INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Acyl carrier protein (ACP) is an important cofactor for de novo fatty acid biosynthesis in all living organisms (Ohlrogge, 1987). It is a small (9 kD) acidic protein with a conserved Ser residue that attaches via a phosphodiester linkage to a 4'-phosphopantetheine prosthetic group. The phosphopantetheine group attaches to fatty acids by way of a thioester linkage. In this manner, ACP is able to carry the acyl chains through the cycles of condensation, reduction, and dehydration steps in fatty acid biosynthesis (Ohlrogge, 1987). Ohlrogge et al. (1979) illustrated the critical role of ACP in de novo fatty acid biosynthesis by demonstrating that antibodies to ACP inhibited fatty acid biosynthesis in spinach (Spinacia oleracea) leaf homogenate. ACP is also a cofactor for stearoyl-ACP desaturase (McKeon and Stumpf, 1982) as well as hydrolase (Ohlrogge et al., 1978; McKeon and Stumpf, 1982) and acyl transferase reactions (Frentzen et al., 1983). In addition to involvement in de novo fatty acid biosynthesis, ACP is a cofactor for other reactions including polyketide synthesis (Revill et al., 1996), transfer of fatty acids to the lipid A portion of lipopolysaccharides of gram-negative bacterial membranes (Brozek et al., 1996), rhizobial capsular polysaccharide biosynthesis (Epple et al., 1998), and the production of Nod factors in rhizobia (Ritsema et al., 1998).

All multicellular plants contain several isoforms of ACP that are expressed either constitutively or in a tissue-specific fashion (Battey and Ohlrogge, 1990). Multiple ACP isoforms have been observed in spinach, castor (Ricinus communis), and soybean (Glycine max; Ohlrogge and Kuo, 1985), barley (Hordeum vulgare; Høj and Svendsen, 1984; Hansen and von Wettstein-Knowles, 1991), Cuphea lanceolata (Kopka et al., 1993), rapeseed (Brassica napus; Safford et al., 1988), and Arabidopsis (Hlouek-Radojcic et al., 1992). Although ACP isoforms have been characterized in various plants, little is known about the functional role of these isoforms in plant lipid and fatty acid biosynthesis or why it appears to be necessary to have more than one isoform. One explanation for multiple ACP isoforms could be tissue-specific activity, which is supported by several examples of tissue-specific ACP isoforms (Ohlrogge and Kuo, 1985; Hlouek-Radojcic et al., 1992) and even organelle-specific ACPs (Shintani and Ohlrogge, 1994). For example, an ACP isoform isolated from pea (Pisum sativum) leaf mitochondria was determined to be involved in de novo fatty acid biosynthesis primarily for the production of lipoic acid (Wada et al., 1997). Furthermore, Song and Allen (1997) isolated a cotton (Gossypium hirsutum) fiber-specific ACP that is likely involved in synthesis of membrane lipids during elongation of cotton fibers. Because some tissues express more than one ACP isoform, one could also reasonably hypothesize that ACP isoforms may have distinct patterns of tissue and developmental expression for the purpose of regulating the balance between fatty acid biosynthesis for housekeeping purposes (membrane lipid biosynthesis) and fatty acid biosynthesis for storage as triglycerides (Ohlrogge and Kuo, 1985).

