Introduction

Microalgae and cyanobacteria are a nearly untapped group of at least 50,000 species (Behrens and Delente 1991). Up to the present day, only a small number of them have been isolated, cultivated, and thus analyzed regarding their metabolic potential (Wijffels 2008). Due to their photosynthetic living, microalgae are exposed to oxidative stress and therefore need to produce antioxidative compounds. Molecular oxygen is essential for aerobic organisms because this molecule acts as an electron acceptor during mitochondria electron transport (Berg et al. 2003; Knook and Planta 1971; Kröger and Klingenberg 1973). In case of an uncompleted reduction of oxygen or due to electron leakage, reactive oxygen species (ROS) such as superoxide anions or peroxides arise. ROS are responsible for physiological dysfunctions which can cause apoptosis due to the lack of antioxidants (Abele 2002).

In the presence of sensibilizators such as chlorophyll, light can cause the formation of hydroperoxides, which can disintegrate homolytically into energy-rich compounds like peroxy or alkoxy radicals. As a result of this and because of an insufficient concentration of antioxidants, an uncontrolled autoxidation of the polyunsaturated fatty acids of the thylakoid membrane can occur. These phenomena are also responsible for apoptosis (Belitz et al. 2001).

Ubiquinones are one of the most important antioxidants known and also act as an electron transporter within the mitochondrial respiratory chain in eukaryotic cells (Berg et al. 2003). Due to the quinoid ring structure, coenzymes Q show high antioxidative capacity as well as electron transport activity. The lipophilic character of those molecules rests upon the isoprenoid side chain, which indicates a species-specific length. The reduced form of coenzyme Q10 (ubiquinole) is capable to pass electrons to radicals in order to eliminate them (Ernster et al. 1992; Lenaz 1985; Lenaz et al. 1999; Lenaz and Genova 2009; Yamamoto and Yamashita 1997). Investigations concerning medical applications of coenzyme Q10 showed positive effects on neurodegenerative diseases such as Chorea Huntington and Parkinson (Beal 2005; Gholipour 2004; Horvath et al. 2003; Smith et al. 2006).

The red microalga Porphyridium purpureum can be found in aquatic as well as in terrestrial habitats, which is the basis for the broad spectra of products of this microalga (Jones et al. 1963). This microalga synthesizes sulfated polysaccharides, phycoerythrin and polyunsaturated fatty acids (PUFAs), such as arachidonic acid and eicosapentanoic acid (Arad et al. 1985, 1988; Cohen 1999; Gantt 1969; Guil-Guerrero et al. 2000). Sulfated metabolites especially, such as polysaccharides and sulfoglycolipids, exhibit antiviral activities and hence are potential candidates for medical and pharmaceutical application (König 2007; Naumann 2009). Contrarily, high concentrations of PUFAs enable use as a neutraceutical (Guil-Guerrero et al. 2000). Furthermore, antioxidative compounds like α-tocopherol and other antioxidants like superoxide dismutase or catalase have been identified in Porphyridium biomass (Chen et al. 2008; Durmaz et al. 2007). In addition, the antioxidative activity of Porphyridium biomass could be proven (Rodriguez-Garcia and Guil-Guerrero 2008).

In order to produce high-value metabolites as coenzyme Q10, monoseptic cultivation of P. purpureum using sterilizable photobioreactors is important to ensure controlled and axenic conditions (Walter et al. 2003).

High concentrations of PUFAs, especially in photosynthetic membranes of microalgae, require antioxidative compounds like ubiquinone, plastoquinone, or α-tocopherol in order to suppress lipid peroxidation. Therefore, a variation of photon flux density (PFD) was investigated concerning its influence on ubiquinone production by P. purpureum.

Materials and methods

Photobioreactor screening modules

Photobioreactor screening modules (PSM) represent in situ sterilizable bubble columns which can be used for microalgae cultivation. Because this kind of photobioreactor can be used in parallel operation, screening as well as optimization investigations can be performed by varying different cultivation parameters such as temperature, PFD, or media components. With an inner diameter of 49 mm, a total of 0.9 L can be used as the cultivation volume. The cylindrical reactor is surrounded by four fluorescence lamps (Lumilux Coolwhite, L18W/840, Osram) ensuring high light yield (surface/volume ratio, 80 m−1; see Fig. 1) (König 2007).

