Adjustment of the pH control settings in the BioLector® XT
Introduction: Challenges in the pH control of microbial cultivation processes
More and more high-performance production strains are being used in microbial cultivation processes. These strains have an ideal pH value, however, they have different requirements to maintain this value. For example, in E. coli fermentations a mixture of acetate, lactate, succinate, formate and ethanol are produced to maintain redox balance. Maintaining a stable pH value during the whole process is challenging and requires a rapid and precise pH control of the cultivation system. Therefore, an adjustable pH control for individual cultivation conditions is very essential. The BioLector® XT provides features to change the pH-control settings according to the requirements of the cultivation process. It combines the scalable BioLector® technology with a microfluidic chip. The benefit of the BioLector® XT is the combination of high throughput batch or fed-batch cultivations in a micro titer plate (MTP) format with online non-invasive monitoring of biomass, fluorescence, DO and online control of pH. 32 wells are available per microfluidic MTP which can be individually controlled for pH and feeding. Thus, all important cultivation process features are combined on a single use plate.
This technical note provides an overview about the functionality of the pH control in the BioLector® XT. The document provides instructions for carrying out a successful pH controlled microbial cultivation process in the BioLector® XT and also reflects on the influence of individual pH control settings on cultivation processes. Please refer Technical Note 002 MTP-Optodes, for a detailed understanding of the pH measurement using the BioLector®. Kindly request your copy via support@m2p-labs.com.
Functionality of the pH control in the BioLector® XT
The pH control in the BioLector® XT is based on a closed control loop of the micro valve pump strokes in the microfluidic MTP. Consequently, it controls the amount of base or acid being pumped into a cultivation well for pH correction.
The closed loop controller consists of a proportionally active p-component (p-band) and the integrally active I-component (integrator). The p-band describes the linear dependency between the actual value of the pH and the correcting variable. In general, lowering the p-band or using higher values for the maximum pump volume (max. volume [μL/cycle]) will result in more acid or base being pumped into the well, which further would result in a faster and stronger response to the pH control. Thus, it is recommended to use small p-band values for highly buffered culture media. Using p-band values that are too low will result in an overshooting of the pH setpoint and hence a poor pH-control.
There is no fixed assignment between the control deviation and the correcting variable for the integrator of the closed-loop controller. Lowering the integrator will result in less acid or base being pumped into the well resulting in a slower and weaker response. The same effect occurs by increasing the p-band. Therefore, the integrator is set to 1.0 by default and it is recommended only to adjust the pH control by changing the p-band and the max. volume.
Adjusting the pH control parameters in the BioLection software
The pH control configuration settings can be accessed and modified via the BioLection software. As a first step, synchronization between the BioLector® XT and the computer is necessary. Then, click on the Experiment Setup tab and the Microfluidics tab in the BioLection software as shown in figure 1.
Figure 1: Main screen of the BioLection 3 software showing the click direction to the pH control configuration menu.
Next, the Microfluidic pH Control Configuration menu opens (see figure 2). This menu includes the default pH control configuration with the opportunity for a modification of the controller settings.
Figure 2: Default settings for the microfluidic pH control in the BioLection 3 software.
The BioLection software provides three default configuration profiles. Alternatively, custom control profiles can be created via the + symbol and the profile name can be renamed if desired. The default profiles have the following configuration settings:
• The slow control profile with a maximum volume of 8.00 μL/cycle and a p-band of 2.00.
• The medium control profile with a maximum volume of 12.0 μL/cycle and a p-band of 1.00
• The fast control profile with a maximum volume of 15.0 μL/cycle and a p-band of 0.05
Besides, the default settings contain the start volume (0.2 μL), the dead band (0.05) and the fluid pressure (50000 Pa). The minimal volume (min. volume) per cycle is preset to 0.1 μL/cycle and it is recommended to avoid any modifications of this setpoint in all pH control profiles. The maximum volume per cycle defines how fast the response of the pH control will be. The higher the max. volume is set the faster is the response of the pH-control. However, the higher the volume, the greater is the risk of overshooting the pH-setpoint.