Although tissue specificity and regulatory functions may partially explain the existence of multiple ACP isoforms, there is increasing evidence that there may be a more complex role for ACP isoforms in fatty acid biosynthesis. First, different ACP isoforms may be expressed for utilization in reactions specific to each isoform. For example, in Streptomyces coelicolor, an ACP involved in fatty acid synthesis could not substitute for an ACP isoform involved in polyketide synthesis (Revill et al., 1996). In addition to specificity for pathways other than fatty acid biosynthesis, it is possible that ACP isoforms may show specificity for enzymes or acyl chains within fatty acid biosynthesis, and thus may influence overall fatty acid composition of oilseed or leaf lipids. Several in vitro experiments provide evidence to support this hypothesis. In spinach leaf and achene tissues, the activity of oleoyl-ACP thioesterase demonstrated specificity for ACP-1^spinach, a leaf ACP, over ACP-2^spinach, a predominant seed ACP (Guerra et al., 1986). However, oleoyl-ACP-2^spinach was a better substrate and had a lower K_m than oleoyl-ACP-1^spinach for chloroplast glycerol-3-phosphate acyl transferase. Furthermore, in fatty acid synthase (FAS) reconstitution assays, Schütt et al. (1998) showed the importance of C. lanceolata ACP-2 for the synthesis of medium-chain fatty acids. The addition of C. lanceolata ACP-1 or ACP-2 to spinach leaf FAS resulted in higher ratios of short- and medium-chain fatty acids to long-chain fatty acids. However, in rapeseed FAS extracts, only the addition of ACP-2 resulted in a 3-fold increase in the short/medium chain to long-chain fatty acid ratio. In addition, Suh et al. (1999) showed that a coriander ⁴-acyl-ACP desaturase, which produces ⁴-hexadecenoic acid, preferred acyl-ACPs purified from coriander endosperm over spinach acyl-ACPs. This specificity was also demonstrated with a ⁶-acyl-ACP desaturase from Thunbergia alata. An ACP purified from coriander favored the production of ⁴-hexadecenoic acid and also favored the elongation of ⁴-hexadecenoic acid to petroselinic acid.

It is clear that there is increasing evidence that enzymes involved in fatty acid biosynthesis favor specific ACPs; thus, ACP isoforms may play a role in determining the fatty acid composition of plant oils. However, as of yet there is a lack of published in vivo evidence to confirm this role for ACP isoforms. The major objective of this research was to determine the influence of ACP isoforms in determining the fatty acid content and composition of oilseed lipids in vivo. In this paper, we report in vivo evidence that ACP plays a role in determining the fatty acid composition of plant lipids.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Overexpression of ACP-1 in Leaf Tissue

The ACP-1 isoform is normally highly expressed in Arabidopsis seed tissue, yet almost absent (not detectable by western-blot analysis) in leaf tissue, where the prominent ACP isoform referred to as the "leaf major isoform" (LMI) as well as ACP-2 and ACP-3 are normally dominant. We transformed Arabidopsis plants with a vector containing the ACP-1 genomic clone (including 400+ bp upstream of the transcription start site) in antisense conformation and driven by the cauliflower mosaic virus (CaMV) 35S promoter (Fig. 1). Twenty independent primary (T2) transgenic plants were generated. Western-blot analysis of the above-described transgenic plants indicated that ACP-1 was overexpressed 3- to 8-fold in leaf tissue compared with both wild type (WT) and transgenic controls (Fig. 2). This phenotype was observed in all transgenic plants produced from the number 6 construct. Because this construct contained the constitutive CaMV 35S promoter, seed tissue of these transgenic plants was also analyzed by western blot, indicating slight, if any, overexpression of ACP-1 in the seed (data not shown). Northern analysis of whole-leaf mRNA from transgenic plants confirmed overexpression of ACP-1 in leaf tissue, demonstrating increased ACP-1 mRNA levels compared with WT and transgenic control plants (Fig. 3).

View larger version (10K): [in this window] [in a new window]   Figure 1.   Design of the binary transformation vector number 6 in pGA748. BL, Left border of Ti plasmid; NPTII, neomycin phosphotransferase gene; CaMV 35S; ACP-1, SalI/BamHI fragment of the Arabidopsis ACP-1 genomic clone (pAD4) in 3'-5' (antisense) conformation; 400+-bp region, the region upstream from transcription start-site of ACP-1; BR, right border of Ti plasmid.

View larger version (76K): [in this window] [in a new window]   Figure 2.   Western blot of ACP proteins in Arabidopsis whole leaf tissue. Lane 1, WT; lanes 2 and 3, transgenic control pGA748; lanes 4 through 11, independent transgenic number 6 plants; lane 12, ACP-1 protein; lane 13, ACP-2 protein. LMI, Leaf major isoform.

