutraceuticals such as ALA, antioxidants, food fiber, gels, and proteins(Winet al.2018).Among known oil crops, chia has the highest ALA content in its seed oil (about 60-71%), and we have previously cloned and functionally identified theShFAD2family andShFAD3family genes from chia(Xueet al.2017, 2018).In this study, we successfully engineered rapeseed for high ALA production using theShFAD2andShFAD3genes from chia.

Fig.1 Expression strategy and transgenic effect of chia FAD2-FAD3 fusion gene.A, α-linolenic acid (ALA) synthesis pathway in seeds.B, chart of vector construction.C, gas chromatography (GC) peaks of seed fatty acids of T2 plants (S23, S23-26-5-3) and non-transformed control (NT, NT-T2).FID, flame ionization detector.D, seed fatty acid profiles of three generations of transgenic lines and corresponding controls.

2.Materials and methods

2.1.Isolation of chia FAD2 and FAD3 coding regions and vector construction

According to our previous studies (Xueet al.2017, 2018),the preferred family membersShFAD2-2(GenBank accession no.KX610644) andShFAD3-2(KX610646) from ALA-king species chia were chosen for ALA engineering.Their coding regions were amplified by high-fidelity PCR and fused together using our newly designed proteinaseintein double-linker LP4-2A (amino acid sequence SNAADEVATLLNFDLLKLAGDVESNPGP, nucleic acid sequence 5´-TCTAACGCTGCTGATGAAGTTGCTACA CTTCTTAACTTTGATCTTCTTAAGCTTGCTGGTGATGT TGAATCTAACCCTGGTCCT-3´ with dicot especially cruciferous codons) to form a fusion geneShFAD2-2-LP4-2A-ShFAD3-2(Fig.1-A; Appendix A).In the platform vector pC2301M1NPB (Fuet al.2017), it was inserted between seed-specific promoter PNapAand terminator TNOSto form recombinant vector pC2301M1NPB-ShFAD2-ShFAD3(S23)viaXbaI-XmaI double-digestion and T4-ligation method.The correct plasmid of S23 was transformed intoAgrobacteriumtumefaciensstrain LBA4404 to obtain engineering strains.

2.2.Plant transformation and screening

S23 construct was introduced intoB.napuscv.ZS10 byA.tumefaciens-mediated transformation, according to Fuet al.(2017).The positive transformation was determined by GUS staining, Basta herbicide resistance,and PCR checking of the regenerated plants, and confirmed transgenic plants were self-pollinated to reach three generations.In T1and T2generations, Mendelian selection was performed to obtain homozygous plants.

2.3.Fatty acid and seed oil analysis

The fatty acid compositions and oil content of mature seeds in transgenic lines and control were assayed by an Agilent 7890B Gas Chromatography (GC) System and near-infrared reflectance spectroscopy (NIRS DS2500,Foss, Denmark), respectively, using the methods described previously (Lianet al.2017).

2.4.qRT-PCR analysis

The qRT-PCR primers (Appendix A) of family genes in rapeseed were designed based on conserved regions of all family members using rapeseed genome data from NCBI (https://ncbi.nlm.nih.gov/) and Genoscope (https://www.genoscope.cns.fr/brassicanapus/).The qRT-PCR with three technical replicates was performed to examine relative expression levels of genes in the developing seeds of the best T2line using the methods described previously at 30 days after flowering (DAF) (Lianet al.2017).qRT-PCR data analysis was carried out as previously described (Livak and Schmittgen 2001).

2.5.Statistical analysis

Data in this study were statistically analyzed with ANOVA,and statistical significance was determined by Student’st-test:＊,P<0.05;＊＊,P<0.01.

