Cyanobacteria and Red Macroalgae as Potential Sources of Antioxidants and UV Radiation-Absorbing Compounds for Cosmeceutical Applications

In recent years, research on natural products has gained considerable attention, particularly in the cosmetic industry, which is looking for new bio-active and biodegradable molecules. In this study, cosmetic properties of cyanobacteria and red macroalgae were analyzed. The extractions were conducted in different solvents (water, ethanol and two combinations of water:ethanol). The main molecules with antioxidant and photoprotective capacity were mycosporine-like amino acids (MAAs), scytonemin and phenolic compounds. The highest contents of scytonemin (only present in cyanobacteria) were observed in Scytonema sp. (BEA 1603B) and Lyngbya sp. (BEA 1328B). The highest concentrations of MAAs were found in the red macroalgae Porphyra umbilicalis, Gelidium corneum and Osmundea pinnatifida and in the cyanobacterium Lyngbya sp. Scytonema sp. was the unique species that presented an MAA with maximum absorption in the UV-B band, being identified as mycosporine-glutaminol for the first time in this species. The highest content of polyphenols was observed in Scytonema sp. and P. umbilicalis. Water was the best extraction solvent for MAAs and phenols, whereas scytonemin was better extracted in a less polar solvent such as ethanol:dH2O (4:1). Cyanobacterium extracts presented higher antioxidant activity than those of red macroalgae. Positive correlations of antioxidant activity with different molecules, especially polyphenols, biliproteins and MAAs, were observed. Hydroethanolic extracts of some species incorporated in creams showed an increase in the photoprotection capacity in comparison with the base cream. Extracts of these organisms could be used as natural photoprotectors improving the diversity of sunscreens. The combination of different extracts enriched in scytonemin and MAAs could be useful to design broad-band natural UV-screen cosmeceutical products.


