Feedback regulation of plastidic acetyl-CoA carboxylase by 18:1-acyl carrier protein in Brassica napus

Abstract

Plant seed oil represents a major renewable source of reduced carbon, but little is known about the biochemical regulation of its synthesis. The goal of this research was to identify potential feedback regulation of fatty acid biosynthesis in Brassica napus embryo-derived cell cultures and to characterize both the feedback signals and enzymatic targets of the inhibition. Fatty acids delivered via Tween esters rapidly reduced the rate of fatty acid synthesis in a dose-dependent and reversible manner, demonstrating the existence of feedback inhibition in an oil-accumulating tissue. Tween feeding did not affect fatty acid elongation in the cytosol or the incorporation of radiolabeled malonate into nascent fatty acids, which together pinpoint plastidic acetyl-CoA carboxylase (ACCase) as the enzymatic target of feedback inhibition. To identify the signal responsible for feedback, a variety of Tween esters were tested for their effects on the rate of fatty acid synthesis. Maximum inhibition was achieved upon feeding oleic acid (18:1) Tween esters that resulted in the intracellular accumulation of 18:1 free fatty acid, 18:1-CoA, and 18:1-acyl-carrier protein (ACP). Direct, saturable inhibition of ACCase enzyme activity was observed in culture extracts and in extracts of developing canola seeds supplemented with 18:1-ACP at physiological concentrations. A mechanism for feedback inhibition is proposed in which reduced demand for de novo fatty acids results in the accumulation of 18:1-ACP, which directly inhibits plastidic ACCase, leading to reduced fatty acid synthesis. Defining this mechanism presents an opportunity for mitigating feedback inhibition of fatty acid synthesis in crop plants to increase oil yield.

Feedback inhibition of biosynthetic pathways occurs when a downstream metabolite accumulates and causes inhibition of an enzyme controlling an early step in its synthesis. Plant fatty acid biosynthesis is feedback-regulated by an unknown mechanism (1–3) and a basic understanding of this process is vital for the successful engineering of this pathway to increase oil yield. A variety of mechanisms contribute to feedback inhibition of fatty acid synthesis in animals, fungi, and bacteria. In rats and yeast, palmitoyl-CoA, an intermediate of fatty acid synthesis, binds to and inhibits acetyl-CoA carboxylase (ACCase) (4). Also in yeast, ACCase and fatty acid synthase (FAS) gene expression are repressed in an acyl-CoA–dependent manner (5). In Escherichia coli, ACCase and the β-ketoacyl acyl carrier protein (ACP) synthase subunit of FAS are inhibited by long-chain (C16-C18) acyl-ACP, an intermediate of fatty acid synthesis (6, 7). Repression of bacterial fatty acid biosynthetic genes (including ACCase) by interaction of long-chain acyl-ACP or acyl-CoA with transcription factors has also been observed (8). Thus, a picture has emerged in which decreased demand for de novo fatty acids is signaled by a variety of signals, including acyl-ACP and acyl-CoA. These metabolites allosterically inhibit fatty acid biosynthetic enzymes and can thereby rapidly decrease the production of malonyl-CoA for use in fatty acid synthesis. When the levels of acyl-ACP and acyl-CoA are not sufficiently reduced, gene expression for the entire biosynthetic pathway can also be repressed.

Tween esters are effective agents for the delivery of fatty acids and have been shown to elicit feedback inhibition in various plant cell culture systems (1–3). Shintani and Ohlrogge (2) hypothesized that feedback possibly occurs through biochemical or posttranslational modification of ACCase and FAS. Purified monomeric ACCases from maize and diatom were inhibited by palmitoyl-CoA (9, 10), but long-chain (C16-C18) acyl-ACPs failed to inhibit partially purified ACCases from castor and pea (11). Decanoyl-ACP did, however, inhibit ketoacyl-ACP synthase activity in crude extracts of canola and spinach (12). The relevance of these results to feedback inhibition in oil-accumulating tissues is unclear because changes in the steady-state pools of acyl-CoA or acyl-ACP during feedback were not measured. The situation is further complicated in dicotyledonous plants because of the presence of structurally distinct heteromeric and monomeric ACCase and FAS systems in the plastid and cytosol, responsible for fatty acid synthesis and elongation, respectively. Whether or not the cytosolic fatty acid elongation pathway is subject to feedback inhibition is unknown. It is also not known if feedback inhibition occurs in oil-seed tissues, which have higher rates of fatty acid synthesis than in vegetative tissues.

