Introduction

Lignans and neolignans are structurally diverse secondary plant metabolites that are widely distributed in the plant kingdom. They consist of two phenylpropane monomers linked through C–C or C–O bonds. Apart from their role in conferring resistance against various opportunistic biological pathogens and predators in vascular plants (Lewis and Davin 1999), they display pharmacologically important properties in mammalian systems (Pool-Zobel et al. 2000; Arroo et al. 2002). The major lignans from flaxseed (Linum usitatissimum), secoisolariciresinol diglucoside and matairesinol, derived from coniferyl alcohol dimerization, are converted into the “mammalian lignans” enterodiol and enterolactone by intestinal bacteria (Wang et al. 2000). These compounds were shown to reduce the incidence rates of breast and prostate cancers by modulating steroidal hormone synthesis (Adlercreutz and Mazur 1997). According to the important pharmacological properties and physiological roles of these lignans in planta, numerous studies have been carried out in an attempt to get a better knowledge of the biological events linked to the biosynthesis and the accumulation of these metabolites (Ford et al. 2001; Sicilia et al. 2003).

In order to investigate lignan metabolism in flax, cell suspension cultures of Linum usitatissimum were established, characterized and their lignan content evaluated. In particular, the accumulation of a phenylpropanoid dimer identified as dehydrodiconiferyl alcohol-4-β-d-glucoside (DCG) (Fig. 1) was evaluated in relation to cell growth and correlated to phytohormone balance of the culture media. This metabolite was first identified in the water soluble part of Euphrasia rostkoviana extract (Salama et al. 1981) and was later isolated from many plant species (Changzeng and Zhongjian 1997; Wang and Jia 1997; Yoshizawa et al. 1990). DCG was also isolated from Citrus limon and its hypotensive effect demonstrated (Matsubara et al. 1991). An antiinflammatory activity was also described for DCG extracted from cell cultures of Plagiorhegma dubium (Arens et al. 1985). Apart from these pharmacological properties, DCG is a plant growth factor discovered when screening plant tumor lines for factors overproduced as a result of oncogenetic transformation (Orr and Lynn 1991).

Fig. 1
figure 1

Chemical structure of dehydrodiconiferyl alcohol-4-β-d-glucoside (DCG)

Full size image

As DCG shows biological activities, a strategy to enhance the production of DCG and the ease of its recovery appeared to be of great interest. One of the strategies currently used to improve the production levels of secondary metabolites from plant cell cultures is cell immobilization (Verpoorte et al. 2002). This technique, which had already proved efficient in our laboratory for this class of compounds (Gillet et al. 2000), was applied to the L. usitatissimum suspension cells.

Materials and methods

General experimental procedures

All solvents were HPLC grade and were obtained from Dislab (France). Semi preparative HPLC was conducted on a TSP 200 system equipped with UV-VIS detector SPD-6AV (Shimadzu). HPLC quantification was carried out on a Shimadzu LC-10AS HPLC system and a Merck L-4250 spectrophotometer. NMR spectra were recorded at 300 K on a BRUKER Avance 300 spectrometer, operating at 300.13 MHz for 1H spectra and 75.47 MHz for 13C spectra. DMSO was used as a solvent and tetramethylsilane (TMS) as an internal standard. Proton resonance assignments and structural elucidation were performed by 1D (1H and 13C) and 2D homo- and heteronuclear experiments (COSY, HSQC, HMBC). ESI-MS spectra were obtained using a Quattro micro mass spectrometer (Micromass, Manchester, UK).

Plant material

Seeds of L. usitatissimum var. Barbara were surface sterilized by successive rinsings in Tween 20 (5 min), EtOH (70% v/v) (5 min), 3.6% sodium hypochlorite (15 min) and then with benlate® (50% benomyl), each rinsing being followed by sterilized water washings (5 min). Seeds were then germinated on solidified medium containing the macro- and microelements of Linsmaier and Skoog (1965) and grown at 23±1°C with a 16 h photoperiod. After 10 days, roots of in vitro plantlets were excised and transferred to several liquid media with different phytohormone balances (FMA = LS medium with 0.5 mg L−1 NAA; LuR = LS medium with 1 mg L−1 NAA and Lu1 = LS medium with 0.5 mg L−1 NAA and 0.5 mg L−1 BAP). All the suspensions were cultured in 300 mL conical flasks filled with 100 mL medium (pH 5.8) and approximately 5 g FW plant material. Flasks were maintained in the dark at 23±1°C on a rotary shaker at 150 rpm (orbital diameter 3 cm).

