Improvement and Scaling Up of Rhamnolipid Production by aPseudomonas aeruginosa Isolate

Abstract

In this study, twenty four Pseudomonas isolates, recovered from thirty soil samples, were screened for RL production using SW agar. Six of these isolates were found to produce RLs, with isolates P4 and P6 showing the largest diameters. These two isolates were selected to be tested for RL production in MSM containing 2% w/v glucose. Higher RL levels were obtained by isolate P6 which was therefore selected for completing the study and identified as P. aeruginosausing 16S rRNA gene sequencing. A relatively limited number of isolates was screened in this study since we managed to recover a promising RL producing P. aeruginosa isolate with a productivity comparable to that of a previously recovered P. aeruginosa isolate obtained through screening of about 1900 isolates in a former study conducted in our laboratory. Several factors affecting RL production by P. aeruginosa isolate P6 in shake flasks were then tested. Upon studying the time course for RL production by P. aeruginosa isolate P6 using MSM containing 2% w/v glucose, a maximum RL concentration of 2.5 g/L was obtained after six days of incubation. The effect of media components and environmental fermentation conditions were then investigated for the aim of optimizing RL production by the selected isolate. Although high RL levels were obtained using 2% v/v soybean oil in SMSM, maximum RL concentration of 6.75 g/L was obtained using 2% v/v glycerol in MSM, called GMSM. Dual carbon source media resulted in lowerRL productivities when compared to media containing one tested carbon source. Concerning the nitrogen source, optimum sodium nitrate concentration for maximum RL production was found to be 2.5 g/L, corresponding to a C/N ratio of 23.83. The time course profile revealed that RL production in GMSM was partly growth associated since production still increased in the stationary phase. By using the method of experimental factorial design and RSM analysis, the optimal fermentation conditions for maximum RL production in batch culturein shake flasks were found to be a temperature of 30°C, a pH of 7.5, an agitation rate of250 rpm and inoculum size of 5% v/v, yielding a maximum RL concentration of 7.54 g/L. ANOVA results for the models revealed that all the tested factors had a significant effect on RL production and biomass, especially inoculum size and pH, and that the model equations were significant and could adequately be used to describe both responses. Regarding fed batch culture using P. aeruginosa isolate P6, shake flask studies revealed that the best setup involved an initial medium containing 0.5 ml glycerol (half the original glycerol concentration) and a feed consisting of 0.5 ml glycerol and 0.125 ml of 0.5 g/ml sodium nitrate (half the original concentration) added every time these substrates were depleted. These results suggest that the growth of P. aeruginosa P6 and RL production were best enhanced by a mixture of substrates in the feed. Hence this setup was selected to be attempted on the fermentor. Furthermore, gamma ray mutagenesis successfully improved the RL productivity by about 1.3 fold reaching 9.62 g/L by the mutant 15GR. The RL values obtained by the parent isolate P6 and its mutant were higher than most shake flask studies cited in literature, therefore, P. aeruginosa isolate P6 and its mutant 15GR can be considered to be promising microrganisms for large scale production of RLs. RL production of the parent isolate P6 and its mutant was subsequently tested in a 14 L laboratory fermentor using GMSM as the production medium and maintaining the temperature at 30°C. Upon studying the kinetic profilesof cell growth and RL production by the parent isolate P6 in batch fermentation, a behavior parallelto that obtained on the shake flask level was obtained, but at higher levels in case of the fermentor. Maximum RL production of 8.64 g/L while maximum biomass of 3.2 g/L was obtained at day 6 of fermentation. Glycerol was consumed at a high rate in the bacterial growth phase followed by a slower rate in the stationary phase, and was fully depleted at the end of the run. Moreover, DO percentage decreased sharply during the growth phase, then increased again during the stationary phase, while pH increased slightly during the growth phase to 7.4, which is related to nitrogen substrate metabolism, then declined to reach the initial level. Different factors were studied to optimize RL production on the fermentor using the parent P. aeruginosa isolate P6. Optimum fermentation conditions in batch fermentation were found to be inoculum size of 5% v/v, initial aeration rate of 1 vvm, temperature of 30°C, initial pH of 7.5 and agitation rate of 350 rpm. These optimized conditions resulted in maximum biomass of 2.97 g/L, maximum RL production of 10.38 g/L at day 6 andthe best fermentation parameters (YP/X = 3.69 