To characterize feedback inhibition of fatty acid synthesis in oil seeds and to elucidate its molecular mechanism, we have used a Brassica napus cv Jet Neuf cell-suspension culture created from microspore-derived embryos. This culture has been used for studies on the synthesis of storage lipids and for purification of native enzymes (13, 14). In contrast to soybean and tobacco cell cultures (1, 2), oil (triacylglycerol or TAG) is present, which contains very-long-chain fatty acids generally found only in seed tissues. With this system we have characterized feedback inhibition in oil-accumulating tissue, and have systematically dissected the inhibition of fatty acid synthesis and present a model to explain its mechanism.

Results

Plastidic ACCase Is Reversibly Inhibited in Response to Tween 80.

Tween 80, containing predominately oleic acid (18:1) (Table 1), was tested for its effects on the rate of fatty acid synthesis in the B. napus suspension cells described above. The rate of fatty acid synthesis was determined by in vivo labeling of cells with ¹⁴C-acetate for 15 min, which fell within the linear range of ¹⁴C incorporation (Fig. S1A). Fig. 1A shows that as soon as 3 h after the addition of 10 mM Tween 80 to the medium, the rate of ¹⁴C-acetate incorporation was reduced by 40%, indicating feedback inhibition of fatty acid synthesis. The degree of inhibition was found to be dose-dependent on the concentration of Tween 80 in the medium (Fig. 1B) and was completely reversible after its removal (Fig. 1C). These results are consistent with a biochemical mode of action for feedback inhibition. To focus on the early events leading to biochemical feedback, all subsequent experiments in this study used 3-h Tween feedings.

Table 1.

Inhibition of ¹⁴C-acetate labeling of lipids by various Tweens

	Composition of Tween (mol %)
	14:0	16:0	16:1	18:0	18:1	Relative 14C-acetate incorporation (% control)
Tween 40	2.3	94.0	ND	3.7	ND	84.4 ± 3.4
Tween 60	2.8	45.4	ND	51.8	ND	85.9 ± 3.4
Tween 80	1.9	7.3	2.4	2.1	86.4	50.9 ± 7.0
Tween 85	2.7	5.4	6.7	2.2	83.0	48.8 ± 5.9
Tween 18:1	ND	ND	ND	ND	100	51.5 ± 6.7

Values are the mean of three repeats ± SD where included. Fatty acid composition was determined by preparing FAMEs of various commercial and custom (Tween 18:1) Tweens.¹⁴C-acetate labeling is relative to a control sample without any Tweens. All values for relative ¹⁴C-incorporation are significantly different from the control as determined by Student’s t test (all P < 0.05). ND, not detected.

Fig. 1.

Inhibition of ¹⁴C-acetate labeling of lipids by B. napus cells in the presence of Tween 80. (A) Incorporation of ¹⁴C-acetate into total lipids by intact cells in the presence or absence of 10 mM Tween 80. (B) ¹⁴C-actetate labeling of total lipids by intact cells after 3 h of exposure to various concentrations of Tween 80. *P < 0.05 determined by Dunnett’s test. (C) ¹⁴C-actetate labeling of total lipids after removal of Tween 80 following a 3-h exposure to 10 mM Tween 80. All data are the mean ± SD (n = 3).

