Characterization and partial purification of acyl-CoA:glycerol 3-phosphate acyltransferase from sunflower (Helianthus annuus L.) developing seeds

Plant Physiology and Biochemistry 48 (2010) 73e80

Contents lists available at ScienceDirect

Plant Physiology and Biochemistry journal homepage: www.elsevier.com/locate/plaphy

Research article

Characterization and partial purification of acyl-CoA:glycerol 3-phosphate acyltransferase from sunflower (Helianthus annuus L.) developing seeds Noemí Ruiz-López a, Rafael Garcés b, John L. Harwood c, Enrique Martínez-Force b, * a

Rothamsted Research, Biological Chemistry Department, Harpenden, Herts AL5 2JQ, UK Instituto de la Grasa, CSIC, Av. Padre García Tejero 4, 41012-Sevilla, Spain c School of Biosciences, Cardiff University, Cardiff CF10 3AX, UK b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 June 2009 Accepted 2 December 2009 Available online 30 December 2009

The glycerol 3-phosphate acyltransferase (GPAT, EC 2.3.1.15) from sunflower (Helianthus annuus L.) microsomes has been characterised and partially purified. The in vitro determination of activity was optimized, and the maximum value for GPAT activity identified between 15 and 20 days after flowering. The apparent MichaeliseMenten Km for the glycerol 3-phosphate was 354 mM. The preferred substrates were palmitoyl-CoA ¼ linoleoyl-CoA > oleoyl-CoA with the lowest activity using stearoyl-CoA. High solubilisation was achieved using 0.75% Tween80 and the solubilised GPAT was partially purified by ion-exchange chromatography using a Hi-Trap DEAE FF column, followed by gel filtration chromatography using a Superose 12 HR column. The fraction containing the GPAT activity was analysed by SDS-PAGE and contained a major band of 60.1 kDa. Finally, evidence is provided which shows the role of GPAT in the asymmetrical distribution, between positions sn-1 and sn-3, of saturated fatty acids in highly saturated sunflower triacylglycerols. This work provides background information on the sunflower endoplasmic reticulum GPAT which may prove valuable for future modification of oil deposition in this important crop. Ó 2009 Elsevier Masson SAS. All rights reserved.

Keywords: Acyltransferase Glycerol 3-phosphate Helianthus annuus Microsomal membrane Sunflower Seeds

1. Introduction Recently, interest in controlling and manipulating the fatty acid composition of vegetable oils has increased appreciably. Attention has primarily focussed on the realization that seeds can store a wide variety of different fatty acids in their triacylglycerols (TAG) and the fact that fatty acid composition and distribution in triacylglycerol molecules determines the physical, chemical and nutritional properties of these oils. In oil seeds, by far the most common lipid is TAG. In higher plants, the biosynthesis of these TAGs is now known to occur by sequential acylations, via a number of different routes, in the endoplasmic reticulum of the cells. The central pathway of synthesis of TAG consists of four reactions, three of them being acylations of the glycerol backbone catalysed by acyl-CoA dependent acyltransferases. In the first step, a glycerol 3-phosphate acyltransferase (GPAT;

Abbreviations: BSA, Bovine serum albumin; CHAPS, 3-[(3echolamidopropyl) dimethylammonio]-1-propane sulfonate; DAF, days after flowering; DAG, diacylglycerol; DAGAT, diacylglycerol acyltransferase; GPAT, glycerol 3-phosphate acyltransferase; LPA, lysophosphatidic acid; LPAAT, lysophosphatidate acyltransferase; PA, phosphatidic acid; TAG, triacylglycerol; TLC, Thin-layer chromatography. * Correspondence to: Enrique Martínez Force, Av. Padre García Tejero, 4, 41012Sevilla, Spain. Fax: þ34 954616790. E-mail address: [email protected] (E. Martínez-Force). 0981-9428/$ e see front matter Ó 2009 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.plaphy.2009.12.001

EC 2.3.1.15) catalyzes the acylation in the sn-1 position of the glycerol 3-phosphate to form lysophosphatidic acid (LPA). In most plants e.g. safflower [16], this enzyme shows a particular preference for saturated fatty acids such as palmitate. Secondly, a lysophosphatidate acyltransferase (LPAAT; EC 2.3.1.51) transfers a fatty acid to the sn-2 position generating phosphatidate (PA). The third enzyme of this pathway is the phosphatidate phosphatase (EC 3.1.3.4), which hydrolyses phosphatidic acid to yield diacylglycerol (DAG), a substrate for phospholipid and glycosylglyceride synthesis. Finally, in the only step specific to TAG synthesis, diacylglycerol is acylated at the sn-3 position by a diacylglycerol acyltransferase (DAGAT; EC 2.3.1.20) [14,35]. The activity of a plant GPAT enzyme was first recognized by Barron and Stumpf [2] in a microsomal fraction from the mesocarp of avocado (Persea americana). Subsequently, additional GPAT activity was revealed in chloroplasts from spinach [17] and in mitochondria from the endosperm of castor bean (Ricinus communis L.) [32]. Nowadays it is known that there are three types of GPAT in plant cells; they are localized in plastids (including chloroplasts), in the endoplasmic reticulum and in mitochondria [9,26]. GPAT in chloroplasts is a soluble protein that is localized in the stroma [17,22]. By contrast, membrane-bound forms are found in mitochondria and the endoplasmic reticulum [9]. GPAT enzymes in the chloroplast, cytoplasm and mitochondria are different proteins and they differ also in their substrate specificities. The

