HPLC-grade chloroform, acetone, acetonitrile, toluene and isopropanol were purchased from the Fisher Scientific Company (Ottawa, Ontario, Canada-Ottawa Admin. Center 112 Colonnade Road Nepean, Ontario K2E 7L6 Canada). Methanol and Ethanol were purchased from Panreac Quimica SA (Barcelona, Spain). Petroleum ether, ethyl ether and acetic acid were from Fisher Scientific SA (Barcelona, Spain). TLC silica plates (silica gel 60 G F254, 20×20 cm, 0.2 mm thickness) were obtained from Merck (Darmstadt, Germany). All standards were acquired from Sigma Aldrich (Madrid, Spain). Other employed chemicals and solvents used in the experiments were purchased from Aldrich (Milwaukee, WI, USA).

The variety of flaxseed “O116” was obtained from “Institut National Recherche Agronomie Tunis” (INRAT), Tunisia. This variety of flaxseed (Linum usitatissimum L.) was grown in restricted zones (15 m×3 m) on the Agronomy farm of the INRAT from the middle of November 2006 until the end of June 2007. Each sample was collected at intervals after a number of days after flowering (DAF) ranging from 7 DAF to 56 DAF. Moisture and seed weight were determined by weighing 100 seeds before and after drying to constant weight in a vacuum oven at 80 °C for 72 h. Seeds lengths were measured using a micrometer.

The total lipids were extracted by the method of Folch et al. (1957) modified by Bligh et al. (1959). Flaxseeds (2.5 g) were washed with boiling water for 5 min to denature the phospholipases (Douce, 1964) and then crushed in a mortar with a mixture of CHCl₃-MeOH (2:1, v/v). Fixing water was added and the homogenate was centrifuged at 3000 rpm for 15 min. The lower chloroformic phase containing the total lipids was dried in a rotary evaporator at 40 °C.

Lipid class separation was performed by TLC according to the method of Mangold (1964) using petroleum ether/ethyl ether/acetic acid (70:30:0.4, v/v/v) as developing solvent. The oil samples were placed on thin-layer chromatographic plates (G60; Merck, Darmstadt, Germany) (20 cm×20 cm×0.2 mm) covered with silica gel using a microsyringe. Spots were revealed by brief exposure to iodine vapor. Iodine was then removed under nitrogen flux and by heating the plates at 60 °C. The TAG spots were then collected.

A 50 mg sample of the extracted lipid was fully dissolved in 5 mL of dichloromethane. LC/MS experiments were performed using an Agilent LC system (Agilent Technologies; Palo Alto, CA-2660 Matheson Boulevard East, Mississauga ON, L4W 5M2) coupled to an Applied Biosystem /MDS Sciex QSTAR Elite Q-TOF MS with a photoionization ion source. The TAG extracts were separated using an Ascentis TM C18 LC column (15 cm×2.1 mm, 3 µm) with C18 guard cartridge (Supelco). The mobile phase consisted of A, acetonitrile and B, isopropanol. A solvent gradient program was run from 30 to 90% B for 25 min. Subsequently, the LC system was programmed to return its initial solvent composition over 0.1 min, followed by a 5 min re-equilibration prior to the following injection. The injection volume was 5 µL and the flow rate was 200 µL·min⁻¹. The entire effluent from the column was directly transferred to the MS through the photoionization interface. MS analysis of the analytes was performed using APPI in the positive ion mode. Toluene was used as dopant delivered at a flow rate of 20 µL·min⁻¹ by an isocratic pump (Agilent Technologies; Palo Alto, CA-2660 Matheson Boulevard East, Mississauga ON, L4W 5M2). Analyst QS 2.0 software was used for data acquisition and analysis. Nitrogen was used as curtain gas, nebulizing gas, drying and lamp gas (4 L·min⁻¹). The mass range recorded was from 50 to 1300 amu. The other optimal instrumental conditions were as follows: photoionization voltage 1300V; curtain gas setting 25; gas 1 setting 20 and gas 2 setting 70; de-clustering potential (DP), 35V; focusing potential (FP), 150; second de-clustering potential (DP2), 15V; and ion source temperature: 400 °C. Estimations of TAG concentrations were achieved using the relative peak areas of each TAG. Although this method is only semi-quantitative, likely overestimating highly unsaturated components relative to saturates, it does allow for the evaluation of the estimation of relative TAG concentrations and the evolution of individual TAG species over the course of seed ripening. As examples of the results, the total ion current (TIC) chromatogram of flaxseed oil is given in Figure 1. The fatty acid composition of triacylglycerols was calculated with the following equation: % FA= [ ( % TAG ) × n . FA ] / 3

