The Brassica Napus Fatty Acid Exporter FAX1-1 Contributes to Biological Yield, Seed Oil Production, and Oil Quality

Background: In the oilseed crop Brassica napus (rapeseed), various metabolic processes inuence seed oil production, oil quality, and biological yield. However, the role of plastid membrane proteins in these traits has not been explored. Results: Our genome-wide association study (GWAS) of 520 B. napus accessions identied the chloroplast membrane protein-localized FATTY ACID EXPORTER 1-1 (FAX1-1) as a candidate associated with biological yield. Seed transcript levels of BnaFAX1-1 were higher in a cultivar with high seed oil content relative to a low-oil cultivar. BnaFAX1-1 localized to the plastid envelope. When expressed in Arabidopsis thaliana, BnaFAX1-1 enhanced biological yield (total plant dry matter), seed yield and seed oil content per plant. Likewise, in the ﬁ eld, B. napus BnaFAX1-1 overexpression lines (BnaFAX1-1-OE) displayed signicantly enhanced biological yield, seed yield, and seed oil content compared with the wild type. BnaFAX1-1 overexpression also up-regulated gibberellic acid 4 (GA4) biosynthesis, which may contribute to biological yield improvement. Furthermore, oleic acid (C18:1) signicantly increased in BnaFAX1-1 overexpression seeds. Conclusion: Our results indicated that the putative fatty acid exporter BnaFAX1-1 simultaneously improved seed oil production, oil quality and biological yield in B. napus, providing new approaches for future molecular breeding. yield. We propose that BnaFAX1-1 is a fatty acid exporter associated with biological yield in B. napus, based on functional annotation analysis, subcellular localization and transcript levels across various B. napus cultivars. Our results revealed that BnaFAX1-1 signicantly contributed to biological yield and improved seed oil production and oil quality. Furthermore, we observed that BnaFAX1-1 may modulate gibberellic acid 4 (GA4) content, offering a potential mechanism for the increase in biological yield. The present study provides an important solution to simultaneously improve biological yield, seed oil content, seed yield and oil quality in B. napus by manipulating a single gene: BnaFAX1-1.


Introduction
Brassica napus (rapeseed) is an important oilseed crop for edible oils, as the oil extracted from its seeds is rich in highly unsaturated fatty acids. In addition, rapeseed oil constitutes a reasonable substitute for diesel fuel as a renewable energy source due to their chemical similarities and high energy potential (Durrett et al., 2008). Increasing rapeseed oil production per hectare would thus increase edible oil and liquid biofuel production. Indeed, substantial work has been invested into improving B. napus seed yield and seed oil content. Similarly, increasing total biological yield of above-ground biomass contributes to increasing seed yield and seed oil production (Lu et al., 2016).
Previous studies revealed that rapeseed seed oil production may be enhanced by overexpressing enzymes or transcription factors involved in various metabolic processes. For example, the overexpression of yeast (Saccharomyces cerevisiae) GLYCEROL-3-PHOSPHATE DEHYDROGENASE (GPD1) under the control of a seed-speci c promoter raised seed oil content (Vigeolas et al., 2007).
Likewise, seed oil content increased in transgenic Arabidopsis thaliana plants overexpressing B. napus WRINKLED1 (WRI1)-like (Liu et al., 2010). The conditional expression of B. napus LEAFY COTYLEDON1 (BnLEC1) and LEC1-LIKE (BnL1L) in developing seeds enhanced seed oil content by 2-20% with no detrimental effects on major agronomic traits (Tan et al., 2011). Finally, the overexpression of B. napus GROWTH REGULATING FACTOR2 (GRF2)-like (BnGRF2) improved seed oil production by regulating cell number and plant photosynthesis (Liu et al., 2012). However, to date, no chloroplast membrane proteins have been characterized that improve both seed oil production and biological yield.
