Polyunsaturated Fatty Acids of Marine Macroalgae: Potential for Nutritional and Pharmaceutical Applications

As mammals are unable to synthesize essential polyunsaturated fatty acids (PUFA), these compounds need to be taken in through diet. Nowadays, obtaining essential PUFA in diet is becoming increasingly difficult; therefore this work investigated the suitability of using macroalgae as novel dietary sources of PUFA. Hence, 17 macroalgal species from three different phyla (Chlorophyta, Phaeophyta and Rhodophyta) were analyzed and their fatty acid methyl esters (FAME) profile was assessed. Each phylum presented a characteristic fatty acid signature as evidenced by clustering of PUFA profiles of algae belonging to the same phylum in a Principal Components Analysis. The major PUFA detected in all phyla were C18 and C20, namely linoleic, arachidonic and eicosapentaenoic acids. The obtained data showed that rhodophytes and phaeophytes have higher concentrations of PUFA, particularly from the n-3 series, thereby being a better source of these compounds. Moreover, rhodophytes and phaeophytes presented “healthier” ∑n-6/∑n-3 and PUFA/saturated fatty acid ratios than chlorophytes. Ulva was an exception within the Chlorophyta, as it presented high concentrations of n-3 PUFA, α-linolenic acid in particular. In conclusion, macroalgae can be considered as a potential source for large-scale production of essential PUFA with wide applications in the nutraceutical and pharmacological industries.

the relative ease of cultivation and harvesting of macroalgae [24], this suggests that these marine photosynthetic organisms can be viable, sustainable sources of PUFA.
Hence, the objective of this work is to assess the potential of several Chlorophyta, Phaeophyta and Rhodophyta algae, found on the Algarve coast (Portugal), as a source of PUFA and/or specifically VLCPUFA. To the authors' knowledge, the fatty acid composition of Cladophora albida, Cladostephus spongiosus, Dictyota spiralis, Bornetia secundiflora and Asparagopsis armata have yet to be published, thus providing an opportunity to widen the range of macroalgae strains with potential nutritional and/or pharmaceutical applications. Although the FA profiles of some of the target species in this work have already been characterized, intra-specific variability is common in macroalgae coming from different geographical locations, resulting in different FA profiles. This might be explained by exposure to diverse abiotic factors (e.g., temperature) that are known to influence the content of PUFA in algae [25,26]. The partial β-oxidation that results in docosapentaenoic acid (DPA) or docosahexaenoic acid (DHA) formation is highlighted in blue. Differences occurring at each step are marked red in the chemical structure. Adapted from Marszalek and Lodish [19].

FAME Concentration
Total FAME concentration ranged from 2.1 in Jania sp. to 13.0 mg/g of dry weight (DW) in Dictyota spiralis (Figure 2). In the Chlorophyta algae, total FAME concentrations varied between 5.2 and 7.5 mg/g, except for Cladophora albida (9.5 mg/g). The Phaeophyta phylum presented significantly higher concentrations of total FAME than the other phyla (5.6-13.0 mg/g), namely in Dictyota dichotoma, D. spiralis, Taonia atomaria and Cladostephus spongiosus (p < 0.05). The lowest concentration of total FAME was recorded in the rhodophytes (p < 0.05), with all species presenting less than 5.5 mg/g. The relative concentration of lipids and corresponding FAME regarding each phylum is in accordance with previous reports [24].

Figure 2.
Total FAME concentration of macroalgae from three different phyla (Chlorophyta, Phaeophyta and Rhodophyta). Error bars represent the standard deviation from four replicates.

