Skip to main content
PMC home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2022 Nov 16;21(2):317–330. doi: 10.1111/pbi.13943

Engineering triacylglycerol accumulation in duckweed (Lemna japonica)

Yuanxue Liang 1, Xiao‐Hong Yu 1, Sanket Anaokar 1, Hai Shi 1, William B Dahl 2, Yingqi Cai 1, Guangbin Luo 3, Jin Chai 1, Yuanheng Cai 1, Almudena Mollá‐Morales 2, Fredy Altpeter 3, Evan Ernst 2,4, Jorg Schwender 1, Robert A Martienssen 2,4, John Shanklin 1,✉
PMCID: PMC9884027  PMID: 36209479

Summary

Duckweeds are amongst the fastest growing of higher plants, making them attractive high‐biomass targets for biofuel feedstock production. Their fronds have high rates of fatty acid synthesis to meet the demand for new membranes, but triacylglycerols (TAG) only accumulate to very low levels. Here we report on the engineering of Lemna japonica for the synthesis and accumulation of TAG in its fronds. This was achieved by expression of an estradiol‐inducible cyan fluorescent protein‐Arabidopsis WRINKLED1 fusion protein (CFP‐AtWRI1), strong constitutive expression of a mouse diacylglycerol:acyl‐CoA acyltransferase2 (MmDGAT), and a sesame oleosin variant (SiOLE(*)). Individual expression of each gene increased TAG accumulation by 1‐ to 7‐fold relative to controls, while expression of pairs of these genes increased TAG by 7‐ to 45‐fold. In uninduced transgenics containing all three genes, TAG accumulation increased by 45‐fold to 3.6% of dry weight (DW) without severely impacting growth, and by 108‐fold to 8.7% of DW after incubation on medium containing 100 μm estradiol for 4 days. TAG accumulation was accompanied by an increase in total fatty acids of up to three‐fold to approximately 15% of DW. Lipid droplets from fronds of all transgenic lines were visible by confocal microscopy of BODIPY‐stained fronds. At a conservative 12 tonnes (dry matter) per acre and 10% (DW) TAG, duckweed could produce 350 gallons of oil/acre/year, approximately seven‐fold the yield of soybean, and similar to that of oil palm. These findings provide the foundation for optimizing TAG accumulation in duckweed and present a new opportunity for producing biofuels and lipidic bioproducts.

Keywords: duckweed, Lemna minor, WRI1, DGAT, triacylglycerol, plant oil

Triacylglycerol accumulation in duckweed (Lemna japonica) increased by 108‐fold to 8.7% of dry weight upon the expression of an inducible CFP‐Arabidopsis WRI1, along with constitutive expression of mouse DGAT2 and a variant sesame OLEOSIN.

graphic file with name PBI-21-317-g004.jpg

Introduction

Lemnaceae, also known as duckweeds, are a family of aquatic monocotyledonous plants with 36 species recognized in five genera (Spirodela, Landoltia, Lemna, Wolffiella, and Wolffia; Acosta et al., 2021; Bog et al., 2019, 2020). Duckweeds are the smallest and fastest‐growing aquatic flowering plants (Les et al., 2002; Wang et al., 2010) with doubling time varying between 16 h and 3 days under optimal conditions (Leng et al., 1995; Rusoff et al., 1980). Their aquatic habit means their growth doesn't displace arable land for conventional crop growth. Robust growth in diverse environments makes duckweeds promising platforms for engineering the accumulation of biofuel and bioproduct feedstocks (Xu et al., 2012b).

The availability of high‐quality genome information is foundational for genetic and metabolic engineering strategies. In recent years, genome assemblies have been published for Spirodela polyrhiza (Michael et al., 2020; Wang et al., 2014), Lemna minor (Van Hoeck et al., 2015), Wolffia australiana (Michael et al., 2020) and Landoltia punctata (Hoang et al., 2019). Of the five duckweed genera, a full genome is so far missing only for Wolffiella, for which the estimated genome size is between ~600 Mb and 1914 Mb (Hoang et al., 2019; Wang et al., 2011). An important online resource that includes the complete sequence of L. japonica is also publicly available (Ernst and Martienssen, 2011). In addition to reliable genome information, efficient stable transformation protocols are needed for the generation of transgenic lines. Protocols based on Agrobacterium‐mediated transformation of callus cultures have been reported for multiple duckweed species (Boehm et al., 2001; Geng et al., 2018; Vunsh et al., 2007; Wang et al., 2021; Yamamoto et al., 2001), but the transformation of L. japonica clone 8627 is the most efficient by far, so we chose this species in these studies (Cantó‐Pastor et al., 2015). This clone was previously misclassified as L. minor (Braglia et al., 2021). Besides the overexpression of foreign genes, the ability to suppress native duckweed genes is also needed. An artificial miRNA approach has been successfully used (Cantó‐Pastor et al., 2015), as well as CRISPR/Cas9‐mediated genome editing (Liu et al., 2019b). Transformation of L. japonica can be achieved within 5–6 weeks with better than 90% efficiency (Cantó‐Pastor et al., 2015).

Triacylglycerol (TAG) has more than twice the energy density of starch. For most biomass crops vegetative tissues comprise a major proportion of the total plant biomass, so the accumulation of TAG in plant vegetative tissues is an appealing strategy to produce high yields of TAG for biofuel uses (Xu et al., 2020; Yee et al., 2021). In duckweeds, almost all the plant biomass is comprised of photosynthetically active tissue in floating leaf‐like fronds, which if genetically modified could potentially produce TAG at high yields per unit land area. A widely used approach to engineer TAG accumulation in vegetative tissues is the so‐called push‐, pull‐ and protect approach (Vanhercke et al., 2014), which refers to the expression of different factors in combination to increase the flux of metabolic precursors to fatty acid biosynthesis (push), increased enzyme capacity to incorporate fatty acids into TAG (pull) and to minimize degradation of TAG from lipid droplets (protect). Accordingly, increased flux into fatty acid synthesis can be achieved by overexpression of WRINKLED1 (WRI1), a transcription factor of the family of APETALA2/ethylene‐responsive element binding proteins, and a well‐stablished master transcriptional activator of fatty acid synthesis (Focks and Benning, 1998; Liu et al., 2019a; Zhai et al., 2017). The heterologous expression of WRI1 homologs from a variety of species has been shown to induce TAG accumulation (Grimberg et al., 2015; Yang et al., 2015). Expression of WRI1 in combination with various other genes was shown to synergistically promote oil synthesis and accumulation (An and Suh, 2015; Beechey‐Gradwell et al., 2019; Lee et al., 2017; Vanhercke et al., 2013, 2019; Zale et al., 2016). The incorporation of fatty acids into TAG can be facilitated by the expression of diacylglycerol acyltransferase (DGAT), which catalyzes the transfer of an acyl group from acyl‐CoA to diacylglycerol (Cases et al., 1998). The expression of DGAT with an engineered plant structural protein (cysteine [Cys]‐oleosin) was shown to result in TAG accumulation in Arabidopsis (Winichayakul et al., 2013). Suppression of lipid degradation can be achieved by overexpression of OLEOSIN (OLE), a structural protein that coats oil bodies in developing seeds and other tissues (Board et al., 2022; Chapman et al., 2012). In combination, overexpression of Arabidopsis WRINKLED1 (AtWRI1), Arabidopsis DGAT1 (AtDGAT1), and sesame OLEOSIN1 (SiOLE) have resulted in levels of TAG up to 15% of DW in tobacco leaves (Vanhercke et al., 2017).

Here we report on two rounds of the design‐build‐test‐learn approach for the engineering of elevated TAG levels in L. japonica. For this purpose, we created a series of constructs to express improved push‐, pull‐ and protect factors and expressed them in different combinations. Concerning the push factor WRINKLED1 (W), we observed morphological changes and reduced growth rate when expressed constitutively. We addressed previously reported protein stability issues and adverse effects of WRI1 expression described herein and elsewhere (Yang et al., 2015) in leaf tissues by creating a fusion of AtWRI1 with Cyan Fluorescent Protein (CFP) and by controlling expression via an estradiol inducible promoter. The other two factors, a highly active form of DGAT from mouse, MmDGAT (D), and a modified OLEOSIN from Sesame indicum (SiOLE(*)) (O) were constitutively expressed. We characterized stable transformed lines containing W, D, or O individually, binary combinations thereof and with all three genes. In combination, TAG levels were increased from 0.08% of DW in the parental wild type (WT) to a mean of 8.7% across several transgenic lines. Our best line accumulated up to 10% (DW) of TAG resulting from synergistic effects of expression of W, D and O.

