Methanol and dichloromethane were purchased from ProQuimicos (Brazil); BHT and TBHQ were purchased from Vogler Ingredients (Brazil); Activated carbon, Potassium sorbate, potassium benzoate, guar gum, potassium iodide, starch and EDTA were purchased from Vetec Química Fina LTDA (Brazil); Vegetable oils used were refined oils without antioxidants, produced by Cargill ^® Company (Brazil), pasteurized whole egg was produced by Fleischmann^® Company (Brazil).

For the development of this work, we used freeze-dried biomass of cyanobacteria Arthrospira platensis grown in our previous study (unpublished data) about the influence of culture conditions on the biomass production, at different conditions of sodium bicarbonate (NaHCO₃), sodium nitrate (NaNO₃) and light intense, which totalize 17 cultivation conditions and plus the standard cultivation condition (S) ( Table ).

Table
Experimental design of growing conditions of Arthrospira platensis.

Eighteen methanolic extracts were prepared by sonication and the solvent was removed from the extract with a rotating evaporator at 35°C. Subsequently, the extracts were purified by open column chromatography using activated carbon as stationary phase. The mobile phase consisted of a gradient of dichloromethane and methanol, in proportions of 9:1 to 1:9 (v:v), which were eluted by 20 ml of each solution. The solvents were subsequently removed with a rotating evaporator.

The antioxidative effect of the extracts on preventing lipid oxidation as compared to the synthetic antioxidant BHT and TBHQ was evaluated using soybean oil added standard antioxidant (100 ppm) or purified extract (1,000 ppm)¹⁴ . Oil without addition of antioxidant or extracts was used as a control. Oil samples were subjected to oxidation induced thermally in Schaal oven test for seven days⁶ and peroxide value (PV) was measured ²¹ .

Based on the results obtained in the first step, an experiment was conducted to evaluate the antioxidative effect of the biomass to oxidation front of a high fat food. Handmade mayonnaise was developed and used like model.

For formulating the mayonnaise, refined vegetable oil without antioxidants (70%) of two types, soybean and sunflower oil, was used. The pasteurized egg was used as a source of lecithin and water. Vinegar (acetic acid) was added as acidulate in the emulsion formulation while maintaining pH low. Provided food additives were added in the mayonnaise, under Brazilian law²² .

The additives were added based on 100 g of product, except TBHQ that was added based on 100 g of oil. The Arthrospira platensis biomass was also added based on 100 g of product, in concentrations of 0.5% and 0.75%. For each type of vegetable oil, four samples of mayonnaise were formulated, namely: control (without added antioxidant or biomass), synthetic antioxidant (TBHQ), Arthrospira platensis biomass 0.5% and Arthrospira platensis biomass 0.75%.

For this assay, 45 g of each sample was stored in plastic tubes Falcon type 50 ml in storage conditions that favored the lipid degradation in accelerated form, under light incidence 25 µmol photon.m^{- 2} .s^{- 1} and temperature of 30ºC ± 1°C, for 14 days²³ . The protect capability of biomass was evaluated in times 0, 7 and 14 days of storage by PV determination by the method of Alemán et al²⁴ with small modifications, so for this analysis, the samples of each time were lyophilized to separate the lipid fraction and thereafter the PV was determined.

The evaluation of the effect of variables (concentration of sodium bicarbonate, sodium nitrate and irradiance) on the protective ability of the extracts derived from the biomass was performed with the aid of Statistica version 8.0. The effect of the variables was considered statistically significant when p < 0.1. The peroxide value results obtained in the evaluation tests of the antioxidative effect of the biomass extracts were statistically analyzed to assess significant differences among samples. Analysis of variance (ANOVA one way) was used. In case of significant difference, the means were compared by Tukey test. Differences were considered statistically significant at p < 0.05. Data were analyzed using the GraphPad Prism version 5.0.

The test to measure the protective ability of the extracts on preventing lipid oxidation was conducted with soybean oil added of the synthetic antioxidant TBHQ or BHT or 18 purified extracts of biomass produced in different cultivation conditions. The oil samples were subjected to oxidation thermally induced and during this process the formation of peroxides occurred, which is one of the intermediates of the mechanism of reaction of lipid deterioration. Thus, after the incubation period, the PV of the samples was measured, to evaluate the ability of the extracts in protecting the oil on preventing oxidation. Figure 1 shows these results. The control is the soybean oil without antioxidant. The non-peroxide formation by the addition of antioxidants TBHQ or BHT or extracts, indicates the ability of these additives to increase the oxidation stability of the oil. The results are expressed in “peroxide value percent of control”.

