Introduction

Arthrospira (Spirulina) sp. is a blue-green filamentous cyanobacterium that is used as dietary supplement [1]. In addition to being rich in proteins (60–70% of biomass dry weight), Arthrospira produces potentially useful compounds including the blue pigment C-phycocyanin (C-PC) [2], the orange-red carotenoids (mainly β-carotene), polyunsaturated fatty acids (e.g., gamma linolenic acid) and the sulfated polysaccharide spirulan. Around 3000 metric tons of Arthrospira is commercially produced annually worldwide [3]. Shallow outdoor open ponds are mainly used for commercial production via photoautotrophic growth, but more controlled production in indoor closed photobioreactors can be used for higher-value Arthrospira products. Photoautotrophic cultures use inorganic carbon, water, mineral salts and natural or artificial light to grow. Arthrospira can be grown also on organic carbon [4, 5] with or without light, but this kind of culture requires expensive carbon nutrients and a rigorous aseptic operation to prevent contamination by unwanted microorganisms which grow rapidly on organic nutrients.

Photoautotrophic production of Arthrospira biomass and metabolites is affected by factors such as temperature [6, 7], light intensity [8], the culture pH [9], and the supply of inorganic nutrients [8, 10,11,12]. The size of the inoculum may also play a role [13]. Given a sufficiency of all nutrients, the light intensity and temperature most influence the production [10, 14, 15].

C-PC and extracellular polymeric substances (EPS) are two important Arthrospira metabolites. C-PC is a water-soluble phycobiliprotein with a light harvesting function. Phycobiliproteins account for nearly 60% of all soluble protein in cyanobacteria [16]. C-PC is an antioxidant [17]. It is extensively used as a pigment in food, cosmetic and nutraceutical industry [2]. Phycobiliproteins are further reviewed in the literature [18, 19].

EPS are heteropolysaccharides, or polysaccharides bound to proteins, uronic acids and other compounds that are yet to be fully characterized. EPS production in cyanobacteria and other microorganisms is hypothesized to protect the cells from stress and harmful conditions [20]. EPS production generally occurs in the stationary phase of culture [21] and also when the cells are subjected to stressful conditions [10, 22]. When light is in excess and growth is prevented by some form of stress, the excess light is used to make EPS which is excreted [23]. EPS are used as food thickeners and for the improvement of food texture. They have many other potential applications [22, 24]. For example, the sulfated polysaccharides are known to have anticoagulant [25], antiviral [26] and antioxidant properties [27].

Accumulation of C-PC in Arthrospira is strongly influenced by light. C-PC occurs in thylakoid membranes as a photosynthetic accessory pigment. Its main function is to channel light energy to chlorophyll a in the chloroplast [19, 28]. The transcripts encoding for phycocyanin subunits and their linker polypeptides are modulated by light level [29]. Cells grown under low light have a higher content of C-PC and certain other components (e.g. allophycocyanin, chlorophyll a, soluble protein, total protein), but reduced levels of glycogen. A low light level also increases the doubling time of cells [29]. Therefore, an optimal light level needs to be established to maximize the production of C-PC. The temperature and the composition of the culture medium are other factors that affect accumulation of C-PC. The nature of the nitrogen source and its concentration in the medium are particularly important [30, 31] as C-PC is a protein and nitrogen is required for the synthesis of the amino acids that make up the C-PC and other key components of the cell. C-PC serves to store nitrogen and is selectively degraded when the cells are starved of nitrogen [32,33,34].

The EPS produced by certain Arthrospira sp. has been characterized. For example, EPS produced by Arthrospira platensis is known to have a molecular weight in the range of 81–98 kDa [20]. This EPS is made of seven neutral sugars (14.9% galactose, 14.3% xylose, 13.2% glucose, 13.2% fructose, 3.7% rhamnose, 1% arabinose, 0.3% mannose) and two uronic acids (13.5% galacturonic acid, 0.9% glucuronic acid) [24]. Factors influencing the production of EPS include light intensity, temperature, pH, age of culture, ultraviolet radiation and the composition of the medium (mainly NaCl, sulfur, iron, phosphate and potassium) [10, 20, 22, 27]. A relatively high temperature and a high light intensity promote the production of EPS [24].