Results

Optimization of gene expression in Lemna japonica

SiOLE was used to test whether the ZmHSP70 intron could enhance its expression. A modified OLEOSIN from Sesame indicum (Cys‐OLEOSIN) with the ZmHSP70 intron (SiOLE(*)) or lacking the intron (SiOLE(S)) were transiently expressed in Nicotiana benthamiana leaves. The SiOLE(*) construct significantly increased TAG content to 0.53% by dry weight (DW) in N. benthamiana leaves relative to 0.34% by DW for SiOLE(S) (Figure 1a). We also assessed the overexpression of maize ZmWRI1b driven by the constitutive 35S long (Cauliflower Mosaic Virus) promoter. Separation of extracted lipids by thin layer chromatography (TLC) showed two transgenic lines (ZmWRI1b‐1 and ZmWRI1b‐2) accumulated 5‐ to 8‐fold more TAG than WT and controls in which enhanced green fluorescence protein (GFP) was overexpressed alone (Figure 1b). Constitutive overexpression of ZmWRI1b severely affected duckweed morphology and slowed their growth (Figure 1c). Consistent with our findings, constitutive expression of WRI1 in monocots has been demonstrated to result in negative physiological effects (Yang et al., 2015), and reduced biomass accumulation (Parajuli et al., 2020). Using what we learned from these experiments, we initiated a second round of the design‐build‐test‐learn cycle making several modifications to our approach. To alleviate issues related to the constitutive expression of WRI1, we switched to inducible expression by the estrogen receptor‐based XVE system (Schlücking et al., 2013). We previously demonstrated that fusing GFP to the N‐terminus of the Arabidopsis WRI1 (GFP‐WRI1) enhanced its stability when expressed in tobacco leaves (Zhai et al., 2017), to further improve AtWRI1 expression, the Anemonia majano CFP was codon optimized for L. japonica and N‐terminally fused to AtWRI1 (CFP‐AtWRI1). Since AtWRI1 is localized to the nucleus (Zhai et al., 2017), we used confocal microscopy to examine an OW (25‐1) line four days after estradiol induction to determine if the optimized CFP‐AtWRI1 correctly localized to the nucleus (Figure 1d). The CFP fluorescence (purple arrow) from OW (25‐1) fronds colocalizes with the free enhanced YFP (EYFP) fluorescence marker protein diffused into the nucleus, demonstrating that the optimized CFP‐AtWRI1 expressed successfully and targets correctly.

Figure 1.

Figure 1

Open in a new tab

Optimization of gene expression in Lemna japonica. (a) Triacylglycerols (TAG) contents in Nicotiana benthamiana leaves transiently expressing the modified SiOLE with ZmHSP70 intron (SiOLE(*)). (b) Thin layer chromatography (TLC) of lipid from two different P35SL:ZmWRI1b transgenic lines and its parental WT and control line (GFP). (c) Morphology of WT and representative ZmWRI1b overexpression plants. Scale bar = 5 mm. (d) Expression and subcellular localization of AtWRI1 with the optimized Anemonia majano Cyan Fluorescent Protein (CFP) fused to the N terminus were assessed by the fluorescence signal of CFP. The CFP fluorescence is shown in blue color and the enhanced YFP (EYFP) is shown in yellow color. The overlapping regions of CFP and EYFP are indicated by purple arrows. Scale bar = 40 μm. Abbreviations: EV, empty vector (L2‐0); SiOLE(*), sesame OLEOSIN with intron (L2‐6); WT, Wild type; ZmHSP70, maize heat shock protein gene. Data are means of three biological replicates and error bars indicate standard deviations. Significant differences between groups (P < 0.05) as identified by one‐way ANOVA and Tukey's multiple comparisons test are shown by different letters.

Generation of single, double and triple combination transformants of SiOLE(*), CFP‐AtWRI1 and MmDGAT

Except for the above‐mentioned CFP‐AtWRI1 under the control of the XVE system to avoid the adverse effects of AtWRI1, Mouse (Mus musculus) DGAT2 (MmDGAT) was used because it is the major metabolic route to TAG accumulation in mouse (Chitraju et al., 2019; Smith et al., 2000; Stone et al., 2004) and we previously reported transient overexpression of the MmDGAT2 in tobacco leaves resulted in significant TAG accumulation (Cai et al., 2019). SiOLE(*) was also created to mitigate the breakdown of TAG from lipid droplets (Hamada et al., 2020; Hsiao and Tzen, 2011; Lin et al., 2002; Tai et al., 2001). An overview of all constructs used in this study is shown in Figure 2. We used Golden Gate assembly (Engler et al., 2014; Weber et al., 2011) to make level one (L1) and level two (L2) constructs (Figure 2) which were used for transformation. An “empty vector” containing BAR and EYFP L2‐0 (EV) was used as a negative control. Constructs or duckweed transgenic lines for single genes for SiOLE(*), CFP‐AtWRI1 and MmDGAT were designated as L2‐6 (O), L2‐22 (W) and L2‐24 (D), respectively. Similarly, binary combinations were named L2‐25 (OW), L2‐27 (OD), and L2‐31 (WD). Finally, the three‐gene combination was named L2‐33 (OWD). All transgenic lines were labeled by their individual line designation as shown in Figure 2. Constructs harboring combinations of SiOLE(*), CFP‐AtWRI1 and MmDGAT were transformed into the callus of WT L. japonica by Agrobacterium‐mediated transformation. After selection with phosphinothricin (PPT), we further screened PPT‐resistant lines for EYFP fluorescence using a microplate reader to eliminate false positive lines (Cantó‐Pastor et al., 2015). Similar areas of fronds were used for this analysis as shown in Figure S1. The EYFP fluorescence intensities of PPT‐resistant lines are shown by a color gradient blue‐white‐red (133–530–20 000 A.U.; Figure S2a). We obtained 6, 5, 7,4, 4, 3, 2 and 5 positive transgenic lines from EV, O, W, D, OW, OD, WD and OWD constructs, respectively (Figure S2b).

Figure 2.

Figure 2

Open in a new tab

Schematic of the gene constructs with different combinations of SiOLE(*), CFP‐AtWRI1 and MmDGAT2. AtWRI1 with an codon optimized N‐terminal cyan fluorescent protein (CFP) fusion under the control of the XVE estradiol‐inducible promoter, MmDGAT2 under the strong constitutive 35S promoter and a modified sesame OLEOSIN (SiOLE(*)) under the control of the strong ZmUbi promoter. The BAR and EYFP genes which were used for transformants screening were driven by Agrobacterium tumefaciens Nos promoter and a long version of 35S promoter (35SL), respectively. The level two L2‐0 construct containing only BAR and EYFP was used as empty vector (EV) control. Scale bar = 1 kb. Abbreviations: At, A. thaliana; BAR, phosphinothricin acetyl transferase; D, MmDGAT2; EYFP, Enhanced yellow fluorescent protein; LB, left border; Mm, M. musculus; O, SiOLE(*); OD, SiOLE(*)+MmDGAT2; OW, SiOLE(*)+CFP‐AtWRI1; OWD, SiOLE(*)+CFP‐AtWRI1+MmDGAT2; P, promoter; Pv, Panicum virgatum; RB, right border; Si, S. indicum; T, Terminator; W, CFP‐AtWRI1; WD, CFP‐AtWRI1+MmDGAT2; Zm, Z. mays.

Induction of expression of CFP‐AtWRI1 by estradiol

To induce TAG accumulation in L. japonica fronds, only transgenic lines containing CFP‐AtWRI1 and the WT negative control (WT+E) were subject to estradiol induction before quantitative lipid analysis. In this pilot experiment, a 5‐day time course of induction with increasing concentrations of estradiol was performed on the OWD line (33‐6). TAG content increased from 2.8% of DW of uninduced samples to a maximum of 8.5% of DW upon estradiol induction (Figure S3). Different estradiol concentrations i.e., 50, 100 and 200 μm for 3, 4 or 5 days resulted in similar levels of TAG accumulation, and 100 μm estradiol treatment for 4 days resulted in the highest TAG accumulation (Figure S3). For subsequent experiments, 4‐day inductions with 100 μm estradiol were used for two independent representative CFP‐AtWRI1 containing transgenic lines per construct as shown in Figure 3.

Figure 3.

Figure 3

Open in a new tab

Lipids content and composition. (a) Total fatty acid content by percentage of Dry Weight (DW). (b) Triacylglycerols (TAG) content by percentage of Dry Weight (DW). (c) lipid profiles of total fatty acid and TAG in different transgenic lines. Two representative lines are show in (a) and (b), single representative lines are show in (c). Fronds were cultivated under half‐strength SH medium and continuous light approximate 100 μmol m−2 s−1 cool white fluorescent light illumination at pH 5.6 and at 23 °C with ambient CO2 level. Data are means of three biological replicates and error bars indicate standard deviations. Significant differences between groups (P < 0.05) as identified by one‐way ANOVA and Tukey's multiple comparisons test are shown by different letters. Abbreviations: E, Estradiol; WT, Lj8627 Wild type.

Overexpression of SiOLE(*), CFP‐AtWRI1 and MmDGAT significantly increases total fatty acid contents and TAG accumulation in L. japonica fronds

The effects of single and combinations of SiOLE(*), CFP‐AtWRI1 (4 days induction) and MmDGAT on the level of total fatty acids (TFA) are shown in Figure 3a. Untransformed and EV‐transformed L. japonica (0‐5, 0‐6) accumulated ~5.4% and ~4.5% of DW of TFA, respectively, and overexpression of OD (27‐2 and 27‐4), WD (31‐1 and 31‐3) and OWD (33‐5 and 33‐6) substantially increased TFA content to 10%, 8% and 15%, respectively (Figure 3a). In control experiments, incubation of WT in media containing 100 μm of estradiol (WT+E) did not result in any changes in TFA content.