Biochemical Composition
The contents of total internal carbon and nitrogen, proteins, carbohydrates, lipids and the C:N ratios of the different species are shown in Table 1. In most of the cases, significant differences were observed among species (p < 0.05) (Table S1A). In cyanobacteria, the highest internal carbon content (>400 mg g −1 DW) was observed in Scytonema sp., N. commune and Anabaena sp., and the lowest content was obtained in C. scopulorum (313 mg g −1 DW). The nitrogen concentration was also higher in Anabaena sp. and Scytonema sp. (85 and 62 mg g −1 DW, respectively) and lower in C. scopulorum (35 mg g −1 DW). The highest C:N ratio was reached in C. scopulorum (9.1) and the lowest in Anabaena sp. (8.5). The highest total protein content estimated from the N content was observed in Anabaena sp. and the lowest in C. scopulorum (407 and 165 mg g −1 DW, respectively), whereas the highest soluble protein content was reached in N. commune (71 mg g −1 DW) and the lowest in C. scopulorum (8 mg g −1 DW). The highest levels of carbohydrates were observed in N. commune (40.8%) and Lyngbya sp. (408 and 334 mg g −1 DW, respectively), and the lowest amount was obtained in Anabaena sp. and Scytonema sp. (185 and 196 mg g −1 DW, respectively), in spite of the presence of a high internal carbon content. N. commune presented the highest lipid content and Anabaena sp. the lowest (21 and 16 mg g −1 DW, respectively).
In red macroalgae, the highest C content was observed in G. corneum and P. umbilicalis (351 and 340 mg g −1 DW, respectively) and the lowest in G. longissima (256 mg g −1 DW). The N content was lower than the observed in cyanobacteria, being the highest concentration obtained in P. umbilicalis and the lowest in O. pinnatifida (45 and 25 mg g −1 DW). The C:N ratio was higher in G. corneum and O. pinnatifida (10.6 and 9.2, respectively) and the lowest in C. rubrum (6.7). Total protein contents were also the highest in P. umbilicalis and the lowest in O. pinnatifida (220 and 123 mg g −1 DW). Soluble proteins followed the same pattern; the highest content was observed in P. umbilicalis and the lowest in O. pinnatifida (34 and 6 mg g −1 DW). The highest carbohydrate content was observed in P. umbilicalis and G. corneum (318 and 275 mg g −1 DW, respectively) and the lowest in C. rubrum (139 mg g −1 DW). Macroalgae presented low lipid contents; all species showed less than 1%. Table 1. Total internal carbon and nitrogen, proteins, carbohydrates and lipid contents (mg g −1 DW) and the C:N ratio of red algae and cyanobacteria expressed as average ± standard deviation (SD) (n = 3). Different letters indicate significant differences among species of cyanobacteria and red algae (ANOVA, p < 0.05, SNK test). The results of phycobiliprotein and scytonemin contents in the different species are shown in Table 2. In most of the cases, significant differences were observed among species (p < 0.05) (Table S1). Both groups of tested organisms presented biliproteins, although different composition and concentrations were observed. Cyanobacteria normally presented more phycocyanin (PC) than phycoerythrin (PE), whereas red macroalgae were enriched mostly in PE. N. commune presented the highest content of PE (14 mg g −1 DW), followed by Scytonema sp. (3.7 mg g −1 DW). The highest content of phycocyanin was also observed in N. commune (25mg g −1 DW) and Scytonema sp. (19mg g −1 DW). C. scopulorum showed a very low content of biliproteins. In red algae, P. umbilicalis presented the Mar. Drugs 2020, 18, 659 5 of 26 highest content of PE (3.4 mg g −1 DW), and O. pinnatifida the lowest (0.07 mg g −1 DW). In general, phycocyanin contents were low in all red algal species.
Considering the cyanobacteria, the higher contents of scytonemin were found in Scytonema sp. and Lyngbya sp. (2.2 and 1.2 mg g −1 DW, respectively). In the other three species, the contents were significantly much lower (<0.5 mg g −1 DW). Table 2. Phycobiliproteins content of cyanobacteria and red algae species and scytonemin content in cyanobacteria species (mg g −1 DW). Pigments were expressed as average ± standard deviation (SD) (n = 3). PE: phycoerythrin; PC: phycocyanin; Scyt: scytonemin. Different letters indicate significant differences among species of cyanobacteria and red algae, analyzed separately (ANOVA, p < 0.05, SNK test). The phenolic compound contents of the different species are shown in Figure 2. Significant differences were observed among species and among solvents (p < 0.05) (Table S1). In most of the cases (83%), the best solvent for the extraction of phenolic compounds was d H 2 O. In cyanobacteria (Figure 2A), the higher contents were observed in the aqueous extracts of Scytonema sp. and N. commune (22.5 and 13 mg g −1 DW, respectively). In red algae, the aqueous extract of P. umbilicalis and the ethanol: d H 2 O (4:1) of O. pinnatifida (aprox. 13 mg g −1 DW in both) showed the highest values ( Figure 2B). The lowest content was observed in C. scopulorum, G. longissima and G. corneum, and in all cases, ethanol was not an effective solvent for the extraction of these compounds. concentrations were observed. Cyanobacteria normally presented more phycocyanin (PC) than phycoerythrin (PE), whereas red macroalgae were enriched mostly in PE. N. commune presented the highest content of PE (14 mg g −1 DW), followed by Scytonema sp. (3.7 mg g −1 DW). The highest content of phycocyanin was also observed in N. commune (25mg g −1 DW) and Scytonema sp. (19mg g −1 DW). C. scopulorum showed a very low content of biliproteins. In red algae, P. umbilicalis presented the highest content of PE (3.4 mg g −1 DW), and O. pinnatifida the lowest (0.07 mg g −1 DW). In general, phycocyanin contents were low in all red algal species. Considering the cyanobacteria, the higher contents of scytonemin were found in Scytonema sp. and Lyngbya sp. (2.2 and 1.2 mg g −1 DW, respectively). In the other three species, the contents were significantly much lower (<0.5 mg g −1 DW). Table 2. Phycobiliproteins content of cyanobacteria and red algae species and scytonemin content in cyanobacteria species (mg g −1 DW). Pigments were expressed as average ± standard deviation (SD) (n = 3). PE: phycoerythrin; PC: phycocyanin; Scyt: scytonemin. Different letters indicate significant differences among species of cyanobacteria and red algae, analyzed separately (ANOVA, p < 0.05, SNK test). The phenolic compound contents of the different species are shown in Figure 2. Significant differences were observed among species and among solvents (p < 0.05) (Table S1). In most of the cases (83%), the best solvent for the extraction of phenolic compounds was dH2O. In cyanobacteria (Figure 2A), the higher contents were observed in the aqueous extracts of Scytonema sp. and N. commune (22.5 and 13 mg g −1 DW, respectively). In red algae, the aqueous extract of P. umbilicalis and the ethanol:dH2O (4:1) of O. pinnatifida (aprox. 13 mg g −1 DW in both) showed the highest values ( Figure 2B). The lowest content was observed in C. scopulorum, G. longissima and G. corneum, and in all cases, ethanol was not an effective solvent for the extraction of these compounds. The quantitative and qualitative MAA contents are shown in Table 3. In relation to the total MAAs content, significant differences were observed among species (p < 0.05) (Table S1). Only two species of cyanobacteria showed relevant concentration of these molecules (Scytonema sp. and Lyngbya sp.), whereas all the red algae presented significant amounts of MAAs. In cyanobacteria, Lyngbya sp. showed the highest content (1.8 mg g −1 DW), whereas in red algae, the highest content of MAAs (5.2 mg g −1 DW) was observed in P. umbilicalis, followed by G. corneum and O. pinnatifida (1.8 and 1.5 mg g −1 DW, respectively). Considering the different types of MAAs detected using HPLC and ESI-MS, Scytonema sp. presented a unique peak at 310 nm (identified as mycosporine-glutaminol). Lyngbya sp. presented a predominant peak at 330 nm (asterina-330, >85%) and a small peak at 320 nm (palythine). P. umbilicalis showed the highest peak at 334 nm (porphyra-334, >85%) and other peaks at 320, 330, 332 and 333 nm (palythine, asterina-330, palythinol and shinorine, respectively). G. longissima only presented a peak at 332 nm (shinorine). G. corneum presented a predominant peak at 330 nm (asterina-330, >60%) and two other peaks at 320 and 332 nm (palythine and shinorine, respectively). C. rubrum presented different peaks at 320, 332 and 340 nm (palythine, shinorine and an unknown absorbing compound, respectively).

Antioxidant Activity
Results of the antioxidant activity are shown in Figure 3. In general, cyanobacteria showed a higher antioxidant activity compared to red macroalgae, reaching values close to 35 µmol TE g −1 DW ( Figure 3A-E), whereas maximal values of red macroalgae were around 15 µmol TE g −1 DW ( Figure 3F-J). In all cases, significant differences (p < 0.05) were found motivated by the interaction between solvents and extraction procedures (Table S1). In most of the cases, the ABTS assay reached higher values than the DPPH assay, and among solvents, different responses were observed. In cyanobacteria, Scytonema sp. presented the highest activity with the ABTS assay (specially using d H 2 O and ethanol: d H 2 O (4:1) as solvent; approx. 35 µmol TE −1 DW). However, with the DPPH method, this species only showed a high value with the aqueous solvent (25 µmol TE g −1 DW) ( Figure 3C). Anabaena sp. showed higher values with the ABTS assay (ethanol: d H 2 O (4:1): 30 µmol TE g −1 DW), but no antioxidant response was detected when DPPH was used ( Figure 3B). N. commune presented the highest antioxidant activity in aqueous solvent (in both ABTS and DPPH assays: 20 µmol TE g −1 DW) ( Figure 3A). Lyngbya sp.

