KR102382101B1 - Method and device for controlling protein production process with microbial fermentation - Google Patents

Abstract

The present invention relates to a method and an apparatus for controlling a protein production process by microbial fermentation using a near infrared spectrophotometer and chemometrics and, more specifically, to a method for simultaneously measuring and monitoring indices of multiple components of a fermented solution in real time utilizing a near infrared spectrophotometer and chemometrics when proteinase K is fermented using yeast (Pichia pastrois) microorganisms and automatically controlling a fermentation process in real time by using electrical signals of each component generated at this time and to an apparatus therefor. The present invention can perform definite process management and maximize productivity.

Description

Method and apparatus for controlling the protein production process by microbial fermentation

The present invention relates to a method and apparatus for controlling a protein production process by microbial fermentation using a Near Infrared Spectrophotometer and Chemometrics. More specifically, the present invention monitors and simultaneously measures and monitors several component indicators in the fermentation broth using a near-infrared spectrometer and chemometrics when fermenting Proteinase K using a yeast ( Pichia pastrois ) microorganism, and each component generated at this time It relates to a method for automatically controlling a fermentation process in real time by utilizing an electrical signal of the and an apparatus for the same.

The fermentation process is a process of inoculating a microorganism into a medium capable of its growth and production of the desired product material, and providing and managing an environment for the microorganism to efficiently produce the desired product material at the same time as growth through consumption of the medium component, Optimization of physical and chemical processes suitable for the physiology of microorganisms is essential.

Among the physical factors related to fermentation process optimization, agitation water, temperature, internal pressure, pH, dissolved oxygen, etc. have already been developed with high-reliability measuring sensors such as high-reliability measuring sensors. control is possible. However, in addition to the above factors normally applied to the current fermentation process management, changes in the concentration of major media components essential for real-time management or optimization of the fermentation process, the growth profile of microorganisms, and the generation trend of target products and by-products over time In case of change, no measuring device (sensor, etc.) or analysis method has been developed to analyze it immediately. For this reason, until now, fermentation process experiments and production site management depend on offline analysis methods or experiences that collect fermentation broth samples and analyze each component individually in a separate, dedicated analyzer. The application of real-time process control is very low compared to chemical processes. Using non-real-time analysis results of biochemical reactions by microorganisms has disadvantages in that efficiency is low in fermentation optimization or management, and a lot of time and manpower are required for analysis.

In addition, recently, microorganisms belonging to many types of Genetically Modified Organisms (GMOs) are being applied to research or production sites, and the level of regulation for the outflow of genetically modified microorganism strains before death to the external environment is becoming stricter, mainly Sampling of the culture medium collected for analysis of the culture process is recognized as one of the main outflow pathways that occur during microbial culture at the production or research site.

On the other hand, recent advances in electronics, information communication and optical devices and the rapid development of artificial intelligence (AI) chemometrics that can accurately and quickly process information on spectroscopy spectra have made it possible to expand the application of spectral analysis technology. Among them, the near-infrared spectroscopy technology can analyze several components continuously and simultaneously in a short time without sample pre-treatment, so multiple components can be analyzed simultaneously without sample pre-treatment in one analysis compared to the existing wet analysis method that has been analyzed for each component can do. Therefore, it is possible to significantly reduce analysis time, cost, and labor, as well as time used for pre-treatment and rapid quantitative analysis without the use of harmful chemicals. Conventionally, near-infrared spectroscopy technology has been mainly used for quality analysis of agricultural products such as non-destructive total analysis of sugar content of apples and citrus fruits (Korean Society of Industrial and Food Engineers Vol. 11, No. 1, p38~48; Journal of the Korean Chemical Society 2005, Vol.49, No. 6; Korean Society of Machine Tools Vol. 15, No. 5), process management of petrochemical processes (Korean Society of Near Infrared Spectroscopy 2001. June 01, P. 1082~1082; Korean Society of Near Infrared Spectroscopy 2002, Nov. 01 , p63-73), etc. have been used.

Korean Patent Application No. 2015-0077576 Korean Patent No. 1605995

An object of the present invention is to solve all of the problems of the prior art described above.

Another object of the present invention is to provide a calibration model capable of monitoring information on changes in fermentation broth components according to fermentation of microorganisms in real time.