Standard protocol for the growth kinetics and DCG quantitation

Fifteen-day-old cells (1 g FW) were gathered and transferred into 20 mL liquid LS medium supplemented with specific hormonal balances (see above) and incubated under the conditions previously described. The cultures were halted after 2, 4, 6, 8, 10 and 15 days. The cells were separated from medium by filtration, rinsed with distilled water and freeze-dried. DW was measured for growth kinetics and 200 mg were used for DCG quantification after extraction with methanol/water (70/30, v/v) (10 mL) for 3 h at 60°C. After filtration, the extracts were made up to 20 mL with distilled water and analysed in triplicate by HPLC using a Kromasil C18 column (5 μm, 250 mm×4 mm, Macherey Nagel, France) eluted isocratically with acetonitrile/0.2% acetic acid/THF (20/79/1, v/v/v) with a flow rate of 0.7 mL min−1 and detected at 280 nm by UV absorption. In these conditions, the DCG retention time was 15 min. Each kinetic was repeated three or four times.

DCG isolation

A freeze-dried and ground sample of cells cultured in FMA medium (5 g) was extracted with methanol/water (70/30, v/v) (200 mL) for 3 h at 60°C. After filtration, the extract was concentrated to 15 mL. DCG was isolated by semi-preparative HPLC using a μBondapack C18 prepacked column [10 μm, 25 mm×100 mm], eluted with acetonitrile/0.2% acetic acid (15/85, v/v), flow rate 8 mL min−1, detection by UV absorption at 280 nm, retention time 25 min. DCG was identified by 1H and 13C NMR and ESI-MS.

Immobilisation procedure

Cell suspensions were filtered using nalgene® filtration unit. One gram FW free cell suspension was mixed with 20 mL sterile alginate solution (1% w/v) according to Charlet et al. (2000), and then dripped into 100 mL of a sterile CaCl2 solution (0.1, 0.45 and 0.8 M), where the Ca-alginate beads were formed by ionotropic gelation. After 30 min, the beads were rinsed five times with sterile deionized water and transferred into 100 mL Lu1 medium.

Biomass determination

Approximately 100 beads were disrupted by adding 10 mL EDTA-phosphate solution (K2HPO4) (0.1 and 0.2 M, respectively) and stirred for 30 min at pH 7.5. The liberated cells were filtered and weighted to measure the final fresh biomass. The final dry biomass was measured after freeze-drying. Each experiment was repeated three times. A growth index (GI) was introduced in order to compare the different growth conditions and is defined as the ratio between final and initial fresh weights.

Cell viability

The cell viability was estimated with FDA staining (Tsuji et al. 1995). Cell viability was determined by the ratio of living (i.e. fluorescent) cells to total cells. A minimum of 1,000 cells were examined for each sample. Microscopic observations were performed using an inverted optical microscope LEICA DMIRP to control cell morphology and viability.

Results and discussion

Initiation and characterization of cell suspension cultures

Cell suspension cultures were initiated from excised roots of aseptically grown L. usitatissimum plantlets. In order to determine efficient conditions for cell growth, the suspensions were subcultured in three media differentiated by their phytohormone balance. FMA and LuR contained the auxin NAA at two different concentrations (0.5 and 1 mg L−1 respectively) whereas Lu1 medium contained a combination of auxin (NAA 0.5 mg L−1) and cytokinin (BAP 0.5 mg L−1). Figure 2 shows the course of culture growth (expressed as g DW per flask) in these three media. The amount of biomass obtained after 15 days of culture was measured. When using only NAA at 0.5 mg L−1 (FMA medium) biomass increased slowly and the total biomass obtained was 0.22 g DW per flask. With double the NAA concentration (LuR medium), growth was slightly enhanced yielding 0.54 g DW per flask which was the highest biomass obtained after 15 days. With a combination of NAA 0.5 mg L−1 and BAP 0.5 mg L−1 (Lu1 medium), cell growth was fast during the first 10 days of culture, however the amount of the biomass after 15 days—only 0.37 g DW per flask—was lower than in the former medium. In all cases after 15 days of culture, the mass dropped and the cell suspension colour turned to brown.

Fig. 2
figure 2

Time course of Linum usitatissimum culture growth in different culture media

Full size image

From these experiments, it can be concluded that all the tested media are efficient in sustaining growth in L. usitatissimum cell suspension cultures. The presence of a cytokinin in the medium enhanced the cell growth, nevertheless the highest biomass at the end of the culture was obtained by increasing the auxin level. Indeed, it has been well established that callus can be initiated with auxin alone as well as with a combination of auxin and cytokinin (Krikorian et al. 1990).

