CRISPR/Cas9-mediated targeted mutagenesis of FAD2-1 gene for oleic acids composition in sunflower

Sunflower ( Helianthus annuus L.) is the most widespread plant used for production of oil among all oilseeds cultivated in Turkey in terms of agricultural area and production. Sunflower with high oleic acid content is always desirable because of benefits on health and industrial use. The enzyme FAD2-1 catalyzes the conversion of oleic acid to linoleic acid in sunflower. The fatty acid composition can be enhanced by gene editing of FAD2-1 gene by CRISPR/Cas9 technique which has been applied recently as a new breeding technique to improve productivity and enhance sustainability in agriculture. Hence, CRISPR/Cas9 genome editing systems were utilized to knockout the FAD2-1 gene in order to increase production of oleic acid content. For this purpose, two low oleic sunflower genotypes were transformed with two sgRNA that target FAD2-1 gene. sgRNA expression cassettes were assembled into the binary vector by Golden Gate assembly in a single reaction which is followed by A. tumefaciens -mediated transformation and in vitro germination. Putatively transformed shoots were selected with A. tumefaciens -mediated the optimized kanamycin concentration (100 mg/L) in the medium. The challenges of transformation of sunflower were summarized and possible solutions were proposed. This study indicated that sunflower still can be modified by CRISPR/Cas9 genome editing system to develop high oleic sunflower.


Introduction
Sunflower (Helianthus annuus L.) is one of the world's most significant oilseed crops with various usage areas (Hu et al., 2010). Sunflowers are categorized in terms of fatty acid composition into high oleic (85%), mid oleic (60-65%), and linoleic (low-oleic). The demand for both production and consumption of high oleic sunflowers has increased in the world, as high oleic sunflower genotypes have various advantages in terms of their valuable oil content for industrial use and human health (Kaya et al., 2007). The oil composition of sunflower can be altered by the genetic modifications made on the fatty acid desaturases 2 (FAD2) gene that facilitates the biological conversion of oleic acid to linoleic acid. Oleic acid composition was elevated up to 75% in Pervenets variety of sunflower with the application of dimethylsulphoxide (DMSO) which is a chemical mutagen (Soldatov et al., 1976). Many inbred lines that are derived from the mutant Pervenets had oleic acid composition up to 90% (Fernandez-Martinez et al., 1993;Miller et al., 1987;Zambelli et al., 2015). Besides, Vick & Miller (1996) reported the development of the high and mid oleic sunflower mutants by treatment with Ethyl-Methane Sulphonate (EMS). Similarly, EMS treatment was also performed by Leon et al. (2013b) to develop high oleic acid mutants. The treatment induced point mutations causing amino acid substitutions and premature stop codon (Leon et al., 2013b). On the other hand, the duplication of the FAD2-1 gene leads to the transcription silencing of the gene and consequently, accumulation of oleic acid (Lacombe et al., 2009;Martinez-Rivas et al., 2001). In addition, Schuppert et al. (2006) also reported that high oleic mutant sunflowers were developed by duplication of the FAD2-1 gene and oleoylphophatidyl choline desaturase induction in sunflower genotypes.
Differently, Leon et al. (2013a) developed two high oleic sunflower genotypes by X-rays treatment of the seeds. The treatment resulted in gene insertions in different regions of the genomes that caused the generation of the premature stop codon and thereby accumulation of oleic acid. Besides the application of chemical mutagens and duplication studies, the recently developed CRISPR/Cas9 technique provides a useful, quick, effective and, more importantly, precise genome editing toolbox and has been applied as a new breeding technique to improve productivity and enhance sustainability in agriculture . Thus, CRISPR/Cas9-mediated targeted mutagenesis of the FAD2 gene for the enhancement of oleic acid production has been studied in a variety of plants. For instance, Jiang et al. (2017) edited FAD2 gene in Camelina sativa L. for the improvement of fatty acid composition. Gene silencing of the FAD2 gene resulted in an increase of the oleic acid content from 16% to 50%. Abe et al. (2018) also improved fatty acid composition by knocking out FAD2 gene in rice while Tian et al. (2020) knocked out FAD2-2 gene in tobacco (Nicotiana tabacum L.). Yuan et al. (2019) also accomplished targeted mutagenesis of FAD2A and FAD2B genes in peanut seeds. Moreover, Okuzaki et al. (2018) also edited FAD2 gene in Brassica napus L. and accomplished the fatty acid composition to be significantly increased. Al Amin et al. (2019) developed high oleic soybean by silencing of FAD2-2 gene with 21% mutation efficiency.
Since the previous studies indicated that knocking out FAD2 gene by CRISPR-Cas9 resulted in elevation of the fatty acid content in significant number of plants, the aim of the study concerned the application of this gene editing technique to knock out FAD2-1 gene in sunflower (H. annuus L.).

