Introduction

Microalgae are an alternative source of a wide variety of high added value compounds with potential applications in food, nutraceuticals, and food supplements. These microorganisms have the ability to accumulate different bioactive compounds including polar lipids, omega-3 fatty acids or carotenoids, depending on the species (Chew et al. 2017; Jacob-Lopes et al. 2019; Katiyar and Arora 2020).

There is extensive detailed literature about the influence on health of compounds such as omega-3 polyunsaturated fatty acids (PUFA) or carotenoids. Among carotenoids, xanthophylls are a class of oxygen-containing carotenoid pigments that outstands due to their biological activities, including mainly antioxidant or anti-inflammatory activities. Fucoxanthin (Fx) is the main carotenoid, particularly a xanthophyll, present in all brown macroalgae. Moreover, it is produced by some microalgae but in high concentrations. Fucoxanthin has also been related in various studies to antidiabetic, antitumor, anti-cancers and hepatoprotective biological properties (Gong and Bassi 2016; Foo et al. 2017; Méresse et al. 2020; Bigagli et al. 2021; Pereira et al. 2021; Ahmed et al. 2022; Pajot et al. 2022; Demets et al. 2023). On the other hand, omega-3 PUFAs, mainly eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), trigger anti-inflammatory cascades in the body and there is scientific evidence linking them with the reduction of risk of developing cardiovascular diseases and neuroinflammatory and neurodegenerative diseases (Van Lo et al., 2016; Kerdiles et al. 2017; Hachem et al. 2020; Balakrishnan et al. 2021; Scheinman et al. 2021; Jayatunga et al. 2022). Moreover, the bioaccessibility of these bioactive compounds depends on the chemical structure in which they are present. Thus, EPA and DHA bound to polar lipids such as glycolipids (GL) or phospholipids (PL) are more bioaccessible than those bound to triglycerides (Castejón and Señoráns 2019; Zhang et al. 2019; Farooq et al. 2020; Fil et al. 2021; Melo et al. 2021).

The traditional sources of omega-3 PUFA are marine fish and krill oils; however, this system encourages marine overexploitation and overfishing. Therefore, new alternative and sustainable sources of bioactive omega-3 fatty acids must be implemented. Microalgae are the only non-animal source that produce EPA and DHA, in addition, these microorganisms normally present these fatty acids bound to polar lipids such as phosphatidylcholine (PC) and lysophosphatidylcholine (LPC), increasing their bioactivity value due to the increase of the bioavailability. This is related to the nature of the specific receptor called Major facilitator superfamily domain-containing protein 2a (Mfsd2a) in which only the PL-DHA (also known as LysoPL) form can cross the blood-brain barrier (Sugasini et al., 2017, 2019; Wang et al. 2018; Sugasini et al. 2019; Yalagala et al. 2019; Zhang et al. 2019; Scheinman et al. 2021). Additionally, studies analyzing Mfsd2a-knockout mice demonstrated the importance of this transporter in the assimilation of DHA at the brain level. Its absence translated into problems related to deficits in learning processes, memory, development of severe microcephaly and an increase in postnatal mortality due to omega-3 PUFA deficiency (Nguyen et al. 2014; Lagarde et al. 2015, 2016; Hachem et al. 2020).

Furhermore, microalgae do not require fertile land for their cultivation with a growth rate between 2-3 times higher than terrestrial plants. This makes them a sustainable and environmentally friendly alternative source of bioactive compounds of high interest (Chew et al. 2017; Barkia et al. 2019; Katiyar and Arora 2020; Tan et al. 2020; Tang et al. 2020; Abu-Ghosh et al. 2021).

The microalga Tisochrysis lutea El M.Bendif & I.Probert is a relatively new species, corresponding to a strain isolated from a Tahitian culture designated as Isochrysis aff. galbana (called "T-Iso") and named Tisochrysis lutea (Alonso et al. 1992a, 1992b; Bendif et al. 2013; Custódio et al. 2014). Tisochrysis lutea has been widely used in aquaculture to feed mollusks, crustaceans and fish in early stages, due to high growth rate and good adaptation to change, and particularly for its pigment and lipid content as it produces a high amount of omega-3 PUFA and fucoxanthin (Ippoliti et al. 2016; Bigagli et al. 2018; 2021; Gallego et al. 2020; Mohamadnia et al. 2022; Pajot et al. 2022).