Photoprotective Capacity
The absorptions of the UV radiation by the creams including algae extracts are shown in Figure 4A. The capacity to protect against this type of radiation measured through two different indices is shown in Table 4. All creams including algal extracts increased the photoprotection when compared with the base cream ( Figure 4A,B). In most of the cases, significant differences (p < 0.05) were observed when comparing the different creams containing algal extracts (Table S1).
In the cream with P. umbilicalis extract, the highest absorptance increase was observed in the UV-A region. Between 335 and 340 nm, the absorptance increased by close to 40% in comparison with the base cream. On the other hand, the cream with Scytonema sp. extract presented a highest absorption increase in the UV-B region. Between 309 and 312 nm, the absorptance increased by 15% in comparison with the base cream ( Figure 4B).
Considering the calculated indexes, the cream with red algae extract reached an SPF of 2.1, which corresponded with 50% of the ESAR, followed by the protection against elastosis (1.5) and UVAPF (1.4) with an ESAR of 24 and 23%, respectively. Using the cyanobacteria extract, the SPF was similar to that obtained with red macroalgae, 1.9 and an ESAR of 47%, whereas the protection against UV-A rays was lower.  Table 4. Sun protection factor (SPF), UV-A protection factor (UVAPF) and biological effective protection factors (BEPFs) against elastosis and photoaging and % of effective solar absorpted radiation (ESAR) against the different action spectra (erythema, PPD, elastosis and photoaging) of the different creams (base cream, creams including algae extracts and control cream). Data are expressed as average ± standard deviation (SD) (n = 4). Different letters indicate significant differences between species and base cream (ANOVA, p < 0.05, SNK test).

Correlations and Principal Component Analyses
Positive correlations (Table S2 and S3) (p < 0.05) were observed between antioxidant activity, mainly ABTS, and different molecules. In cyanobacteria, a high correlation was observed between the ABTS assay (in all solvents) and phenols content (in all solvents), e.g., for ABTS d H 2 1)) also showed positive correlation with polar molecules such as carbonated compounds (r = 0.7), soluble proteins (r = 0.6) and phycocyanin (r = 0.6). On the other hand, scytonemin showed positive correlations with ABTS and DPPH results (mainly with DPPH in ethanol, r = 0.977). DPPH also presented a positive correlation with phenols content.
In red algae, positive correlations were also obtained between ABTS and phenols content, e.g., ABTS vs. phenols in aqueous solvents reached a correlation coefficient r equal to 0.951. However, DPPH results did not show positive correlations with phenols content. The ABTS values obtained for aqueous solvents also showed a positive correlation with nitrogenous compounds: total nitrogen (r = 0.514), total proteins (r = 0.514), soluble proteins (r = 0.696), phycoerythrin (r = 0.576) and MAAs (r = 0.809).
The results of the PCA ( Figure 5) performed with all biochemical analysis and antioxidant activities showed that two principal components explained 96.8% of the variation. The first principal component (PC1) explained 79.1%, whereas the second (PC2) represented 17.7% of the total data variation. Based on the PCA graph and the correlations obtained between all variables and the different principal components (PC-axes) (Table S4), the PC1 was related with the polarity of the molecules, aqueous compounds (such carbohydrates, soluble proteins, phycobiliproteins and antioxidant activity obtained in aqueous solvents) being more correlated with this axis. The variables that most influence this axis are carbohydrates (quadrant 4 of Figure 5) with a correlation index of 0.959, followed by soluble proteins (0.711) and phycobiliproteins (PE and PC, 0.709 and 0.604), respectively, and antioxidant activities under more polar solvents (ABTS.1, ABTS.2 and DPPH.1) with correlation indexes of 0.726, 0.583 and 0.680, respectively (Table S4). The PC1 axis also divided some species in different groups: (1) three red macroalgae (C. rubrum, O. pinnatifida and G. longissima) grouped in quadrant 3 ( Figure 5), (2) the other two red macroalgae and one cyanobacteria (G. corneum, P. umbilicalis and C. scopulorum) grouped in the central area of the axis mainly associated with quadrant 4 and (3) two cyanobacteria (N. commune and Scytonema sp.) grouped in quadrant 2. N. commune and Scytonema sp. are the species more associated with this axis, appearing in the first quadrant ( Figure 5). In relation to the PC2 axis, the variable with the highest correlation is the N content with a correlation index of 0.742; other variables that can explain this axis are ABTS.3 and total carbon (0.640 and 0.590, respectively) (Table S4). This axis only separated clearly one cyanobacteria, Anabaena sp.