Biological yield re ects the total accumulation of photosynthetic products in plant tissues. In plants, photosynthesis and fatty acid (< C18) biosynthesis take place in chloroplasts (Li-Beisson et al., 2013). A recent study revealed that altering carbon metabolism in the chloroplasts of transgenic tobacco (Nicotiana tabacum) plants with high leaf oil levels caused biological yield and oil production to increase. We speculated that photosynthetic products transported across the chloroplast membranes may be critical for biological yield. We previously conducted a genome wide association study (GWAS) of biological yield with 520 B. napus accessions, and identi ed 88 single-nucleotide polymorphisms (SNPs) that were signi cantly associated with this trait (Lu et al., 2016). Few proteins have been reported that localize to the chloroplast envelope and function as transporters regulating carbon metabolism across the chloroplast membranes. In previous work, we showed that Arabidopsis FATTY ACID EXPORTER 1 (AtFAX1) localized to the chloroplast inner envelope and mediated fatty acid export from chloroplasts, an essential step in the biosynthesis of leaf and stem lipids . However, loss of function mutation in AtFAX1 had no effect on seed oil content, seed quality, seed yield, or seed oil production. Therefore, how fatty acids are transported from their site of biosynthesis in the plastids of developing seeds remains unknown.
To determine the contribution of chloroplast envelope proteins to biological yield, which might increase seed oil production in B. napus, we identi ed genes encoding proteins predicted to localize to the chloroplast envelope that mapped near signi cant SNPs from our previous GWAS results in B. napus. Notably, BnaFAX1-1, the B. napus ortholog of Arabidopsis FAX1, mapped near a signi cant SNP associated with biological yield. We propose that BnaFAX1-1 is a fatty acid exporter associated with biological yield in B. napus, based on functional annotation analysis, subcellular localization and transcript levels across various B. napus cultivars. Our results revealed that BnaFAX1-1 signi cantly contributed to biological yield and improved seed oil production and oil quality. Furthermore, we observed that BnaFAX1-1 may modulate gibberellic acid 4 (GA4) content, offering a potential mechanism for the increase in biological yield. The present study provides an important solution to simultaneously improve biological yield, seed oil content, seed yield and oil quality in B. napus by manipulating a single gene: BnaFAX1-1.

Results
Identi cation of chloroplast membrane-localized proteins potentially contributing to biological yield in B.

napus
We previously detected SNPs signi cantly associated with biological yield during a GWAS of 520 B. napus accessions (Lu et al., 2016). We selected 6,627 candidate genes contained within the intervals surrounding 88 signi cant quantitative trait loci (QTLs) associated with biological yield trait. Of those, 29 encoded proteins with potential localization to the chloroplast membrane, as determined by GO analysis (Fig. 1A, Table S1, Table S2). We further narrowed down the list of candidates to potential transporters that may be involved in the export of photosynthetic products. Notably, we identi ed two genes, BnaA04g02480D and BnaA07g17240D, that were closely linked with the signi cant SNP Bn-A07-p12412116 (Lu et al., 2016) and encoded orthologous to the Arabidopsis membrane protein FAX1, known to mediate plastid fatty acid export.
We focused on the characterization of these two FAX1 orthologous genes. The FAX protein family consists of seven members in Arabidopsis, named AtFAX1-7 . To identify potential FAX orthologues in eld mustard (Brassica rapa), wild cabbage (Brassica oleracea) and B. napus, we performed BLAST searches, using the 7 Arabidopsis FAX protein sequences as queries, leading to the identi cation of 9 putative orthologs each in B. rapa (BraFAX) and B. oleracea (BolFAX), and 21 in B. napus (BnaFAX). The physicochemical characteristics (amino acid number, theoretical isoelectric point (pI) values, relative molecular weight and number of transmembrane domains) for BnaFAX proteins are listed in Table S3.
We generated an unrooted neighbor-joining phylogenetic tree based on the 46 protein sequences of FAX family members (Fig. 1B) and discovered that AtFAX1 and six putative BnaFAX1 members (BnaFAX1-1 to BnaFAX1-6) clustered into one branch. To further characterize the B. napus FAX family, we analyzed the chromosomal locations and gene structures of the encoding genes ( Figure S1A, S1B) and predicted the conserved motifs of BnaFAX proteins using the MEME program ( Figure S1C). Of the six B. napus FAX1 members within the same branch as AtFAX1, BnaFAX1-1 (BnaA07g17240D) and BnaFAX1-2 (BnaCnng07490D) were closest to AtFAX1, as evidenced by their very similar gene structures and conserved protein motifs, suggesting that BnaFAX1-1 and BnaFAX1-2 may share the same functions as AtFAX1.