The effects of single gene expression of SiOLE(*), CFP‐AtWRI1 or MmDGAT on TAG accumulation is shown in Figure 3b. Except for O (6‐1 and 6‐5), the CFP‐AtWRI1 or MmDGAT‐ expressing lines showed a significant approximately 7‐fold increase from 0.08% TAG in the parental WT or EV lines to approximately 0.6% for D (24‐2) and W (22‐4; Figure 3b) lines. Similar increases were observed in the previous design cycle for the 35SL::ZmWRI1b lines described above. The overexpression of two‐gene combinations of SiOLE(*), CFP‐AtWRI1 and MmDGAT dramatically increased TAG accumulation by 45‐fold to approximately 3.6% for OD (27‐2, 27‐4), by 25‐fold to approximately 2% for WD (31‐1, 31‐3), and by 7‐fold to approximately 0.6% of TAG for OW (25‐1, 25‐3) relative to parental WT or EV controls (Figure 3b).

In addition to comparing TFA or TAG levels of transgenic lines to WT or EV controls, the same data can also be analyzed to assess synergistic effects. We define synergy as when the effect of combined gene expression exceeds the sum of their effects when expressed singly. With respect to the TFA contents (Figure 3a) there were no such effects visible, but synergistic effects were seen for TAG contents (Figure 3b and Table S1). A detailed analysis of six synergy cases with statistical evaluation is shown in Table 1. For example, the sum of the averages in TAG levels for the O‐, W‐ and D lines in Figure 3b, respectively, minus the contributions of the EV control shown in Figure 3b, amounts to 0.93 (%DW). The average TAG level for the two OWD lines in Figure 3b without EV effects is 7.76% (DW), which is 8.4 times higher than the combined levels of the single transformants (t‐test, P = 4.0 × 10−15, Table 1). The synergy effects are similarly strong and highly significant when comparing the OW and D TAG levels to OWD (8.7 times higher TAG level in OWD), and for comparing O and D to OD (8.8 times higher, Table 1). Other cases that combine W with O, D or OD are less synergistic (with increases of 2.4‐fold or less, Table 1), while combining W and O has merely additive effects (0.94, Table 1). These observations together suggest that the strongest synergy effect arises from combinations that bring O and D together. A similar observation is that among the four assessed combinatorial expressions cases (OW, OD, WD, OWD) the combinations OWD and OD have the highest and second highest measured levels, respectively, both in TFA (Figure 3a) and in TAG (Figure 3b).

Table 1.

Tests for synergy effects for genes expressed separately and in different combinations

Separate genotypes Genotype/condition
O+W O+D W+D OW+D OD+W WD+O O+W+D
Sum of average TAG levels (S1)* 0.54 0.40 0.83 0.90 4.08 2.06 0.93
Combined expression OW OD WD OWD OWD OWD OWD
Average TAG level (S2)* 0.51 3.55 1.96 7.76 7.76 7.76 7.76
Ratio S2/S1 0.94 8.79 2.36 8.65 1.90 3.77 8.35
T −0.93 −18.99 −4.16 −16.53 −8.62 −11.72 −16.79
df 20 20 20 20 20 20 25
P‐value 3.6E‐01 2.9E‐14 4.8E‐04 4.0E‐13 3.6E‐08 2.0E‐10 4.0E‐15
Open in a new tab

Mean values (TAG(%DW)) derived from Table S1. For computation of test statistic T and P‐values see Methods.

*

Minus the contribution of the empty vector control in each case (Table S1).

To visualize lipid droplets, fronds from each transgenic line were stained with boron‐dipyrromethene (BODIPY) and imaged using confocal microscopy. Transgenic lines for the W and D single gene expression showed increased amounts of BODIPY‐stained lipid droplets compared to O and the negative controls (Figure 4). The transgenic lines for two‐gene combinations (OD, OW, WD) showed enhanced accumulation of lipid droplets relative to the genes expressed singly and controls and the OWD lines showed the strongest signal. These findings are consistent with the TAG data shown in Figure 3b. The lipid droplets were relatively small for lines expressing single genes compared to those expressing two or three of the transgenes. The OD and OWD lines have large oil droplets, but the OWD line has more of them.

Figure 4.

Figure 4

Open in a new tab

Confocal images of all transgenic lines and WT. Lipid droplets in all lines were shown at left‐side of each group with yellow color representing BODIPY signal. The right‐side images of each group represent the merged signal of chlorophyll autofluorescence (colored in red) and bright field. Two independent transgenic lines were shown for each gene combination. EV (0‐5, 0‐6), single gene construct O (6‐1, 6‐5), W (22‐4, 22‐9) and D (24‐2, 24‐4), two gene combination constructs OW (25‐1, 25‐3), OD (27‐2, 27‐4), and WD (31‐1, 31‐3), and three gene combination construct OWD (33‐5, 33‐6). Scale bar = 50 μm. Abbreviations: D, MmDGAT; E, Estradiol; EV, empty vector; O, modified SiOLE; W, CFP‐AtWRI1; WT, Lj8627 Wild type.

Transgenic lines show distinct changes in fatty acid profiles

Since heterologous expression of lipogenic genes can alter fatty acid profiles, we determined the profiles of each representative line (Figure 3c). WT and EV lines mainly accumulate C18:3(9,12,15) which accounts for around 50% of TFA, whereas, OD, WD and OWD lines accumulate C18:1(9) and C18:2(9,12) with percentages ranging from 7% (WD (31‐1)) to 39% (OWD (33‐6)) at the expense of C18:3(9,12,15). Analysis of TAG fatty acid profiles showed that C16:0 is present at about 40% in the WT, while it amounts to only 8%, 17% and 6%, respectively in the OD, WD and OWD lines. OD, WD and OWD accumulate 30%, 23% and 31% of C18:1(9), and 45%, 38% and 43% of C18:2(9,12), respectively.

Several gene combinations containing WRI1 had an inhibitory effect on growth rate

In preparation for experimental treatments, lines were cultivated in half‐strength Schenk & Hildebrandt (SH) medium with a 16‐h photoperiod followed by a 3‐day exposure to continuous light. The growth curves were generated by daily measurement of frond areas for 10 days of culture under continuous light (Figure 5a) and the corresponding growth rates are presented in Figure 5b. While the OW (25‐1 and 25‐3) and WD (31‐3) lines were found to grow significantly slower than WT and EV lines, the average growth over 10 days of all other lines showed no significant differences from WT or EV (0‐5). Thus, it might suggest that even background expression of WRI1 can cause growth inhibition, but this is only the case for OW and WD (31‐3), not for W and OWD. Also, since for WD the growth inhibition effect is not consistent (31‐1 vs. 31‐3), we currently cannot make clear conclusions as to the possible effects of WRI1 expression on growth rates. Besides growth rates, morphological effects were also observed. After 10 days the transgenic lines appeared similar to that of WT and EV. However, the fronds of one WD line (31‐3) had a longer frond axis, which gave the appearance of a larger separation between mother and daughter fronds, and the fronds were more triangular‐shaped relative to the more rounded WT fronds (Figure 5c). We also observed that the daughter fronds of OWD (33‐6) have a yellowish tinge compared to the greener colored WT fronds (Figure 5c).

Figure 5.

Figure 5

Open in a new tab

Growth kinetics and morphology. (a) Photoautotrophic growth. (b) Average photoautotrophic growth rate. (c) Morphology comparison. Scale bar = 2 mm. The black line represents WT growth kinetic, and the two green lines show EV (0‐5, 0‐6) control lines growth, and other curves are for other transgenic lines as indicated in the graph. Fronds were cultivated in half‐strength SH medium (pH 5.6) under continuous cool white fluorescent light illumination (approximate 100 μmol m−2 s−1) at 23 °C with ambient CO2 level. Data are means of three biological replicates and error bars indicate standard deviations. Significant differences between groups (P < 0.05) as identified by one‐way ANOVA and Tukey's multiple comparisons test are shown by different letters.

SiOLE(*), CFP‐AtWRI1 and MmDGAT transcript levels

RT‐qPCR was performed to determine the normalized relative transcript levels of the transgenes (Figure 6a–c). The expression level of MmDGAT in lines singly expressing MmDGAT was similar to that when MmDGAT was co‐expressed with SiOLE(*) and/or CFP‐AtWRI1. However, the accumulation of SiOLE(*) transcript was lower when expressed alone than when expressed with MmDGAT or CFP‐AtWRI1. Conversely, the expression of CFP‐AtWRI1 induced by estradiol was higher when expressed alone than when expressed with MmDGAT or SiOLE(*). The protein abundance of CFP‐AtWRI1 in different transgenic lines as determined by Western blot was consistent with its transcriptional expression pattern (Figure 6d). While it is unclear why expression of DGAT or OLEOSIN might be somehow attenuating WRI1 expression, this observation is consistent with the above synergy findings, confirming that combining W with O, D or OD has less synergistic effect than combinations that bring O and D together (Table 1).

Figure 6.

Figure 6

Open in a new tab

Quantitative RT‐PCR analysis of transcriptional expression of SiOLE(*) (a), MmDGAT (b) and CFP‐AtWRI1 (c) in different transgenic lines, as well as the Western blot result of CFP‐AtWRI1 (d). Fronds were cultivated in half‐strength SH medium (pH 5.6) under continuous cool white fluorescent light illumination (approximate 100 μmol m−2 s−1) at 23 °C with ambient CO2 level. Data are means of three biological replicates and error bars indicate standard deviations. Significant differences between groups (P < 0.05) as identified by one‐way ANOVA and Tukey's multiple comparisons test are shown by different letters.

