Plant seed oil plays an important role in human's daily life. It is not only an important source of dietary fat, but also an important source of biofuels and other industrial applications (Cahoon et al., 2007; Lu et al., 2011). The function, properties and suitability of different plant seed oils for specific applications mainly depend on the composition and content of their fatty acids (Wickramarathna et al., 2015). According to the number of double bonds, the unsaturated fatty acids in plant seed oil can be divided into monounsaturated oleic acid (18:1; OA) and polyunsaturated fatty acids (PUFA). Studies have shown that increasing the ratio of OA (18:1) in plant seed oil can improve oxidation stability and has significant health benefits for human body, which is also very important for the production of biodiesel and other renewable resources (Kinney, 1996). At the same time, high PUFA is a viable nutritional option in human diet plans (Hunter, 1990; Williams and Burdge, 2006). For this reason, as the main source of essential fatty acids in the diet, the synthesis of triacylglycerol in plants and the study of related enzymology have been the field of concern to scientists.

During the development of oil crop seeds, most of the monounsaturated oleic acid (18:1) formed in plasts enters phosphatidylcholine (PC) (Shanklin and Cahoon, 1998, Bates et al., 2007). Then, through the desaturation of fatty acid desaturase 2 (FAD2) and FAD3, it forms linoleic acid (18:2) and linolenic acid (18:3) (Browse et al., 1993, Wallis et al., 2002). Since then, PC has been incorporated into TAG, becoming a common form of fat storage in many plant seeds and a major source of essential fatty acids in human dietary nutrition choices (Lu et al., 2009).

Previous studies have identified a key enzyme in seed oil formation, phosphatidylcholine diester choline phosphotransferase (PDCT), which catalyzed the exchange of choline phosphate heads during seed TAG synthesis, leading to the interconversion of PC and DAG (Lu et al., 2009). As a gate enzyme, PDCT provides an important way for 18:1 entering PC desaturation to become 18:2 and 18:3, and for 18:2 and 18:3 desaturation products to return to DAG. In Ninanjie (Arabidopsis thaliana), defects in PDCT reduced the content of polyunsaturated fatty acids (PUFA) in seeds by 40% (Lu et al., 2009). Experiments in Youcai (Brassica napus) also demonstrated the effect of the loss of PDCT activity on the reduction of PUFA accumulation in rapeseed oil. In addition, allogeneously expressing the PDCT gene of Yama (Linum usitatissimum L.) in yeast and Arabidopsis thaliana increased the PUFA level of the mutant (Wickramarathna et al., 2015). These results indicate that PDCT plays an important role in PUFA accumulation in plant seeds. However, little research has been done on PDCTS in other species to see if they have any other function other than the aforementioned role in PUFA accumulation in seeds. In this study, by means of bioinformatics, using six published oil crops-- Huasheng (Arachis hypogea), Ganlanxing Youcai (Brassica napus), Dadou (Glycine max), Ludimian (Gossypium hirsutum) and Zhima (Sesamum indicum), Xiangrikui (Helianthus annuus), and two food crops, Shuidao (Oryza sativa Japonica Group), Yumi (Zea mays), and an aquatic plant, Zhongguolian (Nelumbo nucifera), twenty homologous genes encoding PDCT were identified. Their protein physicochemical properties, gene structure and evolutionary relationship, promoter elements and so on were analyzed. The results of this study provide bioinformatics reference for further exploration of breeding applications of this family gene.

