Appl Microbiol Biotechnol DOI 10.1007/s00253-014-5757-9
BIOENERGY AND BIOFUELS
The effect of degree and timing of nitrogen limitation on lipid productivity in Chlorella vulgaris Melinda J. Griffiths & Robert P. van Hille & Susan T. L. Harrison
Received: 23 January 2014 / Revised: 4 April 2014 / Accepted: 5 April 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Improvements in lipid productivity would enhance the economic feasibility of microalgal biodiesel. In order to optimise lipid productivity, both the growth rate and lipid content of algal cells must be maximised. The lipid content of many microalgae can be enhanced through nitrogen limitation, but at the expense of biomass productivity. This suggests that a two-stage nitrogen supply strategy might improve lipid productivity. Two different nitrogen supply strategies were investigated for their effect on lipid productivity in Chlorella vulgaris. The first was an initial nitrogen-replete stage, designed to optimise biomass productivity, followed by nitrogen limitation to enhance lipid content (two-stage batch) and the second was an initial nitrogen-limited stage, designed to maximise lipid content, followed by addition of nitrogen to enhance biomass concentration (fed-batch). Volumetric lipid yield in nitrogen-limited two-stage batch and fed-batch was compared with that achieved in nitrogen-replete and nitrogenlimited batch culture. In a previous work, maximum lipid productivity in batch culture was found at an intermediate level of nitrogen limitation (starting nitrate concentration of 170 mg L−1). Overall lipid productivity was not improved by using fed-batch or two-stage culture strategies, although these strategies showed higher volumetric lipid concentrations than nitrogen-replete batch culture. The dilution of cultures prior to nitrogen deprivation led to increased lipid accumulation, indicating that the availability of light influenced the rate of lipid accumulation. However, dilution did not lead to increased
Electronic supplementary material The online version of this article (doi:10.1007/s00253-014-5757-9) contains supplementary material, which is available to authorized users. M. J. Griffiths : R. P. van Hille : S. T. L. Harrison (*) Centre for Bioprocess Engineering Research (CeBER), University of Cape Town, Rondebosch, Cape Town, South Africa 7701 e-mail: [email protected]
lipid productivity due to the resulting lower biomass concentration. Keywords Two-stage cultivation . Fed-batch . Algal biodiesel . Nutrient stress . Nitrogen limitation
Introduction Microalgae have several advantages over land-based crops as a potential source of oil for biodiesel. These include the fact that production does not compete for resources with traditional agriculture, their simple unicellular structure allows for a potentially high oil productivity, cultures can be maintained using brack or salt water, nutrients from wastewater and waste CO2 streams, and there is the potential for coproduction of a variety of products (Chisti 2007; Mata et al. 2010; Rodolfi et al. 2009; Sheehan et al. 1998). Although algal biodiesel is technically feasible (Miao and Wu 2006; Xiong et al. 2008), several challenges remain to be overcome for it to be both economically feasible and environmentally desirable. Among the most important of these is increasing lipid productivity in large-scale cultures (Borowitzka 1992; Lardon et al. 2009; Mata et al. 2010; Rodolfi et al. 2009; Sheehan et al. 1998). Algal lipid productivity is affected by the biological potential of the species and the environmental conditions it experiences. The choice of species is thus a key factor (Griffiths and Harrison 2009), along with the design of the culture system to provide the correct conditions (Rodolfi et al. 2009). In microalgae, structural lipids are a primary, growth-associated product. However, storage lipids such as triacyglycerol (TAG), the most suitable class of lipid for biodiesel production (Stephenson et al. 2010), are a secondary metabolite usually formed in large quantities only under adverse growing conditions (Klok et al. 2013). While there has been some success in engineering Cyanobacteria to excrete fatty acids (Liu et al.
Appl Microbiol Biotechnol
2011), lipid storage vesicles are generally contained within cells, up to a maximum proportion of about 60 % DW in Chlorella vulgaris (Griffiths et al. 2012). TAG is therefore a biomass-dependent but nongrowth-associated product, and the product yield is constrained by the biomass concentration. Several environmental factors such as temperature (Converti et al. 2009), light (Rodolfi et al. 2009) and the availability of nutrients such as phosphate, nitrate and silicate (Shifrin and Chisholm 1981) are known to affect microalgal lipid productivity. Nitrogen (N) limitation is the most frequently reported method of enhancing algal lipid content. It is easy to manipulate and has a reliable and strong influence on lipid content in many species (Chelf 1990; Rodolfi et al. 2009; Shifrin and Chisholm 1981). However, N limitation also decreases algal growth rate and final biomass concentration, which may offset the increase in lipid content in terms of overall lipid productivity. As the two key factors in lipid productivity (biomass productivity and lipid content) are maximised under opposite conditions of N availability, this suggests that a two-stage N-supply strategy might be advantageous. For production of secondary microalgal products, such as carotenoids or storage lipids, the use of two production stages to enhance yield has been proposed by several authors (BenAmotz 1995; Hsieh and Wu 2009; Huntley and Redalje 2006; Richmond 2004; Sheehan et al. 1998, Stephenson et al. 2010). For lipid production, the first stage is usually designed to optimise growth, while the second stage provides N-limiting conditions that retard growth and encourage product synthesis. However, the opposite strategy of lipid accumulation followed by cell division could also be employed. The two different strategies are as follows: 1. Optimisation of biomass productivity, then lipid content: an initial N-replete stage, designed to produce maximum biomass concentration as quickly as possible, followed by a N-limited stage, aimed to induce a rapid increase in lipid content while retaining a high biomass concentration. N can either be allowed to deplete naturally in the culture medium (N-limited batch culture) or it can be removed suddenly by transfer of cells to different media (two-stage batch culture). 2. Optimisation of lipid content, then biom ass concentration: initial growth in N-limited medium, designed to produce biomass with a high lipid content, followed by feeding of N in small amounts to allow production of additional biomass, while retaining a high lipid content (fed-batch culture). There is currently no consensus on which N-limitation strategy is optimal for maximising lipid productivity in microalgae. Several studies have reported that N-limited batch culture, at an optimal intermediate starting N concentration,
optimises the tradeoff between growth limitation and lipid accumulation and enhances lipid productivity relative to Nsufficient or N-deficient culture (Griffiths et al. 2014; Hsieh and Wu 2009; Lv et al. 2010; Stephenson et al. 2010). There are conflicting reports as to whether other N-limited culture regimes, such as two-stage, fed-batch, or continuous culture, lead to higher lipid productivities. Pruvost et al. (2009) reported that two-stage batch culture (strategy one) did not improve lipid productivity relative to N-sufficient continuous culture with Neochloris oleoabundans. Similarly, San Pedro et al. (2013) found identical lipid productivities for two-stage batch culture and N-sufficient continuous culture at the optimal dilution rate with Nannochloropsis gaditana, although the fatty acid profile and neutral lipid content were improved in two-stage culture. Takagi et al. (2000) compared fed-batch cultures (strategy two) of Nannochloris sp. to a batch culture started with an equivalent total amount of N. The fed-batch cultures reached the same biomass concentration but had a significantly higher lipid content (50.9 and 51.5 % compared with 31 %). However, the greater lipid content may have been due to the longer culture time in the fed-batch (500 h) compared with the batch culture 200 h). Although not quantified in the study, the longer cultivation time, despite the higher lipid content, would have resulted in a lower overall lipid productivity for the fed-batch cultures compared with the batch. Similarly, Hsieh and Wu (2009), working with a marine Chlorella sp. under urea limitation, reported that fed-batch cultivation did not increase lipid productivity (0.123 g L−1 day−1) significantly compared with N-limited batch culture at the optimal starting urea concentration (0.124 g L−1 day−1). However, they report that a semicontinuous harvesting process where a portion of the culture was removed, and urea renewed at a low concentration (0.025 g L−1) each time the culture achieved early stationary phase, led to a slightly higher maximum lipid productivity (0.139 g L−1 day−1). Several authors have mentioned a potential link between light availability and lipid accumulation under N-limited conditions. In algal cultures, light availability is influenced by culture density. Algal cells self-shade and, in dense cultures, light penetration may be limited to a fraction of the total culture volume closest to the reactor surface. Stephenson et al. (2010), working with C. vulgaris, investigated the effect of altering the cell density of cultures upon transfer to Ndeficient medium and concluded that there was an optimal cell density for lipid production. Su et al. (2011) also report an optimal biomass density for Nannochloropsis oculata at the beginning of the N-limitation stage and concluded that irradiance exhibits a significant influence on lipid production. Klok et al. (2013), working with N. oleoabundans under continuous culture in a turbidostat, at various N-supply and N-dilution rates, showed that TAG productivity could be increased, relative to N-replete continuous culture, through N limitation.
Appl Microbiol Biotechnol
However, this was coupled with a lower biomass yield on light. The authors hypothesised that it should be possible to create an optimal energy imbalance through reduced nutrient supply at constant light input, which would induce TAG accumulation while allowing cell division to continue. No comparison between different N-feeding strategies (Nlimited batch, two-stage batch and fed-batch) for the production of microalgal lipid has been carried out. The present work aimed to investigate the effect of degree and timing of N limitation on the lipid yield and productivity achieved in C. vulgaris cultivated in two-stage batch culture and fedbatch culture and to compare the potential of these different culture strategies. A further aim was to investigate the effect of culture density (and hence light availability) on lipid accumulation.