Discussion

This study demonstrates that duckweed can accumulate significant levels of TFA and TAG as evidenced by the L. japonica 8627 transgenic line OWD (33‐6) which accumulated up to approximately 16% of TFA and 8.7% of TAG on a DW basis. This represents an increase of more than two orders of magnitude of TAG upon the co‐expression of the push, pull and protect factors CFP‐AtWRI1, MmDGAT and SiOLE(*) in the fronds relative to the untransformed parental WT L. japonica. The stability of line OWD (33‐6) is evidenced by the accumulation of equivalent levels of TAG in Figure S3 and Figure 3, which were separated by 7 months and more than 80 generations. In contrast to the three‐gene overexpression, the expression of single, W, D, and O genes did not significantly elevate TFA accumulation compared to the untransformed WT or the EV control. However, the co‐expression of W and D, and of D and O, respectively, resulted in significant increases in both TFA and TAG, relative to the effect of their expression individually. In contrast, the combination of W and O did not result in a significant increase in TFA relative to their expression individually. Combining W and O did also not result in a significant increase in TAG relative to W alone. In light of the push, pull and protect paradigm this indicates that consecutive factors (push and pull, W and D; pull and protect, D and O) act strongly together while push and protect factors (W, O) do not. Evaluating our data for synergy, i.e., quantifying whether the effect of coexpression is more than the sum of effects of individual expression, we find strong synergy effects in TAG accumulation when combining O, W and D, when combining OW and D as well as combining O and D. We further find that combining W and O has no synergy effect. Thus, at least the effectiveness of combining O, W and D and the small effect of combining W and O is consistently found. The latter finding implies that the increased FA synthesis resulting from the expression of the push factor W exceeds the capacity of WT DGAT to transfer FA to TAG. The push, pull and protect paradigm assumes that increased TAG accumulation is accomplished through an increased flux through de‐novo FA synthesis. However, we cannot exclude that TAG is made by the rebalancing of FA chains between membranes and TAG. In cases of expression of O, W and D as well as for coexpression of O and W the TAG levels are well below 1% of DW while in the WT and EV control cases the total FA content ranges between about 4 and 5% of DW. Thus, within uncertainties of experimental error, an undetectable but substantial redistribution of FA from existing membranes to TAG is also possible.

We note that sources for the push‐, pull‐ and protect factors used in these studies and the promoters employed to drive the expression of these genes were chosen somewhat arbitrarily, and it is likely that screening of additional genes encoding each factor for ones that outperform those used in this study in subsequent design‐build‐test‐learn cycles will yield further increases in TAG accumulation. It is also possible that the use of other developmental or tissue‐specific promoters for D and O may result in increases in TAG accumulation. Likewise, for the conditional expression of W to avoid the toxic effects of WRI1 expression by a strong constitutive promoter in monocots (Yang et al., 2015), we chose to drive WRI1 with the estradiol inducible XVE promoter system (Schlücking et al., 2013; Zuo et al., 2000). The use of this system enabled us to successfully obtain and regenerate AtWRI1 transgenic duckweed from callus culture that lacks the morphological changes and reduced growth rates associated with constitutive expression of WRI1. In XVE‐controlled W‐expressing lines we observed high CFP‐AtWRI1 transcript levels after 3 days of estradiol induction, and some chlorosis and cell death after 5 days. It is interesting to note that for unknown reasons levels of W transcripts in lines expressing only W, were 5‐ to 10‐fold higher than that in lines co‐expressing of O, D, or O and D (Figure 6c,d). However, we note that TFA and TAG levels were much higher in lines in which W was coexpressed with D, or O and D, relative to lines in which W was expressed alone, demonstrating that high levels of W were not necessary for maximal oil accumulation in L. japonica. While useful for these pilot studies, the use of an estradiol‐inducible promoter may be undesirable for large‐scale production because of potential toxicity to humans.

TAG levels of 30% of DW have been reported for tobacco (Nicotiana tabacum) leaves (Vanhercke et al., 2017), raising the possibility that similar levels could be achieved in duckweed. Further engineering to achieve higher levels could involve additional genetic manipulation. That WT L. japonica typically accumulates only 0.08% TAG (DW) is consistent with it having a high capacity for lipid degradation and that downregulation of TAG lipases such as SDP1 (Kelly et al., 2013) could result in a doubling of TAG accumulation levels as previously reported for tobacco (Vanhercke et al., 2017). Successful achievement of such levels would result in duckweed biomass having approximately 50% higher TAG than that of soybeans, currently the world's largest source of vegetable oil. In terms of yield per acre per year, soybeans produce approximately 55 gal of oil. Duckweed yields in the southern USA can reach 20 tonnes per acre per year (Leng et al., 1995), which at 10% of its DW as oil, would potentially yield 720 gal of oil, i.e., approximately 13‐fold the yield of soybean. This is more than four times that was reported by Vanhercke et al. (2019) for the best transgenic sorghum lines and similar to that of oil palm. If production lines of duckweed can be further optimized to produce 20% TAG (DW), the theoretical oil yield could reach 1440 gal per acre per year, i.e., it could exceed that of soybean by approximately 25‐fold. The ability of duckweeds to obtain nutrients from agricultural waste streams that normally pollute the environment, adds to their attractiveness as sources of biofuel feedstocks (Xu et al., 2012a; Ziegler et al., 2016).

Besides the accumulation of TFA and TAG, the rate at which the transgenic lines grow is of interest. With a constitutive expression of O and D and background expression of W, a significant reduction in growth was found (Figure 5b). Several recent studies aimed at engineering TAG accumulation in vegetative tissues of various plants reported reduced growth or yield penalty effects (Chu et al., 2020; Mitchell et al., 2020; Parajuli et al., 2020; Vanhercke et al., 2017). Among various possible explanations are: (i) Limitations in the supply of energy cofactors and carbon precursors under conditions of resource diversion to TAG synthesis at high rates. (ii) Energy drain caused by high rates of TAG synthesis and degradation (futile cycling). (iii) Pathway imbalances under high rates of fatty acid and TAG synthesis – for example build‐up of free fatty acids to toxic levels (Yang et al., 2015). (iv) Interference of TAG accumulation with the maintenance of the composition and quantity of cellular membranes. (v) Besides the known gene targets of WRI1 associated with the conversion of carbohydrates into fatty acids, there might be interference of this transcription factor with other processes that could have negative effects. These questions are currently being investigated and it is noteworthy that the lower morphological complexity of duckweed compared to other higher plants, the precision with which physiological conditions can be controlled and growth rates determined, makes Lemna japonica well suited for future studies of this phenomenon.

The FA composition of WT duckweed consists mainly of C18:3, C16:0 and C18:2 (Chakrabarti et al., 2018; Sharma et al., 2019; Tang et al., 2015; Yan et al., 2013). In contrast, oil‐accumulating transgenic lines show an increase of unsaturated FA, including C16:1(9), C18:1(9), C18:2(9,12) and C18:3(9,12,15) in their TAG, and a decrease in C18:3(9,12,15) in their TFA, consistent with a large proportion of their C18:1 or C18:2 being packaged into TAG reducing the availability of C18:2 for further desaturated to C18:3. This trend was strongest in higher TAG accumulation lines, such as in OD, WD and OWD, suggesting TAG biosynthesis is outcompeting polar lipid synthesis for mono‐ and di‐unsaturated FA due in part to the substrate preference of MmDGAT. The resulting reduction in C18:3(9,12,15) in cellular membranes likely contributes to the chlorotic phenotype in the days following the induction of lipid synthesis.

The final committed step in TAG synthesis is the DGAT‐mediated conversion of diacylglycerol (DAG) to TAG. DGATs from different species have distinct activities when heterologously expressed. The MmDGAT from an animal source has a relative high activity in plants compared to other those plant DGATs (Cai et al., 2019). The successful expression of animal enzymes in plants has been reported a number of times, for example, the expression of fish docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA) biosynthetic enzymes show similar levels of activity in plants (Venegas‐Caleron et al., 2010). To date there are few examples of animal‐sourced enzymes used to engineer plant metabolism and whether they can be useful in boosting plant lipid synthesis (Cai et al., 2017). In screening plant and mouse DGATs for their efficacy in plant lipid accumulation, we found the MmDGAT to outperform the tested plant‐sourced enzymes. While this may appear counter intuitive, cross‐kingdom heterologous expression of enzymes can be a useful way of avoiding regulatory feedback mechanisms that are specific to a particular kingdom as previously demonstrated (Stark et al., 1992). Alternatively, some enzymes may have evolved higher rates of catalysis to meet the metabolic demands of their natural hosts. Further study will be needed to determine why MmDGAT is so efficient when expressed in plants.

While work from our group to increase the transformation efficiency of L. japonica was successful (Cantó‐Pastor et al., 2015), the problem of long‐term storage of both WT and transgenic lines remains to be solved. This is because while duckweeds are flowering plants, robust methodologies to induce flowering and harvest seeds have yet to be reported. Thus, subculturing and maintenance at reduced temperatures is currently the best option. Biocontainment of transgenic duckweed is an important concern for the industrial production of high‐oil duckweed. It is envisaged that a system employing a “kill switch” analogous to those previously described (Chan et al., 2016) could be developed to mitigate this issue.

