Heterologous Expression of Jatropha curcas Fatty Acyl-ACP Thioesterase A (JcFATA) and B (JcFATB) Affects Fatty Acid Accumulation and Promotes Plant Growth and Development in Arabidopsis

Plant fatty acyl-acyl carrier protein (ACP) thioesterases terminate the process of de novo fatty acid biosynthesis in plastids by hydrolyzing the acyl-ACP intermediates, and determine the chain length and levels of free fatty acids. They are of interest due to their roles in fatty acid synthesis and their potential to modify plant seed oils through biotechnology. Fatty acyl-ACP thioesterases (FAT) are divided into two families, i.e., FATA and FATB, according to their amino acid sequence and substrate specificity. The high oil content in Jatropha curcas L. seed has attracted global attention due to its potential for the production of biodiesel. However, the detailed effects of JcFATA and JcFATB on fatty acid biosynthesis and plant growth and development are still unclear. In this study, we found that JcFATB transcripts were detected in all tissues and organs examined, with especially high accumulation in the roots, leaves, flowers, and some stages of developing seeds, and JcFATA showed a very similar expression pattern. Subcellular localization of the JcFATA-GFP and JcFATB-GFP fusion protein in Arabidopsis leaf protoplasts showed that both JcFATA and JcFATB localized in chloroplasts. Heterologous expression of JcFATA and JcFATB in Arabidopsis thaliana individually generated transgenic plants with longer roots, stems and siliques, larger rosette leaves, and bigger seeds compared with those of the wild type, indicating the overall promotion effects of JcFATA and JcFATB on plant growth and development while JcFATB had a larger impact. Compositional analysis of seed oil revealed that all fatty acids except 22:0 were significantly increased in the mature seeds of JcFATA-transgenic Arabidopsis lines, especially unsaturated fatty acids, such as the predominant fatty acids of seed oil, 18:1, 18:2, and 18:3. In the mature seeds of the JcFATB-transgenic Arabidopsis lines, most fatty acids were increased compared with those in wild type too, especially saturated fatty acids, such as 16:0, 18:0, 20:0, and 22:0. Our results demonstrated the promotion effect of JcFATA and JcFATB on plant growth and development, and their possible utilization to modify the seed oil composition and content in higher plants.


Introduction
Plant seed oils are one of the most important food sources for human beings, and are important raw materials for the production of cosmetics, soaps, paints, pharmaceuticals, emulsifiers and lubricants, etc. [1,2]. The rapid development of the world economy has led to an increasing demand for energy and the depletion of petroleum resources. Biodiesel is especially 16:0 and 18:0, in various organs, indicating AtFATB plays an essential role in plant growth and seed development as a main determinant in the synthesis of saturated fatty acids and their derivatives in Arabidopsis [29]. They also found that endogenous FATA activity was not increased in the mutant, suggesting no compensation adjustment between FATA and FATB, although previous studies found that the substrate specificities of AtFATA and AtFATB were 18:1 > 18:0 > 16:0 and 16:0 > 18:1 > 18:0, respectively [26,30]. Further study showed that the main metabolic response of the Arabidopsis plants to the disruption of AtFATB was an increased turnover of fatty acids, including synthesis and degradation, in fatb-ko plants [31].
A fata1 fata2 double mutant, generated by crossing two single mutants with a T-DNA insertion in the promoter region of AtFATA1 and AtFATA2, showed a ca. 60% and 50% decrease of the AtFATA1 and AtFATA2 expression levels, respectively, compared with the wild-type Arabidopsis. The fata1 fata2 plants did not show obvious morphological changes, but the seed oil content and fatty acid composition in the dry seeds were affected, including decreased contents of 18:0, 18:1, and 18:2 fatty acids [32]. A T-DNA insertion mutant in the promoter region of AtFATA2, with a significantly decreased expression of AtFATA2, showed longer siliques, more seeds per silique, slightly small seeds, and increased contents of most types of fatty acids except for 24:0 in dry seeds compared with the wild type, indicating that AtFATA2 plays an important role in seed lipid metabolism and silique development [33].
In a previous study, JcFATB1 was isolated from immature seeds of J. curcas and was strongly expressed in immature seeds detected by semi-quantitative RT-PCR. Ectopic overexpression of JcFATB1 in Arabidopsis driven by a seed-specific promoter could significantly increase the contents of saturated fatty acids, and decrease the contents of unsaturated fatty acids [34]. Dani et al. compared the protein sequences of JcFATA and JcFATB with AtFATA and AtFATB using bioinformatic methods and found three potential conserved catalytic active sites, i.e., the catalytic triad of N, H, and C, in JcFATA and JcFATB protein sequences, but functional verification of these potential active sites has not been reported [35]. The expression profile analysis of some key fatty acid enzyme genes showed that the expression levels of JcFATA were increased while the expression of JcFATB was decreased with seed development [36]. In this study, the expression patterns of JcFATA and JcFATB and subcellular localization of JcFATA and JcFATB were firstly analyzed, and their effects on plant growth and development and seed oil contents were further studied by heterologous expression in A. thaliana. Their possible roles in seed oil improvement were also discussed.
The CDS sequences of JcFATA and JcFATB are 1110 and 1257 bp in length, encoding proteins containing 369 and 418 amino acids, respectively. We searched the Jatropha Genome Database [41] using the JcFATA and JcFATB mRNA sequences to obtain their corresponding genome sequences, and the results showed that the genomic sequences of JcFATA and JcFATB are 3680 and 3672 bp in length from the start codon to the stop codon, and JcFATA contains 6 introns and 7 exons, and JcFATB includes 5 introns and 6 exons ( Figure 1A). Phylogenetic analysis of 25 FAT family proteins belonging to 10 families of dicotyledonous flowering plants, including A. thaliana and J. curcas, showed that JcFATA was grouped in the same evolutionary clade as AtFATA1 and AtFATA2 while the remaining four FAT proteins of J. curcas were in the same clade as AtFATB. JcFATA showed the highest homology with HbFATA, and JcFATB showed the highest homology with VfFATB, which were all from Euphorbiaceae plants ( Figure 1B). These results indicated that JcFATA and JcFATB might be functional and important FAT genes in J. curcas, so we conducted further studies on these two genes.
Plant fatty acyl-ACP thioesterase (FAT) proteins are encoded by nuclear genes and targeted to plastids by a transit peptide at the N-terminus. Previous studies have predicted the transit peptide sequences in the N-terminals of JcFATA and JcFATB as shown in Figure S2 according to the previous studies [34,35,42,43]. To determine the subcellular localization of JcFATA and JcFATB, the intact ORF without the stop codon of JcFATA and Jc-FATB were fused to the N-terminus of GFP (green fluorescent protein), respectively, driven by 35S promoter, and then used to transform Arabidopsis green leaf protoplasts for transient expression. In transformed cells, the green GFP signals overlapped with red chlorophyll autofluorescence, which indicated that JcFATA and JcFATB both localized in chloroplasts ( Figure 1C).    proteins. The coding region sequences were aligned using Clustal W, and the evolutionary relationship was analyzed using the neighbor-joining method. Numbers on branches indicate the percentage of replicate trees in which the associated sequences clustered together in the bootstrap test (1000 replicates). The segment under the phylogenic tree is the evolutionary distance, which was computed using the Poisson correction method. The NCBI accession numbers for the FAT proteins of A. thaliana and J. curcas are presented in Table S1. (C) Subcellular localization of the JcFATA-GFP and JcFATB-GFP fusion protein in leaf protoplasts isolated from 8-day-old seedlings of wild-type A. thaliana Col-0. GFP, green fluorescent protein. Scale bars = 2 µm.