Bioactive compounds are traditionally extracted using procedures which use large amounts of organic and harmful solvents with relatively low lipid recoveries (e.g., Folch et al. 1957). Thus, new extraction techniques with alternative and sustainable solvents are necessary to develop novel processes within the framework of Green Chemistry (Zhou et al. 2022). In this way, biodegradable biobased solvents produced from biomass of agro-industrial origin stand out towards traditional solvents. 2-methyl-tetrahydrofuran or 2-methyloxolane (2-me-THF, CAS Number 96-47-9) derived from sugar cane biomass or corn cobs waste is a substitute for solvents such as hexane, accepted for food applications, which may be suitable for extracting lipid compounds such as carotenoids and PUFAs. 2-me-THF is considered a green solvent because it can be produced using renewable resources and has a smaller life cycle impact. Moreover, the relatively high boiling point (80˚C) and low melting point (-137˚C) provide a broad temperature range that other organic solvents forbid, as well as to possess low genotoxicity, being a greener alternative to other hazardous solvents (Slater et al. 2016; de Jesus and Filho 2020; Rapinel et al. 2020a, 2020b; Tang et al. 2020; Blanco-Llamero and Señoráns 2021; Lambré et al. 2022).

Ultrasonic-assisted extraction (UAE) is an advanced extraction technique based on the use of acoustic cavitation effect induced by microwaves in which mass transfer is improved. All these phenomena produce an increase in the lipid extraction yield and a reduction in the volume of solvent and time required to carry out the extraction. In addition, since the technique is carried out at moderate temperatures, oxidation and decomposition of organic compounds are avoided (Blanco-Llamero et al. 2021; Carreira-Casais et al. 2021; Kumar et al. 2021; Zheng et al. 2021; Cisneros-Yupanqui et al. 2023; Islam et al. 2023; Menezes Silva et al. 2023; Sabaruddin et al. 2023).

The aim of the present work is to improve the lipid extraction yield as well as the enrichment in bioactive lipidic compounds such as polar lipids and fucoxanthin using T. lutea as a sustainable and rich source. For this reason, advanced techniques such as ultrasound-assisted extraction is proposed to ease the extraction process, using environmentally friendly protocols and solvents that in turn preserve the bioactivity of these compounds. In this way, the hypothesis is to obtain separated enriched extracts depending on the solvent used and the conditions of extraction.

Materials and methods

Materials

Tisochrysis lutea was obtained from Cianoalgae SL (Madrid, Spain). Potassium hydroxide was supplied by Scharlau (Spain). Lutein 5% was supplied by Hunan Nutramax Inc. (China) and fucoxanthin and 2,2-diphenyl-1-picrylhydrazyl (DPPH) were supplied by Sigma-Aldrich (USA). Chloroform, methanol, absolute ethanol, methyl tert-butyl ether (MTBE), isooctane (2,2,4 trimethylpentane), ethyl acetate and 2-propanol were supplied by Macron Fine Chemicals (Poland). 2-methyltetrahydrofuran (2-methyl-THF) was supplied by Pennakem (USA). All solvents used were analytical grade or HPLC grade. The water used was Milli-Q grade from Millipore Sigma (USA).

Methods

Extraction by traditional Folch method

The Folch extraction was performed following the original method (Folch et al. 1957). 1 g of microalgae was mixed with 20 mL of methanol:chloroform (1:2 v:v), vigorously vortexed, and the mixture was centrifuged for 10 min at 1008 RCF. Subsequently, the organic phase was collected and mixed with 3 mL of distilled water with saturated NaCl, vortexed and centrifuged again. The process was repeated two more times, for a total of three washes. Finally, the chloroform phase containing the extracted lipids was dried with a stream of nitrogen until constant weight. The lipid content was determined gravimetrically and the lipid extraction yield was calculated as percentage by weight of dry biomass (Yield (%) = (extract weight / sample weight) * 100). The lipid extracts obtained were stored in the dark at 4°C until analysis. The experiments were performed in all cases at least in duplicate.