Plants contain ACCase and FAS enzyme systems in both the plastid and cytosol, which are capable of incorporating ¹⁴C-acetate into fatty acids. Comparing the distribution of label in individual fatty acid species can therefore provide information on the relative contribution of these pathways. Tween 80 feeding, although reduces the rate of incorporation of ¹⁴C-acetate into all fatty acids (Fig. 2A), had no effect on the relative distribution of ¹⁴C-label (Fig. 2B). This finding is consistent with reduced de novo fatty acid synthesis inside the plastid rather than reduced cytosolic elongation, which would have resulted in a lower proportion of label in very-long-chain fatty acids (C20-24). ¹⁴C-acetate labeling was also conducted in the presence of haloxyfop, an inhibitor of the cytosolic ACCase involved in fatty acid elongation (Fig. S2). The degree to which ¹⁴C-acetate incorporation is inhibited by haloxyfop is proportional to cytosolic elongation activity, whereas haloxyfop-resistant incorporation is from the de novo pathway within the plastid. Fig. 2C shows that haloxyfop inhibited ¹⁴C-acetate incorporation by the same amount in cultures with or without Tween 80 (i.e., the inhibition curves parallel one another). However, haloxyfop-resistant incorporation (represented by the area below the curves) was reduced by approximately half in cultures with 10 mM Tween 80, showing that Tween 80 feeding specifically causes inhibition of de novo fatty acid synthesis in the plastid. Fatty acid synthesis requires acetyl-CoA and ATP, both of which are also used for sterol synthesis. Therefore, if fatty acid synthesis is inhibited because of substrate limitation, this should also be observed for sterol biosynthesis. After 3 h of Tween 80 feeding, ¹⁴C-acetate incorporation into sterols was 102 ± 4.4% relative to a control, but incorporation into free fatty acids (FFA) was 55.8 ± 5.0%, confirming that the observed inhibition was specific to fatty acid synthesis.

Fig. 2.

Specific inhibition of plastidic ACCase in B. napus cells after 3 h of Tween 80 feeding. (A) Rate of ¹⁴C-acetate incorporation into fatty acids after 3 h of Tween 80 feeding. (B) Relative distribution of ¹⁴C-acetate in fatty acids after 3 h of 10 mM Tween 80 feeding. (C) ¹⁴C-acetate incorporation into total lipids by cells fed various concentrations of haloxyfop for 30 min after 3 h of 10 mM Tween 80 feeding. (D) Incorporation of label from ¹⁴C-malonate and ¹⁴C-acetate into 16:0 fatty acid after 3 h of 10 mM Tween 80 feeding. All data are the mean ± SD (n = 3). *P < 0.05 determined by Student’s t test.

The data indicate that ACCase or FAS in the plastid is inhibited upon the addition of Tween 80. In an attempt to distinguish between these possibilities, we exploited the fact that exogenous malonate can be used by FAS, bypassing the ACCase reaction (15). If ACCase is the only point of inhibition, then the rate of ¹⁴C-malonate labeling of de novo fatty acids should be the same in cultures with or without Tween 80. Cells were labeled with ¹⁴C-malonate for 15 min (same as for ¹⁴C-acetate), which falls within the linear range of ¹⁴C-incorporation into lipids (Fig. S1B). The incorporation of ¹⁴C-malonate into nascent fatty acids (e.g., 16:0 in Fig. 2D and a full panel of C16-C24 fatty acids in Fig. S1C) was not inhibited by Tween 80, in contrast to ¹⁴C-acetate controls. The malonate incorporation data supports the hypothesis that plastidic ACCase is the target for Tween 80 induced feedback inhibition.

Signal of 18:1-ACP Mediates Feedback Inhibition of Plastidic ACCase.

Commercial Tween 80 contains a mixture of fatty acids. To dissect the effects of the individual components, a variety of Tween esters were tested for their effect on fatty acid synthesis. The compositions of individual Tweens are listed in Table 1, along with results from ¹⁴C-acetate labeling experiments. Tween 40 and Tween 60 containing only saturated fatty acids did not inhibit acetate labeling of lipids to the same extent as Tween 80 or 85, containing primarily 18:1. Custom synthesized Tween 18:1 also produced maximum inhibition. Fatty acids from Tween are activated by esterification to free CoA or ACP before being deposited into cellular lipids. Both acyl-CoA and acyl-ACP are inhibitors of ACCase in other organisms, so steady-state pools of these metabolites as well as FFA were measured in cells fed Tween 80. After 3 h of feeding, 18:1 FFA appeared where there was none detected in untreated cells (Fig. 3A). Similarly, both 18:1-ACP and 18:1-CoA doubled in amount upon Tween 80 feeding but most other molecular species went down (Fig. 3 B and C). In a separate experiment designed to distinguish the incorporation of fatty acids from Tween from secondary effects of feedback inhibition, FFA and acyl-ACPs were analyzed in cells fed ¹³C-18:1-Tween. The results showed that at 3 h, 99.1 ± 2.2% of 18:1 FFA and 46.2 ± 4.2% of 18:1-ACP contained ¹³C-oleic acid (Fig. 3), meaning that the increases observed in these metabolites result directly from the uptake of fatty acids from Tween.