JcFATA and JcFATB Are Constitutively Expressed Genes with Similar Expression Profiles
A previous study showed that JcFATA is a constitutively expressed gene as examined by the RT-PCR and GUS reporter system [44]. In this study, we detected the expression profiles of JcFATB using the same materials and method at the same time as the previous study by Liu et al. [44], and we also checked the expression patterns of JcFATA and JcFATB in developing seeds at different stages. As shown in Figure 2A, JcFATB was expressed at relatively low levels in stems, and showed relatively high expression levels in flowers, roots, and leaves, which is very similar to JcFATA [44]. During seed development, the expression of JcFATA and JcFATB increased during early developmental stages (20 to 40 DAF, days after flowering) but reduced during the late stage of seed development (60 DAF), and the highest expression was found in immature seeds at 40 DAF for both genes ( Figure 2B).
In order to study the detailed spatiotemporal expression patterns of JcFATB, the GUS expression vector driven by the JcFATB cognate promoter was transformed into wild-type A. thaliana, and the screened homozygous transgenic lines were used for GUS histochemical staining. We planted the JcFATA-GUS plants used in the previous study [44] and the JcFATB-GUS plants generated in this study under the same condition and then performed GUS staining at the same time for comparison. GUS activity was detected in the radicles and cotyledon leaves of 5-day-old seedlings for both genes. The GUS signals were almost undetectable in the whole hypocotyl of the JcFATA-GUS plant but only undetectable in the middle part of the hypocotyl of the JcFATB-GUS plant (Figures 2(C1) and S1A). For 15-day-old JcFATB-GUS seedlings, high levels of GUS expression were detected in the whole plant except for hypocotyl (Figure 2(C2)), but GUS activity was almost undetectable in the petioles and hypocotyl of 15-day-old JcFATA-GUS seedlings ( Figure S1B). High GUS staining was detected in the full expanded leaf blades, especially in the leaf veins of 30-day-old plants for both genes (Figures 2(C3) and S1C). GUS staining of the inflorescences during the flowering stage showed that almost no GUS activity was detected in the inflorescence rachis while relatively weak and strong GUS signals were detected in flowers during the early developmental stages and late developmental stages, respectively, for both genes (Figures 2(C4) and S1D). For JcFATB-GUS plants, GUS activity was relatively low in fruit pots during different development stages, and blue signals were mainly found in the tips of the siliques and the junction of the silique base and the peduncle (Figure 2(C5)), which is very similar to a previous study [44]. The results showed that JcFATB-GUS plants showed a very similar GUS staining pattern to JcFATA-GUS plants, but JcFATB-GUS showed stronger GUS staining signals in all the tissues detected ( Figures 2C and S1).

