2. Material and methods
2.1. Microorganism, media, and culture conditions
C. bombicola ATCC 22214 was obtained from the Guangdong Culture Collection Center (China) and stored at -80℃ in 20% glycerol solution. The seed medium consisted of 50 g/L glucose, 1 g/L KH2PO4, 4 g/L (NH4)2SO4, 0.5 g/L MgSO4·7H2O, and 10 g/L corn steep liquor (CSL). The seed was cultured in a 1 L baffled shake flask with 200 mL working volume at 200 rpm and 25℃ for 48 h.
The initial fermentation medium in a 5 L bioreactor (Shanghai Guoqiang Bioengineering Equipment Co., Ltd., China) consisted of glucose 100 g/L, KH2PO4 1 g/L, (NH4)2SO4 4 g/L, MgSO4·7H2O 0.5 g/L, CSL 10 g/L. All culture media were sterilized at 115℃ for 30 min. The initial volume of 2.5 L with an inoculum of 3% was cultured at 25℃ for 168 h. Aeration at 0.5 vvm and dissolved oxygen (DO) above 30% of the saturation concentration were maintained by adjusting the agitation in a stepwise manner. A pH of 3.5 was maintained by addition of 4 M NaOH solution, during the whole process. The fed-batch fermentation cycle was 168 h. During fermentation, glucose concentration in the broth was maintained at 30-80 g/L. Oil supplementation rate was 0.5 g/L/h for the first 24 h, after which oil concentration was maintained at 2-15 g/L, according to oil consumption. The semi-continuous fermentation period was 378 h, and SLs was intermittently separated by the separation unit every 72 h. The supplementation rate of glucose and rapeseed oil was optimized according to the sedimentation rate of SLs. After every two in-situseparations, 1/5 of the initial nutrition was supplemented.
2.2. SLs sedimentation mechanism
2.2.1. Influences of oil to SLs ratios on SLs morphology and sedimentation
The ratio of oil to SLs (O/S) (0.075, 0.10, 0.13, 0.15, 0.18, 0.20, 0.22, 0.23, 0.25, 0.30 g/g) was quickly detected and regulated by the low-field nuclear magnetic method (Chen et al., 2019). SLs morphology was observed by optical microscopy (BMDH200 microscope, Sunny Optical Technology Co., Ltd), and the main structures of SLs with different morphologies were analyzed by LC-MS (Chen et al., 2020).
Furthermore, 10 mL of the fermentation broth with different O/S ratios were processed by standing still to observe the SLs sedimentation, and the volume of lower layer was recorded at different time points until it was unchanged during 2 hours.
2.2.2. Measurements of broth viscosity and SLs particle size
Twenty milliliters of broth with different O/S ratios were obtained to measure the viscosity of supernatant with a viscosity meter (Brookfield DV-Ⅱ). Another 5 mL broth was thoroughly mixed, and the SLs particle size was analyzed by nanometer size and ZETA potential analyzer (Beckman Coulter Nano-ZS).
2.3. Effect of ultrasound assistance on SLs precipitation
2.3.1. Effect of ultrasonic ultrasound assistance on cell activity and SLs production
An appropriate amount of fermentation broth was obtained and treated at different ultrasonic times (0, 10, 20, 30, 40, 50, and 60 min) and power (100, 200, and 300 W) in a 10 L ultrasonic cleaning machine (SHT-70al, 40 kHz). And then, the treated broth was diluted at appropriate times and coated on the plate to observe the cell activity. Otherwise, the strains after different ultrasonic treatments were concentrated by centrifugation, resuspended, and inoculated into a medium contained only glucose and oil (50 mL of fermentation broth strains were inoculated into a 500 mL shake flask with 50 mL working volume). And compared strain growth and SLs production.
2.4. Semi-continuous fermentation based on ultrasonic assisted treatment
In semi-continuous fermentation, the optimal O/S was adopted to rational control the sedimentation of SLs for in-situ separation. According to our previously developed SLs separation equipment (Liu et al., 2019), the improved in situ separation platform was shown in Fig 2. The fermentation broth was rapidly pumped into the separation unit, which was placed in the ultrasonic machine, by a peristaltic pump at 5 mL/s. The ultrasonic power and time were controlled at 100 W and 10 min. Subsequently, the SLs was intermittently discharged from port c for 1 min, and the supernatant was refilled from port b to the fermenter until the SLs concentration in the fermenter was lower than 60 g/L. Finally, the separated crude SLs was washed by 1/2 proportion of sterile water (v/v), and thoroughly mixed by stirring. After standing for 20 min, the supernatant was pumped back to the fermenter to recover the biomass, glucose and oil. Due to the loss of biomass after the operation of SLsin-situ separation, medium and sterile water were replenished to the initial volume of 2.5 L.
2.5. Analytical methods
Glucose concentration was analyzed by an enzymatic bio-analyzer (SBA-40C, Shandong Academy of Sciences, China). Oil concentration in the broth was quickly determined by the low-field nuclear magnetic method (Chen et al., 2019). For oil content determination, three parallel broth samples were extracted twice, using the same volume of n-hexane. The upper layer was transferred to another tube and dried in an oven for 24 h, until constant weight. The bottom layer was washed twice with alcohol and dried at 80 ℃ for 24 h to measure dry cell weight (DCW). The concentrations and structures of SLs were determined by HPLC and HPLC-MS, respectively (Chen et al., 2020).
2.6. Data analysis
As the fermentation went on, SLs was gradually accumulated in the broth and once the O/S reached at a certain range, the SLs sedimentation was occurred. The average sedimentation rate was calculated by formula (1). By fitting the height of SLs sedimentation was directly proportional to the concentration of SLs, formula (2) was obtained (Fig. 4A). The sedimentation rate formula related to SLs concentration could be derived in (3).
\(\overset{\overline{}}{V}=\frac{H_{1}-H_{2}}{2t}\) (1)
\(\frac{H_{2}}{H_{1}}=0.0031\times c-0.0001\) (2)
\(\overset{\overline{}}{V}=\frac{H_{1}-H_{1}(0.0031\times c-0.0001)}{2t}\)(3)
Where \(\overset{\overline{}}{V}\) is the average sedimentation rate of SLs in cm/s, H1 is the height of broth in the 10 mL sedimentation tube in cm, H2 is the height of SLs settled in the 10 mL sedimentation tube in cm, t is settling time in s, c is SLs concentration in g/L.
In this study, the sedimentation mechanism and ultrasound assisted efficiency were analyzed by the Stokes law as shown in (4) (Yang et al., 2015).
\(V=\frac{2r^{2}(\rho_{2}-\rho_{1})g}{9\mu}\) (4)
Where V is the sedimentation rate of SLs particles under ideal condition in m/s, r is the radius of SLs particles in m, \(\rho_{1}\) is the density of broth in kg/m3, \(\rho_{2}\) is the density of SLs particles in kg/m3,g is the gravitational acceleration in m/s2\(\mu\) is the viscosity of broth in Pa·s.
Due to the consideration of working volume change by feeding, all the presented data in Figures and Tables were normalized to the initial volume. All experiments were performed in triplicate.