74

N. Ruiz-López et al. / Plant Physiology and Biochemistry 48 (2010) 73e80

soluble enzyme uses acyl-ACP as a substrate in vivo; whilst the membrane-bound GPATs use acyl-CoA thioesters as substrates [10,14]. Most studies have been concerned with the soluble form of the enzyme, which has been purified from several species. Bertrams and Heinz [4] partially purified GPAT from chloroplasts of pea (Pisum sativum) and spinach (Spinacia oleacea) by ion-exchange column chromatography and isoelectric focusing. The degree of purification, in terms of the specific activity, was about 1000-fold for the enzyme from pea and 200-fold for that from spinach. The preparation from pea chloroplasts contained two isoforms. In 1987, Nishida et al. [27] purified GPAT from the chloroplasts of squash (Cucurbita moschata) cotyledons by ACP-affinity column chromatography and ion-exchange column chromatography. Isoelectric focusing revealed the presence of three isoforms with isoelectric points of 6.6, 5.6 and 5.5, respectively. Each of these isoforms was purified to homogeneity with the degrees of purification being 24,000-fold, 40,000-fold and 32,000-fold, respectively. Finally, the crystal structure of the soluble chloroplast acyltransferase from squash was published by Turnbull et al. [33]. Partial purification of the endoplasmic reticulum GPAT was not accomplished until much later. In 1995, Eccleston and Harwood [8] partially purified (150-fold) this enzyme from a microsomal fraction of avocado mesocarp after solubilisation with 3-3-cholamidopropyl dimethylammonio-1-propane sulfonate (CHAPS) and by affinity column chromatography with glycerylphosphorylethanolamine. In 2000, Manaf and Harwood [23] partially purified the cytosolic GPAT from oil palm (Elaeis guineensis) tissues after solubilisation with CHAPS. Purifications of up to 70-fold were achieved with several protein bands being present in the final preparations. To date no membrane-bound GPAT from any plant species has been identified in the databases or purified to homogeneity. In Arabidopsis, eight putative GPAT genes, designated from AtGPAT1 to AtGPAT8, have been described [36]. Only AtGPAT1 and ATGPAT5 have been studied in detail and none of them have shown to play a clear role in seed TAG biosynthesis [3,18]. Sunflower seed oil ranks fourth in world terms and is a major product of Argentinian, Russian and European agriculture. Traditional sunflower oil contains about 60% linoleic acid with little or no a-linolenate. Two important variants have been obtained by traditional breeding that either contain high (70e80%) oleate or an intermediate level (60%) [13]. Because sunflower oil contains TAG molecular species highly enriched in linoleate [28], characterization of the biosynthetic acyltransferases is of special interest. In this report, GPAT activity was studied using a microsomal membrane fraction from sunflower seeds (achenes). Solubilisation and significant purification were achieved. Such studies represent a first step in obtaining detailed knowledge of the sunflower endoplasmic reticulum GPAT in seeds. The understanding of this enzyme, including the possibility of cloning the gene (or genes) responsible for this activity, will be valuable in order to modify TAG biosynthesis in seeds and, thereby, manipulate TAG species in the oil, to improve its physicochemical properties and also open the possibility of increasing oil yields. 2. Results and discussion 2.1. Characterization of microsomal sunflower GPAT Initially, GPAT activity in sunflower seeds was examined in highspeed particulate fractions (microsomes) as GPAT activity is associated with the endoplasmic reticulum and in previous studies [8,23] the majority of the GPAT activity was found in the 105,000-g pellets from fruits or seeds. Several parameters were tested in these fractions with the aim of optimizing GPAT activity determinations.

Fig. 1. Effect of glycerol 3-phosphate concentration on the incorporation of radioactivity into glycerolipids by microsomal GPAT from sunflower seeds. For the assays, 100 mM oleoyl-CoA was used as acyl donor.

Firstly, the apparent Km for glycerol 3-phosphate in developing sunflower seed microsomes was established (Fig. 1). The results gave an apparent MichaeliseMenten Km of about 350 mM and a maximum activity (Vmax) of 2.08 pmol min1 mg1 of protein for the donor substrate. This Km value can be compared to those estimated for other plant species. For example, the microsomal GPAT Km of avocado was of 100 mM [8], olive 1370 mM [31], spinach chloroplast 700 mM [4], and palm callus 3400 mM [23]. In order to avoid glycerol 3-phophate becoming the limiting substrate, further GPAT assays were made using a concentration of 1.5 mM for the glycerol 3-phosphate. Secondly, microsomal protein concentration and incubation times were evaluated. Assays with increased microsomal protein content, from 10 to 300 mg, were undertaken. The GPAT activity increased linearly from 10 to 60 mg of protein (Table 1), therefore a concentration of 50 mg protein in final assay volume of 100 mL was used in further experiments. Time course experiments were then carried out in which incubation times from 2 to 120 min were tested. Product formation increased linearly up to 20 min, hence an incubation time of 15 min was selected as optimum for assaying GPAT activity (Table 1) in sunflower microsomes. When assaying GPAT activity it is important to consider that GPAT works only on monomeric substrates [15] and that acyl-CoAs are amphiphilic molecules, having a structure similar to surfactants. For this reason, molecules of acyl-CoA could form micelles in the aqueous solution when the acyl-CoA concentration is higher than the critical micellar concentration [1] causing very low availability of monomeric donor substrates for the microsomal GPAT. In previous studies, it was proposed that bovine serum albumin (BSA) could bind to acyl-CoAs thus avoiding the formation of micelles [16,31]. In order to determine any beneficial effect of adding additional BSA in sunflower GPAT assays, the incorporation of [14C] glycerol 3-phosphate by microsomal GPAT was determined using different BSA concentrations (data not shown). It was found that Table 1 Optimal incubation characteristics of GPAT from 15 DAF seed microsomes. Microsomal sunflower GPAT Glycerol 3-phosphate Km (mM) Typical Vmax (pmol min1 mg1 protein) Protein dependence (mg ml1) Reaction linearity (min) pH optimum

