Pressurized Liquid Extraction for the Recovery of Carotenoids and Functional Compounds from Green and Orange Dunaliella salina Biomasses

In recent years, intensive research has been conducted on natural carotenoids extraction using several processes. Conventional extraction methods require high amounts of solvents and a long extraction time. However, pressurized liquid extraction demonstrated to be an interesting method. The extraction efficiencies of pressurized liquid for the recovery of carotenoids, from the green and the orange biomasses of the microalga Dunaliella salina DunaDZ1, are described. Organic solvents were tested including ethanol, n -hexane, ethyl acetate and a mixture of n -hexane:ethanol (3:4). Moreover, three extraction temperatures were used (90, 120 and 150 °C) at constant pressure. Extraction efficiency and extracts characterization were conducted. Results have shown that temperature has a positive effect on extraction yield. HPLC characterization showed that β -carotene is the main carotenoid in the orange biomass, and lutein in the green biomass, with the presence of other minor carotenoids in both biomasses. The highest carotenoid amounts were found in the n -hexane orange biomass extract, with β -carotene isomers as the main carotenoid (138.54 and 357.10 mg/g of dry extract, for cis and trans isomers, respectively). Otherwise, extracts obtained at the lowest tested temperature provided the best carotenoid yields. The best results for the antioxidant activity were obtained at 120 °C for orange biomass ethyl acetate extract.


Introduction
Carotenoids are a group of pigments that are gaining more attention due to their potential health benefits and their essential role in preventing human diseases (cardiovascular diseases, macular degeneration and cancers). They are considered as natural antioxidants, which can help maintain the stability and coloring of foods [1][2][3][4]. One of the major interesting carotenoid is β-carotene, which can be found in several natural sources (e.g. carrots, sweet potatoes, apricots, etc.), with the microalgae Dunaliella salina as the main source of this compound (up to 16% of dry matter) [5][6][7].
Some microalgae species are cultivated on a largescale for their ability to produce bioactive compounds with multiple industrial, biotechnological and biorefinery applications [8][9][10]. Numerous species of microalgae are reported to produce different types of carotenoids, e.g. Dunaliella salina, Porphyridium cruentum, Isochrysis galbana, Haematococcus pluvialis, etc. Astaxanthin, β-carotene, lutein, lycopene and canthaxanthin are the main carotenoids found in microalgae, which are receiving increased attention thanks to their biological activity and their potential health benefits [1,[11][12][13].
The green biflagellate microalgae D. salina has pearshape cells and thin elastic plasma membranes. The main characteristic of this species is the lack of a rigid cell wall.
It is considered as good resource for commercial production of β-carotene. D. salina carotenoids are synthesized in response to stress factors such as salinity, light and temperature [14,15]. In the cells, β-carotene is usually present with other carotenoids such as lutein, neoxanthin, zeaxanthin, α-carotene and violaxanthin [5,16].
The choice of extraction method for obtaining carotenoids from microalgae is crucial and should take into consideration several factors, such as the oxidative characteristics of carotenoids which limit the exposure to excess heat and long extraction times [17]. Nowadays, several carotenoids extraction methods have been investigated, which can be classified as conventional: Soxhlet, maceration, etc., and non-conventional methods: pressurized liquid extraction (PLE), supercritical fluid extraction (SFE), pulsed electric field (PEF), etc. [11,[18][19][20]. Most of these methods were tested by several authors on different strains of D. salina, but were only evaluated for their application to the orange biomass [21][22][23][24].
PLE technique uses solvents at high temperatures and pressures to extract various analytes mainly from solid samples. These conditions cause changes in solvent proprieties, enhancing mass transfer rates and decrease solvent surface tension and viscosity and hence the solubility of analytes increases. This technique has the advantage on being fast and requires smaller solvent volumes compared to conventional extraction techniques [25,26]. Therefore, PLE has been used to extract several compounds from different microalgae species, for example, pigments (mainly carotenoids), omega-3 acylglycerols and polyphenols [27][28][29][30].
The aim of this work was to study carotenoids composition and antioxidant activity of different pressurized liquid extracts of green and orange biomass from D. salina strain DunaDZ1 isolated from Zahrez Chergui salt lake in Algeria. PLE carotenoids extraction from D. salina green biomass is, to the best of our knowledge, investigated for the first time.

