- © 2020 American Society of Plant Biologists. All Rights Reserved.
Abstract
The Arabidopsis (Arabidopsis thaliana) fatty acid biosynthesis1 (fab1) mutant has increased levels of the saturated fatty acid 16:0, resulting from decreased activity of 3-ketoacyl-ACP synthase II. In fab1 leaves, phosphatidylglycerol, the major chloroplast phospholipid, contains >40% high-melting-point molecular species (HMP-PG; molecules that contain only 16:0, 16:1-trans, and 18:0 fatty acids)—a trait associated with chilling-sensitive plants—compared with <10% in wild-type Arabidopsis. Although they do not exhibit short-term chilling sensitivity when exposed to low temperatures (2°C to 6°C) for long periods, fab1 plants do suffer collapse of photosynthesis, degradation of chloroplasts, and eventually death. To test the relevance of HMP-PG to the fab1 phenotype, we used transgenic 16:0 desaturases targeted to the endoplasmic reticulum and the chloroplast to lower 16:0 in leaf lipids of fab1 plants. We produced two lines that had very similar lipid compositions except that one, ER-FAT5, contained high HMP-PG, similar to the fab1 parent, while the second, TP-DES9*, contained <10% HMP-PG, similar to the wild type. TP-DES9* plants, but not ER-FAT5 plants, showed strong recovery and growth following 75 d at 2°C, demonstrating the role of HMP-PG in low-temperature damage and death in fab1, and in chilling-sensitive plants more broadly.
In higher plants, the chloroplast membranes that host the light harvesting and electron transport processes of photosynthesis have a characteristically high number of double bonds in the glycerolipid acyl chains. Only ∼10% of the fatty acids that compose the hydrophobic core of the thylakoid bilayer lack double bonds altogether, whereas >80% are polyunsaturated, having two or three double bonds (Ohlrogge et al., 2015). The photosynthetic light reactions produce reactive oxygen species as by-products, and these can degrade polyunsaturated fatty acids, so it is assumed that highly unsaturated membranes are required to support photosynthesis (McConn and Browse, 1998).
The glycerolipids in chloroplast membranes are synthesized by two separate pathways. (Browse et al., 1986; Ohlrogge and Browse, 1995). Synthesis de novo of fatty acids takes place in the stroma of chloroplasts, producing 16:0 esterified to acyl carrier protein (ACP). A large proportion of this 16:0-ACP is elongated by 3-keto-acyl-ACP synthase II (KASII) to 18:0-ACP, which is in turn desaturated by stearoyl ACP desaturase to produce 18:1-ACP (Lindqvist et al., 1996; Carlsson et al., 2002). The fatty acids from 16:0-ACP and 18:1-ACP may be used within the chloroplast in the prokaryotic pathway (Kunst et al., 1988; Kim and Huang, 2004) to produce phosphatidic acid (PA). Some of this PA intermediate is used for synthesis of phosphatidylglycerol (PG; Ohlrogge and Browse, 1995; Wada and Murata, 2007), which is the only chloroplast glycerolipid that is produced solely by the prokaryotic pathway. In some plants, including Arabidopsis (Arabidopsis thaliana), PA is also converted to diacylglycerol (DAG), which is the precursor for the synthesis of the other chloroplast glycerolipids, monogalactosyldiacylglycerol (MGD), digalactosyldiacylglycerol (DGD), and sulfoquinovosyldiacylglycerol (SQD; Browse et al., 1986; Ohlrogge and Browse, 1995; Ohlrogge et al., 2015).
The second route for chloroplast glycerolipid synthesis, the eukaryotic pathway, begins with export of 16:0 and 18:1 from the chloroplast as CoA thioesters. (Li et al., 2015). In the endoplasmic reticulum (ER), these fatty acids are rapidly incorporated into phosphatidylcholine (PC) by acyl exchange (Bates et al., 2007), and are also used (via PA and DAG intermediates) for the synthesis of all the phospholipids of the extrachloroplast membranes of the cell (Ohlrogge et al., 2015). In addition however, the DAG moiety of PC can be returned to the chloroplast and contribute to the production of MGD, DGD, and SQD required for thylakoid synthesis (Benning, 2009; Roston et al., 2012). The ER-to-chloroplast flux of lipid is reversible to some extent (Browse et al., 1989, 1993).
With the exception of the first Δ9 double bond in 18:1-ACP, all the double bonds in the acyl chains are introduced after the initial synthesis of glycerolipid molecules. In Arabidopsis, this involves the action of seven fatty acid desaturases that are integral membrane proteins in the chloroplast and ER (Ohlrogge and Browse, 1995; Wallis and Browse, 2010). Characterization of Arabidopsis fatty acid desaturation (fad) mutants deficient in one or more of these desaturases has shown that the high level of thylakoid unsaturation is essential to photosynthetic function (Murakami et al., 2000; Routaboul et al., 2000). For example, fad2 fad6 double-mutant plants are unable to synthesize polyunsaturated fatty acids and cannot grow autotrophically; however, when grown on Suc as a carbon source, the double mutants are robust plants showing strong leaf and root development (McConn and Browse, 1998). These results indicate that the vast majority of receptor-mediated and transport-related membrane functions required to sustain the organism and induce proper development are adequately supported in the absence of polyunsaturated lipids; photosynthesis is the one process that requires high levels of polyunsaturation. Mutants with smaller changes in unsaturation are often similar to the wild type under typical growth-chamber conditions and reveal their phenotypes only under more extreme conditions (Wallis and Browse, 2002, 2010). Several mutants grow more slowly and become chlorotic at temperatures in the range 2°C to 10°C (Hugly and Somerville, 1992; Routaboul et al., 2000), indicating a role for fatty acid composition in maintaining photosynthesis at these low temperatures.
Like other species native to temperate regions, Arabidopsis is chilling resistant and able to grow at temperatures close to 0°C. By contrast, many tropical and subtropical plant species are chilling sensitive and suffer sharp reductions of photosynthesis and extensive tissue damage after even short exposure to low temperatures. Many of the world’s most important crops, including rice (Oryza sativa), maize (Zea mays), and soybean (Glycine max) are chilling sensitive, so a better understanding of the biochemical and genetic factors contributing to this sensitivity has the potential to enhance sustainable food production (Nishida and Murata, 1996; Iba, 2002; Thakur et al., 2010). One hypothesis proposes that chilling sensitivity is a result of the fatty acid composition of chloroplast PG. It is based on the observation that many chilling-sensitive plants contain >30% of PG molecules with only saturated or trans unsaturated fatty acids—16:0, 18:0, and 16:1-Δ3trans (16:1t)—at both the sn-1 and sn-2 positions of the glycerol backbone, referred to as high-melting-point molecular species (HMP-PG; Murata, 1983; Barkan et al., 2006). This name alludes to the fact that HMP-PG species can induce a phase change from liquid crystalline (typical of biological membranes) to gel phase at temperatures well above 0°C and thereby disrupt membrane and cellular function (Murata and Yamaya, 1984). Chilling-resistant plants have <10% HMP species in chloroplast PG (Murata et al., 1982; Murata, 1983; Roughan, 1985).
