Medium-chain FAs (MCFAs) include FAs of 8-14 carbons in lengths, generated by the hydrolysis of FA from acyl carrier protein between the C8 and C14 stages of elongation via variants of FATB with defined chain length specificities ( Figure 1 ). Lipids containing MCFAs are naturally produced in palm kernel (Elaeis guineensis), coconut (Cocos nucifera), and cuphea genus (Cuphea pulcherrima, Cuphea viscosissima, Cuphea palustris, Cuphea hookeriana), and these plants derived MCFAs serve as potential feedstocks for jet fuel and industrial products such as cosmetics and detergents (Dyer et al., 2008; Kallio et al., 2014; Park et al., 2021). To engineer the production of MCFAs in oilseed crops, FATB variants that specifically hydrolyze C8-C14 FAs from acyl-ACPs were isolated from California bay (Umbellularia californica) and Cuphea and expressed in Arabidopsis thaliana, Camelina sativa, and Brassica napus. Heterologous expression of U. californica FATB1 produced MCFAs primarily consisting of lauric acid (C12:0) up to 21%, 37%, and 50% of seed oil in Camelina, Arabidopsis, and B. napus, respectively (Eccleston et al., 1996; Voelker et al., 1996; Tjellström et al., 2013; Kim et al., 2015b). FATB variants from different Cuphea species showed different efficiencies and substrate chain length specificities when expressed in oilseed plants ( Table 1 ). Expression of FATB2 from C. hookeriana could produce MCFAs ranging from C8 to C14 with capric acid (C10:0) as the most abundant species accounted for approximate 12%, 22%, and 40% of total seed lipids in Camelina, Arabidopsis, and B. napus, respectively (Dehesh et al., 1996; Tjellström et al., 2013; Kim et al., 2015b). FATB1 from C. viscosissima produced 15% MCFAs with chain lengths varying from C8 to C14 in seed oil when expressed in Camelina, whereas FATB2 from C. palustris produced only myristic acid (C14:0) to 24% and 39% of seed oil in Camelina and Arabidopsis, respectively (Tjellström et al., 2013; Kim et al., 2015b). Relatively low levels of MCFAs (1.2%-7.5% of seed oil) were detected in seeds of Arabidopsis and Camelina expressing FATB1, FATB3 or FATB4 from C. pulcherrima, compared with FATBs from other Cuphea species. Based on data collected from Camelina, UcFATB1 and CpFATB2 seem to be the most effective FATB variants in producing MCFAs with UcFATB1 preferentially generating lauric acid (C12:0) and CpFATB2 exclusively producing myristic acid (C14:0). Furthermore, co-expression of ChFATB2 or CpFATB1 with a Cuphea medium-chain-specific 3-ketoacyl-ACP synthase (KAS4) that catalyzes the condensation of acyl-ACP with malonyl-ACP increased MCFA content by up to 40% in B. napus seed oil as compared with that of plants expressing FATB alone (Dehesh et al., 1998). Disruption of acyl-ACP synthetase (AAE15/16) that re-activates FAs released from acyl-ACP in Arabidopsis overexpressing Cuphea FATB further enhanced MCFA accumulation in seeds (Tjellström et al., 2013). In addition to the FAT-type thioesterases containing two “hotdog” folds (Mayer and Shanklin, 2005), another acyl-ACP thioesterase family, acyl-lipid thioesterase (ALT) with a single “hotdog” fold, is generally present in all classes of plants (Kalinger et al., 2020). Overexpression of Arabidopsis ALT1 or ALT4 in Camelina seeds and Nicotiana benthamiana leaves yielded C6–C14 MCFA at a relatively lower level (as much as 3.5% of seed oil) compared to the effective FATB isoforms ( Table 1 ; Kalinger et al., 2021).

