Extraction and Yield Optimisation of Fucose, Glucans and Associated Antioxidant Activities from Laminaria digitata by Applying Response Surface Methodology to High Intensity Ultrasound-Assisted Extraction

The objectives of this study were to employ response surface methodology (RSM) to investigate and optimize the effect of ultrasound-assisted extraction (UAE) variables, temperature, time and amplitude on the yields of polysaccharides (fucose and total glucans) and antioxidant activities (ferric reducing antioxidant power (FRAP) and 1,1-diphenyl-2-picryl-hydrazyl radical scavenging activity (DPPH)) from Laminaria digitata, and to explore the suitability of applying the optimum UAE conditions for L. digitata to other brown macroalgae (L. hyperborea and Ascophyllum nodosum). The RSM with three-factor, four-level Box-Behnken Design (BBD) was used to study and optimize the extraction variables. A second order polynomial model fitted well to the experimental data with R2 values of 0.79, 0.66, 0.64, 0.73 for fucose, total glucans, FRAP and DPPH, respectively. The UAE parameters studied had a significant influence on the levels of fucose, FRAP and DPPH. The optimised UAE conditions (temperature = 76 °C, time = 10 min and amplitude = 100%) achieved yields of fucose (1060.7 ± 70.6 mg/100 g dried seaweed (ds)), total glucans (968.6 ± 13.3 mg/100 g ds), FRAP (8.7 ± 0.5 µM trolox/mg freeze-dried extract (fde)) and DPPH (11.0 ± 0.2%) in L. digitata. Polysaccharide rich extracts were also attained from L. hyperborea and A. nodosum with variable results when utilizing the optimum UAE conditions for L. digitata.


Introduction
Macroalgae are a diverse group of organisms capable of adapting to the extreme marine environmental conditions by producing multiple bioactive compounds. Marine macroalgae are considered a rich source of micro-and macronutrients with antioxidant activities, i.e., minerals, carotenoids, phenolic compounds, proteins and polysaccharides [1].
Macroalgal polysaccharides, particularly fucoidan and laminarin have a wide range of biological activities such as antioxidant, immunostimulatory and anti-microbial both in vitro and in vivo [2,3]. Fucoidans are a family of sulphated fucose-rich polysaccharides, built on a backbone of

