Respiration activity monitoring system for any individual well of a 48-well microtiter plate

Abstract

Background: Small-scale micro-bioreactors have become the cultivation vessel of choice during the first steps of bioprocess development. They combine high cultivation throughput with enhanced cost efficiency per cultivation. To gain the most possible information in the early phases of process development, online monitoring of important process parameters is highly advantageous. One of these important process parameters is the oxygen transfer rate (OTR). Measurement of the OTR, however, is only available for small-scale fermentations in shake flasks via the established RAMOS technology until now. A microtiter plate-based (MTP) μRAMOS device would enable significantly increased cultivation throughput and reduced resource consumption. Still, the requirements of miniaturization for valve and sensor solutions have prevented this transfer so far. This study reports the successful transfer of the established RAMOS technology from shake flasks to 48-well microtiter plates. The introduced μRAMOS device was validated by means of one bacterial, one plant cell suspension culture and two yeast cultures.

Results: A technical solution for the required miniaturized valve and sensor implementation for an MTP-based μRAMOS device is presented. A microfluidic cover contains in total 96 pneumatic valves and 48 optical fibers, providing two valves and one optical fiber for each well. To reduce costs, an optical multiplexer for eight oxygen measuring instruments and 48 optical fibers is introduced. This configuration still provides a reasonable number of measurements per time and well. The well-to-well deviation is investigated by 48 identical Escherichia coli cultivations showing standard deviations comparable to those of the shake flask RAMOS system. The yeast Hansenula polymorpha and parsley suspension culture were also investigated.

Conclusions: The introduced MTP-based μRAMOS device enables a sound and well resolved OTR monitoring for fast- and slow-growing organisms. It offers a quality similar to standard RAMOS in OTR determination combined with an easier handling. The experimental throughput is increased 6-fold and the media consumption per cultivation is decreased roughly 12.5-fold compared to the established eight shake flask RAMOS device.

Keywords: High throughput; Microtiter plate (MTP); Oxygen transfer rate (OTR); RAMOS; μRAMOS.

Fig. 1

a Schematic overview of the developed μRAMOS system for measuring the oxygen transfer rate in each single well. b Single well schematic cross section of the microfluidic MTP cover for OTR measurements. Valves, oxygen sensors, optical fibers, oxygen measuring instruments and capillaries are only shown for one well. All valves as well as the mass flow controller are controlled by a computer. The oxygen partial pressures of the headspaces of every well are detected via oxygen measuring instruments and monitored and analyzed by the computer. The MTP is mounted on an orbital shaker. The capillaries ensure an equal air flow through every well of the MTP. By applying overpressure through the pressure inlets, the elastic switching membrane seals the air inlet and outlet to realize an interruption of the air flow during the cultivation. The fluorescence sensor spot is glued onto the lower side of the cover facing the well headspace. The optical fiber is connected to an oxygen measuring instrument via an optical multiplexer on one end and plugged in the microfluidic cover facing the fluorescence sensor spot on the other end. Due to the transparent material, the oxygen dependent fluorescence behavior of the immobilized fluorescence spot can be excited and non-invasively detected through the material

Fig. 2

Picture of the applied μRAMOS system for measuring the oxygen transfer rate in each single well. Single well schematic cross section of the microfluidic MTP cover for OTR measurements is according to Fig. 1B. The microtiter plate is mounted on an orbital shaker. The fluorescence sensor spot is glued onto the lower side of the cover facing the well headspace and excited via red light. The optical fiber is connected to an optical multiplexer on one end (Fig. 4) and plugged in the microfluidic cover facing the fluorescence sensor spot on the other end

Fig. 3

Illustration of the switching mechanism of the developed μRAMOS technology. a Switch position of the 5/2 valve for providing a under pressure and (b) over pressure above the switching membrane. Two sets of air pumps, capillaries and 5/2 valves are used for switching all 48 inlet valves independently of the 48 outlet valves. Differential switching pressures: - 300 hPa ((a) open valve (blue)), + 600 hPa ((b) closed valve (red)). The pressures are adjusted by the air resistances of the capillaries and the air pump characteristics

Fig. 4

CAD drawing side view (a) and schematic bottom view (b) of the optical multiplexer (8 × 48) for measurements of the oxygen partial pressure in each well. Eight oxygen measuring instruments are rotated against 48 fixed optical fibers. The rotation sequence for oxygen partial pressure measurement through all 48 fibers consists of five 7.5° steps and a reversed – 37.5° step. Complete cycle time is roughly 3.5 s. Only the first 5 cm of the optical fibers are shown in (a) for clarity

