PDAT1 genome editing reduces hydroxy fatty acid production in transgenic Arabidopsis

Abstract

The fatty acids content of castor (Ricinus communis L.) seed oil is 80-90% ricinoleic acid, which is a hydroxy fatty acid (HFA). The structures and functional groups of HFAs are different from those of common fatty acids and are useful for various industrial applications. However, castor seeds contain the toxin ricin and an allergenic protein, which limit their cultivation. Accordingly, many researchers are conducting studies to enhance the production of HFAs in Arabidopsis thaliana, a model plant for oil crops. Oleate 12-hydroxylase from castor (RcFAH12), which synthesizes HFA (18:1-OH), was transformed into an Arabidopsis fae1 mutant, resulting in the CL37 line producing a maximum of 17% HFA content. In addition, castor phospholipid:diacylglycerol acyltransferase 1-2 (RcPDAT1-2), which catalyzes the production of triacylglycerol by transferring HFA from phosphatidylcholine to diacylglycerol, was transformed into the CL37 line to develop a P327 line that produces 25% HFA. In this study, we investigated changes in HFA content when endogenous Arabidopsis PDAT1 (AtPDAT1) of the P327 line was edited using the CRISPR/Cas9 technique. The successful mutation resulted in three independent lines with different mutation patterns, which were transmitted until the T4 generation. Fatty acid analysis of the seeds showed that HFA content decreased in all three mutant lines. These findings indicate that AtPDAT1 as well as RcPDAT1-2 in the P327 line are involved in transferring and increasing HFAs to triacylglycerol.

INTRODUCTION

The fatty acid content of the castor bean seeds (Ricinus communis L.) was reported to be 80-90% ricinoleic acid, 18:1-OH, a hydroxy fatty acid (HFA) (1). HFA derived from castor seeds is a useful raw material for various industrial products such as high-quality lubricants, soaps, paints, coatings, plastics, cosmetic materials (skincare), and pharmaceutical raw materials (2, 3). However, the supply of HFAs from castor seeds has been limited because mechanical harvesting is difficult due to the height of castor plants. In addition, castor seeds contain the toxin ricin and an allergen protein (2S albumin), making castor farming unsuitable (4). Consequently, studies have been conducted to increase HFA levels in other plants by introducing several castor genes that synthesize HFAs (5-10).

Accordingly, a study was conducted to identify the production of HFA at the metabolic engineering level by introducing castor oleate 12-hydroxylase (RcFAH12) into Arabidopsis (Arabidopsis thaliana), a model plant for oil crops (11-13). RcFAH12 is an HFA-synthesis gene that synthesizes the ricinoleic acid from the 18:1 fatty acid of phosphatidylcholine (PC) (14). Because HFA is synthesized from PC in castor, castor has a mechanism to transfer the HFA efficiently from PC to triacylglycerol (TAG) (15-17). Therefore, even if RcFAH12 is introduced into Arabidopsis, HFA cannot be efficiently transferred to TAG because Arabidopsis does not have a PC-editing enzyme of HFA (11-13). To successfully produce HFAs in Arabidopsis, fatty acid metabolism must be redesigned to be similar to that in castor. First, the Arabidopsis fae1 mutant, which cannot synthesize 20:1 fatty acid from 18:1 fatty acid, increases the levels of 18:1 fatty acid, a substrate for HFA (Fig. 1A) (18). By expressing RcFAH12 under the control of a phaseolin promoter in a fae1 mutant, the CL37 and CL7 lines that accumulate 17% HFA in seed oil are generated (Fig. 1B) (12). This is the highest content attained by introducing a single RcFAH12 into the Arabidopsis fae1 mutant.

Fig. 1.

Schematic diagram of lipid metabolism pathway in Arabidopsis transgenic plants. TAG is synthesized via acyl-CoA-dependent Kennedy pathway, which uses the enzymes GPAT, LPAT, PAP, and DGAT. Another route is the acyl-CoA-independent pathway through PC, utilizing the enzymes PDAT and PDCT. The yellow and red boxes indicate Arabidopsis and castor genes, respectively. (A) Arabidopsis fae1 seeds could not produce 20:1, 22:1, and hydroxy fatty acids. (B) Hydroxy fatty acid content increased up to 17% when RcFAH12 was expressed in fae1, which is referred to as the CL37 line (12). (C) Introduction of RcPDAT1-2 into the CL37 line, which is referred to as the P327 line, increases hydroxy fatty acids up to 25% (16). (D) To investigate the function of AtPDAT1 in the hydroxy fatty acid accumulation in the presence of RcPDAT1-2, we mutated AtPDAT1 using the CRISPR/Cas9 system in P327 line.

