Optimization of fermentation conditions for recombinant lipase expression 2 at small scale using response surface methodology for preliminary 3 production in two bioreactor platforms.

9 Background: Pseudomonas lipases are widely used in industrial applications due to its unique biochemical properties, 10 but one of the biggest limitations are the low yields obtained in native strains therefore, organisms as E. coli are used 11 for the recombinant lipase overexpression. However, the recombinant lipase is accumulated as inclusion bodies and 12 it affects biological activity, making that researchers evaluate different fermentation conditions to improve the 13 activity of recombinant enzymes. In this study, a statistical experimental design was implemented to evaluate the 14 effects of temperature, agitation rate and osmolyte concentration on the recombinant lipase activity produced in E. 15 coli BL21 (DE3). Once the significant variables were identified, an optimization by a Response Surface Methodology 16 was applied to maximize the lipase production. 17 Results: The Box-Behnken designs revealed different optimal fermentation conditions for each osmolyte experiment. 19 The glycerol showed the highest specific lipase activity compared to the other osmolytes and 0.1 M of osmolyte 20 glycerol,5°C and 110 rpm showed the highest significant increase on the specific lipase activity and the data fitted 21 the model very well. The validation showed that 452.01 U/mg of specific lipase activity was obtained which was 22 significantly higher compared to the group where no glycerol was added (271.38 U/mg). The relative recombinant 23 lipase expression was 2.7-fold lower at 5°C compared to 25 °C, but at 5°C the lipase activity was significantly higher. 24 In addition, when the 3 L shaken Erlenmeyer Bioreactor power input parameter, the specific lipase activity was not significantly different from that found in Schott (408,4 26 U/mg and 452 U/mg, respectively), which means that this Bioreactor platform should be used for future scale-up 27 processes. 29 Conclusion: Low temperatures, low agitation rates and 0.1 M of glycerol in the autoinduction media enhanced the 30 activity of the recombinant lipase produced in E. coli BL21(DE3). The optimized conditions and the 3 L shaken 31 Erlenmeyer Bioreactor can be used to produce the recombinant enzyme in a higher volume based on the power input 32 parameter. Further studies using this strategy may lead to the identification of optimal culture conditions for a given 33 recombinant enzyme facilitating the large-scale bioprocess implementation.


4
, and more effective than full factorial Designs, because BBD do not contain points in which the variables are 79 simultaneously at the highest or lowest levels and use less number of runs [10,23]. BBD allows identifying 80 significant variables in the process and the optimization of the response using response surface methodology 81 (RSM) [23]. RSM is an efficient tool to optimize the culture parameters and it has been used to improve the

91
But, low temperatures have demonstrated to be one of the most influencing parameters to increase the 92 solubility and enzyme activity of different unstable proteins in E. coli [30]. Most of the studies have worked with 93 temperatures above 23 °C for recombinant E. coli strains grown in auto-inducing media. Nevertheless, this is 94 the first report that enhances the recombinant lipase activity in E. coli BL21 (DE3) grown at temperatures below 95 10 °C using auto-inducing media. The expression changes of the recombinant lipase at low temperatures were 96 also investigated using quantitative real-time PCR (q PCR-RT). This method was chosen to quantify the relative 97 lipase expression versus the 16 S reference gene due to its specificity, accuracy, and sensitivity. The ribosomal 98 RNA 16 S housekeeping gene was used because it is necessary for cell survival and is considered stable in all 99 microorganisms, even under various experimental treatments [31]. Because most of the recombinant lipase 100 studies have mainly focused on cloning and expression of the target protein, which is usually not correctly 101 expressed in the cell, there is still a need for improving the recombinant lipase activity produced in E. coli BL21 102 (DE3) using optimization strategies [32].

103
The optimal culture conditions identified by RSM that enhanced the specific lipase activity in E. coli BL21(DE3) 104 at small scale were used for producing the recombinant enzyme in the two bioreactor platforms at large scale.   Figure 1S). According to the previous result, the ZYM-505 medium was chosen 126 for the inoculum preparation because it enhanced bacterial growth. We also found that cells are in the 127 logarithmic phase between 3-4 hours of incubation, which is the best growth phase in the inoculum for the  x 10 9 with a specific growth rate of 0.16 h -1 . As expected, the temperature had a significant effect on the growth 139 kinetics (Additional File 1: Table S2). With lower temperatures, the lag phase and the specific growth rate are 140 reduced . At the end of the E. coli growth at the three temperatures, cells achieve the stationary phase 141 (Additional File 1: Figure 2S).