One perspective on the role of HMP-PG in plant temperature responses has come from our investigations of the fatty acid biosynthesis1 (fab1) mutant of Arabidopsis. In this mutant, a hypomorphic mutation in the gene encoding KASII reduces elongation of 16:0-ACP to 18:0-ACP (Carlsson et al., 2002), producing plants that have increased levels of 16:0 in all membrane glycerolipids (Wu et al., 1994). In particular, fab1 plants contain HMP-PG at levels (∼40% to 50% of total PG) similar to those of many chilling-sensitive plant species (Wu and Browse, 1995). Nevertheless, the fab1 mutant does not show typical symptoms of chilling sensitivity and is unaffected, in comparison to wild-type controls, by a range of chilling treatments that kill chilling-sensitive plants; instead, fab1 plants only show a collapse of photosynthesis after >10 d of exposure to 2°C, with the plants dying after several weeks at low temperature (Wu and Browse, 1995; Wu et al., 1997).
We have previously screened for genetic suppressors of the fab1 low-temperature phenotype. Most, though not all, of the suppressor mutations substantially reduce the proportion of saturated fatty acids in PG, consistent with the notion that HMP-PG causes eventual death of fab1 plants in the cold (Barkan et al., 2006; Kim et al.,2010; Gao et al., 2015). However, all the suppressors have additional changes, relative to fab1, in the fatty acid compositions of membrane lipids that prevent a clear linkage between reductions in HMP-PG and improved low-temperature survival.
Here, we have taken a new approach to investigating the role of HMP-PG in damage and death of fab1 plants at chilling temperatures by using a 16:0-CoA desaturase from Caenorhabditis elegans, FAT-5 (Watts and Browse, 2000), and a glycerolipid desaturase, DES9*15, derived from a cyanobacterial enzyme by directed evolution (Bai et al., 2016). When expressed in the fab1 mutant background, both the FAT-5 enzyme targeted to the ER and the DES9*15 enzyme targeted to the chloroplast reduced leaf 16:0 to near-wild type levels. The fatty acid compositions of individual leaf lipids in plants of both transgenic lines were very similar, with the sole exception of PG. Plants expressing the FAT-5 desaturase retained high levels of HMP-PG, similar to fab1, while plants expressing the DES9*15 enzyme had HMP-PG lowered to levels close to those of the wild type. Like the fab1 mutant, fab1 plants expressing a 16:0 desaturase in the ER lost photosynthetic function over 28 d of exposure to 2°C and showed little capacity for recovery and growth after longer periods at low temperature. By contrast, plants expressing a 16:0 desaturase targeted to the chloroplast retained substantial photosynthetic function, even after 75 d at 2°C, and were subsequently able to resume growth, flower, and set seed upon return to 22°C. These results provide the most direct evidence yet that high levels of HMP-PG cause gradual loss of photosynthesis and eventual death of plants at chilling temperatures.
RESULTS
Using Desaturases to Lower 16:0 in fab1
Our goal was to use alternative targeting of transgenic 16:0 desaturases to probe the role of HMP-PG in triggering the damage and eventual death of fab1 plants at 2°C. We reasoned that targeting a desaturase to the chloroplast could lead to a reduction in HMP-PG through conversion of 16:0 to cis-unsaturated 16:1 (16:1c), while a desaturase targeted to the ER would lower 16:0 in lipids derived from the eukaryotic pathway, again by conversion to 16:1c. The C. elegans FAT-5 protein is an ER-resident 16:0-CoA desaturase that we have previously expressed in Arabidopsis seeds to lower the saturate content of the oil. In that study, seed-specific expression of FAT-5 resulted in a substantial reduction in 16:0 content of PC and other polar lipids (Fahy et al., 2013). We therefore transformed fab1 plants with the FAT-5 coding sequence under control of the constitutive cauliflower mosaic virus 35S promoter (CaMV35S). Transgenic T1 plants were selected on the basis of BASTA resistance encoded in the transformation vector. Three T1 plants with reduced 16:0 relative to the fab1 parent were identified by fatty acid analysis of leaf samples. T2 seed collected from each of these lines segregated ∼75% BASTA-resistant seedlings, indicating a single site of vector insertion. Homozygous fab1 FAT-5 T3 plants from all three lines (ER-FAT5 lines 13, 21, and 30; Supplemental Table S1) had levels of 16:0 in the leaf fatty acid analyses similar to those of wild-type plants (15% to 16%), and one line, designated ER-FAT5 line 21 (Fig. 1), was chosen for further analysis.
Transgenic 16:0 desaturases have different effects on the leaf fatty acid composition of the fab1 mutant. Plants of four transgenic lines were grown together with fab1 and wild-type controls at 22°C for 25 d before measurement of the overall fatty acid compositions of leaf samples. Data are represented as the mean ± se (n = 4).
To target the FAT-5 protein to the chloroplast, we inserted DNA encoding the chloroplast targeting peptide (TP) of the ribulose bisphosphate carboxylase small subunit protein in frame and immediately 5′ to the FAT-5 coding sequence, and transformed fab1 plants as described above. T1 plants expressing this TP-FAT5 construct had only small increases in the level of 16:1 in the leaves. Three single-insert lines with the highest levels of 16:1 were used to produce homozygous fab1 TP-FAT5 plants (TP-FAT5 lines 7, 24, and 40), but none had any substantial reduction in the level of 16:0 relative to the fab1 parent (Supplemental Table S1). The leaf fatty acid profile of one of these, TP-FAT5 line 40, is shown in Figure 1. This result might be expected given that the FAT-5 desaturase is expected to require 16:0-CoA as its substrate, and such acyl-CoAs are probably not present in chloroplasts (Ohlrogge and Browse, 1995). Principal component analysis (PCA) of the data shown in Figure 1 (Supplemental Fig. S1) confirmed that the overall leaf fatty acid compositions of TP-FAT5 is very close to that of fab1. Because the TP-FAT5 enzyme did not provide any decrease in leaf 16:0 relative to the fab1 parent, these transgenic lines were not investigated further.