Plant material
Sunflower (H. annuus L.) seeds with low oleic fatty acid content were used as the starting material in the present study. The seeds of H. annuus L. lines (2453-A and Deray) were provided by the Republic of Türkiye Ministry of Agriculture and Forestry, Trakya Agricultural Research Institute, Edirne, Turkey. The seeds were planted in soil and grown in a growth chamber at Marmara University, Faculty of Arts and Sciences, Biology Department, Plant Biotechnology Laboratory. The growth conditions were optimum at 25-26°C, 60% humidity, a light intensity of about 6000 lux, and a photoperiod with 16 hours of light and 8 hours of darkness.

Vector construction and cloning to E. coli
The assembly of the expression cassettes into pYLCRISPR/Cas9 binary plasmid was carried out according to Golden Gate Assembly method (Engler et al., 2009) and pHDE-35SCas9-mCherry plasmid by Gibson Assembly method (Gibson et al., 2008). The adaptor sgRNA primers were designed by adding 5'-GTCA-3' (for U3 promoter) and 5'-ATTG-3' (for U6 promoter) onto the 5' end of the sgRNA sequences that were designed and obtained. Digestion/ligation reaction was performed for each sgRNA expression cassette.
pYLsgRNA-AtU3d/LacZ plasmid (Addgene with #66201 catalog number) and pYLsgRNA-AtU6-1 plasmid (Addgene with #66202 catalog number) were used in the reaction. The initial concentrations of pYLsgRNA U6 and pYLsgRNA U3 cassette plasmids were 226.12 ng/µl and 155 ng/µl respectively. The cassette plasmids were diluted to 20 ng/µl (1:11 and 1:9 dilutions). The reaction mixes for 10 µl reactions were prepared in microcentrifuge tubes (0.2 mL) for both U3 and U6 primer mixes. After the reaction was done, the ligation products were 1:10 diluted and used in the first PCR with U-F (5' CTCCGTTTTACCTGTGGAATCG 3') and gR-R (5' CGGAGGAAAATTCCATCCAC 3') primers. PCR products were directly used as the template DNA for the second PCR (overlapping PCR) to make the expression cassettes at the end. The arrangements of Pps/Pgs primer pairs for the two cassettes were as follows: Pps-R/Pgs-2, Pps-2/Pgs-L for the use of pYLCRISPR/Cas9P35S-N and Pps-PmeI-F/Pgs-2, Pps-2/Pgs-SpeI-R for the use of pHDE-35SCas9-mCherry vector. The sequences of the primers used were given in the Table 1. Each PCR product (10 µl) were run on 2% agarose gel electrophoresis to verify the size. The expected size of sgRNA cassettes was 495 bp (360bp + 135 bp). DNA Clean & Concentrator -5 Kit (Zymo Research, catalog no: D4013) was used for the purification of each overlapping PCR product. U3-sgRNA and U6-sgRNA expression cassettes were obtained in the second (Overlapping) PCR and used as DNA templates in Gibson Assembly protocol. First, the second PCR products were cut at BsaI recognition site and combined by PCR reaction. The PCR product was purified with DNA Clean & Concentrator -5 Kit and DNA concentration was quantified using U3/U6-sgRNA PCR product.
NanoDrop spectrophotometer. Another PCR reaction was performed to generate sticky and blunt ends for SpeI and PmeI restriction enzymes on the previous U3/U6-sgRNA PCR product. The PCR product was purified and quantified. NEBioCalculator (https://nebiocalculator) was used to calculate the required mass of insert DNA. According to the result given by NEBioCalculator, 30 ng DNA insert was used for 100 ng vector. The U3/U6 sgRNA fragment was cloned between the PmeI and SpeI sites in pHDE-35SCas9-mCherry CRISPR/Cas9 vector (Gao et al., 2016) by Gibson Assembly. Similarly, U3-sgRNA and U6-sgRNA expression cassettes were prepared in Second (Overlapping) PCR and used as DNA templates in Golden Gate Assembly protocol. The U3/U6 sgRNA fragment was cloned at BsaI sites into pYLCRISPR/Cas9P35S-N vector (Ma et al., 2015) by Golden Gate Assembly. The PCR reaction for the ligation of DNA insert into the vector was performed. The digestion/ligation mix was prepared. The competent E. coli cells were prepared according to Chung et al. (1989). 100 µl competent cells and 4 µl ligation product were mixed in a microcentrifuge tube and mixed gently by flicking the bottom of the tube with finger. The mixture was kept on ice for 5 minutes and then, it was incubated at 42°C in water bath for 60 seconds and transferred on ice for 5 minutes. Then, the mixture was added into 1 mL LB (Luria- Bertani) liquid medium and incubated for 1 hour at 37°C while shaking at 200 rpm. After incubation, the mixture was centrifuged at 800 x g for 5 minutes. The supernatant was poured out, and the pellet was re-suspended in the remaining supernatant. The E. coli cells were spread on LB agar supplemented with 50 mg/L kanamycin and 40 µl of both IPTG (100 µM) and X-gal (20 mg/mL). The agar plates were incubated at 37°C for 24 hours. The presence of insert DNA in the plasmid construct was determined by colony PCR (Woodman, 2008). After the verification of the presence of DNA insert by gel electrophoresis, a sizeable colony was picked with a 100 µl pipette tip and added into 5 mL LB medium supplemented with 5 µl kanamycin (50 mg/mL In summary, 100 µl competent cells A. tumefaciens were thawed in a 1.5 mL microcentrifuge tube on ice. Plasmid DNA (~1.0 µg) was added into the competent cells and mixed by tapping the tube gently. The plasmid DNAcompetent cells mixture was incubated in liquid nitrogen for 5 minutes and then, incubated in 37°C water bath for 5 minutes. The mixture was tapped gently and kept on ice for 2 minutes. Additional 1 mL liquid LB was added to the tube and incubated for 2 hours at 28°C with shaking (200 rpm). The mixture was centrifuged at 3000 rpm for 5 minutes. Approximately 1 mL of supernatant was removed and the bacterial pellet was re-suspended in the remaining supernatant. The A. tumefaciens cells were spread on LB agar plates supplemented with antibiotics. Gentamycin (30 µg/mL) and kanamycin (100 µg/mL) were used for EHA105 strain. Streptomycin (200 µg/mL) and kanamycin (100 µg/mL) were used for LBA4404 strain. The agar plates were incubated at 28°C for 2 days. The glycerol stock of confirmed A. tumefaciens transformants was prepared by mixing the fresh bacterial suspension with 30% glycerol in 1:1 ratio. The bacterial glycerol stock was stored at -80°C.