1 Results and Analysis

1.1 Identification of the PDCT family

The protein sequence of PDCT (AtROD1) reported in Arabidopsis thaliana was used as query, and blast was used to search the protein data of Glycine max, Brassica napus, Helianthus annuus, Gossypium hirsutum, Arachis hypogea, Sesamum indicum, Oryza sativa Japonica Group, Zea mays and Nelumbo nucifera. A total of 31 initial candidate members were obtained, and then 20 candidate PDCT families were obtained by using conservative domain analysis to remove redundancy. The NCBI database revealed 20 genes belonging to the PAP2_Like Superfamily, which is consistent with the PDCT gene found in Bima (Ricinus communis L.). In addition, the four PDCT homologous genes identified in Brassica napus were consistent with the reported identification results, which further confirmed the reliability of our results. According to their positions on chromosomes, we named 20 PDCT homologous genes successively in the form of "species +PDCT+ number". The nucleotide sequence of PDCT gene family was significantly different. The shorter ones were NnPDCT1 (1 630 bp) and NnPDCT2 (1 391 bp) of Nelumbo nucifera, and the longest was BnPDCT4 (8 875 bp) of Brassica napus (Table 1).

Table 1 Basic information of PDCT gene family members in tested 9 plant species

1.2 Analysis of basic physicochemical properties of PDCT family proteins

We then characterized the basic physicochemical properties of 20 PDCT family proteins (Table 2). Although the length of PDCT gene varied greatly among the nine tested species, the protein sequences encoded by PDCT gene ranged from 224 aa to 307 aa. The molecular weight of 20 PDCT protein family members of nine species was 25.13~33.20 kDa, and the amino acids with positive charge were more than those with negative charge. The theoretical isoelectric points are all greater than 7 and alkaline, ranging from 8.15 to 10.06, indicating that the PDCT family proteins are rich in basic amino acids. The results of ProtScale protein analysis showed that the PDCT protein lipid solubility index of nine species was more than 80, and the size was between 89.19 and 107.01, showing high lipid solubility. The hydrophobic index was greater than 0, indicating that PDCT family proteins were hydrophobic. The value of instability coefficient is bounded by 40. If the protein is less than 40, it is stable; if the protein is more than 40, it is unstable. From this, we can see that in the nine species, 8 PDCT family proteins showed instability, and the remaining 12 proteins showed relatively stable.

Table 2 Basic physiological and chemical features of PDCT proteins identified

1.3 Structure and phosphorylation site analysis of PDCT family proteins

We also analyzed the structure and phosphorylation sites of 20 PDCT family proteins (Table 3). The results of PDCT protein secondary structure prediction showed that they had similar characteristics in composition: The α-spiral and random coil structures accounted for 32.86%~52.3% and 33.48%~48.41%, respectively. Transmembrane analysis showed that PDCT protein had four transmembrane helical structures in all dicotyledonous plants. In the monocotyledonous plants rice and maize, the transmembrane structure of PDCT proteins was shown to be 3. Phosphorylation site analysis showed that the phosphorylation sites of PDCT family proteins were relatively small, ranging from 13 to 31, and mainly concentrated in Serine (Ser) and Threonine (Thr), while Tyrosine (Tyr) phosphorylation sites were less (Table 3). In addition, no signal peptides were detected in 20 PDCT family proteins, indicating that PDCT proteins were probably non-secreted proteins.

Table 3 Phosphorylation sites and structures of PDCT proteins

1.4 Functional domain analysis of PDCT family protein sequences

We also predicted the functional domains of 20 PDCT family proteins, and found that there were 15 conserved moieties with high similarity in the PDCT protein sequences (Figure 1A). Among the 15 conserved motifs, motifs 1-6 showed high conserved distribution. In general, although the species differed greatly, the number of conserved motifs varied little, ranging from 7 to 10 (Figure 1B). SMART online detection of conservative motifs found that motif 1 was SOCS_box domain, where motif 2 was PAP2_3 Domain.

Figure 1 Prediction of conserved motifs in PDCT family proteins Note: A: Highly conserved motif patterns of PDCT gene families; B: The distribution and sequence of motifs

1.5 Chromosome localization of PDCT gene in nine species

According to the genome sequence information of PDCT in nine species, its distribution on chromosomes was also mapped (Figure 2). The number of PDCT genes identified was small, and the distribution of PDCT genes on the chromosomes of each species was relatively uniform. In peanut, AhPDCT gene was distributed on chromosome 2 (Chrome02), chromosome 9 (Chrome09) and chromosome 19 (Chrome19), respectively. In Brassica napus, BnPDCT1 and BnPDCT2 were located on genome A, BnPDCT1 was located on chromosome 3 (ChromeA3), BnPDCT2 was located on chromosome 5 (ChromeA5). Both BnPDCT3 and BnPDCT4 were located on chromosome 5 of the C genome (ChromeC5), and more distributions were shown in Figure 2.