Our directed evolution of the DESG 16:0-glycerolipid desaturase from Synechococcus elongatus PCC6301 targeted to the ER of yeast provided several variants with greatly improved activity (Bai et al., 2016). One of these, DES9*15, provides for considerable conversion of 16:0 to 16:1c both in yeast and when expressed in Arabidopsis seeds (Bai et al., 2016). We used this variant, herein called DES9*, in constructs similar to those described above for FAT-5 to target the protein to the ER or to the chloroplast. The T1 fab1 ER-DES9* transgenics were small plants that grew slowly. Fifteen independent T1 transformants were analyzed for leaf fatty acid composition. All contained 27% to 35% 16:1c and levels of 16:0 as low as 9%, which was considerably lower than fab1 (25%) or the wild-type (17%) controls. The leaf fatty acid compositions of three ER-DES9* T3 lines (ER-DES9* lines 2, 11, and 30) are shown in Supplemental Table S1. We selected the T1 plant with highest 16:0 (13%) and a single transgene insert and generated a homozygous line, ER-DES9* line 11, for further characterization. The leaf fatty acid profile of homozygous ER-DES9* line 11 plants contained 40% 16:1 and only 8% 16:0, with large reductions also in the proportion of 16:3, 18:2, and 18:3 fatty acids relative to both fab1 and the wild type (Fig. 1). The PCA results (Supplemental Fig. S1) indicate that the overall leaf fatty acid composition of the ER-DES9* line is very distinct from the other five lines in this experiment. In addition, the plants were substantially smaller than those of the other lines studied as described below.
TP-DES9*, the construct for chloroplast targeting of DES9* in fab1, produced less extreme changes in overall leaf fatty acid composition. Three homozygous, single-insert lines (TP-DES9* lines 38, 47, and 66) contained ∼12% to 13% 16:0 (Supplemental Table S1), and one of these, TP-DES9* line 47 (Fig. 1), was chosen for further analysis. The PCA results (Supplemental Fig. S1) indicate that, in terms of overall leaf fatty acid composition, the TP-DES9* and ER-FAT5 transgenics are more closely related to each other than to other lines in this experiment. The results reported below for temperature responses of ER-FAT5 line 21 were also observed for the other two transgenic lines made with this construct (lines 13 and 30). Similarly, the results reported for TP-DES9* line 47 were also observed in the other two transgenic lines (38 and 66) made with this construct.
Growth and Development of Plants at 22°C
When germinated on potting-mix media under our standard growth chamber conditions (22°C, 120–150 μmol m−2 s−1 photosynthetically active radiation (PAR), and a 16-h photoperiod), the young seedlings of ER-FAT5 and TP-DES9* were similar to those of fab1 and the wild type. After 25 d of growth, rosette plants of the four lines were similar in appearance (Fig. 2A), although plants of the ER-FAT5 line were slightly smaller, with lower dry weight (38.5 ± 1.1 mg plant−1) than plants of fab1 and TP-DES9* (45.5 ± 2.3 and 44.0 ± 2.7 mg, respectively; Fig. 2B). Wild-type plants averaged 52.5 ± 3.8 mg. To assess photosynthetic function in the four lines, we measured the potential quantum yield of PSII, the ratio of variable fluorescence to maximal fluorescence (Fv/Fm). Previous studies have established that Fv/Fm values are well correlated with other measures of photosynthetic function for plants subject to a range of temperature and light regimes (Wu et al., 1997; Vijayan and Browse, 2002; Baker and Rosenqvist, 2004). There were no significant differences in values measured across all four lines (Fig. 2C), with the Fv/Fm of 0.848 ± 0.001 being typical of young, healthy Arabidopsis plants.
Development and photosynthetic fluorescence analysis of wild-type (WT), fab1, and transgenic plants grown at 22°C. A, Representative rosette plants after 25 d of growth. Scale bar = 1 cm. B, Dry weight of 25-d-old plants is represented as the mean ± se (n = 4). Statistical analysis was performed using a one-way ANOVA with Tukey’s honestly significant difference (HSD) mean-separation test. Different letters above each column indicate significant differences (P < 0.001) between plant lines. C, Measurements of Fv/Fm for 25-d-old plants, represented as the mean ± se (n = 6). D, Representative plants after 45 d of growth.
As the experiment continued, plants of the ER-FAT5 and TP-DES9* lines developed slightly more slowly than fab1 but flowered normally (Fig. 2D) and, at maturity, yielded abundant seed. These results indicate that the changes in membrane fatty acid composition brought about by each of these two transgenes have little effect on the growth, development, and function of the plants at 22°C, compared to the fab1 parental line and the wild type.
The growth and appearance of plants of the ER-DES9* line were dramatically different. Rosette plants of this line were very small after 25 d of growth (Fig. 2A), with a mean dry weight of only 14.5 ± 0.8 mg plant−1. Although the Fv/Fm measured on these plants (0.851 ± 0.001) was the same as for the other four lines (Fig. 2C), they continued to grow poorly and produced many fewer flowers and seeds than any of the other lines (Fig. 2D). The much larger changes in overall leaf fatty acid composition caused by expression of the ER-DES9* construct (Fig. 1), together with the poor growth and development of these plants, make this line problematic for understanding the possible role of HMP-PG in the fab1 low-temperature phenotype. For this reason, we chose to focus our investigations on the ER-FAD5 and TP-DES9* lines.
Major Membrane Lipids in ER-FAT5 and TP-DES9* Have Similar Fatty Acid Composition
To further investigate the effects of ER-FAT5 and TP-DES9* expression on the composition of membrane glycerolipids, we extracted lipids from leaves of transgenic plants, as well as fab1 and wild-type controls, and analyzed the fatty acid composition of individual lipids separated by thin-layer chromatography. The fab1 mutant is characterized by increases in 16:0 content of all the chloroplast and extrachloroplast glycerolipids relative to that of the wild type (Fig. 3; Wu et al., 1994). In the major extrachloroplast lipid, PC, expression of either ER-FAT5 or TP-DES9* resulted in substantial accumulation of 16:1c, with decreased levels of 16:0, 18:0, 18:2, and 18:3 relative to both fab1 and the wild type (Fig. 3A). Strikingly, the fatty acid composition of this lipid is very similar in ER-FAT5 and TP-DES9* plants. A very similar pattern is also evident for phosphatidylethanolamine (PE; Fig. 3B), and again expression of either desaturase produces comparable changes in fatty acid composition. Statistical analysis of the data by PCA (Fig. 3, A and B; Supplemental Fig. S2, A and B) indicates that the compositions of PC and PE in ER-FAT5 and TP-DES9* are closely related to one another and very distinct from the composition of the same lipids in the wild type or fab1. Thus, the DES9* enzyme targeted to the chloroplast results in 16:1c accumulation in PC and PE, along with other concomitant changes in fatty acid composition, including reductions in both 18:2 and 18:3.
