Small-scale fed-batch cultivations of Vibrio natriegens: overcoming challenges for early process development

Abstract

Vibrio natriegens is a fast-growing microbial workhorse with high potential for biotechnological applications. However, handling the bacterium in batch processes is challenging due to its high overflow metabolism and mixed acid formation under microaerobic conditions. For early process development, technologies enabling small-scale fed-batch cultivation of V. natriegens Vmax are needed. In this study, fed-batch cultivations in 96-well microtiter plates were successfully online-monitored for the first time with a µTOM device. Using the online-monitored oxygen transfer rate, a scale up to membrane-based fed-batch shake flasks was performed. The overflow metabolism was efficiently minimized by choosing suitable feed rates, and mixed acid formation was prevented. A glucose soft sensor using the oxygen transfer rate provided accurate estimates of glucose consumption throughout the fermentation, eliminating the need for offline sampling. Analyzing the impact of the inducer IPTG on the recombinant production of the enzyme inulosucrase revealed concentration-dependent effects in batch processes. In contrast, fed-batch operating mode resulted in high inulosucrase activity even without induction. Overall, an inulosucrase titer of 80 U/mL was achieved. In conclusion, the advantages of small-scale fed-batch technologies supported by a glucose soft sensor have been demonstrated for early process development for V. natriegens Vmax.

Keywords: Vibrio natriegens; Fed-batch; Glucose soft sensor; Microtiter plate.

Fig. 1

Schematic illustration of the devices used in this work for microtiter plate and shake flask cultivations in fed-batch. a µTOM device used to monitor the oxygen transfer rate (OTR) in 96-well microtiter plate experiments (adapted from [35]). b Single well of a FeedPlate® (adapted from [61]). Glucose crystals are embedded in a silicone matrix. Upon contact with a liquid medium, water penetrates into the silicone matrix, the glucose crystals gradually dissolve, and the glucose solution is released into the medium. c Picture of the µTOM device. d Illustration of a RAMOS flask used with a RAMOS device [36] to monitor the oxygen transfer rate. e Illustration of a membrane-based fed-batch shake flask (adapted from [55]). The reservoir contains a highly concentrated glucose feed solution, separated from the culture broth by a membrane-covered diffusion tip. A flexible tube allows the tip to rotate with the bulk liquid in the flask. Thus, during shaken cultivations, the membrane-covered diffusion tip is always in contact with the culture. Therefore, glucose is released into the culture broth. Due to osmosis, a weak flow of water enters the feed reservoir in reverse direction to the diffusion of the glucose. Proteins and cells are held back in the culture broth because they cannot pass through the membrane due to the molecular weight cut-off. f Picture of the membrane-based fed-batch shake flask

Fig. 2

Calibrating the oxygen transfer to glucose consumption in a 96-deep well plate in a batch cultivation. Non-induced batch cultivation of V. natriegens Vmax pET19b::inuGB-V3 in modified Wilms-MOPS medium (0–5 g/L glucose) with 400 mM MOPS buffer. Initial OD₆₀₀ 0.5, 37 °C, 1000 rpm at 3 mm shaking diameter, oxygen transfer rate monitored using a µTOM device. a 100 µL filling volume in a 96-deep well plate. For clarity, only every third data point is shown as a symbol. Shadows indicate standard deviation for n = 4 replicates. b Linear correlation of the total oxygen consumption (shown in Fig. S3a) to the provided and total consumed glucose. The stars represent online data from microtiter plate cultivations from Forsten et al. [67]. Error bars are hardly visible, as they are mostly within the size of the symbols (Fig. S3b). 95% confidence band of the linear fit (dashed line) is shown as a red shadow

Fig. 3

Application of the calibration from Fig. 2c, d to determine the glucose consumption in a fed-batch cultivation in microtiter plates. Non-induced fed-batch cultivation of V. natriegens Vmax pET19b::inuGB-V3 in modified Wilms-MOPS medium (no additional initial glucose) with 400 mM MOPS buffer. Initial OD₆₀₀ 0.5, 37 °C, 1000 rpm at 3 mm shaking diameter. The oxygen transfer rate was monitored using a µTOM device. 100 µL or 200 µL filling volume in 96-well FeedPlate® (SMFP04001, high release). For washing step see chapter Main culture in batch and fed-batch. a Oxygen transfer rate over time. b Total oxygen consumption over time. c Comparison of glucose consumption determined from the total oxygen consumption (hatched bars) after 18.7 h (see Fig. 2b), technical data from the manufacturer (empty bars) and final OD₆₀₀ (dotted bars). For clarity, only every third data point is shown as a symbol. Shadows indicate minimum/maximum for n = 2 replicates