Biochemical Composition
Algae contain several molecules in their composition with high interest due to their bioactivities. These compounds are normally divided into carbonated and nitrogenated ones. In general, the internal carbon (C) and nitrogen (N) contents and the C:N ratio can be physiological indicators, showing if algae are living in optimal conditions or under nitrogen limitation. In this study, the N content was lower in macroalgae than that in cyanobacteria, probably because the first ones were collected from their natural habitat, whereas the cyanobacteria biomasses were obtained by cultures under N-enriched supply and controlled growth conditions. The N content obtained in this study varied between 2.5 and 4.5% in red macroalgae and between 3.5 and 8.5% in cyanobacteria. Similar results were obtained by other authors in collected macroalgae [36,37]. In microalgae and cyanobacteria, the N percentage can also vary depending on the phase of the culture [38]. The optimal C:N ratio was described as 10 [39], whereas a higher proportion would indicate N limitation. In this study, all species showed a C:N ratio less than or equal to 10, indicating that the species used in this study were not N-limited.
All of the species evaluated in our study presented a high content of total internal carbon, i.e., most species values were higher than 30% of DW, similar to the average C content obtained in algae [40]. Considering C compounds, polysaccharides have acquired a great importance (in this work only the total C and total carbohydrates content were measured, the polysaccharides were not identified or quantified). Red algae present wall polysaccharides such as agars or carrageenans that are used as gelling and stabilizing agents [41,42] and could be used as excipients or functional ingredients in cosmetic formulations [15]. Polysaccharides have also been studied for their antitumoral, antioxidant and immunological activities [43][44][45]. These compounds in cyanobacteria have been less studied, although some authors reported the production of exopolysacharides (EPS) by some cyanobacteria with different bioactivities [46][47][48].
Macroalgae normally present low lipid content (as observed in this study, <2% DW) [49]. Cyanobacteria also showed low values, whereas other authors obtained higher values. In spite of the low values, these organisms can produce high value fatty acids [50]. A possible cosmetic application of lipids in cosmetic products is the avoidance of skin dehydration [15].
Phenolic compounds are C compounds with a role as natural antioxidants, as it has been shown by several authors [51,52]. More recently, they have also been studied for their photoprotective applications [53][54][55]. Brown algae are the main producers of these compounds, however, cyanobacteria and red algae can also produce phenols, although in less quantity [37]. In this work, it was possible to find an optimization of phenolics extraction when different solvents were utilized. In our case, water was the best solvent for their extraction, similarly to other studies [52,56]. In red algae, the highest content was observed in P. umbilicalis and O. pinnatifida (approx. 14 mg g −1 DW). High levels of polyphenols (25 mg GAE (gallic acid equivalent) g −1 DW) were also observed in Porphyra tenera [57], applying similar extraction conditions. Zubia et al. obtained similar results in species from the Gracilariales family [58]. Less information was found in relation to the phenolic content in cyanobacteria. Li [60].
On N compounds, proteins or derivates (peptides or aminoacids) show antioxidant or immune-stimulant properties and can also confer moisture retention in the skin [61]. Total and soluble proteins were in general higher in three species of cyanobacteria (N. commune, Anabaena sp. and Scytonema sp). The soluble proteins represent about 20-25% of total proteins except in Anabaena sp. (9.9%) and C. scopulorum (5%), thus, the level of total protein in Anabaena sp. is much higher than the other species (40.6%). This can be explained by the contribution of structural proteins, which could be higher than in the other cyanobacteria. It is also necessary to remark that total protein was estimated from a nitrogen to protein conversion factor; it would be possible that the applied factor overestimated in the case of Anabaena sp. due to its amino acid profile. Among the red algae, only P. umbilicalis presents high values, close to most cyanobacteria. P. umbilicalis is a species known for its used as human food (e.g., nori) [62]. Nostoc sp. and Anabena sp. also present a high nutritional value and are used as food in some countries [63].
Phycobiliproteins are N molecules that act as accessory pigments responsible for the light harvesting in both group of organisms analyzed. Cyanobacteria presented a high content of biliproteins, mainly phycocyanin. The percentage of biliproteins related to soluble proteins was higher than 30% in N. commune (53.8%), Scytonema sp. (37%) and Lyngbya sp. (31.8%), whereas the other two species presented much lower percentages, e.g., C. scopulorum (18.75%) and Anabaena sp. (7.5%). Other authors obtained similar and higher values of biliproteins than those obtained in this study [64][65][66]. In red algae, the percentage of biliproteins related to soluble proteins was higher in P. umbilicalis (13.3%) and G. corneum (13.6%), in comparison with the other species, e.g., G. longissima (10%), C. rubrum (5.2%) and O. pinnatifida (3.7%). The level of biliproteins was lower than that obtained in collected algae by other authors [67,68]. Biliprotein concentrations can vary depending on environmental factors like irradiance, light quality or nutrients [69][70][71]. The extraction method can also influence the content, e.g., a great number of authors used the freezing and thawing method [72].
Scytonemin, one of the main photoprotective molecules found in cyanobacteria, is a lipophylic and yellow-brown pigment with the capacity to absorb UVR, mainly in the UV-A-violet-blue region with a maximum absorption at 384 nm [73]. This pigment is only presented in the extracellular sheaths of some cyanobacteria [74,75]. All the cyanobacteria tested showed scytonemin in their composition, in contrast to Mushir et al., who did not observe this pigment in Lyngbya sp. or Calothrix brevissema [76]. Other authors also observed scytonemin in the same species tested in this study [77][78][79][80]. In this study the highest level of scytonemin was reached in Scytonema sp. followed by Lyngbya sp., showing similar amounts as those reported in other studies [76,81]. These results can be compared with the UV spectra, in which the scytonemin peaks (380 nm) were visible only in the most polar solvents (ethanol: d H 2 O (4:1) and pure ethanol) and were more prominent in Scytonema sp., Lyngbya sp. and N. commune. The culture or environment conditions can vary the scytonemin content, e.g., nitrogen deficiency can increase scytonemin [82].
Mycosporine-like amino acids (MAAs) are secondary N metabolites also implicated in the protection against the UV radiation and mainly reported in the two groups of organisms tested in this study [83,84]. Most of the species analyzed presented MAAs. In the case of cyanobacteria, only Scytonema sp. and Lyngbya sp. showed quantified peaks in the HPLC. Other authors also observed MAAs in different species of the cyanobacteria genera tested [85][86][87]. Sinha et al. also identified an excreted UV absorbing compound in the culture media of Scytonema sp., although in our study the medium was not analyzed [88]. In red algae, different authors observed the highest content of MAAs in the order Bangiales in which P. umbilicalis is included [89]. The concentrations of MAAs obtained in this study were similar to those observed in the bibliography [89][90][91][92][93][94].
There are different types of MAAs that differ in the chemical formula and the λ max [84,95]. Scytonema sp. was the unique species that presented an MAA with a λ max in the UV-B region. In this study, it was identified as mycosporine-glutaminol for the first time in Scytonema sp., although Rastogi et al. observed mycosporine-glycine and the unknown MAA-334 in this species [72]. Rastogi and Incharoensakdi also observed palythine and asterina-330 in Lyngbya sp. [78]. In red algae, similar patterns were observed by other authors. In P. umbilicalis, different MAAs were detected, the predominant being porphyra-334, as observed in Porphyra spp. or Pyropia spp. [92,96]. In G. longissima, only shinorine was detected, in contrast to other authors who in the same specie observed more types of MAAs [96][97][98]. Briani et al. also studied species of the genus Gelidium obtaining similar results, high content of asterina-330 and shinorine [92]. In the genera Laurencia and Osmundea, a high content of asterina-330 was also observed [92,99]. In C. rubrum, shinorine was also detected, whereas other authors also observed different MAAs in similar species [92,96]. The similarities in the qualitative composition of MAAs found by different authors in similar species suggest that the composition of the main MAAs can be conservative among species, although the percentage of each one can be modified by some environmental factors, such as the presence of UV radiation or a high intensity of solar radiation [99,100].
The different composition of MAAs among species is relevant for cosmeceutical applications, since not all MAAs absorb in the same wavelengths or present the same antioxidant capacity [101][102][103]. The MAAs concentration/composition can be compared with the UV absorption spectra. A peak in the UV-B region (310-312 nm) was only observed in Scytonema sp., whereas the other species present peaks with maximum absorbance between 330-335 nm depending on the predominant MAAs. The highest peak was observed in P. umbilicalis, which means the highest content in MAAs. The UV spectra agreed with the HPLC results. MAAs peaks were similar in most of the solvents, except in pure ethanol. MAAs are water soluble molecules, but these results suggest that only 20% of water can extract approximately the same as 100% water.