354 2.08 Up to 0.06 20 7.5

N. Ruiz-López et al. / Plant Physiology and Biochemistry 48 (2010) 73e80

there was no additional activity as the BSA concentration was increased. This could be due to the fact that the buffer used to prepare the microsomes already contained some BSA, 0.5%, and that this was sufficient for the assay. For this reason, no additional BSA was added in the following enzymatic activity determinations. Previous studies showed that cotyledon microsomal enzymes used for lipid biosynthesis often required divalent cations to achieve a maximum activity [25]. For example, GPAT activity from castor seed endosperm depends on Mg2þ, Ca2þ and Mn2þ [34]. However, the cucumber (Cucumis sativus) GPAT activity was independent of Mg2þ [21] indicating that the requirement for divalent cations depends on the source of the enzyme [16]. With these data in mind and in order to evaluate the effect of divalent cations and EDTA on GPAT activity, sunflower microsomal preparations were incubated with Mg2þ, Mn2þ and EDTA (Table 2). Neither Mg2þ or Mn2þ improved the GPAT activity. The assays with 2 mM Mg2þ or Mn2þ reduced their activity by 13.0 or 18.5%, respectively. Furthermore, the assays carried out with a concentration of 5 mM Mg2þ showed a further (63%) reduction of activity. Interestingly, the divalent cation chelator EDTA (at 10 mM concentration) increased the GPAT activity by 43%. Subsequently, the assay was performed with buffers of different pH from 6.5 to 9.0. Data showed that the pH optimum of the microsomal activity was 7.5 (data not shown). This value is similar to those found for avocado [8], olive [31], and palm [23], where the optimum pH were 7.5, 7.6 and 7.1, respectively. It was also similar to the one observed for chloroplast GPAT activity from pea [4] which was 7.4, but different to the one found for cucumber which had optimal activity in the pH range of 8.8e9.6 [21]. For further GPAT determinations pH 7.5 was chosen as optimum. In order to determine the period, during seed formation, with the highest GPAT activity, microsomal preparations were made from seeds of different ages (Fig. 2). It is important to note that, during sunflower seed formation, not only the lipid but also the storage protein content increases. Therefore, GPAT activity is expressed per seed and is not referenced to protein content. Data obtained from three independent assays showed a maximum GPAT activity value between 15 and 20 days after flowering (DAF), corresponding with the period of active lipid synthesis in sunflower [12]. The concentration effects of palmitoyl-, stearoyl-, oleoyl- and linoleoyl-CoA were tested in microsomes from 15 DAF sunflower seeds. Activity with these four substrates was measured at different substrate concentrations ranging from 0.0 to 0.4 mM (Fig. 3). Data showed that GPAT was able to support incorporation of radioactivity from [14C]glycerol 3-phosphate for all acyl-CoAs, but there was a strong inhibition by substrates at high concentration, probably due to the detergent activity of the acyl-CoAs, making it impossible to determine their Km. Linoleoyl-CoA which is the main fatty acid in TAG from sunflower oil, together with palmitoyl-CoA, gave the highest activity with oleoyl-CoA also yielding significant rates. The lowest GPAT activity was measured in those assays containing stearoyl-CoA which was clearly a poor substrate under the conditions used. These results are consistent with the main presence of the first three acids (but not stearate) at the sn-1

75

Fig. 2. Incorporation of radioactivity from [14C]glycerol 3-phosphate into glycerolipids using microsomal fractions from sunflower seeds of different developmental ages. Data as means (n ¼ 3)  S.D.

position of glycerol in sunflower oil [30]. Although, there is a broad substrate specificity in some plant acyltransferases [4,6,7,23,29], sn-1 acylations often occur with palmitate, oleate or linoleate with the selectivity depending on the plant species examined [14]. In contrast, the chloroplast GPATs can be divided into those that are particularly active with oleate and others with a more general specificity [11]. Finally, we analysed the lipid products which were radiolabelled during the acyl-CoA concentration experiments described above (Fig. 4). These data showed that the enzymes involved in TAG biosynthesis in sunflower seeds, are capable of using linoleate, palmitate or oleate at appreciable rates. In particular, GPAT when incubated with palmitoyl- or oleoyl-CoA showed a similar activity: when these acyl-CoAs were used at low concentrations the microsomes labelled all intermediates, but PA in higher proportion; however, as these acyl-CoA concentrations were increased, radioactivity gradually increased in the LPA fraction suggesting that the

Table 2 Effect of Mg2þ and Mn2þ cations and EDTA on GPAT activity. Data are means of triplicate incubations  SD.