Evaluation of DCG content in the cell suspension cultures

The three cell suspension cultures were analysed using a HPLC method that was developed previously (Charlet et al. 2002) for evaluation of the content of the main flax lignans: matairesinol and secoisolariciresinol diglucoside. The results revealed that these lignans could not be identified in the cell suspensions or in trace amounts only. Another compound accumulated at a high level and its structure was determined by 1H and 13C NMR spectroscopy once isolated by semi-preparative HPLC. These data were identical to those reported by Changzeng and Zhongjian (1997) for dehydrodiconiferyl alcohol-4-β-d-glucoside (Fig. 1). It is noteworthy that we could not detect DCG in the whole plant of Linum usitatissimum.

In the cell suspension cultures, the DCG content was evaluated during 15 days of culture (Fig. 3). When the cell suspensions were grown with a combination of NAA 0.5 mg L−1 and BAP 0.5 mg L−1 in the medium (Lu1 medium), the cellular DCG content increased steadily until day 10 and then decreased. The highest DCG accumulation was obtained with this medium irrespective of the day of culture. For instance, on day 10, 47.7 mg g−1 DW were yielded in this condition versus 20.5 mg g−1 when using NAA alone 0.5 mg L−1 (FMA medium) and 8.3 mg g−1 when the NAA concentration was double (LuR medium). When taking into account the cell biomass, it corresponds to absolute DCG values of 17.97 and 3.39 mg per flask on day 10 for Lu1 and FMA media respectively and of 5.66 mg per flask on day 15 for LuR. These were the maximal values obtained except with LuR medium (NAA 1 mg L−1) in which the maximum occurred on day 15. Other compounds of therapeutic interest have already been produced at a high yield by different cell suspension cultures. This is the case of shikonin by Lithospermum erythrorhizon cells (63 mg g−1 DW) (Yamamoto et al. 2000) or berberine by Coptis japonica cells (132 mg g−1 DW) (Sato and Yamada 1984). Rosmarinic acid production by cell cultures of Coleus blumei has also been achieved on a large scale (360 mg g−1 DW) (Petersen and Simmonds 2003), and sanguinarine was highly produced by cell cultures of Papaver somniferum (50 mg g−1 DW) (Archambault et al. 1996).

Fig. 3
figure 3

Accumulation kinetics of dehydrodiconiferyl alcohol-4-β-d-glucoside in Linum usitatissimum cell suspensions in different culture media

Full size image

During the 15 days of culture, the DCG content was much less fluctuant in the cytokinin- free media, such as FMA and LuR, than it was in Lu1 medium. The average DCG contents in these media were respectively 15.5 mg g−1 (FMA medium) and 6.4 mg g−1 (LuR medium). Indeed, the use of a doubled auxin (NAA) concentration resulted in a loss of DCG content by a factor of 2.5. It is concluded that this phytohormone in higher concentration has a negative effect on DCG accumulation in Linum suspension cells. In media containing the same concentration of auxin, a higher concentration of DCG (2.3 times more) is detected when supplemented with a cytokinin (BAP). This shows a positive effect of the cytokinin on the DCG accumulation. Our results are consistent with that of Liau and Ibrahim obtained on flax callus (Liau and Ibrahim 1973). They have reported the same effects of auxin and cytokinin on p-coumaric and ferulic acids, the latter being a likely precursor of DCG. This parallels some observations made in carrot cell cultures where the auxin down regulated the mRNA level and the enzymatic activity of PAL (phenylalanine ammonia lyase) and CHS (chalcone synthase), two enzymes of the phenylpropanoid pathway (Ozeki et al. 1990a, b).

In contrast, it appeared that a small amount of DCG was released into the media, which represented around 1% of the total DCG content irrespective of the day and the medium of culture (data not shown). This release might be due to cell death. In order to achieve DCG in larger amounts and more easily, immobilization experiments were carried out, since this technique has been shown to be efficient for recovering secondary products in culture media (Dörnenburg and Knorr 1995).

Immobilization of Linum usitatissimum cells

Numerous reports in the literature establish a positive correlation among secondary metabolite synthesis, accumulation and entrapment (Skjåk-Braek and Espevik 1996). Thus cell immobilization using alginate beads presents an option for the increased production of secondary compounds. Moreover, it has been shown previously that the modulated network bead structure can strongly influence the cell behaviour of Solanum aviculare and Nicotiana tabacum cells, especially the production and the excretion of their secondary metabolites (Roisin et al. 1997). In particular, immobilization effects depend on several parameters such as polymer concentration and ionic strength (Nava Saucedo et al. 1996). Thus, different matrices obtained from 1% alginate and various calcium chloride concentrations (0.1, 0.45 and 0.8 M) were used to carry out the first immobilization procedure of L. usitatissimum cells.