Fig. 1
figure 1

Photobioreactor screening modules−left schematic, right picture of a cultivation of P. purpureum, modified (Naumann 2009)

Full size image

Medusa photobioreactor

Scale-up cultivation was performed in 120-L scale, using a Medusa type photobioreactor (König 2007; Walter et al. 2003). The 120-L Medusa type photobioreactor is built of 16 glass tubes with eight pairs of two each connected by a U-tube. The glass tubes are connected by a head space, which is made of stainless steel (VA 1.4301) at the top of the reactor. Light is supplied by concentrically arranged fluorescent tubes, which results in an illuminated surface/volume ratio of 65 m−1 (König 2007; Walter et al. 2003).

Cultivation of P. purpureum in photobioreactor screening modules

Investigations regarding coenzyme Q10 biosynthesis in phototrophic microorganisms were carried out with the unicellular red alga P. purpureum (former Porphyridium cruentum) 1380 1-f (Schlösser 1994). This microalga originates from the Experimental Phycology and Culture Collection of Algae of the University of Goettingen, Germany.

The cultivation of P. purpureum was performed using an artificial seawater medium (Jones et al. 1963). Because this alga tends to an extreme formation of foam, the seawater medium was repeatedly mixed with 1 mL of polyethylene glycol 400 during cultivation. The cultivation took place in photobioreactor screening modules under monoseptic conditions at a temperature of 25°C. CO2 (2.5%) was added to the process gas (synthetic air, 0.9 L min−1). The reactor was irradiated by four concentrically arranged fluorescent lamps providing different PFD with the objective of optimizing coenzyme Q10 biosynthesis: 40, 80, 120, 160, 120 to 180 μmol photons∙m−2 s−1 (increased at optical density of 1.0) and 40 to 250 μmol photons∙m−2 s−1 (increased continuously by keeping the light transmission at the photobioreactor constant using a light controller. Photon flux density was measured on the outside of the reactor in the center of the bubble column (height) diagonal to the column and the fluorescent tubes.

Cultivation of P. purpureum in a Medusa photobioreactor

A scale-up cultivation to investigate the biosynthesis of coenzyme Q10 was performed using the 120-L-scale Medusa photobioreactor. Supply air was supplemented by 2.5% CO2. The temperature was adjusted to 25°C. Photosynthetic PFD was increased continually from 80 to 225 μmol photons∙m−2 s−1 by keeping the light transmission constant using a light controller (Steinau Verfahrenstechnik) which was fixed at one of the glass tubes.

CoQ10 extraction–manual procedure

The separation of cell mass from the culture supernatant by centrifugation (27,000×g, 8°C, 5 min) was performed after the termination of the cultivation, followed by cell disruption using ultraturax treatment (14,000 rpm, 5 min, T18, basic, IKA; Germany). Lipophilic cellular constituents were extracted by n-pentane from lyophilized biomass (25 mL n-pentane∙0.5 g−1 dry weight (dw)). The extract was analyzed by reversed phase high performance liquid chromatography electrospray ionization mass spectrometry (RP-HPLC-ESI-MS) regarding the coenzyme Q10 concentration.

Ubiquinone extraction by Accelerated Solvent Extraction®

In order to optimize and automize the extraction procedure of coenzyme Q10, an Accelerated Solvent Extraction® (ASE) (Dionex, Germany) was used. This extraction method is based on high pressure. Optimization was done by variation of parameters such as pressure (50, 100, 150 bar), extraction time (1 and 5 min), and solvent (n-pentane, methanol, acetone, dioxane). The influence of the rinsing volume (given in percent compared to the volume of the extraction cell) was also investigated. After finishing the extraction procedure, the extraction cell was rinsed with nitrogen. Afterwards, the solvent was evaporated with nitrogen, the extract was resolved by 1 mL n-pentane, and the ubiquinone concentration was analyzed by HPLC.

HPLC analysis

HPLC analysis were performed using a Shimadzu HPLC system (LC-20AT, SPD-10AV) on a C8-column (Eurosphere 100-SC8250∙2 mm, Knauer GmbH, Germany). A UV detector enabled the detection of coenzyme Q10 at a wavelength of 277 nm. Acetonitrile/propan-2-ol 90/10 (v/v) with 3.65 mM ammonium acetate was used as mobile phase at a flow rate of 0.2 mL min−1 (isocratic). The column temperature was adjusted to 30°C. The HPLC system was linked to an ESI-MS system (API 2000, Applied Biosystems, USA; scan type: Q1 MS, declustering potential 70 V, entrance potential 10 V). Measurements were made in the positive mode. The quantification of the CoQ10-containing samples was performed by UV detection applying an external calibration curve by using a commercially available coenzyme Q10 standard (Sigma–Aldrich, Germany).