The hydroxylation of FAs is mediated by the action of FA hydroxylase (FAH), the first functionally divergent FAD2 homolog to be identified ( Figure 1 ). That they are mechanistically related is evidenced by reports that as few as four substitutions between desaturase and hydroxylase can interconvert their functionality (Broun et al., 1998; Broadwater et al., 2002). FAs with hydroxyl groups attached to the acyl chain are useful feedstocks for the formulation of plastics and lubricants (Dyer et al., 2008). The major natural source of hydroxy FAs for industrial uses is castor bean (Ricinus communis), which accumulates 90% hydroxy FAs (mostly ricinoleic acid, C18:1-OH) in its seed oil. The enzyme responsible for synthesizing hydroxy FAs in castor is FAH12 (van de Loo et al., 1995). Heterologous expression of RcFAH12 in Arabidopsis led to accumulation of hydroxy FAs consisting of primarily ricinoleic acid accounted for up to 19% of seed oil in wild type, fad2/fae1, fae1, fad3, or fad3/fae1 backgrounds (Broun and Somerville, 1997; Smith et al., 2003). Similarly, wild-type Camelina expressing RcFAH12 produced approximately 15% hydroxy FAs in seed oil (Aryal and Lu, 2018). Expression of Hiptage benghalensis hydroxylases HbFAH12-1 and HbFAH12-2 in Arabidopsis fad2/fae1 mutant yielded up to 21% and 18% hydroxy FA, respectively, in seed oil (Zhou et al., 2013). In contrast to the plant derived FAH12, a FAH homolog isolated from a fungal pathogen, Claviceps purpurea, produced 25% hydroxy FAs in seed oil when expressed in the Arabidopsis fad2/fae1 mutant (Meesapyodsuk and Qiu, 2008). While RcFAH12 can effectively produce hydroxy FAs in target plants and most plant engineering strategies to date have used RcFAH12 to synthesize hydroxy FAs, searching for a more effective FAH from other species to further enhance the accumulation of hydroxy FAs in bioengineered crops might be productive. Lesquerella (Physaria fendleri), a Brassicaceae species closely related to Arabidopsis and Camelina, produces about 50% of lesquerolic acid (C20:1-OH), an elongated form of ricinoleic acid, in its seed oil (Horn et al., 2016). Thus, Lesquerella represents a promising alternative industrial oilseed for HFA production, and specialized HFA-related factors in Lesquerella represent a promising source for engineering HFA accumulation in other crops (Horn et al., 2016; Chen et al., 2021). A recent study demonstrated the production of erythro-9,10-dihydroxystearate, a vicinal diol by an acyl-ACP desaturase variant via dioxygenase chemistry (Whittle et al., 2020). The identification of additional genes that are more efficient at vicinal diol production may facilitate large scale production of these compounds that are difficult to synthesize chemically.

An epoxy group with its oxygen bridging between adjacent carbons of fatty acyl chains conveys unique chemical reactivity useful for the production of plastics, polymers, coatings, and glues. Epoxy FAs are enriched in seed oils of certain plant species belonging to the Asteraceae and Euphorbiaceae families (Cahoon et al., 2002). Interestingly, the biosynthesis of epoxy FAs in different plant species is catalyzed by different classes of epoxygenase (EPX) enzymes. Those responsible for epoxy FA biosynthesis in Asteraceae species such as Crepis palaestina, Stokesia laevis, and Vernonia galamensis are divergent forms of the FAD2 desaturase, whereas epoxygenases in Euphorbiaceae species such as Euphorbia lagascae are cytochrome P450 enzymes (Bafor et al., 1993; Liu et al., 1998; Cahoon et al., 2002). Despite the distinction of these two classes of EPX, heterologous expression of these enzymes in plants resulted in accumulation of similar levels of epoxy FAs (mostly vernolic