3.Results

In this study, we obtained 5 independent positive transgenic T0plants (Appendix B), which were self-pollinated to generate T1and T2generations.Seed fatty acids of T0-T2plants were analyzed by GC.As shown in Fig.1 and Table 1, the engineered rapeseed lines produced high levels of ALA content in their seed oil.In T0plants, ALA contents in seeds ranged from 12.14 to 36.27% of total fatty acids, and mean ALA content (20.86%) is 2.21 folds of the control (9.42%).n6/n3 (18:2/18:3) ratio in seeds of T0plants ranged from 0.51 to 2.19, and their average value(1.31) was 0.46 fold of the control (2.87).In T1plants,ALA contents in seeds ranged from 14.16 to 35.16% of total fatty acids, and the mean ALA level (23.54%) was 2.68 folds of the control (8.78%).n6/n3 ratio in seeds of T1plants ranged from 0.50 to 1.69, and their average level(0.99) was 0.42 fold of the control (2.37).In T2plants, ALA contents ranged from 16.31 to 37.93% of total fatty acids,and mean ALA content (24.92%) was 3.03 folds of the control (8.22%).n6/n3 ratio in seeds of T2plants ranged from 0.48 to 1.24, and their average value (0.82) was 0.35 fold of the control (2.37).The highest ALA contents of seeds of T0, T1, and T2plants among all tested samples were 38.41, 35.98, and 39.19%, respectively, which were 4.10-4.77 folds of the respective non-transformed controls (NTs).This result showed that the high-ALA trait in transgenic rapeseed was stably inherited over three generations, and there was an increase from seeds of the T0generation to seeds of the T2generation owing to Mendelian selection.

Table 1 Seed fatty acid compositions and n6/n3 ratio of transgenic plants and non-transformed control

To explore the influence of ALA content in transgenic lines on seed oil content, the best lines in five individual transgenic events in three generations were selected to detect their seed oil content using the near-infrared reflectance spectroscopy method.As shown in Appendix C, the seed oil content of T0, T1, and T2plants remained at the same level compared to the controls NT-T0, NT-T1, and NT-T2, respectively, with no significant change.

To reveal the effect ofShFAD2-ShFAD3transgene on the expression of FA-related network (Yanget al.2022), relative expression levels of genes in the developing seeds of the best T2plant at 30 DAF were analyzed by qRT-PCR.High expression levels ofShFAD2-2andShFAD3-2were detected in 30-DAF seeds of S23-26-5-3 (Fig.2-A), and the expression levels of endogenousBnFAD2andBnFAD3genes had no significant change (Fig.2-B).Significant or extremely significant transcriptional up-regulation was observed on FA-pathway key enzymes genesBnACCD,BnFATA,BnSAD,BnSCD,BnDGAT1,BnDGAT2, andBnDGAT3,as well as FA-pathway positive regulatory genesBnWRI1,BnLEC1,BnL1L,BnLEC2,BnABI3,BnbZIP67andBnMYB96(Tanet al.2011; Yanget al.2022; Fig.2-B).In contrast,BnTT1,BnTT2,BnTT8,BnTT16,BnTTG1,andBnTTG2, which encode negative regulators of oil accumulation but positive regulators of flavonoidproanthocyanidin biosynthesis (Wanget al.2014; Lianet al.2017; Xieet al.2020), were all significantly or extremely significantly down-regulated (Fig.2-C).In addition, qRT-PCR assay showed thatShFAD2-2andShFAD3-2were highly expressed in 30-DAF seeds of T2lines of five individual transgenic events (Appendix D),and their expression levels were also associated with the corresponding seed ALA content (Table 1; Fig.1-D).

Fig.2 Relative transcript levels of genes encoding α-linolenic acid (ALA)-related key enzymes (A and B) and transcription factors(C) at 30 days of flowering (DAF) seeds of best T2 plant (S23-26-5-3) and non-transformed control (NT-T2).Values represent mean±SD of three technical replicates.＊, P<0.05; ＊＊, P<0.01.