Ectopic Expression of JcFATA and JcFATB and Their Mutant Versions Affected the Fatty Acid Accumulation in E. coli
Previous study suggested that there may be 3 possible conserved catalytic active sites, i.e., 265N (asparagine), 267H (histidine), and 302C (cysteine) for JcFATA, and 315N 317H, and 352C for JcFATB, similar to AtFATA and AtFATB ( Figure S2) [35]. In order to verify the catalytic activity of the active sites 302C and 352C in JcFATA and JcFATB respectively, we constructed the prokaryotic expression vectors for the wild-type JcFATA and JcFATB, and their mutation versions with a mutation of 302C (TGC, cysteine) to 302F (TTC, phenylalanine) in JcFATA and a mutation of 352C (TGT) to 352F (TTT) in JcFATB named the JcFATA-Mutation and JcFATB-Mutation, respectively ( Figures S2, S3, and 3A,B). Then, each construct and the empty vector were transformed into E. coli strain

Ectopic Expression of JcFATA and JcFATB and Their Mutant Versions Affected the Fatty Acid Accumulation in E. coli
Previous study suggested that there may be 3 possible conserved catalytic active sites, i.e., 265N (asparagine), 267H (histidine), and 302C (cysteine) for JcFATA, and 315N, 317H, and 352C for JcFATB, similar to AtFATA and AtFATB ( Figure S2) [35]. In order to verify the catalytic activity of the active sites 302C and 352C in JcFATA and JcFATB, respectively, we constructed the prokaryotic expression vectors for the wild-type JcFATA and JcFATB, and their mutation versions with a mutation of 302C (TGC, cysteine) to 302F (TTC, phenylalanine) in JcFATA and a mutation of 352C (TGT) to 352F (TTT) in JcFATB, named the JcFATA-Mutation and JcFATB-Mutation, respectively (Figures S2, S3, and 3A,B). Then, each construct and the empty vector were transformed into E. coli strain Rosetta to induce exogenous gene expression. The fatty acid composition was then assayed to test the effects of JcFATA, JcFATA-Mutation, JcFATB, and JcFATB-Mutation on the fatty acid accumulation in E. coli. Compared with the E. coli transformed with pCold I empty vector, the contents of 16:1 and 18:1 unsaturated fatty acids were significantly increased in E. coli transformed with pColdI-JcFATA, and the contents of unsaturated fatty acids were significantly reduced in the E. coli transformed with pColdI-JcFATA-Mutation compared with pCold I and pColdI-JcFATA ( Figure 3C). These results indicated that JcFATA can promote the accumulation of unsaturated fatty acids in E. coli, and 302C may be a key amino acid residue of the enzyme activity in regulating the accumulation of unsaturated fatty acids. Compared with E. coli transformed with pCold I empty vector, the saturated fatty acid contents of 16:0 and 18:0 were significantly increased in E. coli transformed with pColdI-JcFATB. Moreover, the saturated fatty acid contents of 16:0 and 18:0 were significantly reduced in the E. coli transformed with pColdI-JcFATB-Mutation compared with pColdI-JcFATB, and only 18:0 was significantly reduced compared with pCold I ( Figure 3D). These results indicated that JcFATB can promote the production of bacterial saturated fatty acids, and 352C may be a key amino acid residue of JcFATB for the regulation of saturated fatty acid production.

Ectopic Expression of JcFATA and JcFATB Promotes the Growth and Development of A. thaliana
In order to investigate the effect of JcFATA and JcFATB on plant growth and fatty acid accumulation, we constructed the overexpression constructs of JcFATA and JcFATB driven by CaMV 35S promoter, and then the resultant constructs were transformed into Arabidopsis individually. Three independent homozygous lines, i.e., JcFATA-1, JcFATA-2, JcFATA-3 for JcFATA, and JcFATB-1, JcFATB-2, and JcFATB-3 for JcFATB, were obtained by planting and screening with hygromycin from the T3 generation. PCR and semi-quantitative RT-PCR showed the existence of JcFATA and JcFATB transgene and their successful expression in Arabidopsis while the transgene and mRNA of JcFATA and JcFATB were not detected in wild-type Col-0 ( Figure 4A). These homozygous lines were used for further observation and analysis.
We then observed and analyzed some growth parameters during the whole life cycle of wild-type Col-0 plants and the above homozygous JcFATA and JcFATB ectopic expression lines. At 8 DAT (days after transplanting), both JcFATA and JcFATB lines showed a significantly increased root length compared with that of the wild type. Specifically, the JcFATB lines showed a ca. 30-40% increase. Furthermore, the root lengths of JcFATB lines were also significantly longer than the JcFATA lines ( Figures 4B and S4A). Very similar results were also observed for the rosette diameters and the numbers of rosette leaves per plants at 20 DAT, i.e., the JcFATA and JcFATB lines also showed significantly larger and more rosette leaves compared to the wild-type plants ( Figures 4C,D and S4B,C). These results indicate that heterologous expression of JcFATA or JcFATB can promote plant growth during the vegetative stages of Arabidopsis, and JcFATB showed a greater growth promotion effect than JcFATA. contents of 16:0 and 18:0 were significantly reduced in the E. coli transformed with pColdI-JcFATB-Mutation compared with pColdI-JcFATB, and only 18:0 was significantly reduced compared with pCold I ( Figure 3D). These results indicated that JcFATB can promote the production of bacterial saturated fatty acids, and 352C may be a key amino acid residue of JcFATB for the regulation of saturated fatty acid production.