Sunflower transformation
The frozen aliquot of A. tumefaciens which was stored at -80°C, was added into 50 mL of LB broth medium including 50 µg/mL kanamycin and swirled. The culture was covered loosely with sterile aluminum foil. The bacterial culture was incubated overnight at 28°C while shaking (230 rpm) in the shaking incubator. After overnight incubation, the bacterial culture was transferred into 50 mL Falcon centrifuge tube and centrifuged at 2800 x g at 4°C for 20 minutes. The supernatant was discarded without disturbing the bacterial pellet and the bacterial pellet was dissolved in 50 mL of liquid half-MS medium. The density of the bacterial culture was measured at OD600 in spectrophotometer (Nanophotometer® 330 by Implen). The bacterial culture was diluted to 0.6 OD600 with liquid half-MS medium and 150 µM acetosyringone (Sujatha et al., 2012). The sunflower seeds were sterilized according to Dagustu et al. (2010). The sterilized seeds were split into two halves, longitudinally along the embryo axis. The half-seeds were added into A.
tumefaciens culture with 0.6 OD600 and subjected to vacuum infiltration for 10 minutes at 60 kPa. The half-seeds were placed on filter paper for drying. Thereafter, the half-seeds were co-cultured with A. tumefaciens cells (Sujatha et al., 2012). Following the sunflower transformation, the halfcut seeds which were inoculated with A. tumefaciens cells, were removed from the bacterial culture and blotted dry on a filter paper and cocultivated in co-cultivation medium [MS (Murashige and Skoog, 1962) containing 0.5 mg/L BAP, 0.2 mg/L NAA, and 20 mg/L acetosyringone] in dark at 26  2C for 2 days. Following cocultivation, the half-cut seeds were washed with 25 mL ½ MS broth medium including 250 mg/L cefotaxime, three times for 5 minutes each for elimination of A. tumefaciens cells. Then, the halfseeds were transferred to Selection 1 medium (SM1, MS medium containing 0.2 mg/L BAP, 0.01 mg/L NAA, 0.82 mg/L AgNO3, 250 mg/L cefotaxime, and 1 mg/L kanamycin) for shoot induction and the adaxial surface of the seeds was placed in contact with the medium surface. After 2 weeks of cultivation, the shoots were transferred to Selection 2 medium (SM2, MS medium containing 0.2 mg/L BAP, 0.82 mg/L AgNO3, 250 mg/L cefotaxime, and 10 mg/L kanamycin) for two weeks which is followed by their transfer to Selection 3 medium (SM3, MS medium containing 0.1 mg/L BAP, 0.82 mg/L AgNO3, 250 mg/L cefotaxime, and 50 mg/L kanamycin) for shoot elongation. The elongated shoots were transferred to rooting medium (RM, MS medium containing 1.0 mg/L NAA and 250 mg/L cefotaxime) and cultured for 20-25 days. The rooted shoots were transferred to soil and placed in the greenhouse to be acclimated. The plantlets were maintained under high humidity for one week (Sujatha et al., 2012).