Modelling the Extraction of Polysaccharides and Antioxidant Activity
The matrix design and the experimental responses (fucose, total glucans, FRAP and DPPH) for each run are presented in Table 1. There was considerable variation in the results obtained across the different parameters with ranges for: fucose (900.6 to 1257.7 mg/100 g ds), total glucans (774.9 to 1014.4 mg/100 g ds), FRAP (6.6 to 15.1 µM trolox/mg fde) and DPPH (9.8 to 15.3%). The highest yields of fucose (1257.7 mg/100 g ds) were obtained at an ultrasonic amplitude of 70%, at 80 • C for 30 min, while total glucans were maximum (1014.4 mg/100 g ds) using higher amplitudes (100%) and lower temperature and time during UAE (60 • C for 10 min). The antioxidant activities of FRAP (15.28 µM trolox/mg fde) and DPPH (15.10%) were optimum using milder UAE conditions (40% ultrasonic amplitude at 40 • C for 20 min).
A second order polynomial model fitted well to the experimental data ( Table 2) with low standard error and regression co-efficient (R 2 ) values of 0.79, 0.66, 0.64, 0.73 for fucose, total glucans, FRAP and DPPH, respectively. The ANOVA for the response surfaces of Table 2 identified that the linear models were significant for fucose (p < 0.05) and showed a tendency towards significance for FRAP response (p < 0.1). The quadratic model identified a tendency towards significance for DPPH (p < 0.1) and the interactions or cross-products among the extraction parameters studied were non-significant. Thus, only the linear and the quadratic effects of the independent factors were significant on the response surfaces in the current experimental design.
The significance of the three experimental variables affecting the extraction of polysaccharides and antioxidant activity of extracts generated from L. digitata can be determined from the model coefficients, multiple determinations and probabilities generated from the response surface regression (RSREG) and evaluated using ANOVA analysis ( Table 3). The regression coefficients provided in Table 3 indicate the effect of every parameter on the experimental responses; the magnitude of the coefficients is related to the weight of its effect and the positive and negative signs indicate an increase and decrease in the experimental responses, respectively. The time (β2) and amplitude (β3) of extraction significantly affected (p < 0.05) and tended to influence (p < 0.1), respectively, the antioxidant power of the seaweed extracts measured as DPPH. The quadratic effect of amplitude (β33) significantly influenced (p < 0.05) the DPPH radical scavenging activities of the extracts, while the temperature (β11) tended to (p < 0.1) influence the levels of fucose. No significant interactions or cross-products were appreciated for any experimental response, with tendencies to significance (p < 0.1) in the case of total glucans (time and amplitude, β23) and DPPH (temperature and time, β12).
Mar. Drugs 2018, 16, x FOR PEER REVIEW The units of the experimental responses are expressed as follow total glucans (mg/100 g dried seaweed), FRAP (µM trolox/mg DPPH (% radical scavenging activity). The standard errors of the in parentheses. Ϸ Estimated coefficients of the model are: β0 (co coefficients (β1, β2 and β3), quadratic (β11, β22 and β33) and int model referred to the variables X1 (temperature), X2 (time) and 0.05. a Tendency towards significance at p < 0.1.
Estimated coefficients of the model are: β0 (constant coefficient), linear regression coefficients (β1, β2 and β3), quadratic (β11, β22 and β33) and interaction (β12, β23, β13) effects of the model referred to the variables X1 (temperature), X2 (time) and X3 (amplitude). * Significant at p < 0.05. a Tendency towards significance at p < 0.1. Furthermore, contour plots (2D) and response surface plots (3D) were generated from the model equations to visualize the relationship between the UAE variables of extraction and the yields of fucose, total glucans and the antioxidant activities (FRAP and DPPH) of extracts from L. digitata (see Figure 1). These figures provided a visual interpretation of the mutual interactions between the 3 extraction variables and the expected responses (fucose, total glucans, FRAP and DPPH). Each graphic represents the effect of 2 extraction variables on the experimental response when the non-represented extraction variable is kept at its maximum; thus, these figures are a useful tool to predict the optimum extraction conditions and predicted values of polysaccharides and their related antioxidant activities.
Furthermore, contour plots (2D) and response surface plots (3D) were generated from the model equations to visualize the relationship between the UAE variables of extraction and the yields of fucose, total glucans and the antioxidant activities (FRAP and DPPH) of extracts from L. digitata (see Figure 1). These figures provided a visual interpretation of the mutual interactions between the 3 extraction variables and the expected responses (fucose, total glucans, FRAP and DPPH). Each graphic represents the effect of 2 extraction variables on the experimental response when the non-represented extraction variable is kept at its maximum; thus, these figures are a useful tool to predict the optimum extraction conditions and predicted values of polysaccharides and their related antioxidant activities. Contour plots (2D) and response surface plots (3D) of (I) fucose (mg/100 g dried seaweed (ds)); (II) total glucans (mg/100 g ds); (III) FRAP (µM trolox/mg freeze-dried extract (fde)) and (IV) DPPH (%) extracted from Laminaria digitata as a function of (a) time to temperature (amplitude = 100%) (b) temperature to amplitude (time = 30 min) and (c) amplitude to time (temperature = 80 • C).