Results

Microalgae growth

Biotechnology of microalgae always shows high dependency on light supply, especially regarding growth and biosynthesis of bioactive products. Therefore, P. purpureum was cultivated at different photon flux densities using photobioreactor screening modules and an artificial seawater medium. Due to the resulting differences in growth behavior, a variation of time of harvest was chosen. The cultivations were stopped in stationary growth phase. Best results could be achieved by a continuous increase of PFD and by a sudden rise of PFD at an optical density of 1.0. Biomass concentrations of 9.9 (continuous increase of PFD) and 9.5 g  L−1 (sudden increase of PFD), respectively, could be attained. Cultivations of P. purpureum that were performed applying PFDs of 80, 120, and 160 μmol photons∙m−2 s−1 showed biomass concentration of 5.0 (80 μmol photons∙m−2 s−1) and 5.2 g  L−1 (120 and 160 μmol photons∙m−2 s−1), respectively. Further reduction of PFD to 40 μmol photons∙m−2 s−1 led to a biomass concentration that was even lower (4.7 g  L−1) (see Fig. 2).

Fig. 2
figure 2

Growth characteristic of P. purpureum as a function of photon flux density: empty triangles 40 μmol photons∙m−2 s−1 (left y-axis), filled diamonds 80 μmol photons∙m−2 s−1 (left y-axis), empty squares 120 μmol photons∙m−2 s−1 (right y-axis), empty inverted triangles 160 μmol photons∙m−2 s−1 (left y-axis), filled inverted triangles 40–250 μmol photons∙m−2 s−1—continuous increase of PFD (right y-axis), empty diamonds 120–180 μmol photons∙m−2 s−1—sudden increase of PFD (right y-axis), filled triangles Medusa photobioreactor 80–225 μmol∙m−2 s−1 (right y-axis). Error Bars = Standard deviation; n = 3

Full size image

In order to investigate the influence of a scale-up on biosynthesis of ubiquinone by P. purpureum, a cultivation using the 120 L Medusa photobioreactor was performed. To reach high biomass and ubiquinone concentrations, a continuous increase of PFD from 80 to 225 μmol photons∙m−2 s−1 was chosen. Due to significant lower illuminated surface/volume ratio (65 m−1) compared to the photobioreactor screening modules (80 m−1), the initial PFD was set to 80 μmol photons∙m−2 s−1. Considering the growth characteristics of both cultivations, a significant delay in reaching stationary growth phase could be observed in the scale-up cultivation compared to the cultivation in photobioreactor screening modules. The use of different calibration curves (optical density/dry weight) led to marked lower dry weight at the beginning of the cultivation because all cultivations were started at the same optical density (0.1). At the end of the cultivations, a total of 9.9 g  L−1 (cultivation in photobioreactor screening module—continuous increase of PFD) and 14.0 g  L−1 (cultivation in 120 L scale), respectively, could be attained. Therefore, the maximum biomass concentration of the Medusa cultivation was 41% higher, compared to the dry weight achieved using the photobioreactor screening modules (see Fig. 2).

HPLC-ESI-MS analysis ubiquinone biosynthesis

Chromatographic analysis indicated an increase of ubiquinone biosynthesis with ascending PFD. High PFDs showed positive influence on specific as well as on volumetric coenzyme Q10 concentration in P. purpureum. Lower PFDs (40 μmol photons∙m−2 s−1) entailed a ubiquinone concentration of 1.5 μg g −1dw , which was 14.1 times lower compared to the cultivation performed with a continuous increase of light supply (21.2 μg g −1dw ). In contrast, a sudden rise of PFD from 120 to 180 μmol photons∙m−2 s−1 (at an optical density of 1.0) seemed to inhibit coenzyme Q10 biosynthesis (6.1 μg g −1dw ) because the specific ubiquinone concentration was 3.5 times lower compared to a continuous increase of PFD. Investigating PFDs of 120 and 160 μmol photons∙m−2 s−1 specific ubiquinone concentrations of 15.8 and 20.8 μg g −1dw , respectively, were achieved. Optimization of coenzyme Q10 production by P. purpureum could be realized and indicated a volumetric product concentration of 209.2 μg L−1 due to continuous increase of PFD from 40 to 250 μmol photons∙m−2 s−1 (see Fig. 3).