To determine what effect, if any, the six AtFAX1-like genes had on seed oil content, we analyzed their transcript levels across various tissues in one cultivar with high seed oil contents (H, cultivar name: ZS11) and one with low seed oil content (L, cultivar name: ZY821). BnaFAX1-1 was more highly expressed in the H cultivar relative to the L cultivar in all tissues tested (Fig. 1C). By contrast, BnaFAX1-3 (BnaA04g02480D) was barely detectable in either B. napus cultivar. Besides, we also observed the expression levels of 6 members of BnaFAX1 in 6 tissues of a pair of high-and low-seed oil content accessions grown in Chongqing (CQ24, CQ45) and Yunnan (YN24, YN45), among which CQ24 (seed oil content about 43%) and YN24 (seed oil content about 45%) are high-seed oil content (H-SOC) accessions, and CQ45(seed oil content about35%) and YN45 (seed oil content about 37%) are low-seed oil content (L-SOC) accessions. The result is shown in Figure S2, the expression level of BnaFAX1-1 in H-SOC accessions (CQ24, YN24) is higher than that of L-SOC accessions (CQ45, YN45) in the stem (St), leaf (Le), silique pericarps and seeds on the main in orescence of 30 days after owering (30ZP and 30ZS, respectively) and on the primary branch (30CP and 30CS, respectively) ( Figure S2A, S2C). This further con rms the conjecture that BnaFAX1-1 may contribute to the formation of seed high oil content in B. napus. Furthermore, to further determine whether BnaFAX1-1, BnaFAX1-2 are conducive to the formation of high biological yield (Fig. 1A), the seedling leaves of four pairs with extremely high-(P281, P542, P125, P257-HBY) and low-biological yield accessions (P319, P276, P131, P81-LBY) were selected for qRT-PCR analysis, and biological yield dry weight per plant for each accession is shown in Figure S2G. The qRT-PCR results showed that the expression levels of BnaFAX1-1 and BnaFAX1-2 in high-biological yield accessions were higher than those in low-biological yield accessions ( Figure S2E, S2F), which is consistent with the GWAS analysis result (Fig. 1A). Overall, to further determine whether BnaFAX1-1 can increase both seed oil content and biological yield, we further characterized the function of BnaFAX1-1.

Subcellular localization and transcript levels of BnaFAX1-1 in B. napus
To determine the subcellular localization of BnaFAX1-1 in plant cells, we tagged BnaFAX1-1 with green uorescent protein (GFP) and expressed the construct under the control of the constitutive Cauli ower mosaic virus (CaMV) 35S promoter ( Fig. 2A). We transiently transfected Arabidopsis protoplasts with the BnaFAX1-1-GFP construct, using AtFAX1-GFP as a marker for chloroplast envelopes. We observed a ring of uorescence at the periphery of chloroplasts, which is consistent with a plastid envelope localization for BnFAX1-1, as seen previously with AtFAX1 (Fig. 2B). We next measured BnaFAX1-1 transcript levels of in seven tissues across ve stages of development (roots, stems, leaves, owers, buds, seeds, and silique pericarp after owering 7, 14, 21, 30, 40 d). We observed the highest expression level for BnaFAX1-1 in leaves and seeds after 40 days of owering (Fig. 2C). We had previously determined that AtFAX1 was mainly expressed in leaves, but not in seeds . This result suggested that BnaFAX1-1 function may differ from that of AtFAX1, which did not contribute to seed oil accumulation in Arabidopsis.