Control Mn2þ Mg2þ EDTA

Concentration (mM)

GPAT Activity (nmol min1 mg1 protein)

e 2 2 5 10

5.4 4.4 4.7 2.0 7.7

    

0.1 0.1 0.2 1.5 1.4

% 100 81.5 87.0 37.0 143

Fig. 3. Incorporation of radioactivity from [14C]glycerol 3-phosphate into glycerolipids in microsomal fractions from sunflower seeds (15 DAF). when using different acylCoAs at increasing concentrations. Data as means (n ¼ 3)  S.D. (6) Palmitoyl-CoA; (B) Stearoyl-CoA; (>) Oleyl-CoA; (,) Linoleoyl-CoA.

76

N. Ruiz-López et al. / Plant Physiology and Biochemistry 48 (2010) 73e80

Fig. 4. Incorporation of radioactivity from [14C]glycerol 3-phosphate into individual lipid classes with increasing acyl-CoA concentrations using microsomal fractions from sunflower 15 DAF seeds. ( ) TAG þ DAG, (,) PA, (-) LPA.

LPAAT activity in these microsomes was becoming limiting. This effect was more accentuated in assays with palmitate than in oleate incubations. On the other hand, the assays of GPAT with linoleoylCoA showed that most of the radio labelling was in PA at any concentration, indicating a very high LPAAT efficiency for acylating LPA to produce PA. Thus, these data are consistent with linoleic acid representing the 58% of total fatty acids found in the sn-2 position of TAG from sunflower oil [24]. As expected, no incorporation of stearoyl-CoA was detected in DAG or TAG fractions, as well as a very low incorporation in PA fraction indicating that LPAAT and the rest of the microsomal enzymes involved in TAG biosynthesis in sunflower seeds have very low specify for this fatty acid, in keeping also with data obtained specifically for GPAT (Fig. 3). Recently, an asymmetrical distribution of saturated fatty acids, mainly in high-stearic sunflower lines, between position sn-1 and sn-3 of triacylglycerols has been described [24]. Initially, this fact could be due to the different affinity for saturated fatty acids of any of the acyltransferases involved in the transfer of acyl-CoAs to these positions

i.e. GPAT and DAGAT activities. However, the low affinity for stearoylCoA that sunflower GPAT showed indicates that this activity could be the one responsible for the asymmetric distribution. 2.2. Solubilisation of microsomal sunflower GPAT activity As a part of the purification of the GPAT enzyme, some anionic and non-ionic detergents were tested for their abilities to release GPAT activity from microsomal membranes. In this trial, the criterion used to distinguish solubilised membrane proteins after treatment was the presence of enzyme activity in the supernatant fractions, following centrifugation for 1 h at 105,000 g. The GPAT activity was measured in microsomes treated for 15 min on ice with different detergents at different concentrations (Fig. 5). Optimal solubilisation of GPAT activity using CHAPS treatment was achieved in microsomes of both avocado and palm [8,23]. However, it was not successful in solubilising GPAT in sunflower microsomes. Similar results were obtained with Triton X-100. In both cases the

N. Ruiz-López et al. / Plant Physiology and Biochemistry 48 (2010) 73e80

Fig. 5. Solubilisation of GPAT activity from sunflower seeds (15 DAF) using several detergent treatments. Data as means (n ¼ 3)  S.D. (B) Tween20, (6) Tween80, (,) Triton X-100, (7) CHAPS.

activity was reduced to near zero at a detergent concentration of 0.75% (w/v). In contrast, sunflower GPAT activity was solubilised by treatment of microsomes with Tween80 or Tween20 from 0.2 to 0.8% (w/v). Finally, Tween80 at a concentration of 0.75% (w/v) was selected because of its lower interference with protein content determination. The Tween80-solubilised GPAT activity was stable for 24 h at 4  C but after this time started to decrease.

2.3. Partial purification of solubilised GPAT Manaf and Harwood [23] achieved a good quality purification of solubilised oil-palm GPAT by two steps: ion-exchange and gel permeation chromatography. These results directed our attempt at the purification of the membrane-bound protein by anionexchange. Ion-exchange chromatography of solubilised extracts was performed using a continuous salt gradient (0e1 M NaCl) on a DEAE FF column. A representative separation is shown in Fig. 6. GPAT activity was determined in all fractions. The GPAT activity mostly eluted between fractions 10 to 12. These three fractions were pooled, concentrated and chromatographed on a Superose 12 HR column. Fractions of 1 ml were collected. Once more, GPAT activity

77

Fig. 7. Elution profile of proteins after gel filtration chromatography on Superose 12 HR column. The preparation loaded corresponded to a mixture of the fractions 10 to 12 eluted from the DEAF FF column (see Fig. 6). Straight line, optical density (280 nm); and dotted line, GPAT activity from collected fractions. Collected fractions are indicated with horizontal lines above the horizontal axis.