Accordingly, the correct pH control settings for the pH control must be found in advance and need to be adjusted before the experiment protocol is set up. The microbial strain and media conditions have an impact on the pH control because of strain specific acid production or diverse buffer concentrations in the cultivation media. After the pH control configurations have been adjusted, the experiment protocol can be created with the preferred pH control profile. For more information regarding the protocol setup, please refer to chapter 5 of the BioLector® XT user manual or to Tech Note 003.
E. coli batch cultivation example using default medium pH-control profile
The general functionality of the pH-control in the BioLector® XT based on an E. coli batch cultivation experiment is shown in figure 3. Here, the pH control starts after one hour. It is important to await this wetting period since the optodes need to adapt to the medium. The pH value is set to pH 7.00 by adding base (volume B, 3 M NaOH) into the well. The pH setpoint is set to pH 7.00 with a dead band of ± 0.02. As long as the pH value is within this range, the pH control is inactive. That means the addition of acid (volume A) or base (volume B) is stopped. Over the cultivation time the biomass increased exponentially for 10 hours. During the exponential phase more acid is produced by the microorganisms for example due to acetate formation. Thus, more base is needed to maintain the pH setpoint. With the beginning of the stationary phase, the microbial acid production stops. This leads to a pH overshooting by the added base in the stationary phase. Immediately, the acid is pumped into the well to adjust the pH value to the defined pH setpoint. In order to avoid the risk of overshooting, the process parameters can be adjusted with regard to the p-band or the max. volume per cycle pumped into the well.
Figure 3: E. coli BL21 wild type batch cultivation experiment in Wilms-MOPS containing 10 g/L glycerol, 50 mM buffer concentration, using a two-sided pH-control with default medium pH-profile settings. The cultivation took place in a microfluidic FlowerPlate® (MTP-MF32-BOH1) with V0 = 800 μL and n = 1200 rpm at 37 °C using the BioLector® XT.
E. coli batch cultivation examples using modified pH-control profiles
The impact of different p-band values at a fixed max. volume (1 μL/cycle) on the pH during an E. coli batch cultivation is shown in figure 4. With a p-band of 0.1, the pH setpoint is kept within the specified pH value range. The higher the selected p-band, the slower is the response in the pH control. A p-band value of 0.01, 0.1 and 0.5 lead to a fast response, within 15 minutes after the pH-control has started. However, a p-band of 0.01 leads more likely to an overshoot. In comparison, the response of the pH-control with a p-band value of 5 is slower, as a result the pH setpoint is only reached after 1.5 hours. A decreasing drift of the pH value towards the end of the cultivation period is visible when to much acid is produced by the microorganisms, however, the pH control is too slow caused by the higher p-band value and thus a slower response.
Figure 4: Graphical representation of a customized pH-control using 3 M NaOH and 3 M HCl and different p-band values at a fixed max. volume of 1 μL/cycle. The E. coli batch cultivation took place in Wilms-MOPS containing 10 g/L glycerol, 50 mM buffer concentration in a microfluidic FlowerPlate® (MTP-MF32-BOH1) with V0 = 800 μL and n = 1200 rpm at 37 °C using the BioLector® Pro.
This effect of a poor pH control towards the end of the exponential phase can be avoided by choosing a higher max. volume per cycle. In figure 5, the influence of different p-band values at another fixed max. volume (10 μL/cycle) on the pH during an E. coli batch cultivation using Wilms-MOPS cultivation medium with 50 mM buffer concentration is shown. The pH is in range even using a p-band of 0.5. Besides, the drift in the pH value is less significant at a p-band of 5 because a higher max. volume per cycle leads to a faster response in the pH-control.
Figure 5: Graphical representation of a customized pH-control using 3 M NaOH and 3 M HCl and different p-band values at a fixed max. volume of 10 μL/cycle. The E. coli batch cultivation took place in Wilms-MOPS containing 10 g/L glycerol, 50 mM buffer concentration in a microfluidic FlowerPlate® (MTP-MF32-BOH1) with V0 = 800 μL and n = 1200 rpm at 37 °C using the BioLector® XT.