Optimization of the Extraction of Polysaccharides and Antioxidant Activity
The current study focuses on the extraction of both fucose and glucans together along with their antioxidant activity by optimizing time, temperature and amplitude. All extraction parameters were optimized by using a more powerful semi-industrial ultra-sonication device (power 500 W, 20 kHz), compared to a lab grade ultra-sonication device used in previous studies [7,20]. Optimum conditions were determined aiming to maximize the yields of (i) fucose (condition 1), (ii) total glucans (condition 2), (iii) antioxidant activities (FRAP and DPPH) (condition 3) and (iv) yield of polysaccharides and antioxidant activities combined (condition 4). The levels of the three independent parameters used in UAE (temperature, time and ultrasonication amplitude), together with the predicted values and the experimental results obtained from L. digitata extracts are summarized in Table 4. The predicted values of the theoretical model for the four optimum conditions described were confirmed with the experimental data with the exception of the FRAP values, which were lower than the predicted values in both conditions 3 and 4.
The optimum UAE extraction conditions to obtain high yields of fucose from L. digitata were temperature (80 • C), time (30 min) and ultrasonication amplitude (40%; condition 1; Table 4). There is some conflicting data in the literature with regard to the influence of these conditions on the yields of fucose. Previous studies using UAE did not identify an influence of time or amplitude on the fucose content of extracts from A. nodosum [20]. Our results suggested that temperature is a critical factor for getting higher yield of both fucose and glucans along with total antioxidant activity, which was neglected in previous studies [7,20]. Our results are in agreement with Ale et al. [22] wherein the temperature and time of extraction also had an influence on the extraction of fucose from Sargassum spp. using conventional extraction techniques, with optimum extraction conditions obtained at temperatures of 90 • C over a 4 h period. However, previous researchers optimizing UAE conditions to obtain bioactive compounds from plants identified an influence of temperature, time and various ultrasonication parameters (i.e., frequency and power) on the yields of polysaccharides [23,24].
The optimum UAE extraction conditions to obtain high yields of total glucans from L. digitata were temperature (52.5 • C), time (10 min) and ultrasonication amplitude (100%; condition 2; Table 4). High ultrasonication amplitudes were also required to recover glucans from mushroom by-products (Agaricus bisporus) with the highest yields of glucans obtained applying high ultrasonic amplitudes (100 µm) for 15 min, followed by 1 h of precipitation with ethanol [25]. A previous study carried out by Kadam et al. [7] using 0.1 M HCl showed an increased extraction of glucans from L. hyperborea and A. nodosum at 60% of ultrasonic amplitude for 15 min, although the optimization of the UAE parameters was not performed [7].
The mild extraction conditions needed to preserve the antioxidant activities (FRAP and DPPH) of extracts from L. digitata (temperature 40 • C, time 30 min and amplitude 40%; condition 3; Table 4) could be due to the antioxidant power of other thermolabile compounds that could be present in the crude extracts, such as proteins/peptides [1] and polyphenols [26,27]. In fact, previous studies optimizing UAE to achieve phenolic compounds from brown macroalgae (Hormosira banksia) obtained maximum phenolic contents using low temperatures (30 • C) at medium sonication power (60%) for 60 min [28].
The optimum conditions to obtain both high yields of polysaccharides and antioxidant activities were of temperature (76 • C), time (10 min) and ultrasonication amplitude (100%; condition 4; Table 4). To our knowledge there are no studies presented in the literature that aim to optimize the yields of polysaccharides and its antioxidant activities from any species of seaweed. The predicted values were expressed as 95% confidence intervals. b Experimental responses were expressed as mean ± standard deviation of the mean. Number of readings (n = 6). γ The units of the experimental responses are expressed as follows: fucose (mg/100 g dried seaweed), total glucans (mg/100 g dried seaweed), FRAP (µM trolox/mg freeze-dried seaweed extract) and DPPH (% radical scavenging activity).

Application of Optimal UAE Conditions in other Brown Macroalgae
The applicability of the four optimum conditions for L. digitata was subsequently explored to generate polysaccharide rich extracts from other brown macroalgae with commercial value (L. hyperborea and A. nodosum). The contents of fucose, total glucans and antioxidant activities (FRAP and DPPH) of extracts from L. hyperborea and A. nodosum using optimal UAE conditions are compiled in Table 5. L. hyperborea extracts had higher contents of total glucans and DPPH activities, being approximately 10 and 4 fold higher than the values obtained from L. digitata, respectively. A. nodosum extracts showed powerful antioxidant activities (FRAP and DPPH) when compared to both Laminaria species. Previous studies aiming the UAE of fucose and glucans from brown macroalgae achieved extracts containing 87.06 mg fucose/g from A. nodosum [20] and 5.29-6.24 mg glucans/100 mg from L. hyperborea and A. nodosum, although the antioxidant activity of these extracts was not reported [7]. Table 5. Experimental responses obtained using the optimized ultrasound-assisted extraction conditions in brown macroalgae (Laminaria hyperborea and Ascophyllum nodosum). The results are expressed as mean ± standard deviation of the mean (n = 6).

