Introduction

A priority in biotechnological innovation in the last decade has been the exploration of natural resources toward the discovery of new and promising bioactivities, e.g., with therapeutic potential or industrial applications (reviewed in Donia and Hamann 2003; Imhoff et al. 2011; Bonifácio et al. 2014). Surfactants for biomaterials are of particular interest to improve biocompatibility and efficacy, and in this regard, phytochemicals were found to have antioxidant capacities and also to promote osteoblast proliferation and differentiation (Habauzit and Horcajada 2007; Woo et al. 2010; Karadeniz et al. 2014, 2015), which make them suitable for applications related to bone regeneration or increased bone density (Córdoba et al. 2015a,b; Watson and Schönlau 2015). Marine macroalgae, beside their role in the provision of alginates, carrageenans, fibers, and minerals for agrifood (Marsham et al. 2007; Mouritsen 2013; Stiger-Pouvreau et al. 2016), also represent a valuable source for bioactive phytochemicals (reviewed in Imhoff et al. 2011; Stengel and Connan 2015). While macroalgae are largely under-exploited in Europe, they are commonly harvested in Brittany for industrial applications such as cosmetic, food, thalassotherapy, and medicinal products (Bourgougnon and Stiger-Pouvreau 2011). Brown algae are by far the most harvested species (97.2 %), followed by red (2.7 %) and green (0.06 %) algae. Among the green algae, species of the genus Ulva and Enteromorpha are mainly valorized in Brittany (Chambre Syndicale des Algues et Végétaux Marins (CSAVM), personal communications). Phenolic compounds from marine and terrestrial plants exhibit a wide range of chemical structures (Singh and Bharate 2006; Ajila et al. 2011) and a large spectrum of biological activities (Li et al. 2011; Ksouri et al. 2012). For example, they showed antidiabetic (Lee and Jeon 2013), antimicrobial (Eom et al. 2012), antioxidative (Le Lann et al. 2008, 2012; Andrade et al. 2013; Tanniou et al. 2013, 2014; Stiger-Pouvreau et al. 2014; Surget et al. 2015), photoprotective (Surget et al. 2015), antitumoral (Deslandes et al. 2000; Zubia et al. 2009; Montero et al. 2016), anti-inflammatory (Kim et al. 2009; Kang et al. 2013), and radioprotective effects (Liu et al. 2011). Properties of bioceramic-based medical scaffolds have been successfully improved upon supplementation with phenolic compounds from marine origin, in particular toward increased bone tissue regeneration (Yeo et al. 2012; Córdoba et al. 2015a,b). Phlorotannins, i.e., phenolic compounds from brown seaweeds, extracted from Ecklonia brown algal species have also been shown to increase alkaline phosphatase (ALP) activity, mineralization, total protein, and collagen synthesis in human osteosarcoma cells (MG-63 cells) (Ryu et al. 2009; Ali and Hasan 2012; Yeo et al. 2012; Karadeniz et al. 2015), suggesting that they may act as regulators of osteoblast differentiation and bone formation. In the context of the biotechnological valorization of marine algae in Europe and the discovery of new bioactive molecules as potential biomaterials for regenerative medicine, this work intended to get insights into the osteogenic and antioxidant capacities of the phenolic compounds extracted from two green seaweeds, the native Cladophora rupestris and the introduced Codium fragile. In that sense, crude extracts and semi-purified fractions were evaluated for total phenolic content, antioxidative activity, and mineralogenic and proliferative effects, in this case using a bone-derived cell line from gilthead seabream (Sparus aurata) capable of extracellular matrix mineralization (Pombinho et al. 2004). The most active fractions were then tested in vivo, using the zebrafish as a model system to assess mineralization (Laizé et al. 2014).

Material and methods

Regular reagents and materials

Solvents and chemicals (analytical grade) were from Carlo Erba Reagent and Sigma-Aldrich, respectively, unless otherwise stated. Cell culture reagents and plastic consumables were from Invitrogen and Sarstedt, respectively.