acid, C18:1- Δ¹²-epoxy FA) in seed oils ( Table 1 ). Expression of the FAD2-like EPX coding genes from C. palaestina and S. laevis led to accumulation of approximately 2.4%-8% epoxy FAs in seed oils of Arabidopsis and soybean, and the cytochrome P450-type EPX from E. lagascae produced about 8% epoxy FAs in soybean somatic embryos (Singh et al., 2001; Cahoon et al., 2002; Hatanaka et al., 2004; Li et al., 2010). Providing the exogenous CpEPX with more linoleic acid (C18:2) substrate by disrupting FAD3 and FAE1 in Arabidopsis increased the levels of epoxy FAs to 8.6% of seed oil (Zhou et al., 2006). Previous studies suggested that heterologous expression of either type of EPX can reduce the accumulation of linoleic acid in target plants probably caused by decreased activity of the endogenous FAD2 enzyme, and co-expression of EPX with a typical FAD2 dramatically enhanced the production of epoxy FAs to 21% of seed oil in Arabidopsis fad3/fae1 mutant (Singh et al., 2001; Cahoon et al., 2002; Zhou et al., 2006). In addition, epoxy FAs can also be engineered in non-seed tissues. Expression of ElEPX in tobacco (Nicotiana tabacum) calli produced epoxy FAs accounted for 15% of total lipids (Cahoon et al., 2002). Transient expression of SlEPX in petunia (Petunia hybrida) leaves or a FAD2-like EPX from V. galamensis in N. benthamiana leaves resulted in accumulation of 0.5% or 8.7% epoxy FAs in total leaf lipids, respectively (Li et al., 2010; Sun et al., 2022). Moreover, co-expression of VgEPX with VgFAD2 increased the level of epoxy FAs to 13.1% of total lipids in N. benthamiana leaves (Sun et al., 2022). Collectively, divergent classes of EPX from different plant species seem to be equally effective in producing epoxy FAs in seeds and providing more linoleic acid by overexpressing a “typical” FAD2 or disrupting FAD3 and FAE1 is critical for further increasing epoxy FA levels.

Omega-7 unsaturated FAs (ω-7 FAs) are potential feedstocks for the production of octene, a high-demand industrial product used for polyethylene production (Nguyen et al., 2010). Some plants (e.g., milkweed [Asclepias syriaca] and cat’s claw vine [Doxantha unguis-cati]) can naturally produce ω-7 FAs (e.g., palmitoleic acid 16:1Δ⁹ and cis-vaccenic acid 18:1Δ¹¹) by Δ⁹-acyl-ACP desaturase (AAD) with high specificity for 16:0-ACP (Cahoon et al., 1997; Cahoon et al., 1998). Heterologous expression of the milkweed 16:0-ACP desaturase in Arabidopsis failed to produce detectable ω-7 FA, while the Doxantha 16:0-ACP desaturase produced approximately 28% and 9% ω-7 FAs in Arabidopsis and Brassica seed oil, respectively (Bondaruk et al., 2007). An AAD variant with high specificity for converting 16:0-ACP to 16:1 Δ⁹-ACP was selected from a pool of randomized mutants of castor AAD (Cahoon and Shanklin, 2000), and expression of this engineered enzyme (Com25) in Arabidopsis seeds resulted in accumulation of ω-7 FAs to 14% of seed oil (Nguyen et al., 2010). Increasing the level of 16:0-ACP by silencing the 16:0-ACP elongase, β-ketoacyl-ACP synthase II (KASII/FAB1), in fae1 mutant overexpressing Com25 further increased the content of ω-7 FAs to 56% of seed oil. Co-expressing Com25 with two fungal Δ⁹-16:0-CoA desaturases from Stagonospora nodorum (SnD9D) and Aspergillus nidulans (AnD9D) in fab1/fae1 mutant increased ω-7 FA content to 71% of seed oil by desaturating saturated FAs after transfer from the plastid to the ER (Nguyen et al., 2010). A similar strategy co-expressing Com25 and a Δ⁹-16:0-CoA desaturase from Caenorhabditis elegans (FAT5) in Camelina seeds with 16:0-ACP substrate pools increased by silencing genes encoding KASII/FAB1, FAE1, and16:0-ACP thioesterase (FATB) increased ω-7 FAs to 60-65% of seed oil (Nguyen et al., 2015).