Microalgae biomass
Dunaliella salina strain DunaDZ1 was isolated from Zahrez Chergui salt lake in Djelfa province, Algeria. It was identified using molecular tools in a previous study [31]. D. salina was cultivated in f/2 medium using one-step method for the green biomass [32]. Cultures were incubated at 22 ± 2 °C with continuous illumination provided by white lamps (120 μmol photon m −2 s −1 ). However, for the orange biomass (carotenogenesis) a two-step method was used.
The first step (growth phase) consisted of the accumulation of a high biomass level by culturing D. salina under optimal conditions. When the culture had produced a high level of biomass, it was used for the second step (stress phase for carotenoids production). Cultures at the end of the exponential phase were transferred into another medium to induce cells for carotenogenesis (free nitrate medium, at 2 M NaCl). Biomasses were harvested by centrifugation and dried in a vacuum freeze dryer (Chaist). These were then stored in darkness at -20 °C until extraction.

Pressurized liquid extraction of carotenoids
Extractions were carried out with an accelerated solvent extractor Dunix ASE 350 (Sunnyvale, California). Green or orange freeze-dried biomass was mixed in a 1:9 ratio with sea sand (300 to 350 μm of diameter, from Scharlau, Spain). This mixture was added to 10 mL cells. A filter paper was placed at the bottom of the extraction cell and then was automatically filled with solvents. The solvents used were ethanol, n-hexane, ethyl acetate and a mixture of n-hexane:ethanol (3:4). Extractions were performed at three different temperatures: 90, 120 and 150 °C, with a static extraction time of 15 min and pressure of 110 bar. These conditions were selected based on published [11] and unpublished data that aimed to extract carotenoids from green microalgae using PLE. Solvents were removed using HeidolphHei-Vap HB/G3 evaporator at 35 °C. The samples were saturated with N 2 and were stored in darkness at 4 °C until their analysis. Extractions were carried out in duplicate.

Thin Layer Chromatography (TLC) analysis
Silica gel 60 plate (Macherey-Nagel. Germany) was used to perform TLC. 10 μL of each extract at concentration of 20 mg/mL were spotted on the Silica plate. A mixture of petroleum ether and acetone 3:1 (v/v) was employed as the mobile phase. The standard used was β-carotene at a concentration of 5 mg/mL.

HPLC analysis
Analyses of the extracts were performed with an HPLC (Varian ProStar 218) equipped with a diode array detector. Twenty microliters of each extract (2 mg/mL) were injected into the HPLC. This HPLC is equipped with an Eclipse XDB-C18 column (5 μm, 150 × 4.6 mm). The mobile phase used was 100% methanol. The flow rate was kept at 1.5 mL/min for 50 min. Quantification of carotenoids was done by referring to a standard curve, using carotenoid standards (β-carotene, violaxanthin, lutein, α-carotene, zeaxanthin, by Sigma-Aldrich). The detection wavelength was set at 450 nm.

DPPH radical-scavenging activity
DPPH radical-scavenging activity of PLE extracts was measured according to a previously described method of Tepe et al. [33]. A sample of 1 mL (extracts at different concentrations: 300 to 1000 μg/mL) was added to 1 mL 0.004% DPPH solution in methanol. Since our PLE extracts from green and orange biomass absorb at 517 nm, it was necessary to prepare a control. This control contains 1 mL of each sample and 1 mL of methanol. Whereas the blank contains only methanol. The mixture was shaken then incubated in darkness for 30 min. Absorbance was read at 517 nm using a spectrophotometer (Jenway 6705). Analysis was carried out in triplicate.
The DPPH radical activity was calculated using the Eq. (1): where A 0 represents absorbance of DPPH solution (without extract) (DPPH solution:methanol) (v/v), A s represents absorbance of DPPH solution mixed with extracts dilutions.
The results were expressed as IC 50 (concentration providing 50% inhibition). IC 50 values were calculated from the plotted graph of scavenging activity against the concentrations of the samples.