Fig. 4

Membrane-based fed-batch cultivation in shake flasks with different glucose feeding rates. Non-induced fed-batch cultivation of V. natriegens Vmax pET19b::inuGB-V3 in modified Wilms-MOPS medium (no initial glucose) with 400 mM MOPS buffer. 10 mL filling volume in 250 mL RAMOS flasks, initial OD₆₀₀ 0.5, 37 °C, 350 rpm at 50 mm shaking diameter, the oxygen transfer rate was monitored using a RAMOS device. a Oxygen transfer rate over time for different glucose concentrations in the reservoir. For clarity, only every third data point is shown as a symbol. Shadows indicate minimum/maximum for n = 2 biological replicates. The replicate cultivations were conducted independently on different days. b Glucose feed rates corresponding to different reservoir concentrations: either the feed rate was offline measured through HPLC analysis (open bars) or determined via the linear fit (hatched bars) presented in Fig. 2b. Error bars indicate minimum/maximum for n = 2 biological replicates. (c) Offline determined final pH and OD₆₀₀

Fig. 5

Influence of the inducer concentration (IPTG) on inulosucrase production in a batch process in shake flasks. V. natriegens Vmax pET19b::inuGB-V3 in modified Wilms-MOPS medium (20 g/L glucose) with 400 mM MOPS buffer. 8 mL filling volume in 250 mL RAMOS flasks, initial OD₆₀₀ 0.5, 30 °C, 350 rpm at 50 mm shaking diameter. The oxygen transfer rate was monitored using a RAMOS device. Induction using 0, 0.25, or 0.5 mM IPTG at OD₆₀₀ 1.5. a The oxygen transfer rate over time: the arrow indicates the induction time. For clarity, only every third data point is shown as a symbol. Shadows indicate the minimum/maximum for n = 2. b Total glucose consumption after 19 h, determined via linear fit in Fig. 2b. Error bars indicate minimum/maximum for n = 2 replicates. Dotted line: 20 g/L glucose was added to the medium. c Inulosucrase activity after 19 h (determined offline in triplicates). Statistically significant differences were determined via a two-sided t-test, * p < 0.05, ** p < 0.01, *** p < 0.005. d Enzyme yield was calculated using 20 g/L glucose. Error bars were calculated via Gaussian error propagation

Fig. 6

Influence of the inducer concentration (IPTG) on inulosucrase production in a fed-batch process in shake flasks. V. natriegens Vmax pET19b::inuGB-V3 in modified Wilms-MOPS medium (no initial glucose, 300 g/L glucose in reservoir) with 400 mM MOPS buffer. 10 mL filling volume in 250 mL RAMOS flasks, initial OD₆₀₀ 0.5, 30 °C, 350 rpm at 50 mm shaking diameter. The oxygen transfer rate was monitored using a RAMOS device. Induction using 0, 0.25, or 0.5 mM IPTG at OD₆₀₀ 1.5. a Oxygen transfer rate over time, arrow indicates the time of induction. For clarity, only every third data point is shown as a symbol. Shadows indicate the minimum/maximum for n = 2 replicates. b Total glucose consumption after 19 h, determined after linear fit in Fig. 2b. Error bars indicate minimum/maximum for n = 2 replicates. c Inulosucrase activity after 19 h (determined offline in triplicates). d Enzyme yield is calculated from the values in b, c. Error bars were calculated via Gaussian error propagation

Fig. 7

Inulosucrase expression during membrane-based fed-batch cultivation in shake flasks. Fed-batch cultivation of V. natriegens Vmax pET19b::inuGB-V3 in modified Wilms-MOPS medium with no initial glucose with 400 mM MOPS buffer. Induction with 0.25 mM IPTG at OD₆₀₀ 1.5. 10 mL filling volume in 250 mL RAMOS flasks, 300 g/L glucose concentration in feed reservoir, initial OD₆₀₀ 0.5, 37 °C, 350 rpm at 50 mm shaking diameter. The oxygen transfer rate was monitored using a RAMOS device. a Left axis: Oxygen transfer rate over time; a vertical arrow indicates the induction time. For clarity, only every fifth data point is shown as a symbol. Shadow indicates standard deviation for n = 3 replicates. Right axis: Glucose consumption determined via the linear fit presented in Fig. 2b (orange line). b Left axis: Enzymatic activity of the product inulosucrase measured offline. Error bars indicate standard deviation for n = 3 replicates. Right axis: Substrate yield per g of consumed glucose determined via linear fit. c Left axis: OD₆₀₀. Right axis: pH offline measured (blue symbols)