The FAD2 desaturases that produce conjugated FAs by converting Δ⁹ and Δ¹² double bonds to Δ¹¹ and Δ¹³ double bonds are designated as FA conjugases (FADX) ( Figure 1 ; Cahoon et al., 1999; Dyer et al., 2002). The higher oxidation rates of conjugated FAs relative to typical polyunsaturated FAs make them useful as drying agents in paints and inks. Conjugated FAs can also serve as health supplements as they have been reported to have fat-reducing and anticancer effects in animals (Lee et al., 2002; Dyer et al., 2008; Yuan et al., 2014). Natural sources of conjugated FAs include tung tree (Vernicia fordii), Momordica charantia, Impatiens balsamina, and Calendula officinalis, and genes encoding FADX enzymes have been isolated from these plant species and evaluated for their efficacy in producing conjugated FAs in model plants and oilseed crops. Ectopic expression of FADX coding genes from I. balsamina, M. charantia, and C. officinalis in somatic soybean (Glycine max) embryos resulted in production of conjugated FAs to approximately 5%, 18%, and 22% of total FAs, respectively (Cahoon et al., 1999). For engineering approaches carried out in Arabidopsis seeds, mutants with FA desaturase 3 (FAD3) and FAE1 disrupted are used to provide more substrates (linoleic acid) for FADX. Arabidopsis fad3/fae1 mutants expressing FADX genes from V. fordii, M. charantia, and C. officinalis accumulated approximately 6%, 13%, and 15% conjugated FAs in seed oil, respectively (Cahoon et al., 1999; Cahoon et al., 2006). Recent attempts to engineer conjugated FAs in plant vegetative tissues by expressing VfFADX in Arabidopsis successfully produced conjugated FAs (eleostearic acid, 18:3, Δ9c, Δ11t, Δ13t) to 2% of total neutral lipids in leaves (Yurchenko et al., 2017). Among all FADX enzymes tested so far, CoFADX seems to be the most effective enzyme for producing high levels of conjugated FAs in both Arabidopsis and soybean, but different FADX orthologs produce different types of conjugated FAs. Conjugated FAs produced by CoFADX comprise exclusively calendic acid (18:3, Δ8t, Δ10t, Δ12c), while eleostearic acid is the primary conjugated FAs detected in transgenic plants expressing VfFADX or McFADX. In future efforts to engineer conjugated FAs, it will be important to select a FADX that produces high levels and desired types of conjugated FAs.

Very-long-chain polyunsaturated FAs (VLCPUFA) such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are FAs with 20 or 22 carbons and 4 to 6 double bonds. EPA and DHA are valuable nutraceuticals because of their beneficial roles in fetal neuronal system development and cardiovascular diseases prevention (Tocher et al., 2019). Marine fish oils are a major source of EPA, DHA and other ω³-VLCPUFAs. However, the marine-sourced EPA and DHA are insufficient to meet the increasing demand for these FAs in human diet (Tocher et al., 2019). To develop sustainable alternative sources of EPA and DHA, in the past two decades, enormous research effort has been directed at engineering ω³-VLCPUFAs in plant oils, and significant progress has been achieved in producing EPA and DHA in oilseed crops (Napier et al., 2019). There are two pathways that can produce VLCPUFAs from linoleic acids: 1) the conventional pathway including Δ⁶-desaturase (DES), Δ⁶-elongase (ELO), Δ⁵-DES, Δ⁵-ELO, and Δ⁴-DES; and 2) the alternative pathway using Δ⁹-ELO, Δ⁸-DES, Δ⁵-DES, Δ⁵-ELO, and Δ⁴-DES. Both pathways have been introduced into plants and successfully produced EPA and DHA in both seed and vegetative tissues (Qi et al., 2004; Wu et al., 2005; Wood et al., 2009; Petrie et al., 2010). It was