Taguchi method of statistical design of experiments [35,36,37] has proved useful for optimizing many fermentation processes [38,39,40,41,42]. Taguchi method uses orthogonal arrays to reduce the number of factors in an experimental design. Various types of signal (S) to noise (N) ratios (S/N ratios) are then used to establish how well the measured response is attributable to variation in the factor values relative to random experimental noise. Taguchi method facilitates the identification of the influence of individual factors on performance and the optimum levels of factor values using a few well-defined experimental datasets [40].

This study used Taguchi method of experiment design to determine the optimal conditions for producing C-PC and EPS from Arthrospira sp. The identified optimal conditions were confirmed through validation experiments. These conditions were then used in a 30-L raceway pond to scale up the production by nearly 150-fold.

Materials and methods

Microorganism and maintenance

Arthrospira maxima IFRPD 1183 and Arthrospira platensis IFRPD 1208 were obtained from Institute of Food Research and Product Development (IFRPD), Kasetsart University, Bangkok, Thailand. The cultures were maintained at room temperature in Zarrouk’s medium [43]. This medium was also used in the preparation of inoculums for all experiments. The medium contained the following components (per L): 16.8 g NaHCO3, 0.5 g K2HPO4, 2.5 g NaNO3, 1 g K2SO4, 1 g NaCl, 0.2 g MgSO4·7H2O, 40 mg CaCl2, 10 mg FeSO4·7H2O, 80 mg Na2EDTA·2H2O, 1.0 mL of trace elements A5 and 1.0 mL of trace elements B6. The solution A5 contained the following (per L): 2.86 g H3BO3, 1.81 g MnCl2·4H2O, 0.22 g ZnSO4·7H2O, 80 mg CuSO4·5H2O, and 10 mg MoO3. The solution B6 contained the following (per L): 22.9 mg NH4VO3, 96.0 mg K2Cr2(SO4)4·24H2O, 47.8 mg NiSO4·7H2O, 17.9 mg Na2SO4·2H2O, and 44 mg Co(NO3)2·6H2O. The pH of the final medium was adjusted to 9.0.

Inoculum preparation

Cells from the stock culture were transferred to 150 mL of the same medium in a 200-mL microalgal culture tube (32.2 mm internal diameter, 3.9 mm wall thickness, made of optically clear glass) (Fig. 1a–d). The initial concentration of the cells corresponded to an optical density of 0.2 measured at 560 nm (Genesys 20 spectrophotometer; Sigma-Aldrich, St. Louis, MO, USA) against a blank of distilled water. The culture tubes were incubated in a water bath at 30 °C under light (six 36 W cool daylight fluorescent lamps; Philips® Co., Bangkok, Thailand) (Fig. 1b, c). During the photoperiod, the light intensity at the surface of the tubes was 203 μmol photons m−2 s−1. A diurnal cycle of a 16-h photoperiod and 8-h dark was used. The culture tubes were continuously sparged with air mixed with carbon dioxide (2% v/v) at a flow rate of 0.6 vvm. The biomass was harvested on day 3 of culture, filtered and washed with sterile distilled water to remove all nutrients.

Fig. 1
figure 1

The culture systems: a–d 200-mL culture tubes; and e the 30-L raceway pond

Full size image

Optimization of C-phycocyanin and EPS production

The production of C-PC and EPS was first optimized in 200-mL batch culture tubes. Each batch ran for 14 days. Each tube containing 150 mL of the Zarrouk’s medium was inoculated with the preculture to obtain an initial optical density of 0.2 measured at 560 nm.

Taguchi method was used for simultaneous optimization [37] of the production of C-PC and EPS. Six factors (A–F; Table 1), or variables, were optimized at three or two levels. The factors were as follows: (A) light intensity varied at three levels with at a fixed diurnal cycle of a 16-h photoperiod and 8-h dark; (B) the initial culture pH varied at three levels; (C) the cyanobacterial species varied at two levels; (D) the concentration of Zarrouk’s medium (% of full strength) varied at three levels in distilled water; (E) aeration rate of the culture tube with a mixture containing 2% (v/v) carbon dioxide, varied at three levels; and (F) the incubation temperature varied at two levels.