Ectopic Expression of JcFATA and JcFATB Promotes the Growth and Development of A. thaliana
In order to investigate the effect of JcFATA and JcFATB on plant growth and fatty acid accumulation, we constructed the overexpression constructs of JcFATA and JcFATB    To identify whether JcFATA and JcFATB can also improve plant growth and development during the reproductive stages of Arabidopsis, the plant height and number of bloomed flowers for the wild-type and transgenic lines at 28 DAT were recorded and analyzed. The average plant heights of JcFATA and JcFATB lines were significantly increased compared with that in the wild type, and the JcFATB lines showed a significant increase compared to the JcFATA lines, which is very similar to the results obtained during the vegetative stages ( Figures 4D and S4D). At 28 DAT, some of the flower buds flowered, and the numbers of bloomed flowers at this time were slightly increased in 1 and 2 lines of the JcFATA and JcFATB plants, respectively, compared with that in the wild type ( Figures 4F and S4D). At 50 DAT, we observed that the JcFATA and the JcFATB lines produced subtly longer siliques ( Figure S4E,F), and further measurements showed that the average silique lengths of the JcFATA lines were slightly increased compared to the wild type, but the JcFATB lines showed significantly longer siliques compared with the wild type ( Figure 4G). We then analyzed the phenotypes of the mature seeds of the wild-type, JcFATA, and JcFATB lines. As shown in Table 1 and Figure S4G, both JcFATA and JcFATB lines produced significantly bigger and heavier seeds, and the lengths, widths, and grain weights of 500 dry seeds were increased significantly in the JcFATA and JcFATB lines compared with those in the wild type. The grain weight of 500 dry seeds was increased by 14-21% for the JcFATA lines and 38-45% for the JcFATB lines. Our results indicate that the ectopic expression of JcFATA and JcFATB can significantly increase the seed weight of Arabidopsis, and JcFATB showed a greater effect. Values represent means ± SD (standard deviation), and 50 seeds for grain size and 500 seeds for grain weight were measured for each replicate and 3 replicates were used for each line. Data in the same column followed by different letters (a, b, and c) are significantly different at the p ≤ 5% level as determined by Duncan's multiple range test. For each trait, 10 plants were used as 1 replicate.