Figure 1
Peroxide value of soybean oil without antioxidant (control) and added of 100 ppm synthetic antioxidant (TBHQ or BHT) or 1,000 ppm Arthrospira platensis extract (E1: extract one; E2: extract two; E3: extract three; E4: extract four; E5: extract five; E6: extract six; E7: extract seven; E8: extract eight; E9: extract nine; E10: extract ten; E11: extract eleven; E12: extract twelve; E13: extract thirteen; E14: extract fourteen; E15: extract fifteen; E16: extract sixteen; E17: extract seventeen; ES: standard extract). Equal letters with no significant difference (p < 0.05).

It was possible to observe in Figure 1 that the PV of soybean oil was lower by adding 100 ppm synthetic antioxidant TBHQ and BHT and 1,000 ppm Arthrospira platensis extracts E7; E10; E11; E12; E14; E15; E16; E17 and ES, than the control and others treatments. This means that these extracts increased the oxidative stability, being capable to minimize the lipid oxidation that results in peroxide formation. The others extracts were not capable to increase the stability since the PV of oil added of these extracts was significantly equal to the control. Through statistical analysis by analysis of variance (ANOVA) followed by Tukey test, we could verify that the antioxidant TBHQ and BHT gave significantly different protection than the protection promoted by the extracts, except E16 that did not differ statistically by BHT. The antioxidant TBHQ showed the most protective capacity. The oil with added 1,000 ppm of extracts E14 and ES obtained results that did not differ statistically and had the highest antioxidant effect among the tested extracts, followed by E10 and E15 and by BHT and E16 that also did not differ statistically from each other. In the same way, the oil added to extracts E7, E11, E12 and E17 did not differ statistically, but had the lowest antioxidant effect. Therefore, the treatments E14 and ES, in the same period to evaluate the oxidation of soybean oil, showed PV below the other extracts, conferring 76.41% and 79.48% of soybean oil protection on preventing the formation of hydroperoxides, respectively.

The effect of the three variables of cultivation of Arthrospira platensis on the protective ability of the extracts was measured. From these data, analysis of variance was made and the statistically significant factors could be observed, in addition to observing which were the more relevant to antioxidative effect, expressed by the protective capacity measured by the PV. Thus, Figure 2 shows that the variables light and concentration of bicarbonate (NaHCO₃) showed a significant effect (p = 0.085517 and p = 0.096482, respectively), whereas the variable concentration of nitrate (NaNO₃) showed marginally significant effect (p = 0.100306). Thereby, the linear term of the light is the most important variable for antioxidative effect, followed by the quadratic terms of NaHCO₃ and NaNO₃.

Figure 2
Pareto chart for the effect of the variables light, sodium bicarbonate (NaHCO3) and sodium nitrate (NaNO3) on the peroxide value.

Through the modification of culture conditions differences in antioxidative effect of biomass were observed. The conditions that produced biomass extract with more antioxidative effect on preventing oxidation, for the most part, were those with higher light levels (≥ 100 µmol photon.m^{- 2} .s^{- 1} ), regardless of the concentration of NaNO₃ and major NaHCO₃ concentrations (≥ 13.50 g.L^{- 1} ). The biomass growth condition 14 (150 µmol photon.m^{- 2} .s^{- 1} , 1.875 g.L^{- 1} NaNO₃ and 13.5 g.L^{- 1} NaHCO3) showed the best antioxidative effect and the biomass from S condition was significantly equal. The standard condition was produced with low irradiance (50 µmol photon.m^{- 2} .s^{- 1} ), but with maximum concentration of NaHCO₃ (18 g.L^{- 1} ) and NaNO₃ (2.5 g.L^{- 1} ). In this case, the high protection capacity of biomass could be related to the effect of bicarbonate and nitrate concentrations, in accordance with Pareto chart ( Figure 2 ). The larger substrate offer could reflect in the accumulation of substances that influenced the protective capability and that showed a different biosynthetic route from that initiated by oxidative stress induced for high photosynthetic rate, in high light level.

Some studies were conducted to evaluate the best conditions of cyanobacteria cultivation related with its antioxidant potential. Shalaby et al.²⁵ reported that Arthrospira platensis under salt stress conditions, the algal polar extracts displayed the highest antioxidative effect including phycobilins pigments, sulfated polysaccharides and phenolics compounds.