An alternative option to ensure a more stable pH during a batch cultivation is the use of higher buffered cultivation media. Figure 6 shows the scattered light signal and the pH curve using Wilms-MOPS cultivation medium with 200 mM buffer concentration. For the pH-control different p-bands at a max. volume of 1 μL/cycle are chosen. It is shown that the overshooting using a p-value of 0.01 can be avoided completely compared to the pH-control in a lower buffered medium (figure 4). Furthermore, the pH drift at p-bands of 0.5 and 5 are less significant due to the higher buffer concentration leading to a more stable pH value in the cultivation medium.
Figure 6: Graphical representation of a customized pH-control using 3 M NaOH and 3 M HCl and different p-band values at a fixed max. volume of 1 μL/cycle. The E. coli batch cultivation took place in Wilms-MOPS containing 10 g/L glycerol, 200 mM buffer concentration in a microfluidic FlowerPlate® (MTP-MF32-BOH1) with V0 = 800 μL and n = 1200 rpm at 37 °C using the BioLector® XT.
Comparison of the default pH control settings in an E. coli fed-batch cultivation process.
For fed-batch cultivations only one row of reservoir wells on the MTP can be used for pH regulation because the other is used for feeding. Thus, the decision for either base or acid has to be made before starting a cultivation experiment. For microorganisms like E. coli, which are known to produce acidic by-products when grown on glucose or glycerol, choose a base.
The application of the one-sided pH control in an E. coli fed-batch experiment is shown in figure 7. In the upper graph the scattered light signal and the feeding volume (3 μL/h of 500 g/L glycerol) are plotted against the cultivation time. The pH control is performed with both 1 M NaOH or 3 M NaOH using the default pH-control profiles (slow, medium and fast) in both cases. In the scattered light signal the effect using different NaOH concentrations is visible. It shows a relatively lower scattered light signal for cultures with less concentrated NaOH solution used for pH control compared to cultures using higher concentrations. This effect is caused by a stronger dilution since more base volume has to be pumped into the well. During the exponential phase there is an increase in microbial acid production coupled to the microbial growth. This leads to an increase in the amount of base needed. When using 1 M NaOH, the pH setpoint cannot be maintained and a decreasing pH drift is visible. In comparison, when 3 M NaOH is used, the pH drift below the setpoint is less significant. On the other hand, there is a higher overshoot of the pH value using 3 M NaOH compared to 1 M NaOH. This causes a faster stabilization of the pH value in cultures controlled with 1 M NaOH at the beginning of the feeding phase. In addition, the slower the pH control profile is selected, the faster the pH value is stabilized regardless of the NaOH concentration. During the exponential phase, it is shown that the slower the pH control profile is chosen the stronger the pH value drifts below the setpoint. In the case of the slow pH control profile using 1 M NaOH a drift down to pH = 6.82 is visible. In comparison, the drift of the slow pH control profile using 3 M NaOH is less distinctive, but reaches pH = 6.9. After a cultivation time of 10 hours it is visible that an overshooting in the pH value towards the basic range is more intense when a faster pH profile setting is chosen. Furthermore, the higher the NaOH concentration the higher the overshoot. The fast pH profile using 3 M NaOH for example reaches a pH value of 7.16. In comparison, using 1 M NaOH in combination with the fast pH profile leads to a less intense overshoot up to pH = 7.05.
Figure 7: Fed-batch cultivation of E. coli in Wilms-MOPS (containing 10 g/L glycerol and 50 mM buffer concentration) with constant glycerol feed (3 μL/h) and one-sided pH-control with default slow, medium and fast pH-control settings) using 1M and 3 M NaOH. The cultivations took place in a microfluidic FlowerPlate® (MTP-MF32-BOH1) with V0 = 800 μL and n = 1200 rpm at 37 °C using the BioLector® XT.
In conclusion, the BioLector® XT offers the user plenty of flexibility for pH control. The user can choose between different default settings or even create customized pH-control profiles for specific experimental conditions. In general, the needs of the culture used and the existing cultivation conditions must always be taken into account in advance. The pH control must be adapted to the cultivation conditions in order to obtain successful cultivation results and to avoid undesirable side-effects such as biomass dilution or overshooting of the pH value.