Bioactivities (Antioxidant and Photoprotective) and its Relation with Biochemical Composition
Most of the above discussed molecules can act as antioxidant compounds. The general antioxidant activity measured through two different methods based on the free radical scavenging activity was higher in cyanobacteria than in red macroalgae. In cyanobacteria, the highest antioxidant activity was obtained in the order Nostocales (the group of Scytonema sp., Anabaena sp. and Nostoc sp.) as has been reported by other authors [60,[104][105][106]. Considering the red macroalgae, the highest antioxidant activity was reached in the species O. pinnatifida, P. umbilicalis and C. rubrum. Other authors also found a high antioxidant activity in the order Ceramiales (where O. pinnatifida and C. rubrum belong), Gracilariales and Bangiales [37,58,97,107].
The differences among solvents used for the extractions can depend on the molecules extracted in each solvent, i.e., proteins, carbohydrates, MAAs and phycobiliproteins are water soluble, whereas chlorophylls, carotenoids and lipids are soluble inorganic solvents like ethanol, methanol or acetone. Many authors also studied the influence of solvents in antioxidant activity, and different results were observed depending on the species and method applied [97,108,109]. In this work only aqueous and ethanolic solvents were tested due to the non-toxicity needed for cosmetic purposes. The method used also influenced the results. In this study, most of the cases showed the highest antioxidant activity using the ABTS assay [109]. ABTS is dissolved in d H 2 O but can also be dissolved in water based solvents such as ethanol/methanol:water (1:1) [110], whereas DPPH can only be dissolved in organic media, limiting the measure of hydrophilic antioxidants. Some interferences can appear when the absorbance is measured at low wavelength as in the DPPH method [111], and the different molecules presented in the extracts can also influence the results [112].
In general, positive correlations were obtained between the antioxidant activity and the different analyzed molecules. Biliproteins showed a positive correlation with the antioxidant activity [113,114], and two of our tested cyanobacteria presented strong association between them (Scytonema sp. and N. commune, Figure 5). In addition, the proportion of biliproteins related to soluble proteins is very high in the last cited species, 37.0% and 53.8%, respectively, thus providing high content of biliproteins for cosmeceutical applications. In the case of cyanobacteria, scytonemin also showed positive correlations with antioxidant activity, as observed by Matsui et al., who showed the radical-scavenging activity by this molecule [115]. Polyphenols also presented positive correlations with the antioxidant activity in both groups of organisms tested, as has been observed by several studies [116,117]. Finally, MAAs only showed correlation with the antioxidant activity in red algae, as has been previously reported [101,102,118].
The PCA showed that the PC1 axis explained 79.1% and it can be explained by the polarity of the molecules analyzed, with more positive relation with polar extractions (water), whereas the PC2 represents 17.7% of the obtained data and it mostly related to the total nitrogen content. PC1 is mainly related to C content, followed by soluble proteins. The different groups obtained by the PC1 axis seemed to be separated by their content water-soluble compounds (mainly carbonated compounds), the first group, presented in quadrant 3 (negative part of the PC1 axis), being the species with the lowest content and the third group, presented in quadrant 1 (positive part of the axis), the species with the highest content. However, in the PC2 axis only Anabaena sp. was clearly separated due to its high N content. Lyngbya sp. also appeared near to this axis but presented a high variability. Scytonemin presented low and similar correlations with both axes; the fact that red macroalgae did not present this molecule could interfere in the results. Lipids also presented similar correlations with both axes, probably due to their carbonated and lypophylic character. On the other hand, some analyzed molecules did not follow the pattern described before: MAAs should be correlated with the PC2 axis, although these molecules showed a low correlation with both main axes. As for scytonemin, the fact that not all species presented MAAs could interfere in the results. Phenols content only showed correlation with one of the axes (PC1) when it was extracted in water; the other extraction also showed low correlation with both axes. Finally, the species of red algae more taxonomically related are also grouped in the PCA analysis, e.g., the Florideophyceae O. pinnatifida, C. rubrum (Ceramiales order) and G. longissima (Gracilariales order) formed a group, and the other group is formed by P. umbilicalis (Bangiales order) and G. corneum, a Florideophyceae algae of Gelidiales order but closer in the phylogenia to Bangiales than the other studied Florideophyceae algae [119].
The photoprotective capacity was measured in those species that showed a high content of interesting molecules and antioxidant activities (P. umbilicalis and Scytonema sp.). Hydroethanolic (1:1) solvent was selected for this analysis because this solvent might facilitate the extraction of combined molecules, water and organic soluble.
In Figure 4, the peak observed in the cream with P. umbilicalis, in the UV-AII region (320-340 nm), can be related to the MAAs content, increasing more the photoprotection against the effects mostly related to UVA radiation, e.g., using the extract of P. umbilicalis, the ESAR PPD increased 500% and the BEPF PPD 38%, with respect to the base cream. In the case of the cream with Scytonema sp., the high absorptance observed in the UVB region could be due to its MAAs content, which absorbed in the UVB region (mycosporine-glutaminol), instead of porphyra-334 (presence in P. umbilicalis), which absorbed in the UVA region.
The increase of the protection factors (SPF, UVAPF and BEPFs) in comparison with the increase of the ESAR shows a hyperbolic response, as described Schneider et al., i.e., at the beginning of the curve, an increase in the ESAR is expressed in a very small increment in the protection factors values, i.e., creams with a SPF of 2 can absorb 50% of the UV radiation [35]. However, when the capacity to absorb the damage radiation reached 80%, the increase in the SPF was much higher, i.e., a SPF of 5 means that it absorbs 80% of the UVR, a SPF of 15 absorbs around 90% and a SPF of 30 absorbs around 95%. Results obtained in the present work are at the beginning of the curve, reaching 50% of the ESAR.
Many authors have studied the potential of MAAs as UV-filtering compounds [35,93,98,103], but few articles have reported the inclusion of these molecules as UV-absorbing compounds in lotions. De la Coba et al. studied different formulations with MAAs, adding around 5% of purified MAAs and obtained values of SPF that varied from 4.5 to 8.3 [34]. In our study, the % of MAAs included in the creams was much lower, varying from 0.003 to 0.026% due to the dilutions in the cream, explaining the reasons why we found lower values of SPF (around 2). Lawrence et al. found a high UV photoprotection in human keratinocites using a 0.3% palythine [103]. Other authors studied the protection of galenic formulations with MAAs in mouse skin, obtaining positive results [120,121]. MAAs are promising substances in the cosmetic field due to their high stability against UV radiation, pH or temperature [34,78,122], their antioxidant capacity [34,102] and other biological activities such as anti-inflammatory and inhibition of enzymes related to photoaging as collagenase [34,123,124]. In the market, there are some examples of natural antioxidants and photoprotectors based on red algae extracts, e.g., Helionori ® from Gelyma or RonaCare ® RenouMer from Merck, which contains MAAs and presents cosmetic properties. Another example of a natural product is Heliocare ® , based on extracts of the fern Polyplodium leucotomos, enriched in polyphenols with antioxidant and photoprotective properties and known as Fernblock ® [125,126]. Thus, those species with highest level of polyphenols (12-22 mg g −1 DW), e.g., the cyanobacyteria N.commune and Scytonema sp and the red macroalgae P.umbilicalis and O.pinnatifida, could be used as sources of polyphenols for cosmetic products.
Finally, the inclusion of scytonemin in lotions to act as sunscreen has not been reported yet. However, the possibility to use it as photoprotector in cosmetic products have been discussed [127]. In our study, the scytonemin peak was not observed in the creams, probably due to the selected solvent that extracted less pigments in comparison with a less polar solvent such as ethanol or methanol. Scytonemin should be extracted in a solvent compatible with its application in cosmetic products (e.g., ethanol or combinations of water:ethanol) for a better application of this compound as photoprotector. This pigment has been less studied than MAAs, although some properties have been described. It is a stable molecule [78] with positive biological activities (antioxidant or anti-inflammatory) for the skin [115,128], which could be a good option as biological photoprotectors.