Figure 2 Chromosomal distribution of PDCT genes Note: A: Arachis hypogaea; B: Brassica napus; C: Glycine max; D: Gossypium hirsutum; E: Helianthus annuus; F: Sesamum indicum; G: Nelumbo nucifera; H: Oryza sativa Japonica Group; I: Zea mays

1.6 Evolutionary relationship and gene structure analysis of PDCT gene family in nine species

In order to explore the evolution of the PDCT gene family, we constructed an evolutionary relationship tree based on the PDCT protein sequences in the nine species identified in this study and the previously reported PDCT protein sequences in Arabidopsis thaliana, Linum usitatissimum L. and Ricinus communis L.. The results showed that the PDCT genes in the same species were relatively close in evolutionary distance. Further analysis showed that the evolution of AhPDCT2 and AhPDCT3 subtypes in peanut was relatively close to that of AtPDCT1. In Gossypium hirsutum Linn., GhPDCT1 and GhPDCT2, GhPDCT3 and GhPDCT4 were the closest evolutionary distances, respectively. In Brassica napus, BnPDCT1 and BnPDCT4, BnPDCT2 and BnPDCT3 were the closest evolutionary distances, respectively. In addition, the evolutionary relationship also shows that compared with other dicotyledons, sesame and sunflower are closer to monocotyledon rice and corn in evolutionary distance (Figure 3).

Figure 3 Phylogenetic analysis of PDCT genes Note: hypogaea: Kelly dots; Brassica napus: Dark cyan dots; Glycine max: Grey dots; Gossypium hirsutum: Brown dots; Helianthus annuus: Light cyan dots; Sesamum indicum: Pink dots; Ricinus communis: Dark blue dots; Arabidopsis thaliana: Light blue dot; Linum usitatissimum: Orange dots; Nelumbo nucifera: Purple square; Oryza sativa Japonica Group: Brown triangle; Zea mays: Orange triangle

In order to further understand the evolutionary relationship of the PDCT gene family, we analyzed the structure of 20 PDCT genes identified in this study. The results showed that the genetic structure of the same plant had higher similarity in composition, which was consistent with the results of evolutionary analysis. The PDCT genes of dicotyledonous land plants, such as Brassica napus and Arachis hypogaea, mostly contain 3 exons and 2 introns, while monocotyledonous rice and maize and aquatic lotus contain 2 exons and 1 intron. The intron phase of PDCT genes in all nine species was 0, indicating a high degree of conservation (Figure 4).

Figure 4 Gene structure analysis of PDCT Note: The green and purple bar represent the CDS and the un-translated region (UTR), respectively, while the black line represents the intron; Numbers (0, 1, 2) stand for the intron phase

1.7 Analysis of cis-acting elements of PDCT gene promoter

Finally, we conducted online analysis of promoter sequences of 20 PDCT genes to search for possible cis-acting elements. The results showed that the promoters of 20 genes contained 2 836 cis-acting elements, excluding the elements of unknown function. Among them, TATA-box (974) and CAAT-box (720), which were the core promoter elements of transcription, accounted for the largest proportion, followed by stress response elements MYB (112) and MYC (73). In addition, we found 206 cis-acting elements associated with light response, including 14 species of G-box, Box 4, GT1-motif distributed in all identified PDCT genes. The number of elements related to light response of each gene ranged from 4 to 9, among which AhPDCT1 had the largest number (Figure 5).