Increased biological yield and seed oil production by BnaFAX1-1 overexpression in Arabidopsis Next, we phenotyped Arabidopsis lines overexpressing BnaFAX1-1 and compared their growth to wild-type (WT) plants. We noticed that all Arabidopsis BnaFAX1-1 overexpressing lines were slightly larger and produced more biomass than WT plants (Fig. 3B, Table 1). After reaching reproductive maturity, overexpression lines were signi cantly bigger than the WT, with thicker in orescence stalks and more siliques (Fig. 3C, Table 1). A detailed analysis of different tissues and organs in transgenic and WT plants grown for 7 weeks revealed that plant height, rosette fresh and dry weight, fresh and the dry weight of biological yield was signi cantly higher in overexpressing lines relative to WT plants (Table 1). Likewise, stem fresh weight and stem diameter in the transgenic lines were signi cantly increased compared to WT plants. Furthermore, we observed an increase in seed yield per plant in overexpression lines, largely due to an increase in silique number per plant (Table 1). We also determined the total lipid contentof mature seeds, which indicated that overexpression lines accumulated more total lipid content relative to the WT (Fig. 3D). Collectively, these results indicate that the heterologous overexpression of BnaFAX1-1 in Arabidopsis promoted plant growth and development, and led to an increase in seed oil production. 174.9 ± 54* 241.7 ± 51** 172.0 ± 42* *P < 0.05, **P < 0.01, Student's t-test (n = 6-10 ± SD). BYAG: biological yield of above ground organs.
Increased biological yield, gibberellin and leaf lipid contents in B. napus plants overexpressing BnaFAX1-1 To test the effect of BnaFAX1-1 overexpression in B. napus on biomass accumulation, we analyzed the growth kinetics of three independent BnaFAX1-1 overexpression lines selected at random (OE#17, OE#19 and OE#21) (Fig. 4A). We grew all plants hydroponically in Hoagland nutrient solution for 32 d. All BnaFAX1-1 overexpression lines were larger and produced more biomass than their non-transgenic WT control (Fig. 4B). This increase in leaf biomass was re ected in all phenotypes measured: leaf fresh /dry weight and leaf size (including leaf length, leaf width and leaf area; Fig. 4C, 4D). Compared to the WT, overexpression of BnFAX1-1 also resulted in a signi cant increase in total root length, root area, root volume, root fresh and dry weight (Fig. 4D). Overexpression of BnaFAX1-1 in B. napus therefore promoted plant growth.
Toward the identi cation of the potential mechanism linking FAX1 and B. napus growth and biomass improvements, we quanti ed phytohormone contents in the leaves of two transgenic lines (OE#19 and OE#21) and their WT using liquid chromatography followed by tandem mass spectrometry (LC-MS/MS). We observed that gibberellic acid A4 (GA4) accumulated to signi cantly higher levels in transgenic leaves relative to the WT (Fig. 5A). By contrast, the contents of indole-3-acetic acid (IAA), salicylic acid (SA), and jasmonic acid (JA) were similar in the overexpression lines and the WT ( Figure S3). GA4 is a bioactive gibberellin that plays critical roles in plant growth and development (Eriksson et al., 2000). To explore the reason behind the increase in GA4 content in the transgenic lines, we performed transcriptome sequencing from leaves of the transgenic lines (OE#19 and OE#21) and WT. We discovered that the GA4 biosynthetic genes COPALYL DIPHOSPHATE SYNTHASE (CPS), KAURENOIC ACID OXIDASES (KAOs) and GA20 OXIDASE (GA20OX) were more highly expressed in the transgenic lines relative to the WT (Fig. 5B). We validated these results by RT-qPCR (Fig. 5C). These results indicate that overexpression of BnaFAX1-1 led to up-regulated GA4 biosynthesis, which may in turn contribute to biological yield increase in B. napus.
To investigate the consequence of BnaFAX1-1 accumulation in the two selected overexpression lines above on membrane lipid contents, we analyzed lipids from 32-d-old leaves using LC-MS/MS (Fig. 6). We observed a higher lipid content for phosphatidylcholine (PC) and phosphatidylethanolamine (PE) in the leaves of OE#19 and OE#21 plants when compared to WT (Fig. 6). These results revealed that BnaFAX1-1 share the same role as AtFAX1 when ectopically expressed in leaves for the regulation of leaf lipid and biomass accumulation.

BnaFAX1-1 enhances biological yield and seed yield of B. napus plants grown in the eld
To determine if the phenotypes seen in BnaFAX1-1 overexpression lines extended to the eld, we sowed seeds for WT and OE lines in a randomized eld plot design using four plots for each OE line and WT. We investigated growth characteristics in plants at owering stage (grown for 175 d). OE plants were clearly bigger and taller compared to WT, and had produced more leaves on the main stem in the same growth period (Fig. 7A, 7C). In addition, OE plants showed larger leaves at the same position relative to WT plants, which was re ected in increased leaf length, leaf width and leaf area (Fig. 7B, 7C). Lastly, OE lines produced thicker main stems than WT plants; only chlorophyll content and the photosynthetic rate of OE lines were similar to those of the WT (Fig. 7C).