Experimental design of ultrasound-assisted extraction

For UAE, the mixture of solvents with different polarities, temperature and time at three levels were taken as factors, for which a Box Behnken design was used using Statgraphics program, in order to minimize time and costs compared to other designs and to optimize the process. Then, different percentages of two green solvents (from 0 to 100% 2-me-THF in ethanol), different extraction times (20, 30 and 40 minutes) and three temperatures (40, 50 and 60 ºC) were evaluated.

The experimental region was delimited by the three factors mentioned above at three levels in a Box Behnken design: the proportion between the solvents 2-me-THF in ethanol at 0% (-1), 50% (0) and 100% (+1), the temperature at 40 ºC (-1), 50 ºC (0) and 60 ºC (+1) and the time at 20 (-1), 30 (0) and 40 min (+1).

Table 1 shows the design used and the conditions of each extraction performed in triplicate as shown in the Statgraphics program.

Table 1 Extraction conditions by UAE using Box Behnken experimental design

One gram of microalgal biomass was weighed in glass vials and 10 mL of the solvent or its mixture were added, shaken and placed floating in the ultrasonic bath with the previously reached temperature and the time of each extraction was graduated. An Elmasonic S 40H ultrasonic bath (Elma, Germany) was used with automatic time and temperature control, a frequency of 37 kHz and 140 W power. Subsequently the sonicated sample was filtered and the liquid phase, now referred to as the extract, was dried with a rotary evaporator and stream of nitrogen and stored at 4 ºC for subsequent analysis. All experiments were developed in triplicate.

Statistical analysis

The analysis of the statistical differences of the data obtained in the lipid extraction yield was carried out through a simple ANOVA, using Statgraphics XVI.II version 16.02.0004.

Determination of total carotenoid content

A spectrophotometric method was used to determine the concentration of total carotenoids (CTC). The extracts obtained during optimization were dissolved in methanol at a concentration of 0.05 mg mL-1. The absorbance of these solutions was recorded at 470 nm. An external lutein standard calibration curve (0.25–2 mg mL-1) was used to calculate the total carotenoid content in the samples (Gallego et al. 2019). Samples were analyzed at least in duplicate. The results are represented as mg of carotenoids per gram of extract.

Analysis by gas chromatography of the fatty acid content of the extracts

The fatty acid content of the extracts was determined in triplicate by derivatization in an alkaline medium and the fatty acid methyl ester (FAME) obtained were analyzed by gas chromatography coupled to a mass spectrometer (GC-MS). To obtain FAME in basic medium, the ISO TC34/SC 5 Standard was followed: 25 mg of oil were weighed and mixed with 200 μL hexane. 50 μL of a KOH solution in 2 N methanol were added to this mixture and vortexed for 1 min (prepared daily). After an additional reaction time of 5 min the reaction was stopped by the addition of 125 mg NaHSO4·H2O. It was then centrifuged for 5 min at 2800 RCF. Finally, 100 μL of supernatant (containing the FAME) was collected and dissolved in 400 μL hexane (final solution) for later analysis.

The fatty acid analysis was performed on an Agilent Technologies ( USA) 5975 MSD Series gas chromatograph-mass spectrometer with an automatic injector and He as carrier gas. An Agilent Technologies HP-88 capillary column was used, measuring 100 m x 0.25 mm x 0.20 μm. The injection temperature was 250 ºC. The oven was kept at 175 ºC for 8 min. The temperature was then raised at a rate of 3 ºC min-1 to 230 ºC, which was maintained for an additional 10 min. The detector temperature was 230 ºC. The amount of sample injected was 1 μL with a 1:20 split. The mass spectrometer used an ionization potential of 70 eV and an atomic mass range of 30 to 400 μ (atomic mass units).

The fatty acids (FA) in the sample were identified using the NIST Mass-Spectral Library 2.0, expressing the amounts as percentages of the total FA content. The results obtained were compared with updated bibliographic references.