Future Perspectives
The potential application of these organisms as sources of commercial bioactive compounds requires high biomass availability and high biocompounds productivity. Red algae (specially Porphyra genus and Gracilariales or Gigartinales orders) have been cultured for centuries, mainly in Asian countries, for human consumption or for their agar or carrageenan content [62]. These species are normally produced in open sea water using ropes, although they can also be produced under integrated multitrophic aquaculture (IMTA). In this way, algae can use the nitrogen and phosphate coming from the fish pond effluents to produce high valuable compounds [129,130]. As mentioned before, N compounds such as phycobiliproteins or MAAs can be increased under high nutrients availability. On the other hand, cyanobacteria can be isolated and cultured easily in liquid medium, with the possibility to scale up using raceways or thin-layer cascade systems [131]. Bioremediation is also an option for these species. Thereby, algal biomass with high bioactivity could be produced at lower cost and the environmental impact could also be reduced. Apart from the need of a big-scale culture, a high yield extraction, concentration and/or purification (using green techniques) of both interested molecules (MAAs and scytonemin) is a challenge for the natural cosmetic industry [132][133][134], increasing the perspective of utilization of these species for sunscreen formulations.

Conclusions
This study shows the usefulness of extracts of different species of cyanobacteria and red macroalgae as cosmeceuticals due to both their antioxidant and photoprotection capacity. Aqueous and hydroethanolic extracts of different species presented antioxidant activity that can be attributed to different molecules, mainly polyphenols, MAAs and pigments (biliproteins and scytonemin). Hydroethanolic extracts of P. umbilicalis and Scytonema sp. that contain MAAs could increase the photoprotection of sunscreens. The concentration or purification of the interested molecules is needed for better photoprotection. Porphyra spp. are good candidates for cosmetic applications due to their high content of MAAs and cultivation possibilities. In addition, Scytonema sp. can also contribute in the photoprotection of UV-B and UV-A mediated responses due to its MAAs (absorption in the UV-B region) and scytonemin (absorption in the UVA-I region) content. Thus, red algal and cyanobacteria extracts enriched in MAAs and scytonemin can be useful to design broad-band photoprotectors in both UV-A and UV-B regions of the spectra.