To transfer the HFA from PC to TAG, HFA is cleaved from PC and transferred to the acyl-CoA pool, where it is utilized for TAG synthesis via the Kennedy pathway (19). In Arabidopsis, glycerol-3-phosphate acyltransferase 9 (GPAT9), lysophosphatidic acid acyltransferase 2 (LPAT2), and diacylglycerol acyltransferase 1 (DGAT1) are involved in the synthesis of the TAG by Kennedy pathway in seeds (20-22). In castor, RcGPAT9, RcLPAT2, and RcDGAT2 can transfer HFA into TAG via the Kennedy pathway (6, 7, 23, 24). In addition to the Kennedy pathway, phospholipid:diacylglycerol acyltransferase 1 (PDAT1) can transfer the fatty acids of PC to diacylglycerol (DAG) to form TAG (25). Castor has three PDAT (RcPDAT1-1, RcPDAT1-2, RcPDAT2) but only RcPDAT1-2 (or RcPDAT1A) has substrate specificity for HFAs (16, 17). In previous work, we developed the P327 line, which increased the HFA content from 17% to 25%, by adding RcPDAT1-2 to CL37 (16) (Fig. 1C). The endogenous AtPDAT1, which has a similar function to RcPDAT1-2, exists in P327. In this study, we knocked out AtPDAT1 at P327 using CRISPR/Cas9 and investigated whether AtPDAT1 is involved in the accumulation of HFAs in Arabidopsis transformants (Fig. 1D).

RESULTS

Identification of CL37 and P327 lines via gas chromatography analysis

Expression of RcFAH12 in a fae1 mutant of Arabidopsis led to the accumulation of the HFA content up to 17% in the seed oil (called the “CL37 line”) (12); however, this level is substantially lower than that of castor seed oil, which contains 80-90% ricinoleic acid. To increase the HFA content, RcPDAT1-2 was introduced into CL37, and the HFA content increased up to 25% because of the efficient transfer of HFAs from PC to TAG (called the “P327 line”) (16, 17).

Before the experiment, the fatty acid composition of the CL37 and P327 lines was analyzed. The proportion of HFAs (18:1OH + 18:2OH) was 16.2% in the CL37 line and 24.0% in the P327 line.

Selection of AtPDAT1 editing lines and analysis of mutation pattern

To investigate whether AtPDAT1 is related to the synthesis of HFAs in the P327 line, we used the CRISPR/Cas9 system to mutate AtPDAT1, which is 73.4% homologous with RcPDAT1-2. AtPDAT1 (AT5G13640) consists of six exons and five introns (Fig. 2A). Two guide RNA (gRNA) targeting the exon 6 were designed to induce mutations in AtPDAT1 (Fig. 2A). Both gRNAs were expressed under the control of the U6-26 promoter, and Cas9 was expressed under the control of the ubiquitin promoter (Fig. 2B). To confirm that the two gRNAs did not target RcPDAT1-2, the RcPDAT1-2 forward and reverse complement sequence were aligned with the two gRNAs using the clustal W method in the MegAlign program. As a result, 9-11 mismatches were identified between the RcPDAT1-2 forward sequence and the two gRNAs, and 7 mismatches between the reverse complement sequence and two gRNAs (Supplementary Fig. 1). It is predicted that the two gRNA designed in this experiment will target only AtPDAT1 without targeting RcPDAT1-2. The CRISPR/Cas9 vector, including the two gRNAs, was transformed into the P327 line, and T1 transformants (five individuals) with kanamycin resistance were selected (Fig. 2B). Five T1 transgenic plants were advanced to the next generation. However, only two lines that showed a noisy peak in the gRNA target region of AtPDAT1 by Sanger sequencing were advanced to the T3 generation (data not shown). Sanger sequencing of the T3 plants revealed three mutations at gRNA1 and gRNA2 target sites (Fig. 2C). In P327-atpdat1 3-2-2 and 3-2-7 lines, an 8bp deletion (GTAACGGG) occurred near the gRNA2 position. The P327-atpdat1 5-3-2, 5-3-6, and 5-3-7 lines showed the same mutation pattern, with a 1bp (T) insertion around gRNA1 and a 1bp (G) deletion around gRNA2. In P327-atpdat1 5-3-1, 5-3-3, 5-3-10, and 5-3-13 lines, a 1bp (A) insertion around gRNA1 and a 1bp (G) deletion around gRNA2 were observed (Fig. 2C). PDAT has a catalytic S-D-H triad (S254, D573, H626), which is one of the typical features of lecithin:cholesterol acyltransferase (LCAT) (26-28). Comparing the mutated AtPDAT1 amino acid of the P327-atpdat1 lines with wild-type AtPDAT1, P327-atpdat1 3 lines have a premature stop codon at the 647 amino acid, and P327-atpdat1 5 lines have a mutation in H626 of the S-D-H traid and a premature stop codon at the 628 amino acid (Fig. 2D). These results suggest a loss of function of AtPDAT1 in P327-atpdat1 lines. To determine whether mutation of AtPDAT1 affects the HFAs content in the progeny of these lines, the generations were advanced to the T4 generation.