Figure 5 Cis-acting elements structures in the promoter region of PDCT family members

2 Discussion

Unsaturated fatty acids, especially PUFA, play an important role in human healthy diet. Therefore, controlling the synthesis of unsaturated fatty acids to improve the quality of edible oil in crop breeding has become a new direction for oil crop improvement in the future. PDCT has been shown to be a key enzyme for PUFA synthesis in plant seeds during TAG accumulation. For example, defects of PDCT in Arabidopsis thaliana seeds lead to a 40% reduction in PUFA content in mutant seeds (Lu et al., 2009). Studies on Linum usitatissimum and Brassica napus also confirmed the important role of PDCT in unsaturated fatty acid synthesis of oilseed plants (Wickramarathna et al., 2015). At present, PDCT gene identification and preliminary functional studies are limited to a few model plants, and the presence and function of PDCT in other plants have not been analyzed. Therefore, the identification and functional studies of PDCT in plants, especially in oil crops, need to be further strengthened. In recent years, the continuous improvement of species genome information has made it possible to identify and analyze various gene families including the PDCT family in the whole genome. In this study, a total of 20 PDCT genes were identified in 9 different plants including Arachis hypogaea, Brassica napus, Glycine max, Gossypium, Sesamum indicum, Helianthus annuus, Nelumbo nucifera, Oryza sativa and Zea mays by bioinformatics methods. The four PDCT genes identified in Brassica napus were consistent with the published results. The results of this study provide important biological information data for further study of gene function of PDCT family and possible breeding applications.

Our results show that members of the PDCT family have a certain degree of conservation in evolution, which is mainly manifested as: (1) The stability of the properties and structure of the PDCT proteins: the 20 identified proteins are not only similar in the number of amino acids, but also about 30 kDa in molecular weight, showing basic characteristics. The positive charge residue was more than the negative charge residue, and all showed hydrophobic proteins. In addition, PDCT family proteins were similar in the proportion of secondary structure, with α helices and random curls occupying the majority, and threonine and serine dominating the phosphorylation site analysis. (2) Conserved gene structure: Despite the large differences in sequence length, PDCT family genes are relatively similar in structure. Dicotyledonous land plants all have 3 exons and 2 introns, while aquatic plants lotus and monocotyledonous plants rice and maize all have 2 exons and 1 and intron, and the intron phase is highly conserved with 0. (3) Conservation of evolutionary distance: The phylogenetic tree results were consistent with the intraspecific relationships of PDCT family proteins. (4) Conserved domain: PAP2_3 domain exists in both monodicotyledonous plants and terrestrial aquatic plants, and is located near the C-terminal of the protein.

Our results also showed that there were significant differences in the copy number of PDCT family genes in the nine plants tested, but all of them were within 4. The PDCT in food crops (Oryza sativa and Zea mays) and model plants (Arabidopsis thaliana) were single-copy, while the PDCT in oil crops (Arachis hypogaea, Brassica napus, Glycine max, Gossypium, and Helianthus annuus) were multi-copy (2 or more). Interestingly, a single copy of PDCT was found in sesame seeds, but this result deserves further clarification. The results of PDCT gene copy number found in this study were consistent with those previously reported in Arabidopsis thaliana (Lu et al., 2009), Linum usitatissimum (Wickramarathna et al., 2015) and Brassica napus.

Phylogenetic tree analysis showed that these single-copy species are relatively close in evolution, and the PDCT of Sesamum indicum is relatively distant from that of other typical oil crops, such as Ricinus communis, Linum usitatissimum, Arachis hypogaea and Glycine max. These results suggest that PDCT may play a role in the evolution or differentiation of oil plants and food crops, and on the other hand, PDCT in food crops may have different enzyme activities or functions than PDCT in oil crops. Interestingly, comparing the evolutionary distance between PDCT of the only aquatic crop, Nelumbo nucifera, and that of other terrestrial crops, PDCT may also be related to the differentiation of terrestrial and aquatic plants. Such information provides valuable reference for the further study of PDCT gene function and evolution.