Fig. 5

Characteristic segment of a measured single well oxygen partial pressure propagation during cultivation of H. polymorpha RB11 pC10-FMD (P_FMD-GFP). (I) Phase of standard aeration with air (flow phase), 0.8 vvm, 16 min. (II) Phase of closed inlet and outlet valves and determination of the OTR (stop phase), 0 vvm, 3 min. (III) Phase of elevated aeration with supply air to compensate for the prior interruption of air supply (4.8 vvm, high flow phase). The presented oxygen partial pressure decrease during phase (II) corresponds to an OTR of 21.5 mmol L^-1 h^-1. Due to the dynamic equilibrium of oxygen supply and consumption of oxygen, the oxygen partial pressure during phase (I) is reduced and roughly 193 hPa (instead of 209.5 hPa for pure air)

Fig. 6

Comparison of E. coli BL21 (DE3) pRotHi-YFP cultivations using the newly introduced μRAMOS MTP system (─■─) and the standard RAMOS shake flask system (─○─). Mean values of the measured oxygen transfer rates of 48 wells and 4 shake flask cultivations are shown (a). b Detailed propagation of corresponding standard deviations of the measured oxygen transfer rates of the 48 wells and 4 shake flask and cultivations. Cultivation medium: Synthetic Wilms-MOPS auto-induction medium with 0.55 g L^-1 glucose, 2 g L^-1 lactose and 5 g L^-1 glycerol. μRAMOS (MTP) cultivation conditions: 48-well Round Well Plate without optodes, V _L = 800 μL, n = 1000 rpm, shaking diameter d ₀ = 3 mm, 37 °C, flow phase + high flow phase: 17 min, stop phase: 3 min., standard RAMOS cultivation conditions: 250 mL RAMOS shake flask, V _L = 10 mL, n = 350 rpm, d ₀ = 50 mm, 37 °C, flow phase + high flow phase: 25 min, stop phase: 5 min

Fig. 7

Respiration activity of parsley cell suspension cultures in μRAMOS (MTP) and RAMOS (shake flask) treated with salicylic acid (SA) and a 13 amino-acid defense elicitor of Phytophthora sojae (Pep13). Arrows indicate the addition of 100 μM salicylic acid (SA, 72 h) and of a 13 amino-acid defense elicitor of Phytophthora sojae (Pep13, 96 h). Shadows in (a) symbolize standard deviations of eight cultivations. (b) Single shake flask cultivations. Cultivation medium: Gamborg B5 medium with 20 g L^-1 sucrose. Cultivation conditions μRAMOS (a) 48-well Round Well Plate without optodes, V _L = 2000 μL, n = 800 rpm, shaking diameter d ₀ = 3 mm, 25 °C, flow phase + high flow phase: 20 min, stop phase: 10 min, standard RAMOS cultivation conditions (b) 250 mL shake flask, V _L = 50 mL, n = 180 rpm, d ₀ = 50 mm, 25 °C, flow phase + high flow phase: 20 min, stop phase: 10 min

Fig. 8

μRAMOS (a) and RAMOS (b) cultivation of H. polymorpha RB11 pC10-FMD (P_FMD-GFP) under magnesium limitation. b republished data of Kottmeier et al. (2010) [33]. The percentages of magnesium are normalized to the original medium containing 3.0 g L^-1 MgSO₄ · 7H₂O as 100 %. Mean values and corresponding standard deviations (colored shadows) of triple cultivations are shown in (a). Single cultivations are shown in (b). For clarity, only every second data point over time is indicated by the corresponding symbol in all curves. The cultivations containing 0.8 % magnesium were shifted for -1 h due to possible inoculation variances. Cultivation medium: Synthetic Syn-6-MES medium with 10 g L^-1 glycerol. μRAMOS cultivation conditions (a): 48-well Round Well Plate without optodes, V _L = 800 μL, n = 1000 rpm, d ₀ = 3 mm, 30 °C, flow phase + high flow phase: 16 min, stop phase: 4 min., standard RAMOS cultivation conditions (b): 250 mL RAMOS shake flask, V _L = 10 mL, n = 300 rpm, d ₀ = 50 mm, 30 °C, flow phase + high flow phase: 20 min, stop phase: 10 min

Fig. 9

48 different μRAMOS (a) and 8 different RAMOS (b) cultivations of H. polymorpha RB11 pC10-FMD (P_FMD-GFP) under potassium limitation. b republished data of Kottmeier et al. (2010) [39]. The percentages of potassium are normalized to the original medium containing 1.0 g L^-1 KH₂PO₄ and 3.3 g L^-1 KCl as 100 %. Single cultivations are shown. For clarity, only the concentrations of the RAMOS cultivations are indicated by a corresponding symbol within the μRAMOS results. Cultivation medium: Synthetic Syn-6-MES medium with 10 g L^-1 glycerol. μRAMOS cultivation conditions (a): 48-well Round Well Plate without optodes, V _L = 800 μL, n = 1000 rpm, d ₀ = 3 mm, 30 °C, flow phase + high flow phase: 16 min, stop phase: 4 min, standard RAMOS cultivation conditions (b): 250 mL RAMOS shake flask, V _L = 10 mL, n = 300 rpm, d ₀ = 50 mm, 30 °C, flow phase + high flow phase: 20 min, stop phase: 10 min