We next harvested mature plants from the eld to carry out additional measurements. The BnaFAX1-1 OE lines were signi cantly taller than WT plants and bore more effective branches (i.e. branches bearing seeds) per plant (Fig. 8A, 8D-c). Although we observed no differences in the length of the main in orescence between the WT and transgenic plants, the OE lines did exhibit more siliques per main in orescence than in WT plants (Fig. 8B). Similarly, total silique number was signi cantly greater in all OE lines relative to WT (Fig. 8D-f), as were silique length (Fig. 8C) and number of seeds per silique. Together, these results revealed the greater seed yield per plant and biological yield in all OE plants (Fig. 8D), possibly by increasing the number of effective branches and siliques per plant. We did not observe such phenotypes when we overexpressed AtFAX1 in Arabidopsis in our previous work . Therefore, BnaFAX1-1, unlike AtFAX1, may play a vital role in improving seed yield and biological yield in B. napus.
BnaFAX1-1 enhances B. napus seed oil production and improves oil quality We examined total lipid content in seeds from OE lines and WT at 30 d and 45 d after owering, as well as in dry seeds following harvest. Overexpression of BnFAX1-1 in B. napus resulted in an increase in total seed lipid contents at all development stages tested relative to WT (Fig. 9A). We also determined the range of TAG molecular species and total TAG content in BnaFAX1-1 OE lines and WT mature dry seeds. We saw a signi cant rise in the content of many TAG molecular species (TAG 50:2, 50:3, 52:2, 54:2, 54:3, 54:4, 54:5, 56:2, 56:4) and total TAG content in BnaFAX1-1 OE transgenic plants compared to WT (Fig. 9B, C). In addition, an analysis of fatty acid composition of mature dry seeds grown in the eld revealed that oleic acid (C18:1) was signi cantly increased in the transgenic lines, whereas palmitic acid (C16:0), arachidic acid (C20:0) and eicosenic cis (C20:1) were signi cantly reduced. Stearic acid (C18:0), linoleic acid (C18:2) and linolenic acid (C18:3) contents were similar in the OE lines and WT (Fig. 9D). These results demonstrated that overexpression of BnaFAX1-1 effectively increased seed oil production and oleic acid content. These results therefore revealed that BnaFAX1-1 may have important application value in B. napus molecular breeding to improve seed oil content, oil quality, seed yield and biological yield. . Enhancing both seed oil production and seed oil quality is a major goal in B. napus breeding, which can be accomplished by improving seed yield, seed oil content and biological yield, separately or in combination. The present study provides a possible and simple solution to simultaneously improve seed oil production, oil quality, seed yield, seed oil content and biological yield in B. napus by raising the expression levels of a single gene. We identi ed the potential key gene BnaFAX1-1 through the identi cation of putative chloroplast-localized transporters whose encoding genes mapped close to signi cant SNPs from a previous GWAS of 520 B. napus accessions for biological yield. Increasing expression of BnaFAX1-1 improved biological yield and seed oil production, especially oleic acid content, which has not been observed in previous studies.

Discussion
Chloroplasts are the main site of photosynthesis and plant fatty acid production. Recent studies revealed that altering carbon metabolism in the chloroplasts of transgenic oil crops improved biological yield and seed oil production. We hypothesized that export of photosynthates from the chloroplast may contribute to improving biological yield and seed oil production. We therefore focused here on genes encoding proteins predicted to locate to the chloroplast membrane, leading to a rst selection of 29 candidate genes (Table S2). From further analysis of these 29 genes, we identi ed a gene encoding a putative transporter: BnaA07g17240D (BnaFAX1-1). The Arabidopsis ortholog of BnaFAX1-1 is the putative fatty acid exporter AtFAX1, which is crucial for biological yield. Both biological yield and seed oil production increased in 35S:BnaFAX1-1 transgenic B. napus plants grown in the eld relative to WT plants, highlighting the potential of this gene to improve seed oil production in B. napus (Fig. 4, Fig. 6-7).