Biological material

Samples of the green seaweeds Codium fragile (Suringar) Hariot (Ulvophyceae, Bryopsidales, Codiaceae) and Cladophora rupestris (Linnaeus) Kützing (Ulvophyceae, Cladophorales, Cladophoraceae) were collected in rock pools of Porsmeur Bay (Lanildut, France; 48° 28′ 55.67″ N, 4° 46′ 14.76″ E). Codium fragile is an introduced species in Europe, with a worldwide spread. Cladophora rupestris is a native species in Brittany, widely distributed in temperate and cold temperate Atlantic Ocean. After collection and removal of epiphytes, seaweeds were washed with deionized water, frozen, freeze-dried, ground into powder, and stored in the darkness at room temperature to limit the phenolic compound degradation, as described in Le Lann et al. (2008).

Extraction and liquid-liquid semi-purification process

Fifteen gram of dry powder was macerated for 2 h at 40 °C in 150 mL of hydroethanolic solvent (1:1 distilled water/100 % ethanol mixture) under agitation (200 rpm) in the dark. After centrifugation (3000 rpm, 15 °C, 10 min), supernatant was collected and pellet was resuspended in 150 mL of hydroethanolic solvent and macerated for an additional 1 h. This last step was repeated once more, and the three supernatants were pooled and passed through glass fiber filters. Ethanol was removed at 40 °C by vacuum evaporation, and distilled water was added to obtain a 100-mL crude extract (CE). Crude extract was separated into two semi-purified extracts, an aqueous fraction (AF) and an ethyl acetate fraction (EAF), using a protocol adapted from Stiger-Pouvreau et al. (2014) to concentrate polyphenols in EAF. Briefly, lipidic compounds and chlorophyllic pigments were removed through several washes with dichloromethane; then, proteins and carbohydrates were precipitated using pure acetone and 100 % ethanol. Purified extract was then divided into ethyl acetate and aqueous fractions.

Total phenolic content

Total phenolic content (TPC) was determined by spectrophotometry using Folin-Ciocalteu assay (Sanoner et al. 1999) modified according to Le Lann et al. (2008). Wells of a 96-well plate were successfully filled with 20 μL of extracts (dilution ranging from 1 to 0.01 g L−1), 130 μL of distilled water, 10 μL of Folin-Ciocalteu reagent, and 40 μL of sodium carbonate solution (200 g L−1). After agitation, plate was incubated at 70 °C for 10 min, then placed on ice to stop the chemical reaction. Absorbance was measured at 620 nm using a Multiskan MS plate reader (LabSystems), and phenolic content was determined using a standard curve of gallic acid ranging from 0 to 200 μg mL−1. Results are expressed as milligram of gallic acid per gram of dried weight (DW). All measurements were performed in triplicate.

Antioxidant assays

2,2-Diphenyl-1-picrylhydrasyl radical scavenging activity

Radical scavenging activity of extracts/fractions and positive controls was determined using the 2,2-diphenyl-1-picrylhydrasyl (DPPH) assay adapted from Le Lann et al. (2008) and Zubia et al. (2009). This assay tests the capacity of samples to scavenge the synthetic radical DPPH but is independent of sample polarity and highly reproducible (Huang et al. 2005). Briefly, 100 μL of extract/fraction, positive control (ascorbic acid, butylated hydroxyanisole or BHA, 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid or Trolox), or negative controls (distilled water and 100 % ethanol) were added to 100 μL of DPPH solution (36.9 mg L−1) in a 96-well plate. Several dilutions of extracts/fractions and positive controls (ranging from 0.005 to 1 g L−1) were tested. Plate was incubated for 1 h in the dark, and absorbance was measured at 540 nm. Radical scavenging activity of algal extracts was calculated as previously described in Le Lann et al. (2008) and Surget et al. (2015). For each sample, a curve of extract concentration against percent of DPPH inhibition was generated to determine the concentration of extract needed to cause a 50 % reduction of the initial DPPH concentration (IC50 in g L−1). A high IC50 is indicative of a weak radical scavenging activity and vice versa. All measurements were performed in triplicate.