HPLC-DAD analysis of the fucoxanthin content of the extracts

Quantification of fucoxanthin present in the extracts of T. lutea was carried out with an HPLC-DAD Varian ProStar (USA) with UV-Vis diode detector (DAD) at wavelengths of 440 nm and 660 nm. The HPLC system consists of a binary pump and 20 µL manual injector. An Eclipse XDB C18 reverse phase column (150 mm x 4.6 mm and 5 µm) from Supelco (USA) was used. The mobile phase was a mixture between solvent A (water:methanol, 1:5) and solvent B (methanol:isopropanol, 3:1). The mobile phase eluted at a flow rate of 1 mL min-1 with an 8.5 min gradient: at initial time 50% A and 50% B, from 4 min to 8 min 30% A, 70% B and at 8.5 min 50% A and 50% B. For the construction of the calibration curve, analytical grade fucoxanthin was used, which was dissolved in methanol in the range of 0.07-0.6 mg mL-1 and each concentration was analyzed by HPLC (retention time (RT) 2.9 min).

HPLC-ELSD analysis of the polar lipid content of the extracts

HPLC coupled to evaporative light scattering detector (ELSD) was used to determine components with no UV absorption such as lipids with high sensitivity. HPLC-ELSD analysis were performed using an Agilent 1260 Infinity Autoinjector Chromatograph (G1329B) with a quaternary pump (G1311B/C), equipped with an Agilent 385 ELSD (. The chromatographic separation was carried out with a silica normal phase column (250 mm x 4.6 mm i.d., 5 μm) maintained at 30 ºC using a ternary gradient as follows: 0–2 min, 99.5% A and 0.5% B; at t = 6.5 min, 70% A and 30% B; at t = 11 min, 63% A, 27% B and 10% C; at t = 18 min, 99.5% A and 0.5% B; and at t = 20 min, 99.5% A and 0.5% B. Eluent A consisted of 2,2,4-trimethylpentane, eluent B consisted of MTBE, and eluent C consisted of 2-propanol. The flow was kept constant at 1.8 mL min-1 and at a maximum pressure of 600 bar. The optimal signal and resolution of the ELSD detector were achieved with the following conditions: evaporator and nebulizer temperature of 30 ºC, and evaporator gas (N2) at 1.6 standard liter per minute (SLM). In this way, the nebulizer transforms the target components into a fine spray using nitrogen as a carrier gas and the dispersed light is detected. To identify different lipids represented in the sample, standards were injected individually and compared with standards already analyzed in previous studies. In all cases, the measurements were made in duplicate.

Determination of the antioxidant activity of the extracts

The 2,2-diphenyl-1-picrylhydrazyl (DPPH) method was used to determine the antioxidant activity of the different extracts obtained via radical scavenging activity measurement. This method relies on the ability of the antioxidant to stabilize the free radicals that are released by the compound DPPH.

To determine the antioxidant activity, DPPH assays were performed by spectrophotometric assay according to Xia et al., (2013) with some modifications. In this case, 1 mL of the different dilutions of the extracts were added to 1 mL of a 0.004% prepared DPPH solution in methanol (A1). Additionally, a blank was also measured with 1 mL of the extract solution and 1 mL of methanol (A2), and a control that was 1 mL of methanol and 1 mL of the DPPH solution (A0). The samples were left to stand in the dark for 30 min and their absorbance at 517 nm was measured. For determination of DPPH inhibition, the following formula was used: Radical scavenging (%) = [(1 − (A1A2))/A0] ·100. Then, IC50 was calculated as the amount in which 50 % of inhibition of DPPH radicals was achieved. Each measurement was made in duplicate.

Results

Extraction by traditional method and initial characterization of T. lutea

The lyophilized microalgae were used to extract the bioactive compounds using the traditional Folch method. Then, the analysis of fatty acids and carotenoids content, the bioactive compounds of interest in this work, was developed using GC-MS and HPLC-DAD.

The lipid extraction yield was 16.24 ± 0.35 %, with a CTC of 11.41 ± 0.077 mg g-1, of which 2% is that belonging to Fx, with a concentration of 0.23 mg g-1.

As observed in Table 2, several fatty acids were identified using the GC-MS spectral library, being myristic acid, stearidonic acid and DHA the most abundant, which was consistent with recently published studies (Maglie et al. 2021).

Table 2 Fatty acid composition by GC-MS

Experimental design for the optimization of the extraction process of bioactive compounds from T. lutea

The Box Behnken design was applied to study the influence of different parameters (temperature, extraction time and composition of the solvent mixture) involved in the lipid extraction yield of T. lutea using UAE.