Fig. 3.

Quantification of lipid intermediates in B. napus cells after 3 h of Tween 80 feeding. (A) FFA, (B) acyl-ACP, and (C) acyl-CoA content in cells after 3 h of 10 mM Tween 80 feeding. Where present, numbers represent the percent of 18:1 that was ¹³C-labeled after 3 h of 10 mM ¹³C-oleoyl-Tween feeding. All data are the mean ± SD (n = 3). *P < 0.05 determined by Student’s t test.

Because in other systems 18:1-ACP and 18:1-CoA have been shown to act as allosteric regulators (4, 6, 7), we tested their direct effects on ACCase specific activity. First, the in vivo concentrations of these intermediates were calculated based on the values in Fig. 3 and the water content (fresh weight/dry weight, FW/DW) of the cells. The intracellular concentration of 18:1 FFA was estimated to be 0–100 μM. The estimated concentration of 18:1-ACP was 0.6–1.2 μM. However, when taken into account that acyl-ACPs are restricted to the plastid, the volume of which in developing embryos was reported to be 5–20% of the cell (16), the estimate increased to 3.0–24 μM. Total cellular 18:1-CoA was estimated to be 1–3 μM and assuming that the cytosol occupies 10–45% of the cell, this value increased to 2.2–30 μM. These values established ranges of concentrations to test in vitro. As crude cell-culture extracts contained significant FATA thioesterase activity (see SI Materials and Methods) care was taken to ensure minimal degradation of 18:1-CoA or 18:1-ACP during enzyme assays. FFA had no effect on ACCase activity. Inhibition of ACCase was observed when CoA or ACP esterified to 18:1, but not saturated fatty acids, were added at 10 μM concentration (Fig. 4A). When included together in an assay, 18:1-CoA and 18:1-ACP did not cause increased inhibition compared with when each was added alone (Fig. S3). Inhibition with 18:1-CoA and 18:1-ACP was then tested over a range of concentrations and was found to be saturable (Fig. 4B). We note that a highly expressed ACP isoform from B. napus seeds (17) was used for all experiments. The specificity of inhibition to 18:1 acyl moieties is consistent with 18:1-containing Tweens causing maximum feedback. We also tested ACCase inhibition in crude extracts prepared from 25-d-after-flowering canola seeds dissected from siliques. Because of its exclusion from the chloroplast, 18:1-CoA was not tested with this tissue. ACCase from developing seeds was inhibited by 18:1-ACP to a similar extent as ACCase from cell cultures (Fig. 4B).

Fig. 4.

Effects of lipid intermediates on ACCase activity in crude B. napus cell extracts and in developing seeds. (A) Effects of 10 μM acyl-CoAs and acyl-ACPs on ACCase activity in crude cell-culture extracts. *P < 0.01 determined by Dunnett’s test. All other P > 0.10. (B) Inhibition of ACCase from cell cultures (cells) or 25-d-after-flowering seeds (seeds) caused by various concentrations of 18:1-CoA or 18:1-ACP. Values are normalized to the control, which is free ACP. Data are the mean ± SE (n = 3).

To preclude transcriptional and posttranslational contributions to Tween 80–dependent feedback inhibition, enzyme assays were performed. ACCase-specific activity was unchanged relative to a control in crude extracts from cells fed Tween 80 for 3 h (Fig. 5A). ACCase can be inactivated by dephosphorylation (18) or by interaction with PII protein in a 2-oxoglutarate–dependent manner (19). ACCase assays following phosphatase treatment of crude extracts or in the presence of 5 mM 2-oxoglutarate revealed no differences between control and Tween 80–fed cultures (Fig. 5A). Treatment with haloxyfop did not cause a significant reduction in specific activity, indicating that assays with crude extracts are largely specific for plastidic ACCase (Fig. 5B). Tween 80, when added directly to in vitro enzyme assays, did not inhibit ACCase (Fig. 5C). Taken together, these results suggest that transcriptional or known posttranslational mechanisms leading to a reduction in plastidic ACCase specific activity do not account for the observed feedback inhibition.