Methods All cultures of C. vulgaris UTEX 395 were grown in airlift photobioreactors under the same conditions used by Griffiths et al. (2014). The glass and steel reactors were 60 cm high with an external diameter of 10 cm, a draft tube of 5 cm diameter and a working volume of 3.2 L. The medium was 3 N BBM (Bold 1949), with the nitrate concentration manipulated in each experiment as detailed below. Air enriched with 0.29 % CO2 was sparged at 2 L min−1, resulting in a circulation time of approximately 7 s and an overall mass transfer coefficient of 0.0094±0.00026 s−1 (Langley et al. 2012). Continuous illumination (250 μmol m−2 s−1 at the reactor surface, measured using an L-COR light meter, model LI250) was provided by three cool white 18 W fluorescent light bulbs (Osram). Culture temperature was monitored daily and remained constant at 25 °C±1 °C. Starter cultures were grown in 500 mL glass bottles in N-sufficient media for 7 to 10 days before being used to inoculate the airlift reactors at a starting concentration of 0.05 g L−1. Sterile, distilled water was added daily to replace that lost to evaporation. Antifoam (20 μL; Antifoam 204, Sigma-Aldrich) was added to each reactor to reduce foaming. Cultures were grown for 20 days. Batch culture C. vulgaris was inoculated into reactors with an initial media nitrate concentration of 1,200 mg L−1 (N-replete batch culture), 170 mg L−1 (N-limited batch culture), or 40 mg L−1 (very N-limited batch culture) and allowed to exhaust the nitrate naturally. Two-stage batch culture C. vulgaris was cultured under N-replete conditions (starting nitrate concentration of 1,500 mg L−1) for 6 days. On day 6,
biomass was harvested by centrifugation at 1,520×g for 10 min. The biomass was resuspended to the original volume (3.2 L) in nitrate-free medium or medium containing 170 mg L−1 nitrate. The starting nitrate concentration of 170 mg L−1 was chosen because it produced the highest volumetric lipid concentration in N-limited batch culture in previous work (Griffiths et al. 2014). Cultivation was continued in the original reactor to day 20. Fed-batch culture Under the growth conditions described above, two different starting nitrate concentrations and four different nitrate feeding regimes were tested. The two starting nitrate concentrations (40 and 170 mg L−1) were chosen as these were the concentrations that led to the most rapid lipid accumulation and highest overall lipid productivity respectively, in batch culture in previous work (Griffiths et al. 2014). Nitrate depletion in the media was found to occur within 1.2 and 2.2 days in the 40 and 170 mg L−1 cultures, respectively. Additional nitrate was fed as a concentrated aliquot in the make-up water, replacing that lost to evaporation, hence the volume of the cultures was not altered. In the first and second experiments, concentrated nitrate solution was added, to result in a final nitrate concentration in the reactors of either 40 or 80 mg L−1 (assuming no residual nitrate), every second day, between days 4 and 10. In the third and fourth experiments, the nitrate concentration in the reactors was adjusted (assuming no residual nitrate) to 40 mg L−1 either every 4 days (between days 4 and 16) or every 8 days (on days 8 and 16) (Table 1). Diluted cultures Previously, Griffiths et al. (2014) found that cultures with a higher starting nitrate concentration (e.g. 420 mg L−1 nitrate) attained a higher biomass concentration and accumulated lipid more slowly than those with a lower starting nitrate concentration. This suggested that the lower availability of light in Table 1 Nitrate feeding regimes for fed-batch cultures Fed-batch experiment
Starting nitrate concentration (mg L−1)
Nitrate addition (mg nitrate L−1 reactor volume)
Feeding frequency (days)
1 1 2 2 3 3 4 4
40 40 170 170 40 40 170 170
40 80 40 80 40 40 40 40
2 2 2 2 4 8 4 8
Appl Microbiol Biotechnol
dense algal cultures may limit lipid accumulation. To test this, cultures were diluted to different biomass concentrations on day 5. The more diluted cultures should experience less mutual shading and a higher light intensity. C. vulgaris was grown for 5 days under N-replete conditions. Biomass was harvested by centrifugation at 1,520×g for 10 min and resuspended in fresh, N-free medium, to volumes one, two and three times the original culture volume, yielding cultures that were undiluted, or diluted 1:1 and 1:2 with respect to the original culture and grown for a further 15 days. Assays The nitrate concentration in the media was monitored daily by UV spectrophotometry. Filtered samples were diluted to less than 12 mg L−1 nitrate. Optical density (OD) at 220 nm was measured in a quartz cuvette and quantified using a nitrate standard curve (Clesceri et al. 1998). Biomass was measured by OD at 750 nm every day and by dry weight (DW) every third day according to Griffiths et al. (2011). Individual calibration curves of OD as a function of DW were constructed for each culture and used to calculate DW from OD. Cell lipid content (measured as total fatty acid content) was quantified daily by direct transesterification and gas chromatography (Griffiths et al. 2010). Calculations Volumetric lipid concentration (PVOL) was calculated as the product of biomass concentration (X) and lipid content (P). Lipid productivity (QP) was calculated as PVOL at each time point divided by the time from inoculation. Maximum parameters were defined as the highest value reached within the 20day culture period. Yield of lipid on nitrate was calculated as the PVOL at the end of the culture period divided by the total amount of nitrate taken up by the cells during cultivation. In the case of the diluted cultures, the amount of consumed nitrate was divided by the dilution factor to account for the cells that were discarded.
Results Strategy 1: two-stage batch culture (biomass then lipid) C. vulgaris was grown under N-replete conditions for 6 days. On day 6, biomass was harvested and resuspended in medium containing either 0 or 170 mg L−1 nitrate. Cultivation was continued to day 20. Cultures in which the nitrate concentration had been reduced to zero or 170 mg L−1 continued to accumulate biomass up to day 20, although at a more linear rate than the equivalent nutrient replete batch culture (starting nitrate concentration 1,200 mg L−1 with no transfer, dashed
line, Fig. 1a). Maximum biomass concentrations of 1.9 and 2 g L−1 were reached in the 0 and 170 mg L−1 nitrate two-stage cultures, respectively (Fig. 1a), compared with 2.4 g L−1 in the N-replete batch culture. In N-replete batch culture, the fatty acid content remained between 10 and 12 % throughout the culture period (Fig. 1b). The fatty acid content of the twostage cultures increased slowly from 11 % at day 6 to 28 % (0 mg L−1) and 21 % DW (170 mg L−1) at day 20 (Fig. 1b). In comparison, the N-limited batch culture with an initial nitrate concentration of 170 mg L−1 reached a lipid content of 50– 60 %. The volumetric lipid concentration (PVOL) of both twostage cultures increased linearly over the culture period to a maximum of 525 and 434 mg L−1 for the 0 and 170 mg L−1 nitrate cultures respectively at day 20. This is compared with a maximum PVOL of 281 mg L−1 in N-replete batch culture and 790 mg L−1 in the N-limited batch culture with the highest lipid productivity (starting nitrate concentration of 170 mg L−1, Fig. 1c). Initially, both two-stage cultures had a lower average lipid productivity (QP) than the N-replete batch culture, but from day 10, this situation was reversed (Fig. 1d). The nitrate limited culture with a starting concentration of 170 mg L−1 had, on average, double the lipid productivity of the two-stage and the N-replete cultures. Strategy 2: fed-batch culture (lipid then biomass) In the first fed-batch experiment, C. vulgaris cultures were grown in medium containing 40 mg L−1 nitrate and fed with an additional 40 or 80 mg L−1 nitrate every second day between days 4 and 10. Results are compared with those of single-stage batch culture with 40 mg L−1 nitrate (dashed lines) and 170 mg L−1 nitrate (solid lines). Biomass concentration increased during and after N feeding and was greater in the culture fed with a higher concentration of nitrate (Fig. 2a). The maximum biomass concentrations reached (1.3 and 1.6 g L−1 in the cultures fed 40 and 80 mg L−1 nitrate, respectively) were over two to three times that in N-limited batch culture with 40 mg L−1 nitrate (0.5 g L−1). Lipid content was decreased by nitrate supplementation. Fed-batch cultures began accumulating lipid at the same rate as the batch culture, but lipid content decreased upon N feeding, remaining at an average of 24 and 14 % DW in the cultures fed 40 and 80 mg L−1, respectively, until feeding was stopped at day 10. Once N-feeding ceased, lipid content increased slowly again but did not reach the high lipid content (max. 60 % DW) of the batch cultures (Fig. 2b). As a result of the lower lipid content, the volumetric lipid concentration and the average lipid productivity of the fed-batch cultures was initially lower than that of the batch cultures. Once feeding ceased, the lipid yield and productivity of the fed-batch cultures was greater than that of the 40 mg L−1 nitrate batch culture, due to the higher biomass concentration but lower than that of the 170 mg L−1 nitrate batch culture (Fig. 2c, d).
Appl Microbiol Biotechnol
Fig. 1 a Biomass (markers) and media nitrate concentration (lines with markers), b lipid content, c volumetric lipid concentration (PVOL) and d average lipid productivity (QP) over 20 days in C. vulgaris grown initially under N-replete conditions, harvested on day 6 (indicated by a vertical line) and resuspended in fresh media with 0 (filled triangles) or
170 mg L−1 (filled squares) nitrate before continued cultivation. Results are compared with those of N-replete batch culture (starting nitrate concentration, 1,200 mg L−1, filled diamonds) and N-limited batch culture (starting nitrate concentration, 170 mg L−1, solid line)
In the second experiment, the same feeding regimes were carried out in cultures with a starting nitrate concentration of 170 mg L−1. Again, the final biomass concentration was higher in fed-batch cultures, although the growth rate between days 2 and 8 was slower in fed-batch than 170 mg L−1 batch culture (Fig. 3a). In contrast to the first experiment, feeding with 80 mg L−1 showed no improvement over 40 mg L−1. Again, nitrate feeding retarded lipid accumulation, with 80 mg L−1 having a greater effect than 40 mg L−1 (Fig. 3b). Volumetric lipid productivity in fed-batch cultures was lower than the 170 mg L−1 batch culture, due to the lower lipid content (Fig. 3c). Lipid productivity in the fed-batch cultures was lower during the period of nitrate feeding (days 4 to 10) and increased thereafter as the lipid content increased (Fig. 3d). The lipid productivity of the culture fed with 40 mg L−1 was greater than that fed 80 mg L−1 every 2 days. In the third experiment, started at 40 mg L−1 nitrate, feeding of 40 mg L−1 nitrate was carried out at two different intervals: every 4 days and every 8 days. Cultures with more infrequent feeding showed a more gradual increase in biomass (Fig. 4a). Growth rate increased after feeding (days 8 and 16) in the culture fed every 8 days. Lipid accumulation was retarded by nitrate feeding, but to a lesser degree the less frequent the
feedings (Fig. 4b). In the culture fed every 4 days, lipid content remained constant (between 34 and 38 % DW) throughout the period of nitrate feeding. In the culture fed every 8 days, lipid content decreased for 4 days after feeding at day 8, but then increased again to the same level as the batch culture. Upon feeding at day 16, the lipid content of the culture decreased as biomass concentration increased. Volumetric lipid productivity was greater in the culture fed every 4 days (621 mg L−1) than that fed every 8 days (574 mg L−1) or 2 days (490 mg L−1; Fig. 2c), but still less than that achieved in batch culture at a starting nitrate concentration of 170 mg L−1 (Fig. 4c). The cultures fed every 4 or 8 days had a more constant lipid productivity (Fig. 4d) relative to those fed every 2 days, which showed a sharp dip in productivity during additional nitrate feeding (e.g. Fig. 1d). The average lipid productivity was greater than the 40 mg L−1 batch culture but less than the 170 mg L−1 batch culture. In the fourth experiment, the same feeding regimes were repeated with a starting nitrate concentration of 170 mg L−1. In these cultures, feeding with 40 mg L−1 nitrate every 4 or 8 days did not have a significant effect on biomass concentration (maximum biomass concentration was 1.7 and 1.6 g L−1 in the cultures fed every 4 and 8 days, respectively, as opposed to
Appl Microbiol Biotechnol
Fig. 2 a Biomass (markers) and media nitrate concentration (lines), b lipid content, c volumetric lipid concentration (PVOL) and d average lipid productivity (QP) over 20 days in C. vulgaris grown under fed-batch conditions, with starting nitrate concentration of 40 mg L−1, fed either
40 mg L−1 (filled triangles) or 80 mg L−1 (filled squares) nitrate on days 4, 6, 8 and 10. The results of single-stage batch culture with 40 mg L−1 nitrate (dashed line) and 170 mg L−1 nitrate (solid line) are shown for comparison
1.5 g L−1 in the normal batch culture) (Fig. 5a). Lipid accumulation, however, was retarded and remained relatively constant (between 19 and 31 % DW) over the period of feeding (Fig. 5b). In the culture fed every 8 days, lipid content increased between feeds, but to a lesser extent than in the 40-mg L−1 starting nitrate culture (Fig. 4b). As a result of the lower lipid content, the volumetric lipid concentration of the fed-batch cultures was lower than that of the 170mg L−1 batch culture (Fig. 5c). The lipid productivity of the fed-batch cultures started at 170 mg L−1 nitrate was initially lower than those started at 40 mg L−1, but reached a constant productivity earlier than those fed more frequently (Fig. 5d). Figure 6 plots the biomass and volumetric lipid concentrations for all the fed-batch experiments, as well as N-limited batch cultures with starting nitrate concentrations of 40 and 170 mg L−1, in order of the total amount of nitrate added to the culture over the 20-day cultivation period. Below a critical N level (approximately 200 mg L−1 nitrate), the total biomass concentration was limited by the N availability, and the lipid content of the biomass was relatively high. Above this N level, the biomass and PVOL achieved were similar, regardless of
when the N was added. Lipid content decreased with increasing N fed. Effect of light availability The biomass concentration after centrifugation and resuspension in N-free medium was 1.1, 0.7 and 0.5 g L−1 for the undiluted culture and cultures diluted 1:1 and 1:2, respectively. All three cultures continued to increase in biomass at similar rates, reaching maximum biomass concentrations of 2.3, 1.8 and 1.4 g L−1, respectively (Fig. 7a). The more dilute the culture, the faster the lipid accumulation rate. The most dilute culture (1:2) reached the highest lipid content (39 % DW), followed by the culture diluted 1:1 (33 % DW) and then the undiluted culture (25 % DW) (Fig. 7b). Although the diluted cultures achieved a higher lipid content, this was offset by their lower biomass concentration. As a result, the volumetric lipid yield of all three cultures was similar (reaching a maximum of 565, 555 and 545 mg L−1 in undiluted, 1:1 and 1:2, respectively) (Fig. 7c). This was less than the PVOL of the best N-limited batch culture (starting nitrate concentration, 170 mg L−1). The lipid productivity (QP) was initially lower
Appl Microbiol Biotechnol
Fig. 3 a Biomass (markers) and media nitrate concentration (lines), b lipid content, c volumetric lipid concentration (PVOL) and d average lipid productivity (QP) over 20 days in C. vulgaris grown under fed-batch conditions, with starting nitrate concentration of 170 mg L−1, fed either
40 mg L−1 (filled triangles) or 80 mg L−1 (filled squares) nitrate on days 4, 6, 8 and 10. The results of single-stage batch culture with 40 mg L−1 nitrate (dashed line) and 170 mg L−1 nitrate (solid line) are shown for comparison
in the diluted cultures after dilution but from day 12 onwards, they matched the QP of the undiluted culture (Fig. 7d).