Fig. 2.

Knock-out of AtPDAT1 in P327 lines using CRISPR/Cas9 system. (A) Genomic DNA structure of AtPDAT1. AtPDAT1 has six exons and five introns. Two guide RNA were targeted to exon 6. (B) CRISPR/Cas9 vector map for knock-out of AtPDAT1. spCas9 is controlled by ubiquitin promoter and two guide RNA is controlled by U6-26 promoter. (C) Sanger sequencing result of P327-atpdat1 lines compared with Col-0. There are three mutation types (−GTAACGGG, +T and −G, +A and −G). (D) The protein sequence of AtPDAT1 between Col-0 and P327-atpdat1 lines. Red letters, yellow box, and asterisk (*) indicate altered protein sequence, one part of catalytic S-D-H triad, and stop codon, respectively.

Fatty acid analysis of P327-atpdat1 lines

The fatty acid composition in seeds was determined for each of the three different mutation types in the P327-atpdat1 line. Two independent lines of each of the three P327-atpdat1 mutants were grown until the T4 generation and the fatty acid composition of the T5 seeds was analyzed (Fig. 3). The content of 18:1-OH and 18:2-OH were 18.2% and 3.8%, respectively, in the P327 line but decreased to 14.3-16.9% and 2.5-2.8%, respectively, in the P327-atpdat1 line (Fig. 3A, B). P327 line accumulated total HFA content at average of 22% (Fig. 3C). However, P327-atpdat1 lines showed a decrease in HFA content, ranging from 17.1% to 19.4%. Notably, the average HFA content was slightly higher than that of CL37 (Fig. 3C). All the mutants had higher levels of 18:1 fatty acid compared to the P327 (Fig. 3D). The 18:1 fatty acid content in P327 was 23.4%, on average, whereas the corresponding values in the P327-atpdat1 lines were between 29.5% and 33.9% (Fig. 3D). There were no significant differences in 18:2 content between the P327-atpdat1 and P327 (Fig. 3E). The P327 line had an average 18:3 content of 10%, whereas the P327-atpdat1 lines had a slightly lower 18:3 content of 7.6-9.0% (Fig. 3F). These results suggest that AtPDAT1, a transformant endogenous gene, is involved in HFAs accumulation. The increase of HFA content in the P327 line was predicted to be a cooperative effect with AtPDAT1 rather than a single effect of exogenous RcPDAT1-2. Taken together, these results suggest that AtPDAT1 is an essential enzyme for HFA accumulation in the host plant of the transformant.

Fig. 3.

Fatty acid analysis of T5 seeds in six transgenic lines with three different mutation types. Comparison of fatty acid composition between P327 and P327-atpdat1 lines. (A) 18:1-OH fatty acid (B) 18:2-OH fatty acid (C) Total hydroxy fatty acid (18:1-OH + 18:2-OH) (D) 18:1 fatty acid (E) 18:2 fatty acid (F) 18:3 fatty acid. Data are indicated as box and whisker plots (n ≥ 5). Statistical significance is indicated by different letters using one-way ANOVA with Tukey’s multiple comparison test (P < 0.05).