Methods

Materials and callus induction

A duckweed strain, Lemna japonica 8627 used in this study was originally obtained from the Rutgers Duckweed Stock cooperative of Rutgers University. The duckweed was cultivated in SH‐modified medium (Schenk and Hildebrandt, 1972) with 1% (w/v) sucrose (SHS), pH5.6 (plantMedia, Dublin, OH, USA) for 2–3 weeks under 100 μmol m−2 s−1 light illumination in a 16‐h photoperiod and at 23 °C. Calli were induced by incubating duckweed fronds on induction medium (IM; Cantó‐Pastor et al., 2015). Briefly, 4.4 g/L Murashige and Skoog (MS) basal salts (Murashige and Skoog, 1962) plus 3% (w/v) sucrose, with 5 μm 2,4‐dichlorophenoxyacetic acid (2,4‐D) and 0.5 μm thidiazuron (TDZ) was used as IM, pH5.6. After one month, the fastest growing calli were propagated in nodule production medium (NPM), containing 4.4 g/L MS basal salts, 3% (w/v) sucrose, 10 mm 2,4‐d and 20 mm 6‐benzylaminopurine (BAP).

Codon optimization of cyan fluorescence protein (CFP)

A CFP gene from Anemonia majano with relatively narrow excitation and emission spectra was codon optimized based on a 60% GC content for expression in L. japonica (Figure S4). To further improve gene expression, a 10 bp Kozak sequence (CTCGAGGCCG; Kozak, 1987, 1989) from the Spirodela polyrhiza WRI1 gene (Asrani et al., 2018) was fused to the start codon of CFP.

Vector construction

The ZmWRI1b coding sequence along with UTR regions was amplified from ZmWRI1b plasmid cloned into the pB7WG2D Gateway destination vector behind the 35SL promoter.

To enhance the expression level of MmDGAT and SiOLE(*), an intron from the maize 70 KDa heat shock protein 1_8 (ZmHSP70) followed by a Kozak sequence were introduced upstream of each of the ORFs. To prevent gene silencing in stably transformed lines, different promoters and terminators were used for each of the transgenes. Phosphinothricin acetyl transferase (BAR) was used as the selection maker to identify transgenics.

GoldenGate modular cloning components of the MoClo Toolkit (Weber et al., 2011; Werner et al., 2012) and the MoClo Plant Parts Kit (Engler et al., 2014) from Addgene (http://www.addgene.org/www.addgene.org) were used for vector constructions. The AGL1 Agrobacterium strain and a modified level 2 accept vector (pICSL4723‐just‐L2) were used for these studies. To clone genes into Golden Gate vectors, we changed those nucleobases in the genes with BsaI, BpiI or BsmBI restriction sites. To combine these different elements, a PCR‐based method was used to make two new Level zero acceptor vectors which with TACT‐GATG (for 5′ untranslated sequence (5U) part) and GATG‐GCTT (for coding sequence 1 (CDS1) part) sticky ends after digestion with BsaI. With the concern of WRI1 adverse physiological effects on duckweed, the inducible XVE system (Schlücking et al., 2013; Zuo et al., 2000) was used, which is regulated by estradiol. An optimized CFP with relative narrow excitation and emission spectrum was synthesized by GENEWIZ (https://www.genewiz.com/en) and fused to the N‐terminal of AtWRI1 to get CFP‐AtWRI1, which was under the control of bacterial LexA operator fused upstream (−46 to +12) of the 35S minimal promoter and A. tumefaciens gene7 terminator. The XVE fusion gene was driven by Arabidopsis ubiquitin 10 promoter (PAtUbi10) and with A. tumefaciens ocs terminator (TOCS; Figure 2). MmDGAT expression was driven by the strong constitutive 2x35S promoter (P2x35S) and terminated by the PolyA signal of 35S (Figure 2). The modified sesame Cys‐OLEOSIN (SiOLE(*)) was driven by the strong maize polyubiquitin 3 promoter (PZmUbi) and terminated by the switchgrass ubiquitin 1 terminator (TPvUbi1; Figure 2). Different combination of SiOLE(*), CFP‐AtWRI1 and MmDGAT (Figure 2) were constructed according to the method described (Engler et al., 2014; Weber et al., 2011; Werner et al., 2012). Names and sequences of primers, template plasmid, PCR product length, annealing temperature, DNA ligase, restriction digestion enzymes and antibiotic selection markers are shown in Table S2.

Agrobacterium preparation and transient expression in tobacco (Nicotiana benthamiana) leaves

Constructed expression vectors were transformed into Agrobacterium tumefaciens AGL1 strain using a BioRad micropulser electroporator. Transformed AGL1 containing, EV, SiOLE(*) or SiOLE(S) (Figure 1a) were cultivated in Luria‐Bertani (LB) medium with appropriate antibiotics to OD600 = 1.0. After centrifugation, cell pellets were resuspended in infiltration buffer containing 10 mm MES, pH5.6, 10 mm MgCl2, and 200 μm acetosyringone (AS) and adjusted to OD600 = 0.8. The cells were incubated for 2 h at room temperature with gentle shaking (Norkunas et al., 2018), and an equal volume of Agrobacterium cells containing P19 gene (Voinnet et al., 2003) was included in each infiltration mixture. The mixture was infiltrated into the leaves of 4‐week‐old N. benthamiana plants. Following infiltration, the plants were left at 28 °C for an hour before returning to the growth chamber (Wood et al., 2009).

Duckweed callus transformation

We performed the transformation of duckweed as previously described (Cantó‐Pastor et al., 2015), for the ZmWRI1b lines, and with minor modifications for the OWD lines. Agrobacterium tumufaciens GV3101 (ZmWRI1b) or AGL1 (OWD) cells containing different constructs were cultivated with appropriate antibiotics and 100 μm AS at 28 °C to OD600 around 0.8. Then cells were spun down and resuspended in the infection medium (10 mm MgCl2, 1% sucrose and 200 μm AS) and cultured for 2 h. Finally, cells were diluted with infection medium to OD600 = 0.2. Calli from Lemna japonica were cut into small pieces and immersed in Agrobacterium for 5 min, then vacuum infiltrated for 8 min at 34 kPa, followed by 5 min at normal pressure and vacuum for another 8 min. Then, calli were cultivated on NPM medium with 100 μm of AS for 3 days under dark at 23 °C.

Selection and regeneration of transgenic plants

After three days of co‐cultivation, the infected calli were transferred onto regeneration and selection medium (Liu et al., 2019b). Briefly, Gamborg's B5 basal medium supplemented with 1% sucrose, 4.65 μm kinetin, 2.57 μm indole‐3‐acetic acid (IAA), 10 mg/L PPT and 400 mg/L timentin were added. Calli were cultured under a 16‐h photoperiod of 50 μmol m−2 s−1 of light. Regenerated fronds were proliferated on SHS liquid medium.

Fluorescence screen of transformed plants

All transformed plants were verified by detecting YFP fluorescence using a TECAN SPARK 20M multimode microplate reader (TECAN US, Durham, NC: Tecan Genios, Grödig, Austria). The excitation wavelength was set at 485 nm with a bandwidth of 20 nm, and the emission wavelength was set at 535 nm with 20 nm of bandwidth.

RNA isolation and cDNA preparation

Total RNA from L. japonica was extracted using Direct‐zol RNA miniprep kit (Zymo Research Corp, Orange, CA, USA) following the manufacturer's protocol. The first strand cDNA of each line was synthesized from 2 μg of total RNA using the SuperScript IV VILO Master Mix with ezDNase kit (ThermoFisher Scientific, Waltham, MA, USA) following the manufacturer's protocol.

Reverse transcription‐ quantitative PCR (RT‐qPCR)

RT‐qPCR was carried out using SsoAdvanced Universal SYBR Green Supermix (Bio‐Rad, Hercules, CA, USA) following the instruction from the manufacturer to analyze transgene expression. The L. japonica α‐tubulin coding gene was used as the reference gene (Cantó‐Pastor et al., 2015). Primers for qPCR are described in Table S2.

Lipids extraction and analysis by Thin‐Layer Chromatography (TLC) and Gas Chromatography‐Mass Spectrometry (GC‐MS)

For oil extraction, duckweed lines were treated with 100 μm estradiol in half‐strength SH medium and continuous light for 4 days, and 100 mg fresh fronds of each line were lyophilized using the dry FreeZone system (Labconco, Kansas City, MO, USA) for 2 days. A total of 800 μL of methanol:chloroform:formic acid (2:1:0.1, v/v/v) and 20 μg of C17:0 TAG standard were added into each sample followed by shaking for 2 h. Finally, 400 μL of 1 m KCl with 0.2 m phosphoric acid (H3PO4) was added. Then centrifuged at 17 000  g for 5 min, and 30 μL of chloroform phase were used for total fatty acids analysis. For TAG analysis, 75 μL of lipid extract was separated by thin layer chromatography (TLC), and subsequently, the TAG fraction was scraped off and trans‐methylated with 1 mL 12% (w/w) boron trichloride (BCl3) in methanol for 90 min at 85 °C (Zhai et al., 2018), except for the ZmWRI1b experiment, 100 μL of total lipid were loaded and development with hexane:diethyl ether:acetic acid (80:20:1, v/v/v) on TLC plate, followed air dried and exposed in iodine vapor. Fatty acid methyl esters (FAMEs) were analyzed on GC‐MS with an Agilent J&W DB 23 capillary column (30 m × 0.25 mm × 0.25 μm) as described previously (Yu et al., 2014).