Reducing power

Reducing power (RP) of seaweed extracts was determined following the method in Zubia et al. (2009). RP assay implies an electron transfer and tests the capacity of antioxidants to reduce Fe3+/ferricyanide complex to the ferrous form (Huang et al. 2005). A volume of 25 μL of extracts/fractions, negative controls (water and ethanol), or positive controls (ascorbic acid, BHA, Trolox) was mixed with 25 μL of sodium phosphate buffer (0.2 M; pH 6.6) and 25 μL of potassium ferricyanide (1 %; w/v). Mixture was incubated for 20 min at 50 °C. After cooling on ice, 25 μL of 10 % (w/v) trichloroacetic acid, 100 μL of distilled water, and 20 μL of 0.1 % (w/v) FeCl3 were added, and mixture was incubated at room temperature for 10 min. Absorbance was measured at 620 nm. Capacity of extracts/fractions to reduce Fe3+ was determined from EC50 in g L−1 (calculated by interpolation of a regression linear curve) and corresponds to the effective concentration of the sample to obtain an optical density equal to 0.5. All measurements were performed in triplicates.

Cultures of bone-derived cell line VSa13

Cell culture maintenance

Gilthead Seabream bone-derived cell line VSa13 (Pombinho et al. 2004) was used to evaluate mineralogenic and proliferative capacities of seaweed extracts. Cells were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10 % fetal bovine serum (FBS), 1 % penicillin–streptomycin, 1 % L-glutamine, and 0.2 % fungizone and incubated at 33 °C in a 10 % CO2-humidified atmosphere (Marques et al. 2007). Cells were sub-cultured 1:4 twice a week using trypsin-EDTA solution (0.2 % trypsine, 1.1 mM EDTA, pH 7.4).

Cell exposure to seaweed molecules

Dry extracts/fractions from green macroalgae were dissolved in distilled water, 50 % ethanol (in distilled water), or 100 % ethanol according to their polarity to prepare ×1000 stock solutions. CE, EAF, and AF stock solutions were added to the culture medium to achieve concentrations of 0.05, 0.5, 5, 50, 100, or 250 μg mL−1 for EAF and 0.5, 5, 50, or 100 μg mL−1 for CE and AF. Medium supplemented with extracts/fractions was 0.2 μm filtered before applied to the cell culture.

Cytotoxicity assay

Cells were seeded in 96-well plates at a density of 104 cells well−1 and further cultured until confluence. Culture medium was replaced with fresh medium containing either the vehicle (control) or algal extracts and renewed every 3–4 days. Cell viability was assessed after 9- and 18-day exposure using the Cell Proliferation Kit XTT (AppliChem). Viability of the cells exposed to seaweed extracts was calculated as percentage of survival in comparison with cells cultured in respective controls.

Proliferation assay

Cells were seeded in 96-well plates at a density of 1.5 × 103 cells well−1. After 24 h, culture medium was replaced with fresh medium containing either the vehicle (control) or seaweed extracts and renewed every 3–4 days. Cell proliferation was determined after 9 days of exposure using the Cell Proliferation Kit XTT (AppliChem). Results are presented as percentage of cell proliferation, in comparison with respective controls.

Extracellular matrix mineralization

Cells were seeded in 24-well plates at a density of 5 × 104 cells well−1. Extracellular matrix (ECM) mineralization was induced in confluent cultures by supplementing culture medium with 50 μg mL−1 L-ascorbic acid, 10 mM β-glycerophosphate, and 4 mM calcium chloride (differentiation medium). Seaweed extracts or vehicles were added to differentiation medium and renewed twice a week. After 17 days of culture, mineral deposition was revealed through alizarin red S (AR-S; Sigma-Aldrich) staining and quantified by spectrophotometry (Stanford et al. 1995). Results are expressed as percentage of ECM mineralization, relative to the respective controls.

Quantification of in vivo mineralization

Zebrafish larvae were exposed to increasing concentrations of EAF fractions of green macroalgae, C. fragile (1, 5, and 10 μg mL−1) and C. rupestris (5, 50, and 100 μg mL−1), and control (ethanol), from 9 to 11 days post-fertilization (dpf). Briefly, from 9 to 11 dpf, larvae were reared in a 24-well dish (2 larvae well−1) placed in the dark, to avoid photodegradation of the compounds, at 28.5 °C. Each well was filled with 1 mL of embryo medium (Westerfield 2007) supplemented with EAF, the most active fraction. Medium was renewed once a day, and larvae were fed with Artemia nauplii (EG strain, INVE) twice a day. After 48-h exposure, larvae were euthanatized with 168 μg mL−1 of tricaine (MS-222; Sigma-Aldrich) and fixed in 4 % paraformaldehyde in phosphate-buffered saline (PBS) for 16 h at 4 °C. After two washes with PBS, larvae were stained with 0.01 % alizarin red S solution for 30 min, washed twice with distilled water, and maintained in 25 % glycerol until analysis. AR-S fluorescence in zebrafish larvae was imaged under a MZ75 stereomicroscope (Leica) equipped with a green filter (excitation filter 546/10 nm, barrier filter 590 nm) and a F-view camera (Olympus). Mineralized area of the operculum was determined from the morphometric analysis of the fluorescence images using ImageJ software (National Institutes of Health) and normalized with the area of the head. Zebrafish larvae were also exposed to 1 fg mL−1 of vitamin D (calcitriol; Sigma-Aldrich) as a positive control.

