This investigation provided useful information for improving the production of DHA and the quality of oil from BL10 through the optimization of culture conditions. The first method is the adoption of 0.5% diluted seawater for the cultivation of BL10. In our study, cultures cultivated by 0.5% diluted seawater yielded the greatest production of high-quality oil in BL10 cells. This approach would require fermenters with anti-corrosion coating, because chloride levels in 0.5% diluted seawater are highly corrosive to stainless steel. In addition, the decrease in dry biomass and increase in cellular debris we observed with this treatment (compared with BL10 cultivated in 1.0% and 0.2%) suggests that the extremely high quantities of oil accumulating in cells could result in cell disruption. As a result, strong agitation that could shear cells must be avoided when lipid content is high. For cases in which anti-corrosion fermenters are unavailable, sodium sulfate could be used as an alternative to sea salts in order to reduce chloride ions in the medium. Unfortunately, the addition of sodium sulfate reduces the quality of the oil produced as well. To salvage the quality of oil, the fed-batch method (instead of the batch cultivation method) could be employed.
This investigation also provided useful information related to using BL10 microalgae as bioreactors to produce specific lipids, such as DHA-containing phospholipids, which have unique biotechnological value in the enrichment of live prey to improve the growth of fish larvae (Gisbert et al.
2005), and as a food supplement to reduce the risk of developing dementia (Schaefer et al.
2006). Our preliminary data shows that, after 18 hours of growth, 88% of the total lipids harvested from cells were phospholipids, in which 62% of the fatty acid was composed of DHA. These results could serve as a protocol for the production of DHA-containing phospholipids from BL10 culture. A similar concept which used a glucose-deficient medium to stimulate the conversion of DHA-rich triacylglycerol to DHA-containing phospholipids in a thraustochytrid-like microorganism (strain B12) has been mentioned in previous study (Okuyama et al.
2007).
Prior research has investigated the environmental conditions that influence the composition of fatty acids and the accumulation of lipids in strains of Aurantiochytrium, and concluded the following:
1. Temperature: In a study on
Aurantiochytrium sp. strain mh0186, Taoka et al (
2009) discovered that an elevation in temperature from 10°C to 35°C results in a concomitant decrease in both DHA/DPA and DHA/PA ratios. Furthermore, in an investigation of
Aurantiochytrium sp. strain OUC-88, Zhu et al (
2007) reported an increase in the production of the two linear and odd-numbered fatty acids C15:0 and C17:0 at higher temperatures (37°C).
2. Salinity and salt species:
Yaguchi et al (1997) determined that the optimal salinity concentration for the cultivation of
Aurantiochytrium limacinum strain SR21 to achieve maximum biomass ranges between 50% and 200% salinity of sea water. These results are similar to those found for
Aurantiochytrium strain OUC-88 (Zhu et al.
2007). In addition, the lipid content in OUC-88 increased slightly from 41.34 to 48.97%, following a decrease in salt concentration from 3.6 to 0.9%. Palmitic acid accounted for 43.38% of the TFA when this microalga was cultivated at 3.5%, and dropped to 32.62% at a salinity level approaching 0%. However, a different pattern in the production of fatty acids was observed in
Aurantiochytrium mangrovei strain Sk-02, which has low TFA content in a weak saline environment (between 0 and 1.0%), and higher, stable TFA content in strong saline environment (between 1.0 and 6.8%) (Unagul et al.
2006).
3. Substrate concentration: Glucose levels as high as 100 gL
-1 resulted in the retardation of growth in
Aurantiochytrium strains MP2 (Wong et al.
2008) and SR21 (Yaguchi et al.
1997).
4. Growth stage: In a study of
Aurantiochytrium sp. strain T66, Jakobsen et al (
2008) revealed that the DHA/DPA ratio is slightly higher in the lipid-accumulation phase than in the cell division phase. Chi et al (
2009) later reported that the cell size of
A. limacinum strain SR21 is stable during cell division, but continually increases during the lipid accumulation stage.
5. Oxygen level: Jakobsen et al (
2008) suggested that lipid accumulation can be initiated by limiting the availability of O
2, which also results in an increase in the DHA/DPA ratio. Chi et al (
2009) further reported that aeration has a negative influence on the accumulation of fatty acids in
A. limacinum strain SR21.
6. Response to cerulenin: Hauvermale et al. (
2006) investigated the influence of cerulenin on the synthesis of fatty acids in
Schizochytrium sp. strain ATCC20888 (our preliminary study suggested that it could be re-classified into
Aurantiochytrium due to its close phylogenetic relationship with other known
Aurantiochytrium strains). This study revealed a significant reduction in the synthesis of saturated fatty acids when cerulenin levels exceeded 1