Fig. 3
figure 3

Specific (white) and volumetric (black) coenzyme Q10 concentration of lipophilic extracts of P. purpureum subject to photon flux density. Error Bars = Standard deviation; n = 3

Full size image

Accelerated Solvent Extraction®

Optimization and automization of the extraction of ubiquinone from microalgae cells was investigated applying a high-pressure extraction devise (ASE®) under variation of different parameters like pressure, rinsing volume, solvents and extraction time. ASE® was performed using harvested biomass of P. purpureum cultivated in a 120-L Medusa photobioreactor. Thereby, product recovery could be enhanced by increasing the rinsing volume up to 40%. Highest product yield under variation of extraction pressure was reached at 100 bar. Higher pressure such as 150 bar leads to a decrease in ubiquinone recovery. HPLC results showed an influence of extraction time on product yield, with positive effects of short extraction time (1 min). In applying optimized extraction parameters (pressure, time, and rinsing volume), a variety of solvents was investigated showing best results for methanol compared to n-pentane, acetone, dioxane, and propan-2-ol. Compared to standard extraction procedure (15 min stirring using n-pentane), an increase of product recovery from 10.33 up to 141.23 μg g −1dw (1.96 mg L−1) could be realized by utilization of ASE® (methanol, 100 bar, 40% rinsing volume, 1 min), which was equivalent to a space–time yield of 79 μg L−1 d−1 (see Fig. 4).

Fig. 4
figure 4

Coenzyme Q10 concentration of lipophilic extracts of P. purpureum extracted by an ASE® system (n-pentane: white, black: variation of solvent)

Full size image

Discussion

Microalgae growth

Cultivations of P. purpureum in photobioreactor screening modules showed a great dependency of PFD on growth. A continuous increase of PFD led to the highest biomass concentration and cell density, whereas lower PFDs resulted in a delay of reaching the stationary growth phase and in lower biomass concentrations. A sudden increase of PFD from 120 to 180 μmol photons∙m−2∙s−1 also showed a delay in reaching the stationary growth phase, whereas the final biomass concentration was similar to the cultivation performed by continuous increase of PFD, which showed the highest biomass concentration during these investigations. Thepenier and Gudin (1985) and Jones et al. (1963) found that (based on cell number) a continuous light supply is favored by P. cruentum (purpureum) compared to light/dark cycles (16/8 h), while no significant differences could be detected in biomass concentrations. Jones et al. (1963) also found that higher PFDs showed a positive influence on growth of P. cruentum. These results could be confirmed by the present study.

Due to the setup of the PSM, a light supply of nearly 100% can be achieved, whereas about 16% of the culture broth cultivated in a Medusa photobioreactor (120 L) is not fed by light (at a definite time), which was the reason for the choice of a higher starting PFD when this parameter was increased continuously. During the exponential growth phase, major differences are obvious between PSM and Medusa photobioreactor cultivations as a steeper exponential phase could be observed using PSM. Furthermore, the transitional phase was much shorter. Due to the smaller illuminated surface/volume ratio, the stationary growth phase was reached later in 120 L Medusa photobioreactor (after 20 days of cultivation). Present results correlated with Thepenier and Gudin (1985) who found out that a doubling of illuminated surface area showed positive results on the growth of P. cruentum. Moreover, the space–time yield was nearly twice as high using PSM (1.03 g  L−1 d−1–continuous increase of PFD) compared to Medusa photobioreactor (0.56 g  L−1 d−1–continuous increase of PFD). Despite a much lower space–time yield achieved in the Medusa photobioreactor, higher biomass concentration could be reached compared to cultivation performed in the photobioreactor screening module (both continuous increase of PFD) because the growth curve was shifted backwards during cultivation in the Medusa photobioreactor, which could also be the result of lower illuminated surface area.

Light and photobiological oxygen production might lead to oxidative stress, which can be the reason for a decrease of cell growth. Current research did not indicate reduced cell growth due to increase of PFDs under the cultivation conditions used. Obviously, P. purpureum is adapted to photooxidative stress due to biosynthesizing a sufficient amount of antioxidants. The photobioreactors used can be regarded as an appropriate system cultivating microalgae P. purpureum under chosen cultivation parameters.

