PhD Thesis Defence Presentations - Vasilis Andriopoulos

Microalgae have been extensively studied by biologists for at least 80 years, while during the last decades they attracted the interest of biotechnologists due to their high growth rate and attractive biochemical composition, which, depending on the species and cultivation conditions, can be suitable for the production of fuel, food or feed, or medicinal and cosmetic products. While initially biofuel was the main focus of microalgal research, recently interest in that sector has been diminished, with food, feed and pharmaceutical or cosmetic applications taking its place. One category of molecules that was only reported during the last 20 years in microalgae is phenolic compounds. Measured both with quantitative and qualitative techniques, this class of molecules has the potential to make microalgal biorefinery more feasible, introducing one an additional high-value product next to pigments and omega-3 fatty acids. The goal of this thesis was to study the production of phenolics and other antioxidants from selected marine microalgae, while at the same time maintaining a high protein and omega-3 content, suitable for food or feed purposes.

The selected marine species were Chlorella minutissima, Dunaliella salina, Nannochloropsis oculata, Isochrysis galbana and Tisochrysis lutea, which had an attractive phenolic content as well as nutritious biomass composition according to the literature. The first step was to screen the proximate biomass composition and phenolic content of those species. To that end, cultivation under saturating light intensity and controlled temperature (25°C) was performed, while biomass was collected at two stages, the early and late stationary phases. C. minutissima and N. oculata displayed the highest growth rates, while all microalgae were rich in protein. Only N. oculata contained C20:5_n3 (EPA) in significant amount, while only I. galbana, and T. lutea contained C22:6_n3 (DHA). Biomass of all species was extracted sequentially with water and methanol to assess its antioxidant content. While the methanolic extract of C. minutissima was the most potent in terms of the Folin-Ciocalteau assay, a quantitative method for the measurement of phenolics in plant samples, chromatographic analysis showed a complete absence of phenolics from all samples, with the exception of minor peaks in T. lutea extracts. The conclusion was that pigments such as chlorophyll might be significant interfering compounds in the FC assay, while the biomass composition of N. oculata and T. lutea held promise for the production of nutritious biomass and extracts of pigments and omega-3 FAs.

To further explore the potential of antioxidant production from N. oculata and T. lutea different stressors were studied, specifically salinity, light intensity and the lack of aeration. A fractional factorial design with two levels was used, applying salinities of 38 and 60ppt, and aeration of 2.4 or 0 vvm, while the levels of light intensity were different for the two species due to their different tolerances. Significant effects of lack of aeration were uncovered for N. oculata, with an increase of protein content and decrease of saturated FA content, while the apparent phenolic content according to the FC assay also increased without a correlation to pigments. On the contrary, T. lutea did not display positive effects under stress conditions, while its apparent phenolic content varied with pigment content. Conclusively, the lack of aeration might be a useful tool in the optimization of N. oculata cultivation, while further research is required on the compounds contributing to its apparent phenolic content.

The next step was to optimize critical cultivation parameters, specifically temperature, pH, light period and illuminated surface to volume ratio, for the production of nutritious biomass and antioxidant extract from N. oculata. At the same time an extraction protocol was developed in collaboration with the Laboratory of Pharmacognosy and Chemistry of Natural Products (LPCNP), Department of Pharmacy, University of Patras. Like before, a fractional factorial approach was adopted, while two sets of experiments were performed, one in photobioreactors with online pH control via acid or base addition, and one in shake-flasks with daily adjustment of pH to a setpoint. Regression analysis to the combined data from the two experiments showed effects of temperature and light intensity that agree with the literature, for example the negative effects on these two parameters on EPA fraction in total lipids, while a more appealing biomass composition was attained in photobioreactors, which had a lower illuminated surface to volume ratio, at the expense of growth rate, which was higher in shake-flasks. Among the extraction methods that were assessed from LPCNP, ultrasound-assisted ethanolic extraction was chosen, due to the high protein content of the residual biomass, and the high pigment and EPA content of the extracts. The powdery texture and the practically zero moisture content of the residual biomass after freeze-drying were also very appealing for industrial use like incorporation to food products. Using the selected extraction method on the biomass from the optimization experiments showed an interesting effect of temperature, with a higher pigment content in extracts of biomass cultivated under high temperature despite the lower pigment biomass content. That could relate to a weakened cell wall or to a modified packing of pigments in the chloroplast and requires further examination. Co-optimizing the cultivation conditions for high protein and EPA biomass content as well as for high pigment extract content, showed that cultivation under 20-27.5°C, 16-18 hours of light and pH ~ 7.7 in photobioreactors would yield biomass with protein content of 40-60 % AFDW and EPA content of 3.2-4.5 % AFDW, while producing extract with at least 386 mg Chla and 86 mg carotenoids L^-1.

Further investigation of the effects of cultivation conditions on the fatty acid content and profile on N. oculata was performed, using novel data from various cultivation systems and the available literature. A new regression method, the high dimensional selection with interactions algorithm, was equipped to assess a very large number of predictor parameters and their interactions, something that would not be possible with least squares regression. The findings confirmed well known effects such as those of temperature, while less studied effects such as that of potassium on palmitic acid were also confirmed, despite the absence of potassium-related studies in the training data. The most significant novel effect was that of aeration on EPA fraction in total lipids. Absence of aeration can modify the fatty acid content, decreasing saturated FAs while maintaining rather constant EPA content in biomass. More research is required to untangle the effects of aeration from that of mixing and clarify whether increase of oxygen under non-aerated conditions enhances fatty acid desaturation. Validation with external data showed a good performance of the models for total lipid content and EPA fraction in total lipids under stable conditions, while under outdoor conditions the EPA content was underestimated, with the overall trend being correctly predicted however. Adding more data to the training dataset should be a target for future improvement. Optimization showed a great promise of N. oculata for EPA production, with up to ~ 11 % AFDW EPA in biomass and 0.12 g L^-1 EPA achieved under optimal conditions, while the lack of aeration might be equipped to decrease the saturated FAs and achieve a more appealing dietary profile of the biomass. 

As a final step, the phenolic content of microalgal was revisited. Evaluating the literature on the phenolic content of microalgae according to both the Folin-Ciocalteau method and chromatographic analysis showed several issues with the reported findings and the methodology used. On the one hand, the interference of pigments, free-amino acids and nucleotides on the FC assay has been almost universally neglected, while antioxidant mechanisms of the cells such as the glutathione system have been studied to some extent but not in their effect to the FC assay. On the other hand, chromatographic evidence for phenolics in microalgae might be compromised by contamination from external sources via adsorption to the cell wall or absorption and biotransformation from the cells. The lack of two critical enzymes that initiate the phenylpropanoid pathway (PP) from most microalgae also makes the presence of cinnamic acids, anthocyanins and other products of PP enigmatic. Exclusion of phenolics from all growth media and accurately reported methods should be used in future research, while the removal of phenolics from waste streams by microalgae is very promising.

In conclusion, N. oculata has a great promise for the development of a biorefinery based on an ethanolic extract rich in pigments, EPA and possibly other antioxidants and a residual biomass with an excellent protein content and residual EPA, while the lack of aeration can be further explored as a finishing step for the improvement of the fatty acid profile and antioxidant content. Meanwhile, the phenolic content of microalgae might not be of great practical interest, considering the abundance of plant waste streams rich in phenolics, while the removal of phenolics by microalgae is promising, and the boosting effects of phenolics in microalgal metabolism is also an interesting research subject.