Fig. 5.

ACCase specific activity in crude cell extracts. (A) ACCase activity in control, phosphatase treated (CIP), and 2-oxoglutarate (2-OG) supplemented crude extracts from cells grown with or without Tween 80 for 3 h. (B) Effect of 10 μM haloxyfop on ACCase activity when added to crude extracts. (C) Effect of Tween 80 on ACCase activity when added directly to enzyme assays. Samples from B and C were grown without Tween in the culture medium. Data are the mean ± SD (n = 3). Significance between + and − Tween samples (A), and between controls and haloxyfop (B) or Tween 80 (C) treated samples was calculated using Student’s t-test and all P > 0.05.

Discussion

Data from work presented here identifies 18:1-ACP as the signal molecule and plastidic ACCase as the target for biochemical feedback inhibition of fatty acid biosynthesis in an oleogenic plant tissue. That 18:1-ACP is the signal responsible for rapid and reversible feedback inhibition is supported by the observations that: (i) feeding experiments with 18:1-containing Tween esters cause maximal inhibition; (ii) 18:1-ACP accumulates in vivo during feedback inhibition; (iii) ACCase enzyme activity displays saturable in vitro inhibition upon supplementation of 18:1-ACP at the physiological concentrations directly measured in this study; and (iv) although we demonstrate the in vitro inhibitory effects of both 18:1-ACP and 18:1-CoA on ACCase, the involvement of 18:1-CoA in feedback inhibition is discounted because its measured level in this study is insufficient to cause significant inhibition and because long-chain acyl-CoAs have not been detected in plastids (20). That the plastidic multimeric isoform of ACCase is the specific target of in vivo feedback inhibition is supported by the observations that: (i) feedback inhibition affects de novo fatty acid synthesis in the plastid, and not fatty acid elongation in the cytosol; (ii) experiments showing that haloxyfop, an inhibitor of cytosolic ACCase, causes an additive effect together with 18:1-Tween inhibition; (iii) that acetate, but not malonate, incorporation into fatty acids is inhibited by Tween feeding, thus ruling out FAS as the target of inhibition; (iv) the level of inhibition of ACCase by 18:1-ACP observed in vitro was equivalent to that measured in vivo; and (v) the plastidial homomeric isoform of ACCase is not expressed in oilseed embryos (21). Moreover, we extended our analysis to developing canola seeds and also observed inhibition of ACCase by 18:1-ACP. Based on these results, we propose a mechanism in which the concentration of 18:1-ACP controls plant fatty acid synthesis through its biochemical inhibition of plastidic ACCase (Fig. 6). This proposed biochemical mechanism is also consistent with the observations that the levels of enzymatic activities measured in extracts from inhibited cultures were equivalent to untreated controls, demonstrating that transcriptional and posttranslational mechanisms are not responsible for the observed feedback inhibition. Inhibition of the multimeric plant ACCase by 18:1-ACP is consistent with the biochemical feedback mechanism of E. coli (6) and the bacterial lineage of plastidic ACCase.

Fig. 6.

Model for proposed mechanism of feedback inhibition of fatty acid synthesis. Plastidic ACCase is inhibited by 18:1-ACP. This metabolite is a product of de novo fatty acid synthesis, or can be synthesized from exogenous fatty acids provided by Tween 18:1. Reactions that can produce or consume, and therefore participate in the regulation of 18:1-ACP concentration, are indicated with arrows.