The lipid extraction yield was considered the variable of response. Table 3 shows the conditions and lipid extraction yield obtained for the 15 extraction conditions.

Table 3 Experimental design including the different factors and the values of the response variables (lipid extraction yield) studied in the optimization of UAE

As observed in Table 3, lipid extraction yields were higher when the percentage of ethanol was 100% (0% 2-me-THF), followed by the mixtures of ethanol and 50% 2-me-THF, obtaining the worst results when only 2-me-THF was used as extraction solvent, as would be expected based on preliminary results (data not shown).

To determine the influence exerted by the extraction factors (solvent or D, temperature or T and time or Ti) and its three levels in the performance variable, the data were adjusted to a quadratic model whose formula (Eq. 1) allowed determining the regression coefficients, their interactions and double interactions through the p-value. This statistical model adjusted the experimental data of the performance percentage with R2: 98.82 and error of 1.01, which were reliable results to validate the quadratic model used, determine the influence of each extraction factor on performance with the Pareto diagram and the effect of the levels of each factor with the response surfaces.

$${\mathrm{\% Yield}= - 3.50875 + 0.0762917}^{*} {\mathrm{D }+ 0.410917}^{*} {\mathrm{T }+ 0.542875}^{*}{\mathrm{Ti }- 0.00134417}^{*}{{\text{D}}2 - 0.00122}^{*}{{\text{D}}}^{*}{\mathrm{T }- 0.000615}^{*}{{\text{D}}}^{*}{{\text{Ti}}+0.000420833}^{*}{{\text{T}}2 -0.01085}^{*}{{\text{T}}}^{*}{\mathrm{Ti }- 0.000154167}^{*}{{\text{Ti}}}^{2}$$
(1)

Through the Pareto diagram (Fig. 1) it was observed that the solvent factor (A) and its double interaction in the model (AA), exerted a significant influence on the percentage of yield in the extraction by UAE. Its blue color indicates that it was a main effect of decrease in % yield, while temperature and time did not have a significant effect. However, the gray color in the temperature bar indicates an effect of increasing the lipid extraction yield, not significant.

Fig. 1
figure 1

Pareto chart reflecting the standardized effect of draw factors for UAE on draw performance

In the response surfaces (Fig. 2) the central color scale indicates an increase or decrease in performance (y-axis); from blue to red increases performance. It was observed that the increase in the 2-me-THF solvent decreases the % yield regardless of the extraction time. In relation to temperature, the increase in its level at the shortest extraction time (20 min), and the lowest percentage of solvent, both produced the highest yield.

Fig. 2
figure 2

Response surface for lipid extraction yield as a function of extraction factors obtained by UAE for T. lutea at different times (a: 20 min, b: 30 min, c: 40 min)

According to the quadratic model used, the optimized conditions were designated based on the lipid extraction yields for the three factors analyzed (% of 2-me-THF, temperature and time) with which a maximization of the yield percentage would be achieved, within the studied experimental region. These values ​were 1.92 % 2-me-THF in ethanol, at an extraction temperature of 59.5 °C for 20 min. Additionally, the statistical program used predicted a lower performance limit of 17.38 % and an upper limit of 23.24 % with 95 % confidence.

To corroborate this prediction, the UAE extraction was carried out experimentally, according to the indicated conditions and the lipid extraction yield obtained for the optimal extract was 17.83 ± 1.12 %. The said value was within the limits predicted by the quadratic model used. In addition, the results were compared with those obtained with the traditional Folch method, which had a yield of 16.24 ± 0.35 %.

Chemical characterization of the UAE extracts of T. lutea

Based on the results obtained through the response surfaces, some representative extracts from different extraction conditions were selected to analyze their content in Fx, CTC, DHA and polar lipids. Additionally, the extraction of bioactive compounds from T. lutea microalgae was carried out under two alternative conditions to experimental design, in order to compare the extracts at 50 ºC, keeping the extraction time fixed (30 minutes) and the percentage of solvent (0 and 100%). For this reason, experiments 16 and 17 were also developed to compare the data under the same conditions. Data are shown in Table 4. The results were compared with the traditional Folch extraction and with the optimized extraction (1.92 % 2-me-THF in ethanol, at an extraction temperature of 59.5 °C for 20 min).