Table 1 Experimental factors and their levels for optimizing C-PC and EPS production
Full size table

Qualitek-4 software (Nutek Inc., Bloomfield Hills, MI, USA) was used to identify the trial experimental profiles shown in Table 2. The experimental data (y i ) were converted to a signal-to-noise (S/N) ratio [37]. The objective was to maximize the values of the product (C-PC, EPS) concentration, productivity and yield on biomass. Therefore, the S/N ratio was calculated for larger-the-better performance criterion, as follows:

$$ S/N = - \log_{10} \left( {\frac{{\sum {\left( {\frac{1}{{y_{i}^{2} }}} \right)} }}{n}} \right) $$
(1)

where y i is the measured data in experiment i for a set levels of the factors and n is the number of trials [37]. The expected values of observations (Y opt), i.e., the concentration, the productivity and yield of C-PC and EPS, were calculated using the following equation:

$$ Y_{\text{opt}} = \bar{T} + \sum {(\bar{F}_{i} - \bar{T})} $$
(2)

where \( \bar{T} \) and \( \bar{F}_{i} \) are the grand averages of the S/N ratios and the factor averages at each factor level, respectively. The main effect was calculated as the difference between the maximum and minimum values of the factor averages in each factor level. The percent main effect of each factor was calculated as the percentage of its main effect divided by the sum of the main effects of all the factors. Details of Taguchi’s statistical optimization are further discussed in the literature [36, 37, 40].

Table 2 The factor levels in the experimental runs
Full size table

C-PC and EPS production in 30-L open raceway pond

Once the optimal cyanobacterial species and the culture conditions (i.e., the light intensity, the initial culture pH, the concentration of Zarrouk’s medium, the flow rate of air/CO2 mixture, and the culture temperature) had been identified in 200-mL culture tubes, these conditions were used to grow the bacterium in the 30-L open raceway pond for 14 days (Fig. 1e). The pond had semicircular ends (0.2 m radius), a total length of 1.2 m and a width of 0.4 m. The length of the central dividing baffle was 0.8 m. The static depth of the culture was 0.07 m. A 6-bladded paddlewheel (diameter = 0.26 m) was used for mixing. The raceway was continuously sparged at a specified flow rate with a filtered (0.2 μm Teflon membrane filter; Sartorius) aeration gas (air containing 2% v/v CO2). An aquarium stone diffuser was used as sparger. The sparger was placed in front of the paddlewheel, 0.435 m from the far end of the pond. The paddlewheel was operated continuously at a rotational speed of 25 min−1.

The temperature of the pond was controlled using an electrical heater. The pond was illuminated by eight 36 W cool daylight fluorescent lamps (Philips® Co., Bangkok, Thailand) located above it. A light–dark cycle of a 16-h photoperiod and an 8-h dark period was used.

Analytical methods

The biomass concentration (g L−1) during culture was measured gravimetrically. A 10 mL sample of the culture broth was filtered through of 0.65 µm Sartorius-membrane filter. The recovered biomass was washed twice each time with 10 mL of distilled water. The biomass was then dried to constant weight in an oven at 105 °C, cooled to room temperature in a desiccator and weighed.

For standardizing the inoculum, a sample of the culture broth was serially diluted with the fresh medium and the optical density (OD560) of the dilutions was measured at 560 nm against a blank of distilled water. The biomass concentration (C b, g L−1) was correlated to the optical density using equations of the form C b = f·OD560 where f was 0.887 (R 2 = 0.999) for A. maxima and 0.826 (R 2 = 0.994) for A. platensis.

The nitrogen concentration in the cell-free culture supernatant was determined by brucine colorimetric method [44]. The pH of the culture broth was measured using a calibrated pH meter.

The C-PC was extracted from the filtered and washed fresh biomass. A 10 mL sample of the culture broth was filtered through a plankton cloth (50 μm pore size). The biomass on the cloth was washed twice with 0.9% NaCl (10 mL per wash). The biomass was then quantitatively transferred to a centrifuge tube using 25 mL of 0.1 M sodium phosphate buffer (pH 7). The resulting suspension was mixed well and incubated at 37 °C for 18 h in the dark. This suspension was centrifuged (10,000×g, 4 °C, 20 min) and the supernatant was recovered. The spectrophotometric absorbance of the supernatant was measured at 620 nm against a blank of 0.1 M sodium phosphate buffer, pH 7 (Miyakawa K, Siam Algae Co., Ltd., Thailand; personal communication, 2014). The concentrations were calculated using the following equations (Miyakawa K, Siam Algae Co., Ltd., Thailand; personal communication, 2014):

$$ {\text{C-PC}}\;({\text{mg}}) = \frac{{A_{620} \times 1000 \times 25}}{6500} $$
(3)
$$ {\text{C-PC}}\;({\text{mg}}\;{\text{g}}^{ - 1} ) = \frac{{{\text{C-PC}}\;({\text{mg}}) \times 1000}}{w} $$
(4)

where A 620 was the absorbance measured at 620 nm and w (mg) was the dry weight of the water-washed biomass recovered from a 10 mL sample of the broth identical to the one used in recovering the fresh biomass for extraction.