Biological Material
Ten species of both cyanobacteria and red macroalgae were employed in this study. Cyanobacterium clonal strains were provided by the Culture Collection at the Spanish Bank of Algae (Taliarte, Gran Canaria, Canary Islands) ( Table 5). Cultures were scaled up to 1 L flasks, under controlled laboratory conditions (irradiance: 100 µmol photons m −2 s −1 ; photoperiod: 16:8 (L:D); temperature: 23 ± 2 • C) and continuous aeration supplied with CO 2 pulse addition at a rate of 1 min every hour. Culture medium was BG11. Biomass samples were harvested and freeze-dried (6.5 L Labconco, Kansas, MO, USA) before carrying out the extractions. Red macroalgae were collected in coastal areas of Cadiz and Malaga in the Southern Iberian Peninsula (Table 5), transported to the laboratory in a portable fridge at 4 • C, washed and cleaned for removal of sediments and epiphytes, frozen at −40 • C and freeze-dried (Cryodos, Telstar, Barcelona, Spain). Criteria for the selection of these species were (1) biomass availability, both by collection in nature or culture under controlled conditions, and (2) the potential content of bioactive compounds with photoprotection capacity.  Absorbances or optical densities in the UV region (290 to 400 nm) of the different extracts were measured using a UV-visible spectrophotometer (UV Mini-1240, Shimadzu, Columbia, MD, USA). Extracts in the different solvents were used to compare the content of phenolic compounds and the antioxidant activity. Aqueous extracts were also used for the quantification of soluble proteins, phycobiliproteins and mycosporine-like amino acids (MAAs).

Total Carbon and Nitrogen
Total internal carbon and nitrogen contents were determined from dry biomass of red algae and cyanobacteria using a CNHS LECO-932 elemental analyzer (St. Joseph, MI, USA) in the Research Support Central Services (SCAI, University of Málaga, Málaga, Spain). Results were expressed as mg g −1 DW.

Proteins
Total proteins were calculated by using the nitrogen to protein conversion factor, 4.92 for macroalgae [36] and 4.78 for cyanobacteria [38]. Soluble proteins were quantified according to the Bradford method [135]. Twenty µL of the aqueous extracts (item 2.2) were mixed with 780 µL of phosphate buffer (0.1 M, pH 6.5) and 200 µL of Biorad solution (Bio-rad, Feldkirchen, Germany). After 15 min of incubation at room temperature and darkness, absorbances were measured at 595 nm. Bovine serum albumin was used as standard. In both cases, results were expressed as mg g −1 DW.

Carbohydrates
Carbohydrates were quantified according to the phenol-sulfuric acid method [136]. Five mg of DW were homogenized in 5 mL of 1 M H 2 SO 4 and incubated for 1 h at 100 • C; after cool down, extracts were centrifuged (2000× g; 10 min). For the reaction, 1 mL of the supernatant was mixed with 1 mL of phenol 5%, incubated for 40 min at room temperature, and mixed with 5 mL of concentrated H 2 SO 4 . Absorbances were measured at 485 nm. Glucose was used as a standard and results were expressed as mg g −1 DW.

Lipids
Total lipids were determined spectrophotometrically using the sulpho-phospho-vanillin (SPV) method [137]. Twenty mg of DW were homogenized in 1 mL of methanol. After centrifugation (2000× g, 10 min), 100 µL of the supernatant were mixed with 1 mL of H 2 SO 4 and incubated for 10 min at 100 • C. After cool down, 2.5 mL of phospho-vainillin solution (0.6 g of vainillin, 100 mL of ethanol and 400 mL of H 3 PO 4 85%) were added to the mix, which was incubated for 15 min at environmental temperature and measured spectrophotometrically at 530 nm. Trioleine was used as standard and results were expressed as mg g −1 DW.

Phycobiliproteins and Scytonemin
Aqueous extracts (item 2.2) were used for phycobiliproteins determination. In red algae, the quantification was made according to the formula described by Beer and Eshel, and in the case of cyanobacteria, the one proposed by Bennet and Bogorad [138,139].
For the extraction of scytonemin in cyanobacteria, 20 mg of DW was extracted in 1 mL of acetone. The quantification was made according to the formula described by García-Pichel and Castenholz, using the extinction coefficients reported by García-Pichel et al. [74,140].
Results of both compounds were expressed as mg g −1 DW.

Phenolic Compounds
Phenolic compounds were determined from the different extracts using the Folin-Ciocalteu method [141] with some modifications. Briefly, 100 µL of the different extracts (item 2.2) were mixed with 700 µL of distilled water and 50 µL of Folin-Ciocalteu phenol reagent (Sigma-Aldrich, St. Louis, MO, USA). After vortexing, 150 µL of Na 2 CO 3 20% were added and samples were mixed again. The mixtures were incubated for 2 h at 4 • C in darkness. Absorbances were measured at 760 nm. Phloroglucinol was used as standard and results were expressed as mg g −1 DW.

Mycosporine-Like Amino Acids (MAAs)
MAAs were determined using HPLC (high performance liquid chromatography) according to Korbee-Peinado et al., with some modifications of Chaves-Peña et al. [96,142]. Seven hundred microliters of the aqueous extracts (item 2.2) were dried under vacuum. The dry extract was re-suspended in 700 µL of methanol, filtered through a 0.2 µm membrane and analyzed using HPLC (Waters 600 HPLC, Waters Cromatografía, Barcelona, Spain). MAAs were detected using a Luna C8 column (Phenomenex, Aschaffenburg, Germany), applying an isocratic flow of 0.5 mL min −1 and a mobile phase of 1.5% methanol and 0.15% acetic acid in ultrapure water. The detection was made using a photodiode array (PDA) detector at 310 and 330 nm. Secondary standards were used for the identification of MAAs, and the quantification was performed using the molar extinction coefficients (ε) of the different MAAs [143]. Results were expressed as mg g −1 DW.
The MAAs were also identified via positive electrospray ionization mass spectrometry (ESI-MS) (Orbitrap Q-Exactive, Thermo Scientific S.L., Bremen, Germany) in the Research Support Central Services (SCAI, University of Malaga, Spain).