Fatty acid composition of major leaf lipids of wild-type, fab1, and transgenic plants. Data for PC (A), PE (B), MGD (C), DGD (D), SQD (E), and PG (F) are represented as the mean ± se (n = 3).
The ER-FAD5 and TP-DES9* plants also have very similar fatty acid profiles for three of the chloroplast glycerolipids, MGD, DGD, and SQD (Fig. 3, C–E). The fab1 mutation increases 16:0 in MGD by only a small amount (Fig. 3C), and in the fab1 background, both transgenes result in 16:1 accumulation in this lipid, mainly at the expense of 16:3, 18:2, and 18:3. In DGD, by contrast, the fab1 mutation increases 16:0 from 18.0% ± 1.5% in the wild type to 30.1% ± 1.4% of the fatty acids in this lipid. Expression of the ER-FAT5 desaturase in fab1 lowers the level of 16:0 in DGD to 19.4% ± 0.8%, while the TP-DES9* enzyme reduces it to 15.9% ± 0.9% (Fig. 3D). In SQD, both desaturases also bring about large declines in 16:0, to 38.5% ± 2.0% for ER-FAT5 and 34.1% ± 1.7% in TP-DES9*, compared to 54.1% ± 1.6% in fab1 and 47.6% ± 2.0% in the wild type (Fig. 3E). In DGD and SQD, expression of either desaturase causes a small increase of 18:1 and small decreases of 18:2 and, for TP-DE9*, 18:3. PCA analysis for MGD and DGD indicates that the compositions of these lipids from the two transgenic lines are more closely related to each other than to the wild type or fab1, while SQD shows greater separation (Supplemental Fig. S2, C–E)
The PG fatty acid composition also breaks the pattern of close similarity in the lipid compositions of ER-FAT5 and TP-DES9* lines. The fab1 mutation increases 16:0 in PG to 42.4% ± 0.3% compared to 29.9% ± 0.6% in the wild type. The ER-FAT5 desaturase reduces 16:0 only slightly, to 39.6% ± 0.8%, with accumulation of 16:1c to 6.2% ± 0.8% in PG. By contrast, in the TP-DES9* plants, 16:0 in PG is reduced to 24.5% ± 0.3% and 16:1c increased to 22.4% ± 0.6% (Fig. 3F). The results from PCA analysis (Supplemental Fig. S2F) reflect this larger separation of the two transgenic lines.
The results from these lipid analyses were encouraging because they indicate that the compositions of both chloroplast and extrachloroplast membrane lipids in our two transgenic desaturase lines were very similar to each other, with the exception of PG. Importantly, the data in Figure 3F indicate the possibility that PG in ER-FAT5 plants will contain a high proportion of HMP-PG, similar to the fab1 mutant, while in TP-DES9* plants, HMP may be reduced to levels closer to that of the wild type.
TP-DES9* Reduces HMP-PG, but ER-FAT5 Does Not
To get a second perspective on the composition of membrane lipids, and to quantify the proportion of HMP molecular species in PG of our four lines, we submitted samples for lipidomics analysis to the KS Lipidomics Research Center (Esch et al., 2007). For all the major chloroplast and extrachloroplast glycerolipids, with the exception of PG and SQD, the lipidomics data (Table 1; Supplemental Table S2) confirm the comparable effects of expressing either ER-FAD5 or TP-DES9* in the fab1 mutant. Relative to the fab1 parent, both desaturases resulted in substantial decreases in molecular species containing 16:0, as well as in the molecular species containing combinations of 18:2 and 18:3. These were replaced by molecular species containing 16:1c, the product of the desaturases. Importantly, the molecular species distribution within each of these membrane lipids is very similar in ER-FAD5 and TP-DES9* plants (Table 1; Supplemental Table S2). The close relationships is confirmed by the results of PCA (Supplemental Fig. S4, A–F), which shows the compositions of these two lines clustered together and very distinct from the wild type and fab1. The molecular species composition of PG breaks this pattern. Consistent with previous studies (Wu and Browse, 1995; Gao et al., 2015), the lipidomics analyses show that the fab1 mutation causes an increase in HMP-PG. In this experiment, HMP-PG accounts for 51.5% ± 1.4% of the total PG molecular species compared with only 6.4% ± 0.4% in the wild type (Table 2). This calculation is based on identification of the 32:0 and 32:1 molecular species as 16:0/16:0 and 16:0/16:1t fatty acid combinations, since PG in the wild type and fab1 contain only very low levels of 16:1c (Fig. 3F).
Values are mean percentages of the total in each lipid calculated from lipidomics data in Supplemental Table S2.
Values are mean percentages of the total (± SE) calculated from lipidomics data from four biological replicates (Supplemental Table S2). The total of species with two 18-carbon fatty acids is shown as 36:X.
Expression of ER-FAT5 reduces the sum of 32:0 and 32:1 by only a modest amount, to 45.2%. Because PG in ER-FAT5 contains 6.2% 16:1c and 18.5% 16:1t (Fig. 3), we needed to consider the likelihood that a portion of the 32:1 molecules contained 16:1c as the monounsaturated fatty acid and are therefore not an HMP-PG species. The 16:1t (Δ3-trans) and 16:1c (Δ9-cis) acyl chains fragment differently in the mass spectrometer (Hsu et al., 2007), because the Δ3 double bond leads to preferential loss of 16:1t as a ketene (yielding an [M-H-236] ion) rather than loss of the fatty acid, as happens with 16:1c, which has its double bond at the Δ9 position. In the case of 16:0/16:1-PG from ER-FAT5, the analysis indicates that less <15% of the molecules contain 16:1c, meaning that HMP-PG in these plants is at least 39.7% ± 0.3% of the total PG. As indicated by ANOVA (Supplemental Fig. S3), this is significantly lower than HMP-PG in fab1 but is still in the middle of the range found for chilling-sensitive plants (Murata, 1983; Roughan, 1985).
Only the TP-DES9* desaturase brings about a substantial decline in HMP-PG. In Table 2, the sum of 32:0 plus 32:1 species is 13.1%. Calculations based on the abundance of [M-H-236] in the mass spectrum of the 32:1 peak indicated that no more than half of the molecules contained 16:1t. Applying this correction to the original lipidomics analyses, the HMP-PG in these plants is calculated to be <7.4% ± 0.1% of total PG, close to the 6.4% ± 0.3% found for wild-type PG in these analyses.