4.Discussion

Modern people generally suffer from ALA deficiency because most staple food oils have low or deficient ALA content.Dietary ALA is rich in some non-staple oil crops(Ciftciet al.2012; Yuanet al.2022), such as chia, perilla,flax, and tree peony.However, they have low yields and poor environmental adaptability, so they cannot be widely cultivated and abundantly produced.Enrichment of ALA in staple oil cropsviametabolic engineering is of great importance.Among known oil crops, chia seed oilhas the highest ALA content (about 60-71%), and thus,chiaFAD2andFAD3genes should be ideal target genes for ALA engineering.Previous studies have cloned and functionally identified theShFAD2family andShFAD3family genes from chia (Xueet al.2017, 2018).In this study, we successfully engineered canola for high ALA production using theShFAD2andShFAD3genes from chia.The highest ALA contents of seeds of T0, T1, and T2plants among all tested samples were 38.41, 35.98,and 39.19%, respectively, which were 4.10-4.77 folds of the respective NTs.The high-ALA trait of transgenic rapeseed is stably inherited within three generations,and there existed an increase from seeds of the T0generation to seeds of the T2generation due to Mendelian selection.Seed oil contents of T0, T1, and T2plants had no significant differences with the corresponding controls.This is also consistent with the result of engineering docosapentaenoic acid (DPA) and DHA production inBrassicajuncea(Belideet al.2022).

In previously reported metabolic engineering of PUFA in plants, the transformation of multiple genes was mainly achieved in two ways: multiple cassettes in a plasmid (Wuet al.2005; Belideet al.2022) and serial retransformation (Qiet al.2004).However, the former was prone to potential gene silencing, and the latter required different selection markers for each transformation.Gene fusion technologyviaLP4-2A peptide linker is an alternative multigene transformation method, but to date,it has not been used in engineered PUFA production in plants.Here, for the first time, we successfully achieved ALA enrichment in rapeseed using chiaFAD2-FAD3gene fusion with a modified LP4-2A peptide linker.From the perspective of the degree of enhancement of ALA, its selfcleaving efficiency should be high, so this technology can also be applied to plant synthetic biology and molecular breeding in other fields.

Significant or extremely significant transcriptional upregulation was observed on FA-pathway key enzyme genes, as well as FA-pathway positive regulatory genes (Tanet al.2011; Yanget al.2022).In contrast, genes encoding negative regulators of oil accumulation but positive regulators of flavonoid-proanthocyanidin biosynthesis(Wanget al.2014; Lianet al.2017; Xieet al.2020; Ronget al.2022) were all significantly or extremely significantly down-regulated.This means that the introduction of two structural genes (ShFAD2-2andShFAD3-2) promoted primary metabolism and inhibited secondary metabolism simultaneously and remodeled both positive and negative loci of the FA-related network with a trend to promote ALA deposition in transgenicB.napusseeds.This interesting phenomenon deserves further study.

This is the first report on the metabolic engineering of ALA in canola, which produces a transgenic canola material with the highest level of ALA of 39.19%.The high-ALA rapeseed germplasm created by the seedspecific expression ofFAD2andFAD3from chia(Appendix E) here has great application potential.Its seed oil can be blended into ALA-poor vegetable oils,e.g., those from palm, peanut, sesame, cotton seed, and sunflower, to adjust the n6/n3 ratio (the golden ratio is 4:1).Besides, it can be used for sustainable industrial ALA production with significant cost/price reduction to meet market demand.Compared with flax, perilla, chia, and tree peony, its advantages are greater suitable planting potential regions and higher yield output potentials.

5.Conclusion

For the first time, we created the ALA-rich rapeseed germplasmviaseed-specific expression of chiaShFAD2-ShFAD3fusion gene.The highest ALA levels of seeds of T0-T2plants were 35.98-39.19%, which were 4.10-4.77 folds of the respective controls.These findings have significant application prospects in food nutrition and ALArelated industries.Moreover, directly and indirectly, the foreignShFAD2-ShFAD3fusion gene remodeled both positive and negative loci of the whole FA-related network in transgenicB.napusseeds.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (31871549, 32001441 and 32272015), the Chongqing Research Program of Basic Research and Frontier Technology, China(cstc2015jcyjBX0143), the Fundamental Research Funds for the Central Universities, China (XDJK2020C038), the National Key R&D Program of China (2016YFD0100506),and the Young Eagles Program of Chongqing Municipal Commission of Education, China (CY220219).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on https://doi.org/10.1016/j.jia.2023.05.018