Statistical analysis
Statistical analyses were performed using analysis of variance (ANOVA) and Tukey's test. The software used was SPSS 21.0 (SPSS Inc., Chicago, USA). Pearson's correlation coefficient correlation was calculated to determine the relationship between parameters.
3 Results and discussion 3.1 PLE yields PLE extraction yields from green and orange D. salina biomass are presented in Fig. 1. Extraction yields values are closely related to temperature. Fig. 1 demonstrates that temperature enhances the extraction yields for both biomasses. The highest yield was obtained for ethanol extracts at 150 °C and this is for both D. salina biomasses, green (17.83%) and orange (8.68%). It is clear that the temperature increased the solubility of solutes in the solvent. The extraction yield of D. salina biomass also improved in the presence of polar solvent (ethanol).
Previous works have shown the effect of temperature to enhance yields, it improves mass transfer rates, promoting solutes extraction from biomass to the solvent. This is explained by the increase of the solute solubility and diffusion coefficients; it also reduces solvent surface tension and viscosity [11,17,34].

PLE extracts analysis 3.2.1 TLC analysis
Thin layer chromatograms are shown in Fig. 2. The retention factor (Rf ) value of each spot was calculated and then compared with the scientific literature. The TLC analysis of green biomass revealed several green spots, which are characteristic of chlorophylls. Grey spots are also detected at different Rf, with a high intensity for n-hexane and ethyl acetate extracts. Spots intensity is higher with the highest temperature tested 150 °C. These spots are pheophytins, which are chlorophylls degradation products. Pheophytin is a chlorophyll molecule lacking a central Mg 2+ ion after heating or at low pH [35]. The same remarks have been announced by Jaime et al. [11], for n-hexane PLE extracts from the microalgae Haematococcus pluvialis. Carotenoids are also revealed in D. salina green biomass at different Rf, with lutein (Rf ~ 0.21) as the dominant carotenoids.
However, for D. salina orange biomass, yellow to orange spots are revealed, with β-carotene as the main pigment (Rf ~ 0.9). Furthermore, spots intensity depends on the solvent used. As shown in Fig. 2, almost all chlorophyll spots have disappeared. Pheophytins are also present but with less intensity compared with the green biomass. It is well known that D. salina cells turn to orange aplanospores when subjected to stress.

Analysis and quantification of carotenoids
Carotenoid identification was determined by the retention time and the UV-visible spectral produced, and they were compared with scientific literature. In this study only carotenoids are quantified.
Carotenoid concentration and temperature effects are shown in Table 1, lutein is the main carotenoid in the green biomass, with the higher amount (279.7 mg/g dry extract) was obtained for the n-hexane extract at 90 °C. However, to our knowledge, there are no studies on carotenoids from the green biomass of D. salina, which is characterized by the presence of a large amount of lutein. In addition to lutein, other xanthophylls (violaxanthin and zeaxanthin) were found in D. salina green extracts, with an amount between 36 and 91.7 mg/g dry extract. Furthermore, α-carotene and β-carotene were also detected and presented amounts ranging from 37.1 to 62.6 mg/g dry extract. As shown, and among all solvents tested, temperature has a positive effect on carotenes extraction (α and β), except n-hexane,  in which the highest amount is obtained at the lowest temperature tested (90 °C). Otherwise, temperature has a negative effect on the other carotenoids (xanthophylls), and the best temperature is the lowest tested (90 °C). Carotenoids concentration order in the green biomass is as follows: lutein > violaxanthin > α-carotene > β-carotene. Nonetheless, HPLC results of D. salina orange biomass show trans β-carotene as the main carotenoid. The highest amount of this pigment (357.1 mg/g dry extract) is revealed in n-hexane extract at 90 °C. Cis β-carotene is also present in the orange biomass at lower levels than trans β-carotene, except for ethanolic extracts, where the amount of cis β-carotene exceeds trans β-carotene when extraction was done at 120 and 150 °C.
Carotenoids profiles and concentration from the D. salina orange biomass are summarized in Table 2. In general, the highest carotenoid amounts are obtained at the lowest temperature tested (90 °C). The α-carotene is also detected in D. salina orange biomass, with the higher level (57 mg/g dry extract) in the n-hexane extract at 90 °C. Lutein is also present in the orange biomass with the higher level (102.3 mg/g dry extract) for n-hexane at 120 °C, this level is lower than the obtained for the green biomass. Violaxanthin is also identified in the orange biomass with the higher level (48 mg/g dry extract) in hexanic extract at 90 °C. Moreover, two other carotenoids are detected in the orange biomass but not identified. Quantities of these pigments vary from 36.5 to 51.2 mg/g dry extract.
The microalga D. salina DunaDZ1 produces not only β-carotene, but is considered also as a good source of lutein which can help to prevent degenerative diseases as for example age-related macular degeneration [36].
It is important to note that, amounts of carotenoids found depend greatly on not only the microalgae species or strain studied but also on the culture conditions (whether the one or two step approach were used, the stress factor used, etc.). Salinity, light intensity, nitrate concentration, photoperiod are the main factors determining biomass composition [40,41].