Analysis of putative transformed shoots by PCR
For molecular analysis of putative transformed shoots, the shoots were harvested. Doyle & Doyle (1987) protocol was used for the genomic DNA extraction from the leaves. The target gene FAD2-1, was amplified by PCR using target specific primers and then sequenced. In addition, the presence of kanamycin resistance gene was also verified by PCR using KANF (5' CTATTCGGCTATGACTGGGC 3') and KANR (5' AGGCATCGCCATGTGTCACG 3') primers. The PCR products were run on 2% agarose gel. Moreover, in order to confirm T-DNA integration, another PCR reaction was performed with using primers from both inside: Pps F1

Data collection and analyses of A. tumefaciensmediated sunflower transformation efficiency
Two technical replicates were used in in vitro germination studies. The percentage rates of survival of half-seeds co-cultured with A. tumefaciens cells in co-cultivation medium were calculated. In addition, the results of shoot and root formations in SM1, SM2, SM3, and RM were recorded. Co-cultivation was repeated twice for each cultivar (total 4 transformation groups) with 150 half cut seeds in each. In total, 600 half-seeds were used for agroinfection.

FAD2-1 gene amplication and sequencing
The approximate length of PCR fragments obtained by amplification of FAD2-1 gene was between 700-750 bp (Figure 1). The standard curve was made in Microsoft Excel to determine the length of DNA fragments according to the 1kb and 100 bp DNA Ladder. The determined size of DNA fragment was 718 bp. Then, DNA fragments were sequenced and no SNPs were found. Thus, it was confirmed that FAD2-1 gene sequences were the same in both genotypes and correspond to the sequences available in NCBI database. The sequence of FAD2-1 gene that was amplified from Deray and 2453-A genotype was given in Figure 2.

Gene Editing of FAD2-1 Gene
A variety of factors including utilization of two different A. tumefaciens strains (EHA105 and LBA4404), two different low oleic genotypes of sunflower (2453-A and Deray), and two different assembly techniques (Golden Gate and Gibson assembly) was assessed to increase transformation efficiency in this study.
Even though, 315/fw sequence has 1 off-target in genome with 3 mismatches, it is acceptable. The sequences were examined for GC% content in NEBcutter V2.0 and 50-70% GC pairs was accepted to ensure maximum stability. The GC% content of 497/rev is 57% and 315/fw is 70%. Additionally, the availability of restriction enzymes was checked by NEBcutter V2.0. 497/rev has target sites for DpnI, PvuI, BsiEI, MboI, DpnII, Sau3AI, TaqI and 315/ fw has target sites for Bs1I, HinfI, HpaII, MspI, BsaWI, MspJI, LpnPI. After NEBcutter analyses, the secondary structures of the sequences were examined by the mFold Web Server RNA Folding Version 2.3 Energies to eliminate the target sequences which might form hairpins or contain longer than six bp stems. The results showed that the selected sequences do not form hairpins and have no longer than six bp stems which inhibit the system (data not shown).