was assayed in all eluted fractions (Fig. 7). Fraction 7 gave the major peak of specific GPAT activity which eluted slightly after the main protein peak. The apparent molecular mass of the GPAT protein was determined after calibration of the Superose 12 HR column with known standards and using the elution volume of the GPAT activity. The GPAT peak corresponded to proteins, ranging between 30 and 45 kDa. Table 3 shows the recovery of each purification step. A purification of 16-fold was achieved after these two chromatographic steps. This modest purification was partly due to total losses in enzyme activity during purification but also by co-election of GPAT with a major protein peak (Fig. 7). Fractions 6e9 (Fig. 7) were concentrated and run on SDS-PAGE to examine the protein bands present (Fig. 8). The main peak of activity (Fraction 7) showed a major band of 60.1 kDa. This was also present in Fraction 8 which contained lower GPAT activity but not in Fraction 6 which did not have detectable activity. Surprisingly, there was relatively little protein present in Fraction 7 in the range 30e45 kDa which had been indicated by gel filtration to contain the GPAT. However, estimations of molecular masses in the presence of detergent are known to be problematic, especially by gel filtration [20]. Therefore, we were not unduly surprised at an apparent lack of consistency between our data for gel filtration and SDS-PAGE. Nevertheless, the gel filtration step was effective, since the activity was concentrated mainly in one of the fractions. It is known that the chloroplastic GPATs have molecular masses ranging from 30 to 40 kDa in pumpkin (C. moschata) cotyledons [27] to 42 kDa in pea and spinach [4] (see Frentzen and Wolter [11] for a detailed discussion). In contrast, microsomal preparations from avocado mesocarp gave three main proteins associated with GPAT activity in the range 54e70 kDa [8], while that from palm tissues revealed candidate peptide bands of 56e67 kDa [23]. Taking Table 3 Purification of microsomal glycerol 3-phosphate acyltransferase from 15 DAF sunflower seeds. Purification step

Fig. 6. Elution profile of GPAT activity after ion-exchange chromatography on a DEAE FF column. The loaded preparation corresponds to the 15 DAF seed microsomal proteins solubilised after a treatment with Tween80 (0.75%, w/v) for 15 min. Straight line, optical density (280 nm); dotted line, GPAT activity from collected fractions; and dashed dotted line, NaCL gradient. Collected fractions are indicated with horizontal lines above the horizontal axis.

Yield Purification Total activity Protein Specific activity (nmol min1 mg1) (%) (-fold) (nmol$min1) (mg)

Microsomal 6.8 fraction Solubilised 3.5 GPAT Ion-exchange 2.9 þ Gel filtration

1400

4.8

100.0

1.0

535

6.6

52.2

1.4

37

79.5

43.5 16.5

78

N. Ruiz-López et al. / Plant Physiology and Biochemistry 48 (2010) 73e80

25/15  C (day/night), with a 16-h photoperiod, photon flux density of 200 mmol m2 s1. Seeds were sampled at 15 days after flowering or at other different stages when needed. 4.3. Preparation of microsomal membranes/high-speed particulate fractions

Fig. 8. Analysis by SDS-PAGE of GPAT purified from microsomal fraction of sunflower seeds (15 DAF). Lane ST, protein molecular-mass standards. Lanes Fr.6 to Fr.9, fractions obtained by ion-exchange followed by gel filtration chromatography. The arrow indicates the protein band that could contain GPAT activity.

all this into account, it is probable that the 60.1 kDa band contained the sunflower endoplasmic reticulum GPAT. However, further experimentation is needed to confirm this. 3. Conclusions This work has highlighted valuable information about the microsomal sunflower GPAT. The activity has been partly characterised, solubilised and modesty purified from developing sunflower seeds. The GPAT was notably active with linoleoyl-CoA and its low activity with stearoyl-CoA may explain the low levels of stearate at the sn-1 position in sunflower triacylglycerols. Other properties including a possible molecular mass of about 60 kDa are consistent with previously-reported data for different plant endoplasmic reticulum GPATs. In future studies, the understanding gained in this work should provide useful underpinning information for further characterization of the sunflower endoplasmic reticulum GPAT. This may prove valuable for future modification of the characteristics of oil deposition in this important crop.

