The current study showed that Cc extracts have, in general, high phenolic content. The highest phenolic contents were found in leaf extracts with a concentration range of 5.22 ± 0.1 and 9.25 ± 0.29 mg GAE g⁻¹, in DWL and GCL, respectively. This phenolic concentration range is similar to previously reported Cc leaf methanolic extracts [30] and is comparable to that demonstrated for different aromatic and medicinal plants [31]. Concerning the antioxidant activity, our results showed that all extracts exhibit both reduction and DPPH free radical scavenging activities. Statistical analysis showed that for phenolic content, reduction capacity and DPPH scavenging activity there is significant interaction between botanical variety, harvest time and plant part; this indicates that the effect of one factor is not consistent at all combinations of the two other factors. Therefore, an unambiguous interpretation of the main effect is not possible. Thus significant interactions between factors were analyzed over the level of plant parts. Concerning the phenolic content, in leaf extracts, there is no difference between dry and green plants; however, there is significant difference between extracts obtained from wild and cultivated cardoon. These results showed that within leaf extracts the most important factor for phenolic content is the biological variety; extracts from cultivated plants are richer in phenols than those obtained from wild plants. Thus, cultivated plants are the most profitable to be used as source of natural phenolic compound. Regarding the reduction capacity and scavenging activity, leaf extracts are the most active. In addition, there is no interaction between harvest time and biological variety. This suggests that the antioxidant activity of leaf extracts does not depend on harvest time or biological variety. GCL extract presents the highest DPPH scavenging capacity.

In stalks, there is no differences between phenolic content of extracts from wild and cultivated Cc plants, however the phenolic content of extracts from green and dry plants are significant different. Extracts obtained from green plants are richer in phenols. Thus, within stalks extracts harvest time is the most important factor for phenol content. Regarding the reduction capacity and scavenging activity there are differences between green vs. dry and wild vs. cultivated Cc plants extracts. These results suggest that both factors are determinate for extracts anti-oxidant activity. Both for reduction capacity and scavenging activity extracts obtained from dry plants are more effective than those obtained from green plants.

Concerning extracts obtained from inflorescences there is statistical difference within harvest time and biological variety for phenolic content, reduction capacity and scavenging activity for both harvest time and plant variety. This suggests that both factors are important for determining the phenolic content and the antioxidant activity of the extracts. Extracts obtained from mature inflorescences present the highest levels of phenolic compounds; these extracts are also the most effective antioxidant (reduction capacity and scavenging activity).

In addition, there is a correlation between total phenolic concentration and antioxidant capacity, suggesting that extracts bioactivity depends, at least in part, on total phenolic contents. The highest concentration of phenols was found in cultivated leaf extracts, these extracts were also the more active antioxidants. Although differences between extracts activity in vivo are more tenuous than those found in vitro; leaf extracts are the more effective in protecting yeast against H₂O₂ damage. Our results showed that the antioxidant activity of Cc extracts in vitro can be extrapolated to in vivo systems. Furthermore, given that using an equal phenolic concentration the antioxidant effect of different Cc extracts is similar, these observations reinforce the hypotheses that antioxidant activity of Cc extracts strongly depends on phenol concentrations. Thus, Cc leaves extracts are a rich source of antioxidant compounds that could be used in preventive therapies in a number of diseases that are described to be related to oxidative stress, including cancer [32].

During the second stage of carcinogenesis (promotion) tumor cells acquire the capacity to undergo uncontrolled cell division. Most cancer therapies currently used target this phase of carcinogenesis; cancer chemotherapy relies mostly on cytotoxic drugs, which inhibit tumor cell proliferation and cause cell death. Our results showed that only four of the 12 Cc extracts tested are active in decreasing tumor cell viability. Thus, there is no correlation between total phenolic contents and the ability of Cc extracts to decrease cell viability. Unlike the antioxidant activity, total phenolic content does not seem to be essential for the anti-proliferative activity of Cc extracts; we hypothesize that the physiological mechanisms involved in decreasing tumor cell viability may require specific phenols. Alternatively, anti-proliferative effects of Cc extracts may be due to compounds other than phenols. Preliminary studies aiming to identify GCL bioactive phytochemicals in GCL are in progress; this extract was fractionated by solid phase extraction (SPE) and high performance liquid chromatography (HPLC) and the resultant fractions were tested for their anti-proliferative activity (data not published). Following this approach, a fraction with high bioactivity was obtained. Liquid chromatography coupled with mass spectrometry (LC-MS) analysis revealed that this fraction has a major peak which identification is ongoing.