View larger version (15K): [in this window] [in a new window]   Figure 3.   Percentages (based on WT amount) of ACP-1 sense mRNA in whole leaf RNA. PS, Transgenic control; 6c, 6d, and 6g, independent transgenic number 6 plants. Values were normalized using -actin as a control for equal loading.

These results were completely unexpected because this construct was designed for antisense inhibition of expression of ACP-1. Based on literature review, we believe that this result is due to a region of the CaMV 35S promoter that may act as an enhancer element; driving expression of nearby genes by influencing promoters within specific distances, either up- or downstream of the CaMV 35S promoter element (Odell et al., 1988). To support the hypothesis that the CaMV 35S promoter was enhancing the 400+-bp promoter region of the ACP-1 gene and driving expression in sense orientation, we designed constructs containing this ACP-1 clone without the CaMV 35S promoter and also without the 400+-bp region of the promoter. At least 10 transgenic plants from each of these constructs were produced and analyzed. Our hypothesis was supported by western analysis of these additional transgenic plants indicating normal levels of expression of ACP-1 in leaf tissue (data not shown).

Seeds from primary transgenic plants were screened on kanamycin media. They exhibited a Mendelian segregation pattern. Healthy green plants with roots were transplanted to soil; after 4 to 5 weeks, leaves were collected for western analysis. Western analysis of T3 plants demonstrated the same phenotype as T2 plants, indicating that ACP-1 overexpression is stable. Although we are still in the process of generating homozygous plants from construct number 6, one independent homozygous T3 plant has been generated; it displays the ACP-1 overexpression phenotype that has been observed in previous generations.

ACP-1 Overexpression Affects Leaf Fatty Acid Composition

An exciting aspect of this transformation is that it allowed us to observe the effects of overexpression of an ACP isoform on leaf fatty acid composition. Leaf fatty acid content and composition of T2 and T3 transgenic plants was analyzed by gas chromatography. When leaf fatty acid composition from 13 of 20 independent T2 transgenic plants (seven of the original T2 plants were not analyzed due to desire to look at further generations) was pooled and compared with the leaf fatty acid composition of six WT plants, small but significant alterations in leaf fatty acid composition were apparent. Significant (P < 0.05) increases in the levels of 18:3 as well as decreases in 18:1, 18:2, and 16:3 were noted in transgenic plants containing increased ACP-1 when compared with WT plants. There was no significant change in the total leaf lipid content. In addition, there were no other visible phenotypical differences between transgenic and WT plants. Mature seeds were also analyzed, but no significant alterations in lipid content or fatty acid composition were observed.

Leaf fatty acid analysis was also carried out on T3 transgenic plants. As stated above, the phenotype of plants exhibiting overexpression of ACP-1 carried through to the second generation, so we expected the fatty acid phenotype to carry through as well. From this analysis, we found six out of 20 plants for which the fatty acid phenotype was seen in the second generation (Table I). Although there was some variation from plant to plant in leaf fatty acid composition, the transgenic plants demonstrated significantly higher levels of 18:3 and lower 16:3 compared with WT plants grown at the same time indicating that in some plants it is a stable phenotype. The fatty acid composition of transgenic controls did not differ from that of WT plants.

To investigate whether the overexpression of ACP-1 in leaf tissue resulted in alterations of leaf lipid components, polar lipid classes were separated and quantitated by HPLC. Preliminary analysis of monogalactosyldiacylglyceride (MGDG), digalactosyldiacylglyceride, phosphatidylcholine, phosphatidylglycerol, and phosphatidylethanolamine did not reveal any differences in relative amounts of these polar lipid classes (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