The EPS concentration was determined in the culture supernatant using the method developed by Moore and Tischer [45]. Briefly, a 5 mL of sample of the culture broth was filtered (Whatman filter paper No. 2) to remove the cells. The filtrate was vigorously mixed (vortex mixer) with two volumes of 95% ethanol and incubated overnight at 4 °C. This mixture was centrifuged (10,000×g, 4 °C, 20 min) and the supernatant was discarded. The EPS pellet was recovered and washed (2 × 5 mL) with deionized water by shaking manually for 1 min. The centrifugation step was repeated. The pellet was recovered and suspended in 5 mL of deionized water. The EPS concentration in this suspension was determined using phenol–sulfuric acid method [46].

For specified samples of the washed biomass of the validation experiments, the content of C-PC, the total protein, total carbohydrate and the total chlorophylls in the biomass were determined. The total C-PC was determined as specified above. The total protein was determined by Kjeldahl method [47]. The total carbohydrate was determined using the phenol–sulfuric acid method [46] and total chlorophylls were determined by a modified method of Bennett and Bogorad [48].

Only the samples from the open pond were measured for phosphorus concentration in the culture supernatant. Phosphorus was measured using the vanodomolybdate colorimetric method [49].

Calculations

The specific rates and the volumetric rates of growth and formation of products were calculated as previously described [50].

Results and discussion

Growth in 200-mL culture tubes

The growth and product formation data for the two cyanobacteria grown separately under various conditions (Table 2) for 14-days are summarized in Table 3.

Table 3 Summary of the growth and production kinetic parameters for 14-day cultures of Arthrospira a
Full size table

The runs 3, 6, 13 and 16 (Table 3) achieved relatively high values of the biomass concentration (6.27–7.39 g L−1) within 4–10 days. These runs were carried out at high light levels (factor A; 135–203 µmol photons m−2 s−1) using concentrated media (factor D; 75–100% of full strength) and temperatures of 30–35 °C (factor F) (Table 2). The biomass in the runs conducted at the above specified high light levels in relatively concentrated media (D ≥ 75%) also generally displayed high specific growth rates (µ) and short doubling times (t d) (Table 3). The biomass grown under low light (e.g. runs 9–12; A = 68 µmol photons m−2 s−1; Table 2) had lower specific growth rates (μ ≤ 0.26 d−1; Table 3). Therefore, the concentration of nutrients in the medium (i.e., factor D) and the light intensity (factor A) were clearly important in influencing growth and biomass production. An overly dilute culture medium (i.e., a 50% concentration of Zarrouk’s medium) limited growth to ≤3.88 g L−1 within 14 days (Table 3).

These results were generally consistent with earlier reports of an increasing final biomass concentration and growth rate with increasing light level [10, 24, 51]. Both the incubation temperatures used proved satisfactory for both species of Arthrospira (Table 3) although in earlier studies a temperature of 35°C had reduced growth of A. platensis growth [9, 52].

Influence of factors on C-PC and EPS production (main effects)

Data on the maximum concentration, the volumetric productivity and yield of C-PC and EPS are summarized in Table 3. Based on the data shown, the optimal conditions for maximizing the production of C-PC and EPS were quite different. Highest concentration of C-PC (C C-PC ≥ 545 mg L−1) was found in the experimental runs 3, 6 and 9 (Table 3). In contrast, the runs 2, 4 and 15 afforded the highest concentrations of EPS (C EPS ≥ 124 mg L−1). The highest productivity of C-PC (Q C-PC = 70–75 mg L−1 d−1) occurred in experimental runs 1, 6, 14 and 15 (Table 3) while the highest productivity of EPS (Q EPS = 8.9–11.0 mg L−1 d−1) occurred in runs 2, 4 and 15. Therefore, EPS production tended to be maximal under a combination of high light (A = 203 µmol photons m−2 s−1), high concentration (D = 50–75% of full strength) of Zarrouk’s medium, and a high temperature (F = 35 °C), irrespective of the initial culture pH and the microbial species used. In contrast, C-PC concentration tended to be maximal under a combination of high light (A = 203 µmol photons m−2 s−1), high temperature (F = 35 °C) and high concentration (D ≥ 50%). This notwithstanding, the yield of C-PC on biomass was maximal under relatively low light (A = 135 µmol photons m−2 s−1 in runs 5, 6, 7; A = 68 µmol photons m−2 s−1 in runs 9, 10 and 11), irrespective of the settings of the other factors.