The third stage of carcinogenesis is tumor progression, this stages involves metastasis and angiogenesis. Metastases are the cause of 90% of human cancer deaths [33]; the ability of a cancer cell to undergo migration and invasion allows tumor cells to escape from the primary tumor mass and spread to other parts of the body [34]. In addition, tumor growth relies on angiogenesis in order to receive an adequate supply of oxygen and nutrients. Tumors cannot grow more than 1–2 mm without blood supply for nutrient supply and waste removal [35]. GCL extract is effective in inhibition of breast tumor cell migration; however, metastasis involves other processes beyond the capacity of tumor cell to migrate. Further studies are needed to evaluate the in vivo effect of GCL extract on tumor cell metastization. Our results using the in vivo angiogenesis zebrafish model, point out GCL extract as a promising source of anti-angiogenic compounds. However, other experimental models should be use to evaluate the putative effect of Cc leaves extract on tumor induced angiogenesis.

Methanol, yeast extract and sodium acetate were purchased from Merck (Darmstadt, Germany). The Folin-Ciocalteu reagent, sodium carbonate, gallic acid, potassium ferricyanide, ferric chloride, dextrose, peptone, 2-tert-butyl-4-methoxyphenol (BHA), 2,2-diphenyl-1-picrylhydrazyl (DPPH), Paraformaldheydo (PFA) Trichloroacetic acid (TCA) were obtain from Sigma-Aldrich (St. Louis, MO, USA); Dulbecco’s modified Eagle’s medium (DMEM), glucose, L-glutamine, fetal bovine serum (FBS), penicillin-streptomycin mixture solution, trypsin (5 g·L⁻¹)—EDTA and trypan blue were purchased from Lonza (Verviers, Belgium).

Plants were collected at two different stages; during flowering (green plants) and after fructification (Cc plants dried naturally in field). Both wild (Cc var. sylvestris) and cultivated (Cc var. altilis) green Cc plants were collected in June 2009; dried plants were collected in August 2009. Wild plants were collected at Campo Experimental whereas cultivated plants were collected at Centro Hortofrutícola, both from Instituto Politécnico de Beja (south of Portugal). Green plants were storage at −80 °C; dried plants were kept in a dark dried container, at room temperature (RT) until used.

Twelve different extracts were prepared; green wild leaves (GWL), green cultivated leaves (GCL), green wild stalks (GWS), green cultivated stalks (GCS), green wild inflorescences (GWI), green cultivated inflorescences (GCI), dry wild leaves (DWL), dry cultivated leaves (DCL), dry wild stalks (DWS), dry cultivated stalks (DCS), dry wild inflorescences (DWI) and dry cultivated inflorescences (DCI). Dry (100 g) and fresh tissues (100 g) were cut into small pieces and left hydrating in water (100 g/L) overnight at RT. Thereafter, the water volume set to 400 mL in each recipient and 600 mL of methanol was added to achieve 60% methanol. Tissues were left in 60% methanol for 1 h, and then were homogenized with a blender and ultrasonicated (JP Selecta, Barcelona, Spain) for 5 min. Extracts in 60% methanol were then filtered using filter paper, followed by 0.2 μm filtration (VWR, vacuum filtration system 0.2 μm). The obtained extracts were completely evaporated (Laborota 4000 Efficient, Heidolph instruments GmbH, Schwaback, Germany) dissolved in water and stored at −80 °C until use. For dry matter (DM) determination, 10 g of fresh plants tissues was left in an oven at 70 °C and weight was recorded until stabilization. Measurements were made in triplicates. Concentrations for “green tissues extracts” were converted into dry matter. Extracts concentration was expressed in micrograms of Cc tissue DM per milliliter (μg Cc mL⁻¹ DM).