% FA: Fatty acid type percentage in individual TAG

% TAG: individual TAG percentage

n: Number of FA type esterified in individual TAG

TAGs were analyzed by HPLC on a SCHIMADZU apparatus equipped with an RP-18 stainless steel column (250 mm length×4.6 mm internal diameter), a guard column and a refractive index detector (differential refractometer). The mobile phase was acetonitrile/acetone 1:1 v/v. The flow rate was isocratically controlled at 1.5 mL·min⁻¹. Ten microliters of sample were injected at each run. The identification of TAG peaks was made by comparison of their retention times with authentic standards. The concept of the equivalent carbon number (ECN) was used to rationalize the retention time of TAG. The ECN is defined as the total number of carbon atoms (CN) in the FA acyl chains minus twice the number of double bonds (N) per molecule: ECN=CN-2N. Quantification of the contents of the TAG species (expressed in % of total TAG) was done using the following formula: content (%)=PAi/TPA, where PAi is the peak area of an individual TAG and TPA is the peak area of the total TAG. A typical HPLC chromatogram of flaxseed oil is shown in Figure 2.

The AOCS Official Method was employed for the determinations of iodine value (IV) in the oil samples (AOAC, 1993).

Oxidative stability was evaluated by the Rancimat method (Gutiérrez, 1989). Stability was expressed as the oxidation induction time (hours), measured with the Rancimat 743 apparatus (Metrohm Co., Basel, Switzerland), using an oil sample of 2 g warmed to 100 °C and an air flow of 10 L·h⁻¹.

The calorific value was estimated (in Kcal·100 g⁻¹) by multiplying the percentage of crude protein, crude lipid and carbohydrate by the recommended factor (4, 9 and 4.125 respectively) used in vegetable analysis (Hunt et al., 1987).

Digestible crude protein was estimated as follows: Digestible crude protein (g)=(protein (g)×0.96–4.21) (Barrett and Larkin, 1977).

Protein fractions were isolated by sequentially extracting flaxseed with different solvents (the seed to solvent ratio was 1:10, w/v, in each case) according to the modified Osborne scheme (Osborne, 1924), as described by Chavan et al. (2001): Samples (1 g) were dispersed in 10 ml of distilled water by stirring with a magnetic stirrer and extracted over 45 min periods at room temperature (25 °C). The suspension was then centrifuged at 5,000×g for 15 min, and the resultant supernatant was filtered. The residues were re-extracted twice more with the same solvent and the recovered filtrates were combined and designated the ‘‘water-soluble fraction”. The residue was then extracted successively with a 0.5 M NaCl solution (pH 7.0) of 70% (v/v) ethanol at 65 °C in a shaking water bath, and 0.1 M sodium hydroxide in order to separate the total seed proteins into albumin, globulin, prolamin and glutelin fractions, respectively. The filtrates containing the desired protein fractions were dialyzed against distilled water for 48 h at 4 °C and separately lyophilized. The protein content of each fraction was determined by the micro-Kjeldahl procedure (AOAC, 1984).

Weighed flaxseed (500 mg seeds for each assay) were homogenized with four volumes of extraction buffer containing 50 mmol·L⁻¹ Tris-HCl (pH 7.0), 3 mmol·L⁻¹ CaCl₂, and 4 mmol·L⁻¹ NaCl and centrifuged at 8000 g for 20 min. The supernatant was centrifuged at 20 000 g for 20 min. A 200-mL aliquot of the final supernatant was mixed with 50 mL of β-limited dextrin made from potato starch with β-amylase (Kohno and Nanmori, 1991). After incubation at 37 °C for 5 min, 0.5 mL of a 0.1% iodine–0.3% potassium iodide solution (I₂-KI), and 2 mL distilled water were added to the reaction mixture. The change in color was recorded at 620 nm with the extraction buffer as the blank control. A unit of activity was defined as the amount of enzyme that caused a decrease in absorbance of 0.1 at 620 nm under the assay conditions (Chen et al., 1998).