Antioxidant activity
The antioxidant potential of the PLE extracts was estimated and results are summarized in Table 3. PLE extracts obtained from orange and green D. salina biomass were analyzed to evaluate antioxidant activity and the results are expressed as IC 50 values (μg of extract per mL). Results showed that the orange biomass is the most active. Temperature affects the IC 50 values, the best activity was found at 120 °C.
The lowest IC 50 (important antioxidant activity) is obtained in the ethyl acetate extract, with 194 ± 0.22 and 336.2 ± 0.59 μg/mL, for orange and green biomasses, respectively.
The highest IC 50 was obtained at 90 °C for the orange biomass and at 150 °C for the green one. The difference  [42]. Otherwise, the IC 50 of D. salina extracts were highest then IC 50 of other reference compounds with 6.3 µg/mL and 2.5 µg/mL, vitamin C and butylated hydroxytoluene (BHT), respectively [43].
Statistical analysis showed non-significant positive correlation between IC 50 and total carotenoids from the green biomass (r = 0.107; p = 0.741; p > 0.05). However, for the orange biomass a non-significant negative correlation was revealed (r = -0.236; p = 0.461; p > 0.05). Carotenoids are known as an antioxidant agent. Protecting cells and tissues against damaging reactive oxygen species (ROS). They have a high free-radical scavenging proprieties thanks to the elongation in the conjugated double bond system. Carotenoid pigments are responsible for a major part of antioxidant activity in D. salina. They have a high free-radical scavenging proprieties thanks to the elongation in the conjugated double bond system. But other molecules such as phenolic compounds, polyunsaturated fatty acids and polysaccharides may also be present and consequently contribute to the increase of antioxidant activity [44,45].
Otherwise, the non-significant correlation between total carotenoids and DPPH methods could be explained by the efficiency weakness of the method used to estimate the antioxidant activity (DPPH), as it is reported by several authors [46,47].
In another study conducted on the antioxidant activity of Moroccan microalgae, the best result (IC 50 = 283 μg/mL) was found in a Dunaliella sp. strain [43]. However, Cakmak et al. [48] reported IC 50 values ranging from 450 to 3460 μg/mL for extracts obtained by different solvents from a strain of D. salina.
Moreover, Hu et al. [49] have demonstrated that the antioxidant activity of carotenoids extracted from D. salina is higher than the antioxidant activity of synthetic carotenoids, which opens numerous ways to utilize and commercialize these compounds.

Conclusions
PLE extracts of D. salina DunaDZ1 were found to contain major carotenoids, trans β-carotene for the orange and lutein for the green biomass, as well as other minor carotenoids. PLE hexanic extracts has shown the highest trans β-carotene yield (357.10 mg/g) in the orange biomass, and the highest lutein yield (279.67 mg/g) in the green biomass. The best temperature for PLE extraction from both biomasses is the lowest tested (90 °C). However, better antioxidant activity was obtained in the orange biomass extracted with ethyl acetate in comparison with the green one. The non-significant correlation between carotenoid amount and antioxidant activity could be explained by whether the weak efficiency of the method used and/ or by the presence of other compounds in the biomass which contribute to improve this activity. In this study we have shown that both biomasses, green and orange, from D. salina DunaDZ1 are of great interest considering the industrial and commercial applications of β-carotene and lutein. Furthermore, PLE has advantages compared to traditional methods for being shorter in time and using small quantities of solvents.