The objective of our research is to determine the role of multiple ACP isoforms in plant fatty acid biosynthesis. Immunoblot analysis of Arabidopsis tissues indicates that this plant contains at least five isoforms of ACP (Hlouek-Radojcic et al., 1992). Shintani and Ohlrogge (1994) have also characterized a mitochondrial ACP from Arabidopsis. ACP-1, ACP-2, and ACP-3 are considered to be constitutive ACPs, meaning they are expressed in all tissues, although not necessarily in equimolar amounts. A fourth isoform is the major isoform found in leaves referred to as the LMI, whereas a fifth isoform appears to be restricted to seeds (Hlouek-Radojcic et al., 1992). ACP-1 is more highly expressed in seed tissue than in the leaf or root. In this study, we have generated transgenic Arabidopsis plants that display 3- to 8-fold overexpression of ACP-1 protein in leaf tissue, where normally there is negligible expression of this isoform. Northern-blot analysis of transgenic leaf tissue confirmed that ACP-1 mRNA levels were also increased. There was no overexpression of ACP-1 protein in transgenic plants where either the 400+-bp region upstream of the ACP-1 transcription start site or the CaMV 35S promoter were deleted from the construct. Thus, overexpression of ACP-1 is most likely due to an interaction between enhancer elements in the CaMV 35S promoter and the 400+-bp promoter region of ACP-1, driving transcription in the sense orientation.

Because ACP-1 expression is normally low in Arabidopsis leaf tissue and high in the seed, overexpression of ACP-1 in leaves of number 6 transgenic plants indicates that in designing the number 6 construct a region of the promoter responsible for seed-specific expression of ACP-1 may have been eliminated. Several cis-acting enhancer and/or repressor elements have been localized in the promoter region of Arabidopsis ACP-2 and ACP-3, and these appear to control tissue-specific as well as developmental regulation of expression of these proteins (Baerson et al., 1994). We currently are conducting a functional analysis of the 5' region of ACP-1 upstream from the transcription start site via GUS fusion assays to get further insight into the key elements of the promoter region responsible for tissue-specific expression of ACP-1. Timing and tissue specificity of transgene expression is critical in optimizing attempts to manipulate the fatty acid content and composition of oilseed lipids through genetic engineering. Therefore, understanding how promoter elements regulate tissue-specific expression of proteins involved in fatty acid biosynthesis could enhance efforts to genetically engineer oilseed lipid biosynthetic enzymes for agricultural or industrial purposes.

Overexpression of ACP-1 in leaf tissue resulted in an increase in linolenic acid (18:3) and a decrease in 16:3. This provides the first in vivo evidence that overexpression of an ACP isoform in plants can influence its lipid fatty acid composition. There were no evident alterations in the fatty acid composition of seed lipids, which is consistent with ACP expression as there was no significant change of ACP profile in seeds. In addition, there was no significant change in the total lipid content in either leaf or seed tissue. Post-Beittenmiller et al. (1989b) expressed spinach ACP-1, the predominant ACP in spinach leaves, in transgenic tobacco (Nicotiana tabacum) plants at levels 2- to 3-fold higher than endogenous tobacco ACPs. This overexpression of ACP did not affect the lipid content of transgenic tobacco leaves. In addition, there were no significant changes in fatty acid composition of the transgenic tobacco leaf. Spinach ACP-1 is a leaf-predominant ACP (Battey and Ohlrogge, 1990). Spinach and tobacco leaves are very similar in fatty acid composition; therefore, expression of a spinach leaf ACP in tobacco leaf may not be expected to alter fatty acid composition. Post-Beittenmiller et al. may have seen alterations in tobacco leaf fatty acid composition if levels of spinach ACP-1 expression had been comparable to what was achieved in this study. In this case, a seed-predominant ACP was expressed in leaf tissue, where it is normally not expressed. Although it remains to be proven, it is possible that these changes are due to alterations in the ACP isoform profile, and not just the changes in total ACP content. The observation that overexpression of ACP in leaf tissue does not alter fatty acid content indicates that there may be a threshold level of ACP concentration such that increasing ACP levels beyond this threshold does not alter the amount of fatty acid synthesisized.

If ACP isoforms play a role in the determination of plant fatty acid composition, the question remains as to the mechanism by which ACP isoforms impart their influence. One possible mechanism may be specificity of ACP isoforms to fatty acid chain length. Because many ACP isoforms differ considerably in their amino acid sequence, even when they are from the same plant species (Hlouek-Radojcic et al., 1992), their structures may also differ, resulting in potential impacts on their functionality. In this case, ACP-1 could be more stable to elongation of its acyl chain from C16 to C18 fatty acids than other isoforms found in the leaf. Because the proportion of ACP-1 to other leaf isoforms was dramatically increased, a higher percentage of acyl chains may be attached to ACP-1 and subsequently elongated from C16 to C18.