Researchers found that the antioxidative effect of the extracts obtained from cyanobacteria biomass is influenced by the strain of Arthrospira used for cultivation and the composition of the growth medium. In their study, they evaluated antioxidative effect of ethanolic extracts and reported that higher values of antioxidative effect were achieved when the culture medium had reduced concentrations of NaNO₃ and/or NaHCO₃²⁶ . In the present study, the biomass of extract E14 was cultivated with reduced concentrations of NaNO₃ and NaHCO₃.

Other study demonstrated the existence of significant interactions with the content of phenolic compounds of methanolic extracts of Arthrospira platensis and antioxidant potential. The authors demonstrated that higher NaNO₃ concentrations (1.875 and 2.500 g.L^{- 1} ) promoted higher levels of phenolic compounds in Arthrospira platensis grown at 35°C. These NaNO₃ concentrations are equal to culture conditions of biomass obtained for E14 and ES, respectively, which showed a better antioxidative effect on preventing oxidation for E14, than for other Arthrospira platensis extracts²⁷ .

Araújo et al.¹⁴ studied the occurrence of antioxidant compounds in Anabaena PCC 7119 cyanobacteria which was cultivated under different light intensities (100, 200 and 300 µmol photon.m^{- 2} .s^{- 1} ). These researchers found good antioxidative effect in methanolic extracts from the cultures with incident light 100 and 200 µmol photon.m^{- 2} .s^{- 1} , irradiance values similar to those studied in this work, which were tested in soybean oil.

As there was no significant difference between the 14 and S tests that showed the best protective capacity on preventing oxidation in soybean oil, the standard assay (S) was selected for the development of the evaluation test to protective capacity of biomass on preventing the oxidation of mayonnaise.

Figure 3 shows the results of PV on 0, 7 and 14 days for the mayonnaise samples made with soybean and sunflower oil. It can be seen that the initial PV (time 0) of mayonnaise produced with soybean oil is higher than the mayonnaise produced with sunflower oil; this is probably due to the fatty acid composition of these oils. In the study made by Castelo Branco and Torres²⁸ , the results showed that more than 50% of the fatty acid in soybean and sunflower oil are polyunsaturated, but the sunflower oil has a higher content of monounsaturated. Monounsaturated fatty acids are less susceptible to oxidation than polyunsaturated²⁹ , justifying the highest concentration of peroxides in the soybean oil used to prepare the mayonnaise.

Figure 3
Peroxide Value of mayonnaise prepared with (A) soybean oil and (B) sunflower oil.

In the Brazilian law there is not a limit of PV to mayonnaise, but 10 mEqO₂.Kg^{- 1} is the maximum level for PV to vegetable refined oil³⁰ . Based on this law, all samples were classified as suitable for consumption by PV.

Based on statistical analysis, it was observed that in mayonnaise made with soybean oil the control sample (without synthetic antioxidant or biomass of Arthrospira platensis ) at the end of the storage period (time 14) did not differ significantly from initial PV (time 0). This result is probably related to the protective capability of the components present in soybean oil, such as tocopherols. Castelo Branco³¹ investigated the influence of the initial composition of vegetable oils on the oxidative stability and total antioxidant capacity; it was observed that the soybean oil has a high content of γ-tocopherol, being also the oil with a higher content of total tocopherol, which is naturally present in vegetable oils and is a natural effective antioxidant as also described by Chaiyasit et al² .

Even with the mayonnaise prepared with soybean oil, it was observed that TBHQ was able to preserve the photo-degradation of mayonnaise by 14 days of storage, while the biomass of Arthrospira platensis at concentration of 0.5% was able to preserve only for the first 7 days. The statistical analysis shows that the PV measured on the seventh day, to mayonnaise added of 0.5% biomass, had no significant difference of PV measured for mayonnaise added of TBHQ. This demonstrates that the protection offered by the biomass was equal to that conferred by TBHQ, during the first 7 days. This result also demonstrates the ability of this biomass for minimizing lipid oxidation during 7 days of storage, under the test conditions. It is worth emphasizing that in this experiment mayonnaise was kept in conditions that accelerated the degradation rate (25 µmol photon.m^{- 2} .s^{- 1} , 30° C), which does not correspond to those in which the product is usually kept for marketing, differing mainly on the light intensity. Thus, it can be inferred that the addition of biomass in this concentration could be able to protect the product for a longer period of time, enhancing the potential use of biomass, in that concentration, as a substitute for synthetic antioxidant.