later found that, besides the minimum five enzymes required for DHA biosynthesis in plants, introduction of a highly active Δ¹²-DES from Lachancea kluyveri and a Δ¹⁵/ω³-DES with broad substrate specificity from Pichia pastoris could effectively increase the ratio of ω³/ω⁶ FAs and therefore the level of DHA in seed oil (Petrie et al., 2012; Petrie et al., 2014). To optimize the engineering strategy to produce high levels of EPA and DHA in plants, DES and ELO enzymes from a wide variety of species including algae, fungi, oomycetes, mosses, animals, and flowering plants were introduced into plants to test their efficacy in producing EPA and DHA in various plant species (Abbadi et al., 2004; Qi et al., 2004; Wu et al., 2005; Wood et al., 2009; Petrie et al., 2010; Petrie et al., 2012; Petrie et al., 2014; Han et al., 2020; Han et al., 2022). The optimal enzyme combination that achieved the highest levels of EPA and DHA (over 20%) in seed oil reported to date consists of a Δ¹²-DES from Phytophthora sojae, Δ¹⁵/ω³-DES from Phytophthora infestans, Δ⁶-DES from Ostreococcus tauri, Δ⁶-ELO from Physcomitrella patens, Δ⁵-DES from Thraustochytrium sp., Δ⁵-ELO from O. tauri, and Δ⁴-DES from Ostreococcus RCC809 (Han et al., 2020; Han et al., 2022).

Cyclopropane FAs (CPAs), such as dihydrosterculic acid (9, 10-methylene octadecanoic acid) and lactobacillic acid (11, 12 methylene octadecanoic acid), are specialized FAs that contain a cyclopropane group (three-carbon carbocyclic ring) within the carbon chain. They are found in bacteria and certain plant species such as Litchi chinensis. The highly strained and reactive carbocyclic ring of CPA readily opens to form methyl-branched fatty acids, which exhibit unique physical and chemical properties such as low melting temperatures, resistance to oxidation, and propensity for self-polymerization, making CPAs suitable for application in lubricants, paints, and coatings (Carlsson et al., 2011). Ectopic expression of genes encoding cyclopropane synthase (CPS), the enzyme catalyzes the conversion of monounsaturated FAs to CPAs ( Figure 1 ), successfully produced CPAs in plants that normally lack these compounds. Intriguingly, the CPS gene from Escherichia coli is more effective than CPS homologs isolated from plant species for synthesizing CPAs in plants ( Table 1 ). Engineering of CPAs in Arabidopsis and Camelina seeds were carried out in mutant lines with reduced FA desaturase 2 (FAD2) and FA elongase 1 (FAE1) that accumulate increased levels of oleoyl substrates for CPA production. Expression of CPS genes cloned from Sterculia foetida and cotton (Gossypium hirsutum) in Arabidopsis fad2/fae1 mutant lines led to the accumulation of CPA to between 0.05% and 1% of total seed FAs, respectively (Yu et al., 2011). In contrast, the expression of E. coli CPS resulted in nearly 10% CPA accumulation in seed oil in Arabidopsis and Camelina (Yu et al., 2014; Yu et al., 2018). Recently, researchers tested the ability of CPS enzymes to synthesize CPAs in plant vegetative tissues by transiently expressing cotton or E. coli CPS genes in N. benthamiana leaves and found that GhCPS1 produced up to 1% CPA of total leaf FAs while EcCPS expression led to the accumulation of CPA up to 3.7% of total lipids (Okada et al., 2020). The levels of CPAs in leaf lipids can be further elevated to 4.8% and 11.8% by silencing the expression of endogenous NbFAD2 in leaves expressing GhCPS1 and EcCPS, respectively (Okada et al., 2020). Therefore, future strategies to engineer CPA accumulation in plant seed and vegetative tissues will exploit EcCPS rather than plant CPS variants.