The effect of ACP on fatty acid composition alternatively may be more related to the interaction of ACP isoforms with enzymes involved in fatty acid biosynthesis rather than with the fatty acids themselves. Several in vitro studies have shown specificity of thioesterases and desaturases for ACP isoforms (Guerra et al., 1986; Suh et al., 1999). One possible explanation for our results may be that stearoyl-ACP desaturase, which catalyzes the formation of the first double bond in a C18 fatty acid, may show specificity for ACP-1. If stearoyl-ACP desaturase shows specificity for ACP-1 over other ACP isoforms, and more acyl chains are attached to ACP-1, then more fatty acids would be desaturated to oleic, and then, perhaps by a feed-forward mechanism, be further desaturated to 18:2 and 18:3. Specific interaction of ACP-1 with the subsequent desaturases is not possible because fatty acids are assembled into galacto- or glycerolipids either in the chloroplast or in the endoplasmic reticulum (ER) prior to desaturation from oleic to linoleic and linolenic acid (Roughan and Slack, 1982; Norman and St. John, 1986).

In addition to possible interactions with stearoyl-ACP desaturase, ACP-1 may interact with other fatty acid biosynthesis or lipid assembly enzymes in the chloroplast resulting in changes in the ratio of fatty acids partitioned to either the prokaryotic or eukaryotic pathways of lipid biosynthesis. In the prokaryotic pathway, which takes place in the chloroplast, C16 fatty acids are incorporated onto the sn-2 position of glycerolipids due to the preference of chloroplast monoacylglycerol-3-phosphate acyl transferase for this acyl chain length (Frentzen et al., 1983). Glycerolipids originating from the eukaryotic pathway, on the other hand, contain primarily C18 fatty acids due to glycerol-3-phosphate acyl transferase and monoacylglycerol-3-phosphate acyl transferase specificity for C18 fatty acids (Frentzen, 1990). In Arabidopsis leaf tissue, approximately one-half of the glycerolipids are packaged in the ER, the eukaryotic pathway, and one-half in the prokaryotic pathway (Browse and Somerville, 1991). Therefore, one possible explanation of our results may be that the increase in ACP-1 protein expression results in an increase in the amount of fatty acids sent to the eukaryotic pathway. In the Arabidopsis act1 mutant (Kunst et al., 1988), which is completely deficient in activity of glycerol-3-phosphate acyltransferase, the first step in the prokaryotic pathway, there is an increase in the ratio of C18 to C16 fatty acids. A similar but less dramatic increase in the ratio of C18 to C16 fatty acids is also noted in our transgenic plants. A decrease in the amount of MGDG might also explain a decrease in 16:3; however, in preliminary analyses, we did not detect any differences in MGDG or other polar lipid components between WT and transgenic plants.

There are several possible interactions that ACP-1 could be involved in to increase the fatty acids sent through the eukaryotic pathway. One possibility is that acyl-ACP-1 may be favored by 3-ketoacyl-ACP synthase II, which catalyzes the elongation of 16:0 to 18:0 (Shimakata and Stumpf, 1982). This would result in more fatty acids being elongated to C18, and thus more sent to the eukaryotic pathway in the ER. Another possibility may be partiality of oleoyl-ACP thioesterase or hydrolase for acyl-ACP-1; if more 18:1 is hydrolyzed from ACP, then more free fatty acid will be available for transfer out of the chloroplast. Finally, a decrease in the interaction of ACP with plastidial forms of either glycerol-3-phosphate acyl transferase or lysophosphatidic acid acyltransferase activity could impair the amount of fatty acid sent through the prokaryotic pathway, resulting in an increase in C18 fatty acids in leaf lipids.