μM. Furthermore, increasing cerulenin levels to greater than 25

μM resulted in a near total blockage in the synthesis pathways of both saturated and non-saturated fatty acids.
This study discovered considerable differences between BL10 and other Aurantiochytrium strains with regard to the synthesis of fatty acids and the accumulation of lipids, including the following:
1. Response to salinity and salt species: The optimal salinity for the production of biomass and DHA by BL10 is 0.5%, which is considerably lower than that of other strains of
Aurantiochytrium (Yaguchi et al.
1997; Zhu et al.
2007). In addition, the effect of salinity on fatty acid production differs considerably as well (Zhu et al.
2007; Unagul et al.
2006), since the known strains of
Aurantiochytrium are not particularly susceptible to the influence of salinity; however, reducing the salinity could induce a dramatic increase in the oil production of BL10. Given that BL10 is capable of maximum DHA production in a weaker saline environment (which is less corrosive to stainless steel fermenters), BL10 is actually a more promising candidate for the commercial production of DHA.
2. Substrate concentration: increasing glucose levels to as high as 150

gL
-1 did not influence the growth rate of BL10. Thus, BL10 has a greater tolerance for high glucose levels in the culture medium than other species of
Aurantiochytrium (Yaguchi et al.
1997; Wong et al.
2008).
3. Growth stages: BL10 has a considerably higher DHA/DPA ratio during the cell division phase than during the lipid accumulation phase. This trend is the opposite of that observed in
Aurantiochytrium strain T66 (Jakobsen et al.
2008). A continuous decrease in the size of BL10 during cell division stage also differs from that of strain SR21, which maintains stable cell size throughout the stage (Chi et al.
2009).
4. Tolerance to cerulenin: BL10 is clearly more tolerant to cerulenin treatment than ATCC20888 (Hauvermale et al.
2006). Treatment with 25

μM cerulenin arrested the synthesis of fatty acids in ATCC20888; however, this treatment had far less influence on the fatty acid composition and DHA production of BL10.
Our results revealed a number of other interesting characteristics related to the composition of fatty acids and the production of lipids by BL10 in response to environmental changes, including the following:
1. The influence of sodium sulfate on DHA/PA ratio.
2. Dramatic up-regulation in the synthesis of C15:0 and C17:0 when cell division approached termination. The up-regulation in the synthesis of C15:0 and C17:0 was first reported for
Aurantiochytrium cultivated under conditions of high temperature and low salinity (Zhu et al.
2007), revealing the possible anti-stress role of the two fatty acids. An increase in the two fatty acids in BL10 toward the termination of cell division is much more significant than under conditions of low salinity; therefore, we speculate that the two fatty acids may play a more important role in shifting the condition of cell physiology from cell proliferation in the cell division stage to energy storage in the lipid accumulation stage, and since synthesis of C15:0 and C17:0 needs a precursor: propionate, a metabolic intermediate from degradation of various amino acids, such as valine and isoleucine as well as α-ketobutyric acid (the metabolic products of methionine and threonine) (Smith and Macfarlane
1997; Vlaeminck et al.
2006), as a result, it has been suggested that there is a sudden increase in the turnover of intrinsic proteins as the cell division stage approaches termination, followed by an increase in concentration of propionate and the final up-regulation in the synthesis of C15:0 and C17:0.