Ubiquinone biosynthesis

HPLC analysis demonstrated an increase of coenzyme Q10 biosynthesis due to rising PFD during cultivations in photobioreactor screening modules. Increase of PFD presumably led to higher oxygen production by microalgae as a result of an assumed increase of photosynthetic activity and hence to oxidative stress inside the cells. The reason for this might be photooxidation of membrane-bound unsaturated fatty acids. Produced oxygen can be transformed to singlet status by intracellular sensibilizators like chlorophyll a, which are activated by light. Singlet oxygen reacts to hydroperoxides due to a cyclo addition reaction with unsaturated fatty acids. Hydroperoxides can cause cell damage ROS due to homolytical decay (Belitz et al. 2001; Xu et al. 1999; Yang et al. 2003). As a result, microalgae cells produce antioxidants like coenzyme Q10, α-tocopherol, or plastoquinone (Carballo-Cárdenas et al. 2003; Dionex 2006; Hundal et al. 1995). Membranes of P. purpureum possess oxidation-sensitive PUFAs like eicosapentanoic acid. Direct oxygenation of these fatty acids, due to photooxidation and subsequent formation of alkoxy or peroxy radicals, led to a radical chain reaction of lipid peroxidation. Formed radicals are stabilized mesomerically resulting in an increase of their half life enabling further peroxidation (Belitz et al. 2001; Klyachko-Gurvich et al. 1999). Suppression of such peroxidation of membrane-bound lipids requires biosynthesis of lipophilic antioxidants like coenzyme Q10. This proposed mechanism explains the increase of ubiquinone concentration in cells of P. purpureum as a result of increase of PFD. By contrast, an increase of PFD from 120 to 180 μmol photons∙m−2 s−1 (at an optical density of 1.0) indicated a decrease of ubiquinone biosynthesis. One reason for this phenomenon might be that plastoquinone was produced instead.

Analysis of ubiquinone biosynthesis also revealed a significant increase of space–time yield of coenzyme Q10 based on a surface/volume ratio of 80 m−1 (photobioreactor screening module–continuous increase of PFD; space–time yield = 21.7 μg L−1 d−1) compared to a surface/volume ratio of 65 m−1 (Medusa photobioreactor—continuous increase of PFD; space–time yield = 5.8 μg L−1 d−1), due to a much longer cultivation duration in the Medusa photobioreactor. Current results suggested a positive influence of high PFDs as well as high surface/volume ratios on ubiquinone production by P. purpureum.

Downstream processing using ASE®

Using different solvents and varying pressure, extraction time and rinsing volume, coenzyme Q10 was extracted from P. purpureum cells by an automized method called Accelerated Solvent Extraction®. Compared to the standard extraction procedure (stirring 15 min, room temperature, atmospheric pressure, 25 mLn-pentane g −1dw ), an optimization of product recovery by a factor of 14 could be achieved using appropriate extraction parameters. A rise in solvent volume resulted in an enhancement of product recovery in a range from 10% to 40% rinsing volume. Due to the problem of solvent evaporation, larger amounts of solvent did not seem useful.

Variation of the extraction pressure indicated an increase of ubiquinone concentration up to 100 bar. Higher pressures showed a decrease of coenzyme Q10 yield. Pressures higher than 100 bar possibly prevented entire wetting of the biomass by the solvent.

Applying n-pentane as a solvent, the influence of the extraction between time 1 and 5 min was investigated. An extension of the extraction duration had a negative effect on ubiquinone yield. One explanation might be the temperature sensitivity of coenzyme Q10. The temperature could not be adjusted to low values using this special apparatus (demonstration device). The optimized parameters based on the usage of n-pentane were used afterwards to identify an optimal solvent for the ubiquinone extraction from P. purpureum biomass. Since methanol causes membrane disintegration more than any other solvent investigated, treatment with methanol indicated a significant enhancement of product recovery. Contrary to ASE® conditions (100 bar), methanol could not be used for the standard extraction procedure because ubiquinone is nearly insoluble in methanol at atmospheric pressure. Thus, automations and optimization of the product extraction using ASE® was successful.

Experiments on the variation of PFD during the cultivation of the red alga P. purpureum indicated an increase of coenzyme Q10 biosynthesis as the PFD rose. A prerequisite for this analysis is an axenic cultivation of the alga in sterilizable photobioreactor screening modules. These provide the technical prerequisite for further analyses of the metabolic potential of phototrophic microorganisms with regard to bioactive agent research. Furthermore, an optimization of the product recovery could be demonstrated using Accelerated Solvent Extraction®.