g/g; YP/S = 0.46 g/g; YX/S = 0.135 g/g and PV = 0.072 g/L/h) when compared to other batch fermentation runs carried out in this study. Upon testing different fermentation modes in the fermentor, fed batch fermentation using the setupregimenthat proved to be the best in shake flasks resulted in a maximum biomass of 3.85 g/L, maximum RL production of 13.61 g/L and a 12 % improvement of YP/Xreaching 4.14 g/g. Semicontinuous fermentation was also attempted in the fermentor where one fourth of the culture medium was replaced with fresh medium after three days of fermentation. Although this strategy resulted in lower maximum biomass of 2.23 g/L, RL levels (10.85 g/L) were nearly the same as that obtained using batch fermentation. This resulted in a 41% increase in RL yield to reach the highest YP/X of 5.22 g/g. However, volumetric productivities of both fed batch and semicontinuous modes were lower than that obtained using batch fermentation due to the longer fermentation time required in these runs. Batch, fed batch and semicontinuous culture modes in the fermentor were also carried out using P. aeruginosamutant 15GR. Batch fermentation led to a lower biomass of 2.71 g/L, higher RL levels of 11.56 g/L and yields (YP/X and YP/S) and PV which were about 1.2 fold higher than those obtained by the parent isolate.In fed batch fermentation, using the same setup that resulted in maximum production by the parent, the highest RL level of 15.9 g/L was obtained in a shorter time than that obtained by the parent. This enhanced YP/X to reach 5.18 g/g which was about 1.3 fold greater than that obtained using the parent isolate under fed batch fermentation.Furthermore, semicontinuous mode using a different strategy than that used for the parent isolate was attempted, where one fourth of the total volume was replaced but after maximum RL levels were reached (after 6 days) and again after 10 days. This resulted in a maximum RL production of 14.67 g/L, maximum biomass of 2.7 g/L and the highest YP/X of 5.73 g/g. After that, kinetic models for cell growth, product formation and substrate consumption by the parent isolate P6 and its mutant 15GR were established. Results showed that the proposedmodels were capable of accurately predicting theexperimental results pertaining to cell growth, glycerol utilization and RL production, with R2 ≥ 0.98. Finally, in an attempt to further improve RL yields, RL production by P. aeruginosa mutant 15GR using SSF was carried out. A mixture of sugarcane bagasse and sunflower seed mealproved to be the optimum substrate, resulting in a maximum RL production of 28.25 g/L IS (56.5 g/kg IDS) after 6 days of incubation at 30°C, using GMSM as impregnating solution. A time course profile was carried out to determine the time required for maximum RL production and was found to be 10 days, resulting in 31.65 g/L-IS (63.3 g/kg-IDS). This value is over 3.5 fold higher than that obtained using SLF in shake flasks, which proves the efficiency of SSF.Increasing the glycerol concentration in impregnating solution to 5% v/v further improved RL yields to 37.25 g/L IS (74.5 g/Kg IDS). On the basis of RSM, it was possible to select the experimental conditions that led to maximum RL production using SSF, which were found to be an inoculum size of 1% v/v, temperature of 30°C and initial pH of 8, resulting in a RL concentration of 46.85 g/L IS.ANOVA results revealed that these 3 factors had a significant effect on RL concentration, especially temperature, and that the significant model equation derived by RSM could adequately be used to describe the RL production by SSF.Repeating the time course profile using these optimized conditions revealed that RL production showed a very similar profile to that obtained under unoptimized conditions, where maximum production was attained at day 10, but this time reaching 46.85 g/L IS. This value was about 5.5 fold higher than that obtained by SLF in shake flasks (8.45 g/L) and about 3 fold higher than the highest RL production obtained by SLF on the fermentor (15.9 g/L) using the tested mutant. Taken together, it can be concluded that the present study succeeded in: (i) recovery of a promising P.aeruginosa RL producing isolateand improving its RL productivity by random mutation, (ii) achieving high RL productivity at two SLF scales, shake flask and laboratory fermentor, for both the parent isolate and its improved mutant, (iii) developing two microbial mass culture methods exemplified by SLF and SSF for RL production by the improved mutant and (iv) mathematical description of the RL production process in the fermentor using batch mode for both the parent isolate and its mutant in terms of biomass formation, RL production and substrate consumption. The results obtained provide clear evidence and solid rationale for the future ongoing scale up of the RL production process by the parentP. aeruginosa isolate P6 and/or its improved mutant 15GR.