Discussion Strategy 1: two-stage batch culture
Comparison In order to determine the optimum culture regime in terms of lipid productivity, the best results obtained in each of the different strategies were compared on the basis of maximum biomass (X), lipid content (P), volumetric lipid concentration (PVOL) and lipid productivity (QP) as well as yield of lipid on nitrate (Table 2). The best culture strategy, in terms of PVOL, QP and yield on nitrate, was N-limited batch culture with 170 mg L−1 nitrate. The second best, in terms of PVOL, was fed-batch with the feeding strategy of 40 mg L−1 every 4 days. Two-stage batch culture improved the lipid productivity the least of the N-limitation strategies tested. Although none of the two-stage or fed-batch cultivation strategies improved on the lipid productivity of N-limited batch culture, they were all better than N-replete batch culture, which was the worst strategy in terms of PVOL, QP and yield of lipid on nitrate, despite having the highest final biomass concentration.
In two-stage batch culture, the goal of the first stage was to optimise biomass productivity, followed by a second, Nstarvation stage to increase lipid content. After the sudden removal or reduction of nitrate in the medium on day 6, C. vulgaris was able to continue growing for up to a further 10 days and accumulated lipid only gradually. Lipid accumulation is a function of the N content of the cells (Griffiths et al. 2014). The continued cell growth and lack of lipid accumulation after N removal may be supported by intracellular stores of N (Li et al. 2008; Lourenço et al. 2004). During the initial N-replete phase, the two-phase cultures consumed approximately 500 mg L−1 nitrate, almost three times the amount available to the batch culture with a starting nitrate concentration of 170 mg L−1, but the same amount of biomass was produced in each (approximate;y 0.85 g L−1), therefore the cells in two-phase culture had higher intracellular N reserves upon nitrate depletion. By utilising reserves of inorganic N (NO3, NO2 and NH4), as well as breaking down N-containing
Appl Microbiol Biotechnol
Fig. 4 a Biomass (markers) and media nitrate concentration (lines), b lipid content, c volumetric lipid concentration (PVOL) and d average lipid productivity (QP) over 20 days in C. vulgaris grown under fed-batch conditions, with starting nitrate concentration of 40 mg L−1, fed
40 mg L−1 nitrate every 4 days (filled triangles) or 8 days (filled squares). The results of single-stage batch culture with 40 mg L−1 nitrate (dashed line) and 170 mg L−1 nitrate (solid line) are shown for comparison
macromolecules such as protein, nucleic acid and pigment, cells may be able to delay the major metabolic shifts associated with severe N limitation. This is a disadvantage from the point of view of optimising lipid productivity, as it causes a delay in lipid accumulation after N removal. The concentration of nitrate in the first, N replete, phase was varied to see whether this influenced the growth and lipid content after N removal. It was hypothesised that the initial N concentration could affect the amount of N stored and hence the physiology of the cells after N removal. It was found that the concentration of nitrate during the initial N-replete growth phase could vary between 500 and 1,000 mg L−1 without effect (Fig. S1 in the Electronic supplementary material). It was hypothesised that the shear stress involved in centrifugation of the two-stage batch cultures may have had an effect of the growth of the cells. This was tested by comparing cultures in which the nitrate was naturally exhausted after 5 days, with those in which the culture medium was changed from N-replete to N-free media via centrifugation on day 5. The growth of the centrifuged culture was not adversely affected (Fig. S2 in the Electronic supplementary material). The lipid yield and productivity of the culture transferred from N-replete to N-free media was higher than that transferred to media with an intermediate level of N (170 mg L−1
nitrate) (Table 2). This is in contrast to the report of Stephenson et al. (2010) who investigated the effect of various nitrate and cell concentrations on lipid productivity after transfer to N-limited medium. It was found that maximum average lipid productivity in C. vulgaris (46 mg L−1 day−1) was achieved when cultures were initially grown in N-sufficient medium, and then transferred to medium with an intermediate nitrate concentration (200 mg L−1). This (and the higher lipid productivity reported) could be due to different culture conditions, particularly the higher illumination levels used by Stephenson et al. (2010). Strategy 2: fed-batch culture Fed-batch cultures were inoculated at a limiting nitrate concentration, in order to maximise lipid content, but at a low biomass concentration. Small quantities of nitrate were added at various time points in an attempt to enhance biomass concentration while maintaining a high lipid content. The amount and timing of N feeding in fed-batch culture had a significant effect on biomass productivity and lipid accumulation during the culture period, although, over the range tested here, not on the final product yield. A higher starting nitrate concentration (170 rather than 40 mg L−1) decreased
Appl Microbiol Biotechnol
Fig. 5 a Biomass (markers) and media nitrate concentration (lines), b lipid content, c volumetric lipid concentration (PVOL) and d average lipid productivity (QP) over 20 days in C. vulgaris grown under fed-batch conditions, with starting nitrate concentration of 170 mg L−1, fed
40 mg L−1 nitrate every 4 days (filled triangles) or 8 days (filled squares). The results of single-stage batch culture with 40 mg L−1 nitrate (dashed line) and 170 mg L−1 nitrate (solid line) are shown for comparison
the response of the cells to additional nitrate fed (Figs. 3 and 5), but the overall productivity was enhanced by a higher biomass concentration. Feeding more nitrate (80 rather than 40 mg L−1) resulted in greater biomass accumulation but also a greater decrease in lipid content (Figs 2 and 3). Longer gaps between nitrate feeding (every 4 or 8 days instead of every 2),
decreased the lipid content less and increased the biomass concentration more gradually (Figs. 2 and 4). A low starting nitrate concentration (40 mg L−1), combined with feeding of a small amount of nitrate (40 mg L−1) every 4 days, maintained a steady increase in biomass concentration without compromising the lipid content as severely. There was a
Fig. 6 Summary of maximum volumetric lipid concentration (black bars) and total biomass concentration (sum of black and grey bars) achieved during fedbatch experiments with different starting nitrate concentrations (40 or 170 mg L−1, fourth row), fed either 40 or 80 mg L−1 nitrate (third row) at different time intervals (every 2, 4 or 8 days, second row). The experiments are shown in order of total amount of nitrate added to the culture (mg L−1, first row)
Appl Microbiol Biotechnol
Fig. 7 a Biomass, b lipid content, c volumetric lipid concentration (PVOL) and d average lipid productivity (QP) over 20 days in C. vulgaris grown initially under N-replete conditions, harvested on day 4.8 (indicated by a vertical line) and resuspended in fresh media undiluted (filled
triangles) and diluted 1:1 (filled squares) or 1:2 (filled diamonds) before continued cultivation. The results for single-stage batch culture with 170 mg L−1 nitrate (solid line) are shown for comparison
critical total N supply (approximate;y 200 mg L−1 nitrate over 20 days) below which growth was constrained by availability of N and lipid content was increased. Above 200 mg L−1 total nitrate supply over 20 days, biomass and lipid production was similar regardless of the timing of N feeding. This suggests that the critical variable in achieving high lipid productivity in fed-batch culture is the total amount of N fed and not the quantity or timing of feeding. The N-limited batch culture fed
170 mg L−1 nitrate gave the highest PVOL. This starting nitrate concentration is just below the critical total nitrate feed, resulting in a similar final biomass concentration, but enhanced lipid content. Takagi et al. (2000) also reported that the timing of nitrate feeding (during log phase, linear phase or stationary phase) during fed-batch had negligible effect on the lipid content of Nannochloris sp., although biomass concentration was higher
Table 2 Comparison of different culture regimes in terms of maximum biomass (X), lipid content (P), volumetric lipid concentration (PVOL) and lipid productivity (QP) as well as yield of lipid per gram nitrate used
N-replete batch (starting NO3, 1,200 mg L−1) N-limited batch (starting NO3, 170 mg L−1) Two-stage batch (1,500 to 0 mg L−1) Two-stage batch (1,500 to 170 mg L−1) Fed-batch 40 mg L−1 every 4 days (starting 40 mg L−1) Fed-batch 40 mg L−1 every 4 days (starting 170 mg L−1) Light experiment (undiluted) Light experiment (diluted 1:1) Light experiment (diluted 1:2)
Max X (g L−1)
Max P (% DW)
Max PVOL (mg L−1)
Max QP (mg L−1 day−1)
Yield of lipid on nitrate (g g−1)
2.4 1.5 1.8 2.0 1.5 1.7 2.3 1.8 1.4
12 54 26 20 41 35 25 33 39
281 790 477 391 621 588 565 555 545
25 41 25 21 33 34 30 30 29
0.2 4.6 1.1 0.7 3.1 1.8 1.3 2.6 3.9
Appl Microbiol Biotechnol
with later feed times. They found that feeding of small amounts of nitrate (0.9 mM, equivalent to 56 mg L−1) during the log phase increased the biomass concentration to twice that obtained without feeding, while lipid content was maintained at a high level compared with N-replete batch culture. They did not report lipid productivities, however, calculating these from the final biomass concentration and lipid content, PVOL was not improved compared with N-limited batch culture at starting nitrate concentration between 250 and 600 mg L−1 nitrate. The fact that none of the fed-batch strategies tested outperformed N-limited batch culture at the optimal starting nitrate concentration is also in agreement with Hsieh and Wu (2009), who reported that additional feeding of urea during the growth curve of a marine Chlorella sp. prolonged cell growth and increased the final biomass concentration, however, the higher the amount of urea fed, the lower the lipid content. They showed that small amounts of additional urea could enhance lipid productivity (up to 123 mg L−1 day-1), above that of cultures without additional feeding, but not above the maximum found in N-limited batch culture (124 mg L−1 day1 ) with an optimal starting urea concentration of 0.1 g L−1, equivalent to 205 mg L−1 nitrate. This value agrees closely with the critical total nitrate feed concentration found in this work (200 mg L−1). Effect of light availability Cultures diluted to a lower biomass concentration in N-free medium accumulated lipid more rapidly and to a greater degree than the undiluted culture. This supports the hypothesis that lipid accumulation is light limited in dense cultures. Rodolfi et al. (2009) also found that increasing the illumination supplied to a flat panel reactor enhanced both the biomass productivity and the lipid content of Nannochloropsis. Klok et al. (2013) reported that, under N-limited conditions, an increase in the average light intensity led to more and larger lipid bodies at the same N supply rate in N. oleoabundans. From light attenuation measurements in the airlift reactors at the biomass densities and incident light intensities used, it is assumed that all incident light was absorbed by both the diluted and undiluted cultures, yielding similar volumetric light absorption in each case. In the diluted cultures, more light would have been absorbed per cell, increasing specific lipid production, however, due to the identical volumetric light absorption, overall volumetric lipid production remained constant. These results support the work by Stephenson et al. (2010) who noted that cell density at the beginning of the second stage of two-stage N-limited culture affects the rate of lipid accumulation. Cells transferred to N-deficient media at a lower cell density accumulated more lipid; however, similarly to this work, due to lower cell concentrations, the overall productivity of cultures inoculated at lower cell density was
