Identification of triacylglycerol remodeling mechanism to synthesize unusual fatty acid containing oils

Plant storage oils (TAG) can contain over 450 unusual fatty acids with nutritional and industrial value, but engineering their high-level accumulation in major oilseeds has been limited by metabolic bottlenecks, particularly the PC-derived DAG pathway that conflicts with membrane compatibility of unusual fatty acids such as hydroxy fatty acids (HFA). Prior work in Physaria fendleri suggested a TAG remodeling pathway that converts initially formed 1HFA-TAG into 2HFA-TAG without routing HFA through membrane PC. Key unknowns included whether distinct DGATs catalyze the first and second acylation steps, whether sn-2,3-DAG intermediates are utilized, what lipase selectively removes common fatty acids while sparing HFA, and how lipase activity avoids net TAG turnover. This study aims to identify and characterize the enzymes and mechanisms underlying TAG remodeling in P. fendleri and to evaluate their utility for engineering HFA accumulation in transgenic plants.

Plant TAG synthesis proceeds via de novo DAG (Kennedy pathway) and PC-derived DAG. PC is the site of fatty acid desaturation and many modifications, facilitating flux of polyunsaturates into TAG in species like soybean, canola, Arabidopsis, and Camelina. However, many unusual fatty acids disrupt membrane properties, making PC-dependent flux a bottleneck for engineered accumulation, as shown for ricinoleic acid in Arabidopsis and Camelina. Castor bean circumvents this by producing HFA on PC, then using acyl editing and Kennedy pathway DAG to assemble 3HFA-TAG. Prior transcriptomics in P. fendleri showed expression patterns consistent with PC-derived DAG utilization (including PDCT), yet isotopic labeling indicated a remodeling route: PC-derived sn-1,2-DAG forms 1HFA-TAG (HFA at sn-3), which is converted over time into 2HFA-TAG (HFA at sn-1,3), implying a lipase-mediated formation of sn-2,3-DAG followed by re-acylation. Enzymes mediating this remodeling had not been identified.

• Candidate identification: Re-evaluation of P. fendleri seed transcriptomes to identify DGAT1/2/3 and TAG lipase candidates, focusing on expression during seed development.
• Yeast functional assays: Expression of PfeDGAT1/2/3 in TAG-deficient Saccharomyces cerevisiae H1246 to test TAG synthesis; feeding hydrolyzed Physaria oil to assess HFA-TAG production. Microsome preparation from DGAT-expressing yeast for in vitro assays.
• In vitro DGAT assays: Assayed acyl-CoA substrate specificity (18:1-, 18:1OH-, 20:1-, 20:1OH-CoA) with [14C]-labeled dioleoyl-DAG vs [14C]-1HFA-rac-DAG (from [14C]2HFA-TAG). Tested DAG stereochemistry using sn-1,2-diolein vs rac-diolein with [14C]18:1-CoA or [14C]20:1OH-CoA to evaluate sn-1,2 vs sn-2,3 DAG utilization.
• Subcellular localization: Generated GFP fusions of PfeDGAT1, PfeDGAT2, PfeTAGL1; transiently expressed in Nicotiana benthamiana leaves; confocal microscopy for ER localization and co-localization with mCherry-ER marker; in High Oil tobacco lines, co-localized with Nile Blue-stained oil bodies to assess ER vs oil body association.
• Protein–protein interactions: Split-ubiquitin yeast two-hybrid (DGAT1/2 as baits; TAGL1, OBL1, SDP1, DGAT1/2 as preys) and BiFC in N. benthamiana to test interactions and subcellular localization of interacting complexes.
• Lipase biochemical characterization: Expressed truncated His-tagged PfeTAGL1 in E. coli, purified under native conditions; in vitro TAG lipase assays using [14C]triolein across pH, time, and enzyme concentration; competitive assays mixing [14C]triolein with unlabeled triolein, 1HFA-TAG, or 2HFA-TAG to assess acyl selectivity; in vivo expression in yeast with [14C]acetate labeling to monitor TAG, DAG, and FFA.
• RNAi knockdown in P. fendleri: Seed-specific RNAi constructs for PfeDGAT1, PfeDGAT2, PfeTAGL1; plant transformation; T2 seed oil content and fatty acid composition; qRT-PCR validation and assessment of compensatory expression of DGAT1/2, PDAT1, TAGL1.
• Heterologous engineering in Arabidopsis: Seed-specific expression of PfeDGAT1, PfeDGAT2, PfeTAGL1 (and PfeSDP1 control) in RcFAH Arabidopsis background producing HFA; measured total oil, total HFA, whole-seed FA profiles; separated TAG fractions (0HFA-, 1HFA-, 2HFA-TAG) by HPLC and analyzed FA composition by GC-FID.
• Analytics and statistics: TLC/phosphor imaging, GC-FID, HPLC fractionation; statistical tests (one- or two-way ANOVA, multiple t-tests) with significance at p<0.05.