Table 4 Chemical characterization of the selected extracts from UAE with experimental design compared to traditional method

In general, the higher lipid extraction yields were obtained using ethanol as solvent (0% 2-me-THF). Nevertheless, the higher amounts of Fx and DHA were obtained in 2-me-THF extracts. Comparing these results to traditional Folch method, although the lipid extraction yield was high, it was not efficient for Fx extraction in T. lutea.

Carotenoids and fucoxanthin content

When analyzing the carotenoid profile of the extracts by means of HPLC-DAD, it was observed that ethanol was not selective in Fx extraction since both chlorophylls and carotenoids were also obtained. In this case, the characteristic spectrum of carotenoids only appears when measuring at 440 nm while chlorophylls present a peak at 440 nm but also at 660 nm, and both appeared when using ethanol as green solvent (spectra not shown) despite the fact ethanol is highly polar. However, extraction of chlorophylls and other non-polar compounds such as PUFA was also previously described using ethanol (Delbrut et al. 2018; Pajot et al. 2023). Furthermore, it was observed that the concentration of Fx increased when the extraction temperature with ethanol also increased, reaching values of 7.2 mg g-1 when the extraction temperature was 60 ºC. The optimal extract, which mainly presents ethanol in the solvent mixture, followed the same dynamics as the rest of the extracts, and Fx concentration was also low, around 3 mg g-1 of extract.

On the contrary, when 2-me-THF was used in the extraction procedure, chlorophylls were not extracted, and Fx and other carotenoids (not identified) were extracted in higher concentrations. In this way, an extract enriched in carotenoids of interest where Fx predominates was achieved by using 2-me-THF as a biobased alternative solvent.

In addition, when extracting with the traditional Folch method, the Fx concentration was the lowest obtained and the extraction process was not selective since chlorophylls were also extracted, emphasizing the importance of advanced extraction techniques that improve not only yields of extraction but also the composition of the extracts obtained.

Since ethanol has been widely used in the literature for the extraction of carotenoids from different microalgae, including Fx from T. lutea (Kim et al. 2012; Menezes Silva et al. 2023), the mixture of ethanol and 2-me-THF for the extraction of bioactive compounds of interest was also studied (Zarrinmehr et al. 2022). In this case, the mixtures of ethanol and 50% 2-me-THF fail to increase the concentration of Fx in the extract, with values similar to those obtained using only ethanol, but minimizing the extracted chlorophylls.

The total carotenoid content did not vary between the extracts either, since the wavelength at which the extracts are analyzed in the spectrophotometer (470 nm) is not selective for carotenoids and also detects chlorophylls. As previously described, the extracts that were extracted with ethanol had many chlorophylls in their composition. On the other hand, those extracted with 2-me-THF had more carotenoids. In all cases, the content varied between 14 and 22 mg g-1, being in this case the highest concentration in the optimized extract, and the lowest in the extract obtained with Folch.

DHA content

DHA content was analyzed by GC-MS. The highest percentage values for DHA (from 15.3 to 17.49 %) over total lipids were obtained with the highest percentages of 2-me-THF, together with Folch, possibly due to the polarity of the solvents used in these extractions. Indeed, 2-me-THF is known for having affinity for non-polar compounds, such as fatty acids in form of triacylglycerols. Moreover, the mixture between chloroform and methanol (2:1) used in Folch enables the extraction of a wide range of lipid classes due to the nature of the solvents, including non-polar substances. This could be explained due to the higher composition in chloroform compared to methanol. In mixtures with 50 % composition of non-polar solvents such as 2-me-THF and 50 % of polar solvents like ethanol, the mixture provoked a “mild” polarity which was not as efficient as that obtained with chloroform and methanol. Therefore, the percentage of DHA was not as high as those obtained in the forementioned conditions (14.95 % compared to 17.49 %). Although there were differences between the extracts, all of them were in values between 12 and 17 %, including the optimal extract and the one obtained with Folch.