lower. Maximum productivity was achieved in the culture with the highest cell concentration. A single dilution at the time of media transfer did not improve lipid productivity in this work but enhancing light availability in other ways (e.g. a draw and feed strategy, increasing the overall ambient light intensity or enhancing mixing) may improve both biomass and lipid yield. Caution should be used when dramatically increasing light intensity, as the reduced pigment content found in N-limited cultures (Griffiths et al. 2014) could lead to enhanced susceptibility to photoinhibition. While the two-stage batch and fed-batch culture regimes were not fully optimised in this work, limiting the strength of conclusions that can be drawn, none of the multistage cultivation strategies tested were found to improve lipid productivity above that measured in N-limited (170 mg L−1) batch culture where the nitrate became exhausted naturally in the medium during the exponential growth phase. If two-stage batch culture were carried out where the second, nitratelimited, stage was initiated after exactly 170 mg L−1 nitrate was consumed in stage one, it is likely that the lipid productivity would be very similar to a batch culture with an initial nitrate concentration of 170 mg L−1. However, transfer of culture in any two-stage system would require additional media as well as energy input and labour costs for pumping and biomass harvesting, increasing the costs and environmental footprint of the process. Two-stage cultivation is therefore not recommended for microalgal biodiesel production, unless the conditions of the second stage are cheaper or less energy intensive to maintain, for example, in transfer from a closed reactor to an open pond. In designing conditions for the second stage, it should be borne in mind that cells require conditions of sufficient light and carbon supply for lipid formation. This implies that two-stage processes that do not provide optimal light conditions in the second, lipid accumulation stage will not be optimal. Ideally, reactor design and cultivation strategies should aim to maintain optimal light provision throughout the culture period. Hsieh and Wu (2009) showed that a strategy of semicontinuous cultivation, where 25 % or the culture was removed and 0.025 g L−1 urea added every day from days 3 to 7, produced a higher lipid productivity (139 mg L−1 day−1) than either batch or fed-batch without culture removal. The removal of culture may have facilitated light penetration, and hence lipid accumulation, by reducing cell density. Klok et al. (2013) have also shown that the TAG productivity of N. oleoabundans could be enhanced by N limitation under continuous culture in a turbidostat, relative to N replete continuous culture. Continuous or semicontinuous culture at the critical N-feed rate to balance biomass and lipid productivity, and high light intensities, may be a promising strategy for further investigation.
Appl Microbiol Biotechnol
Although lipid productivity has a direct effect on the economic viability of algal biodiesel production, a high productivity in itself is not sufficient to guarantee success. Other factors such as ease of cultivation, resistance to contamination, and effects on downstream processing should also be taken into account. Restriction of N supply may discourage contamination by preventing growth of other algal species; however, N-limited cultures may prove less robust in the outdoor environment. In addition, N-limited cells may have different densities or cell surface properties and result in lower final biomass concentrations, which could affect harvesting. In conclusion, all N-limited culture strategies improved the volumetric lipid yield relative to N-replete culture. However, transfer of cells from N-replete to N-limited media (two-stage batch) or gradual feeding of N (fed-batch) did not improve the lipid yield on nitrate over that of cultures allowed to exhaust a low level of nitrate (170 mg L−1) naturally (N-limited batch culture). The lower-than-expected lipid productivity in twostage culture was due to a delay in lipid accumulation, postulated to be due to the utilisation of intracellular N reserves, and a decreased rate of lipid accumulation, possibly due to light limitation in dense cultures. N-limited batch and fed-batch cultures showed the highest yield of lipid per gram nitrate used, potentially lowering the cost and environmental impact associated with the provision of fixed N.
Acknowledgments This work is based upon research supported by the South African National Energy Development Institute (SANEDI), the South African Research Chairs Initiative (SARChI) of the Department of Science and Technology, the National Research Foundation (NRF) and the Technology Innovation Agency (TIA). The financial assistance of these organisations is hereby acknowledged. Any opinion, finding and conclusion or recommendation expressed in this material is that of the authors and SANEDI, SARChI, TIA or the NRF do not accept any liability in this regard.
References Ben-Amotz A (1995) New mode of Dunaliella biotechnology: two-phase growth for β-carotene production. J Appl Phycol 7(1):65–68 Bold H (1949) The morphology of Chlamydomonas chlamydogama sp. nov. Bull Torrey Bot Club 76:101–108 Borowitzka MA (1992) Algal biotechnology products and processes— matching science and economics. J Appl Phycol 4:267–279 Chelf P (1990) Environmental control of lipid and biomass production in two diatom species. J Appl Phycol 2(2):121–129 Chisti Y (2007) Biodiesel from microalgae. Biotechnol Adv 25(3):294– 306 Clesceri LS, Greenberg AE, Eaton AD (1998) Standard methods for the examination of water and wastewater, 20th edn. American Public Health Association, Washington DC Converti A, Casazza AA, Ortiz EY, Perego P, Del Borghi M (2009) Effect of temperature and nitrogen concentration on the growth and lipid content of Nannochloropsis oculata and Chlorella vulgaris for biodiesel production. Chem Eng Process 6:1146–1151
Griffiths MJ, Harrison STL (2009) Lipid productivity as a key characteristic for choosing algal species for biodiesel production. J Appl Phycol 21(5):493–507 Griffiths MJ, van Hille RP, Harrison STL (2010) Selection of direct transesterification as the preferred method for assay of fatty acid content of microalgae. Lipids 45(11):1053–1060 Griffiths MJ, Garcin C, van Hille RP, Harrison STL (2011) Interference by pigment in the estimation of microalgal biomass concentration by optical density. J Microbiol Methods 85(2):119–123 Griffiths MJ, van Hille RP, Harrison STL (2012) Lipid productivity, settling potential and fatty acid profile of eleven microalgal species grown under nitrogen replete and limited conditions. J Appl Phycol 24(5):989–1001 Griffiths MJ, van Hille RP, Harrison STL (2014) The effect of nitrogen limitation on lipid productivity and cell composition in Chlorella vulgaris. Appl Microbiol Biotechnol 98(5):2345–2356 Hsieh C-H, Wu W-T (2009) Cultivation of microalgae for oil production with a cultivation strategy of urea limitation. Bioresour Technol 100(17):3921–3926 Huntley ME, Redalje DG (2006) CO2 mitigation and renewable oil from photosynthetic microbes: a new appraisal. Mitig Adapt Strateg Glob Chang 12(4):573–608 Klok AJ, Martens DE, Wijffels RH, Lamers PP (2013) Simultaneous growth and neutral lipid accumulation in microalgae. Bioresour Technol 134:233–243 Langley N, Harrison STL, van Hille RP (2012) A critical evaluation of CO2 supplementation to algal systems by direct injection. Biochem Eng J 68:70–75 Lardon L, Hélias A, Sialve B, Steyer JP, Bernard O (2009) Life-cycle assessment of biodiesel production from microalgae. Environ Sci Technol 43(17):6475–6481 Li Y, Horsman M, Wang B, Wu N, Lan CQ (2008) Effects of nitrogen sources on cell growth and lipid accumulation of green alga Neochloris oleoabundans. Appl Microbiol Biotechnol 81(4):629– 636 Liu X, Sheng J, Curtiss R (2011) Fatty acid production in genetically modified cyanobacteria. Proc Natl Acad Sci U S A 108(17):6899– 6904 Lourenço S, Barbarino E, Lavín P, Lanfer Marquez U, Aidar E (2004) Distribution of intracellular nitrogen in marine microalgae: calculation of new nitrogen-to-protein conversion factors. Eur J Phycol 39(1):17–32 Lv J-M, Cheng L-H, Xu X-H, Zhang L, Chen H-L (2010) Enhanced lipid production of Chlorella vulgaris by adjustment of cultivation conditions. Bioresour Technol 101(17):6797–6804 Mata TM, Martins AA, Caetano NS (2010) Microalgae for biodiesel production and other applications: a review. Renew Sustain Energy Rev 14(1):217–232 Miao X, Wu Q (2006) Biodiesel production from heterotrophic microalgal oil. Bioresour Technol 97(6):841–846 Pruvost J, Van Vooren G, Cogne G, Legrand J (2009) Investigation of biom ass and lipids production with Neochloris oleoabundans in photobioreactor. Bioresour Technol 100(23): 5988–5995 Richmond A (2004) Microalgal culture. Biotechnology and applied phycology. Blackwell Science, Oxford Rodolfi L, Chini Zittelli G, Bassi N, Padovani G, Biondi N, Bonini G, Tredici MR (2009) Microalgae for oil: strain selection, induction of lipid synthesis and outdoor mass cultivation in a low-cost photobioreactor. Biotechnol Bioeng 102(1):100–112 San Pedro A, Gonzalez-Lopez CV, Acién FG, Molina-Grima E (2013) Marine microalgae selection and culture conditions optimization for biodiesel production. Bioresour Technol 134:353–361 Sheehan J, Dunahay T, Benemann JR, Roessler P (1998) A look back at the U.S. Department of Energy's Aquatic Species Program:
Appl Microbiol Biotechnol Biodiesel from algae. Closeout report. National Renewable Energy Lab, Department of Energy, Golden, Report number NREL/TP-58024190, dated July 1998 Shifrin NS, Chisholm SW (1981) Phytoplankton lipids: interspecific differences and effects of nitrate, silicate and light-dark cycles. J Phycol 17:374–384 Stephenson AL, Dennis JS, Howe CJ, Scott SA, Smith AG (2010) Influence of nitrogen-limitation regime on the production by Chlorella vulgaris of lipids for biodiesel feedstocks. Biofuels 1(1): 47–58