Furthermore, we determined that the expression of GA biosynthetic genes was up-regulated, in agreement with the higher GA4 levels measured in the OE lines (Fig. 5). GA biosynthesis is derived from geranylgeranyl diphosphate (GGDP), which is synthesized in the chloroplast. GGDP is then converted to ent-kaurene by ent-COPALYL DIPHOSPHATE SYNTHASE (CPS) and ent-KAURENE SYNTHASE (KS). entkaurene is then exported from the chloroplast to complete GA biosynthesis. Fatty acid export may affect carbon metabolism in plastids and further affect ent-kaurene contents or export from plastids. GA biosynthesis was reported to improve plant biological yield (Eriksson et al. 2000). Therefore, BnaFAX1-1 may indirectly affect GA biosynthesis, resulting in the observed increase in biological yield and seed yield, as seen in BnaFAX1-1 overexpression lines.
Increasing seed yield and seed oil content are two major approaches to enhance seed oil production. In B. napus, seed yield is largely determined by three yield component traits: silique number per plant, seed number per silique, and the weight per thousand seeds (Lu et al., 2017). In the present study, the increase in seed yield measured for B. napus BnaFAX1-1-OE lines was mainly due to silique number per plant and seed number per silique, as we detected no differences in the weight per thousand seeds between BnaFAX1-1-OE and WT plants (Fig. 8C). Seed oil content is another essential contributor to seed oil production, which constitutes the basis for B. napus economic importance. Overexpressing BnaFAX1-1 in Arabidopsis and B. napus led to a signi cant rise in seed oil content relative to WT (Fig. 3D, 9A-9C). Therefore, overexpressing BnaFAX1-1 had great potential practical value to increase seed oil production in B. napus.
In our previous study, Arabidopsis lines overexpressing FAX1 showed a pronounced increase of lipids in owers and leaves . However, the fax1 loss of function mutation had no effects on seed oil content or seed yield . In the green alga Chlamydomonas (Chlamydomonas reinhardtii), the overexpression of either CrFAX1 or CrFAX2 resulted in the accumulation of intracellular TAG (Li et al., 2019). Recently, AtFAX2 and AtFAX4 were reported to be seed-speci c transporters mediating seed embryo fatty acid export for seed oil content accumulation in Arabidopsis (Li et al., 2020). These results reveal that the tissue-speci city of FAX expression may contribute to lipid accumulation in speci c tissues. Notably, we determined that BnaFAX1-1 is highly expressed in siliques during the seedlling stage in a cultivar with high seed oil content, when compared to another cultivar with low seed oil content, indicating a correlation between BnaFAX1-1 function and the regulation of seed oil content and seed yield (Fig. 1C). The present study reveals that BnaFAX1-1 mainly mediated seed plastid fatty acid export for the accumulation of seed TAG during the seed-lling stage, leading to the measured increase in seed oil content in the overexpression lines. By contrast, AtFAX1 had no effect on seed TAG accumulation in Arabidopsis, as AtFAX1 is expressed at low levels during the seed-lling stage.
Oleic acid (C18:1) is an important unsaturated fatty acid component of B. napus oil with high nutritional value and good thermal stability. Oil with high oleic acid content can reduce the risk of cardiovascular disease in overweight individuals (Rudkowska et al., 2006) and effectively prevent arteriosclerosis (Nicolosi et al., 2004). In addition, it is highly resistant to oxidization and degradation at high temperature and enjoys a long shelf life (Talcott et al., 2005). Such oleic acid-rich oils also emit little to no smoke when heated to high temperatures and reduce cooking times (Miller et al., 1987), display high stability during frying, and imbue frying products with excellent aroma. Aside from cooking applications, oleic acid is also a raw material for biodiesel production (Piazza and Foglia, 2001). In the present study, we determined that oleic acid (C18:1) content signi cantly increased in BnaFAX1-1-OE seeds, while other plastid-derived fatty acids remained unchanged relative to WT (Fig. 9C). In our previous study, AtFAX1 affected the content of plastid-derived fatty acids in Arabidopsis leaves and owers, but not in seeds . Arabidopsis ATP-BINDING CASSETTE A9 (ABCA9) mediated fatty acids and acyl-CoA import into the endoplasmic reticulum for TAG accumulation. However, the fatty acid composition and contents were not affected by a loss of function in AtABCA9 (Kim et al., 2013). Our results suggest that BnaFAX1-1 may selectively mediate the export of speci c plastid-derived fatty acids, and will be the subject of future work.