Statistical analysis

Results are presented as the mean of at least three replicates with standard deviation (SD). Homogeneity of variance was tested with the Bartlett’s test at the 0.05 significance level. As the sample panel was small for the in vitro tests, a Kruskal-Wallis test was performed to highlight potential significant difference. In the case of significant difference between the data, non-parametric multiple comparisons (Behrens-Fisher test) were tested using the nmpc package. Statistical analysis of ECM mineralization assay and in vivo assay was performed with GraphPad Prism 4 (GraphPad, USA) using one-way ANOVA followed by Dunnett’s or Tukey’s multiple comparisons tests, respectively, to determine statistical differences among groups. Differences were considered statistically significant for p < 0.05.

Results

Total phenolic contents and antioxidant activities

TPC was determined in the CE and semi-purified fractions (EAF and AF) of C. fragile and C. rupestris (Table 1). While all the three fractions prepared from C. rupestris presented high phenolic contents (ranging from 15.833 ± 0.103 to 21.726 ± 0.899 mg GAE g−1 DW; Table 1) with EAF showing the highest TPC, only EAF prepared from C. fragile exhibited a high TPC value, 10-fold higher than CE content (22.381 ± 0.206 and 2.202 ± 0.103 mg GAE g−1 DW, respectively; Table 1). AF prepared from C. fragile exhibited the lowest TPC (0.298 ± 0.103 mg GAE g−1 DW; Table 1), indicating an absence of aromatic compounds in this fraction. Antioxidant activities of each extract and those from the three positive controls (Trolox, BHA, and ascorbic acid) were determined through the measurement of DPPH radical scavenging activity and reducing power and expressed as IC50 and EC50, respectively. In both assays, antioxidant activity was significantly higher in the ethyl acetate fractions of both species when compared to crude extracts and aqueous fractions (Table 1). Even though, EAF antioxidant activities were lower than the positive controls tested. The lowest antioxidant activity (i.e., high IC50 and EC50) was always found in AF. Our data indicates that highest antioxidant activities were associated with EAF, which have the highest phenolic content, whereas lowest antioxidant activities were associated with aqueous fractions, exhibiting the lowest phenolic content.

Table 1 Phenolic content and antioxidant activities of Cladophora rupestris and Codium fragile extracts. Phenolic content, DPPH scavenging activity, and reducing power were determined for crude extracts and semi-purified fractions and commercial standards (ascorbic acid, BHA, and Trolox). Values are expressed as mean ± SD (n = 3)
Full size table

Proliferative and mineralogenic activities of seaweed extracts

Seaweed extracts were further characterized for their capacity to alter cell proliferation and ECM mineralization (Figs. 1 and 2). Cytotoxicity of the extracts/fractions was first evaluated to determine non-toxic concentrations to be tested in subsequent assays. In that sense, cell viability was evaluated at 9 days (endpoint to assess cell proliferation) and 18 days (endpoint to assess extracellular matrix mineralization) using a wide range of extract/fraction concentrations and presented as cell survival rates. Survival rates were not concentration dependent and always above 80 % for all samples/concentrations at 9 days (Fig. 1b) and above 75 % at 18 days (Fig. 2b). All concentrations were further used to evaluate proliferative and mineralogenic activities of the extracts.