PCA results for the lipidomics data in Table 2 and Supplemental Table S2, confirmed that with respect to the molecular species composition of most membrane glycerolipids, ER-FAD5 and TP-DES9* are most closely related to one another and more distant from both the wild type and fab1. However, for PG and SQD, the relationships among the lines are shifted, with the ER-FAT5 composition being closer to fab1 than to TP-DES9* (Supplemental Fig. S3).
Only the TP-DES9* Desaturase Rescues the fab1 Mutant in the Cold
From our lipid analyses, we conclude that plants of the TP-DES9* and ER-FAT5 lines are substantially similar in the fatty acid composition of their membrane glycerolipids, except that the ER-FAT5 desaturase does not substantially lower the proportion of HMP molecular species in PG relative to the fab1 parental line, while the TP-DES9* enzyme restores HMP-PG to near wild-type levels. We therefore used these lines in experiments to probe the role of HMP-PG in the low-temperature phenotype that is observed in fab1 plants (Barkan et al., 2006; Gao et al., 2015). To test the initial responses to low-temperature treatment, we grew plants of ER-FAT5, TP-DES9*, fab1, and the wild type for 21 d at 22°C before transferring them to 2°C. After 20 d at 2°C, the fab1 and ER-FAT5 plants were slightly more chlorotic than plants of the wild-type and TP-DES9* lines (Fig. 4A). Despite this relatively subtle difference in appearance, fluorescence analysis indicated that the potential quantum yield of PSII had declined considerably in both fab1 (Fv/Fm = 0.299 ± 0.057) and ER-FAT5 (Fv/Fm = 0.295 ± 0.046) plants (Fig. 4B). By contrast, Fv/Fm in the wild type (0.808 ± 0.005) and TP-DES9* (0.755 ± 0.014) remained close to values measured for plants of all the lines grown at 22°C (Fig. 2C). After a further 7 d at 2°C, plant appearance was unchanged, but both fab1 and ER-FAT5 plants showed a further decline in Fv/Fm to 0.113 ± 0.041 and 0.134 ± 0.046, respectively (Fig. 4C). The Fv/Fm for TP-DES9* had declined significantly to 0.561 ± 0.046, while wild-type plants averaged 0.817 ± 0.006. In this experiment, ER-FAT5 plants responded the same as fab1 controls, consistent with previous studies of the fab1 mutant (Barkan et al., 2006; Gao et al., 2015). For the TP-DES9* line, Fv/Fm was similar to values found in other fab1 suppressor lines (Barkan et al., 2006).
Phenotype and photosynthetic fluorescence analysis of wild-type (WT), fab1, and transgenic plants after transfer to 2°C. A, Representative plants after 20 d at 2°C. Scale bar = 1 cm. B, Measurements of Fv/Fm for plants after 20 d at 2°C. C, Measurements of Fv/Fm for plants after 28 d at 2°C. Data in B and C are represented as the mean ± se (n = 4). Statistical analysis was performed using a one-way ANOVA with Tukey’s honestly significant difference (HSD) mean-separation test. Different letters above each column indicate significant differences (P < 0.001) between plant lines.
Previous studies of fab1 have shown that mutant plants could recover Fv/Fm and growth when returned to 22°C after as long as 35 d at 2°C (Wu et al., 1997), but lost the ability for recovery after 70 to 80 d at 2°C. We therefore ran a second experiment in which plants of the four lines were exposed to 2°C for 75 d. At the end of this time, both wild-type and TP-DES9* plants had senescent older leaves with green younger leaves at the center of the rosettes (Fig. 5A). The TP-DES9* plants showed less growth and development than the wild type, and Fv/Fm in TP-DES9* plants averaged 0.690 ± 0.028, compared with 0.815 ± 0.005 for the wild type (Fig. 5B). Nevertheless, both TP-DES9* and wild-type plants rapidly resumed growth after being returned to 22°C, and 10 d after transfer plants of both lines had bolted and begun to produce flowers and seeds (Fig. 5C). By contrast, fab1 plants had no visible green tissue, Fv/Fm close to zero, and showed no recovery after return to 22°C. ER-FAT5 plants also appeared dead at the end of the cold treatment, with Fv/Fm <0.1 (Fig. 5, A and B). However most of these plants apparently retained some living tissue because they produced a small number of green leaves and eventually bolted during the 10-d recovery period at 22°C (Fig. 5C).
Damage to and recovery of transgenic plants after 75 d at 2°C. A, Representative plants after 75 d at 2°C. Scale bar = 1 cm. B, Measurements of Fv/Fm for plants after 75 d at 2°C, represented as the mean ± se (n = 4). Statistical analysis was performed using a one-way ANOVA with Tukey’s honestly significant difference (HSD) mean-separation test. Different letters above each column indicate significant differences (P < 0.001) between plant lines. C, Representative plants after 10 d recovery at 22°C.
DISCUSSION
While plants native to temperate and cold climates are able to grow and develop at temperatures close to freezing, many tropical and subtropical plants undergo sharp reductions in photosynthesis rates and suffer tissue damage after short periods at temperatures in the range 0° to 10°C (Wu and Browse, 1995; Allen and Ort, 2001). Chilling-sensitive species include maize, soybean, rice, and other major food crops, and chilling damage is a significant contributor to harvest losses when these crops are grown in temperate regions (Iba, 2002; Allen and Ort, 2001; Thakur et al., 2010). Understanding the mechanism(s) of chilling injury is key to increasing crop yields in the extended ranges where these and other tropical crops are now grown.
Because many chilling-sensitive plants show a similar set of symptoms after low-temperature exposure, it is tempting to consider that a single primary lesion is involved that results in a cascade of secondary events that produce the chilling-sensitive phenotype. An attractive early hypothesis proposed that glycerolipids in the membranes of chilling-sensitive plants with low levels of fatty acid unsaturation would undergo a transition from the fluid, liquid-crystalline phase to the more rigid gel state at chilling temperatures. This phase change would result in disruption of membrane function and lead to cellular and tissue damage (Lyons and Raison, 1973; Raison, 1973). Relevant to this concept, several mutants of Arabidopsis deficient in polyunsaturated fatty acids (PUFAs)—fad5, fad6, fad7 fad8, and fad3 fad7 fad8—all show chlorosis and reduced growth rates relative to the wild type when grown at 4°C (Hugly and Somerville, 1992; Routaboul et al., 2000). These mutants are also more susceptible to photoinhibition (Vijayan and Browse, 2002), a characteristic shared with some chilling-sensitive species (Allen and Ort, 2001). However, these Arabidopsis mutants are nevertheless able to complete their life cycles at chilling temperatures (Hugly and Somerville, 1992; Routaboul et al., 2000), indicating that high PUFA levels are important to photosynthetic function at low temperatures, but are not a major determinant of chilling sensitivity.