Chlorophyta
The six species of the phylum Chlorophyta analyzed represent three different orders: Bryopsidales (Codium sp. and C. fragile), Cladophorales (Cladophora albida and Chaetomorpha sp.), and Ulvales (Enteromorpha sp. and Ulva sp.). The most abundant FA in this phylum were palmitic (C16:0), myristic (C14:0), behenic (C22:0), palmitoleic (C16:1n-7), oleic (C18:1n-9c) and linoleic (C18:2n-6) acids (Table 1). These results are consistent with those found in the related literature for the same genera although for different species [27][28][29]. All species analyzed in this study presented considerably higher amounts of saturated fatty acids (SFA) than those reported previously. While in this study the SFA were more than 50% of the total detected FA, in the literature SFA relative amounts varied between 25% and 38% [27][28][29][30]. Conversely, the total concentration of PUFA of the Chlorophyta in this study ranged between 17% and 35%, significantly lower than those reported by other authors (37%-64%) [27][28][29][30]. Nevertheless, the C 18 fatty acids were generally the dominant PUFA in all lipid profiles analyzed. Linoleic acid (LA; C18:2n-6) was the main PUFA of most chlorophytes. The only exception was Ulva sp., in which higher percentages of ALA (16%) were detected, in comparison to LA (5.7%). This ALA content is in accordance with earlier publications in which this FA was considered as characteristic of the Ulvales [24,[26][27][28]. The lipid profiles of Codium sp. and Chaetomorpha sp. were the richest in terms of unsaturated fatty acids, while Enteromorpha sp. had the lowest PUFA content. Codium sp. was the only representative of this phylum in which γ-linolenic acid (GLA; C18:3n-6) was detected although it has been previously reported that species of the same genera contain this FA in minimal amounts (0.2%-2.3%) [27,28]. Both Codium species presented relatively high concentrations of n-3 hexadecatrienoic acid (C16:3n-3). Though this is a common FA within the Codium genus [28,29,31], the remaining analyzed species showed only trace amounts. EPA (20:5n-3) was detected in all chlorophytes at medium concentrations. In this phylum, EPA content ranged between 1% and 4% of the total fatty acid content and among the three analyzed phyla showed a trend similar to that of total PUFA, as also noticed by other authors [24]. DHA was only detected in C. albida (0.8%) and is fairly absent from this phylum, being reported in the literature at percentages lower than 1% [27][28][29].
This variability was also reported by other studies in which SFA relative amounts ranged between 26% and 71% [27,32]. Similarly to what is described in the literature, the most abundant SFA in all strains studied were myristic and palmitic acids (Table 3) [27,29,30,32]. Total MUFA content of this phylum was lower than 10% of the total FA profile, except for A. armata and B. secundiflora, which presented slightly higher concentrations. Palmitoleic (C16:1n-7) and oleic (C18:1n-9c) acids were once more the main MUFA (Table 3). Other authors have described relatively higher amounts of MUFA in other rhodophytes although these were consistently the least representative of all FA [27,29,30,32]. Regarding PUFA, contrary to the other two phyla studied, linoleic acid (C18:2n-6) is not a major PUFA, reaching only 2% of the total FA. The most abundant PUFA in this phylum were AA (C20:4n-6) and EPA (20:5n-3), which are usually the most predominant FA in red algae [26,27,30,32,36]. However, as in SFA, there was a wide range in the total concentration of PUFA detected within species of this phylum: A. armata displayed only 5%, whereas in Peyssonnelia sp. PUFA content reached 52% of the total FA. Wide variability of PUFA content was also found by Graeve et al. [29], who reported a range in PUFA contents of 18%-63% in red algae from Arctic and Antarctic waters. Similarly, Li et al. [27] described PUFA contents ranging from 8% to 55% in rhodophytes from the Bohai Sea. Contrary to the other two phyla, in Rhodophyta, C 18 are not the most representative PUFA. In fact, rhodophytes and phaeophytes exhibited considerably higher concentrations of C 20 PUFA than chlorophytes [3,27,30,32,37]. In this study, except for A. armata, all strains exhibit considerably high amounts of EPA (15% and 27% of total FA). In the literature, Rhodophytes are commonly reported has good EPA producers, which suggests that red algae may be the best source of this nutritionally important fatty acid [27,29,30,32,38]. Peyssonnelia sp. was the only representative of this phylum in which DHA (C22:6n-3) was detected. In fact, this species presented significantly higher concentrations of this fatty acid than all algae studied (p < 0.05), reaching nearly 5% of the total FA (Table 3). DHA is often not found in red algae, or when present exists at low concentrations. Other authors have reported relative DHA amounts of 0.3% to 1.5% for several rhodophytes, which are significantly lower than those reported for Peyssonnelia in this study [29,38]. Table 3. Fatty acid profile of the rhodophytes Jania sp., Pterocladiella capillacea, Asparagopsis armata, Peyssonnelia sp., and Bornetia secundiflora. Values are given as means of total FAME percentage ± standard deviation (n = 4). n.d., not detected.

