Biotransformation of ricinoleic acid into g-decalactone by Yarrowia lipolytica : bioprocess optimization

The biotechnological production of γ-decalactone (a peach-like aroma compound) by biotransformation of ricinoleic acid carried out by microorganisms is an interesting process to produce the aroma with a “natural” label, which is valuable, considering the preference of consumers. Although there are many works described in the literature about this subject, several factors in the process remain to fully understand and, consequently, to optimize. One of these factors is the effect of oxygen in the overall process. Thus, this work initially aimed to study the oxygen mass transfer phenomenon from gas to the biotransformation medium, an oil-in-water emulsion stabilized by a non-ionic surfactant, Tween 80. The oil is simultaneously the substrate of the process and it works as an oxygen carrier, since the solubility of this compound is higher in the oil than in the aqueous phase. The influence of each operation parameter (aeration rate and presence and concentration of surfactant agent and organic phase) on the variables involved in the oxygen transfer (gas-liquid interfacial area, a; liquid-side mass transfer coefficient, kL; and volumetric mass transfer coefficient, kLa) was analyzed in a bubble column and in an airlift reactor. Results demonstrated that in the bubble column the increase of aeration rates is positive for both gas-liquid interfacial area and mass transfer due to the increase of turbulence and gas hold-up. The surfactant concentration had a positive effect on the interfacial area since it reduced the gas bubbles size and it had a negative effect upon kL because its molecules are located at the gas-liquid interface, obstructing the oxygen mass transfer. Regarding the oil concentration, it had a negative effect upon the interfacial area but it improved kL, since it causes a new distribution of surfactant in the medium, decreasing its concentration in the gas-liquid interface. The overall result was a negative effect of the organic phase upon kLa. In the airlift reactor, it was observed that the increase of the aeration rates had a negative effect on kL. This was attributed to differences in the liquid distribution inside the airlift reactor. Since the main goal of this work was to optimize the production of γ-decalactone, two different ricinoleic acid sources (methyl ricinoleate, MR, and castor oil, CO) were tested, in different concentrations, as substrates of the process. Moreover, different cell inoculation strategies were attempted, differing among each other in the washing or not of the cells. The results revealed that the use of non-washed cells is more beneficial for the aroma production, independently of the substrate used; and a concentration of 30 g L-1 MR was the most adequate among the range tested, since it allowed the highest γ-decalactone productivity (14.9 mg L-1 h-1). This substrate revealed also to be a lipase inducer. The use of CO as substrate of the process allowed to achieve almost 2 g L-1 of aroma but the process was rather slow, resulting in low productivities. It was then hypothesized an insufficient oil hydrolysis and an enzymatic hydrolysis was attempted with different commercial enzymes and operating conditions (temperature and pH). Lipozyme TL IM, pH 8 and 27 °C were selected as the most efficient lipase and operating conditions, respectively, to hydrolyze CO. The results obtained using CO previously hydrolyzed by the selected lipase were compared with the results obtained in experiments in which the enzymatic hydrolysis occurred during the biotransformation and in experiments without adding lipase, indicating that the process was faster when lipase was involved in any form, but the aroma concentrations were lower, resulting in similar productivities. The droplets size of both oils was characterized by laser granulometry in emulsions with different oil concentrations. The impact of the presence of cells on droplets size was also analyzed as well as the relevance of washing inoculum cells. The granulometry of emulsions was related with γ-decalactone production and it was observed that, in the presence of non-washed cells, the smaller droplets disappeared, with both oils, which increased γ-decalactone concentration, suggesting that the access of cells to the substrate occurs by their adhesion around larger oil droplets. Experiments in a stirred bioreactor using 30 g L-1 MR (concentration at which the highest aroma productivity was achieved) and different aeration and agitation rates demonstrated the direct influence of oxygen transfer rate on the production of γ-decalactone and of another compound, 3-hydroxy-γ-decalactone, that can also accumulate in the medium. The accumulation of this compound indicates a deviation in the metabolic pathway of γ-decalactone production, decreasing its yields. A response surface methodology was used to optimize pH (6.17) and dissolved oxygen concentration (44.4%) for the aroma production. These operating conditions were applied in two fed-batch strategies: with constant medium feeding rate and with intermittent feeding. Both strategies were compared with the traditional batch mode in terms of overall productivity and yield in respect to the substrate. Although the productivity was considerably higher in the batch mode, the level of substrate conversion to both lactones was greater in the intermittent fed-batch, allowing the accumulation of high aroma concentrations (6.8 g L-1 γ-decalactone and 10.0 g L-1 3-hydroxy-γ-decalactone). Finally, the production of aroma was attempted in an airlift bioreactor due to the advantages of this type of bioreactor, mainly in terms of high power economies, the non-mechanical agitation which avoids damage to cells and the higher mass transfer coefficients attained. The highest γ-decalactone production was obtained at an air flow-rate of 1 L min-1. The aeration rate increase of 5-fold lead to lower aroma concentrations. However, the time needed to reach the peak of production was also reduced, resulting in higher productivities.