Ultrasound-Assisted Extraction (UAE)
The pre-treatment of the seaweed samples and the UAE process performed in this study is presented schematically in Figure 2

Ultrasound-Assisted Extraction (UAE)
The pre-treatment of the seaweed samples and the UAE process performed in this study is presented schematically in Figure 2   Ultrasound-assisted extraction (UAE) of the samples was performed using semi-industrial grade UIP500hdT ultrasonic processor (maximum nominal power 500 W, 20 kHz, Hielscher Ultrasound technology, Teltow, Germany). The extraction variables temperature ( • C), time (min) and amplitude (%) were adjusted according to the matrix design described in detail in Section 3.5. Each extraction condition was performed in duplicate, the seaweed residues were filtrated through Whattman ® number 3 (GE Healthcare, Buckinghamshire, UK) and the supernatants combined. The combined extracts were freeze-dried in an industrial scale freeze-drier (FD80 model 119, Cuddon Engineering, Blenheim, New Zealand), vacuum sealed and stored at −20 • C until further analysis.

Composition of the Macroalgal Extracts
All composition analyses were performed in duplicate. The fucose and total glucan concentrations of the macroalgal extracts were analysed together with their antioxidant activity (ferric reducing antioxidant power (FRAP) and 1,1-diphenyl-2-picryl-hydrazyl (DPPH)) as described in the following sections:

Fucose Determination
Fucoidan contents were estimated performing fucose measurements as described by Dische and Shettles [31] with slight modifications. Briefly, 1 mL of fucose standards (ranging from 0.005 to 0.1 mg/mL) and macroalgal extracts at appropriate dilutions were added to 4.5 mL of a mixture 1:6 of water:sulfuric acid. The mixtures were warmed for 10 min at 22 • C followed by placing the samples 10 min at 100 • C in a water bath. The samples and standards were cooled at room temperature, 0.1 mL of 3% cysteine hydrochloride were added and stored for 60 min at room temperature. The absorbance of the standards and extracts were read at 396 (A 396 ) and 430 nm (A 430 ) in a microplate reader (Epoch, BioTek, Winooski, VT, USA). The fucose content of the samples was determined against the fucose standard at effective absorbance of A 396 -A 430 . The fucose values were expressed as mg fucose per 100 g dried seaweed (ds).

Total Glucan Determination
The total glucan contents of the macroalgal extracts were determined enzymatically using the enzymatic kit K-YBGL (Megaenzyme International Ltd., Bray, Ireland) according to the manufacturer's instructions. Briefly, 100 mg of dried and milled samples and positive control (yeast β-glucan) were weighed and mixed with 1.5 mL of concentrated HCl (37% w/v). The samples were mixed and warmed at 30 • C for 45 min followed by the addition of 10 mL of distilled water and incubation in a shaking water bath (100 • C, 100 rpm and 2 h). Samples were cooled at room temperature, neutralized with 2 M KOH and adjusted to 100 mL with sodium acetate buffer (pH 5.0). Samples were centrifuged at 1500 g during 10 min and the supernatants collected. Duplicate subsamples of each supernatant (0.1 mL) were mixed thoroughly with 0.1 mL of a solution containing exo-1,3-β-glucanase (20 units (U)/mL) and β-glucosidase (4 U/mL) and incubated in a water bath at 40 • C for 60 min. Each subsample, together with blanks (0.2 mL sodium acetate buffer pH 5.0) and glucose standards (0.1 mL of glucose standard (1 mg/mL) and 0.1 mL of acetate buffer pH 5.0), were incubated with 3 mL of glucose-oxidase-peroxidase-reagent (GOPOD) at 40 • C for 20 min. The absorbance of glucose standards and subsamples were read at 510 nm against reagent blank (UVmini-1240, Shimadzu, Kyoto, Japan). Total glucans were calculated using Mega-Calc™ provided by Megazyme (Megaenzyme International Ltd., Bray, Ireland). The total glucan values were expressed as mg total glucans per 100 g ds.

