CRISPR/Cas9-Induced fad2 and rod1 Mutations Stacked With fae1 Confer High Oleic Acid Seed Oil in Pennycress (Thlaspi arvense L.)

Abstract

Pennycress (Thlaspi arvense L.) is being domesticated as an oilseed cash cover crop to be grown in the off-season throughout temperate regions of the world. With its diploid genome and ease of directed mutagenesis using molecular approaches, pennycress seed oil composition can be rapidly tailored for a plethora of food, feed, oleochemical and fuel uses. Here, we utilized Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)/Cas9 technology to produce knockout mutations in the FATTY ACID DESATURASE2 (FAD2) and REDUCED OLEATE DESATURATION1 (ROD1) genes to increase oleic acid content. High oleic acid (18:1) oil is valued for its oxidative stability that is superior to the polyunsaturated fatty acids (PUFAs) linoleic (18:2) and linolenic (18:3), and better cold flow properties than the very long chain fatty acid (VLCFA) erucic (22:1). When combined with a FATTY ACID ELONGATION1 (fae1) knockout mutation, fad2 fae1 and rod1 fae1 double mutants produced ∼90% and ∼60% oleic acid in seed oil, respectively, with PUFAs in fad2 fae1 as well as fad2 single mutants reduced to less than 5%. MALDI-MS spatial imaging analyses of phosphatidylcholine (PC) and triacylglycerol (TAG) molecular species in wild-type pennycress embryo sections from mature seeds revealed that erucic acid is highly enriched in cotyledons which serve as storage organs, suggestive of a role in providing energy for the germinating seedling. In contrast, PUFA-containing TAGs are enriched in the embryonic axis, which may be utilized for cellular membrane expansion during seed germination and seedling emergence. Under standard growth chamber conditions, rod1 fae1 plants grew like wild type whereas fad2 single and fad2 fae1 double mutant plants exhibited delayed growth and overall reduced heights and seed yields, suggesting that reducing PUFAs below a threshold in pennycress had negative physiological effects. Taken together, our results suggest that combinatorial knockout of ROD1 and FAE1 may be a viable route to commercially increase oleic acid content in pennycress seed oil whereas mutations in FAD2 will likely require at least partial function to avoid fitness trade-offs.

Keywords: CRISPR; MALDI-MSI; Thlaspi arvense; oilseed; oleic acid; pennycress; polyunsaturated fatty acid; triacylglycerol.

FIGURE 1

Simplified scheme for fatty acid modification in Brassicas. Fatty acid (FA) elongation of 18:1-CoA is carried out by a protein complex containing the FA elongase, FAE1. Lysophosphatidylcholine acyltransferase (LPCAT) shuttles acyl groups between PC and the acyl-CoA pool. The polyunsaturated FA pathway involves desaturation of 18:1 (esterified to PC) by the FA desaturases, FAD2 and FAD3. 18:1-CoA and acyl-PC are the substrates for FAE1 and the FA desaturases, respectively. PDCT (phosphatidylcholine (PC):diacylglycerol (DAG) cholinephosphotransferase) is also named REDUCED OLEATE DESATURATION (ROD1) that modulates polyunsaturated FA (PUFA) content by interconverting PC and DAG, transferring 18:1 into PC for desaturation, and 18:2 and 18:3 into the triacylglycerol (TAG) biosynthetic pathway thereby increasing PUFA content in TAG (Lu et al., 2009). Choline phosphotransferase (CPT) may also mediate the interconversion of DAG and PC but is not shown here for simplicity.

FIGURE 2

Fatty acid composition of seed TAGs for the different lipid mutants versus WT. Graph shows mean weight percents of the different lipid species. Error bars represent standard deviations of the means where n = 2 or 3 biological reps and 3 technical reps. Those without error bars are single samples. See Supplementary Table 1 for values and statistical differences.

FIGURE 3

Plant growth and seed yield comparisons. (A–F) Graphs showing average (A) time until first flower opening, (B) time until final flower opening, (C) duration of flowering, (D) plant height upon senescence, (E) total weight of seeds per plant, and (F) single seed weight. Bars are standard deviations. Different letters above columns represent significant differences between genotypes (one-way ANOVA; Tukey test, p < 0.01). n = 11 biological reps. For (E,F), all seeds from each plant were measured.

FIGURE 4

Amounts of seed germination over a 10-day period of the different lipid mutants versus wild type grown at 22°C. One hundred and fifty seeds from each line were plated on three agar growth media plates (50 seeds per plate; each plate treated as a biological rep). The data for single mutants are graphs as dashed lines, whereas double mutants and wild type (WT) are graphed as solid lines. Error bars represent standard deviations of the means. Values and significant differences can be found in Supplementary Table 2.

FIGURE 5

Mass spectrometry (MS) imaging of selected phosphatidylcholine (PC) molecular species in wild-type and CRISPR mutant T. arvense seeds. MS images are presented as false colored images with each pixel representing an individual scan rastered over the imaged seed section (microscopy image above, scale bar = 500 μm). The color intensity scale ranges from green (low) to red (high) as the mol% value of the metabolite class. Each row of images are set to the same mol% scale to show relative differences in abundance and localization between each seed type.

FIGURE 6

Imaging of selected triacylglycerol (TAG) molecular species (54 vs. 62 carbon TAG) in wild-type and mutant fad2, fae1, fad2 fae1, rod1, and rod1 fae1 T. arvense mature seed sections. MS images are presented as false colored images with each pixel representing an individual scan rastered over the imaged seed section (microscopy image left, scale bar = 500 μm). The color intensity scale ranges from green (low) to red (high) as the mol% value of the metabolite class. Each image mol% value is set individually (color scale below image) to show relative differences in abundance and localization between each seed type.

FIGURE 7

Measurements of triacylglycerol (TAG) molecular species composition (A) and total content (B) within wild-type Spring 32-10 (dark gray) and CRISPR/Cas9 mutant T. arvense seeds fad2 (red), fae1 (green), fad2 fae1 (dark blue), rod1 (orange), and rod1 fae1 (light blue). Error bars represent standard deviations of the means. (n = 5, ±SD, one-way ANOVA; Tukey test, p < 0.01).

FIGURE 8

Measurements of phosphatidylcholine (PC) molecular species composition (A) and total PC content (B) within wild-type Spring 32-10 (dark gray) and CRISPR/Cas9 mutant T. arvense seeds fad2 (red), fae1 (green), fad2 fae1 (dark blue), rod1 (orange), and rod1 fae1 (light blue). Error bars represent standard deviations of the means (n = 5, ±SD, one-way ANOVA; Tukey test, p < 0.01).