The plants were sown under field conditions and harvested at intervals of three weeks. Based on these condi- tions, three different stages of maturity could be used for the study: roots at 9 weeks (9 w), 6 weeks (6 w) or 3 weeks (3 w). The extracts were prepared according to Luyengi et al. [8] as modified by Matscheski et al. [9] . Plant material (20 g) was ground with liquid nitrogen in a mortar and extracted with 180 ml of 100% methanol in a water bath (15 min, 70˚C) by using a reflux condenser. The solution was cooled, filtered and evaporated almost to dryness. The extract was resuspended in 8 ml distilled water and partitioned with ethyl acetate. After being dried, the ethyl-acetate-soluble fraction was dissolved in absolute ethanol to provide a stock solution of 100 mg/ml. For cell culture tests, this solution was diluted to the following final concentrations: 1000, 500, 100, 50, 10, 1, 0.1 and 0.01 µg/ml.

By using Py-FIMS, we were able to analyse the components of the root extracts [10] . An aliquot of 5 µl of the extract was transferred to a quartz crucible. With an applied heating rate of 6.66 K per scan, the biological mac- romolecules in the extract were fragmented into their chemical subunits. This thermal evaporation occurred rap- idly and with a minimal mass of substance. The fragments were correlated with the known structure of sub- stances of interest. The molecular fragments of the thermal degradation was ionised in the following high elec- tric field (FI Emitter +4.8 kV, Counter electrode –5.5 kV). The gaseous ions were accelerated, separated, elec- trically intensified and registered in a Finnigan MAT 900 mass spectrometer. During the heating period from 110˚C to 700˚C and the period of analysis of 12 minutes, 91 mass spectra at an interval from 50 to 700 Dalton were registered and recorded in a data system. For each of the single scans, the absolute and relative ion intensi- ties of 14 classes of chemical compounds were calculated by summation of the ion intensities of 4 to 127 indi- cator signals. The marker signal selection was based on the concepts of Schulten et al. [11] , Hempfling et al. [12] and Hempfling and Schulten [13] . Adaptation and validation of a linum-based mass spectrometric database were achieved by an exhaustive literature survey with tentative assignments and the measurement of selected pure compounds. This procedure was carried out for each of three replicate measurements per sample. The results were averaged for statistical analyses.

The human breast cancer cell line MCF7 was obtained from the Department of Human and Animal Cell Culture, Braunschweig, Germany. Cells were cultivated in Dulbecco’s modified Eagle’s medium (DMEM, Sigma-Al- drich-Chemie, Germany). This medium was supplemented with 4.5% glucose; L-glutamine; 2.6 g/l HEPES pH 7.4; 3.0 g/l NaHCO₃; 10% foetal calf serum; 1% penicillin; 1% streptomycin and 0.5% amphotericin B. The MCF7 cell line was cultured in an incubator under physiological conditions (37˚C; 5% CO₂, humidified atmos- phere).

For proliferation and cytotoxicity tests, a cell concentration of 5 × 10⁵ cells/ml was used. The breast cancer cells (100 µl/well) were plated onto 96-well micro-plates. After 24 hours, the flax root extracts at different concentrations (0.01 µg/ml - 1000 µg/ml) and all controls were added. 17β-oestradiol and tamoxifen (anti-oestrogen) were used as positive controls (final concentrations: 50 µg/ml; 1 µg/ml; 0.1 µg/ml). Both controls were commercially acquired from Sigma (Germany). One of the negative controls was the solvent ethanol (Roth; Germany), the other one being the untreated MCF7 cell suspension.

For measuring cytotoxicity, the Cytotoxicity Detection Kit LDH (Roche Diagnostics, Germany) and, for cell proliferation, the BrdU assay (colorimetric, Roche Diagnostics) were used according the manufacturer’s protocol.

Cell culture tests were performed in quadruplicate and repeated at least four times. Statistical analysis was exe- cuted by using the Student’s t-test for a comparison of the means. Data are presented as means ± standard devia- tion. All values with P ≤ 0.01 differ significantly from the negative control (ethanol) and are denoted with an as- terisk (^*). Significance between different stages of maturity (P ≤ 0.05) is symbolised with a hatch sign (^#).

Py-FIMS allowed the characterisation of substance classes in the extracts being tested. The analysis showed a high fraction of free saturated and unsaturated fatty acids. Sterols and triterpenes were detected in all three stag- es, but their total ion intensity (TII) was higher in 3 w than in 6 w and 9 w extracts (Figure 1).

The lignan fraction (0.3%) in 3 w extracts was striking. Phenols and lignin/lignan monomers were verified, as were lignin dimers. In the 6-week-old flax root, we detected lignans (2.2%), lignin dimers, phenols and lignin/

lignan monomers. The 9 w root showed the highest TII of lignans (3%) and lignin dimers but the proportion of fatty acids was the lowest compared with the other two stages. Flavonoids were also found in all three flax roots, increasing with the maturity of the flax (0.4%, 0.8% and 1.3%). The percentage of groups of phenolic compounds, potentially containing phytoestrogens, was much higher in the extract of the 9 w (23%) root than in 6 w (16.3%) and 3 w (6.8%) roots.

Figure 2 illustrates the thermograms of total ion intensity (inset upper right) and summed averaged Py-FI mass spectra of the flax root extracts. At first glance, all depict similar patterns. However, on closer inspection, differences can be noted. The thermograms at 3 w (Figure 2(a)) and 9 w (Figure 2(c)) had clear multimodal curvature, indicating fractions of different thermal stability. The first volatilization maximum for 3 w occurred at 200˚C. The same was detected for 6 w roots (Figure 2(b)) but without a clear peak and without further curves. The thermogram of the 9 w root had a maximum in the range 130˚C to 180˚C and a second around 250˚C to 300˚C. The maximum intensities were different and increased with the age of the roots. The signal patterns of the summed Py-FI mass spectra also behaved differently, with masses m/z > 500 at 6 w; these were lower or not pronounced in the other two root groups. However, we also found that the same mass signals differed in their TII (e.g. 180, 248, 256, 280 340, 400 and 414). The mass spectra and thermograms at 3 w and 9 w seemed simi- lar.

A significant cytotoxic effect was detected for all three stages of maturity at the highest extract concentration (1000 µg/ml). The extracts of 3 w and 9 w roots caused a cytotoxicity of more than 50%. For the concentrations of 500 µg/ml, a cytotoxicity of about 35% was determined. At other concentrations, no cytotoxic effect was measured (Figure 3). A comparison between the three stages of maturity showed statistical significance for 3 w/6 w (1000, 100 and 0.01 µg/ml) and 3 w/9 w (0.01 µg/ml).

In addition to cytotoxicity, the influence of extracts on cell proliferation was of interest (Figure 4). Cell pro- liferation was significantly inhibited at high concentrations of all three extracts in an age-dependent manner. The 9 w root extract decreased cell growth by up to 50%, the 6 w extract by up to 40% and the 3 w extract by up to 15%. Incubation of the cells with extracts at 100 µg/ml and 50 µg/ml resulted in an inhibition of nearly 15% (6 w and 9 w). A significant difference as noted between 3 w and 9 w (100 and 500 µg/ml) and 3 w/6 w (100 µg/ml). At low concentrations (1, 0.1 and 0.01 µg/ml), the extract of the 3-week-old flax root promoted cell proliferation to 130%. In the case of 0.01 µg/ml, this was statistically significant compared with the results of 6 w extract. The other stages of maturity did not show this proliferating influence.