For characterization of the GPAT activity, microsomal fractions were prepared as follows. Seeds from sunflower capitulum were pealed and then ground using a pre-cooled pestle and mortar in Buffer A containing 0.33 M sucrose, 1% BSA in 0.1 M potassium phosphate buffer (pH 7.2) at a sunflower achene:buffer (g/ml) ratio of 2:1. The homogenate was filtered through two layers of Miracloth, diluted ten times in Buffer A and spun at 20,000 g for 10 min. The resulting supernatant (with the floating fat layer removed) was filtered again and spun at 105,000 g for 90 min to yield a microsomal membrane pellet. The pellets were resuspended by pipetting up and down in Buffer A (without BSA) at a concentration of 3e6 mg protein ml1. For solubilisation and purification of the GPAT, microsomal fractions were prepared as follows. Seeds from sunflower capitulum were pealed and ground using a pre-cooled mortar in Buffer B containing 0.33 M sucrose in 0.1 M Bis-Tris propane buffer (pH 7.0) at a sunflower achene:buffer (g/ml) ratio of 20:1. Protease inhibitors (aminocaproic acid, benzamidine-HCl and phenylmethylsulfonylfluoride) were added to Buffer B at final concentrations of 5 mM, 1 mM, and 1 mM, respectively. The homogenate was filtered through two layers of Miracloth, diluted ten times in Buffer B, and spun at 20,000 g for 10 min. The resulting supernatant was filtered again and spun at 105,000 g for 90 min. The pellets were resuspended as before in Buffer B at a concentration of 3e6 mg protein ml1. All steps were carried out at 4  C. For both preparations, BCA Protein Assay Kit (Pierce, Rockford, IL, USA) was used to assay the protein content. Bovine serum albumin was used as standard. Aliquots were frozen at 80  C until required. 4.4. Assay of GPAT Incubations were carried out in 1.5 ml Eppendorf tubes for 15 min in a 30  C water bath with gentle shaking. The standard incubation mixture contained L-[U-14C]glycerol 3-phosphate (3700 Bq, 150 nmol), 300 mM linoleoyl-CoA, 0.5% BSA and 50 mg microsomal proteins in either 50 mM potassium phosphate (pH 7.5) or 50 mM Bis-Tris propane buffer. Reactions were made up to a final volume of 100 ml and stopped at different time points (standard incubation 15 min) with the same volume of 0.15 M acetic acid. Lipids were extracted as described below. 4.5. Lipid extraction

4. Experimental 4.1. Chemicals L-[U-14C]Glycerol 3-phosphate, ammonium salt (specific activity 5.59 GBq mmol1) was obtained from Amersham Pharmacia Biotech. All other chemicals and solvents were of the best available grades and were purchased from Sigma Chemical Co, Boehringer Mannheim, Whatman International or Amersham Pharmacia Biotech. 4.2. Plant material Plants from a standard sunflower line RHA-274 were used in this study. The sunflower plants were cultivated in growth chambers at

Lipids were extracted from the enzymatic assays using a method adapted from Bjerve et al. [5] that prevents lysophosphatidate loss. An equal volume of a mixture containing butan-1-ol/water/acetic acid (1:0.95:0.05, by vol.) was added to the aqueous mixture as one phase and vortexed. An equal volume of butan-1-ol was added immediately and mixed again. After phase separation by spinning at 2000 g for 7 min, the upper butanol phase was stored and the remaining aqueous phase was re-extracted with an equal volume of upper phase that had been equilibrated with an artificial aqueous phase. Extracted lipids were stored at 20  C under a nitrogen atmosphere. Radioactivity was determined in aliquots of the combined butanol extracts using a LS-6500 Multipurpose Scintillation Counter (Beckman) to give total incorporation rates. The

N. Ruiz-López et al. / Plant Physiology and Biochemistry 48 (2010) 73e80

counts were corrected for quenching using the external standard channels ratio method.

4.6. Lipid analysis Thin-layer chromatography (TLC) was performed using precoated silica-gel plates (Merck, Damstadt, Germany) activated before use for 1 h at 100  C. Polar lipids were separated using a solvent system of chloroform/methanol/acetic acid/water (85:15:10:3.5, by vol.). Non-polar lipids were separated with a solvent system of hexane/diethyl ether/acetic acid (80:20:1, by vol.). Usually lipids were separated using a one-dimensional double-development technique with the polar solvent run threequarters up the plate followed by the non-polar solvent. Lipids on TLC plates were visualised with iodine vapour and were identified by co-chromatography with authentic standards. Incorporation of radioactivity from radiolabelled precursors into total lipids or individual classes were determined using Ecoscint scintillant (National Diagnostics) and a LS-6500 Multipurpose Scintillation Counter (Beckman). All samples were corrected for sample quenching using an external standard.

79

4.9. Analysis by SDS-PAGE Protein samples were concentrated using the Nanosep 10 K Omega (Pall) and supplemented with 0.5 volumes of loading buffer (0.25 M TriseHCl pH 6.8, 4% SDS, 27% sucrose, 0.3% glycine, 0.1% bromophenol blue and 10% b-mercaptoethanol). Samples were denatured by heating at 90  C for 5 min. The SDS-PAGE analysis was performed according to Laemmli [19] using polyacrylamide NuPageÒ Novex Bis-Tris Gels 4e12% gradient (Invitrogen). Electrophoresis was carried out at a constant voltage between 80 and 175 V in an XCell ll mini-Cell (Novex) tank. After electrophoresis, the proteins were fixed to the gels using water/methanol/acetic acid (72/20/7.4, by vol). Gels were stained with 0.5% (w/v) Coomassie Brilliant Blue G (Sigma Aldrich) in water/methanol/acetic acid (72/ 20/7.4, by vol) and destained in water/methanol/acetic acid (72/20/ 7.4, by vol) by gentle shaking. As reference proteins, the SDS-PAGE low-range molecular weight standards (Bio-Rad) were used. Acknowledgements We are very grateful to Richard Haslam for his advice and reviewing this manuscript. This work was supported by the MEC and FEDER, project AGL2005-00100.