143
As expected with mesophilic microorganisms, temperature not only affected growth rates of E. coli BL21 (DE3),

152
The total recombinant lipase production and the lipase activity were also affected by temperature. The total 153 production of the recombinant lipase at 5°C was 1.5 mg. When the temperature was increased from 5°C to 15°C 154 / 25°C, a rise of 0.8 mg (1.5-fold) and 3.5 mg (3.3-fold) was obtained, respectively. The semi-quantification on 155 the SDS-PAGE gel showed that 1.5 mg and 2.3 mg of total recombinant lipase are produced by 220 mg and 350 156 mg of wet biomass when cells are grown at 5°C and 15 °C, respectively and up to 5 mg of the target enzyme are 7 lipase / wet biomass) was higher at 25°C compared to 5°C (1.16 and 0.68%, respectively) However, these 159 results were obtained from the total extract, it means that the amount of recombinant lipase produced in the 160 soluble and non-soluble fraction was used for the quantification. Our results in the soluble fraction showed that 161 decreasing the temperature from 15°C to 5°C the enzyme activity is increased in 11,10 U/mL (16 %), and 162 decreasing the temperature from 25°C to 5°C the enzyme activity is increased in 25,31 U/mL (47 %) ( Figure   163 2A).
164 Figure 2B shows the monitoring of enzyme activity during the E. coli cultures grown at 5 °C, 15 °C and 25 °C.

165
During the whole culture, the recombinant lipase activity was higher in cells grown at 5°C compared to those at 166 15°C. The results obtained demonstrate that reducing temperatures enhances the recombinant lipase activity 167 produced in E. coli BL21 (DE3).

169
To determine the effect of osmolytes on the specific lipase activity, different osmolytes were added to the 170 autoinduction media and the results were compared with a control group where none osmolyte was present 171 ( Figure 3). The one-way ANOVA test demonstrated that the osmolytes significantly affected lipase specific 172 activity with a p-value < 0.05. To determine which osmolyte had a significant effect a Dunnett test was 173 performed (Additional File 1: Table S4-S5). The results showed that when glycerol and glycine are added to the 174 culture media the specific lipase activity increase by 7 % and 16 %, respectively, but the increase was not

206
The p-value= 0.04 from the glycerol and sorbitol models indicates that the model explains the data behaviour.

207
While the p-value= 0.052 for glycine implied that the model was not significant. The lack-of-fit values of 0.056, 208 0.054 and 0.249 for glycerol, sorbitol and glycine models indicate that were not significant ( To understand the interactions of fermentation conditions and to find the optimal conditions required for 211 maximum specific lipase activity, 3-D response surface curves were plotted. Response Surface graphs allowed 212 to establish a relationship between the response variable and the experimental levels of the significant 213 variables. Figure    which was close (difference of 11.1%) to the predicted value 301.4 U/mg (Additional file: Figure 4S).

226
Additionally, the validated experiments showed that when glycerol was added to culture media, the specific  Once the recombinant lipase was obtained using the optimal fermentation conditions, a semi-quantification was 237 performed. We found that the amount of lipase produced in the soluble fraction was very low and high amount 238 of the enzyme was produced in the insoluble fraction as inclusion bodies (Additional file: Figure

254
The q PCR-RT performed to investigate the differences in lipase expression when E. coli BL21 (DE3) was grown 255 at 25°C, 15°C and 5°C (Control) showed that levels of lipase mRNA were markedly elevated when the 256 recombinant lipase was produced at 25°C and 15°C compared with 5°C (2.7-fold and 1.3-fold, respectively).

257
These measurements confirm the low production of the recombinant lipase observed in SDS-PAGE gels at low 258 temperatures and suggest that the lipase gene is highly expressed at higher temperatures compared to the 259 lower ones ( Figure 5). It is crucial to highlight here that even though the expression level of the recombinant 260 lipase at 5°C is lower compared to 15°C and 25 °C E. coli cultures, the specific lipase activity was higher at 5°C 261 11 than 25°C as reported previously (Figure 2) , confirming that the enzyme productivity (U/mg) is higher at low 262 temperatures.

264
Once the optimal fermentation conditions were validated in Schott, we used the same optimal conditions to 265 produce the recombinant lipase in two different bioreactor platforms to evaluate if the enzyme productivity 266 and bacterial growth at the end of the culture remained the same as in Schott. The first platform, a 3 L shake 267 flask Erlenmeyer and the second one a 2 L Bioreactor. x 10 8 , respectively) and a higher the specific growth rate compared to Schott (0.003 h-1 and 0.001 h -1 , 274 respectively). On the other hand, the specific enzyme activity was slightly lower in Shake Flask Erlenmeyer 275 compared to Schott (408,4 U/mg and 452,0 U/mg, respectively). However, this difference was not significant, 276 which means that in terms of enzyme activity, the optimal culture conditions could be effectively used in shaken 277 flask bioreactors with higher volumes, up to 3 L ( Figure 6) (Additional File 1: Table S12-S12).

278
Concerning the production of the recombinant lipase in the 2 L Bioreactor, we found a lower number of viable 279 colonies compared to the Schott (1,32 x 10 8 and 1,76 x 10 8 , respectively) and consequently a significant 280 reduction in the specific lipase activity from 452 U/mg in Schott to 337,9 U/mg in the 2 L Bioreactor (Figure 7).