Total phenolic contents were determined using the Folin-Ciocalteu method [36]. Briefly, samples (100 μL) were mixed with Folin-Ciocalteu reagent (1.4 mL, diluted tenfold) and left for 5 min at RT. Then, 1.5 mL of sodium carbonate (60 g·L⁻¹) was added and the mixture was then incubated at RT in the dark for 60 min. The absorbance was read at 725 nm in a Double-beam UV-Vis scanning spectrophotometer (Helios alpha, Thermo scientific, Bremen, Germany).

For the calibration curve different concentrations of gallic acid (500, 400, 300, 200, 100, 50, 10 µg·mL⁻¹) were prepared in 60% methanol and treated as the samples. Total phenolic was expressed as milligrams of gallic acid equivalents per gram of DM (mg GAE g⁻¹ DM). All samples were analyzed in three replications.

The reducing power was determined according to a previous described procedure [28]. Briefly, a serial dilution of the extracts was performed (100, 50, 30, 15, 10, 5 and 1 μg of Cc mL⁻¹) and mixed with 2.0 mL of 200 mM sodium phosphate buffer (pH 6.6) followed by 2 mL of 1% potassium ferricyanide (w/v). The mixture was incubated in a water bath at 50 °C for 20 min. After incubation 2.0 mL of 10% TCA (w/v) was added and the mixture was centrifuged at 3000 rpm, for 10 min at 4 °C (Hermle Z323K, Hermle Labor Technik, Wehingen, Germany). The upper layer (1.0 mL) was mixed with 1.0 mL of 0.1% of ferric chloride (w/v), and the absorbance was measured spectrophotometrically at 700 nm in a Double-beam UV-Vis scanning spectrophotometer. Tests were carried out in triplicate. Increased absorbance of the reaction mixture indicates greater reducing power. The extract’s effective concentration (EC), providing 0.5 of absorbance (EC₅₀), was calculated from the graph of absorbance registered at 700 nm against the correspondent extract concentration.

The radical scavenging activity of Cc extracts against DPPH free radical was assessed according to previously described [37,38]. A stock methanolic solution of 6 × 10⁻⁵ M DPPH was prepared. A serial dilution of the extracts was performed (100, 50, 30, 15, 10, 5 and 1 μg of Cc mL⁻¹). Extract samples (300 μL) at different concentrations, a blank (containing methanol instead of DPPH solution) and a control (containing extracts solvent instead of the extract) were mixed with DPPH methanolic solution (2.7 mL). After mixing, they were left in dark for 60 min at RT. DPPH radical inhibition was measured at 517 nm against the blank. Tests were carried out in triplicate.

DPPH scavenging effect was calculated as percentage of DPPH discoloration using the following equation: DPPH scavenging activity (%) = [1 − (Abs of sample/Abs DPPH control)] × 100

The percentage of scavenging activity was plotted against the sample concentration to obtain the EC₅₀, defined as the concentration of sample necessary to cause 50% inhibition.

Saccharomyces cerevisiae were cultured in YPD nutrient medium (1% yeast extract; 2% dextrose; and 2% peptone) at 30 °C under continuous stirring (130 rpm) (Orbital Shaker incubator, TEQ, Massamá, Portugal) for 24 h. Cells were incubated at an optic density (OD)₆₀₀ of 0.3. Yeast was incubated with Cc extracts (GCL, GCS and GCF) in a concentration of 25 μg GAE mL⁻¹ during 1 h. After this incubation period 0.1 mM H₂O₂ was added to the incubation medium and OD₆₀₀ was immediately read (OD₆₀₀ at time 0), yeast growth was monitored measuring the OD₆₀₀ every hour for 16 h. In addition, cells were incubated with BHA (25 μg·mL⁻¹) under the same conditions; BHA was used as reference compound. Two distinct controls were preformed; a positive control in which cells grew in standard conditions (just YPD medium) and a negative control in which cells grew in YPD medium containing H₂O₂. Cellular growth was accessed plotting OD₆₀₀ against incubation time and GC₅₀ (time required to reach 50% of maximum growth) was calculated from the obtain plots. The assay was performed in three independent experiments.