4.7. Solubilisation of GPAT Microsomal fractions equivalent to a protein concentration of 2 mg/ml were incubated with four solubilising agents (CHAPS, Triton X-100, Tween20 and Tween80) at different concentrations in a 50 mM Bis-Tris propane buffer (pH 7.0) containing 1 mM benzamidine-HCl, 5 mM aminocaproic acid, and 0.3 M KCl. Solubilisation was carried out for 15 min at 4  C with gentle stirring. After incubation, the solubilised material was isolated by centrifuging at 105,000 g for 1 h. The supernatant (containing solubilised GPAT) and the pellet were assayed to measure the specific activities and recovery. The supernatant was kept at 4  C but used within 24 h for further purification of the enzyme activity, as required. 4.8. Partial purification of GPAT All steps were carried out at 4  C. The solubilised material was applied to a 1 ml column Hi-Trap DEAE FF (Amersham Biosciences) equilibrated with 10 vol. of 20 mM Bis-Tris propane (pH 9.0) containing Tween80 (0.2%, w/v), 5 mM aminocaproic acid and 1 mM benzamidine-HCl (Buffer C). Unbound protein was eluted with Buffer C (10 column vol.) at a constant flow rate of 1 ml min1. The column was eluted with a linear gradient from 0 to 1 M NaCl in Buffer C (flow rate 1 ml min1). The eluate was collected in 1 ml fractions in which GPAT activity was determined. Fractions containing acyltransferase activity (from 10 to 12) were combined and concentrated using a Nanosep 10 K Omega (Pall). After concentration, the active fraction was applied to a Superose 12 HR 10/30 (Amersham Biosciences) column (15 cm long, 1 cm i.d.) which had been equilibrated with 5 vol of Buffer D containing Tween80 (0.1%, w/v) and 0.15 M KCl in 75 mM Bis-Tris propane (pH 7.0). Elution was carried out using Buffer D at a constant flow of 0.5 ml min1. The eluate was collected in 1 ml fractions in which GPAT activity was determined. In both separations, protein elution was monitored by measuring the absorbance at 280 nm. Fractions with acyltransferase activity were characterised immediately due to the fact that, under these conditions activity was unstable to prolonged storage. However, the fractions maintained constant rates of GPAT activity for at least 24 h (data not shown) and were always used in this period.

References [1] R.E. Barden, W.W. Cleland, 1-acylglycerol 3-phosphate acyltransferase from rat liver. J. Biol. Chem. 244 (1969) 3677e3684. [2] E.J. Barron, P.K. Stumpf, Evidence for separate pathways for biosynthesis of saturated and unsaturated fatty acids by an avocado enzyme system. J. Biol. Chem. 237 (1962) 613e614. [3] F. Beisson, Y. Li, G. Bonaventure, M. Pollard, J.B. Ohlrogge, The acyltransferase GPAT5 is required for the synthesis of suberin in seed coat and root of Arabidopsis. Plant Cell 19 (2007) 351e368. [4] M. Bertrams, E. Heinz, Positional specificity and fatty-acid selectivity of purified sn-glycerol 3-phosphate acyltransferases from chloroplasts. Plant Physiol. 68 (1981) 653e657. [5] K.S. Bjerve, L.N.W. Daae, J. Bremer, Selective loss of lysophospholipids in some commonly used lipid-extraction procedures. Anal. Biochem. 58 (1974) 238e245. [6] G.M. Carman, G.F. Harrington, R. Amegah, Microsomal associated glycerophosphate acyltransferase activity in germinating soybeans. J. Food Biochem. 5 (1981) 185e195. [7] V.S. Eccleston, J.L. Harwood, Acylation reactions in developing avocado fruits. in: P.J. Quinn, J.L. Harwood (Eds.), Plant Lipid Biochemistry, Structure and Utilisation. Portland Press, London, 1990, pp. 178e180. [8] V.S. Eccleston, J.L. Harwood, Solubilization, partial purification and properties of acyl-CoAeglycerol-3-phosphate acyltransferase from avocado (Persea americana) fruit mesocarp. Biochim. Biophys. Acta 1257 (1995) 1e10. [9] M. Frentzen, Comparison of certain properties of membrane-bound and solubilized acyltransferase activities of plant microsomes. Plant Sci. 69 (1990) 39e48. [10] M. Frentzen, Acyltransferases and triacylglycerols. in: T.S. Moore Jr. (Ed.), Lipid Metabolism in Plants. CRC Press, Boca Raton, London, 1993, pp. 195e220. [11] M. Frentzen, F.P. Wolter, Molecular biology of acyltransferases involved in glycerolipid synthesis. in: J.L. Harwood (Ed.), Plant Lipid Biosynthesis, Fundamentals and Agricultural Applications. Cambridge University Press, Cambridge, U.K, 1998, pp. 247e272. [12] R. Garcés, M. Mancha, Oleate desaturation in seeds of two genotypes of sunflower. Phytochemistry 28 (1989) 2593e2595. [13] F.D. Gunstone, J.L. Harwood, Occurrence and characterisation of oils and fats. in: F.D. Gunstone, J.L. Harwood, A.J. Dijkstra (Eds.), The Lipid Handbook, third ed. Taylor and Francis, Boca Raton, 2007, pp. 37e69. [14] J.L. Harwood, Recent advances in the biosynthesis of plant fatty acids. Biochim. Biophys. Acta 1301 (1996) 7e56. [15] G. Hübscher, Glyceride metabolism. in: S.J. Wakil (Ed.), Lipid Metabolism. Academic Press, New York, 1970, pp. 279e370. [16] K. Ichiara, sn-Glycerol-3-phosphate acyltransferase in a particulate fraction from maturing safflower seeds. Arch. Biochem. Biophys. 232 (1984) 685e698. [17] J. Joyard, R. Douce, Site of synthesis of phosphatidic acid and diacylglycerol in spinach chloroplasts. Biochim. Biophys. Acta 486 (1977) 273e285. [18] H.U. Kim, Y. Li, A.H. Huang, Ubiquitous and endoplasmic reticulum-located lysophosphatidyl acyltransferase, LPAT2, is essential for female but not male gametophyte development in Arabidopsis. Plant Cell 17 (2005) 1073e1089. [19] U.K. Laemmli, Cleavage of structural proteins during the assembly of the head bacteriophage. Nature 227 (1970) 680e685.