DISCUSSION

The production of HFAs in Arabidopsis was accomplished by the introduction of RcFAH12. A single expression of RcFAH12 in Arabidopsis increased the HFA up to 17% but this is not enough when compared with that of the castor, which contains 80-90% HFA in seed oil (11-13). This is because even though RcFAH12 converts 18:1 of the sn-2 position of PC to 18:1-OH, there is no DGAT or PDAT to transfer HFA into TAG (16, 17, 23). In addition, the 1-HFA-DAG synthesized in the CL37 line is not efficiently transferred to PC (29). Accordingly, when RcPDAT1-2, which synthesizes TAG by transferring HFA at the sn-2 position of PC to DAG, was expressed in the CL37 line, HFA content increased from 17% to 25% (named as “P327”) (16, 17).

In this study, we aimed to elucidate the relationship between the transgene RcPDAT1-2 and endogenous AtPDAT1 in HFA accumulation in the P327 line (Fig. 1). Among the P327-atpdat1 lines, the 3-2-2 and 3-2-7 lines, which had an 8bp deletion not affecting the catalytic triad (S-D-H), exhibited a minimal reduction of HFA content, showing a decrease of 18.8-19.4% compared to P327. Meanwhile, the P327-atpdat1 5-3-2, 5-3-6, 5-3-3, and 5,3-10 lines, where the catalytic triad was disrupted due to indel mutations, showed a further decrease in HFA content to 17.1-18.1% (Fig. 2, 3). These results suggest that AtPDAT1 plays a role in the transfer of HFA from PC into TAG and that RcPDAT1-2 may have additionally enhanced the accumulation of HFA. In other words, the mutation of AtPDAT1 in P327 by CRISRP/Cas9 demonstrated that AtPDAT1, which was previously thought to be a competitor of RcPDAT1-2, works in tandem with RcPDAT1-2 in the synthesis of HFAs. In a previous study, Dauk et al. attempted to investigate the functions of AtPDAT1, AtPDAT2, and AtDGAT1 in synthesizing HFAs in Arabidopsis (30). RcFAH12 was expressed in wild type (WT), EMS mutant of atdgat1, and T-DNA insertion lines of atpdat1 and atpdat2, respectively. As a result, the total HFA content of seeds was 6% in all the transgenic lines with no difference between them (30). Therefore, Dauk et al. concluded that AtPDAT1, AtPDAT2, and AtDGAT1 do not play major roles in the transfer of HFA from PC to TAG (30). However, mutation of AtPDAT1 in the P327 line decreases the HFA content from 22% to up to 17-19% (Fig. 3C). For the other fatty acid composition, the atpdat1-FAH12 line showed a slight increase in 18:1 content and a decrease in 18:2 and 18:3 contents that compared to that of the WT-FAH12 line (30). In this study, the P327-atpdat1 line also exhibited an increase in 18:1 content and a decrease in 18:3 content but there was no difference in 18:2 content (Fig. 3D-F). The difference in the HFA content change in the atpdat1 mutant may be due to the difference in HFA content produced by the transformant used in the experiment and the presence or absence of RcPDAT1-2.

Although research has focused on the introduction of castor genes to enhance HFA production (5, 6, 16, 17), studies on the knockout of Arabidopsis genes competing with castor genes using CRISPR/Cas9 should be conducted simultaneously. Endogenous gene mutations increase the content of target substrates for fatty acid metabolism. For example, the knockout of AtFAE1, which synthesizes 20:1 fatty acid from 18:1 fatty acid, increases HFA content by increasing 18:1, which is a substrate for RcFAH12 (5). To investigate the relationship between transgene RcDGAT2 and endogenous AtDGAT2 in the accumulation of HFA, the atdgat1 mutant was crossed with the CL7 RcDGAT2 line, which expresses RcDGAT2 in the CL7 line (31). The total HFA content was increased to 29% in the atdgat1 CL7 RcDGAT2 line compared to the CL7 RcDGAT2 line (31). These results suggest that RcDGAT2 competes with AtDGAT1 in terms of HFA accumulation. In addition, atpdat1-amiRNA/RcPDAT1A (RcPDAT1-2) vector was introduced into the atdgat1 CL7 RcDGAT2 line to check whether there was a competition between RcPDAT1-2 and AtPDAT1 in HFA accumulation (31). The total HFA content was 32% in atdgat1 CL7 RcDGAT2 atpdat1-amiRNA RcPDAT1A lines, which was not different from that of the atdgat1 CL7 RcDGAT2 RcPDAT1A line. In previous work, when both AtPDAT1 and AtDGAT1 are disrupted in Arabidopsis, they do not survive (32). Therefore, seeds in which AtPDAT1 was strongly suppressed by RNAi in the atdgat1 CL7 RcDGAT2 line would not germinate (31). This is a possible explanation for why HFA was not different in atdgat1 CL7 RcDGAT2 atpdat1-amiRNA RcPDAT1A line (31).