As calculated above, the amount of DCG produced by using a combination of NAA 0.5 mg L−1 and BAP 0.5 mg L−1 (Lu1 medium) was up to three times higher than when using one of the two other media. For this reason, Lu1 medium was chosen to carry out the immobilization experiments.

The experiments were evaluated by viability tests, determinations of biomass and measurements of DCG inside the cells and excreted into the medium. Cells were monitored at all four culture conditions (0.1, 0.45, 0.8 M CaCl2 and control) and at different days. The free cell suspension was taken as the control: the free cells mainly consisting of aggregates showed a spherical shape, and the cell viability was estimated to be about 100% by FDA staining. Initially with 0.1 M CaCl2, it was noticed that the beads were homogeneous with immobilized cells having a spherical shape, however the cells began to leak into the suspension after 12 days of growth. Nevertheless cell viability remained optimal since staining was observed in all cells. When the concentration of CaCl2 was increased to 0.45 M, the bead matrix became inhomogeneous but 70% of the immobilized cells were still fluorescent. With 0.8 M CaCl2, cells were no longer viable. This phenomenon may be explained by the hypertonicity of the medium due to the high ionic concentrations, which favour cell plasmolysis (Charlet et al. 2000).

As expected, the matrix bead structure strongly affected the physiological behaviour of immobilized L. usitatissimum cells (Gillet et al. 2000). The cell growth was evaluated by calculating the growth index GI. The best GI of approximately seven was obtained with the lowest concentration of 0.1 M CaCl2. This GI, very close to that of the control (8.1), showed that these immobilization conditions were not detrimental to the cell growth. Increasing the CaCl2 concentration to 0.45 M induced a GI decrease to 1.8. With 0.8 M no growth was observed. These results showed that an increase in CaCl2 concentration has a negative effect on the L. usitatissimum cell growth which is probably due to the tighter bead structure. Previous investigations on the immobilization of cell cultures, e.g. N. tabacum (Gillet et al. 2000) have also shown a decrease in dry biomass linked to a slower cell growth. The entrapment conditions, not favourable for increasing biomass, may be more suitable for the accumulation of DCG.

DCG production in immobilized cells

The produced and excreted DCG levels were determined by HPLC analysis as described above. As can be seen in Fig. 4, the production of DCG in free cell conditions was approximately 25 mg g−1 DW on day 6 and 48 mg g−1 DW at day 10. Intra and extracellular concentrations of DCG were evaluated in the two tested conditions (0.1, 0.45 M) on day 6 and day 10.

Fig. 4
figure 4

Effect of CaCl2 concentrations on DCG accumulation and excretion (mg g−1 DW) in Linum usitatissimum cell suspensions on day 6 and day 10

Full size image

The data showed that whatever the entrapment conditions, the total DCG contents varied from 28.0 to 60.0 mg g−1 DW in immobilized cells. On day 6 after immobilization the amount of DCG remained similar to that of the control. On day 10, the DCG production was approximately 20% higher compared to the control at 0.45 M CaCl2. This could be due to the effect of immobilization which can provide better cell to cell contact and also help in cell differentiation to produce secondary metabolites (Komaraiah et al. 2003). Besides, immobilization also influenced the excretion of DCG since higher extracellular recoveries were observed under both tested conditions. The secretion of the metabolite into the surrounding medium was enhanced by a factor of 25 in the case of 0.45 M on day 10 for instance. These results proved that immobilization can effectively influence cell physiology by offering the opportunity to increase the excretion of metabolites of biological interest with reduced costs for the extraction and purification of the product.

Conclusion

As DCG remains undetectable in whole Linum plant extracts, the suspension cells cultivated in the presence of both auxin and cytokinin accumulates this compound at a level of up to 49 mg g−1 DW. Although immobilization experiments do not significantly enhance the total DCG production, it is possible to significantly increase the release of DCG into the medium. If suspension cell cultures of L. usitatissimum were to be used for exploiting the therapeutic value of DCG on a larger scale, the latter property would be of interest for seriously reducing costs of product recovery. The product could be recovered easily after separation of the entrapped cells by direct solvent extraction, which can be considered of practical importance for industrial applications.