In the mayonnaise added of 0.75% biomass, it was observed that the PV increased significantly until the seventh day of storage, followed by a decrease of the same to the fourteenth day of storage, showing results significantly equal to the initial value. This fact occurs because the PV gives a measure about the degree of oxidation produced by peroxide and hydroperoxide that react with potassium iodate³² , but the peroxides are produced during an initial stage of oxidation and in the following stages (propagation and termination) they are complexed to form secondary products that are more stable² , resulting in a PV reduction. Abreu et al.³² explain that the PV tends to increase during the initial stages of oxidation because it is when the hydroperoxide rate formation is higher than the decomposition rate, but when a maximum value is achieved, the PV decreases due to lower accessibility of substrate and to instability of peroxide molecules.

The increase of biomass concentration, 0.5% to 0.75%, did not represent a greater capacity to protect the product of the photo-oxidation, which would signify that this increment, beyond promoting higher concentration of antioxidant substances at mayonnaise, also represents the increase of concentration of substances that could be degraded in the same conditions, such as chlorophyll and carotenoids. This degradation produces instable products, among which reactive oxygen species, that oxidize other molecules, like lipid³³ , which can explain why the PV increases in mayonnaise with 0.75% biomass.

The results of the mayonnaise prepared with sunflower oil show that in the control sample there was lipid degradation throughout the storage period, being more intense after the seventh day. This degradation could be related with the greater vulnerability of sunflower oil, as demonstrated by Castelo Branco³¹ that verified the initial quality of vegetable oil and found that sunflower oil shows higher PV than that showed by the soybean oil (1.14 mEq O₂.Kg^{- 1} and 0.38 mEqO₂.Kg^{- 1} , respectively). In other results in the same study, it was verified that sunflower oil showed greater susceptibility to oxidation and lower oxidative stability.

TBHQ ensures protection to mayonnaise prepared with sunflower oil until seven days of storage, showing degradation after this time, but this degradation was significantly lower compared to control, which showed on the fourteenth day the highest degradation among all tested samples of sunflower oil. For samples with 0.5% and 0.75% biomass, it was observed that the PV in the seventh day has a significant increase, followed by reduction at the fourteenth day of storage. The biomass, independent of the added concentration, was not able to protect from the photo-degradation the mayonnaise produced with sunflower oil.

Some studies were performed to investigate the oxidative stability of emulsions, such as the study of Pavlović et al.³⁴ that evaluates the effect of ascorbic acid and EDTA in emulsified dressings salad made with different types of sunflower oil (with different tocopherol profiles) and stored for three months at 25°C protected from light. The results showed that in the sample control (without antioxidant) the PV increased rapidly during the first month of storage; then the increase was slower, probably due to formation of secondary oxidation products. Another study about oxidative stability of emulsion was performed by Altunkaya et al³⁵ , and evaluated mayonnaise added with grape seed extract, which is rich in phenolic compounds. They investigated the oxidative alterations in mayonnaise by various methods, including PV, during storage in conditions that protect from light at 20° C for 8 weeks. The mayonnaise added with grape seed extract showed PV significantly lower compared with the control mayonnaise, after 8 weeks of storage. However, the progress in oxidation follows a pattern that is normally found in the auto-oxidation, such as initial increase in lipid peroxidation products followed by a subsequent increase in secondary products of lipid oxidation.

Castelo Branco and Torres³⁶ described that the tocopherol profile of refined vegetable oils is one of the determinants of total antioxidant capacity, being the concentration of γ- and δ-tocopherol responsible for this feature. When evaluating different samples of refined vegetable oils, the following values were found in soybean oil : 93.3 to 129.4 and 24.0 to 49.7 mg.100g^{- 1} of γ- and δ-tocopherol, respectively, while for sunflower oil the values were 4.50 to 5.97 and 0.23 to 0.57 mg.100g^{- 1} of γ- and δ-tocopherol, respectively. The findings of Castelo Branco and Torres³⁶ study are complementary to that found in this work, pointing to a possible synergistic effect of tocopherols present in soybean oil with substances that promote antioxidative effect of biomass, promoting the protection of mayonnaise produced with soybean oil. These results also contribute to elucidate the behavior observed in mayonnaise produced with sunflower oil. As sunflower oil inherently has the lowest total antioxidant capacity and is therefore more prone to oxidation, the substances conveyed through the biomass, irrespective of the added concentration, were not able to promote oxidation prevention of mayonnaise in this study’s conditions.

* E-mail: beatriz.correa@macae.ufrj.br