Polar lipid composition

The structure in which omega-3 PUFA are found is essential to improve the bioactivity and bioavailability of these bioactive fatty acids. Therefore, the polar lipid profile of the selected extracts was studied by HPLC-ELSD (Table 5). Extracts with different percentage of solvent used (0, 50 and 100% 2-me-THF in EtOH) but same extraction time and temperature (50 ºC and 30 min) are represented as extracts 1, 2 and 3, corresponding to 0%-50 ºC-30min, 50%-50 ºC-30min and 100%-50 ºC-30min, in order to resume the results expressed before only in three important enriched extracts. It is important to know that the hereinafter called extracts 1, 2 and 3 are equivalent to experiments 17, 10 and 16 of Table 4, respectively. Results were compared to optimal extraction and Folch extract.

Table 5 Lipid composition by HPLC-ELSD of the selected extracts compared with the optimal extract and the one obtained by Folch method

Extract 1: 0%-50 ºC-30min; Extract 2: 50%-50 ºC-30min; Extract 3: 100%-50 ºC-30min.

Due to the importance of polar lipids related to several benefits for human health (Zheng et al. 2019; Fil et al. 2021; Melo et al. 2021), focus on phospholipids (PL) and glycolipids (GL) was made.

Although the percentage of DHA present in these three extracts (extracts 1, 2 and 3 equivalent to experiments 17, 10 and 16 of Table 4) practically did not vary as previously expressed, differences were observed in the lipid profile and composition of polar lipids depending on the solvents and the methodology used. For Folch extract, non-polar lipids were mostly present in the form of TAG, and no phospholipids were extracted.

When analyzing the different extracts with the same extraction time and the same temperature, it was observed that the higher the percentage of 2-me-THF used, the higher non-polar compounds such as TAGs were obtained, with results similar to those obtained with the Folch method. In all the three cases studied, regardless of the solvent used, a reduced percentage of PC was obtained that was not achieved using Folch.

Thus, the most interesting extracts based on their polar lipid profile were obtained using high concentrations of ethanol using UAE (0% 2-me-THF in ethanol, equivalent to 100 % ethanol) and optimal. With the optimized conditions it was possible to enrich the extract in phospholipids of great added value, such as PC and even LPC, decreasing the percentage of TAGs present. Thus, GL and PL were extracted with advanced methodologies coupled to the use of highly polar solvents, such as ethanol.

Antioxidant activity of selected extracts with rich composition

The most interesting extracts enriched in bioactive compounds of high value found in this work were studied and compared using the radical scavenging DPPH assay to evaluate their antioxidant activity. Figure 3 shows the IC50 data calculated for each extract. All the values are expressed as mean ± standard deviation (SD). As previously explained these data represent the amount of extract necessary to inhibit oxidation by 50%. Thus, the lower IC50 value represents the extract with higher antioxidant activity.

Fig. 3
figure 3

IC50 values of selected extracts compared to optimal and Folch extract analyzed by DPPH assay. Extract 1: 0%-50 ºC-30min; Extract 2: 50%-50 ºC-30min; Extract 3: 100%-50 ºC-30min. Results expressed as average (mg mL-1) ± standard deviation (SD) of duplicates.

No significant differences were shown between the IC50 value of extracts 2 and 3, and the optimal extract. In this context, these three extracts had a great composition in Fx and other carotenoids, with low quantity of chlorophylls, as well as high content in polar lipids, all of them related to antioxidant activities.

On the other hand, when extracting by UAE with 100% ethanol (0% 2-me-THF), 0.645 mg mL-1 is needed to reach the IC50, results that are consistent with those obtained in the chemical characterization of these extracts as it has a lower amount of Fx. Moreover, Folch extract presented an intermediate IC50 of 0.331 mg mL-1 due also to the composition shown before. In this case, Folch extract rendered the lowest concentration of Fx, but adversely a high concentration of DHA that may also exert antioxidant activity.

Discussion

The results of this work show the potential of the microalga T. lutea to obtain bioactive compounds of interest using sustainable sources and environmentally friendly processes such as advanced extraction techniques and green bioderived solvents instead of traditional methods that use harmful organic solvents.

In this sense, not only an increase of the lipid extraction yield is important to start using alternative and advanced procedures industrially, but also the extract composition and selectivity of the extraction for the target compounds regarding the solvent employed.