Antioxidant Activity Ferric Reducing Antioxidant Power (FRAP)
The ferric reducing ability related antioxidant potential of the extracts was studied using the FRAP method described by Benzie and Strain [32] modified by Bolanos de la Torre, et al. [33]. Solutions containing the extracts (1 mg/mL) were prepared in Milli Q water. Trolox at concentrations ranging from 15-420 µM were used as standard. The FRAP working solution was freshly prepared by mixing 10:1:1:1.4 of acetate buffer (300 mM, pH 3.6), ferric chloride (20 mM in Milli Q water), 2,4,6-Tripyridyl-s-Triazine (TPTZ) (10 mM in 40 mM HCl) and Milli Q water, respectively. The reaction was initiated in a Greiner CELLSTAR ® 96 flat bottom microplate by adding 280 µL of FRAP working solution to 20 µL of the test compound (extracts at 1 mg/mL) or standard. The samples were incubated at 37 • C in the dark for 30 min and the absorbances were read at 593 nm. The FRAP values were expressed as µM trolox equivalents per mg (freeze-dried seaweed extract) fde.

1,1-Diphenyl-2-Picryl-Hydrazyl (DPPH) Radical Scavenging Activity
The DPPH inhibition assay was performed according to the method described by Nicklisch and Waite [34] with slight modifications. Briefly, freeze-dried macroalgal extracts and positive control (ascorbic acid) were dissolved at a concentration of 1 mg/mL in 0.1 M citrate phosphate buffer pH = 5 with 0.3% (v/v) Triton X-100. The initial absorbance values of the tested samples, positive control and blank solutions (190 µL) were read in a Greiner CELLSTAR ® 96 flat bottom in a microplate reader (Epoch, BioTek, Winooski, VT, USA). The reaction was started by adding to each well 10 µL of a 2 mM solution of DPPH in methanol to give a final DPPH concentration of 100 µM in each well. The plates were incubated in the dark at room temperature for 30 min and the final absorbance of the reaction was read at 515 nm. The initial absorbance readings were subtracted from the final readings and the% radical scavenging activities were calculated using the following equation: % DPPH inhibition = ((Abs Blank − Abs Inhibitor)/Abs Blank) × 100 (5) where Abs Blank is the absorbance of the DPPH solution without any test compounds and the Abs Inhibitor is the absorbance of the tested samples or positive control after the reaction takes place.

Experimental Design and Statistical Analysis
The optimization of the extraction of bioactive compounds from L. digitata was performed using RSM. A Box-Behnken Design with 3 independent variables, each at 4 levels, was employed in this study, requiring a total of 17 experiments for the optimization of the UAE variables. The experimental order was randomized and the levels of the independent variables temperature (40-80 • C), time (10-30 min) and ultrasonic amplitude (40-100%) were coded and listed with the original values in Table 6. The experimental design matrix and the extraction yields of fucose (mg/100 g ds), total glucans (mg/100 g ds), FRAP (µM trolox/mg fde) and DPPH (%) are compiled in Table 1. The results were analysed using response surface regression (RSREG) (SAS version 9.2) fitted to the following second-order polynomial model: where, Y is the predicted response (fucose, total glucans, FRAP and DPPH); β0 is the constant coefficient; β i is the linear coefficient; β ii is the quadratic coefficient; β ij is the cross product coefficients; X i and X j are independent variables. Plots combining contour (2D) and response surface (3D) were generated using Design Expert (v.11) software. The plots show the variation in the responses obtained from multiple combinations of 2 independent variables while holding one of the components constant in the second-order polynomial model. The validity of the model was determined by comparing the experimental and predicted values.

Conclusions
Ultrasound-assisted extraction (UAE) was studied for the extraction of polysaccharides (fucose and glucans) and antioxidant activities (FRAP and DPPH) from L. digitata. Response surface methodology was employed to investigate the effect of the UAE variables (temperature, time and ultrasonic amplitude) on the macroalgal extracts to enhance the yields of polysaccharides and its antioxidant activities. The UAE parameters studied showed significant influence on the levels of fucose, FRAP and DPPH. Levels of 1060.75 mg/100 g ds, 968.57 mg/100 g ds, 8.70 µM trolox/mg fde and 11.02% were obtained for fucose, total glucans, FRAP and DPPH respectively at optimized conditions of temperature (76 • C), time (10 min) and ultrasonic amplitude (100%) using 0.1 M HCl as solvent. The UAE conditions described were then applied successfully to other economically relevant brown macroalgae (L. hyperborea and A. nodosum) to obtain polysaccharide rich extracts. This study demonstrates the applicability of UAE to enhance the extraction of bioactive polysaccharides from various macroalgal species.