Su C-H, Chien L-J, Gomes J, Lin Y-S, Yu Y-K, Liou J-S, Syu R-J (2011) Factors affecting lipid accumulation by Nannochloropsis oculata in a two-stage cultivation process. J Appl Phycol 23:903–908 Takagi M, Watanabe K, Yamaberi K, Yoshida T (2000) Limited feeding of potassium nitrate for intracellular lipid and triglyceride accumulation of Nannochloris sp. UTEX LB1999. Appl Microbiol Biotechnol 54(1):112–117 Xiong W, Li X, Xiang J, Wu Q (2008) High-density fermentation of microalga Chlorella protothecoides in bioreactor for microbiodiesel production. Appl Microbiol Biotechnol 78(1):29–36
BIOENERGY AND BIOFUELS
The effect of degree and timing of nitrogen limitation on lipid productivity in Chlorella vulgaris Melinda J. Griffiths & Robert P. van Hille & Susan T. L. Harrison
Received: 23 January 2014 / Revised: 4 April 2014 / Accepted: 5 April 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Improvements in lipid productivity would enhance the economic feasibility of microalgal biodiesel. In order to optimise lipid productivity, both the growth rate and lipid content of algal cells must be maximised. The lipid content of many microalgae can be enhanced through nitrogen limitation, but at the expense of biomass productivity. This suggests that a two-stage nitrogen supply strategy might improve lipid productivity. Two different nitrogen supply strategies were investigated for their effect on lipid productivity in Chlorella vulgaris. The first was an initial nitrogen-replete stage, designed to optimise biomass productivity, followed by nitrogen limitation to enhance lipid content (two-stage batch) and the second was an initial nitrogen-limited stage, designed to maximise lipid content, followed by addition of nitrogen to enhance biomass concentration (fed-batch). Volumetric lipid yield in nitrogen-limited two-stage batch and fed-batch was compared with that achieved in nitrogen-replete and nitrogenlimited batch culture. In a previous work, maximum lipid productivity in batch culture was found at an intermediate level of nitrogen limitation (starting nitrate concentration of 170 mg L−1). Overall lipid productivity was not improved by using fed-batch or two-stage culture strategies, although these strategies showed higher volumetric lipid concentrations than nitrogen-replete batch culture. The dilution of cultures prior to nitrogen deprivation led to increased lipid accumulation, indicating that the availability of light influenced the rate of lipid accumulation. However, dilution did not lead to increased
Electronic supplementary material The online version of this article (doi:10.1007/s00253-014-5757-9) contains supplementary material, which is available to authorized users. M. J. Griffiths : R. P. van Hille : S. T. L. Harrison (*) Centre for Bioprocess Engineering Research (CeBER), University of Cape Town, Rondebosch, Cape Town, South Africa 7701 e-mail: [email protected]
lipid productivity due to the resulting lower biomass concentration. Keywords Two-stage cultivation . Fed-batch . Algal biodiesel . Nutrient stress . Nitrogen limitation
Introduction Microalgae have several advantages over land-based crops as a potential source of oil for biodiesel. These include the fact that production does not compete for resources with traditional agriculture, their simple unicellular structure allows for a potentially high oil productivity, cultures can be maintained using brack or salt water, nutrients from wastewater and waste CO2 streams, and there is the potential for coproduction of a variety of products (Chisti 2007; Mata et al. 2010; Rodolfi et al. 2009; Sheehan et al. 1998). Although algal biodiesel is technically feasible (Miao and Wu 2006; Xiong et al. 2008), several challenges remain to be overcome for it to be both economically feasible and environmentally desirable. Among the most important of these is increasing lipid productivity in large-scale cultures (Borowitzka 1992; Lardon et al. 2009; Mata et al. 2010; Rodolfi et al. 2009; Sheehan et al. 1998). Algal lipid productivity is affected by the biological potential of the species and the environmental conditions it experiences. The choice of species is thus a key factor (Griffiths and Harrison 2009), along with the design of the culture system to provide the correct conditions (Rodolfi et al. 2009). In microalgae, structural lipids are a primary, growth-associated product. However, storage lipids such as triacyglycerol (TAG), the most suitable class of lipid for biodiesel production (Stephenson et al. 2010), are a secondary metabolite usually formed in large quantities only under adverse growing conditions (Klok et al. 2013). While there has been some success in engineering Cyanobacteria to excrete fatty acids (Liu et al.
Appl Microbiol Biotechnol
2011), lipid storage vesicles are generally contained within cells, up to a maximum proportion of about 60 % DW in Chlorella vulgaris (Griffiths et al. 2012). TAG is therefore a biomass-dependent but nongrowth-associated product, and the product yield is constrained by the biomass concentration. Several environmental factors such as temperature (Converti et al. 2009), light (Rodolfi et al. 2009) and the availability of nutrients such as phosphate, nitrate and silicate (Shifrin and Chisholm 1981) are known to affect microalgal lipid productivity. Nitrogen (N) limitation is the most frequently reported method of enhancing algal lipid content. It is easy to manipulate and has a reliable and strong influence on lipid content in many species (Chelf 1990; Rodolfi et al. 2009; Shifrin and Chisholm 1981). However, N limitation also decreases algal growth rate and final biomass concentration, which may offset the increase in lipid content in terms of overall lipid productivity. As the two key factors in lipid productivity (biomass productivity and lipid content) are maximised under opposite conditions of N availability, this suggests that a two-stage N-supply strategy might be advantageous. For production of secondary microalgal products, such as carotenoids or storage lipids, the use of two production stages to enhance yield has been proposed by several authors (BenAmotz 1995; Hsieh and Wu 2009; Huntley and Redalje 2006; Richmond 2004; Sheehan et al. 1998, Stephenson et al. 2010). For lipid production, the first stage is usually designed to optimise growth, while the second stage provides N-limiting conditions that retard growth and encourage product synthesis. However, the opposite strategy of lipid accumulation followed by cell division could also be employed. The two different strategies are as follows: 1. Optimisation of biomass productivity, then lipid content: an initial N-replete stage, designed to produce maximum biomass concentration as quickly as possible, followed by a N-limited stage, aimed to induce a rapid increase in lipid content while retaining a high biomass concentration. N can either be allowed to deplete naturally in the culture medium (N-limited batch culture) or it can be removed suddenly by transfer of cells to different media (two-stage batch culture). 2. Optimisation of lipid content, then biom ass concentration: initial growth in N-limited medium, designed to produce biomass with a high lipid content, followed by feeding of N in small amounts to allow production of additional biomass, while retaining a high lipid content (fed-batch culture). There is currently no consensus on which N-limitation strategy is optimal for maximising lipid productivity in microalgae. Several studies have reported that N-limited batch culture, at an optimal intermediate starting N concentration,
optimises the tradeoff between growth limitation and lipid accumulation and enhances lipid productivity relative to Nsufficient or N-deficient culture (Griffiths et al. 2014; Hsieh and Wu 2009; Lv et al. 2010; Stephenson et al. 2010). There are conflicting reports as to whether other N-limited culture regimes, such as two-stage, fed-batch, or continuous culture, lead to higher lipid productivities. Pruvost et al. (2009) reported that two-stage batch culture (strategy one) did not improve lipid productivity relative to N-sufficient continuous culture with Neochloris oleoabundans. Similarly, San Pedro et al. (2013) found identical lipid productivities for two-stage batch culture and N-sufficient continuous culture at the optimal dilution rate with Nannochloropsis gaditana, although the fatty acid profile and neutral lipid content were improved in two-stage culture. Takagi et al. (2000) compared fed-batch cultures (strategy two) of Nannochloris sp. to a batch culture started with an equivalent total amount of N. The fed-batch cultures reached the same biomass concentration but had a significantly higher lipid content (50.9 and 51.5 % compared with 31 %). However, the greater lipid content may have been due to the longer culture time in the fed-batch (500 h) compared with the batch culture 200 h). Although not quantified in the study, the longer cultivation time, despite the higher lipid content, would have resulted in a lower overall lipid productivity for the fed-batch cultures compared with the batch. Similarly, Hsieh and Wu (2009), working with a marine Chlorella sp. under urea limitation, reported that fed-batch cultivation did not increase lipid productivity (0.123 g L−1 day−1) significantly compared with N-limited batch culture at the optimal starting urea concentration (0.124 g L−1 day−1). However, they report that a semicontinuous harvesting process where a portion of the culture was removed, and urea renewed at a low concentration (0.025 g L−1) each time the culture achieved early stationary phase, led to a slightly higher maximum lipid productivity (0.139 g L−1 day−1). Several authors have mentioned a potential link between light availability and lipid accumulation under N-limited conditions. In algal cultures, light availability is influenced by culture density. Algal cells self-shade and, in dense cultures, light penetration may be limited to a fraction of the total culture volume closest to the reactor surface. Stephenson et al. (2010), working with C. vulgaris, investigated the effect of altering the cell density of cultures upon transfer to Ndeficient medium and concluded that there was an optimal cell density for lipid production. Su et al. (2011) also report an optimal biomass density for Nannochloropsis oculata at the beginning of the N-limitation stage and concluded that irradiance exhibits a significant influence on lipid production. Klok et al. (2013), working with N. oleoabundans under continuous culture in a turbidostat, at various N-supply and N-dilution rates, showed that TAG productivity could be increased, relative to N-replete continuous culture, through N limitation.
Appl Microbiol Biotechnol
However, this was coupled with a lower biomass yield on light. The authors hypothesised that it should be possible to create an optimal energy imbalance through reduced nutrient supply at constant light input, which would induce TAG accumulation while allowing cell division to continue. No comparison between different N-feeding strategies (Nlimited batch, two-stage batch and fed-batch) for the production of microalgal lipid has been carried out. The present work aimed to investigate the effect of degree and timing of N limitation on the lipid yield and productivity achieved in C. vulgaris cultivated in two-stage batch culture and fedbatch culture and to compare the potential of these different culture strategies. A further aim was to investigate the effect of culture density (and hence light availability) on lipid accumulation.