In conclusion, increasing the levels of ACP-1 in leaf tissue has resulted in increased 18:3 and decreased 16:3 in leaf lipids. This is the first in vivo evidence that overexpression of an ACP isoform can influence plant fatty acid composition. This knowledge, together with continuing in vivo studies, biochemical studies, and our future efforts to obtain structure/function information about ACP isoforms will further clarify the role of ACP isoforms in determining fatty acid composition. Information obtained will lead to a better overall understanding of how plants regulate the fatty acid composition of their lipids, and could be utilized to optimize and tailor ongoing attempts to modify oilseed lipid content and composition through genetic engineering.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

All chemicals were obtained from Sigma-Aldrich (St. Louis) unless otherwise stated.

DNA Constructs and Plant Transformation

DNA manipulation and transformation of Escherichia coli DH5 were performed according to standard protocols (Sambrook et al., 1989). The Arabidopsis ACP-1 genomic clone pAD4 (Post-Beittenmiller et al., 1989a) in pBluescript KS(+) (Stratagene, La Jolla, CA) was obtained from the laboratory of John Ohlrogge (Michigan State University, East Lansing). A 1.8-kb SalI/HindIII fragment of pAD4 was ligated into pBluescript KS(+) and designated as pAD412. This was cut with SalI and BamHI and ligated into the XhoI/BglII sites of the binary vector pGA748, which is derived from pGA643 (An et al., 1988), in antisense conformation behind the CaMV 35S promoter. This plasmid also contains the NPTII gene for kanamycin resistance and the Nos terminator. The resulting binary transformation vector was designated number 6 (Fig. 1).

Transgenic Plant Selection and Growth Conditions

The binary transformation vector was transferred into Agrobacterium tumefaciens strain C58C1 (pMP90) by electroporation. Arabidopsis ecotype Columbia transformants were generated by A. tumefaciens-mediated transformation using the vacuum infiltration (Bechtold et al., 1993) or the dip transformation method (Clough and Bent, 1998). Arabidopsis plants were also transformed with the vector pGA748 to serve as a transgenic control. Primary transformants were generated by screening T1 seeds on agar plates (pH 5.7) consisting of 4.3 g L¹ Murashige and Skoog basal salt mixture, 0.5 g L¹ MES (4-morpholineethanesulfonic acid), 9.6 g L¹ agar (Becton-Dickinson, Sparks, MD), and 50 mg L¹ kanamycin.

Twenty independent T2 transgenic plants were generated by the selection described above. Once roots formed, the seedlings were transferred to soil and grown in a temperature-controlled room at 22°C under conditions of 16-h light and 8-h dark. Leaves were harvested for analysis 6 weeks after germination. Because we were particularly interested in the fatty acid composition of the transgenic leaf tissue, which is often affected by variables such as temperature, humidity, and light, growth and analysis of T3 transgenic plants were carried out under tightly controlled conditions. Seeds from T2 plants were germinated on selection agar as described above to produce T3 plants, whereas WT seeds were germinated on Murashige and Skoog agar containing no antibiotic. Once roots developed, T3 seedlings were randomized and transplanted to soil in 32-well flats. Due to the large number of plants and the time taken for analysis, four to five independent transgenic lines were grown in random pattern in flats together with four WT plants and four plants from each of one to two independent transgenic control lines (seven total independent transgenic control lines). These experiments were repeated to include all of the individual number 6 transgenic plant lines as well as all of the transgenic controls. Plants were grown in a growth chamber (Percival Scientific, Boone, IA) at 22°C under conditions of 16-h light, 8-h dark, and 60% (v/v) humidity.