One specific model for phase-change effects concerns PG, the major chloroplast phospholipid. PG is identified in the crystal structure of the PSII light-harvesting complex and is essential for chloroplast development and photosynthetic function (Nussberger et al., 1993; Hagio et al., 2002; Liu et al., 2004), two processes that are inhibited by chilling (Liu et al., 2018). The proportion of HMP-PG molecular species has been correlated with the severity of chilling sensitivity (Murata, 1983; Roughan, 1985; Wada and Murata, 2007); typically, chilling-resistant plants contain <10% HMP-PG, whereas many chilling-sensitive plants have >30% HMP-PG. Results supporting this model include the detection of liquid-crystalline-to-gel phase transitions in PG isolated from some chilling-sensitive plants (Murata and Yamaya, 1984). Transgenic approaches that alter HMP-PG levels also support their role in contributing to the severity of plant chilling damage (Murata et al., 1992; Wolter et al., 1992; Moon et al., 1995; Ishizaki-Nishizawa et al., 1996).
The fab1 mutation causes a Leu-337-Phe substitution in KASII, the 3-ketoacyl-ACP synthase responsible for elongation of 16:0-ACP during fatty acid synthesis (Carlsson et al., 2002). This results in a 40% reduction in KASII activity and increased proportions of 16:0 in all major leaf glycerolipids of fab1 plants (Wu et al., 1994). In particular, HMP-PG is increased to 40% to 50% of the total PG, equivalent to values typical of many chilling-sensitive plants (Murata, 1983; Roughan, 1985; Wu and Browse, 1995). Nevertheless, fab1 plants survived a range of chilling treatments that quickly damaged and killed chilling-sensitive plants (Wu and Browse, 1995). Furthermore, the fab1 mutant is not more susceptible than wild-type Arabidopsis to photoinhibition during short-term experiments under high light (1,200 μmol quanta m−2 s−1 PAR) and low temperature (2°C). fab1 plants only begin to show a decline in photosynthetic capacity (as indicated by the fluorescence parameter Fv/Fm) starting 10 d after transfer to 2°C. The mutant plants will eventually die at 2°C, but they succumb only after >2 months (Figs. 3 and 4; Barkan et al., 2006); before this, they retain a substantial capacity for recovery (Wu et al., 1997).
These results clearly indicate that HMP-PG cannot be the only cause of damage in chilling-sensitive plants, because fab1 plants only die after prolonged exposure to 2°C. Instead, our results, together with those of other researchers (Wu and Browse, 1995; Jones et al., 1998; Thakur et al., 2010), suggest an alternative theory of chilling sensitivity and resistance. During plant evolution in consistently warm tropical and subtropical habitats, there would have been no selection against traits that compromise growth at low temperatures. Many such traits may have evolved, especially if they confer even a small selective advantage at higher temperatures. Some of these evolved character traits might affect plant performance only after extended cold treatment, whereas others might result in damage on the much shorter time scale normally associated with chilling sensitivity. The progressive dispersa1 of angiosperms to colder regions would have required the elimination of traits that compromised plant performance in these cooler environments. Under this proposal, any particular chilling-sensitive species is likely to possess severa1 (or many) traits that restrict its geographical range. This concept accommodates the possibility of other mechanisms driving chilling sensitivity, such as disruptions to diurnal gene expression, redox processes, or metabolic pathways (Jones et al., 1998; Allen and Ort 2001; Baier et al., 2019), and it is supported by the data from plant breeding experiments indicating that improvements in chilling tolerance are typically multigenic and may be specific to particular stages in the plant life cycle (Venema et al., 2005; Thakur et al., 2010).
Although evidence from other studies supports a role for HMP-PG in causing the damage and death of fab1 plants at 2°C, it has been difficult to evaluate the contribution of HMP-PG relative to other changes in the fatty acid composition of membrane glycerolipids. For example, all the glycerolipids in the fab1 mutant have decreased PUFAs relative to the wild type (Wu et al., 1994). We have previously isolated and characterized a series of suppressor lines in the fab1 background, screened on the basis of plant survival at 2°C (Barkan, et al., 2006). Most, though not all, of these suppressors showed substantial reduction in HMP-PG relative to the fab1 parent (Barkan et al., 2006; Kim et al., 2010; Gao et al., 2015). However, all the suppressor mutations are in enzymes of the prokaryotic pathway, and all show changes in glycerolipid content and/or membrane unsaturation (Gao et al., 2015) that complicate comparisons with fab1 and any simple correlation between the reduced HMP-PG content and plant survival at 2°C. Thus, while the suppressors have provided important insights into lipid synthesis by the prokaryotic pathway and the relationships between lipid metabolism and plant function (Barkan et al., 2006; Kim et al., 2010; Gao et al., 2015), more evidence is required to establish the role of HMP-PG in producing the fab1 phenotype.
In this study, we used four constructs targeting two 16:0 desaturases to either the ER or chloroplast to produce transgenic fab1 lines. The C. elegans 16:0-CoA desaturase targeted to the chloroplast, line TP-FAT5, did not support fatty acid desaturation, while the engineered 16:0-glycerolipid desaturase targeted to the ER, line ER-DES9*, produced extremely high amounts of 16:1c in leaves (Fig. 1), and resulted in transgenic plants that grew poorly at 22°C (Fig. 2). However, expression of either the ER-FAT5 or TP-DES9* desaturase in fab1 led to reduction of 16:0 in leaf tissue to levels close to or a little lower than the wild-type level (Fig. 1). The proportions of other major fatty acids in the leaf profiles were similar in ER-FAT5 and TP-DES9* plants (Fig. 1). We completed a detailed characterization of individual glycerolipids from wild-type, fab1, ER-FAT5, and TP-DES9* plants, both in terms of overall fatty acid composition (Fig. 3) and through lipidomics analysis of the molecular species composition of each lipid (Tables 1 and 2, Supplemental Table S2). It is an advantage to our goal that both transgenes produce similar changes in composition for the major chloroplast and extrachloroplast membrane glycerolipids—PC, PE, MGD, and DGD.
The similarity in lipid composition between ER-FAT5 and TP-DES9* plants evident in these results is somewhat surprising, since the ER-FAT5 desaturase acts on 16:0-CoA on the eukaryotic pathway in the ER, while the TP-DES9* enzyme is expected to act on glycerolipid products of both the prokaryotic and eukaryotic pathways on chloroplast membranes. However, data from analyses of the fad3, fad6, fad7, and other Arabidopsis mutants indicate that transfer of lipids from the ER to the chloroplast is reversible to some extent (Browse et al., 1989, 1993), so products of the TP-DES9* desaturase, like those of the FAD6, FAD7, and FAD8 desaturases, can appear in glycerolipids of the extrachloroplast membranes.