Confocal microscopy

Leica TCS SP5 confocal scanning laser microscope was used to detect the fluorescence of the CFP and the EYFP to verify the subcellular localization of CFP‐AtWRI1 in duckweed cells. Fluorescent proteins were excited by a laser at 458 nm for CFP and 514 nm for EYFP, and the emission signal was collected between 459–498 nm for CFP and between 563–605 nm for EYFP. To visualize lipid droplets, fronds from different lines were cultivated under continuous light for 4 days (for CFP‐AtWRI1 containing lines 100 μm of estradiol was added), stained with 10 μg/mL BODIPY 493/503 (Invitrogen, Eugene, OR, USA) and imaged using a Leica TCS SP5 laser scanning confocal microscope with excitation at 488 nm and emission at 492–508 nm for BODIPY and 700–784 nm for chlorophyll autofluorescence.

Western blot analysis of CFP‐AtWRI1 expression

For immunoblot analysis, total protein was extracted from different lines as described previously (Zhai et al., 2017). Generally, proteins were extracted from 2.5 mg fresh weight of each sample and separated by SDS‐PAGE using Tris‐MOPS‐SDS running buffer. Then proteins were transferred to BioTrace NT nitrocellulose membrane and immunoblotted with anti‐WRI1 antibodies (1/1000) from rabbit (Zhai et al., 2017).

Growth rate analysis of Lemna japonica cultures

The growth rate per day (r) was calculated based on the equation from (Naumann et al., 2007) for exponential growth:

r=lnAt−lnA0t−t0,

where A t and A 0 are the frond area at time t and start point (t 0), respectively. We used the time points 0 and 10 days for the calculation.

Statistical analysis of synergistic effects

For statistical analysis of synergistic effects for the expression of W, D and O, we applied a modified t‐test as described by (Demidenko and Miller, 2019). For example, for a WD double gene expression event, the null hypothesis is defined as the mean WD TAG content minus the sum of the mean TAG contents of W and D lines, minus the mean TAG content of the empty vector control.

Accession numbers

The GenBank accession number for genes reported in this paper are as follows: AtWRI1 (AT3G54320), Cys‐OLEOSIN (AAD42942), ZmHSP70 (Zm00001d012420), ZmWRI1b (GRMZM2G174834), MmDGAT (AF384160), CFP (AF168421)

Conflicts of interest

The authors declare no conflicts of interest.

Author contributions

J.Sh., Y.L., X.Y., A. M.‐M., E.E., R.M., and J.Sc. designed experiments. Y.L. performed experiments, analyzed data and made figures and table with support from X.Y., W.D., X.Y., A. M.‐M., and E.E. performed some of the callus induction, transformation and regeneration. S.A., Yi.C., Yu.C., G.L., W.D., A.M.‐M and FA contributed to construction of recombinant DNA vectors. H.S. and J.Sc. optimized oil accumulation conditions and analyzed growth rates. S.A., Yi.C., Yu.C., and X.Y. carried out transient overexpression in tobacco. J.C. performed fluorescence screening of transformed plants. Yi.C. helped with confocal microscopy. Y.L. and J.Sh., wrote the article with input from X.Y. S.A. Yi.C., Yu.C., J.Sc., E.E., and R.M. All authors approved the article.

Supporting information

Figure S1 Representative images of duckweed.

Figure S2 Screening of L. japonica transformants using plate reader.

Figure S3 TAG accumulation in duckweed treated with different concentrations of estradiol for different time periods.

Figure S4 Sequence and features of optimized CFP.

Table S1 TAG content as percent (DW) in WT and different transgenic lines of Lemna japonica.

Click here for additional data file. (1.9MB, pdf)

Table S2 Golden gate clone and qPCR primers.

Click here for additional data file. (22.2KB, xlsx)

Acknowledgements

This work was primarily supported by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research program under Award Numbers DE‐SC0018244 (Y.L., H.S., W.D., E.E., J.Sc., R.M., J.Sh), and DE‐SC0018254 (S.A., Yu.C.), and in part by the DOE Center for Advanced Bioenergy and Bioproducts Innovation (U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research under Award Number DE‐SC0018420; Yi.C., G.L., F.A. J. Sh. and X.Y) for providing recombinant DNA vectors and gene constructs that contributed to this work. The Division of Chemical Sciences, Geosciences, and Biosciences, Office of Basic Energy Sciences, US Department of Energy (grant DOE KC0304000; J.C. and J. Sh.) for contributing gene constructs to this work. The confocal microscopic study at the Center for Functional Nanomaterials at Brookhaven National Laboratory was supported under Office of Basic Energy Sciences of DOE with contract no. DE‐SC0012704 (to CFN). We thank Dr. Mircea Cotlet and Dr. Changcheng Xu for assistance with confocal microscopy. We also thank Dr. Zhiyang Zhai and Dr. Jantana Keereetaweep for sharing the GFP‐WRI1 gene and the anti‐WRI1antibody. We also thank Dr. Peter Rogowsky (Université de Lyon) for sharing the ZmWRI1b plasmid.

Data availability

All data generated or analyzed during this study are included in this article and in its Supporting Information. Source data are provided with this paper.