Methods All cultures of C. vulgaris UTEX 395 were grown in airlift photobioreactors under the same conditions used by Griffiths et al. (2014). The glass and steel reactors were 60 cm high with an external diameter of 10 cm, a draft tube of 5 cm diameter and a working volume of 3.2 L. The medium was 3 N BBM (Bold 1949), with the nitrate concentration manipulated in each experiment as detailed below. Air enriched with 0.29 % CO2 was sparged at 2 L min−1, resulting in a circulation time of approximately 7 s and an overall mass transfer coefficient of 0.0094±0.00026 s−1 (Langley et al. 2012). Continuous illumination (250 μmol m−2 s−1 at the reactor surface, measured using an L-COR light meter, model LI250) was provided by three cool white 18 W fluorescent light bulbs (Osram). Culture temperature was monitored daily and remained constant at 25 °C±1 °C. Starter cultures were grown in 500 mL glass bottles in N-sufficient media for 7 to 10 days before being used to inoculate the airlift reactors at a starting concentration of 0.05 g L−1. Sterile, distilled water was added daily to replace that lost to evaporation. Antifoam (20 μL; Antifoam 204, Sigma-Aldrich) was added to each reactor to reduce foaming. Cultures were grown for 20 days. Batch culture C. vulgaris was inoculated into reactors with an initial media nitrate concentration of 1,200 mg L−1 (N-replete batch culture), 170 mg L−1 (N-limited batch culture), or 40 mg L−1 (very N-limited batch culture) and allowed to exhaust the nitrate naturally. Two-stage batch culture C. vulgaris was cultured under N-replete conditions (starting nitrate concentration of 1,500 mg L−1) for 6 days. On day 6,
biomass was harvested by centrifugation at 1,520×g for 10 min. The biomass was resuspended to the original volume (3.2 L) in nitrate-free medium or medium containing 170 mg L−1 nitrate. The starting nitrate concentration of 170 mg L−1 was chosen because it produced the highest volumetric lipid concentration in N-limited batch culture in previous work (Griffiths et al. 2014). Cultivation was continued in the original reactor to day 20. Fed-batch culture Under the growth conditions described above, two different starting nitrate concentrations and four different nitrate feeding regimes were tested. The two starting nitrate concentrations (40 and 170 mg L−1) were chosen as these were the concentrations that led to the most rapid lipid accumulation and highest overall lipid productivity respectively, in batch culture in previous work (Griffiths et al. 2014). Nitrate depletion in the media was found to occur within 1.2 and 2.2 days in the 40 and 170 mg L−1 cultures, respectively. Additional nitrate was fed as a concentrated aliquot in the make-up water, replacing that lost to evaporation, hence the volume of the cultures was not altered. In the first and second experiments, concentrated nitrate solution was added, to result in a final nitrate concentration in the reactors of either 40 or 80 mg L−1 (assuming no residual nitrate), every second day, between days 4 and 10. In the third and fourth experiments, the nitrate concentration in the reactors was adjusted (assuming no residual nitrate) to 40 mg L−1 either every 4 days (between days 4 and 16) or every 8 days (on days 8 and 16) (Table 1). Diluted cultures Previously, Griffiths et al. (2014) found that cultures with a higher starting nitrate concentration (e.g. 420 mg L−1 nitrate) attained a higher biomass concentration and accumulated lipid more slowly than those with a lower starting nitrate concentration. This suggested that the lower availability of light in Table 1 Nitrate feeding regimes for fed-batch cultures Fed-batch experiment
Starting nitrate concentration (mg L−1)
Nitrate addition (mg nitrate L−1 reactor volume)
Feeding frequency (days)
1 1 2 2 3 3 4 4
40 40 170 170 40 40 170 170
40 80 40 80 40 40 40 40
2 2 2 2 4 8 4 8
Appl Microbiol Biotechnol
dense algal cultures may limit lipid accumulation. To test this, cultures were diluted to different biomass concentrations on day 5. The more diluted cultures should experience less mutual shading and a higher light intensity. C. vulgaris was grown for 5 days under N-replete conditions. Biomass was harvested by centrifugation at 1,520×g for 10 min and resuspended in fresh, N-free medium, to volumes one, two and three times the original culture volume, yielding cultures that were undiluted, or diluted 1:1 and 1:2 with respect to the original culture and grown for a further 15 days. Assays The nitrate concentration in the media was monitored daily by UV spectrophotometry. Filtered samples were diluted to less than 12 mg L−1 nitrate. Optical density (OD) at 220 nm was measured in a quartz cuvette and quantified using a nitrate standard curve (Clesceri et al. 1998). Biomass was measured by OD at 750 nm every day and by dry weight (DW) every third day according to Griffiths et al. (2011). Individual calibration curves of OD as a function of DW were constructed for each culture and used to calculate DW from OD. Cell lipid content (measured as total fatty acid content) was quantified daily by direct transesterification and gas chromatography (Griffiths et al. 2010). Calculations Volumetric lipid concentration (PVOL) was calculated as the product of biomass concentration (X) and lipid content (P). Lipid productivity (QP) was calculated as PVOL at each time point divided by the time from inoculation. Maximum parameters were defined as the highest value reached within the 20day culture period. Yield of lipid on nitrate was calculated as the PVOL at the end of the culture period divided by the total amount of nitrate taken up by the cells during cultivation. In the case of the diluted cultures, the amount of consumed nitrate was divided by the dilution factor to account for the cells that were discarded.
Results Strategy 1: two-stage batch culture (biomass then lipid) C. vulgaris was grown under N-replete conditions for 6 days. On day 6, biomass was harvested and resuspended in medium containing either 0 or 170 mg L−1 nitrate. Cultivation was continued to day 20. Cultures in which the nitrate concentration had been reduced to zero or 170 mg L−1 continued to accumulate biomass up to day 20, although at a more linear rate than the equivalent nutrient replete batch culture (starting nitrate concentration 1,200 mg L−1 with no transfer, dashed
line, Fig. 1a). Maximum biomass concentrations of 1.9 and 2 g L−1 were reached in the 0 and 170 mg L−1 nitrate two-stage cultures, respectively (Fig. 1a), compared with 2.4 g L−1 in the N-replete batch culture. In N-replete batch culture, the fatty acid content remained between 10 and 12 % throughout the culture period (Fig. 1b). The fatty acid content of the twostage cultures increased slowly from 11 % at day 6 to 28 % (0 mg L−1) and 21 % DW (170 mg L−1) at day 20 (Fig. 1b). In comparison, the N-limited batch culture with an initial nitrate concentration of 170 mg L−1 reached a lipid content of 50– 60 %. The volumetric lipid concentration (PVOL) of both twostage cultures increased linearly over the culture period to a maximum of 525 and 434 mg L−1 for the 0 and 170 mg L−1 nitrate cultures respectively at day 20. This is compared with a maximum PVOL of 281 mg L−1 in N-replete batch culture and 790 mg L−1 in the N-limited batch culture with the highest lipid productivity (starting nitrate concentration of 170 mg L−1, Fig. 1c). Initially, both two-stage cultures had a lower average lipid productivity (QP) than the N-replete batch culture, but from day 10, this situation was reversed (Fig. 1d). The nitrate limited culture with a starting concentration of 170 mg L−1 had, on average, double the lipid productivity of the two-stage and the N-replete cultures. Strategy 2: fed-batch culture (lipid then biomass) In the first fed-batch experiment, C. vulgaris cultures were grown in medium containing 40 mg L−1 nitrate and fed with an additional 40 or 80 mg L−1 nitrate every second day between days 4 and 10. Results are compared with those of single-stage batch culture with 40 mg L−1 nitrate (dashed lines) and 170 mg L−1 nitrate (solid lines). Biomass concentration increased during and after N feeding and was greater in the culture fed with a higher concentration of nitrate (Fig. 2a). The maximum biomass concentrations reached (1.3 and 1.6 g L−1 in the cultures fed 40 and 80 mg L−1 nitrate, respectively) were over two to three times that in N-limited batch culture with 40 mg L−1 nitrate (0.5 g L−1). Lipid content was decreased by nitrate supplementation. Fed-batch cultures began accumulating lipid at the same rate as the batch culture, but lipid content decreased upon N feeding, remaining at an average of 24 and 14 % DW in the cultures fed 40 and 80 mg L−1, respectively, until feeding was stopped at day 10. Once N-feeding ceased, lipid content increased slowly again but did not reach the high lipid content (max. 60 % DW) of the batch cultures (Fig. 2b). As a result of the lower lipid content, the volumetric lipid concentration and the average lipid productivity of the fed-batch cultures was initially lower than that of the batch cultures. Once feeding ceased, the lipid yield and productivity of the fed-batch cultures was greater than that of the 40 mg L−1 nitrate batch culture, due to the higher biomass concentration but lower than that of the 170 mg L−1 nitrate batch culture (Fig. 2c, d).
Appl Microbiol Biotechnol
Fig. 1 a Biomass (markers) and media nitrate concentration (lines with markers), b lipid content, c volumetric lipid concentration (PVOL) and d average lipid productivity (QP) over 20 days in C. vulgaris grown initially under N-replete conditions, harvested on day 6 (indicated by a vertical line) and resuspended in fresh media with 0 (filled triangles) or
170 mg L−1 (filled squares) nitrate before continued cultivation. Results are compared with those of N-replete batch culture (starting nitrate concentration, 1,200 mg L−1, filled diamonds) and N-limited batch culture (starting nitrate concentration, 170 mg L−1, solid line)
In the second experiment, the same feeding regimes were carried out in cultures with a starting nitrate concentration of 170 mg L−1. Again, the final biomass concentration was higher in fed-batch cultures, although the growth rate between days 2 and 8 was slower in fed-batch than 170 mg L−1 batch culture (Fig. 3a). In contrast to the first experiment, feeding with 80 mg L−1 showed no improvement over 40 mg L−1. Again, nitrate feeding retarded lipid accumulation, with 80 mg L−1 having a greater effect than 40 mg L−1 (Fig. 3b). Volumetric lipid productivity in fed-batch cultures was lower than the 170 mg L−1 batch culture, due to the lower lipid content (Fig. 3c). Lipid productivity in the fed-batch cultures was lower during the period of nitrate feeding (days 4 to 10) and increased thereafter as the lipid content increased (Fig. 3d). The lipid productivity of the culture fed with 40 mg L−1 was greater than that fed 80 mg L−1 every 2 days. In the third experiment, started at 40 mg L−1 nitrate, feeding of 40 mg L−1 nitrate was carried out at two different intervals: every 4 days and every 8 days. Cultures with more infrequent feeding showed a more gradual increase in biomass (Fig. 4a). Growth rate increased after feeding (days 8 and 16) in the culture fed every 8 days. Lipid accumulation was retarded by nitrate feeding, but to a lesser degree the less frequent the
feedings (Fig. 4b). In the culture fed every 4 days, lipid content remained constant (between 34 and 38 % DW) throughout the period of nitrate feeding. In the culture fed every 8 days, lipid content decreased for 4 days after feeding at day 8, but then increased again to the same level as the batch culture. Upon feeding at day 16, the lipid content of the culture decreased as biomass concentration increased. Volumetric lipid productivity was greater in the culture fed every 4 days (621 mg L−1) than that fed every 8 days (574 mg L−1) or 2 days (490 mg L−1; Fig. 2c), but still less than that achieved in batch culture at a starting nitrate concentration of 170 mg L−1 (Fig. 4c). The cultures fed every 4 or 8 days had a more constant lipid productivity (Fig. 4d) relative to those fed every 2 days, which showed a sharp dip in productivity during additional nitrate feeding (e.g. Fig. 1d). The average lipid productivity was greater than the 40 mg L−1 batch culture but less than the 170 mg L−1 batch culture. In the fourth experiment, the same feeding regimes were repeated with a starting nitrate concentration of 170 mg L−1. In these cultures, feeding with 40 mg L−1 nitrate every 4 or 8 days did not have a significant effect on biomass concentration (maximum biomass concentration was 1.7 and 1.6 g L−1 in the cultures fed every 4 and 8 days, respectively, as opposed to
Appl Microbiol Biotechnol
Fig. 2 a Biomass (markers) and media nitrate concentration (lines), b lipid content, c volumetric lipid concentration (PVOL) and d average lipid productivity (QP) over 20 days in C. vulgaris grown under fed-batch conditions, with starting nitrate concentration of 40 mg L−1, fed either
40 mg L−1 (filled triangles) or 80 mg L−1 (filled squares) nitrate on days 4, 6, 8 and 10. The results of single-stage batch culture with 40 mg L−1 nitrate (dashed line) and 170 mg L−1 nitrate (solid line) are shown for comparison
1.5 g L−1 in the normal batch culture) (Fig. 5a). Lipid accumulation, however, was retarded and remained relatively constant (between 19 and 31 % DW) over the period of feeding (Fig. 5b). In the culture fed every 8 days, lipid content increased between feeds, but to a lesser extent than in the 40-mg L−1 starting nitrate culture (Fig. 4b). As a result of the lower lipid content, the volumetric lipid concentration of the fed-batch cultures was lower than that of the 170mg L−1 batch culture (Fig. 5c). The lipid productivity of the fed-batch cultures started at 170 mg L−1 nitrate was initially lower than those started at 40 mg L−1, but reached a constant productivity earlier than those fed more frequently (Fig. 5d). Figure 6 plots the biomass and volumetric lipid concentrations for all the fed-batch experiments, as well as N-limited batch cultures with starting nitrate concentrations of 40 and 170 mg L−1, in order of the total amount of nitrate added to the culture over the 20-day cultivation period. Below a critical N level (approximately 200 mg L−1 nitrate), the total biomass concentration was limited by the N availability, and the lipid content of the biomass was relatively high. Above this N level, the biomass and PVOL achieved were similar, regardless of
when the N was added. Lipid content decreased with increasing N fed. Effect of light availability The biomass concentration after centrifugation and resuspension in N-free medium was 1.1, 0.7 and 0.5 g L−1 for the undiluted culture and cultures diluted 1:1 and 1:2, respectively. All three cultures continued to increase in biomass at similar rates, reaching maximum biomass concentrations of 2.3, 1.8 and 1.4 g L−1, respectively (Fig. 7a). The more dilute the culture, the faster the lipid accumulation rate. The most dilute culture (1:2) reached the highest lipid content (39 % DW), followed by the culture diluted 1:1 (33 % DW) and then the undiluted culture (25 % DW) (Fig. 7b). Although the diluted cultures achieved a higher lipid content, this was offset by their lower biomass concentration. As a result, the volumetric lipid yield of all three cultures was similar (reaching a maximum of 565, 555 and 545 mg L−1 in undiluted, 1:1 and 1:2, respectively) (Fig. 7c). This was less than the PVOL of the best N-limited batch culture (starting nitrate concentration, 170 mg L−1). The lipid productivity (QP) was initially lower
Appl Microbiol Biotechnol
Fig. 3 a Biomass (markers) and media nitrate concentration (lines), b lipid content, c volumetric lipid concentration (PVOL) and d average lipid productivity (QP) over 20 days in C. vulgaris grown under fed-batch conditions, with starting nitrate concentration of 170 mg L−1, fed either
40 mg L−1 (filled triangles) or 80 mg L−1 (filled squares) nitrate on days 4, 6, 8 and 10. The results of single-stage batch culture with 40 mg L−1 nitrate (dashed line) and 170 mg L−1 nitrate (solid line) are shown for comparison
in the diluted cultures after dilution but from day 12 onwards, they matched the QP of the undiluted culture (Fig. 7d).