Incorporation of specialty FA into TAG requires specialized enzymes that can recognize the specialized FA for catalyzing multiple steps of TAG assembly ( Figure 1 ). In the glycerolipid biosynthesis pathway, The ER-localized LPAT catalyzes the transfer of FAs from the acyl-CoA pool to LPA to form PA, which serves as a key intermediate for channeling FAs into TAGs and membrane lipids. Lysophosphatidylcholine acyltransferase (LPCAT) incorporates FAs into PC. PDCT and PLC catalyze the conversion of PC to DAG and thereby allow FAs esterified to PC to enter the DAG pool, which are subsequently converted to TAG by DGAT and PDAT. Below we describe some variants of these enzymes that are specialized for incorporating different types of specialty FAs into TAGs.

Acetyl-TAGs (acTAGs) are specialty TAGs with an acetate esterified to the sn-3 position in place of a long-chain fatty acid. Oils containing acTAGs exhibit reduced viscosity and therefore have high value in a wide variety of industrial applications such emulsifiers and lubricants. Specialized DGATs (DAcT) responsible for acTAG biosynthesis were isolated from Euonymus alatus and Euonymus fortune, plants that naturally produce acTAGs in their seeds. Heterologous expression of EaDAcT resulted in accumulation of 40% and 52% of acTAG in seeds of Arabidopsis and Camelina, respectively (Durrett et al., 2010). EfDAcT, functioning more efficiently than EaDAcT, produced an average of 72% acTAG in transgenic Camelina seeds (Alkotami et al., 2021). Interestingly, co-expression of CnLPAT from MCFA-containing coconut and EaDAcT from acTAG-enriched E. alatus in camelina plants expressing UcFATB1 produced acTAGs with MCFAs, which have yet to be found in nature, suggesting a potential synthetic-biology strategy for creating novel lipid structures in plants (Bansal et al., 2018).

DGAT enzymes catalyze the final committed step of TAG biosynthesis, and an efficient DGAT is key to enhancing TAG accumulation in plants ( Figure 1 ). Zienkiewicz et al. (2017) screened six out of 12DGATs from Nannochloropsis oceanica, a microalga that produces high amounts of TAGs, and identified DGAT5 (DGTT5) as the most efficient isoform for restoring TAG synthesis in a TAG synthesis-deficient mutant of yeast. Transient expression of NoDGTT5 in N. benthamiana leaves led to a 2-fold increase in TAG, and stable expression of NoDGTT5 in Arabidopsis increased leaf TAG contents by 6-fold and boosted seed oil content by 50% (Zienkiewicz et al., 2017). In another study of DGATs from microalga, DGAT1 from Chlorella ellipsoidea increased the oil content by 8–37% and by 12–18% in seeds of Arabidopsis and B. napus, respectively (Guo et al., 2017). In addition, mouse (Mus musculus) DGAT2, the predominant DGAT responsible for TAG biosynthesis in mouse, when transiently expressed in N. benthamiana leaves, produced over 20-fold more TAG than that of control leaves (Cai et al., 2019). Recently, a study of Cyperus esculentus, a unique plant accumulating large amounts of TAG in its underground tubers, revealed that its heterologous expression in N. tabacum increased the TAG content to 5.5% of leaf dry weight (DW), which is 7.2-fold and 1.7-fold higher than that in wild-type leaves and leaves expressing AtDGAT1, respectively (Gao et al., 2021). Moreover, CeDGAT2-2 expression resulted in a substantial increase in the proportion of oleic acid in N. tabacum leaves (Gao et al., 2021). In another study, heterologous expression of Arabidopsis DGAT1 reportedly led to a 7-fold increase in TAG contents in N. tabacum leaves (Bouvier-Navé et al., 2000). That the DGATs tested in different studies were driven by different promoters, expressed either transiently or stably, and tested in different plant tissues and species, precludes us from making meaningful comparisons for assessing the relative efficacy of DGATs from different sources. Thus, it would be useful to evaluate all promising DGATs under same conditions and in same tissues and target organisms.