In conclusion, we identi ed a fatty acid export protein, BnaFAX1-1, which mediates fatty acid export from plastids in developing seeds. The overexpression of BnaFAX1-1 signi cantly upregulated seed oil content, oil quality, seed yield and biological yield in B. napus. BnaFAX1-1 overexpression lines resulted in an upregulation of GA4 biosynthesis, indicating that BnaFAX1-1 overexpression may in uence GA biosynthesis, leading to the observed increase in biological yield and seed oil production. Furthermore, BnaFAX1-1 contributed to the accumulation of oleic acid, an unsaturated fatty acid of high economic value, in seeds, thus improving oil quality. The present study provides an important solution to simultaneously improve rapeseed seed oil production, seed oil content, seed yield, biological yield and seed oil quality by modulating the expression of BnaFAX1-1. We propose that BnaFAX1-1 should be a potential target for B. napus molecular breeding in the future.

Experimental Procedures
Phylogenetic analyses, protein properties and sequence analyses of the FAX family We downloaded the sequences of FAX proteins from Arabidopsis (Arabidopsis thaliana), eld mustard (Brassica rapa), wild cabbage (Brassica oleracea), and B. napus to generate a on multiple protein sequence alignments using the integrated MUSCLE program in MEGA7.0 (Kumar et al., 2016). We also generated a phylogenetic tree of the FAX family with MEGA7.0 software using the Neighbor-Joining (NJ) method and bootstrap analysis (1,000 replications

Vector constructs and plant transformation
We PCR-ampli ed the complete coding sequence (CDS) of the BnaFAX1-1 gene from B. napus cDNA using the primer pair BnaFAX1-1 FP(XbaI)+ BnaFAX1-1 RP(SacI) (primers used in this study are listed in Table S4). We then subjected the PCR product to restriction digest with XbaI and SacI and ligated the puri ed digested product into the pCAMBIA2301M vector (a modi ed version of pCAMBIA2301 available in our laboratory) to generate pCAMBIA2301-35S:BnaFAX1-1. We introduced the pCAMBIA2301-35S:BnaFAX1-1 construct into Agrobacterium (Agrobacterium tumefaciens) strain GV3101 for transformation of Arabidopsis Col-0 accession by the oral dip method (Clough and Bent, 1998). We selected positive transgenic by germinating seeds on Murashige and Skoog (MS) medium supplemented with 10 mg/L Basta. We also transformed B. napus (Westar cultivar) hypocotyls with Agrobacterium bearing the pCAMBIA2301-35S:BnaFAX1-1 construct based on a protocol described previously (Lu et al., 2013). We selected transgenic calli for growth on MS medium supplemented with Basta, and identi ed transgenic plants by PCR using BnaFAX1-1-specific primers (F35S3ND+ BnaFAX1-1 RP(SacI), and BnaFAX1-1 FP(XbaI)+NOS5ND) (Table S4). We then transferred positive 35S:BnaFAX1-1 transgenic plants to soil for seed setting and phenotyping.

Plant growth conditions and trait measurements
For Arabidopsis, we surface-sterilized seeds from Col-0 and homozygous lines overexpressing BnFAX1-1 (BnaFAX1-1-OE) and sowed them on half-strength MS medium agar plates supplemented with 1% sucrose. We incubated plates in the dark at 4°C for 2 d before releasing them in a plant incubator set to a light/dark photoperiod of 16 h light/8 h dark. After 7 d, we transferred seedlings to soil and allowed them to grow in the incubator under the same conditions. After 7 weeks, we measured a number of phenotypes on all plants: stem fresh weight (mg/cm; 1 cm from the bottom of 2 nd internode of the primary inflorescence stem), stem diameter (mm; from bottom part of 2 nd internode of the primary inflorescence stem), silique length, plant height, rosette fresh/dry weight, and fresh/dry weight of biological yield above ground per plant. We also collected the number of siliques per plant, the number of seeds per silique, the weight of 1,000 seeds and the seed yield per plant after seeds maturation.