Our results also showed that there are a large number of stress response elements and light response elements in the promoters of PDCT family genes, suggesting that the expression of PDCT may be influenced by many growth and environmental stress signals, especially light response signals. Therefore, it is reasonable to believe that PDCT may also play an important role in plant growth and environmental stress, especially in light response.

3 Materials and Methods

3.1 Data download and verification

In this study, proteome sequences, genome sequences and gene annotation files of Arachis hypogea (V3), Brassica napus (V2), Glycine max (V2.1), Gossypium hirsutum (V1), Helianthus annuus (V1), Sesamum indicum (V1), Nelumbo nucifera (V1.1), Oryza sativa Japonica Group (V4) and Zea mays (V4) were downloaded from NCBI (National Center of Biotechnology Information) database (https://ftp.ncbi.nlm.nih.gov/genomes/), respectively.

The Arabidopsis thaliana PDCT (AtROD1) protein sequence reported in the literature was used as query and Blast 2.5.0 software (Camacho et al., 2009) was used to search the PDCT gene homologous proteins of nine species (E-value=1e^-10). After manual de duplication of screening results, the initial protein of PDCT family was obtained, and then the initial protein sequence was uploaded to CDD online website (http://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi) and SMART website (http://smart.embl-heidelberg.de/) for preliminary functional domain detection and analysis. After manual de redundancy, candidate members of PDCT family were obtained for subsequent analysis.

3.2 Analysis of PDCT Family Proteins

The candidate proteins obtained by the above steps were analyzed, and the molecular weight and theoretical isoelectric point of the obtained 20 PDCT family proteins were predicted using Compute pI/MW (https://web.expasy.org/compute_pi/) from ExPASy. Online software ProtParam (https://web.expasy.org/protparam/) predicted first-order physicochemical properties such as charge residual base, total atomic number, instability coefficient, fat solubility index and hydrophobic index. PHD (https://npsaprabi.ibcp.fr/cgibin/npsa_automat.pl?page=/NPSA/npsa_phd.hml) method was used to predict the secondary structure of proteins in the PDCT family. SignalP-5.0 (http://www.cbs.dtu.dk/services/SignalP/) realized the detection of protein signal peptide. TMHMM website (http://www.cbs.dtu.dk/services/) achieved the analysis of transmembrane structure. NetPhos 3.1 Server (www.cbs.dtu.dk/services/NetPhos/) provided phosphorylation site detection. Finally, MEME (http://meme-suite.org/) was used to predict and analyze the conservative motif of PDCT protein sequences of nine species.

3.3 Gene structure, chromosome localization, evolutionary relationship and promoter analysis of PDCT

The location information of PDCT genes in nine species was mapped on chromosomes by TBtools software. MEGA-X software (Kumar et al., 2018) realized the construction of phylogenetic tree: After multi-sequence alignment of PDCT protein sequences by ClustalW (Gap Opening Penalty=10, Gap Extension Penalty=0.2, Delay Divergent Cutoff=30%), the evolutionary tree was obtained by maximum likelihood method (Bootstrap=1000, Partial deletion).

The genetic structure of PDCT was mapped using GSDS (http://gsds.cbi.pku.edu.cn/) online software. PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) implemented the analysis of promoter cis-acting elements. And using TBtools software to draw the structure diagram.

Authors’ contributions

RXR was the performer of the experimental research and the writer of the paper draft. HJ and YFZ participated in data collation, results discussion and paper draft writing. SJX was the project leader, experimental designer and modifier of the paper. All authors read and approved the final manuscript.