References

  1. Acosta, K. , Appenroth, K.J. , Borisjuk, L. , Edelman, M. , Heinig, U. , Jansen, M.A.K. , Oyama, T. et al. (2021) Return of the Lemnaceae: duckweed as a model plant system in the genomics and postgenomics era. Plant Cell, 33, 3207–3234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. An, D. and Suh, M.C. (2015) Overexpression of Arabidopsis WRI1 enhanced seed mass and storage oil content in Camelina sativa . Plant Biotechnol. Rep. 9, 137–148. [Google Scholar]
  3. Asrani, K.H. , Farelli, J.D. , Stahley, M.R. , Miller, R.L. , Cheng, C.J. , Subramanian, R.R. and Brown, J.M. (2018) Optimization of mRNA untranslated regions for improved expression of therapeutic mRNA. RNA Biol. 15, 756–762. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Beechey‐Gradwell, Z. , Cooney, L. , Winichayakul, S. , Andrews, M. , Hea, S.Y. , Crowther, T. and Roberts, N. (2019) Storing carbon in leaf lipid sinks enhances perennial ryegrass carbon capture especially under high N and elevated CO2 . J. Exp. Bot. 71, 2351–2361. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Board, A.J. , Crowther, J.M. , Acevedo‐Fani, A. , Meisrimler, C.N. , Jameson, G.B. and Dobson, R.C.J. (2022) How plants solubilise seed fats: revisiting oleosin structure and function to inform commercial applications. Biophys. Rev. 14, 257–266. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Boehm, R. , Kruse, C. , Voeste, D. , Barth, S. and Schnabl, H. (2001) A transient transformation system for duckweed (Wolffia columbiana) using Agrobacterium‐mediated gene transfer. J. Appl. Bot. 75, 107–111. [Google Scholar]
  7. Bog, M. , Appenroth, K.‐J. and Sree, K.S. (2019) Duckweed (Lemnaceae): its molecular taxonomy. Front. Sustain. Food Syst. 3, 117. [Google Scholar]
  8. Bog, M. , Sree, K.S. , Fuchs, J. , Hoang, P.T. , Schubert, I. , Kuever, J. , Rabenstein, A. et al. (2020) A taxonomic revision of Lemna sect. Uninerves (Lemnaceae). Taxon, 69, 56–66. [Google Scholar]
  9. Braglia, L. , Lauria, M. , Appenroth, K.J. , Bog, M. , Breviario, D. , Grasso, A. , Gavazzi, F. et al. (2021) Duckweed species genotyping and interspecific hybrid discovery by tubulin‐based polymorphism fingerprinting. Front. Plant Sci. 12, 270. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cai, Y. , McClinchie, E. , Price, A. , Nguyen, T.N. , Gidda, S.K. , Watt, S.C. , Yurchenko, O. et al. (2017) Mouse fat storage‐inducing transmembrane protein 2 (FIT2) promotes lipid droplet accumulation in plants. Plant Biotechnol. J. 15, 824–836. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Cai, Y.Q. , Whitehead, P. , Chappell, J. and Chapman, K.D. (2019) Mouse lipogenic proteins promote the co‐accumulation of triacylglycerols and sesquiterpenes in plant cells. Planta, 250, 79–94. [DOI] [PubMed] [Google Scholar]
  12. Cantó‐Pastor, A. , Mollá‐Morales, A. , Ernst, E. , Dahl, W. , Zhai, J. , Yan, Y. , Meyers, B. et al. (2015) Efficient transformation and artificial mi RNA gene silencing in Lemna minor . Plant Biol. 17, 59–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cases, S. , Smith, S.J. , Zheng, Y.‐W. , Myers, H.M. , Lear, S.R. , Sande, E. , Novak, S. et al. (1998) Identification of a gene encoding an acyl CoA:diacylglycerol acyltransferase, a key enzyme in triacylglycerol synthesis. Proc. Natl Acad. Sci. USA, 95, 13018–13023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Chakrabarti, R. , Clark, W.D. , Sharma, J.G. , Goswami, R.K. , Shrivastav, A.K. and Tocher, D.R. (2018) Mass production of Lemna minor and its amino acid and fatty acid profiles. Front. Chem. 6, 479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Chan, C.T.Y. , Lee, J.W. , Cameron, D.E. , Bashor, C.J. and Collins, J.J. (2016) ‘Deadman’ and ‘Passcode’ microbial kill switches for bacterial containment. Nat. Chem. Biol. 12, 82–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Chapman, K.D. , Dyer, J.M. and Mullen, R.T. (2012) Biogenesis and functions of lipid droplets in plants: thematic review series: lipid droplet synthesis and metabolism: from yeast to man. J. Lipid Res. 53, 215–226. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Chitraju, C. , Walther, T.C. and Farese, R.V. (2019) The triglyceride synthesis enzymes DGAT1 and DGAT2 have distinct and overlapping functions in adipocytes. J. Lipid Res. 60, 1112–1120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Chu, K.L. , Jenkins, L.M. , Bailey, S.R. , Kambhampati, S. , Koley, S. , Foley, K. , Arp, J.J. , Czymmek, K.J. , Bates, P.D. and Allen, D.K. (2020) Shifting carbon flux from non‐transient starch to lipid allows oil accumulation in transgenic tobacco leaves. BioRxiv .
  19. Demidenko, E. and Miller, T.W. (2019) Statistical determination of synergy based on bliss definition of drugs independence. PLoS One, 14, e0224137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Engler, C. , Youles, M. , Gruetzner, R. , Ehnert, T.‐M. , Werner, S. , Jones, J.D. , Patron, N.J. et al. (2014) A golden gate modular cloning toolbox for plants. ACS Synth. Biol. 3, 839–843. [DOI] [PubMed] [Google Scholar]
  21. Ernst, E. and Martienssen, R.A. (2011) Lemna.org ‐ Lemnaceae Genome Repository. http://www.lemna.org
  22. Focks, N. and Benning, C. (1998) wrinkled1: A novel, low‐seed‐oil mutant of Arabidopsis with a deficiency in the seed‐specific regulation of carbohydrate metabolism. Plant Physiol. 118, 91–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Geng, Q.Q. , Li, T. , Li, P.L. , Wang, X. , Chu, W.J. , Ma, Y.N. , Ma, H. et al. (2018) The accumulation, transformation, and effects of quinestrol in duckweed (Spirodela polyrhiza L.). Sci. Total Environ. 634, 1034–1041. [DOI] [PubMed] [Google Scholar]
  24. Grimberg, Å. , Carlsson, A.S. , Marttila, S. , Bhalerao, R. and Hofvander, P. (2015) Transcriptional transitions in Nicotiana benthamiana leaves upon induction of oil synthesis by WRINKLED1 homologs from diverse species and tissues. BMC Plant Biol. 15, 1–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Hamada, S. , Kishikawa, A. and Yoshida, M. (2020) Proteomic analysis of lipid droplets in Sesamum indicum . Protein J. 39, 366–376. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hoang, P.T.N. , Schubert, V. , Meister, A. , Fuchs, J. and Schubert, I. (2019) Variation in genome size, cell and nucleus volume, chromosome number and rDNA loci among duckweeds. Sci. Rep. 9, 13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Hsiao, E.S.L. and Tzen, J.T.C. (2011) Ubiquitination of oleosin‐H and caleosin in sesame oil bodies after seed germination. Plant Physiol. Biochem. 49, 77–81. [DOI] [PubMed] [Google Scholar]
  28. Kelly, A.A. , van Erp, H. , Quettier, A.‐L. , Shaw, E. , Menard, G. , Kurup, S. and Eastmond, P.J. (2013) The SUGAR‐DEPENDENT1 lipase limits triacylglycerol accumulation in vegetative tissues of Arabidopsis. Plant Physiol. 162, 1282–1289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kozak, M. (1987) An analysis of 5′‐noncoding sequences from 699 vertebrate messenger RNAs. Nucleic Acids Res. 15, 8125–8148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kozak, M. (1989) The scanning model for translation: an update. J. Cell Biol. 108, 229–241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Lee, K.R. , Kim, E.H. , Kim, K.H. , Park, J.S. and Kim, H.U. (2017) Vegetable oil production in vegetative plant tissues. Plant Biotechnol. Rep. 11, 385–395. [Google Scholar]
  32. Leng, R. , Stambolie, J. and Bell, R. (1995) Duckweed‐a potential high‐protein feed resource for domestic animals and fish. Livestock Res. Rural Dev. 7, 36. [Google Scholar]
  33. Les, D.H. , Crawford, D.J. , Landolt, E. , Gabel, J.D. and Kemball, R.T. (2002) Phylogeny and systematics of Lemnaceae, the duckweed family. Syst. Bot. 27, 221–240. [Google Scholar]
  34. Lin, L.J. , Tai, S.S.K. , Peng, C.C. and Tzen, J.T.C. (2002) Steroleosin, a sterol‐binding dehydrogenase in seed oil bodies. Plant Physiol. 128, 1200–1211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Liu, H. , Zhai, Z. , Kuczynski, K. , Keereetaweep, J. , Schwender, J. and Shanklin, J. (2019a) WRINKLED1 regulates BIOTIN ATTACHMENT DOMAIN‐CONTAINING proteins that inhibit fatty acid synthesis. Plant Physiol. 181, 55–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Liu, Y. , Wang, Y. , Xu, S. , Tang, X. , Zhao, J. , Yu, C. , He, G. et al. (2019b) Efficient genetic transformation and CRISPR/Cas9‐mediated genome editing in Lemna aequinoctialis . Plant Biotechnol. J. 17, 2143–2152. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Michael, T.P. , Ernst, E. , Hartwick, N. , Chu, P. , Bryant, D. , Gilbert, S. , Ortleb, S. et al. (2020) Genome and time‐of‐day transcriptome of Wolffia australiana link morphological minimization with gene loss and less growth control. Genome Res. 31, 225–238. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Mitchell, M.C. , Pritchard, J. , Okada, S. , Zhang, J. , Venables, I. , Vanhercke, T. and Ral, J.P. (2020) Increasing growth and yield by altering carbon metabolism in a transgenic leaf oil crop. Plant Biotechnol. J. 18, 2042–2052. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Murashige, T. and Skoog, F. (1962) A revised medium for rapid growth and bio assays with tobacco tissue cultures. Physiol. Plant. 15, 473–497. [Google Scholar]
  40. Naumann, B. , Eberius, M. and Appenroth, K.‐J. (2007) Growth rate based dose–response relationships and EC‐values of ten heavy metals using the duckweed growth inhibition test (ISO 20079) with Lemna minor L. clone St. J. Plant Physiol. 164, 1656–1664. [DOI] [PubMed] [Google Scholar]
  41. Norkunas, K. , Harding, R. , Dale, J. and Dugdale, B. (2018) Improving agroinfiltration‐based transient gene expression in Nicotiana benthamiana . Plant Methods, 14, 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Parajuli, S. , Kannan, B. , Karan, R. , Sanahuja, G. , Liu, H. , Garcia‐Ruiz, E. , Kumar, D. et al. (2020) Towards oilcane: engineering hyperaccumulation of triacylglycerol into sugarcane stems. GCB Bioenergy, 12, 476–490. [Google Scholar]
  43. Rusoff, L.L. , Blakeney, E.W., Jr. and Culley, D.D., Jr. (1980) Duckweeds (Lemnaceae family): a potential source of protein and amino acids. J. Agric. Food Chem. 28, 848–850. [DOI] [PubMed] [Google Scholar]