Vector construction and cloning
Two PCR reactions were set for each sgRNA (U3 497 and U6 315) with using UFRP and FPgR-R primer sets. Promoter and target regions were amplified using UFRP primer set. Additionally, sgRNA scaffold and target regions were amplified using FP  gR-R primer set. The expected size of the product of the PCR reaction 1 (in which UF and RP primers used) was 360 bp. The band sizes obtained as expected: 135 bp for scaffold region and 360 bp for scaffold region including target (Figure 4a). The product of the second PCR was 490 bp as expected (Figure 4b).  Similarly, pYLCRISPR/Cas9 binary vector was commonly used in CRISPR Cas9 gene editing studies on different plants such as soybean (Cheng et al., 2019), saffron (Chib et al., 2020), Medicago truncatula Gaertn. (Zhang et al., 2020), grape (Wang et al., 2018), and cotton (Gao et al., 2017).
After the PCR reaction, the product was used directly in E. coli transformation. The efficiencies of Gibson and Golden Gate assembly techniques were tested and compared. pHDE-35SCas9-mCherry CRISPR/Cas9 vector, which was used in Gibson Assembly, has PmeI and SpeI restriction sites while PmeI restriction enzyme generates blunt ends, SpeI restriction enzyme generates sticky ends. Therefore, sticky and blunt ligations were required in pHDE-35SCas9-mCherry CRISPR/Cas9 vector. Since the ligation efficiency of blunt ends is much lower than the one for sticky ends, blunt ends are not preferred. However, colony PCR results after E. coli transformation was subjected to gel electrophoresis and no bands could be obtained with Gibson assembly. Thus, the result of gel electrophoresis indicated that sgRNA expression cassettes could not be ligated into the vector or the ligation efficiency was very low. Differently, pYLCRISPR/Cas9 binary vector was applied in Golden Gate Assembly. The ligation product (490 bp, as expected) was observed in colony PCR ( Figure 5), hence, this method was chosen for the assembly of the sgRNA expression cassettes. The vector has BsaI restriction sites which generate sticky ends, thereby probably increase the ligation success. Plasmid was isolated from one colony and sequenced. The sequencing result, which was given in Figure 6, confirmed the presence of the correct sequence of cassettes in the plasmid.

A. tumefaciens transformation
Following the transformation, A. tumefaciens cells were spread on LB agar plates supplemented with adequate antibiotics: gentamycin and kanamycin for EHA105 strain and streptomycin and kanamycin for LBA4404 strain. Kanamycin is used for the selection of the vector. Additionally, gentamycin and streptomycin are used for selection of EHA105 and LBA4404 A. tumefaciens strains. Therefore, the transformed A. tumefaciens colonies were selected with these antibiotics.

Sunflower transformation
After vacuum infiltration, the half-seeds were cocultivated with A. tumefaciens cells in cocultivationmedium for two days in dark at 28°C. The survival rates of the seeds after transformation varied from 87% to 98% (Table 2). The highest survival rate was obtained in the with the co-cultivation of the half-seeds with A. tumefaciens EHA105 strain. Moreover, multiple shoots were obtained (1.5-1.7 shoot/explant) with 80% of rooting success (Figure 7) in SM1 and RM media, respectively. According to our preliminary studies, the cutting of the seeds was found to be an important step that had a direct impact on in vitro germination and the efficiency of the transformation as well (unpublished data).
The half-seeds were kept with A. tumefaciens culture in sterile tubes and subjected to vacuum infiltration for 30 minutes at 60 kPa as described by Sujatha et al. (2012). According to the study of Sujatha et al. (2012), vacuum infiltration was the treatment that had positive effect on transformation efficiency and 30 minutes were reported as the optimal time period of the treatment. Therefore, vacuum infiltration was applied for 30 minutes to enhance agro infection in this study. Yellow highlighted region represents U6 promoter sequence. Red highlighted region represents sgRNA sequence.

Analysis of putative transformed shoots by PCR
PCR results of FAD2-1 ( Figure 8) and kanamycin (Figure 9) genes demonstrated that both genes were successfully integrated in putative transformants. Moreover, PCR product was obtained with Pps F1 ve Pgs R2 primers as they amplified the T-DNA region while no band was seen with PmeI and Pgs R2 primers since PmeI primer amplified the outer region of T-DNA in the pYLCRISPR vector (data not shown).
This result showed that T-DNA (including the FAD2-1 and kanamycin genes) was successfully transformed to sunflower and the presence of the PCR product was not resulted from A. tumefaciens contamination since no PCR product was obtained with PmeI and PgsR2 primers.