In PG also, the proportions of most fatty acids are similar in ER-FAT5 and TP-DES9* plants. Only small differences are evident in the percentages of 16:1t, 18:0, 18:1, 18:2, and 18:3 fatty acids (Fig. 3). However, ER-FAT5 plants have 39.6% 16:0 in PG, which is close to the 42.4% found in PG of fab1 plants. By contrast, PG from TP-DES9* plants contains only 24.5% 16:0, which is matched by accumulation of 22.4% 16:1c; ER-FAT5 PG only contains 6.2% 16:1c (Fig. 3). These differences translate into a substantial dichotomy in the content of HMP-PG for our four experimental lines (Table 2). Similar to many chilling-sensitive plants, the fab1 mutant and ER-FAT5 line both contain high levels of HMP-PG, while wild-type and TP-DES9* plants contain <10%, which is in the range found in chilling-resistant plants.
The overall similarity in fatty acid composition of chloroplast and extrachloroplast lipids between ER-FAT5 and TP-DES9* plants and the contrasting proportions of HMP-PG provide an excellent opportunity to test the significance of HMP-PG to the damage and death of fab1 plants, and thereby test the relevance of HMP-PG to chilling sensitivity in plant species more broadly. During an initial period of 4 weeks at 2°C, ER-FAT5 plants with high HMP-PG exhibited the same leaf chlorosis and dramatic decline in Fv/Fm seen in fab1 controls (Fig. 4). Plants of the TP-DES9* line with low HMP-PG remained similar in appearance to the wild type (Fig. 4A), although Fv/Fm had declined significantly relative to that of the wild type after 28 d at 2°C (Fig. 4B). When cold treatment was extended out to 75 d, the TP-DES9* plants maintained Fv/Fm at ∼85% of wild-type Fv/Fm, and while the plants were smaller than wild-type plants, they retained healthy green tissue at the center of the rosette (Fig. 5, A and B). Importantly, TP-DES9* plants recovered and rapidly resumed growth after being returned to 22°C, and 10 d later were flowering and setting seed (Fig. 5C). As observed previously (Barkan et al., 2006; Gao et al., 2015), fab1 plants died before plants were returned to 22°C; however, ER-FAT5 often showed some limited recovery 10 d after transfer to 22°C (Fig. 5C). It is possible that the modest reduction in HMP-PG brought about by the expression of the ER-FAT5 enzyme in fab1 allows for the limited survival observed in plants of this line.
The lower Fv/Fm in TP-DES9* plants (Figs. 4B and 5B) and smaller size of these plants relative to the wild type (Fig. 5, A and C) suggest that the plants are not fully reverted to wild type. This may be because TP-DES9* plants have substantially reduced levels of PUFAs both in the total-leaf fatty acid profile (Fig. 1) and in individual chloroplast lipids (Fig. 3). For example, in MGD from the wild type, PUFAs are 91.2% of the fatty acids, but this is reduced to 74.9% in TP-DES9* MGD; similarly, PUFAs are reduced from 74.5% to 46.3% in DGD and from 38.5% to 26.4% in PG. Previous studies have shown that reduction in PUFA levels is associated with reduced photosynthetic function and photoinhibition, especially at low temperatures (Kanervo et al., 1997; Routaboul et al., 2000; Allen and Ort, 2001; Vijayan and Browse, 2002). Our results do, however, demonstrate the central role of HMP-PG in the fab1 low-temperature phenotype, because lipids from ER-FAT5 and TP-DES9* plants have similar fatty acid composition except for the very different levels of HMP-PG (Fig. 3; Tables 1 and 2).
The results we report here provide improved clarity to the observation that HMP molecular species account for >30% of PG in many chilling-sensitive species, while no chilling-resistant plants have been- found to have such high levels of HMP-PG (Murata, 1983; Roughan, 1985). Our work has important implications for understanding and ameliorating low-temperature damage in chilling-sensitive species such as rice, maize, and soybean. As discussed above, our findings on the fab1 mutant here and previously (Wu and Browse, 1995; Wu et al., 1997; Vijayan and Browse, 2002; Barkan et al., 2006; Gao et al., 2015) are consistent with a broad range of studies (Jones et al., 1998; Allen and Ort 2001; Venema et al., 2005; Thakur et al., 2010; Baier et al., 2019) that support the notion that chilling-sensitive species have acquired many traits during evolution in warm climates that are incompatible with growth in periodically cold environments. Thus, our results clearly indicate that reducing HMP-PG in cold-sensitive species is a prerequisite to improving long-term performance and survival at low temperatures. However, this alone will probably not provide for robust chilling tolerance, since other traits contribute to the more rapid photoinhibition and tissue damage typically seen in these species after chilling.
Finally, our results support the view that high HMP-PG, while incompatible with long-term growth and survival of plants subject to chilling temperatures in cooler climates, may confer an advantage to plants growing at high temperatures in tropical and subtropical climates. While this is a reasonable hypothesis, we have not been able to detect any measurable improvement in the performance of the fab1 mutant over wild-type Arabidopsis at elevated temperatures (Vijayan and Browse, 2002; Routaboul et al., 2012).
MATERIALS AND METHODS
Plant Material and Growth Conditions
The wild-type control was Arabidopsis (Arabidopsis thaliana) ecotype Columbia-0 (Col-0). The fab1 mutant is in the Col-0 background and was used as the parental line for transformation with constructs expressing the desaturases. Seeds were sown on potting mix, transferred to 4°C for 48 h, then cultivated at 22°C with 16 h of light at 120 to 150 μmol quanta m−2 s−1 PAR. For chilling treatment, plants were transferred to growth chambers at 2°C under continuous light at 120 to 150 μmol quanta m−2 s−1 PAR. After 75 d at 2°C, plants were returned to 22°C for 10 d.
Constructs Encoding the Desaturases and Plant Transformation
DNA sequences encoding the FAT-5 16:0-CoA desaturase from Caenorhabditis elegans (Watts and Browse, 2000) and the DES9*15 16:0-glycerolipid desaturase produced by directed evolution (Bai et al., 2016) were each recloned into a pENTR vector using the TOPO Cloning Kit (Invitrogen) according to the manufacturer’s instructions. After confirmation of the sequences, they were transferred by Gateway cloning into the binary plasmid PB2GW7 under control of the Cauliflower mosaic virus 35S promoter (Karimi et al., 2002). Overlap PCR was used to generate constructs containing the chloroplast TP of the ribulose bisphosphate carboxylase small subunit protein (Roesler, et al., 1997) fused in frame and immediately 5′ to the coding sequence of each desaturase, and these constructs were also cloned into the PB2GW7 vector. Vector constructs were transferred into Agrobacterium tumefaciens GV3102 and transformed into fab1 mutant plants using the floral dip method (Clough and Bent, 1998). Transgenic T1 plants were selected on the basis of BASTA resistance encoded by the vector. For each construct, analysis of leaf fatty acid composition from 5 to 15 individual T1 plants allowed the identification of lines with reductions in 16:0 content, and subsequent pedigree analysis allowed the identification of three independent, single-insert lines for each construct.