Discussion Strategy 1: two-stage batch culture
Comparison In order to determine the optimum culture regime in terms of lipid productivity, the best results obtained in each of the different strategies were compared on the basis of maximum biomass (X), lipid content (P), volumetric lipid concentration (PVOL) and lipid productivity (QP) as well as yield of lipid on nitrate (Table 2). The best culture strategy, in terms of PVOL, QP and yield on nitrate, was N-limited batch culture with 170 mg L−1 nitrate. The second best, in terms of PVOL, was fed-batch with the feeding strategy of 40 mg L−1 every 4 days. Two-stage batch culture improved the lipid productivity the least of the N-limitation strategies tested. Although none of the two-stage or fed-batch cultivation strategies improved on the lipid productivity of N-limited batch culture, they were all better than N-replete batch culture, which was the worst strategy in terms of PVOL, QP and yield of lipid on nitrate, despite having the highest final biomass concentration.
In two-stage batch culture, the goal of the first stage was to optimise biomass productivity, followed by a second, Nstarvation stage to increase lipid content. After the sudden removal or reduction of nitrate in the medium on day 6, C. vulgaris was able to continue growing for up to a further 10 days and accumulated lipid only gradually. Lipid accumulation is a function of the N content of the cells (Griffiths et al. 2014). The continued cell growth and lack of lipid accumulation after N removal may be supported by intracellular stores of N (Li et al. 2008; Lourenço et al. 2004). During the initial N-replete phase, the two-phase cultures consumed approximately 500 mg L−1 nitrate, almost three times the amount available to the batch culture with a starting nitrate concentration of 170 mg L−1, but the same amount of biomass was produced in each (approximate;y 0.85 g L−1), therefore the cells in two-phase culture had higher intracellular N reserves upon nitrate depletion. By utilising reserves of inorganic N (NO3, NO2 and NH4), as well as breaking down N-containing
Appl Microbiol Biotechnol
Fig. 4 a Biomass (markers) and media nitrate concentration (lines), b lipid content, c volumetric lipid concentration (PVOL) and d average lipid productivity (QP) over 20 days in C. vulgaris grown under fed-batch conditions, with starting nitrate concentration of 40 mg L−1, fed
40 mg L−1 nitrate every 4 days (filled triangles) or 8 days (filled squares). The results of single-stage batch culture with 40 mg L−1 nitrate (dashed line) and 170 mg L−1 nitrate (solid line) are shown for comparison
macromolecules such as protein, nucleic acid and pigment, cells may be able to delay the major metabolic shifts associated with severe N limitation. This is a disadvantage from the point of view of optimising lipid productivity, as it causes a delay in lipid accumulation after N removal. The concentration of nitrate in the first, N replete, phase was varied to see whether this influenced the growth and lipid content after N removal. It was hypothesised that the initial N concentration could affect the amount of N stored and hence the physiology of the cells after N removal. It was found that the concentration of nitrate during the initial N-replete growth phase could vary between 500 and 1,000 mg L−1 without effect (Fig. S1 in the Electronic supplementary material). It was hypothesised that the shear stress involved in centrifugation of the two-stage batch cultures may have had an effect of the growth of the cells. This was tested by comparing cultures in which the nitrate was naturally exhausted after 5 days, with those in which the culture medium was changed from N-replete to N-free media via centrifugation on day 5. The growth of the centrifuged culture was not adversely affected (Fig. S2 in the Electronic supplementary material). The lipid yield and productivity of the culture transferred from N-replete to N-free media was higher than that transferred to media with an intermediate level of N (170 mg L−1
nitrate) (Table 2). This is in contrast to the report of Stephenson et al. (2010) who investigated the effect of various nitrate and cell concentrations on lipid productivity after transfer to N-limited medium. It was found that maximum average lipid productivity in C. vulgaris (46 mg L−1 day−1) was achieved when cultures were initially grown in N-sufficient medium, and then transferred to medium with an intermediate nitrate concentration (200 mg L−1). This (and the higher lipid productivity reported) could be due to different culture conditions, particularly the higher illumination levels used by Stephenson et al. (2010). Strategy 2: fed-batch culture Fed-batch cultures were inoculated at a limiting nitrate concentration, in order to maximise lipid content, but at a low biomass concentration. Small quantities of nitrate were added at various time points in an attempt to enhance biomass concentration while maintaining a high lipid content. The amount and timing of N feeding in fed-batch culture had a significant effect on biomass productivity and lipid accumulation during the culture period, although, over the range tested here, not on the final product yield. A higher starting nitrate concentration (170 rather than 40 mg L−1) decreased
Appl Microbiol Biotechnol
Fig. 5 a Biomass (markers) and media nitrate concentration (lines), b lipid content, c volumetric lipid concentration (PVOL) and d average lipid productivity (QP) over 20 days in C. vulgaris grown under fed-batch conditions, with starting nitrate concentration of 170 mg L−1, fed
40 mg L−1 nitrate every 4 days (filled triangles) or 8 days (filled squares). The results of single-stage batch culture with 40 mg L−1 nitrate (dashed line) and 170 mg L−1 nitrate (solid line) are shown for comparison
the response of the cells to additional nitrate fed (Figs. 3 and 5), but the overall productivity was enhanced by a higher biomass concentration. Feeding more nitrate (80 rather than 40 mg L−1) resulted in greater biomass accumulation but also a greater decrease in lipid content (Figs 2 and 3). Longer gaps between nitrate feeding (every 4 or 8 days instead of every 2),
decreased the lipid content less and increased the biomass concentration more gradually (Figs. 2 and 4). A low starting nitrate concentration (40 mg L−1), combined with feeding of a small amount of nitrate (40 mg L−1) every 4 days, maintained a steady increase in biomass concentration without compromising the lipid content as severely. There was a
Fig. 6 Summary of maximum volumetric lipid concentration (black bars) and total biomass concentration (sum of black and grey bars) achieved during fedbatch experiments with different starting nitrate concentrations (40 or 170 mg L−1, fourth row), fed either 40 or 80 mg L−1 nitrate (third row) at different time intervals (every 2, 4 or 8 days, second row). The experiments are shown in order of total amount of nitrate added to the culture (mg L−1, first row)
Appl Microbiol Biotechnol
Fig. 7 a Biomass, b lipid content, c volumetric lipid concentration (PVOL) and d average lipid productivity (QP) over 20 days in C. vulgaris grown initially under N-replete conditions, harvested on day 4.8 (indicated by a vertical line) and resuspended in fresh media undiluted (filled
triangles) and diluted 1:1 (filled squares) or 1:2 (filled diamonds) before continued cultivation. The results for single-stage batch culture with 170 mg L−1 nitrate (solid line) are shown for comparison
critical total N supply (approximate;y 200 mg L−1 nitrate over 20 days) below which growth was constrained by availability of N and lipid content was increased. Above 200 mg L−1 total nitrate supply over 20 days, biomass and lipid production was similar regardless of the timing of N feeding. This suggests that the critical variable in achieving high lipid productivity in fed-batch culture is the total amount of N fed and not the quantity or timing of feeding. The N-limited batch culture fed
170 mg L−1 nitrate gave the highest PVOL. This starting nitrate concentration is just below the critical total nitrate feed, resulting in a similar final biomass concentration, but enhanced lipid content. Takagi et al. (2000) also reported that the timing of nitrate feeding (during log phase, linear phase or stationary phase) during fed-batch had negligible effect on the lipid content of Nannochloris sp., although biomass concentration was higher
Table 2 Comparison of different culture regimes in terms of maximum biomass (X), lipid content (P), volumetric lipid concentration (PVOL) and lipid productivity (QP) as well as yield of lipid per gram nitrate used
N-replete batch (starting NO3, 1,200 mg L−1) N-limited batch (starting NO3, 170 mg L−1) Two-stage batch (1,500 to 0 mg L−1) Two-stage batch (1,500 to 170 mg L−1) Fed-batch 40 mg L−1 every 4 days (starting 40 mg L−1) Fed-batch 40 mg L−1 every 4 days (starting 170 mg L−1) Light experiment (undiluted) Light experiment (diluted 1:1) Light experiment (diluted 1:2)
Max X (g L−1)
Max P (% DW)
Max PVOL (mg L−1)
Max QP (mg L−1 day−1)
Yield of lipid on nitrate (g g−1)
2.4 1.5 1.8 2.0 1.5 1.7 2.3 1.8 1.4
12 54 26 20 41 35 25 33 39
281 790 477 391 621 588 565 555 545
25 41 25 21 33 34 30 30 29
0.2 4.6 1.1 0.7 3.1 1.8 1.3 2.6 3.9
Appl Microbiol Biotechnol
with later feed times. They found that feeding of small amounts of nitrate (0.9 mM, equivalent to 56 mg L−1) during the log phase increased the biomass concentration to twice that obtained without feeding, while lipid content was maintained at a high level compared with N-replete batch culture. They did not report lipid productivities, however, calculating these from the final biomass concentration and lipid content, PVOL was not improved compared with N-limited batch culture at starting nitrate concentration between 250 and 600 mg L−1 nitrate. The fact that none of the fed-batch strategies tested outperformed N-limited batch culture at the optimal starting nitrate concentration is also in agreement with Hsieh and Wu (2009), who reported that additional feeding of urea during the growth curve of a marine Chlorella sp. prolonged cell growth and increased the final biomass concentration, however, the higher the amount of urea fed, the lower the lipid content. They showed that small amounts of additional urea could enhance lipid productivity (up to 123 mg L−1 day-1), above that of cultures without additional feeding, but not above the maximum found in N-limited batch culture (124 mg L−1 day1 ) with an optimal starting urea concentration of 0.1 g L−1, equivalent to 205 mg L−1 nitrate. This value agrees closely with the critical total nitrate feed concentration found in this work (200 mg L−1). Effect of light availability Cultures diluted to a lower biomass concentration in N-free medium accumulated lipid more rapidly and to a greater degree than the undiluted culture. This supports the hypothesis that lipid accumulation is light limited in dense cultures. Rodolfi et al. (2009) also found that increasing the illumination supplied to a flat panel reactor enhanced both the biomass productivity and the lipid content of Nannochloropsis. Klok et al. (2013) reported that, under N-limited conditions, an increase in the average light intensity led to more and larger lipid bodies at the same N supply rate in N. oleoabundans. From light attenuation measurements in the airlift reactors at the biomass densities and incident light intensities used, it is assumed that all incident light was absorbed by both the diluted and undiluted cultures, yielding similar volumetric light absorption in each case. In the diluted cultures, more light would have been absorbed per cell, increasing specific lipid production, however, due to the identical volumetric light absorption, overall volumetric lipid production remained constant. These results support the work by Stephenson et al. (2010) who noted that cell density at the beginning of the second stage of two-stage N-limited culture affects the rate of lipid accumulation. Cells transferred to N-deficient media at a lower cell density accumulated more lipid; however, similarly to this work, due to lower cell concentrations, the overall productivity of cultures inoculated at lower cell density was
