*2.3. Fermentative Limonene Production in a Stirred-tank Reactor*

The first attempt to produce limonene with *E. coli* BL21 (DE3) pJBEI-6410 and glycerol as the sole carbon source in a two-liquid phase fed-batch setup was carried out in a 3.1 L reactor with 1 L of M9 minimal medium and 0.5 L of the organic carrier solvent diisononyl phthalate (DINP). A carbon limited exponential feed was applied with a calculated growth rate of 0.18 h−<sup>1</sup> (Figure 4A,B), which was based on a substrate specific biomass production rate obtained from an initial batch cultivation. Heterologous gene expression was induced with the addition of 0.1 mM IPTG after two hours of fed-batch cultivation. After 10 h of exponential growth, no further increase in biomass was observed. The growth rate for this period was 0.15 h−<sup>1</sup> and a final cell dry weight (CDW) of 28.7 g·L−<sup>1</sup> was achieved. No accumulation of glycerol was detected until this time point. Acetate formation was suppressed during growth. After 11 h, growth stopped, and the exponential feed was set constant. During the next 15 h, glycerol accumulated, followed by acetate formation, leading to final concentrations of 5.7 g·L−<sup>1</sup> glycerol and 2.8 g·L−<sup>1</sup> acetate. The ammonium concentration increased during the fermentation from 0.4 to 1.9 g·L<sup>−</sup>1, probably due to the pH regulation with ammonia as a result of acetate and carbon dioxide formation. The specific activity of limonene synthesis increased after induction with IPTG and reached a maximum of 2.6 U gCDW<sup>−</sup><sup>1</sup> by the end of exponential growth. However, significant limonene formation was still observed for the non-growing cells, and a final limonene concentration of 4.4 g·Lorg<sup>−</sup><sup>1</sup> was achieved. Concentrations of limonene were determined for the organic phase volume, because of the significant dilution of the aqueous phase due to the addition of the feed solution. The cell growth and the production of limonene seemed to be limited by a yet unidentified mechanism, which might be the accumulation of limonene or another compound related to the biosynthesis of limonene—such as an intermediate or metabolite—or a limitation caused by the depletion of a medium component.

**Figure 4.** Two-liquid phase fed-batch fermentation with *E. coli* BL21 (DE3) pJBEI-6410 in M9 minimal medium. Cell dry weight (CDW) (), limonene concentrations (-) in the organic phase, glycerol (-), acetate (), and ammonium () concentrations were determined at regular intervals. The specific activities (•) were calculated for distinct time points throughout the fermentation time. The feed rate is displayed as well (dotted line). (**A)** and (**B)** display the initial fed-batch fermentation (D = 0.18 h<sup>−</sup>1), whereas (**C**) and (**D**) display the optimized fed-batch fermentation with a lower feed rate (D = 0.15 h<sup>−</sup>1) and additional trace element supply. The error bars for CDW relate to two independent measurements.

To test whether intracellular limonene, terpene intermediates, or the expression of heterologous pathway genes influence cell growth, a glycerol-limited fed-batch fermentation was carried out without the induction of the heterologous pathway (data not shown). Although heterologous gene expression was not induced, 222 mg·Lorg−<sup>1</sup> limonene was produced during the exponential growth. These small amounts are probably caused by leaky expression of the pathway genes. Nearly identical CDW was achieved, while 20-fold less limonene was produced. Therefore, toxification by intermediates of the MVA pathway or intracellular limonene can be excluded.

In order to overcome the growth limitation, the parameters for the fermentation were changed to permit prolonged exponential growth, coupled with higher limonene concentrations. A carbon limited exponential feed was set up with a calculated growth rate of 0.15 h−1, which is lower than in the previous experiments (0.18 h−1) (Figure 4C,D). Furthermore, additional US\* trace elements were supplied. After 6 and 16 h of cultivation, 2 and 1 mL of US\* trace element solution were spiked, respectively. The exponential growth could be maintained for nearly 17 h, with a growth rate of 0.12 h<sup>−</sup>1. Thus, the addition of trace elements prolonged the growth phase, and a CDW of 48.8 g·L−<sup>1</sup> was reached, whereas the second trace element spike did not affect the cells in their stationary phase. The specific activity quickly increased reaching a maximum of 1.8 U gCDW−<sup>1</sup> after 5 h, before it decreased with ongoing fermentation. In contrast to in previous experiments, the specific activity could be maintained

above 1 U gCDW<sup>−</sup><sup>1</sup> for a time period of more than 17 h. The prolonged production phase and higher biomass concentration resulted in 7.3 g·Lorg−<sup>1</sup> limonene after 24 h by the end of the fermentation. To our knowledge, this is the highest limonene titer reported so far.
