The influence of growth rate-controlling feeding strategy on the surfactin production in Bacillus subtilis bioreactor processes | Microbial Cell Factories

Chemicals and standards

Chemicals used in this study, if not otherwise stated, were purchased from Carl Roth GmbH & Co. KG (Karlsruhe, Germany). Surfactin standards (≥ 98% purity) were obtained from Sigma-Aldrich Laborchemikalien GmbH (Seelze, Germany).

Bacterial strain, media and conditions for fed-batch cultures

The Bacillus subtilis strain BMV9 (spo0A3; trp⁺; sfp⁺; ΔmanPA) was used in this study [5, 10]. In comparison to Klausmann et al. [5], the only difference to JABs32 is a removal of the erm^R resistance cassette from the manPA knockout region in BMV9. The media compositions used for precultures or fermentation processes were previously described by Klausmann et al. [5]. In brief, the first preculture was performed in LB-medium, while a chemically defined mineral salt medium was used for the subsequent second preculture as well as the final bioreactor fermentation culture [5].

The shake flask cultivations were carried out in an incubator shaker (NewbrunswickTM/Innova 44, Eppendorf AG, Hamburg, Germany) at 37 °C and 120 rpm. The bioreactor cultures were performed in a 30 L fermenter (ZETA GmbH, Graz/Lieboch, Austria) filled with 12 L batch medium. For protection against overfoaming, the bioreactor was connected with a foam trap described previously by Klausmann et al. [5]. The following parameters were set to a temperature of 37 °C, a pH value of 7.0 and an initial stirrer speed of 300 rpm using three Rushton turbines. Dissolved oxygen was adjusted to a minimum of 50% by adjusting the stirrer speed and aeration rate. After inoculation of 12 L of batch medium to an initial OD₆₀₀ of 0.1, the cells were cultured at constant parameters of 37 °C, a pH of 7 and an aeration rate of 10 L/min until glucose was depleted as the sole carbon source. The associated cellular adaptation of metabolism for consumption of acetate as an alternative carbon source produced in non-affecting concentrations during the batch phase led to a slower metabolic rate and a characteristic increase of pO₂. In this way, a real-time measurement was available for the identification of the feeding start within 1 min before the cell suspension entered the transient and stationary phase. This characterization and identification of the feeding start was previously described and experimentally established by Henkel et al. [21]. Afterwards, the initial aeration rate of 10 L/min was adjusted stepwise from 15 to 72 L/min when a 50% (w/w) glucose solution was added. The details of the cultivation process in both shake flask and one-step bioreactor fermentation have been described by Klausmann et al. [5]. One difference to be mentioned is the fact that instead of an additional supply of ammonia as nitrogen source, which was initially provided at 1 g/L in the batch medium, the further addition of ammonium was ensured via pH control by adding 20% (v/v) NH₃ solution, which was maintained at a constant level of around 1 g/L ammonium (Figure S1).

The initial feeding rate F₀ for the glucose feed was calculated directly after the batch phase according to the formula below. The initial feeding rate was used to calculate the feed rate F(t) at every time point (t) of the fed-batch.

In these equations, F₀ is the initial feed rate (kg/h); F(t) the exponential feed rate at every time point t (kg/h); µ the targeted growth rate set to 0.075, 0.15, 0.2, 0.25, 0.3 or 0.4 h^− 1; m the maintenance coefficient set to 0.05 g/(g*h); Y_X/S the conversion yield of glucose to biomass in the batch phase (g/g); C_{X, Batch} the biomass concentration at feed start (g/L); V₀ the bioreactor volume in the batch (L) and C_{S, Feed} the glucose concentration in the feed (g/L).

Sample analysis

The samples taken during cultivation were centrifuged at 3890 g for 10 min at 4 °C (Multifuge X3R, Thermo Fisher Scientific, Waltham, USA). The cell-free supernatants were used to quantify glucose with an enzymatic assay kit (R-Biopharm AG, Darmstadt, Germany), acetate with an enzymatic assay kit (R-Biopharm AG, Darmstadt, Germany) and ammonium with a photometric assay kit (Merck KGaA, Darmstadt, Germany). For the calculation of cell dry weight (CDW), samples (10 mL) from bioreactor approaches were taken. Cells were separated by centrifugation at 3890 g for 10 min at 4 °C. After washing in 9% (w/v) NaCl solution, cell pellets were dried at 110 °C for 24 h (memmert UF 110, Memmbert GmbH & Co.KG, Schwabach, Germany). After analysing 11 representative bioreactor samples from the fed-batch process, a mean correlation factor of 0.232 was calculated between CDW and the experimentally determined OD₆₀₀ values.

Surfactin quantification

Surfactin was only measured from the cultivation broth. Therefore, surfactin produced was quantified using high-performance thin-layer chromatography (HPTLC) (CAMAG AG, Muttenz, Switzerland). All experimental details were described by Geissler et al. [22]. In brief, 2 mL of the cell-free supernatant was used for a threefold extraction with chloroform/methanol (2:1). The pooled solvent layers were dried using a rotary evaporator (RVC 2–25 Cdplus, Martin Christ Gefriertrocknungsanlagen GmbH, Osterode am Harz, Germany) at 40 °C and 10 mbar. After dissolving in 2 mL methanol, the samples were applied in 6-mm bands to a silica HPTLC plate. Migration at a distance of 60 mm was performed with a mixture of chloroform/methanol/water (65:25:4) as mobile phase before surfactin detection at 195 nm [22]. A surfactin standard (Sigma Aldrich, Seelze, Germany) was used for quantification.

Data analysis

All experiments were carried out in biological duplicates. The yield of product per biomass (Y_P/X), product per substrate (Y_P/S), biomass per substrate (Y_X/S), specific productivity (q_P/X) and specific substrate-product conversion rate (q_P/S) were calculated for the feeding phase, excluding parameters from the batch phase, with the equations below. For these calculations, the parameters time (t), biomass (X), glucose as substrate (S) and surfactin as product (P) were used. More specifically, the time point of maximum surfactin concentration was used to determine the Y_P/S and Y_P/X values, while Y_X/S was determined at time point of maximum biomass formation.

Fitting of experimental data and production parameters

All graphs were generated using OriginPro 2022b (OriginLab Corporation, Northampton, United States) software. The data series for acetate, ammonium, CDW, glucose, and surfactin were analyzed utilizing polynomial functions. The production parameters were analyzed using the Simple Fit tool with existing functional equations (polynomial or exponential) to maximize the coefficient of determination.