Fig. 1
figure 1

Effect of crude extract and semi-purified fractions of Codium fragile and Cladophora rupestris on VSa13 cell proliferation (a) and survival (b) at 9 days. Control values were determined in vehicle-treated cells and set to 100 % (dotted line). Proliferation and survival data are presented as mean values ± standard deviation, n > 5. The asterisks indicate the values significantly different from the control value according to Berhens Fisher’s test (**p < 0.01, ***p < 0.001). CE crude extract, EAF ethyl acetate fraction, AF aqueous fraction

Full size image
Fig. 2
figure 2

Effect of crude extract and semi-purified fractions of Codium fragile and Cladophora rupestris on VSa13 extracellular matrix (ECM) mineralization (a) and survival (b) at 18 days. Control values were determined from vehicle-treated cells and set to 100 % (dotted line). Mineralization and survival data are presented as mean values ± standard deviation, n = 4 (a) and n = 5 (b). The asterisks indicate the values significantly different from the control value according to Berhens Fisher’s test (**p < 0.01, ***p < 0.001). CE crude extract, EAF ethyl acetate fraction, AF aqueous fraction

Full size image

Effect of seaweed molecules on cell proliferation

For both C. fragile and C. rupestris, none of the extracts/fractions had a pro-proliferative effect at the concentrations tested (Fig. 1a). At the highest concentrations (50 and 100 μg mL−1), the AF of C. rupestris presented a statistically significant negative effect of 10.7 and 13.6 %, respectively, compared to the control (Fig. 1a), likely indicating a toxic effect of the highest doses tested, also supported by the fact that cell survival was also negatively affected with the same doses (Fig. 1b).

Effect of the EAF fractions on ECM mineralization

Alizarin red S staining revealed that mineralization of VSa13 extracellular matrix was strongly increased (80 % over control cultures) upon chronic exposure to the highest concentration (50 μg mL−1) of both ethyl acetate fractions (Fig. 2a). In the case of C. rupestris EAF, pro-mineralogenic effect increased with concentration indicating a dose dependency. Aqueous fraction of C. fragile (50 μg mL−1) significantly increased mineral deposition (30 % over control cultures), but no increase was observed at higher concentration, indicating that this result might not be of biological significance. To further validate the pro-mineralogenic effect of ethyl acetate fractions, new EAFs prepared from the same algal powder were prepared and evaluated for their capacity to increase ECM mineralization using an increased range of concentrations 5, 50, 100, and 250 μg mL−1. Highest concentrations of C. fragile EAF (100 and 250 μg mL−1) were toxic to the cells (data not shown) and therefore were not tested for mineralogenic activity. At 5 and 50 μg mL−1, new EAF of C. fragile induced mineral deposition by 50 and 370 %, respectively, when compared to the control, indicating a dose-dependent effect (Fig. 3a). Similarly, new EAF of C. rupestris stimulated ECM mineralization in a dose-dependent manner, reaching a 630 % increase over the control at the highest concentration (250 μg mL−1) (Fig. 3b). These data further confirmed the mineralogenic potential of ethyl acetate fractions prepared from both green macroalgae. For the same concentrations, the effect was higher in the second assay (recent extracts).

Fig. 3
figure 3

Effect of ethyl acetate fraction (EAF) of Codium fragile (a) and Cladophora rupestris (b) on VSa13 extracellular matrix (ECM) mineralization. Pictures of cell cultures treated with either the vehicle (V) or different EAF concentrations and stained with alizarin red S are shown above each graph. Level of ECM mineralization in vehicle-treated cell cultures was set to 100 % (dotted line). Mineralization data are presented as mean values ± standard deviation, n = 4. The asterisks indicate the values significantly different from the vehicle value (one-way ANOVA followed by Dunnett’s post hoc test; *p < 0.05, **p < 0.01). N.D. not determined