The EPS productivity was extremely low (Q EPS ≤ 1.2 mg L−1 d−1; runs 9–12, Table 3) under low light (A = 68 µmol photons m−2 s−1) irrespective of the value of all the other factors. The best case EPS productivity (runs 2 and 4; Table 3) was 26-fold higher than in the worst case productivity (run 9, Table 3).

A statistical analysis of the percentages of main effect of the factors on the concentration, volumetric productivity and yield of C-PC and EPS, is shown in Table 4. The concentration of Zarrouk’s medium (factor D) and temperature (factor F) were the most significant factors in relation to C-PC concentration. While the light intensity (factor A) and the initial culture pH (factor B) were most influential in affecting the C-PC productivity and yield.

Table 4 The percentages of main effect of each factor on the production of C-PC and EPS
Full size table

The light intensity (factor A), the temperature (factor F) and the concentration of Zarrouk’s medium (factor D) were the three main influences on EPS concentration, productivity and yield (Table 4). The choice of the cyanobacterium (factor C) had the least impact on EPS production (main effect < 3%; Table 4).

Concentrations of certain elements (mainly sulfur, nitrogen, phosphorus, carbon, and iron) in the culture medium have been previously reported to influence synthesis of pigments including C-PC and allophycocyanin by Arthrospira sp. and other cyanobacteria [53]. In the present study, the concentration of these elements varied in the medium as a result of changes in values of factor D, the concentration of Zarrouk’s medium, but the relative ratios of the key elements excluding carbon were constant in all experiments. In earlier studies, the concentrations of nutrients were not simultaneously varied with the environmental factors (i.e., light intensity, pH, temperature and aeration rate) as in the present work.

Analysis of variance

Analysis of variance (ANOVA) was used to determine the contributions of individual factors to variation in responses relating to the production of C-PC (Table 5) and EPS (Table 6).

Table 5 Analysis of variance (ANOVA)a of factors affecting the production of C-PC
Full size table
Table 6 Analysis of variance (ANOVA)a of factors affecting the production of EPS
Full size table

Temperature (factor F) had a statistically significant (p < 0.05) effect on the concentration of C-PC (Table 5) in keeping with data in Table 4. Both light intensity (factor A) and temperature (factor F) showed statistically significant effects on the concentration (p < 0.01) and productivity of EPS (p < 0.10) (Table 6). The results were consistent with the previously discussed percentages of main effect (Table 4).

C-PC concentration was most strongly influenced by temperature (factor F) and the concentration of Zarrouk’s medium (factor D) (Table 5). In contrast, the light intensity (factor A) and the cyanobacterial species (factor C) had the most impact on C-PC productivity (Table 5) although the impact of choice of species was relatively minor.

Concerning EPS production (Table 6), the light intensity (factor A) had the most impact on its concentration, productivity and yield. Temperature (factor F) affected production to a much lesser level (Table 6). All the other factors (i.e., the initial culture pH; the choice of species; the concentration of Zarrouk’s medium; and aeration rate) did not individually influence EPS production much. Overall, nearly 80% of the variation in data could be explained by changes in light intensity (factor A) and the remaining 20% was attributed to settings of the other factors (Table 6).

Optimum conditions for production of C-PC and EPS

Taguchi method uses the signal-to-noise (S/N) ratio (Table 3) as the principal criterion for identifying the optimal conditions [37]. High values of the S/N ratio are taken to indicate the optimal conditions (Fig. 2). As shown in Fig. 2, the optimal conditions for the production of C-PC (Fig. 2a–c) and EPS (Fig. 2d–f) were different. The optimal conditions of maximizing the concentration of C-PC were found to be the following: a light intensity of 68 µmol photons m−2 s−1, an initial pH of 10.0, use of A. maxima IFRPD1183 as the producer, a Zarrouk’s medium strength of 100%, a temperature of 30 °C and aeration at a flow rate of 0.6 vvm using air mixed with 2% v/v CO2.