A previous study of feedback inhibition in tobacco vegetative tissue reported the rate of production of intermediates of fatty acid synthesis, but not their steady-state pool size (2). Reduced synthesis of long-chain ACPs upon Tween feeding led the authors to conclude that they were unlikely to be involved in feedback inhibition. We measured steady-state levels of metabolites and found that FFAs, 18:1-ACP, and 18:1-CoA concentrations actually increased as a result of incorporation of fatty acids from Tween. The FFAs that accumulated were exclusively from Tween 80 and are likely the result of its hydrolysis upon entering the cells. A plastid localized acyl-ACP synthetase can esterify exogenous fatty acids to ACP (22). Although the amount of acyl-ACP reported in this work is higher than that of spinach leaves, the composition of individual molecular species is similar (23). This quantitative difference may be attributable to the embryo-like identity and associated higher rate of fatty acid synthesis of the cells used here. In addition, long-chain (C16-C18) acyl-ACP concentration in E. coli is as high as 63 μM (7) and the tissue used here is estimated to have 7.5–60 μM. Therefore, the acyl-ACP concentrations reported here fall within the range of previously published values. Total acyl-ACP concentration was reduced by feedback inhibition, which is consistent with the previously reported reduced rate of acyl-ACP synthesis (2). This finding is likely the result of limited substrate availability for the synthesis of acyl-ACPs resulting from reduced ACCase activity. The biochemical mechanism of ACCase inhibition by acyl-ACP is unknown, even for E. coli ACCase (6). Inhibition of ACCase by acyl-ACP only occurred when (i) species specific ACP isoforms are used and (ii) DTT, which is an activator of ACCase (23), is included in the assay buffer (6, 11, and present study). Moreover, the apparent IC₅₀ values for 18:1-CoA and 18:1-ACP were similar and we did not observe increased inhibition by the addition of both 18:1-ACP and 18:1-CoA, implying a common mode of action that is dependent on both the identity of the fatty acid and the presence of a thioester bond.

The reduced rate of fatty acid synthesis during Tween feeding could result from a shortage of free ACP and CoA because of thioesterification of exogenous fatty acids. If this were the case, reduced incorporation of both ¹⁴C-acetate and ¹⁴C-malonate into fatty acids would be observed. However, only the incorporation of ¹⁴C-acetate was reduced. ATP is required by ACCase and its availability could influence enzyme activity. Exogenous long-chain acyl-CoAs have been shown to reduce fatty acid synthesis in isolated plastids by inhibition of ATP import (24). However, for the experiments presented here, the combined amount of long-chain acyl-CoAs was unaffected by Tween feeding, making it unlikely that ATP import was inhibited, as well as supporting the notion that CoA was not limiting. In addition, sterol biosynthesis and growth rate of the cells, both of which would be adversely affected by limited ATP supply, were unaffected by Tween feeding. Plastidic ACCase is regulated by a variety of factors in vivo (25) and the identification of plastidic ACCase as the target of feedback inhibition is consistent with its regulatory role in chloroplast fatty acid synthesis (26). Inhibition of FAS would be predicted to result in the accumulation of malonyl-CoA. In the absence of FAS activity, malonyl-CoA would be a dead-end product in the plastid, and therefore inhibition of ACCase can be considered to be more efficient than that of FAS. In the cytosol, malonyl-CoA is required for flavonoid biosynthesis and loss of cytosolic ACCase activity is lethal (27). This function of cytosolic ACCase may explain its evident immunity to the effects of feedback.

Our observations of the seed-like cell line used herein demonstrates that feedback inhibition of ACCase can occur in tissues where fatty acid synthesis is a primary metabolic function and may explain why overexpression of ACCase resulted in marginal increases in fatty acid production in seeds (28). Paradoxically, it also implies that oil seeds may have evolved means of reducing feedback to achieve high oil content. The metabolic predisposition of seeds to synthesize TAG could explain why higher concentrations of Tween 80 were needed to induce feedback in B. napus cells than was required for tobacco (2). Feedback inhibition is relieved in E. coli by overexpression of a thioesterase (29) and in B. napus indirect reduction in 18:1-ACP by overexpression of a medium-chain thioesterase resulted in higher rates of fatty acid synthesis (30). Conversely, in Arabidopsis reduction of FATA thioesterase gene expression causes reduced fatty acid accumulation in developing seeds (31). Feedback inhibition may also explain the shared control of oil accumulation between synthesis and assembly in B. napus (3). Diacylglycerol acyltransferase (DGAT), which consumes acyl-CoA in the cytosol, exerts control over oil accumulation (32) and overexpression of this enzyme in Arabidopsis results in enhanced oil content (33). Increased fatty acid synthesis is a logical prerequisite for elevated oil content. Conversely, the Arabidopsis asi1(tag1) mutant deficient in DGAT has less oil than wild-type and reduced incorporation of ¹⁴C-acetate into lipids, indicating reduced fatty acid synthesis (34). Premised on these results and the present study, DGAT might exert control over oil accumulation by consuming acyl-CoA in the cytosol, thus driving vectorial export of de novo fatty acids from the plastid and thus attenuating feedback inhibition.