Full size image

In vivo effect of the ethyl acetate fractions on bone-mineralized area

To further study the osteogenic action of both phenolic-enriched seaweed extracts, zebrafish larvae were exposed to ethyl acetate fractions, and in vivo bone formation and mineralization were assessed through the histomorphometric analysis of the operculum of exposed versus control larvae. After 48 h of exposure, larvae were immersed in alizarin red S, a calcium-specific dye, to stain mineralized bone structures (Gavaia et al. 2000; Bensimon-Brito et al. 2016), which at this age (11 dpf) are mainly localized in the head. The operculum is among the first structures to develop and also straightforward to identify (Laizé et al. 2014), and its area was measured to assess EAF osteogenic activity. As expected from previous studies (Fleming et al. 2005), vitamin D (used here as a positive control) increased the mineralized area of zebrafish operculum approximately 1.5-fold over vehicle, therefore validating our experiment (Fig. 4). Ethyl acetate fractions of both C. fragile and C. rupestris also increased the mineralized area of zebrafish operculum (Fig. 4). At 1, 5, and 10 μg mL−1, C. fragile EAF significantly increased the mineralized area up to approximately 1.8-fold change over the vehicle (Fig. 4a). Higher concentrations were tested but were lethal to the larvae (data not shown). At 50 and 100 μg mL−1, C. rupestris EAF significantly increased the mineralized area of zebrafish operculum, up to approximately 1.6-fold change over the vehicle (Fig. 4b). Our data shows that both ethyl acetate fractions enriched in phenolic compounds increased the mineralized area of zebrafish operculum at similar or higher levels than the positive control vitamin D.

Fig. 4
figure 4

Effect of ethyl acetate fraction (EAF) of Codium fragile (a) and Cladophora rupestris (b) on the mineralization of zebrafish operculum area. Larvae at 9 days post-fertilization were exposed for 48 h to increasing EAF concentrations, to vehicle (V), or to 1 fg mL−1 of vitamin D (positive control). Mineralization data are presented as mean values ± standard deviation, n ≥ 3. The asterisks indicate the values significantly different from the vehicle value (one-way ANOVA followed by Tukey’s post hoc test; *p < 0.05, **p < 0.01, ***p < 0.001)

Full size image

Discussion

Marine macroalgae represent an under-exploited source of phenolic compounds with high potential for therapeutic or industrial applications. In this work, we explored the bioactive potential of extracts prepared from two marine Ulvophyceae, C. fragile and C. rupestris, focusing on antioxidant and osteogenic activities. The fractionation process proved to be efficient, since the phenolic content was higher in EAF from both species, although this process seems to depend on the species studied, since the aqueous fraction from C. rupestris, but not from C. fragile, presented residual levels of phenolic compounds. The difference in the polarity of phenolic compounds produced by each species could explain the differences that occurred on each fraction. To further characterize the extracts/fractions prepared from green macroalgae, their antioxidant capacity was evaluated through two simple, fast, reliable, and classical biochemical methods—the DPPH and reducing power tests—both measuring electron transfers (Huang et al. 2005). As antioxidant molecules can act through several and complex mechanisms (Frankel and Meyer 2000), the combination of two antioxidant tests allowed us to better highlight the in vivo complexity of interactions among antioxidants in foods and biological systems (Frankel and Meyer 2000). Our results revealed that independently of the assay or the species considered, EAF fraction exhibited the highest in vitro antioxidant activities, which was also the fraction with highest content in phenolic compounds (Table 1), in agreement with previous studies indicating the antioxidative properties of phenolic compounds (Fujii et al. 2013; Tanniou et al. 2013; Stiger-Pouvreau et al. 2014; Surget et al. 2015). Since antioxidants counteract oxidative stress promoted by the increase of reactive oxygen species (ROS) and reducing power assay depends on the redox (reduction-oxidation) potential of the compounds present in a sample, we propose that phenolic compounds present in the marine macrophyte EAF are probably highly efficient antioxidants against ROS (e.g., peroxyl or hydroxyl radicals) because of their low redox potential as proposed by Zhu et al. (2002). Moreover, oxidative stress plays a key role in bone cell function (reviewed in Wauqier et al. 2009), and recent studies have evidenced bone mineralization/remodeling as one of the many biological processes influenced by natural compounds with antioxidant activity (Karadeniz et al. 2015; Córdoba et al. 2015a,b; Watson and Schönlau 2015). In agreement, our in vitro data clearly demonstrated the mineralogenic properties of C. fragile and C. rupestris EAF, the fractions with higher phenolic content and higher antioxidant activity. In vitro assays on VSa13 cells exposed to all the extracts and purified fractions evidenced that only EAF from both species increased ECM mineralization, probably by promoting cell differentiation since no proliferative effect was observed. A pro-mineralogenic effect of phenolic compounds from higher plants was previously shown in mouse osteoblastic MC3T3-E1 cells by promoting calcium deposition alone (Hagiwara et al. 2011) or in combination with increased cell viability, alkaline phosphatase activity, and collagen synthesis (Ding et al. 2010). Similar pro-mineralogenic effects were obtained with phlorotannins on the human fetal lung fibroblast MRC-5 cells (Ryu et al. 2009) or human osteosarcoma MG-63 cells (Ryu et al. 2009; Yeo et al. 2012). Several in vitro studies also evidenced the role of antioxidant molecules in restoring bone formation affected by oxidative stress through the inhibition of osteoblast differentiation (reviewed in Wauquier et al. 2009). In this regard, ROS were found to inhibit osteoblast ECM mineralization by inducing an alteration of osteogenic gene expression and an up-regulation of antioxidant systems (Mody et al. 2001; Arai et al. 2007). Also, Rao et al. (2008) discussed the stimulatory effect of polyphenol complex on osteoblast activity concomitant with a decrease of hydrogen peroxide. This suggests that compounds exhibiting antioxidant capacities could be effective by reducing tissues and intracellular ROS levels in prevention of bone diseases like osteoporosis (Wauqier et al. 2009; Rao and Rao 2015).