In the future, experiments including (1) seed-specific overexpression of AtPDAT1 in CL37, (2) knockout of AtPDAT1 in CL37 or CL7 lines, and (3) knockout of AtPDAT1 in CL7 RcDGAT2, may indicate the potential involvement of AtPDAT1 in increasing HFA levels in transgenic plants.

MATERIALS AND METHODS

Plant material and growth condition

The CL37 and P327 lines were used as controls (12, 16). For germination, all the seeds were sterilized, treated with 70% EtOH for 1 min and 0.5% (w/v) NaOCl for 5 min, and washed five times with distilled water. Seeds were subjected to stratification at 4°C for 3 days. Seeds were planted in 1/2 MS medium containing 1% sucrose and cultured in a culture chamber at 23°C, 16h/8h light/dark photoperiod. The light intensity was 100 μmol·m⁻²·s⁻¹. Arabidopsis with true leaves were transferred to the soil and grown in a growth chamber under the same conditions as described above.

CRISPR/Cas9 vector construction

Two gRNAs (5’-GCGTCCCGCGCCCTTCCAAC-3’ and 5’-TATATCAGACCCGTTACCTC-3’) targeting AtPDAT1 were designed using the CRISPR/P web tool (33). Each gRNA was cloned into the pICH47751 and pICH47761 vectors, including the Arabidopsis U6 promoter. To create the final vector, pICH47751, pICH47761, pICH47732 (including NOSp::NPTII::OCST), and pICH47742 (including 35Sp::Cas9::NOST) were cloned into the pAGM4723 vector using the golden gate reaction (34). Finally, the 35S promoter of Cas9 was replaced with a ubiquitin promoter to improve Cas9 efficiency.

Arabidopsis transformation

CRISPR vector was transformed into the Agrobacterium strain GV3101. A 300 μl agrobacterium cells were inoculated in 300 ml LB media and incubated overnight at 28°C. The LB media was centrifuged at 3,800 g for 10 min and resuspended in a solution containing 1% sucrose and 0.05% Silwet L-77. The prepared Agrobacterium cell suspension was then transformed into Arabidopsis using the floral dipping method (35). Arabidopsis seeds were placed in 1/2 MS medium containing 1% sucrose and 50 μg/ml kanamycin, and the surviving plants were selected.

Identification of AtPDAT1 mutation in P327 lines

Kanamycin-resistant individuals were planted in the soil, and genomic DNA was extracted from the leaves. A pair of primers (forward primer: 5’-CGTGGCAAGACAAGATTCAA-3’, reverse primer: 5’-CAGCTTCAGGTCAATACGCT-3’) that could cover the target regions of the two gRNAs were designed, and PCR was performed to amplify AtPDAT1. The PCR product was eluted using a PCR purification kit, and Sanger sequencing was performed to confirm the mutation pattern in AtPDAT1.

Fatty acid analysis

Arabidopsis seeds including 5% H₂SO₄ and toluene were incubated in water bath (85°C) for 2 h, after which 1 ml 0.9% NaCl and n-hexane were added, and the mixture was centrifuged at 330 g for 2 min. The supernatant containing fatty acid methyl ester (FAME) was transferred to a 6 mL tube and purged with nitrogen gas. Lastly, 200 μl of n-hexane was added, and the mixture was transferred to the GC vial for analysis. A DB-23 column (30 m × 0.25 mm, 0.25 μm film, Agilent, USA) was used, and analysis was performed using a GC-2030 instrument (Shimadzu, Japan). The GC oven temperature was increased from 190°C to 230°C at a rate of 5°C/min.

Statistical analysis

Statistical analyses were conducted to determine the statistically significant differences. One-way ANOVA with Tukey’s multiple comparison test was conducted using GraphPad Prism. The lowest value was set as letter a, and when there was a difference of P < 0.05, it was displayed in alphabetical order.

Funding Statement

ACKNOWLEDGEMENTS This work was supported by grants from the Mid-Career Researcher Program of the National Research Foundation of Korea (NRF-2020R1A2C2008175) and New Breeding Technologies Development Program (RS-2022-RD009977), and Program of the Rural Development Administration, Republic of Korea.