Ectopic Expression of JcFATA and JcFATB Affects Seed Fatty Acid Accumulation in Arabidopsis
Due to the high expression of JcFATA and JcFATB in J. curcus seeds at 40 DAF and the considerable phenotypical changes in the seeds of the JcFATA and JcFATB ectopic expression lines, we further examined the fatty acid composition of seed storage lipids in wild-type and ectopic expression lines of Arabidopsis. Compositional analyses of seed oil revealed that, except 20:1, the contents of other unsaturated fatty acids, including 16:1, 18:1, 18:2, 18:3, and 22:1, were significantly increased in the dry seeds of JcFATA lines compared with those in the wild type ( Figure 5A). Moreover, the contents of 18:1, 18:2, and 18:3 were increased by 69-95%, 58-68%, and 52-70%, respectively, in the JcFATA lines compared with the wild type (Table S2). Notably, the contents of the 18:0 and 20:0 saturated fatty acids were also significantly increased in the JcFATA lines. These results suggested that JcFATA can significantly promote the accumulation of fatty acids, especially the accumulation of unsaturated fatty acids. Similarly, most of the fatty acids were increased in the dry seeds of the JcFATB lines compared with those in the wild type except that 22:1 was decreased significantly ( Figure 5B). The saturated fatty acids 16:0, 18:0, 20:0, and 22:0 were significantly increased by 84-108%, 124-145%, 65-92%, and 96-170%, respectively, in the JcFATB lines compared with the wild type (Table S3). Regarding the unsaturated fatty acids, 18:1, 18:2, and 18:3 were also significantly increased at relatively lower levels than the saturated fatty acids. The significantly increased contents of fatty acids, especially the saturated fatty acids, in the mature seeds of A. thaliana indicate that JcFATB has a significant stimulative effect on the accumulation of fatty acids, especially on the accumulation of saturated fatty acids. As seen from the data shown in Tables S2 and S3
A previous study showed that the contents of the total seed oil and the four major fatty acids, i.e., palmitic acid (16:0), stearic acid (18:0), oleic acid (18:1), and linoleic acid (18:2), varied significantly in 19 different accessions of J. curcas [49], which indicated that the seed oil content and composition could be improved through conventional breeding and biotechnology. Furthermore, Jatropha seed oil contains a high content of polyunsaturated fatty acids, which may reduce the oxidation stability of the oil, and can also lead to an increase in nitrogen oxide emissions after burning [50,51]. The ideal biodiesel should contain more monounsaturated fatty acids, such as oleic acid (18:1), rather than polyunsaturated fatty acids, such as linoleic acid (18:2) and linolenic acid (18:3), so the composition of fatty acids in seed oil has yet to be improved [52]. FATs are specific for substrate selection and have a certain determining effect on the types of fatty acids produced in seeds and can be used as potential target genes for the improvement of To deduce whether the seed oil contents were altered in the transgenic Arabidopsis lines expressing JcFATA and JcFATB compared with the wild-type Col-0, we calculated the percentages of the fatty acids presented in Figure 5 in the total seed oils by analyzing the fatty acid GC-MS (gas chromatograph-mass spectrometer) data. Surprisingly, the percentages of almost all the saturated fatty acids and unsaturated fatty acids did not change significantly in the JcFATA lines compared with those in the wild type, and the JcFATB lines only showed significantly increased percentages for 18:0 and a significant decrease for 22:1 (Table S4). Moreover, the total percentages were evidently increased and decreased for the saturated fatty acids and the unsaturated fatty acids, respectively, in the JcFATB lines. However, the total percentages for the saturated and unsaturated fatty acids did not change much in both the JcFATA lines and JcFATB lines (Table S4). We also calculated the GC-MS peak ratios of the total saturated and unsaturated fatty acids with the standard substance (ethyl decanoate), and we found that the ratios were increased significantly for both the JcFATA lines and JcFATB lines compared with that in the wild type (Table S5). These data suggest that the seed oil content might be increased in the Arabidopsis line with ectopic expression of JcFATA and JcFATB. To prove this, further experiments are required.
A previous study showed that the contents of the total seed oil and the four major fatty acids, i.e., palmitic acid (16:0), stearic acid (18:0), oleic acid (18:1), and linoleic acid (18:2), varied significantly in 19 different accessions of J. curcas [49], which indicated that the seed oil content and composition could be improved through conventional breeding and biotechnology. Furthermore, Jatropha seed oil contains a high content of polyunsaturated fatty acids, which may reduce the oxidation stability of the oil, and can also lead to an increase in nitrogen oxide emissions after burning [50,51]. The ideal biodiesel should contain more monounsaturated fatty acids, such as oleic acid (18:1), rather than polyunsaturated fatty acids, such as linoleic acid (18:2) and linolenic acid (18:3), so the composition of fatty acids in seed oil has yet to be improved [52]. FATs are specific for substrate selection and have a certain determining effect on the types of fatty acids produced in seeds and can be used as potential target genes for the improvement of seed oil in plants. Several studies have showed that FATA can increase the accumulation of oleic acid (18:1) and linoleic acid (18:2) in several plant species [26,[53][54][55].
FATs are expected to localize in plastids or chloroplasts, and the previous study showed that AtFATA2 localized in chloroplasts [33]. In this study, we found that JcFATA and JcFATB both localized in chloroplasts as detected by transient expression in Arabidopsis leaf protoplasts, indicating they may have similar functions to their Arabidopsis homologs. However, we also found that the fluorescence signal patterns of JcFATA and JcFATB were not identical, and the fluorescence signals for JcFATA-GFP showed an uneven pattern, which is very similar to AtFATA2, while the signals of JcFATB-GFP were more contiguous in a single chloroplast. Previous studies indicated that the proteins localized in the outer or inner membrane of chloroplasts usually show uneven fluorescence signals [33,[56][57][58]. However, the GFP fluorescence signals of thylakoid proteins are usually uniformly distributed throughout the chloroplast [58]. The difference in their localization in chloroplasts suggests that FATA and FATB may reside in different positions to exert their hydrolysis function [33,[57][58][59][60].
FATB in Cuphea hyssopifolia is widely expressed in the developing embryos [61]. At-FATB is widely expressed in all organs but has the highest expression level in flower organs [28]. JcFATB transcripts were detected in all the organs examined, with the highest expression in the developing seeds at 32 DAF [34]. The expression of JcFATA was increased with the development of the endosperm and then decreased during seed maturation [62]. In this study, we found that the spatiotemporal expression patterns of JcFATA and JcFATB were similar to those of FATA and FATB in Arabidopsis and other plants, suggesting that JcFATA and JcFATB may act as main acyl-ACP thioesterases and share similar functions in the various tissues and developmental stages of Jatropha, especially in fatty acid synthesis. This may, in turn, make FATs essential for plant viability by affecting fatty acid metabolism as shown by previous studies [32,37]. In the present study, longer and bigger seeds were produced by Arabidopsis plants with ectopically expressed JcFATA or JcFATB driven by CaMV 35S promoter, and the increased fatty acid contents and changed seed oil compositions in mature seeds may in turn affect seed development. Previous research and the present study suggest that JcFATA and JcFATB may play an important role in seed development by participating in fatty acid synthesis [34].