Fatty Acid Extraction and Analysis
The overall fatty acid compositions of leaf tissues were determined as previously described (Wu et al., 1994), except that gas chromatographic (GC) analysis was carried out on an Agilent 6890 Series GC using an EC-WAX, 30.0 m × 0.53 μm × 1.20 μm capillary column (www.grace.com). The GC program to separate the fatty acid methyl esters started at 170°C and ramped to 200°C at 5°C min−1 followed by a 1.5°C min−1 ramp to 210°C and a final ramp of 25°C min−1 to 250°C.
Glycerolipid Extraction and Analysis
The more detailed analyses of lipid and fatty acid composition were performed as described previously (Wu et al., 1994). Aliquots of the lipid extract were separated by one-dimensional thin-layer chromatography on (NH4)2SO4-impregnated silica gel G (Wu et al., 1994) using acetone:benzene:water (30:10:2.7 [v/v/v]; Khan and Williams, 1997). Lipids were rendered visible under UV light by exposure to 0.005% (w/v) primulin in 80% acetone, after which the individual lipid classes were collected. To determine the fatty acid composition and the relative amounts of individual lipids, the silica gel from each spot was transferred to a screw-capped tube and fatty acid methyl esters were prepared after addition of a known amount of 17:0 PC as internal standard. The samples were heated for 1 h at 80°C in 1 mL of 2.5% (v/v) H2SO4 in methanol. After cooling to room temperature and addition of 1.5 mL of 0.9% (w/v) NaCl solution and 1 mL of hexane, fatty acids were extracted into the hexane phase by shaking and the tubes were centrifuged at low speed. Samples (1 μL) of the hexane phase were separated by gas chromatography as described above.
Lipidomics Analyses
For lipidomics analysis, lipids from four independent samples were extracted according to the instructions provided by the KS Lipidomics Research Center (http://www.k-state.edu/lipid/lipidomics/), and analyses of individual lipid molecular species were performed at that facility on a triple quadrapole mass spectrometer (Esch et al., 2007). Identification of a molecular ion and a headgroup fragment ion allows identification of individual molecular species of the different glycerolipids. To assess the contribution of 16:1t and 16:1c isomers to the 16:0/16:1 molecular species of PG (molecular ion m/z = 719), the Lipidomics Center conducted constant neutral loss scanning of the tandem mass spectrometry spectra for PG to allow calculation of the ratio of 16:1c to 16:1t (Hsu et al., 2007). Because the Δ3 double bond in 16:1t facilitates loss of this fatty acid as a ketene (mass 236 atomic-mass units), it allows identification of 16:0/16:1t molecules by a diagnostic peak at m/z = 485 (719–235). Since wild-type PG contains <2% 16:1c, comparing the [M-236] for 16:0/16:1-PG signal to the [M-153] signal (that is quantified using an internal standard) gave a normalization ratio (5.3 ± 0.2) that was then used to calculate the approximate proportions of 16:1c and 16:1t in this molecular species from the ER-FAT5 and TP-DES9* samples. Original data from the lipidomics analyses have been archived on the Research Exchange at Washington State University and are available at: https://research.libraries.wsu.edu/xmlui/handle/2376/16922.
Measurements of Chlorophyll Fluorescence
Chlorophyll fluorescence from rosette leaves was analyzed by a Fluorescence Monitoring System (Hansatech). The Fv/Fm, representing the potential quantum yield of PSII, was measured after leaves on intact plants were dark adapted at 22°C for 30 min (Barkan et al., 2006).
Statistical Analyses
For the experiments in comparing the lipid compositions of our lines, we undertook Principal Component Analyses using the prcomp function in R (Venables and Ripley, 2013). This function uses singular value decomposition of the fatty acid compositions of the lines. The quantitative data in Figures 2, 4, and 5 were analyzed by one-way ANOVA with Tukey’s honestly significant difference (HSD) mean-separation test using the criterion of P < 0.001 for statistically significant differences.
Accession Numbers
Sequence data from this article can be found at Arabidopsis.org under accession number AT1G74960 (FAB1), in the Uniprot database under accession number Q4411 (DES9), and in the National Center for Biotechnology Information library under accession number AAF97548 (FAT-5).
Supplemental Data
The following supplemental materials are available.
Supplemental Table S1. Leaf fatty acid composition of T3 plants of fab1 lines expressing each of the four different 16:0-desaturase constructs.
Supplemental Table S2. Molecular species analysis of leaf lipids from wild-type, fab1, ER-FAT5, and TP-DES9* plants.
Supplemental Figure S1. PCA of the overall leaf fatty acid compositions in each line shown in Figure 1.
Supplemental Figure S2. PCA of the fatty acid composition of major leaf lipids in each line shown in Figure 3.
Supplemental Figure S3. The HMP-PG content of each line shown in Table 2.
Supplemental Figure S4. PCA of lipidomic species of each leaf lipid in each line shown in Supplemental Table S2.
Acknowledgments
We thank Dr. John Ohlrogge (Michigan State University) for providing DNA encoding the ribulose bisphosphate carboxylase small subunit targeting peptide and Dr. Shuangyi Bai (Washington State University) for helpful discussions. The lipid analyses described in this work were performed at the Kansas Lipidomics Research Center Analytical Laboratory.
Footnotes
J.G., J.G.W., and J.B. conceived and designed the research; J.G. and D.L. carried out and interpreted all experimental work; and all authors contributed to data analysis and manuscript preparation.
↵1 This work was supported by the National Science Foundation (grant no. IOS–1555581) and the U.S. Department of Agriculture, Prosser Irrigated Agricultural Research Center at Washington State University (Hatch umbrella project no. 1015621). Instrument acquisition and lipidomics method development for this study were supported by the National Science Foundation (grant nos. EPS 0236913, MCB 1413036, MCB 0920663, DBI 0521587, and DBI 1228622), Kansas Technology Enterprise Corporation, K-IDeA Networks of Biomedical Research Excellence of the National Institutes of Health (grant no. P20GM103418), and Kansas State University.
↵[OPEN] Articles can be viewed without a subscription.
- Received September 9, 2020.
- Accepted September 24, 2020.
- Published October 7, 2020.
REFERENCES