Fig. 2
figure 2

Signal-to-noise (S/N) ratios for the various factors (A–F, Table 1) and levels. The level of each factor is shown as a numerical value after the factor on the x-axes. The factors were: A, light intensity; B, initial culture pH; C, the cyanobacterial species; D, concentration of Zarrouk’s medium; E, aeration rate; and F, temperature. Effects on: a C-PC concentration; b C-PC productivity; c C-PC yield on biomass; d EPS concentration; e EPS productivity; and f EPS yield on biomass. Optimal conditions are indicated by the peak values of the S/N ratio

Full size image

The optimal conditions of maximizing the concentration of EPS were found to be the following: a light intensity of 203 µmol photons m−2 s−1, an initial pH of 9.5, use of A. maxima IFRPD 1183 as the producer, a 50% strength of Zarrouk’s medium, a temperature of 35 °C and an aeration rate of 0.2 vvm using air mixed with 2% v/v CO2. A higher temperature of 35 °C promoted EPS production while a lower temperature of 30 °C promoted growth and C-PC production. Overall, A. maxima IFRPD 1183 proved to be a better producer of both C-PC and EPS relative to A. platensis IFRPD 1208.

The equations shown in Tables 5 and 6 were used to estimate the expected values (Y opt) of concentrations, productivities and yields of the two products under various conditions. Under the above specified optimal conditions, the predicted or expected values (Table 7) of the concentration, productivity and yield of C-PC were 758.1 mg L−1, 86.6 mg L−1 d−1 and 190.0 mg g−1, respectively. Similarly, the predicted values of the concentration, productivity and yield of EPS (Table 7) were 154.0 mg L−1, 10.9 mg L−1 d−1 and 37.8 mg g−1, respectively.

Table 7 Summary of C-PC and EPS production by A. maxima IFRPD 1183
Full size table

The above results were generally consistent with the earlier observations. For example, a lower light intensity has been reported to favor the production of C-PC [54,55,56,57]. A high light results in fewer thylakoids per cell and a reduced content of phycobiliproteins in the cell [29, 51, 58]. Earlier work with A. platensis grown under low light showed the cells to contain three species of phycocyanin mRNA but in high light one of these species disappeared [29]. As in our work, a high light intensity has been found to generally favor the production of the biomass as well as EPS [10, 22].

Confirmation of C-PC and EPS production under the optimal conditions

Once the optimal production conditions had been identified (see previous section), confirmation experiments were carried out using these conditions to validate the predicted values of the concentrations, productivities and yields of C-PC and EPS (Table 7). The confirmation experiments were carried out in 200-mL culture tubes using only A. maxima IFRPD1183. The results are shown in Fig. 3. Based on this figure, the maximum concentrations, productivities and the yields of C-PC and EPS are shown in Table 7 (experimental results). As shown in the table, the measured data generally agreed well with the expected (or predicted) data (Table 7). This validated the earlier identified optimal production conditions.

Fig. 3
figure 3

Culture profiles of A. maxima IFRPD 1183 under optimized conditions for: a C-PC production; and b EPS production. Profiles of biomass concentration, pH, residual nitrogen, C-PC concentration and EPS concentration are shown for the confirmation experiments carried out in 200-mL culture tubes

Full size image

Optimization of the culture conditions was of clear benefit. Under optimized conditions in 200-mL culture tubes, the volumetric rate of C-PC production was 85.5 mg L−1 d−1 (Table 7). In earlier studies of photoautotrophic production of C-PC by Arthrospira sp., the volumetric production rate has ranged from 10 to 70 mg L−1 d−1 [56, 59, 60], suggesting suboptimal production conditions. Generally, quite high production rates can be achieved for some Arthrospira metabolites under heterotrophic and mixotrophic culture conditions, compared to photoautotrophic conditions. This notwithstanding, in the present photoautotrophic study, the maximum observed volumetric rate of C-PC production was nearly fourfold higher than has been reported for an optimized mixotrophic culture of A. platensis [5].

For EPS, the highest observed concentration of 153.8 mg L−1 (Table 7) was attained in 14-days, giving a volumetric productivity of nearly 11 mg L−1 d−1. Other studies have reported a somewhat higher final concentration of 210 mg L−1 after 21-days of culture [24], but this corresponds to a productivity of 10 mg L−1 d−1, a value comparable to ours.

Biomass compositions

The composition of the biomass of A. maxima IFRPD1183 harvested at the end of the cultures (day 8 for C-PC production; day 14 for EPS production) of the validation experiments (see above) is shown in Table 8.