MDA-MB-231 human breast cancer cell line was obtained from American Type Culture Collection, and maintained in DMEM supplemented with 10% heat-inactivated FBS and 1% penicillin-streptomycin mixture solution. The cells were incubated in an atmosphere of 95% air and 5% CO₂ at 37 °C (C150, Binder GmbH, Tuttlingen, Germany).

Cell viability was accessed using tetrazolium compound based CellTiter 96^® AQueous One Solution Cell Proliferation (MTS) assay (Promega). MDA-MB-231 cells were seeded in 96-well plates at 2 × 10⁵ cells mL⁻¹. After 24 h, cells were gently washed with 1X PBS, and then incubated with culture medium containing Cc extracts at various concentrations (10, 50, 100, 250, 500, 750, 1000 μg·mL⁻¹) for 48 h. After the exposure period MTS assay was performed according to the manufacturer’s instruction.

IC₅₀, defined as the extract concentration necessary to cause 50% inhibition of cell viability, was calculated plotting the percentage of cell viability compared to a control (no extract added) against the different Cc extracts concentration.

To evaluate cell motility, MDA-MB-231 cells (2.5 × 10⁵ cells mL⁻¹) were seeded in 6-well plates and grown to 80%–90% confluence. After aspirating the growth medium, a linear wound 1 mm wide was made by scratching the monolayer with a sterile 200 μL micropipette tip. Cellular debris was removed by gently washing twice with 1× PBS. MDA-MB-231 cells were incubated with GCL extract at 35, 70, 140 and 280 μg of Cc leave mL⁻¹ DM for 24 h. Wound healing of the cells was observed and photographed under phase-contrast microscope at time 0 and 24 h. Analyses were performed in triplicate. The area of the wound was measured in control and treated cells using the Motic Images Plus 2.0 software [39]. The percentage of wound healed was then calculated using the formula 100 (final area/initial area × 100%).

Adult zebrafish animals were maintained at 28 °C in a re-circulating water system (Tecnoplast, Milan, Italy). Zebrafish Tg(fli1:EGFP) parents crossed with wild-type (wt) parent pairs under 14 h:10 h (l:d) and made to spawn. Fertilized eggs were collected at 1-cell stage and incubated in E3 medium in Petri dishes at 28.5 °C. At 15 hours post-fertilization (hpf) [40] animals were screen and collected for green fluorescent protein (GFP) positive cells under a fluorescent SXZ 7 stereoscope system (Olympus, Tokyo, Japan).

Several dilutions of GCL extract were prepared in E3 medium as follows: 30, 15, 7.5, 1.5 and 0.15 mg GCL mL⁻¹ E3 medium. Five experimental groups (with GCL extract) and a control (E3 medium) were assembled in a 6-well cell culture plate (Stardstar, Berlin, Germany).

Seven GFP-positive animals previously selected were incubated at 28.5 °C in each experiment group and control until 30 hpf [40]. At this time, animals were fixed at room temperature (RT) for 1 h in 1% PFA/1× PBS pH 7. Animals were rinsed 3× in 1× PBS pH 7 and immediately imaged in a Zeiss (Jena, Germany) Z2 microscope system. The number of intersegmental vessels and arteries developing in the trunk was counted. These data were used to plot a dose response curve as previously described [41].

Each experiment was performed in triplicate. Results were expressed as means ± SD. Three-way analysis of variance (ANOVA; Sigma plot, San Jose, CA, USA) was used to test for differences between extracts activity (reduction capacity and DPPH scavenging activity) and total phenolic contents. One Way ANOVA was used to test for differences between extracts effect on cell migration capacity and cell viability). Where differences did exist, the source of the differences at a P < 0.05 significance level was identified by all pairwise multiple comparison procedure (Tukey test). Pearson Product Moment Correlation was applied to search for correlation between extracts activity and total phenols concentration.