  44. Schenk, R.U. and Hildebrandt, A. (1972) Medium and techniques for induction and growth of monocotyledonous and dicotyledonous plant cell cultures. Can. J. Bot. 50, 199–204. [Google Scholar]
  45. Schlücking, K. , Edel, K.H. , Köster, P. , Drerup, M.M. , Eckert, C. , Steinhorst, L. , Waadt, R. et al. (2013) A new β‐estradiol‐inducible vector set that facilitates easy construction and efficient expression of transgenes reveals CBL3‐dependent cytoplasm to tonoplast translocation of CIPK5. Mol. Plant, 6, 1814–1829. [DOI] [PubMed] [Google Scholar]
  46. Sharma, J. , Clark, W.D. , Shrivastav, A.K. , Goswami, R.K. , Tocher, D.R. and Chakrabarti, R. (2019) Production potential of greater duckweed Spirodela polyrhiza (L. Schleiden) and its biochemical composition evaluation. Aquaculture, 513, 734419. [Google Scholar]
  47. Smith, S.J. , Cases, S. , Jensen, D.R. , Chen, H.C. , Sande, E. , Tow, B. , Sanan, D.A. et al. (2000) Obesity resistance and multiple mechanisms of triglyceride synthesis in mice lacking Dgat. Nat. Genet. 25, 87–90. [DOI] [PubMed] [Google Scholar]
  48. Stark, D.M. , Timmerman, K.P. , Barry, G.F. , Preiss, J. and Kishore, G.M. (1992) Regulation of the amount of starch in plant tissues by ADP glucose pyrophosphorylase. Science, 258, 287–292. [DOI] [PubMed] [Google Scholar]
  49. Stone, S.J. , Myers, H.M. , Watkins, S.M. , Brown, B.E. , Feingold, K.R. , Elias, P.M. and Farese, R.V., Jr. (2004) Lipopenia and skin barrier abnormalities in DGAT2‐deficient mice. J. Biol. Chem. 279, 11767–11776. [DOI] [PubMed] [Google Scholar]
  50. Tai, S.S.K. , Lee, T.T.T. , Tsai, C.C.Y. , Yiu, T.J. and Tzen, J.T.C. (2001) Expression pattern and deposition of three storage proteins, 11S globulin, 2S albumin and 7S globulin in maturing sesame seeds. Plant Physiol. Biochem. 39, 981–992. [Google Scholar]
  51. Tang, J. , Li, Y. , Ma, J. and Cheng, J.J. (2015) Survey of duckweed diversity in Lake Chao and total fatty acid, triacylglycerol, profiles of representative strains. Plant Biol. 17, 1066–1072. [DOI] [PubMed] [Google Scholar]
  52. Van Hoeck, A. , Horemans, N. , Monsieurs, P. , Cao, H.X. , Vandenhove, H. and Blust, R. (2015) The first draft genome of the aquatic model plant Lemna minor opens the route for future stress physiology research and biotechnological applications. Biotechnol. Biofuels, 8, 188. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Vanhercke, T. , El Tahchy, A. , Shrestha, P. , Zhou, X.R. , Singh, S.P. and Petrie, J.R. (2013) Synergistic effect of WRI1 and DGAT1 coexpression on triacylglycerol biosynthesis in plants. Febs Lett. 587, 364–369. [DOI] [PubMed] [Google Scholar]
  54. Vanhercke, T. , El Tahchy, A. , Liu, Q. , Zhou, X.R. , Shrestha, P. , Divi, U.K. , Ral, J.P. et al. (2014) Metabolic engineering of biomass for high energy density: oilseed‐like triacylglycerol yields from plant leaves. Plant Biotechnol. J. 12, 231–239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Vanhercke, T. , Divi, U.K. , El Tahchy, A. , Liu, Q. , Mitchell, M. , Taylor, M.C. , Eastmond, P.J. et al. (2017) Step changes in leaf oil accumulation via iterative metabolic engineering. Metab. Eng. 39, 237–246. [DOI] [PubMed] [Google Scholar]
  56. Vanhercke, T. , Belide, S. , Taylor, M.C. , El Tahchy, A. , Okada, S. , Rolland, V. , Liu, Q. et al. (2019) Up‐regulation of lipid biosynthesis increases the oil content in leaves of Sorghum bicolor. Plant Biotechnol. J. 17, 220–232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Venegas‐Caleron, M. , Sayanova, O. and Napier, J.A. (2010) An alternative to fish oils: metabolic engineering of oil‐seed crops to produce omega‐3 long chain polyunsaturated fatty acids. Prog. Lipid Res. 49, 108–119. [DOI] [PubMed] [Google Scholar]
  58. Voinnet, O. , Rivas, S. , Mestre, P. and Baulcombe, D. (2003) An enhanced transient expression system in plants based on suppression of gene silencing by the p19 protein of tomato bushy stunt virus. Plant J. 33, 949–956. [DOI] [PubMed] [Google Scholar]
  59. Vunsh, R. , Li, J.H. , Hanania, U. , Edelman, M. , Flaishman, M. , Perl, A. , Wisniewski, J.P. et al. (2007) High expression of transgene protein in Spirodela . Plant Cell Rep. 26, 1511–1519. [DOI] [PubMed] [Google Scholar]
  60. Wang, W. , Wu, Y. , Yan, Y. , Ermakova, M. , Kerstetter, R. and Messing, J. (2010) DNA barcoding of the Lemnaceae, a family of aquatic monocots. BMC Plant Biol. 10, 1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Wang, W. , Kerstetter, R.A. and Michael, T.P. (2011) Evolution of genome size in duckweeds (Lemnaceae). J. Bot. 2011, 1–9. [Google Scholar]
  62. Wang, W. , Haberer, G. , Gundlach, H. , Gläßer, C. , Nussbaumer, T. , Luo, M. , Lomsadze, A. et al. (2014) The Spirodela polyrhiza genome reveals insights into its neotenous reduction fast growth and aquatic lifestyle. Nat. Commun. 5, 1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Wang, K.‐T. , Hong, M.‐C. , Wu, Y.‐S. and Wu, T.‐M. (2021) Agrobacterium‐mediated genetic transformation of taiwanese isolates of Lemna aequinoctialis . Plants, 10, 1576. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Weber, E. , Engler, C. , Gruetzner, R. , Werner, S. and Marillonnet, S. (2011) A modular cloning system for standardized assembly of multigene constructs. PLoS One, 6, e16765. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Werner, S. , Engler, C. , Weber, E. , Gruetzner, R. and Marillonnet, S. (2012) Fast track assembly of multigene constructs using Golden Gate cloning and the MoClo system. Bioeng. Bugs. 3, 38–43. [DOI] [PubMed] [Google Scholar]
  66. Winichayakul, S. , Scott, R.W. , Roldan, M. , Hatier, J.‐H.B. , Livingston, S. , Cookson, R. , Curran, A.C. et al. (2013) In vivo packaging of triacylglycerols enhances Arabidopsis leaf biomass and energy density. Plant Physiol. 162, 626–639. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Wood, C.C. , Petrie, J.R. , Shrestha, P. , Mansour, M.P. , Nichols, P.D. , Green, A.G. and Singh, S.P. (2009) A leaf‐based assay using interchangeable design principles to rapidly assemble multistep recombinant pathways. Plant Biotechnol. J. 7, 914–924. [DOI] [PubMed] [Google Scholar]
  68. Xu, J. , Cheng, J.J. and Stomp, A.M. (2012a) Growing Spirodela polyrrhiza in swine wastewater for the production of animal feed and fuel ethanol: a pilot study. CLEAN–Soil Air Water, 40, 760–765. [Google Scholar]
  69. Xu, J.L. , Zhao, H. , Stomp, A.M. and Cheng, J.J. (2012b) The production of duckweed as a source of biofuels. Biofuels, 3, 589–601. [Google Scholar]
  70. Xu, X.Y. , Akbar, S. , Shrestha, P. , Venugoban, L. , Devilla, R. , Hussain, D. , Lee, J.W. et al. (2020) A synergistic genetic engineering strategy induced triacylglycerol accumulation in potato (Solanum tuberosum) leaf. Front. Plant Sci. 11, 215. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Yamamoto, Y.T. , Rajbhandari, N. , Lin, X.H. , Bergmann, B.A. , Nishimura, Y. and Stomp, A.M. (2001) Genetic transformation of duckweed Lemna gibba and Lemna minor . In Vitro Cell. Dev. Biol. Plant, 37, 349–353. [Google Scholar]
  72. Yan, Y. , Candreva, J. , Shi, H. , Ernst, E. , Martienssen, R. , Schwender, J. and Shanklin, J. (2013) Survey of the total fatty acid and triacylglycerol composition and content of 30 duckweed species and cloning of a Δ6‐desaturase responsible for the production of γ‐linolenic and stearidonic acids in Lemna gibba . BMC Plant Biol. 13, 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Yang, Y. , Munz, J. , Cass, C. , Zienkiewicz, A. , Kong, Q. , Ma, W. , Sanjaya, S. et al. (2015) Ectopic expression of WRINKLED1 affects fatty acid homeostasis in Brachypodium distachyon vegetative tissues. Plant Physiol. 169, 1836–1847. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Yee, S. , Rolland, V. , Reynolds, K.B. , Shrestha, P. , Ma, L. , Singh, S.P. , Vanhercke, T. et al. (2021) Sesamum indicum Oleosin L improves oil packaging in Nicotiana benthamiana leaves. Plant Direct, 5, e343. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Yu, X.‐H. , Prakash, R.R. , Sweet, M. and Shanklin, J. (2014) Coexpressing Escherichia coli cyclopropane synthase with Sterculia foetida lysophosphatidic acid acyltransferase enhances cyclopropane fatty acid accumulation. Plant Physiol. 164, 455–465. [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. Zale, J. , Jung, J.H. , Kim, J.Y. , Pathak, B. , Karan, R. , Liu, H. , Chen, X.H. et al. (2016) Metabolic engineering of sugarcane to accumulate energy‐dense triacylglycerols in vegetative biomass. Plant Biotechnol. J. 14, 661–669. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Zhai, Z. , Liu, H. and Shanklin, J. (2017) Phosphorylation of WRINKLED1 by KIN10 results in its proteasomal degradation, providing a link between energy homeostasis and lipid biosynthesis. Plant Cell, 29, 871–889. [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Zhai, Z. , Keereetaweep, J. , Liu, H. , Feil, R. , Lunn, J.E. and Shanklin, J. (2018) Trehalose 6‐phosphate positively regulates fatty acid synthesis by stabilizing WRINKLED1. Plant Cell, 30, 2616–2627. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Ziegler, P. , Sree, K. and Appenroth, K.‐J. (2016) Duckweeds for water remediation and toxicity testing. Toxicol. Environ. Chem. 98, 1127–1154. [Google Scholar]
  80. Zuo, J. , Niu, Q.‐W. and Chua, N.‐H. (2000) An estrogen receptor‐based transactivator XVE mediates highly inducible gene expression in transgenic plants. Plant J. 24, 265–273. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1 Representative images of duckweed.

Figure S2 Screening of L. japonica transformants using plate reader.

Figure S3 TAG accumulation in duckweed treated with different concentrations of estradiol for different time periods.

Figure S4 Sequence and features of optimized CFP.

Table S1 TAG content as percent (DW) in WT and different transgenic lines of Lemna japonica.

Click here for additional data file. (1.9MB, pdf)

Table S2 Golden gate clone and qPCR primers.

Click here for additional data file. (22.2KB, xlsx)

Data Availability Statement

All data generated or analyzed during this study are included in this article and in its Supporting Information. Source data are provided with this paper.

Articles from Plant Biotechnology Journal are provided here courtesy of Wiley

RESOURCES