Table 8 Typical composition of dry biomassa of A. maxima IFRPD 1183
Full size table

The biomass produced under conditions optimized for the production of C-PC obviously had much more C-PC (Sample A with nearly fourfold more C-PC; Table 8) compared to the biomass produced under conditions optimized for producing EPS (Sample B, Table 8). (EPS is not shown in Table 8 as it is not an intracellular product.) The C-PC rich biomass also had nearly 2.5-fold more total protein compared to the biomass Sample B. This is expected as C-PC is a phycobiliprotein and such proteins make up nearly 60% of the total soluble protein in cyanobacterial cells [16]. The higher carbohydrate content (nearly twofold higher; Table 8) of the biomass Sample B from cultures optimized for EPS production is also expected as EPS are heteropolysaccharides and a lot of carbohydrate must be made for producing them. In addition, the carbohydrate-rich Sample B had been grown under a higher light intensity (nearly threefold higher than Sample A). High light levels are known to increase the amount of carbohydrate/glycogen in the cell [54, 55, 61].

In terms of total chlorophylls, the C-PC rich Sample A was nearly 2.4-fold richer in chlorophylls compared to Sample B. This too is consistent with expectations as Sample A had been grown under low light that favors accumulation of chlorophylls as well as C-PC [57]. More of these pigments are made under low light as the cell attempts to capture more of this essential resource for survival.

The composition of the cell is also affected by the supply of certain key nutrients. For example, under nitrogen limitation Arthrospira sp. accumulate lipids [62] so long as carbon is available and photosynthesis can occur. Similarly, phosphorus limitation has been shown to reduce the amount of A. platensis biomass produced, but greatly elevate the carbohydrate content of the biomass [63].

Samples A and B (Table 8) were grown at temperatures of 30 and 35 °C, respectively, and temperature can also affect biomass production and its composition [9]. This is because the rate of photosynthesis and, therefore, the rate of growth depend on temperature [64].

Raceway pond cultivation

The optimal conditions for maximized production of C-PC and EPS as established in 200-mL culture tubes were next used in the 30-L raceway pond batch cultures. Each batch ran for 14-days. A. maxima IFRPD1183 was the cyanobacterium used in all cases.

For the production of C-PC, the conditions were as follows: a light intensity of 68 µmol photons m−2 s−1, an initial pH of 10.0, a Zarrouk’s medium strength of 100%, a temperature of 30 °C and an aeration rate of 0.6 vvm using air mixed with 2% v/v CO2.

The following conditions were used for producing the EPS: a light intensity of 203 µmol photons m−2 s−1, an initial pH of 9.5, a 50% strength of Zarrouk’s medium, a temperature of 35 °C and an aeration rate of 0.2 vvm using air mixed with 2% v/v CO2.

The culture profiles of production of C-PC and EPS are shown in Figs. 4 and 5, respectively. In the raceway, the concentrations of the biomass, the C-PC (Fig. 4) and EPS (Fig. 5) were substantially less relative to the results obtained in 200-mL culture tubes (Fig. 3). Thus, the maximum biomass concentration was 7.9-fold lower during C-PC production and it was 2.4-fold lower during EPS production. Similarly, in the raceway the concentration of C-PC was nearly 11.5-fold lower (Fig. 4) compared to the data obtained in the culture tubes (Fig. 3a). In the raceway, the concentration of EPS was nearly 4.2-fold less (Fig. 5) relative to the culture tube data (Fig. 3b). All these effects were primarily related with the reduced production of the biomass in the raceway compared to the culture tubes.

Fig. 4
figure 4

Culture profiles of A. maxima IFRPD 1183 in the 30-L raceway pond under optimal conditions for the production of C-PC. Profiles of biomass, nitrogen, phosphorous, the pH and C-PC are shown

Full size image
Fig. 5
figure 5

Culture profiles of A. maxima IFRPD 1183 in the 30-L raceway pond under optimal conditions for the production of EPS. Profiles of biomass, nitrogen, phosphorous, the pH and EPS are shown

Full size image

Although the pH in the raceway cultures (Figs. 4, 5) varied, it was generally within the limits of 9 and 10 during C-PC production (Fig. 4) and this range is optimal for growth of Arthrospira spp. [11, 65]. During EPS production, the pH did exceed 10, but was always less than 11, conditions that Arthrospira spp. tolerate well. During EPS production the pH rose (Fig. 5) as the flow rate of gas mixture was low (=0.2 vvm) compared to the flow rate of 0.6 vvm used during C-PC production (Fig. 4). A low flow rate of the gas mixture combined with a rapid consumption of carbon dioxide (because of a high light level and high biomass concentration during EPS production, Fig. 5) apparently resulted in an insufficiency of dissolved carbon dioxide in the culture and hence the alkalinization.