Antioxidant Activity (ABTS and DPPH)
The antioxidant activity of the different extracts was evaluated through two different methods based on the free radical scavenging activity.
The ABTS assay was performed according to Re et al. [144] with some modifications. The ABTS radical cations (ABTS +• ) were generated via a reaction of 7 mM ABTS (2,2 -azino-bis(3ethylbenzothiazoline-6-sulfonic acid) and 2.45 mM K 2 S 2 O 8 in phosphate buffer (0.1M, pH:6.5). This reaction was stored for 12-16 h at room temperature to ensure the complete formation of the radical. ABTS +• solution was diluted with phosphate buffer until the absorbance at 727 nm was around 0.75 ± 0.05. For the reaction, 50 µL of the different extracts (item 2.2) were mixed with 950 µL of the diluted ABTS +• . The mixture was incubated for 8 min at room temperature and darkness, and absorbances were measured at 727 nm.
The DPPH assay was made according to Brand-Williams et al. [145] with some modifications. For the reaction, 200 µL of the different extracts (item 2.2) were mixed with 1 mL of the DPPH • (2,2-diphenyl-1-picrylhydrazyl) solution (0.06 mM of DPPH in methanol 80%). After 30 min of incubation at room temperature and darkness, absorbances were measured at 517 nm.

Photoprotection Capacity
The photoprotection capacities were measured in vitro in those species in which the highest content of interest molecules was observed. The selected species were one cyanobacterium (Scytonema sp.) and one red macroalga (P. umbilicalis). Hydroethanolic (1:1) extracts were used for this analysis. Extracts (item 2.2) were concentrated under vacuum in a rotary evaporator (Büchi R-210, Switzerland) from 50 to 200 mg mL −1 . Extracts were incorporated in an oil-water cream (25% w/w). One hundred milligrams of the creams were spread in a polymethylmethacrylate (PMMA) plate (Schönberg, Hamburg, Germany), with a surface of 50 mm × 50 mm and a roughness (Ra) in the range of 4.5-5.2 µm. Creams were distributed on the plates with gloved fingertips for 30 s (approx.); after that, plates were allowed to dry for 15 min in darkness. Transmission through plates was measured using a spectrophotometer (UV-2700i, Shimadzu, Duisburg, Germany) with an integrating sphere. The sun protection factor (SPF), UVA protection factor (UVAPF) and biological effective protection factors (BEPFs) considering other action spectra (elastosis and photoaging) were calculated using the following equation In both formulae, Act.Sp (λ) = action spectra ( Figure 6); E (λ) = spectral irradiance of a sunny midday in a summer day (June) in Malaga (W m −2 ); T (λ) = transmittance values at each wavelength (0-1); A (λ) = absorptance values at each wavelength (0-1).
The different relative action spectra used in this study are shown in Figure 6. A commercial cream with SPF 10 that contains different chemical filters (butyl methoxydibenzoylmethane, ethylhexyl methoxycinnamate, methylene bis-benzotriazolyltetramethylbutylphenol and octocrylene) and titanium dioxide as a physical filter was used as positive control. In this case, 32 mg were spread on the plate according to COLIPA (2011) and Pissavini et al. (2018) [29,30].
The different relative action spectra used in this study are shown in Figure 6. A commercial cream with SPF 10 that contains different chemical filters (butyl methoxydibenzoylmethane, ethylhexyl methoxycinnamate, methylene bis-benzotriazolyltetramethylbutylphenol and octocrylene) and titanium dioxide as a physical filter was used as positive control. In this case, 32 mg were spread on the plate according to COLIPA (2011) and Pissavini et al. (2018) [29,30].

Statistical Analyses
The software STATISTICA (V.7) was used for the statistical analysis. Analyses of variance (ANOVA) were done to compare data obtained. Two-way ANOVA were performed to detect significant differences in the antioxidant activity values, comparing solvents and methods. One-way ANOVA were used to evaluate species effects in all the biochemical compounds and antioxidant activity among species (cyanobacteria and red macroalgae were compared separately), the extraction Figure 6. Normalized action spectra responses of different harmful effects related to skin damages driven by UV radiation and used in this study for the determination of the photoprotection capacity.

Statistical Analyses
The software STATISTICA (V.7) was used for the statistical analysis. Analyses of variance (ANOVA) were done to compare data obtained. Two-way ANOVA were performed to detect significant differences in the antioxidant activity values, comparing solvents and methods. One-way ANOVA were used to evaluate species effects in all the biochemical compounds and antioxidant activity among species (cyanobacteria and red macroalgae were compared separately), the extraction of phenolic compound in each solvent and the photoprotection capacity. A Student-Newman-Keuls (SNK) post-hoc test was performed to compare the data after significant differences in the ANOVA. Homogeneities of variances were tested using the Cochran test and via visual inspection of the residuals. Correlations among dependent variable data were calculated using Pearson's correlation analysis.
The software PRIMER 6 v.6.1.13 & PERMANOVA+ v.1.0.3 was used for the principal component analysis (PCA), performed to detect patterns among the biochemical composition and antioxidant activities of the different species. Data were correlated with each PC-axis using Pearson's correlation coefficient.
Supplementary Materials: The following are available online at http://www.mdpi.com/1660-3397/18/12/659/s1, Table S1: ANOVA summary Tables obtained after the comparison of (A) biochemical composition of the different species (cyanobacteria and Rhodophyta separately), (B) phenolic compounds content extracted in each solvent, (C) antioxidant activity in the different solvents and using the two methods and (D) photoprotective capacity among species. DF: degrees of freedom. Figure Table S2: Pearson coefficient (r) between the different variables analyzed in cyanobacteria. Table S3: Pearson coefficient (r) between the different variables analyzed in red macroalgae. Table S4: Pearson coefficient (r) between the PC axis and the analyzed variables in both group of organism tested.