281
Even though the specific growth rate was similar in Bioreactor and Schott (0.002 h -1 and 0.001 h -1 , respectively), 282 the viability was much lower than in Schott. hand, the inoculum's logarithmic phase was selected because in the main cultures, the lag phase is shortened 290 and exhibits the maximal growth rate [41].

291
It is widely known that the yield and specific growth rate of many bacteria are affected by temperature [42].

292
The optimal growth temperature for E. coli BL21 (DE3) is 37 °C with a μmax = 1.05 h -1 , however, our study 293 performed experiments below the optimal temperature to enhance the activity of the recombinant enzyme [43].

294
Our results showed that low temperatures exhibited lower specific growth rates compared with higher 295 temperatures, this is explained because the reaction rates are lowered, conducing to decrease growth and

315
Bacterial culture temperature is one of the most crucial factors that must be optimized for increasing the 316 recombinant enzyme activities and yields because it influences cell growth and protein synthesis [49]. Our 317 study showed that the optimal temperature for glycerol and sorbitol experiments was the lowest temperature 318 5°C, probably due to different events occurring at this condition like induction of cold-shock proteins (CspA)

322
Another variable evaluated was the osmolyte concentration, that did not show a significant effect in the glycerol 323 and sorbitol models. However, when we analyzed separately the osmolyte effect on the specific lipase activity 324 we found that glycerol and glycine increased significantly the response variable. Glycerol may have stabilized

337
The agitation rate was another variable optimized in this study. We found that low agitation rates (110 rpm)

345
The optimization results showed that the lipase expression at the lowest temperature and agitation rate, and    is increased [67] 381 In conclusion, our study results suggest that the optimization of fermentation conditions in E. coli significantly 382 enhanced the enzyme productivity in terms of specific lipase activity. Low temperatures, low agitation rates 383 and 0.1 M of glycerol maximized the specific lipase activity. Even though, the recombinant lipase relative 384 expression was lower at low temperatures compared to high temperatures, the lipase activity was higher when 385 E. coli BL21(DE3) was grown at 5°C compared to 25°C. On the one hand, when the recombinant lipase was 386 produced in the shaken Erlenmeyer Bioreactor using the same optimal culture conditions for Schott 387 experiments, the enzyme productivity was not significantly different, suggesting that using this bioreactor can 388 be a useful alternative for lipase production for industrial purposes. On the other hand, it is essential to consider 389 the type of impeller for Bioreactor and to establish a threshold to avoid shear stress during cultivation to 390 maintain the lipase productivity and cell growth identical to the small scale. This study represents a reliable and

437
The total recombinant lipase production yields were calculated for each temperature from the densitometric 438 analysis in the Image J software (mg of total recombinant lipase/ mg wet biomass).  Table S1). The predicted lipase specific activity was calculated using first and second-order

554
In brief, the real-time PCR was performed in a CFX96 touch system (Bio-rad, California, United States) using the

562
The optimal annealing temperatures of the LipA and 16S primers were evaluated and the appropriate target

568
The test was used to choose the appropriate RNA concentration in the original samples for 16 S and 569 LipA genes. NTC were run as a negative control of the reaction for each set of primers.

571
Recombinant lipase production and growth curves in two Bioreactor platforms

572
The optimal fermentation conditions obtained from the BBD that showed the highest specific lipase activity 573 were used in the 2-L bioreactor (IKA EUROSTAR 200 P4 Control) and the 3 L Shake Flask Bioreactor to produce 574 23 the recombinant lipase in higher volumes. Both bioreactors were inoculated with 10% of the total effective 575 volume. For the 2 L Bioreactor containing 1,6 L of autoinduction medium, 10% of the pre-inoculum was 576 inoculated in 160 mL of ZYM medium to prepare the inoculum. This inoculum was used to inoculate the 2 L 577 bioreactor that was maintained at 5°C, 131 rpm for 120 h.

578
For the Shake Flask Bioreactor containing 600 mL of autoinduction medium, 10% of the pre-inoculum was 579 inoculated in 60 mL of ZYM medium to prepare the inoculum. This inoculum was used to inoculate the Shake 580 Flask Bioreactor that was maintained at 5°C, 162 rpm, for 120 h. The specific lipase activity was calculated at 581 the end of both cultures and cell growth was monitored during the whole cultivation.

582
In this work, the power input (P/V) in shake flasks was assumed to be the same as reported by Gamboa and

655
Error bars are presented in percentage.

678
Erlenmeyer and 2 L Bioreactor. Table S12. ANOVA Test to evaluate the effect of flasks on specific lipase activity at the three different flasks 679 schott, 3 L shaken Erlenmeyer and 2 L Bioreactor. Table S13. Dunnet Test to evaluate which flask did show significative differences in the 680 specific lipase activity at the three different flasks schott, 3 L shaken Erlenmeyer and 2 L Bioreactor.