lower. Maximum productivity was achieved in the culture with the highest cell concentration. A single dilution at the time of media transfer did not improve lipid productivity in this work but enhancing light availability in other ways (e.g. a draw and feed strategy, increasing the overall ambient light intensity or enhancing mixing) may improve both biomass and lipid yield. Caution should be used when dramatically increasing light intensity, as the reduced pigment content found in N-limited cultures (Griffiths et al. 2014) could lead to enhanced susceptibility to photoinhibition. While the two-stage batch and fed-batch culture regimes were not fully optimised in this work, limiting the strength of conclusions that can be drawn, none of the multistage cultivation strategies tested were found to improve lipid productivity above that measured in N-limited (170 mg L−1) batch culture where the nitrate became exhausted naturally in the medium during the exponential growth phase. If two-stage batch culture were carried out where the second, nitratelimited, stage was initiated after exactly 170 mg L−1 nitrate was consumed in stage one, it is likely that the lipid productivity would be very similar to a batch culture with an initial nitrate concentration of 170 mg L−1. However, transfer of culture in any two-stage system would require additional media as well as energy input and labour costs for pumping and biomass harvesting, increasing the costs and environmental footprint of the process. Two-stage cultivation is therefore not recommended for microalgal biodiesel production, unless the conditions of the second stage are cheaper or less energy intensive to maintain, for example, in transfer from a closed reactor to an open pond. In designing conditions for the second stage, it should be borne in mind that cells require conditions of sufficient light and carbon supply for lipid formation. This implies that two-stage processes that do not provide optimal light conditions in the second, lipid accumulation stage will not be optimal. Ideally, reactor design and cultivation strategies should aim to maintain optimal light provision throughout the culture period. Hsieh and Wu (2009) showed that a strategy of semicontinuous cultivation, where 25 % or the culture was removed and 0.025 g L−1 urea added every day from days 3 to 7, produced a higher lipid productivity (139 mg L−1 day−1) than either batch or fed-batch without culture removal. The removal of culture may have facilitated light penetration, and hence lipid accumulation, by reducing cell density. Klok et al. (2013) have also shown that the TAG productivity of N. oleoabundans could be enhanced by N limitation under continuous culture in a turbidostat, relative to N replete continuous culture. Continuous or semicontinuous culture at the critical N-feed rate to balance biomass and lipid productivity, and high light intensities, may be a promising strategy for further investigation.
Appl Microbiol Biotechnol
Although lipid productivity has a direct effect on the economic viability of algal biodiesel production, a high productivity in itself is not sufficient to guarantee success. Other factors such as ease of cultivation, resistance to contamination, and effects on downstream processing should also be taken into account. Restriction of N supply may discourage contamination by preventing growth of other algal species; however, N-limited cultures may prove less robust in the outdoor environment. In addition, N-limited cells may have different densities or cell surface properties and result in lower final biomass concentrations, which could affect harvesting. In conclusion, all N-limited culture strategies improved the volumetric lipid yield relative to N-replete culture. However, transfer of cells from N-replete to N-limited media (two-stage batch) or gradual feeding of N (fed-batch) did not improve the lipid yield on nitrate over that of cultures allowed to exhaust a low level of nitrate (170 mg L−1) naturally (N-limited batch culture). The lower-than-expected lipid productivity in twostage culture was due to a delay in lipid accumulation, postulated to be due to the utilisation of intracellular N reserves, and a decreased rate of lipid accumulation, possibly due to light limitation in dense cultures. N-limited batch and fed-batch cultures showed the highest yield of lipid per gram nitrate used, potentially lowering the cost and environmental impact associated with the provision of fixed N.
Acknowledgments This work is based upon research supported by the South African National Energy Development Institute (SANEDI), the South African Research Chairs Initiative (SARChI) of the Department of Science and Technology, the National Research Foundation (NRF) and the Technology Innovation Agency (TIA). The financial assistance of these organisations is hereby acknowledged. Any opinion, finding and conclusion or recommendation expressed in this material is that of the authors and SANEDI, SARChI, TIA or the NRF do not accept any liability in this regard.
References Ben-Amotz A (1995) New mode of Dunaliella biotechnology: two-phase growth for β-carotene production. J Appl Phycol 7(1):65–68 Bold H (1949) The morphology of Chlamydomonas chlamydogama sp. nov. Bull Torrey Bot Club 76:101–108 Borowitzka MA (1992) Algal biotechnology products and processes— matching science and economics. J Appl Phycol 4:267–279 Chelf P (1990) Environmental control of lipid and biomass production in two diatom species. J Appl Phycol 2(2):121–129 Chisti Y (2007) Biodiesel from microalgae. Biotechnol Adv 25(3):294– 306 Clesceri LS, Greenberg AE, Eaton AD (1998) Standard methods for the examination of water and wastewater, 20th edn. American Public Health Association, Washington DC Converti A, Casazza AA, Ortiz EY, Perego P, Del Borghi M (2009) Effect of temperature and nitrogen concentration on the growth and lipid content of Nannochloropsis oculata and Chlorella vulgaris for biodiesel production. Chem Eng Process 6:1146–1151
Griffiths MJ, Harrison STL (2009) Lipid productivity as a key characteristic for choosing algal species for biodiesel production. J Appl Phycol 21(5):493–507 Griffiths MJ, van Hille RP, Harrison STL (2010) Selection of direct transesterification as the preferred method for assay of fatty acid content of microalgae. Lipids 45(11):1053–1060 Griffiths MJ, Garcin C, van Hille RP, Harrison STL (2011) Interference by pigment in the estimation of microalgal biomass concentration by optical density. J Microbiol Methods 85(2):119–123 Griffiths MJ, van Hille RP, Harrison STL (2012) Lipid productivity, settling potential and fatty acid profile of eleven microalgal species grown under nitrogen replete and limited conditions. J Appl Phycol 24(5):989–1001 Griffiths MJ, van Hille RP, Harrison STL (2014) The effect of nitrogen limitation on lipid productivity and cell composition in Chlorella vulgaris. Appl Microbiol Biotechnol 98(5):2345–2356 Hsieh C-H, Wu W-T (2009) Cultivation of microalgae for oil production with a cultivation strategy of urea limitation. Bioresour Technol 100(17):3921–3926 Huntley ME, Redalje DG (2006) CO2 mitigation and renewable oil from photosynthetic microbes: a new appraisal. Mitig Adapt Strateg Glob Chang 12(4):573–608 Klok AJ, Martens DE, Wijffels RH, Lamers PP (2013) Simultaneous growth and neutral lipid accumulation in microalgae. Bioresour Technol 134:233–243 Langley N, Harrison STL, van Hille RP (2012) A critical evaluation of CO2 supplementation to algal systems by direct injection. Biochem Eng J 68:70–75 Lardon L, Hélias A, Sialve B, Steyer JP, Bernard O (2009) Life-cycle assessment of biodiesel production from microalgae. Environ Sci Technol 43(17):6475–6481 Li Y, Horsman M, Wang B, Wu N, Lan CQ (2008) Effects of nitrogen sources on cell growth and lipid accumulation of green alga Neochloris oleoabundans. Appl Microbiol Biotechnol 81(4):629– 636 Liu X, Sheng J, Curtiss R (2011) Fatty acid production in genetically modified cyanobacteria. Proc Natl Acad Sci U S A 108(17):6899– 6904 Lourenço S, Barbarino E, Lavín P, Lanfer Marquez U, Aidar E (2004) Distribution of intracellular nitrogen in marine microalgae: calculation of new nitrogen-to-protein conversion factors. Eur J Phycol 39(1):17–32 Lv J-M, Cheng L-H, Xu X-H, Zhang L, Chen H-L (2010) Enhanced lipid production of Chlorella vulgaris by adjustment of cultivation conditions. Bioresour Technol 101(17):6797–6804 Mata TM, Martins AA, Caetano NS (2010) Microalgae for biodiesel production and other applications: a review. Renew Sustain Energy Rev 14(1):217–232 Miao X, Wu Q (2006) Biodiesel production from heterotrophic microalgal oil. Bioresour Technol 97(6):841–846 Pruvost J, Van Vooren G, Cogne G, Legrand J (2009) Investigation of biom ass and lipids production with Neochloris oleoabundans in photobioreactor. Bioresour Technol 100(23): 5988–5995 Richmond A (2004) Microalgal culture. Biotechnology and applied phycology. Blackwell Science, Oxford Rodolfi L, Chini Zittelli G, Bassi N, Padovani G, Biondi N, Bonini G, Tredici MR (2009) Microalgae for oil: strain selection, induction of lipid synthesis and outdoor mass cultivation in a low-cost photobioreactor. Biotechnol Bioeng 102(1):100–112 San Pedro A, Gonzalez-Lopez CV, Acién FG, Molina-Grima E (2013) Marine microalgae selection and culture conditions optimization for biodiesel production. Bioresour Technol 134:353–361 Sheehan J, Dunahay T, Benemann JR, Roessler P (1998) A look back at the U.S. Department of Energy's Aquatic Species Program:
Appl Microbiol Biotechnol Biodiesel from algae. Closeout report. National Renewable Energy Lab, Department of Energy, Golden, Report number NREL/TP-58024190, dated July 1998 Shifrin NS, Chisholm SW (1981) Phytoplankton lipids: interspecific differences and effects of nitrate, silicate and light-dark cycles. J Phycol 17:374–384 Stephenson AL, Dennis JS, Howe CJ, Scott SA, Smith AG (2010) Influence of nitrogen-limitation regime on the production by Chlorella vulgaris of lipids for biodiesel feedstocks. Biofuels 1(1): 47–58
Su C-H, Chien L-J, Gomes J, Lin Y-S, Yu Y-K, Liou J-S, Syu R-J (2011) Factors affecting lipid accumulation by Nannochloropsis oculata in a two-stage cultivation process. J Appl Phycol 23:903–908 Takagi M, Watanabe K, Yamaberi K, Yoshida T (2000) Limited feeding of potassium nitrate for intracellular lipid and triglyceride accumulation of Nannochloris sp. UTEX LB1999. Appl Microbiol Biotechnol 54(1):112–117 Xiong W, Li X, Xiang J, Wu Q (2008) High-density fermentation of microalga Chlorella protothecoides in bioreactor for microbiodiesel production. Appl Microbiol Biotechnol 78(1):29–36