For B. napus growth in hydroponics, we germinated seeds for B. napus WT (Westar cultivar) and BnaFAX1-1-OE lines in glassware covered with three layers of wet lter paper. We .. Brie y, we mixed samples in 500-750 μL 2-propanol/water/concentrated hydrochloric acid (2: 1: 0.002 v/v/v) for extraction and it was shaken at 100rpm for 30min at 4℃.We then added 1mL dichloromethane, and it also was shaken at 100rpm for 30min at 4℃, then centrifuged the mixture and collected the supernatant. We repeated this extraction procedure three times, combined all supernatants and dried them under nitrogen ow. Finally, we added 200 µL of methanol/0.1% formic acid aqueous solution (1: 1 v/v) to resuspend the pellet, ltered it through an organic lter and placed each sample into the injection tube. We analyzed phytohormones by liquid chromatography followed y tandem mass spectrometry (LC-MS/SM) (QTRAP 6500+) using the MRM approach described by Pan et al (2010) Brie y, LC uses a binary solvent system. The mobile phase is methanol and 0.05% formic acid. We selected an Eclipse plus C18 (5μm, 2.1 * 150 mm) chromatographic column. The ow rate was controlled at 300 µL / min and the column temperature was 30°C, 10 µL per injection. We used a gradient elution, with the initial gradient of methanol of 10%, held for 2 min, and gradually increased to 10 min and maintained at 90% for 5 min. At 15.1 min, we reduced methanol to the initial gradient and held for 7 min. We added external phytohormone standards for gibberellic acid, abscisic acid, indole-3-acetic acid, salicylic acid and jasmonic acid to calculate the level of each phytohormone in the samples.

Lipid analyses
We extracted lipids from 32-d-old leaves and analyzed lipid content by LC-MS/MS using the method reported previously (Lu et al. 2018). We determined the levels and molecular species for phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), phosphatidylinositol (PI), phosphatidylserine (PS), phosphatidic acid (PA), diacylglycerol (DAG) and triacylglycerol (TAG) as previously described (Welti et al., 2002). For each experiment, we sampled six plants in six technical replicates; we used 50-100 mg fresh leaf tissue to extract lipids.
We quanti ed fatty acid composition using a gas chromatograph coupled with a ame ionization detector (GC-FID). We immersed 3-5mg dry seed samples in 1.5 mL methanol containing 1.5% H 2 SO 4 and 0.01% BHT. We then added 25 µL heptadecanoic acid triglyceride (C17: 0 TAG) with 5.4 µmol/L in tubes as internal standard. We incubated the tubes in a water bath at 90°C for 1 h before allowing them to cool to room temperature. We then added 1 mL H 2 O and 1 mL chromatographic hexane, mixed well, and centrifuged the samples at 1,000 rpm for 10 min. Subsequently, we transferred 0.8 mL of the upper phase to a new glass tube and dried it under nitrogen ow. Finally, we added 0.4 mL hexane to dissolve fatty acid methyl esters, and injected 1 µL of the ester solution into the GC with the detector temperature set to 280°C, oven temperature to 170°C for 2 min, and then increased by 3°C /min up to 210°C.
We determined total lipid content in seeds using a previously described method (Ma et al., 2013). Briefly, we immersed 50 mg dry seeds in 1 mL methanol (chromatography grade) and 2 mL 2% (m/v) NaOH in glass tubes, after which we mashed the seeds with glass rods and vortexed for 10 min on a vortex shaker.
Next, we placed the glass tubes containing the seed mixture in a 60°C water bath for 1 h before allowing the tubes to cool to room temperature. We then added 2 mL chloroform (chromatography grade) and vortexed again for 10 min. After centrifugation of the mixture at 5000g for 5 min at room temperature, we transferred the lower phase (chloroform) to a dry weighed glass tube, added 1mL hexane (chromatography grade) to the remaining upper layer and vortexed again for 10 min. Availability of data and material All data generated or analysed during this study are included in this published article and its supplementary information les