• DGAT activities and substrate/stereoselectivity: PfeDGAT1 and PfeDGAT2 restored TAG synthesis in yeast (DGAT3 did not). In vitro, PfeDGAT2 showed higher activity with HFA-containing substrates and preferentially utilized 1HFA-DAG and 20:1OH-CoA, while PfeDGAT1 preferred non-HFA dioleoyl-DAG. Stereoselective assays showed PfeDGAT1 selectively used sn-1,2-DAG, whereas PfeDGAT2 utilized both sn-1,2- and sn-2,3-DAG with a preference for sn-2,3-DAG.
• Localization and interactions: PfeDGAT1 and PfeDGAT2 localized to ER; PfeTAGL1 localized to ER and puncta around oil bodies (enriched near oil body edges), consistent with ER–oil body junctions. PfeSDP1 and PfeOBL1 coated oil bodies but not ER. Split-ubiquitin Y2H and BiFC demonstrated PfeDGAT1 self-interaction (dimerization) and interaction with PfeTAGL1 in the ER; PfeDGAT2 showed no detectable interactions with tested partners.
• Lipase function and selectivity: PfeTAGL1 is an active TAG lipase (pH optimum ~7), increasing DAG and FFA and reducing TAG in yeast; purified enzyme hydrolyzed [14C]triolein in a time- and concentration-dependent manner. Competitive assays indicated PfeTAGL1 preferentially removes common fatty acids over HFAs from TAG, increasing [14C]triolein degradation in the presence of 1HFA- and even more with 2HFA-TAG. By contrast, PfeSDP1 preferentially removed HFA from HFA-TAG (TAG turnover role).
• Genetic evidence in P. fendleri: Seed-specific RNAi knockdown of PfeDGAT1, PfeDGAT2, or PfeTAGL1 reduced total seed oil and HFA content, confirming roles in HFA-TAG accumulation and TAG remodeling.
• Engineering in Arabidopsis: Seed expression of PfeDGAT1 or PfeDGAT2 in HFA-producing RcFAH Arabidopsis increased total oil and HFA; PfeDGAT1 generally gave larger increases in these heterologous systems. PfeTAGL1 expression alone produced the largest increases in both total oil and HFA, and shifted TAG FA composition toward increased PUFA across TAG classes, consistent with remodeling and re-entry of released FA into PC for desaturation. PfeDGAT2 particularly enhanced 20:1 and 20:1OH incorporation in 2HFA-TAG.
• Mechanistic model: TAG remodeling in P. fendleri proceeds via consecutive actions of PfeDGAT1 (sn-3 acylation of PC-derived sn-1,2-DAG to form 1HFA-TAG), PfeTAGL1 (removal of sn-1 common FA to form sn-2,3-DAG with sn-3 HFA), and PfeDGAT2 (acylation to form sn-1,3 2HFA-TAG). Evidence supports formation of an ER-localized metabolon involving PfeDGAT1 and PfeTAGL1 near oil body–ER junctions.

The study resolves key unknowns underlying TAG remodeling in P. fendleri by demonstrating that two DGATs with distinct substrate and stereochemical preferences catalyze sequential acylation steps, and that a unique ER-localized TAG lipase (PfeTAGL1) interacts with DGAT1 and selectively removes common fatty acids, enabling conversion of 1HFA-TAG to 2HFA-TAG without deleterious TAG turnover. The ability of PfeDGAT2 to utilize sn-2,3-DAG provides the enzymatic basis for the second acylation in the remodeling cycle. Protein localization and interaction data support a metabolon model at ER–oil body interfaces, allowing channeling of specific DAG pools and efficient remodeling. Genetic knockdowns validate in vivo roles of DGAT1, DGAT2, and TAGL1 in oil and HFA accumulation. Heterologous expression in Arabidopsis highlights practical implications: enhancing TAG remodeling components increases both oil yield and HFA content, alleviating feedback inhibition on de novo FA synthesis observed when HFA are inefficiently incorporated into TAG. Together, these findings address the bottleneck of PC-derived DAG in unusual FA accumulation and illustrate how plants can maintain membrane integrity while accumulating unusual FA in TAG.

This work identifies and characterizes the enzymatic machinery of TAG remodeling in Physaria fendleri: PfeDGAT1 (sn-1,2-DAG selective), PfeTAGL1 (ER-localized, common-FA-preferring TAG lipase interacting with DGAT1), and PfeDGAT2 (HFA- and sn-2,3-DAG-preferring). The proposed sequential pathway explains efficient accumulation of 2HFA-TAG while avoiding membrane incompatibility. The components functionally increase oil and HFA levels in engineered Arabidopsis, demonstrating an engineering strategy to control seed oil composition via remodeling. The findings suggest TAG remodeling may operate in other species that accumulate unusual FA at sn-1,3 without PC intermediates, and that co-expression of acyl-selective lipases with DGATs can be leveraged to tailor oil compositions. Future work should assess the prevalence of DAG stereoselectivity among DGATs across species, define the broader interactome and dynamics of ER lipid metabolons, and test remodeling-based engineering in agronomic oilseeds.

The study relies on heterologous systems (yeast, Nicotiana benthamiana, and Arabidopsis) for functional assays, which may not fully recapitulate native substrate pools and metabolon organization. Quantitative in vivo flux through remodeling steps in P. fendleri was inferred from prior isotopic tracing and supported here by enzyme properties and genetics rather than direct time-resolved flux measurements in this study. It remains unclear how generalizable the observed DGAT stereoselectivities are across species; further characterization is needed to establish whether DGAT1/2 from other plants share similar sn-1,2 vs sn-2,3 DAG preferences. Structural mechanisms of selectivity and the precise composition/dynamics of the proposed ER–oil body metabolon were not resolved.