Multivariate Analysis
Principal Components Analysis (PCA) was performed on the 21 detected fatty acids as a proportion of total fatty acid content, to assess the relationship between the 17 strains belonging to the three different phyla. The levels of capric (C10:0), myristic, hexadecatrienoic (n-3 and n-6) and ALA were removed from the multivariate analysis due to their low variation, therefore promoting a more reliable analysis. The first two components explain 46.6% of the total variation, 26% for PC1 and 20.5% for PC2 (both significant). In the loading scatter plot of the variables (FA; Figure 3), AA, EPA, DHA, pentadecylic (C15:0), behenic (C22:0), lignoceric (C24:0) and hexadecadienoic (C16:2n-6) acids were the most discriminant variables along PC1, while palmitic (C16:0), oleic (C18:1n-9c), elaidic (C18:1n-9t) and dihomo-γ-linolenic (C20:3n-6) acids were the main discriminant FA along PC2. In this plot, FA are distributed according to their saturation: most SFA are placed on the upper quadrants while most PUFA grouped in the lower left quadrant. In the scores plot, species clustered according to their phylum (Figure 4), suggesting that each phylum has a distinct FA profile and supporting earlier evidence that lipid composition may be a biochemical marker for each taxonomic group [24,36,39]. Besides its evident application in phylogenic studies, this feature may help to assign different algae to distinct diets. In this sense, the placement of the rhodophytes and phaeophytes in the left quadrants, and of all chlorophytes in the right quadrants, indicate that the first two phyla display a lipid profile clearly enriched in pentadecylic, stearic, EPA, DHA, AA and eicosenoic (C20:1) acids. Within the Chlorophyta phylum, Codium strains are located further to the right due to their increased content in n-3 hexadecatrienoic (C16:3n-3), behenic (C22:0) and lignoceric (C24:0) acids. The Rhodophyta and Phaeophyta species are separated only along PC2, with the relative proportions of AA and dihomo--linolenic acid as the main discriminant factors (Figure 4).