Acknowledgments

This research was jointly supported by the National Natural Science Foundation of China (31971907) and the Subject Innovation and Talent Introduction Program of Higher Education (111 Project, B14016)

References

Bates P.D., Ohlrogge J.B., and Pollard M., 2007, Incorporation of newly synthesized fatty acids into cytosolic glycerolipids in pea leaves occurs via acyl editing, J. Biol. Chem., 282(43): 31206-31216

https://doi.org/10.1074/jbc.M705447200

PMid:17728247

Browse J., McConn M., James D., and Miquel M., 1993, Mutants of Arabidopsis deficient in the synthesis of alpha-linolenate. Biochemical and genetic characterization of the endoplasmic reticulum linoleoyl desaturase, J. Biol. Chem., 268(22): 16345-16351

https://doi.org/10.1016/S0021-9258(19)85427-3

PMid:8102138

Cahoon E.B., Shockey J.M., Dietrich C.R., Gidda S.K., Mullen R.T., and Dyer J.M., 2007, Engineering oilseeds for sustainable production of industrial and nutritional feedstocks: solving bottlenecks in fatty acid flux." Curr. Opin. Plant Biol., 10(3): 236-244

https://doi.org/10.1016/j.pbi.2007.04.005

PMid:17434788

Camacho C., Coulouris G., AvagyanV., Ma N., Papadopoulos J., Bealer K., and Madden T.L., 2009, BLAST+: architecture and applications, BMC Bioinformatics, 10: 421

https://doi.org/10.1186/1471-2105-10-421

PMid:20003500 PMCid:PMC2803857

Hunter J.E., 1990, n-3 fatty acids from vegetable oils, Am. J. Clin. Nutr., 51(5): 809-814.

https://doi.org/10.1093/ajcn/51.5.809

PMid:1970702

Kinney A.J., 1996, Designer oils for better nutrition, Nat. Biotechnol., 14(8): 946.

https://doi.org/10.1038/nbt0896-946

PMid:9631031

Kumar S., Stecher G., Li M., Knyaz C., and Tamura K., 2018, MEGA X: Molecular Evolutionary Genetics Analysis across Computing Platforms, Mol. Biol. Evol., 35(6): 1547-1549

https://doi.org/10.1093/molbev/msy096

PMid:29722887 PMCid:PMC5967553

Lu C.F., Napier J.A., Clemente T.E., and Cahoon E.B., 2011, New frontiers in oilseed biotechnology: meeting the global demand for vegetable oils for food, feed, biofuel, and industrial applications, Curr. Opin. Biotechnol., 22(2): 252-259

https://doi.org/10.1016/j.copbio.2010.11.006

PMid:21144729

Lu C., Xin Z., Ren Z., Miquel M., and Browse J., 2009, An enzyme regulating triacylglycerol composition is encoded by the ROD1 gene of Arabidopsis, Proc Natl Acad Sci USA, 106(44): 18837-18842

https://doi.org/10.1073/pnas.0908848106

PMid:19833868 PMCid:PMC2774007

Shanklin J. and Cahoon E.B., 1998, Desaturation and Related Modifications of Fatty Acids1, Annu Rev. Plant Physiol. Plant Mol. Biol., 49: 611-641

https://doi.org/10.1146/annurev.arplant.49.1.611

PMid:15012248

Wallis J.G., Watts J.L., and Browse J., 2002, Polyunsaturated fatty acid synthesis: what will they think of next? Trends Biochem. Sci., 27(9): 467

https://doi.org/10.1016/S0968-0004(02)02168-0

PMid:12217522

Wickramarathna A.D., Siloto R.M., Mietkiewska E., Singer S.D., Pan X., and Weselake R.J., 2015, Heterologous expression of flax PHOSPHOLIPID:DIACYLGLYCEROL CHOLINEPHOSPHOTRANSFERASE (PDCT) increases polyunsaturated fatty acid content in yeast and Arabidopsis seeds, BMC Biotechnol., 15: 63

https://doi.org/10.1186/s12896-015-0156-6

PMid:26123542 PMCid:PMC4486708

Williams C.M. and Burdge G., 2006, Long-chain n-3 PUFA: plant v. marine sources, Proc. Nutr. Sci. 65(1): 42-50

https://doi.org/10.1079/PNS2005473

PMid:16441943