To get more insights on the high potential of C. fragile and C. rupestris phenolic compounds as pro-mineralogenic/osteogenic phytochemicals for biomedical applications, we confirmed our in vitro data through an in vivo assay based on developing zebrafish larvae. Indeed, the mineralized area of the operculum (a dermal bone derived from neural crest cells; Eames et al. 2012) was increased in fish exposed to EAF fractions, at similar levels promoted by vitamin D (positive control), indicating a pro-osteogenic activity of the compounds present in both C. fragile and C. rupestris phenolic-enriched fractions, validating our in vitro data, and reinforcing the pro-mineralogenic and osteogenic activities of these fractions. Importantly, these data also evidences the effect of the antioxidant phenolic-enriched fractions here studied in an in vivo system and their ability to impact on bone formation, an important feature for future biomedical applications. In fact, osteoporotic patients exhibit lower bone mineral density associated with higher oxidative stress (Almeida et al. 2007), whereas the activity of the main antioxidant enzymes is diminished (Ozgocmen et al. 2007). Mice models also showed that treatment with antioxidant enzymes inhibits bone loss (Lean et al. 2005). Recently, several studies have highlighted the therapeutic use of natural antioxidants, e.g., as dietary supplements in bone diseases such as osteoporosis or as nanocoatings to functionalize biomaterials and improve for instance current bone implants (Córdoba et al. 2015a,b; Karadeniz et al. 2015; Watson and Schönlau 2015). Thus, alternative approaches in the treatment of bone-related diseases are still needed and the potential of combined polyphenols, minerals, and vitamins which act in increasing the antioxidant capacity of the complex and thus reducing the oxidative stress correlated to bone loss as been suggested as a viable option (Rao and Rao 2015).

As a conclusion, the ethyl acetate fraction, which is rich in phenolic compounds, evidenced high antioxidant activity and high mineralogenic effect, promoting ECM mineralization of osteoblast-like cells in vitro and increasing the mineralized area of the zebrafish operculum in vivo. Thus, this study emphasized the biological potential of marine macroalgae and especially the use of invasive species, such as C. fragile, for future biomedical applications related to the improvement of bone status. To our knowledge, this is the first report of pro-mineralogenic activity in extracts/fraction prepared from two green macroalgae C. fragile and C. rupestris. Moreover, we demonstrated also the suitability of using zebrafish as a model for the screening of osteogenic and mineralogenic natural products. Purified extracts generated in our study and exhibiting multiple biological activities present interesting potential for industrial applications, especially in nutraceutics to prevent bone loss but also in medicine, as for example, to complement existing treatments of post-menopausal osteoporosis (Wauquier et al. 2009; Rao and Rao 2015).

Even though the identification of the active phenolic compound(s) responsible for EAF fraction activities was not the objective of this study, compounds such as bromophenols, coumarin, or derivative from vanillic acids are known to be produced by green algae (Stengel et al. 2011) and could be the pro-mineralogenic compounds in these extracts. In that regard, future studies should be conducted to isolate, identify, and test separately the bioactive compounds present in EAF of both Ulvophyceae, C. fragile and C. rupestris, in order to better understand their mechanisms of action.