Composition of microalgae extracts was successfully studied and compared to select the best procedure to obtain extracts with potential application as nutraceuticals. When analyzing the content of interesting omega-3 PUFA found in T. lutea, results were consistent with those reported in the literature. T. lutea is considered an interesting source for the production of total fatty acids (up to 320 mg g-1dw (dry weight)). Tisochrysis lutea is of interest due to its high composition in ALA (17.8 ± 0.1 mg g-1 dw), SDA (46.2 ± 1.4 mg g-1 dw), EPA (1.8 ± 0.1 mg g-1 dw) and DHA (with contents that vary between 23.0 ± 0.4 mg g-1 dw analyzed by direct transmethylation of the biomass (Delbrut et al. 2018) and 19.89 mg g-1(Gao et al. (2022)). In this way, SDA, DHA, ALA and EPA were obtained in the same proportions as shown herein.

Among omega-3 PUFA, DHA stands out due to its health benefits. EFSA has carried out various scientific evaluations regarding benefits of EPA and DHA consumption, with health claims approved for years. Thus, there is a specific health claim for DHA indicating that it contributes to the maintenance of normal brain function and a beneficial effect for the consumer that is obtained with an ingestion 250 mg of DHA daily. Hence, T. lutea could be used to develop food formulations or nutraceuticals aimed to improve the accumulation of DHA in brain. If so, using advanced environmentally clean techniques together with the use of green solvents it was possible to extract the great amount of the DHA present in the microalgae biomass (around 15% DHA on its composition depending on culture conditions).

Additionally, the polar lipid profile of the extracts obtained is decisive when selecting the extracts for biological activity tests or for their possible use in food ingredients, since it is not only important the enrichment in antioxidant bioactive compounds but also in other kind of bioactive molecules such as polar lipids. Interestingly, co-extraction of Fx and PUFA was also previously reported in literature (Delbrut et al. 2018). Results showed that not only the choice of solvent used is important, but also the use of advanced extraction techniques that improve the composition of the extracts obtained, as is the case with UAE. By using UAE coupled to biodegradable solvents, the extraction of polar lipids such as PL and GL was satisfactorily obtained.

Moreover, extraction of Fx was also improved related to Folch method which uses harmful organic solvents. Under normal culture conditions Gao et al. (2022) reported that the microalga T. lutea can produce up to 20.3 mg g-1 of Fx (dry weight). However, the accumulation of Fx on microalgae biomass will highly depend on culture conditions, mainly quantity and quality of light, nutrient starvation and also storage (Alkhamis and Qin 2015; Rasdi and Qin 2015; Sánchez-Saavedra et al. 2015; Ippoliti et al. 2016; Marchetti et al. 2018; Huang et al. 2019; Gao et al. 2020; Maglie et al. 2021). In this work, a maximum recovery of 12 mg g-1 was obtained using UAE and green solvents, similarly to those obtained in previous reports (Kim et al. 2012; Delbrut et al. 2018). On the contrary, Folch extracts was not able neither for Fx nor for polar lipids extraction.

After optimization of extraction conditions, selected extracts due to composition in high-added value compounds as well as with high lipid recovery were evaluated regarding their antioxidant activity. Thus, with these bioactivity tests, the importance of choosing a good extraction technique such as UAE was emphasized. The use of green solvents in advanced extraction techniques enabled to selectively extract bioactive compounds from T. lutea biomass reducing the amount of time and solvent used, as well as developing more environmental-friendly procedures to obtain enriched extracts with compounds oh high-added value. The selectivity for the extraction of compounds of interest from the two studied solvents was different, with 2-me-THF being the best for obtaining extracts enriched in carotenoids including Fx, DHA, and polar lipids. Extracts enriched in DHA coupled to polar lipids and Fx had shown promising antioxidant activity for food applications.

In future works, the evaluation of other antioxidant compounds known in T. lutea such as brassicasterol, stigmasterol or fucosterol (Mayer et al., 2021) which have a remarkable antioxidant capacity will be studied. In addition, best extracts will be studied using antioxidant and anti-inflammatory in vitro tests as well as for the preparation of microcapsules with interest in food industry and further applications.