Nutritional and Pharmaceutical Applications
PUFA are vital components in human nutrition and are known to have several beneficial effects for human health. A diet intake of PUFA, including both n-3 and n-6 fatty acids, is known to modulate inflammatory processes among other cell functions. Although many of the species analyzed in this work displayed high amounts of SFA, some Phaeophyta and Rhodophyta species exhibited higher concentrations of PUFA, and PUFA/SFA ratios higher than 1 (H. scoparia, 1.46; T. atomaria, 1.33; C. spongiosus, 1.77; Peyssonnelia sp., 1.33). The lowest PUFA/SFA ratios were observed in algae from the phylum Chlorophyta (0.27-0.68). It appears that this phylum has a lower potential, comparing to the other two phyla studied, as a nutritional source of PUFA for human consumption. The results presented herein are in agreement with previous studies in which rhodophytes and phaeophytes displayed higher concentrations of unsaturated fatty acids as compared with chlorophytes [24]. However, not all PUFA are associated with the promotion of health benefits. For example, in the inflammation process, eicosanoids derived from n-6 PUFA are generally considered as pro-inflammatory or as promoters of other cell harmful effects, whereas n-3 PUFA derivatives are considered less inflammatory or even anti-inflammatory [8,40,41]. Since the biosynthetic pathway of these fatty acids relies on the same enzymes for n-3 and n-6 PUFA, the health promoting effects are dependent on the n-6/n-3 ratio of PUFA obtained through diet. The World Health Organization (WHO) recommends a ∑n-6/∑n-3 ratio lower than 10 [21,23]. In this study, almost all algae can be considered as a good source of dietary PUFA, since they showed ratios ranging between 0.29 and 6.73. The exception was Chaetomorpha sp., in which the ∑n-6/∑n-3 ratio was the highest from all the studied species (31.25) and in D. spiralis in which no n-3 fatty acids were detected.
Besides an appropriate nutritional profile, these macroalgae can also be exploited for pharmaceutical purposes. Many of the PUFA detected throughout this work are considered powerful molecules against several diseases and are already used in different biomedical applications. For example, several reports suggest that n-3 fatty acids, mainly EPA and DHA, may have a significant potential in the treatment of autoimmune and inflammatory diseases [42]. In this study, Rhodophyta was the phylum with the highest percentage of n-3 fatty acids (16%-27% of total FAME), followed by Phaeophyta (0%-15%), in which significant amounts of n-3 were also present. Aside from Ulva sp. that had 18% of n-3 FAME, Chlorophyta macroalgae presented the lowest values of n-3 fatty acids (1%-9%). Conversely, the detected n-6 fatty acids were lower in rhodophytes (8%-15%), due to the low concentration of linoleic acid, except for Peyssonnelia sp., where n-6 concentration was approximately 28% of total FAME. Phaeophytes showed the highest contents of n-6 fatty acids (23%-44%), whereas chlorophytes presented mid-range values (6%-27%). Considering the absolute concentrations of PUFA in the various species tested, Ulva sp., T. atomaria, C. spongiosus, Peyssonnelia sp. and B. secundiflora possess the highest contents of n-3 PUFA, 1.07, 1.38, 1.19, 1.06 and 1.42 mg/g, respectively. Apart from Ulva sp., in which ALA dominated, the n-3 profile of the remaining strains was essentially composed of EPA. DHA was not a major PUFA in any of the algae studied in this work. Nevertheless, Peyssonnelia sp. exhibited a relatively high content of DHA, 0.22 mg/g of dry biomass, coupled with an EPA concentration of 0.84 mg/g. A variety of potential applications are described for EPA and DHA, which hold significant potential for pharmaceutical purposes, namely cancer treatment, asthma, psoriasis, rheumatoid arthritis, antibiotic, inflammatory bowel disease, depression, allergies, cardiovascular diseases, among others [1,20]. More recently PUFA proved to have a strong potential in drug delivery; in addition to the described cytotoxicity of a few PUFA, PUFA enable a more efficient penetration of specific molecules through the cell membranes of tumor cells, due to their unique lipophilic characteristics [43]. In fact, several studies show that tumor cells display faster PUFA intake than normal cells, as demonstrated for the conjugated taxoid DHA-paclitaxel [12].
The nutritional and pharmaceutical benefits of PUFA, however, contrast with the increasing difficulty in finding sustainable sources of n-3 VLCPUFA, which traditionally were obtained from fish and fish oil. Declining fish stocks caused by decades of overfishing [44] makes ever more urgent to find non-traditional alternatives for the western world. As VLCPUFA are usually absent from terrestrial higher plants [45], traditional crops can also be excluded as viable sources of these FA. Though this deficiency can be overcome by applying genetic engineering, transgenic foods are not always well accepted by the general public. Therefore, n-3 VLCPUFA are typically associated with marine organisms, and algae, as the basis of the marine trophic chain, come out as a very promising source of VLCPUFA. In fact, large scale farming of marine algae has been accomplished successfully for hundreds of years [46]. Approximately 220 algal species are currently cultivated and harvested all over the world for different purposes [47]. Though mostly used as food for human consumption, particularly in Asia, macroalgae are also the primary source of hydrocolloids such as agar, carrageenan and alginate, which have numerous industrial applications, such as gelling, stabilizing or binding agents [47][48][49]. The next step could well be the sustainable exploitation of marine macroalgae as alternative sources of VLCPUFA, not only in Asia, but also in the western world.

