The experiments concerning the sucrose concentration showed best growth and fatty acid production in a range of 20 to 40 g l-1 (Figures 2A and B). Lower as well as higher concentrations led in flask experiments to a decrease of accumulation. The culture that was grown in presence of 240 g l-1 sucrose stored even less fatty acids than the culture grown in presence of 5 g l-1 though at the latter concentration all carbon was depleted during the exponential growth phase. Due to the very low growth inhibition of sucrose even at concentrations as high as 240 g l-1, it seems unlikely that the bacterial metabolism is strongly affected. Moreover, the growth behavior of the culture containing 240 g l-1 sucrose (840 mOsm l-1 of the complete medium) is very similar to the growth in medium with an osmotic concentration of 800 mOsm l-1 (compare Figures 2B and G), so that the negative effect of high sucrose concentrations seems to be due to the enhanced osmotic pressure. Tolerance towards comparably high sucrose concentrations has also been shown in other organisms, for example in Bacillus sp. or in Saccharomyces cerevisiae (Belghith et al. 2012;Ando et al. 2006).
The tolerance of high sucrose concentrations is interesting also with regard to the report of (Kurosawa et al. 2010), who showed that the tolerance towards high glucose concentrations was paralleled by a long lag-phase, which could be reduced only by larger sizes of the inocula. A possible explanation for this different behavior is that sucrose belongs to the compatible solutes, while glucose potentially interferes with the bacterial metabolism. Additionally, the osmotic concentration of the defined medium with 240 g l-1 glucose and 13.4 g l-1 (NH4)2SO4 (Kurosawa et al. Kurosawa et al. 2010) amounts to 1600 mOsm l-1 and according to our data is at the limit of an acceptable growth rate.
Altering the ammonium concentration in flask experiments caused strong differences in the final OD. However, in the first 30 hours of cultivation only the cultures with ammonium concentrations lower than 0.4 g l-1 showed reduced growth, while concentrations of up to 1.4 g l-1 seemed not to be growth limiting (Figure 2C). Concerning the final fatty acid concentration there is a clear difference between the cultures that contain up to 0.4 g l-1 ammonium and the cultures with higher concentrations. This difference is easily explained by the fact that the cultures with low fatty acid concentration have not consumed the nitrogen in the medium, and consequently the fatty acid accumulation as TAG was not induced. In the cultures with a reduced ammonium concentration, the final OD correlated with the ammonium concentration (Figures 1 and 2C). With 0.4 g l-1, an OD of 24 was reached, while 0.2 g l-1 and 0.1 g l-1 led to final ODs of about 12 and 6. Growth with only 20% of the standard concentration of ammonium (0.07 g l-1, provided as 0.2 g l-1 NH4Cl) gave a final OD of 20% of the control experiment. Thus it seems reasonable that the differences in the final OD can be explained either by a reduced cell growth, caused by too low ammonium concentrations, with normal fatty acid production or by a maximal cell growth but without a significant fatty acid production in the case of ammonium concentrations higher than 0.4 g l-1.
It was shown that the final OD of R. opacus was positively affected by high concentrations of both, MgSO4 or MgCl2 (Figure 2D). In comparison with the good growth and fatty acid synthesis of R. opacus under reduced MgSO4 concentrations (Figure 1), it is obvious that the sulfate concentration (0.08 g l-1 in the basic medium and 0.02 g l-1 under reduced conditions) is not limiting in the flask experiments and the higher OD with additional MgSO4 is caused by the magnesium ions. The fatty acid content and profile were not significantly altered, but the accumulation appeared to be enhanced in the presence of 0.12 g l-1 and 0.51 g l-1 magnesium and slightly decreased in the presence of higher magnesium concentrations in the culture medium. A continued cell division of R. opacus during the nitrogen limitation can be ruled out since the lack of nitrogen limits the DNA replication and protein biosynthesis. However, it is possible that the higher OD reflects a continued cell elongation in presence of enhanced magnesium concentrations, which might be due to an ongoing membrane synthesis.
The small differences in TAG storage between the control and 5-times reduced or up to 50-times increased magnesium concentrations indicate that the availability of magnesium ions as cofactor for fatty acid synthesis was sufficient under all tested conditions or had only a minor effect. The osmotic concentration of the growth medium in these experiments was in the range of 253 to 376 mOsm l-1 and thus should not be the cause for the observed differences in final OD and fatty acid content (compare with Figure 2G).
In the limitation experiments, a reduction of the phosphate concentration led to a 20% increase in OD and a 10% increase in fatty acid content. Since a reduction of the phosphate concentration severely decreased the buffer capacity of the medium, the pH was manually adjusted to that of the control culture. Without this pH adjustment, the final OD and fatty acid content were very low and similar to a culture with an initial pH of 6.8 (Figure 2F), which is probably due to a rapid decrease of the pH value (data not shown). In the following experiments, enhanced levels of phosphate (2- and 4-times of the standard concentration) were investigated to enhance the buffer capacity, but were not found to be suitable to improve cell growth or fatty acid accumulation. In fact, the growth rate in the first 40 hours was reduced, and the onset of TAG storage (marked by a second increase in OD between 40 and 70 hours; see Figure 2E) was delayed in the cultures with the highest phosphate concentration. Since the osmotic concentration of the cultures with 13.6 g l-1 phosphate (450 mOsm l-1) is about the same as in the cultures with 120 g l-1 sucrose concentration (488 mOsm l-1; Figure 2B) and the cultures containing NaCl to get an osmotic concentration of 450 mOsm l-1 (Figure 2G), the delay in growth should not be caused by the enhanced osmotic concentration.
A possible reason for the slower growth rate and delay in accumulation could be a regulatory role of the phosphate ions. In E. coli it has been shown that phosphate concentrations above 37 mM (equal to 3.5 g l-1) trigger the maintenance of cellular viability in the stationary phase (Schurig-Briccio et al. 2008), and in Clostridium perfringens the phosphate concentration influences the onset of sporulation (Philippe et al. 2006). In R. opacus the intracellular phosphate concentration could be involved in determining the energy status and thus inhibit the TAG synthesis.
Flask experiments with different initial pH values have shown that a good fatty acid production is possible when the initial pH is in the range of 7.1 to 8 (Figure 3E). In the first 32 hours, a strong decrease of the pH was observed, and the main fatty acid biosynthesis in the flask experiments seems to occur at a pH higher than 5. Lower pH values, as in the culture with an initial pH of 6.8 are not suitable for a continued growth or fatty acid accumulation (Figures 2F and 3E).
Overall this study provided further insights in the nutritional requirements of the lipid producing bacterium R. opacus. The results can be used for the optimization of the fermentation process in stirred tank reactors with the aim to produce high amounts of fatty acids as an alternative source for biodiesel or renewable diesel. In this regard the high salt tolerance is of importance, since the addition of strong acids or bases for a pH control during fermentation leads to an increase of the osmotic pressure of the medium.