Prokaryotic expression studies showed that plant FATs could influence the fatty acid accumulation of E. coli. Ectopic expression of a plant FAT gene encoding a medium-chainspecific FAT, named BTE from Umbellularia californica, in both a normal and a fatty acid synthesis-deficient mutant of E. coli had a limited impact on the normal strain and a huge impact on the mutant strain on the fatty acid composition [63]. E. coli expressing CsFATA of Coriandrum sativum showed an increased unsaturated fatty acid content, indicating that CsFATA plays a certain role in promoting the formation of unsaturated fatty acids [30]. E. coli can also be used for the production of free fatty acids by blocking the fatty acid elongation process caused by ectopic expression of FAT genes from other species [64]. A previous study found that ectopic expression of seven exogenous FAT genes, including AtFATA, in E. coli led to significantly increased production of free fatty acids, and the fatty acid compositions of the strains with different FATs showed substrate specificity, suggesting that FATs can be engineered and introduced into E. coli to produce free fatty acids [65]. Ectopic expression of a FatB gene cloned from Diploknema butyracea (Madhuca butyracea) in E. coli caused a significant increase in the content of 16:0 saturated fatty acids in the supernatant [66]. In this study, we found that ectopic expression of JcFATA and JcFATB in E.  [66]. Ectopic expression of JcFATA or JcFATB might have the same effect on the CFA synthesis of the cell membrane in E. coli, such as for AhFatA [67].
As the main FATs, FATA and FATB are required for the final step of de novo fatty acid biosynthesis, which determines the metabolic flux of fatty acid metabolism in plants, and are therefore essential for plant survival [67]. Our results showed that the ectopic expression of JcFATA and JcFATB produced longer roots and siliques, larger and more rosette leaves, more flowering buds, and greater plant height, and the JcFATB lines showed greater increases in these phenotypes compared with the JcFATA lines ( Figure 4). Corresponding to the stimulative effects on plant growth, we found that most of the fatty acids of the mature seeds were increased due to the ectopic expression of JcFATA and JcFATB, and the significant increase in some fatty acids, such as 18:0, 18:1, 18:2, and 18:3 for the JcFATA lines and 16:0, 18:0, and 20:0 for the JcFATB lines, might be responsible for the plant growth promotion ( Figure 5). The biosynthesis and supply of saturated fatty acids and unsaturated fatty acids are essential in plant growth and development [29]. Although we did not measure the fatty acid contents in other organs, we expect their fatty acid contents would be increased similar to that observed in mature seeds considering their phenotypic changes. The phenotypic changes in the transgenic Arabidopsis plants in this study may be due to metabolic responses to the changed fatty acid contents and compositions. Protein Sacylation, especially palmitoylation, is an important post-translational modification, which is essential for the regulation of activity and localization of membrane-related signaling proteins [68]. As a reversible modification of the cysteine residues of target proteins, protein S-acylation plays important roles in multiple aspects of protein function, such as localization, stability, trafficking, and conformation. The increased contents of 16:0 and other fatty acids may affect signal transduction in multiple growth and development processes by promoting protein S-acylation [69].
One interesting work showed that overexpression of Arabidopsis ceramide synthase genes LOH1, LOH2, and LOH3 under the control of the CaMV 35S promoter in Col-0 led to differentially altered growth and extensive changes in sphingolipid metabolism, in which LOH1 and LOH3 overexpression lines showed a significant increase in plant size and biomass with little alteration in the sphingolipid composition or content on a tissue mass basis. However, LOH2 overexpression led to severely dwarfed plants with differentially altered sphingolipid profiles in their rosettes compared with the wild type [70]. Moreover, an interesting study in cotton and Arabidopsis found that the very-long-chain fatty acids (VLCFAs) with saturated fatty acids, especially C24:0, play an important role in cell elongation and expansion by activating ethylene biosynthesis and signaling. Comprehensive lipid analysis indicated that linolenic (18:3) and palmitic (16:0) acid are the most abundant fatty acids in the development of cotton fibers [71]. The significant decrease in 22:1 in JcFATB ectopic Arabidopsis lines could be caused by the increase in the accumulation of C16 and C18 fatty acids caused by JcFATA overexpression, and determination of the VLCFAs content might explain this question in our further study. Overexpression of AtFAAH, which encodes the fatty acid amide hydrolase that is responsible for hydrolyzing N-acylethanolamines, a group of fatty acid derivatives of ethanolamine, into ethanolamine and their corresponding free fatty acids, under the control of CaMV 35S promoter, showed a significant promotion effect on plant growth [72]. From these studies, we could speculate that the overexpression of some genes related to fatty acid metabolism or related pathways may promote plant growth and development.
Vegetable oils and their derivatives have important economic value and have been used as important industrial raw materials and nutritional sources. Thus, significant attention has been given to the yield and quality of seed oils. In the effort to improve seed oil, several genes have been used in previous studies; however, with an improved seed oil content, some genes showed inhibition effects on plant growth and development [73,74]. Ectopic expression of a fatty acid dehydrogenase gene increased the content of α-linolenic acid (18:3) in soybean seeds, but this was accompanied by a serious decrease in the total oil content [75,76]. Several studies showed that FATs can impact on the production of glycerolipds by affecting the export of acyl moieties to ER in Arabidopsis [28,29,[31][32][33]. Although JcFATA and JcFATB had different effects on the fatty acid composition of seed oils when ectopically expressed in A. thaliana, these two genes can promote the growth and development of plants, with almost the same effects on various organs. Our results suggested that FAT genes are potential genes that can be used for seed oil improvement, and the FAT genes from other plant species could also be used. In this study, we used the constitutive CaMV 35S promoter to drive the expression of JcFATA and JcFATB instead of the seed-specific promoter used in other studies, and we also obtained very promising results. Increased unsaturated fatty acids may promote the chilling tolerance of plants as described previously, which suggests that FATA might be used to improve cold tolerance and modify the seed oil composition, similar to other genes involved in fatty acid synthesis [75][76][77][78][79][80][81][82]. Overexpression of JcFATA may provide a possible potential to achieve a combined increase in unsaturated fatty acids, seed yield, and chilling tolerance in J. curcas. Simultaneous overexpression of FATA and FATB with different expression levels might also be a strategy for modifying the seed oil content in different plants.