Statistical analysis was performed using the Proc ANOVA in SAS (software version 8). Results are presented as the means ± SD from three triplicates of each sample.

The stages and seeds lengths are given in Table 1. Flaxseed development can be divided into 8 stages. This study used seed size and color to determine time of maturity. The plant growth and especially the flowering were followed daily to provide a precise determination of the maturation stage which proved to be a very important parameter (Attoumbré et al., 2011). Seed size increased to a peak value at S7–S8 stages (Table 1). The average length of mature flaxseeds was 6.0–6.5 mm. At stage S8, seeds were considered as mature according to the characteristic brown color of mature seeds.

The changes in proximate composition during flaxseed development are shown in Table 5. The moisture content (74% at 7 DAF) decreased continuously in weight as maturity progressed to obtain a maximum of 5% at ripe seed (56 DAF). The gradual decline in relative water content is likely due to the accumulation of dry matter as it is in agricultural species (Lee et al., 2008). The protein levels increased gradually from 8% at 7 DAF to 24% at complete maturity, which was attributed to the rapid synthesis of oil and carbohydrates in early seed development (Saldivar et al., 2011). The results indicate that flaxseed is a good source of protein and lipids for human consumption. The ash content showed a continuous increase during flaxseed development in that the highest level of ash (3.20%) was observed at complete maturity. This observation finds great support in the literature (Mueller et al., 2010). The level of total carbohydrates (by difference) fluctuated and no definite trend was obvious, although the highest amount (31%) was detected at complete maturity. The total energy values for the seeds analyzed indicated that the mature seeds (49 DAF) which contained significantly the highest oil content had more total energy (2400 Kj·100 g⁻¹) than the other stages. Flaxseed could therefore be a good supplement as a source of energy in feed formulation. The estimated digestible crude protein ranged from 3% (7 DAF) to 19% (49 DAF). From the foregoing discussion, it may be concluded that flaxseed is potentially a good source of dietary energy and protein.

Osborne (1924) classified seed protein into four groups on the basis of their different solubilities: albumins (water soluble), globulins (soluble in dilute salt solutions), prolamins (soluble in aqueous alcohols) and glutelins (dilute acid). Flaxseed accumulates albumin as a major protein (Table 6). The highest percentage in albumin (65%) was observed at 28 DAF. The level of globulins increased at the beginning of seed development and decreased before the end of seed maturation. Both albumins and globulins represented the most abundant soluble protein (from 62 to 87% of total soluble proteins) during flaxseed development. Since albumins are mostly enzyme proteins (Bewley and Black, 1978), the synthesis of enzymes during ripening might be responsible for the apparent increase in the content of the water-soluble fraction. Prolamins declined rapidly during the initial 2 to 4 weeks of flaxseed development and then remained in a relatively constant content. Glutelins followed a similar pattern but then increased throughout the final ripening stages. Our results show that at full maturity, protein was a mixture of albumins (53%), globulin (33%), prolamin (3%) and glutelin (11%). These findings are in agreement with the results of Wanasundara et al. (1999) who reported that the major classes of flaxseed proteins were albumins and globulins.

A low α-amylase activity was observed at the beginning of seed maturation (Figure 3). However, from 28 DAF, the α-amylase activity in flaxseed increased to be more than 0.3 units/10mg. α-amylase was involved in starch breakdown at later development stages (Monma et al., 1991). The highest α-amylase activity was recorded at 49 DAF. After that, the activity of α-amylase decreased again. This decrease may be explained by the decrease in the break-down of starch at late developmental stages and with concomitant lower soluble sugar. Flaxseed has low starch content at complete maturity (1–2%) (Epaminondas et al., 2011). Relatively high α-amylase activity was concomitant with the lower starch content in flaxseed. It has an active role in the hydrolysis of the starch during seed development. It may also be responsible for the maintenance of requisite water potential, by providing solute sugars during seed development (Murtaza and Asghar, 2012).