Determination of Alterations in ACP-1 Levels

Tissue from transgenic plants was analyzed for alterations in ACP protein levels by western-blot analysis as in Battey and Ohlrogge (1990). Leaves (25-50 mg) or seed tissue (40 seeds) were ground in 5× volumes of MOPS [3-(N-morpholino)-propanesulfonic acid] buffer (pH 6.8) that consisted of 50 mM MOPS (FisherBiotech, Fairlawn, NJ) with 10 mM dithiothreitol added fresh. The suspension was centrifuged and the supernatant was mixed with an appropriate volume of native sample buffer. Equal volumes were separated by native polyacrylamide (13% [w/v], 40:1 acylamide:bis-acrylamide) gel electrophoresis as described by Rock and Cronan (1981). Sample buffer consisted of (for 4×): 0.25 M Tris, pH 6.8 (Research Organics Inc., Cleveland), 40% (v/v) glycerol, 0.05% (w/v) bromphenol blue, and 20 mM dithiothreitol (added fresh). Gels were transferred to nitrocellulose membranes (Schleicher and Schuell, Keene, NH) with a Trans-Blot SD semidry transfer cell (Bio-Rad, Hercules, CA). ACP protein was detected with rabbit anti-spinach (Spinacia oleracea)-ACP polyclonal antibodies and developed with goat-anti-rabbit IgG-alkaline phosphatase conjugate (Kirkegaard and Perry Laboratory, Gaithersburg, MD). Pictures of the western blots were taken with a Kodak DC 290 (Eastman Kodak Co, Rochester, NY) digital camera, and Kodak 1D Image Analysis Software was used to compare the intensity of bands for determination of relative levels of various ACP isoforms.

Determination of Alterations in ACP-1 mRNA Levels

ACP-1 mRNA was analyzed by northern analysis. Total RNA was isolated from 100 mg of leaf tissue using TRIzol total RNA isolation reagent (Gibco BRL, Gaithersburg, MD) following the manufacturer's protocol with the following modifications to reduce polysaccharide contamination: 1, After homogenization, the suspension was centrifuged in a tabletop microcentrifuge at 13,000 rpm for 15 min to remove insoluble plant material; and 2, instead of precipitation with 0.5 mL of isopropanol, RNA was precipitated with 0.25 mL of isopropanol and 0.25 mL of a solution containing 0.8 M sodium citrate and 1.2 M sodium chloride. An estimate of RNA concentration was made by measuring A₂₆₀, and 10 µg of total RNA was loaded onto formaldehyde agarose gels for separation by electrophoresis. RNA was then transferred to nylon membranes overnight in 20× SSC (3 M sodium chloride, 0.3 M sodium citrate, pH 7.0). For probing, a SalI/BamHI fragment of pAD4 was ligated into pBluescript KS, then linearized with XhoI. RNA polymerase (Gibco BRL) was used to generate a single stranded [-³²P]-UTP-labeled RNA probe. Membranes were also probed with -actin to ensure equal loading. Membranes were hybridized overnight at 42°C and after washing were exposed to radiography film for 24 h. Intensity of bands was used for determination of relative levels of ACP-1.

Determination of Fatty Acid Composition and Content

Leaf lipid analysis was performed on WT, transgenic control, and transgenic plants grown side by side in the same 32-well flat. Two to three young but fully expanded leaves (45 d after planting) were harvested from each plant and total lipid was extracted from tissue with chloroform:methanol (1:2, v/v) and 0.15 M acetic acid as in the method of Bligh and Dyer (1959). An aliquot of total lipid extract was removed and heptadecanoic acid (17:0) was added as an internal standard. The lipid was dried under nitrogen and fatty acid methyl esters were made by heating at 95°C in 3 N methanolic HCl (Supelco, Bellefonte, PA) for 40 min. After cooling to room temperature, fatty acid methyl esters were extracted with hexane. Samples were injected onto a 5890 series II gas chromatograph (Hewlett-Packard, Palo Alto, CA) equipped with a flame ionization detector and a DB-225 0.25 µm i.d. capillary column (J&W Scientific, Folsom, CA). Chromatography conditions consisted of an initial temperature of 180°C for 3 min, ramping to 230°C at 3°C min¹, and holding at 230°C for 5 min. Peaksimple software (SRI Instruments, Torrance, CA) was used for data collection and integration. Fatty acid data was imported into Excel and fatty acid percentages (mol %) of transgenic plant lines versus WT and transgenic control lines were compared using analysis of variance by Statistical Analysis System (SAS Inc., Cary, NC). Means of individual plant lines were compared using the test of least square means.