Plant Materials and Growth Conditions
For J. curcas, a genotype named M-19 collected from Yunnan Province of China was used in this study [83,84]. The mature tree plants were grown on the farm of South China Agricultural University in Guangzhou, China. Arabidopsis thaliana L. ecotype Columbia (Col-0) was used in this study. The wild-type seeds of Col-0 were sterilized in 2% sodium hypochlorite for 15 min, and then rinsed 3 times with sterile water and inoculated on 1/2 MS medium [85]. Seeds from transgenic plants were planted on the same basic medium containing 50 mg/L hygromycin. After treatment in darkness at 4 • C for 48 h, the seeds were germinated at 22 • C under a 16 h day/8 h night photoperiod with a light intensity of 120 µmol/m 2 s and 80% relative humidity in an artificially controlled plant growth chamber. The seedlings were transferred to soil 7 days after germination and grown under the same growth conditions. All mediums were supplemented with 100 mg/L myo-inositol (Sigma-Aldrich, St. Louis, MO, USA) and 2.5% sucrose, adjusted to pH 5.8-6.0, and solidified with 0.6% agar prior to autoclaving at 1.4 kg cm −2 for 20 min.

Phylogenetic Analysis of FAT Proteins
The protein and mRNA sequences of FATA and FATB from J. curcas and A. thaliana were downloaded from National Center for Biotechnology Information [86] and the NCBI accession numbers for each sequence used in this study are listed in Table S1. BLAST analysis was performed to search the homologous proteins in other species using Jc-FATA and JcFATB as the query, respectively. In total, 12 FATAs and 6 FATBs from other dicotyledonous plant species were chosen and used for phylogenetic analysis, includ- Multiple alignments of the protein sequences were generated with Clustal X [87]. A phylogenetic tree was built by the neighbor-joining method with MEGA 5.05 using 1000 bootstrap replicates [88].

RNA and cDNA Preparation
Total RNA of the flowers after 5 DAF (days after flowering); roots, leaves, and stems from mature plants; seeds at different developmental stages, i.e., 20, 40, and 60 DAF of J. curcas; and 30-day-old rosette leaves of A. thaliana was extracted using a Plant RNA Kit (Solar Technologies, Gaithersburg, MD, USA) according to the manufacturer's protocol. Total RNA (1 µg) free of DNA was used for cDNA synthesis using M-MLV Reverse Transcriptase (Promega, Madison, WI, USA) according to the manufacturer's protocol.

Semi-Quantitative RT-PCR
The primer pairs specific for JcFATA and JcFATB were designed using Primer Premier 5.0 software (PREMIER Biosoft, San Francisco, CA, USA), and validated to generate single PCR products with the expected sizes. For analysis of the expression profiles of JcFATA and JcFATB in J. curcas by semi-quantitative RT-PCR, 1 µL of a 1:1 diluted RT reaction product was used as template in a 20 µL reaction volume with primer pairs JcFATA-RTSF and JcFATA-RTSR to amplify a 376 bp coding sequence of JcFATA and JcFATB-RTSF and JcFATB-RTSR to amplify a 324 bp coding fragment of JcFATB. The house-keeping gene JcActin7 (XM_012232498.2) was used as the reference gene. For amplification, the primers JcACT7-SF and JcACT7-SR were used, which can amplify a 393 bp PCR fragment. For analysis of the expression of JcFATA and JcFATB in Arabidopsis by semi-quantitative RT-PCR, the same primer pairs and RT-PCR reaction system were used, except that the constitutive expression gene AtActin2 (NM_112764.4) was used as the internal control, and the primers AtACT2-F and AtACT2-R were used for amplification, which amplified a 191 bp PCR fragment. All the primers used in this study are listed in Table S6.

Histochemical GUS Assay
A modified pCAMBIA1300 vector with the CaMV 35S promoter, GUS gene, and NOS terminator in the multiple cloning sites was used for GUS fusion construction [89]. A 2234 bp genomic sequence upstream of the ATG start codon of JcFATB was amplified from M-19 by PCR using the primer pairs JcFATB-GUS-F and JcFATB-GUS-R. Then, the CaMV 35S promoter of the above pCAMBIA1300-GUS vector was replaced as described for the construction of the JcFATA-GUS fusion in which a 2271 bp promoter region of JcFATA was used [44]. The resulting construct was transformed into Agrobacterium tumefaciens stain EHA105 by the freeze-thaw method [90]. Transformation of A. thaliana Col-0 plants was performed using the floral dip method as previously described [91]. The whole seedlings or tissue cuttings of the wild-type and transgenic Col-0 plants at different developmental stages were stained in 2 or 5 mL tubes. GUS staining was performed as described previously [92]. After staining and decoloration, samples were observed and photographed with a stereomicroscope LeicaMZ16 (Leica, Wetzlar, Germany).

Subcellular Localization Analysis
The intact coding region sequences of JcFATA and JcFATB without a stop codon were amplified by RT-PCR with the primer pairs of JcFATA-PGFP-F and JcFATA-PGFP-R and JcFATB-PGFP-F and JcFATB-PGFP-R, and inserted into the transient expression vector pUC18-35S-eGFP between the CaMV 35S promoter and the GFP (green fluorescent protein) gene, generating an in-frame fusion for each gene. The leaves of 8-day-old Arabidopsis seedlings were cut into 1-2 mm pieces using a fresh sharp blade and used for protoplast preparation. Protoplasts were quantified using a hemocytometer under a microscope, and the fusion constructs for each gene and the control expression vector was introduced into the protoplasts as described [93]. GFP signals were observed under a fluorescence microscope OLYMPUS MF30 (Olympus Corporation, Tokyo, Japna) with the excitation and emission filters Ex480 ± 20/DM505/BA535 ± 25 and Ex535 ± 25/-DM565/BA645 ± 37.5 for GFP and chlorophyll auto-fluorescence, respectively. All fluorescence images obtained were processed with LSM 5 Image Browser (Carl Zeiss AG, Oberkochen, Germany).