This study was designed to address early events in biochemical feedback inhibition. However, feedback is persistent during prolonged Tween feeding, and by analogy with other systems may involve several distinct mechanisms capable of rapid and then persistent response to oversupply of fatty acids. Nevertheless, the mechanism of biochemical feedback inhibition presented here provides a potential target for designing strategies for increasing fatty acid synthesis in plants. Precedent for a strategy involving defeating the biochemical inhibition of a single regulatory enzyme comes from the expression of a feedback-insensitive ADP-Glc pyrophosphorylase in potatoes, which resulted in increased starch content (35).

Materials and Methods

Cell Culture Growth.

B. napus cv Jet Neuf suspension cell cultures were grown in NLN medium, as previously described (36). Explicit details can be found in SI Materials and Methods.

Lipid Extraction and Quantification.

Total lipids were extracted from ∼100 mg FW of frozen cells by homogenizing twice in 500 μL of methanol:chloroform:formic acid (20:10:1 vol/vol). The organic solvent was extracted with 500 μL of 1 M KCl, 0.2 M H₃PO₄. The organic phase was dried under N₂ and suspended in hexane. Lipids were separated by TLC with hexane:diethylether:acetic acid (80:20:1, vol/vol) or with acetone:toluene:water (91:30:7 vol/vol) and 0.15 M (NH₄)₂SO₄ impregnated plates. Fatty acid methyl esters (FAMEs) were prepared with 1 mL of 12% (wt/wt) BCl₃ in methanol for 1 h at 85 °C, and were extracted with 1 mL of water and 2 mL hexane. FAMEs were quantified by GC-MS, as previously described (37). ¹³C-fatty acids were detected and quantified as previously described (38). Heptadecanoic acid was routinely used for internal standard.

¹⁴C and ¹³C Labeling.

All radioisotopes were obtained from American Radiolabeled Chemicals. Labeling was done on cells 5 d after subculturing with a cell density of ∼20 mg FW mL⁻¹ with medium refreshed 16 h prior. Typically, 1-mL aliquots of cells were labeled with either 0.2 μCi of 1,2-¹⁴C-acetate (50-60 mCi⋅mmol⁻¹) or 2-¹⁴C-malonate (40-60 mCi⋅mmol⁻¹) for 15 min at 25 °C with shaking. Haloxyfop (Sigma) was dissolved in DMSO and added to cultures 30 min before labeling. Lipids were extracted and separated by TLC as described above. Radioactivity was detected and quantified by phosphorimaging. Incorporation of label into fatty acids species was determined as previously described (22). The 1-¹³C-oleoyl-Tween synthesis is described in SI Materials and Methods.

FFA, Acyl-CoA, and Acyl-ACP Extraction.

FFAs were extracted from tissue by quenching ∼300 mg of frozen cells in 2 mL boiling isopropanol for 5 min. Once cooled, 2 mL of 0.9% NaCl was added and lipids were extracted twice with 4 mL of hexane. FFA were separated by neutral lipid TLC, made into FAMEs, and analyzed as described above. Acyl-CoAs were extracted from ∼15 mg FW of cells and quantified as previously described (39). Acyl-ACPs were extracted and quantified as previously described (23), with modifications: (i) TCA precipitated proteins were dissolved with three extractions of Mops buffer, and (ii) internal standards used were 11:0-CoA (Sigma) and 17:0-ACP made from spinach ACP as described in SI Materials and Methods.

Enzyme Assays.

ACCase activity was measured as the acetyl-CoA dependent incorporation of ¹⁴C-NaHCO₃ into acid-stable products as previously described (40). Details of ACCase assays, acyl-ACP synthesis, and thioesterase assays can be found in SI Materials and Methods.