As only a small amount of biomass was produced in the raceway, the key nutrients of nitrogen and phosphorous were barely consumed during production of C-PC (Fig. 4) and, therefore, the culture was free of any nutritional stress compared to the case of 200-mL culture tubes. This is further confirmed by the continuous growth of the biomass during almost the entire culture period in the raceway (Fig. 4). Similarly, during the production of EPS, there was barely any depletion of phosphorus, but nearly all nitrogen was consumed (Fig. 5). This suggests that because of a higher biomass concentration in the culture tubes, the culture medium in the tubes under optimal conditions for EPS production may have been fully depleted of nitrogen and a nitrogen stress likely contributed to enhanced production of EPS in the tubes. Comparing the raceway cultures in Figs. 4 and 5, the almost complete consumption of nitrogen in Fig. 5 and not in Fig. 4 is simply because of lower initial nitrogen concentration used in the production of EPS (Fig. 5). (The optimal conditions for producing EPS had been found to be a half-strength Zarrouk’s medium, as noted earlier.)

In practice, the data obtained under the optimized condition in 200-mL culture tubes represent a practical upper limit on the achievable production performance of A. maxima IFRPD1183. Any scaled up production system with a culture depth much greater than the 32 mm diameter of the culture tubes will have a reduced performance on account of a severely reduced light penetration in the culture. For a given concentration of biomass in a culture broth and a given incident light level at its surface, the light intensity declines exponentially with depth [66]. The depth of culture fluid in a full scale commercial raceway for biomass production is typically 0.25–0.30 m [66, 67]. Therefore, a commercial raceway culture will have a further reduction in productivity relative to the 0.07 m deep raceway used in the present work. In fact, the biomass concentration in a typically commercial raceway does not exceed about 0.5 g L−1 [66, 67]. In the present study, the concentrations in the raceway were higher (Figs. 4, 5) on account of a shallow depth compared to a commercial raceway.

In principle, tubular photobioreactors may be used to produce A. maxima IFRPD1183 and its metabolites at productivities comparable to the values found in this work in 200-mL culture tubes under the optimized operational conditions. Tubular photobioreactors [68,69,70,71] essentially consist of arrays of long transparent tubing with each tube being typically less than 0.1 m in diameter. Photobioreactors with tubing of nearly the same diameter as a 200-mL culture tube may also be used. The cost of artificial illumination may be reduced using light-emitting diodes (LEDs) [56, 72, 73] that consume much less power compared to conventional lights.

Conclusions

Optimal conditions for the production of C-PC and EPS were different. In 200-mL culture tubes the concentration of C-PC was maximized in a 14-day batch culture carried out under the following conditions: a light intensity of 68 µmol photons m−2 s−1 (a diurnal cycle of 16-h photoperiod and 8-h dark period), an initial pH 10, the full strength (100%) Zarrouk’s culture medium, an aeration rate of 0.6 vvm (air mixed with 2% v/v CO2) and a culture temperature of 30 °C. In contrast, a high light level and a low concentration of inorganic salts favored production of EPS. The EPS concentration in a 14-day batch culture was maximized under the following conditions: a light intensity of 203 µmol photons m−2 s−1 with the earlier specified light–dark cycle; an initial pH of 9.5; a 50% strength of Zarrouk’s medium; an aeration rate of 0.2 vvm (air mixed with 2% v/v CO2); and a temperature of 35 °C. In both cases, A. maxima IFRPD1183 appeared to be a better producer compared to A. platensis IFRPD 1208, although the differences between the two cyanobacteria were small. Raceway cultures performed poorly relative to the 200-mL culture tubes because of insufficiency of light and possibly a limited supply of carbon dioxide. For production under the recommended optimized culture conditions, the use of a tubular photobioreactor with 32 mm diameter tubes is likely to match the results obtained in 200-mL culture tubes. Any culture system with a depth >32 mm is likely to have a reduced performance compared to the results reported here.