Construction of JcFATA and JcFATB Site-Directed Mutagenesis Vectors
The prokaryotic expression vectors for JcFATB (pColdI-JcFATB) were constructed first as previously described for JcFATA (pColdI-JcFATA), which was also used in this study [94].

Analysis of the Fatty Acid Composition of E. coli
To induce JcFATA and JcFATB expression in E. coli, the strain Rosseta (DE3) (TransGen Biotech, Beijing, China) was transformed with pCold I, pColdI-JcFATA, pColdI-JcFATA-Mutation, pColdI-JcFATB, and pColdI-JcFATB-Mutation, respectively. Cultures were grown at 15 • C to OD 600 of 0.6, and then 0.5 mmol/L of IPTG were added to induce the expression of cloned genes. The cells were then collected by a 10-min centrifugation at 4000 rpm, and then resuspended in 5 mL of ddH 2 O. After a 10-min centrifugation at 4000 rpm, 2 mL of NaOH-methanol solution were added followed by incubation in a 100 • C water bath for 40 min. the solution was then left to cool to room temperature. Two volumes of HClmethanol solution (80 • C) were added into the tube, incubated in an 80 • C water bath for 40 min with shaking at 80 rpm, and then rapidly cooled to 25 • C in an ice box. Then, 1 mL of n-hexane (Sigma-Aldrich, St. Louis, MO, USA) was added into the cell mixture solution and vortexed for 2 min. After a 5-min centrifugation at 4000 rpm, the upper supernatant was transferred to a new tube. After N 2 evaporation, FAMEs were dissolved and assayed according to the method used for mature seeds in this study.

Construction of the Overexpression Vectors of JcFATA and JcFATB and Arabidopsis Transformation
The coding sequence (1110 bp) of JcFATA and the coding sequence (1257 bp) of JcFATB were amplified by RT-PCR with the primer pairs JcFATA-F and JcFATA-R, and JcFATB-F and JcFATB-R, and the PCR products were gel-purified, digested, and inserted downstream of the CaMV 35S promoter cloned into the binary vector pCAMBIA1390. The two recombinant vectors pCAMBIA1390-35S-JcFATA and pCAMBIA1390-35S-JcFATB were transformed into A. tumefaciens stain EHA105, and then introduced into A. thaliana Col-0 to obtain the transgenic Arabidopsis lines ectopically expressing JcFATA or JcFATB via the floral dip method [91].

Phenotypic Observation and Analysis
For phenotypic observation of the wild-type and transgenic Arabidopsis plants, we obtained three independent homozygous lines for Arabidopsis plants ectopically overexpressing both JcFATA and JcFATB by screening the survival rate of the T3 generation in 1/2 MS solid medium containing 50 mg/L hygromycin. After screening, the homozygous lines and the wild-type plants were planted and grown on the same medium without hygromycin under the same conditions. After 7 days, some of the plants were transplanted onto new medium to measure the root length, and all other plants were transplanted into small pots with sterilized nutritional soil and vermiculite at a 1:1 volume ratio. Four plants were planted in each pot. The phenotypes, including the root length, rosette diameter, rosette number, plant height, flowering efficiency, silique length, seed number, seed size, and seed weight, of the transgenic plants and wild-type plants were observed and recorded at 8, 20, 28, and 50 DAT. Regarding the rosette diameter, the largest pair of rosette leaves was used, and the number of rosette leaves was also recorded at 20 DAT. At 28 DAT, the plant height and the flowering efficiency were assayed. The flowering efficiency was designated as the percentage of bloomed flower buds in all the flower buds, i.e., the bloomed ratios. The silique length was measured at 50 DAT, and 10 siliques on the middle part of the main stem for each plant were used. The seed length and width were measured with a micrometer under a microscope, and 50 seeds were measured for each replicate and 3 replicates were used for each line. For each trait, 10 plants were used as 1 replicate and 3 replicates were analyzed in total.

Fatty Acid Analysis of the Mature Seeds of Arabidopsis
Fatty acid methyl esters (FAMEs) were prepared from dry seeds of the wild-type control and the transgenic lines as described previously with minor modification [33]. For each sample, 10 mg of dry seeds were transferred into a 5 mL centrifuge tube, and 1 mL of 5% methanol-sulfuric acid solution and 25 µL of 0.2% dibutyl hydroxytoluene dissolved in methanol were added. Then, the tubes were placed in a 90 • C water bath for 90 min. After incubation, the tubes with samples were placed on ice, and 2 mL of n-hexane and 1.5 mL of 0.9% NaCl were added into each tube. Then, the upper liquid was transferred to a new tube. After N 2 evaporation, FAMEs were dissolved in 200 µL of n-hexane with 0.216 ng of ethyl decanoate (Sigma-Aldrich, St. Louis, MO, USA) as an internal standard and transferred to a GC vial. The samples were filtered by a 0.22 µm filter (organic phase) and assayed by GC-MS. FAMEs were separated using a 30 m + 0.25 mm DB-23 capillary column with helium as the carrying gas in an Agilent Technologies 7890A Gas Chromatograph, and detected by a flame ionization detector at 230 • C. The program was 170 • C for 1 min followed by an increase of 4 • C/min to 250 • C, which was maintained for a further 3 min, and the column flow was 7.4 mL/min at a split vent ratio of 10:1. FAMEs were identified by comparison with the retention times of reference standards. FAMEs were quantified by comparing the areas of major peaks with those of internal standards.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.