80

N. Ruiz-López et al. / Plant Physiology and Biochemistry 48 (2010) 73e80

[20] T.M. Lane, D.G. Rhodes, Determination of size, molecular weight and presence of subunits. in: M.P. Deutscher (Ed.), Methods in Enzymology: Guide to Protein Purification, vol. 182. Academic Press, New York, 1990, pp. 566e587. [21] B.A. Macher, C.P. Brown, T.T. McManus, J.B. Mudd, Studies on phospholipid synthesizing enzyme activities during growth of etiolated cucumber cotyledons. Plant Physiol. 55 (1975) 130e136. [22] T.A. McKeon, P.K. Stumpf, Purification and characterization of the stearoyl-acyl carrier protein desaturase and the acyl-acyl carrier protein thioesterase from maturing seeds of safflower. J. Biol. Chem. 257 (1982) 2141e2147. [23] A.M. Manaf, J.L. Harwood, Purification and characterisation of acyl-CoA, glycerol 3-phosphate acyltransferase from oil palm (Elaeis guineensis) tissues. Planta 210 (2000) 318e328. [24] E. Martinez-Force, N. Ruiz-Lopez, R. Garcés, The determination of the asymmetrical stereochemical distribution of fatty acids in triacylglycerides. Anal. Biochem. 334 (2004) 175e182. [25] P. Mazliak, A. Jolliot, C. Bonnerot, Biosynthesis and metabolism of phospholipids. in: J.F.G.M. Wintermans, P.J.C. Kuiper (Eds.), Biochemistry and Metabolism of Plant Lipids. Elsevier, Amsterdam/New York, 1982, pp. 89e98. [26] N. Murata, Y. Tasaka, Glycerol-3-phosphate acyltransferase in plants. Biochim. Biophys. Acta 1348 (1997) 10e16. [27] I. Nishida, M. Frentzen, O. Ishizaki, N. Murata, Purification of isomeric forms of acyl-[acyl-carrier-protein]-glycerol-3-phosphate acyltransferase from greening squash cotyledons. Plant Cell Physiol. 28 (1987) 1071e1079. [28] H.M.D. Noor Lida, K. Sundram, W.L. Sien, A. Aminak, S. Mamot, TAG composition and solid fat content of palm oil, sunflower oil and palm kernel oil

[29]

[30]

[31]

[32] [33]

[34] [35] [36]

blends before and after chemical interesterification. J. Am. Oil Chem. Soc. 79 (2002) 1137e1144. R.J. Porra, Formation of phospholipids from sn-glycerol-3-phosphate and free fatty acids or their derivatives by homogenates of soybean cotyledons. Phytochemistry 18 (1979) 1651e1656. J. Reske, J. Siebrecht, J. Hazebroek, Triacylglycerol composition and structure in genetically modified sunflower and soybean oils. J. Am. Oil Chem. Soc. 74 (1997) 989e998. A.J. Rutter, J. Sánchez, J.L. Harwood, Glycerolipid synthesis by microsomal fractions from fruits and tissue cultures of olives. Phytochemistry 46 (1997) 855e862. S.A. Sparace, T.S. Moore, Phospholipid-metabolism in plant-mitochondriasub-mitochondrial sites of synthesis. Plant Physiol. 63 (1979) 963e972. A.P. Turnbull, J.B. Rafferty, S.E. Sedelnikova, A.R. Slabas, T.P. Schierer, J.T. M. Kroon, I. Nishida, N. Murata, J.W. Simon, D.W. Rice, Crystallization and preliminary X-ray analysis of the glycerol-3-phosphate 1-acyltransferase from squash (Cucurbita moschata). Acta Crystallogr. D Biol. Crystallogr. 57 (2001) 451e453. B. Vick, H. Beevers, Phosphatidic acid synthesis in castor bean endosperm. Plant Physiol. 59 (1977) 459e463. R.J. Weselake, Storage lipids. in: D.J. Murphy (Ed.), Plant Lipids: Biology, Utilisation and Manipulation. Blackwell Publishing, Oxford, U.K, 2005, pp.162e225. Z. Zheng, Q. Xia, M. Dauk, W. Shen, G. Selvaraj, J. Zou, Arabidopsis AtGPAT1, a member of the membrane-bound glycerol-3-phosphate acyltransferase gene family, is essential for tapetum differentiation and male fertility. Plant Cell 15 (2003) 1872e1887.