FAME Preparation
Lipids and free fatty acids were converted to the corresponding FAME, according to a modified protocol of Lepage and Roy [50]. This method is based on the direct transesterification with acetyl chloride/methanol, followed by direct extraction of the lipidic phase into hexane. Briefly, 0.1 g of algal biomass was weighed and treated with 1.5 mL of derivatization solution (methanol/acetyl chloride, 20:1, v/v), in reaction vessels. The biomass was disrupted with an IKA Ultra-Turrax disperser and afterwards 1 mL of hexane was added and the mixture heated for 1 hour at 100 °C . After cooling in an ice bath, 1 mL of distilled water was added and the organic phase was removed and dried with anhydrous sodium sulfate. The extracts were then filtered and stored at −20 °C until further analysis.

Determination of FAME Profile by GC-MS
FAME were analyzed on an Agilent GC-MS (Agilent Technologies 6890 Network GC System, 5973 Inert Mass Selective Detector) equipped with a DB5-MS capillary column (25 m × 0.25 mm internal diameter, 0.25 µm film thickness, Agilent Tech) using helium as carrier gas. Samples were injected at 300 °C and the temperature profile of the GC oven was 60 °C (1 min), 30 °C min −1 to 120 °C , 5 °C min −1 to 250 °C , and 20 °C min −1 to 300 °C (2 min). For the identification and quantification of FAME, the total ion mode was used. A -Supelco ® 37 Component FAME Mix‖ (Sigma-Aldrich, Sintra, Portugal) was used as a standard and separate calibration curves were generated for each of the FAME in this standard. When there was no standard available, the calibration curve of the most similar FAME in terms of structure was used. Values were expressed as mg/g of dry weight.

Statistical Analysis
Obtained results are presented as mean ± standard deviation; all analyses were performed in quadruplicate. Differences between species or phyla were assessed using analysis of variance (one-way ANOVA) while differences between FAME groups were determined using the t-test for independent samples. Significant differences were considered when p < 0.05 by means of the statistical program StatSoft STATISTICA (release 7.0). PCA was used to compare the FAME profiles of different macroalgae, as previously published by other authors [24,36,39]. PCA is used to transform an original set of potentially correlated variables into a reduced set of uncorrelated variables that are called principal components. These components are obtained in order of decreasing importance. The first principal component explains the most variance; the second principal component explains the next degree of variance. For each sample, PCA calculates a -score‖ and two-dimensional plots of the scores for the first two principal components, which may reveal clusters and trends in data. Principal component analysis was performed in Umetrics SIMCA-P software (release 12.0.1).

Conclusions
VLCPUFA are well known bioactive molecules with important nutritional and pharmaceutical applications. The FA content of the macroalgae studied displayed similar signatures within species of the same phylum. However, when chlorophytes, phaeophytes and rhodophytes were compared among each other, distinct FA profile clusters per phylum were observed. Representatives of the Rhodophyta and Phaeophyta had significantly higher concentrations of VLCPUFA, namely the C 20 , AA and EPA. In particular, Ulva sp. had a lipid profile particularly enriched in the n-3 fatty acid ALA (16%), whereas T. atomaria, C. spongiosus and B. secundiflora presented the highest contents of EPA (>1 mg/g). DHA was not a dominant VLCPUFA in the studied species, although Peyssonnelia sp had 0.22 mg/g of this FA, nearly 5% of the total FAME. In combination with 0.84 mg/g of EPA the latter macroalga makes an excellent and balanced source of n-3 VLCPUFA. Taken together, these results suggest that most macroalgae may provide human beings with beneficial FA in their diets if used as nutrient sources in food products commonly consumed in the western world.