Another object of the present invention is to automatically control the fermentation process in real time in response to real-time information on changes in fermentation broth components according to microbial fermentation.

Another object of the present invention is to be able to control the microbial fermentation outside the microbial fermentation system by monitoring the microbial fermentation in situ.

A representative configuration of the present invention for achieving the above object is as follows.

According to an aspect of the present invention, there is provided a method for real-time control of a protein production process by microbial fermentation, comprising the steps of (i) providing a calibration model obtained using two or more near-infrared wavelength regions specific for each of one or more fermentation broth component indicators , and (ii) monitoring each change in the fermentation broth component index with a light source in the two or more near-infrared wavelength regions is provided.

According to another aspect of the present invention, there is provided a fermentation system for producing a protein by microbial fermentation, (i) using two or more near-infrared wavelength regions specific for each of the fermentation broth component indicators to monitor a change in one or more fermentation broth component indicators. A fermentation system including a control unit for real-time monitoring and control of a protein production process by microbial fermentation is provided.

In addition to this, another method for implementing the present invention, another system, and a computer-readable recording medium for recording a non-transitory computer program for executing the method are further provided.

The method for real-time control of the protein production process by microbial fermentation according to the present invention monitors changes in key component indicators of the fermentation process in real time and automatically controls the fermentation process in real time by using the electrical signals of each component generated at this time. It has the effect of maximizing accurate process control and productivity. In addition, since the method of the present invention can analyze various components simultaneously in real time without sampling during fermentation culture, when genetically modified microorganisms are used in the fermentation process, the problem of external leakage of such microorganisms can be fundamentally solved.

1 is a schematic diagram showing a near-infrared spectrometer and a fermentation process.
Figure 2 shows the raw spectrum (Raw Spectrum) (Wavenumber: cm-1) of the Proteinase K fermentation broth.
3 shows the selected region (Calibration, Wavenumber: cm-1) of the Proteinase K (protein) concentration spectrum of the Proteinase K fermentation broth.
4 shows a proteinase K concentration measurement spectrum (1st Derivative) of a Proteinase K fermentation broth.
5 shows a calibration model of the proteinase K concentration of the Proteinase K fermentation broth.
6 shows a calibration model of validation of Proteinase K concentration of Proteinase K fermentation broth.
7 shows a selection region (Calibration Wavenumber: cm-1) of the concentration spectrum of glycerol, a carbon source of Proteinase K fermentation broth.
Figure 8 shows the measurement spectrum (first differential) of the concentration of glycerol, a carbon source of Proteinase K fermentation broth.
9 shows a calibration model of the glycerol concentration, a carbon source of Proteinase K fermentation broth.
10 shows a calibration model of validation of glycerol concentration, which is a carbon source of Proteinase K fermentation broth.
11 shows the selected region (Calibration, Wavenumber: cm-1) of the concentration spectrum of methanol, which is the carbon source of Proteinase K fermentation broth.
12 shows a measurement spectrum (primary differential) of methanol, which is a carbon source, of Proteinase K fermentation broth.
13 shows a calibration model of the methanol concentration, which is a carbon source of Proteinase K fermentation broth.
14 shows a validation calibration model of methanol concentration, which is a carbon source of Proteinase K fermentation broth.
15 shows a selected region (Calibration, Wavenumber: cm-1) of an absorbance (optical density) spectrum indicating the cell growth of Proteinase K fermentation broth.
16 shows an absorbance measurement spectrum (first differential) showing the cell growth of Proteinase K fermentation broth.
17 shows a calibration calibration model of absorbance indicating the cell growth of Proteinase K fermented broth.
18 shows a validation calibration model of absorbance indicating the cell growth of Proteinase K fermented broth.
19 is a calibration model to monitor Proteinase K concentration, carbon source glycerol and methanol concentration, and absorbance (OD) trend indicating the growth of bacteria in an online state by connecting a near-infrared spectrometer to the fermenter, and to manage the concentration of the carbon source The results of automatic input and control of phosphorus glycerol and methanol in real time are shown.

Hereinafter, the present invention will be described in more detail.

According to an aspect of the present invention, there is provided a method for real-time control of a protein production process by microbial fermentation, comprising the steps of (i) providing a calibration model obtained using two or more near-infrared wavelength regions specific for each of one or more fermentation broth component indicators , and (ii) monitoring each change in the fermentation broth component index with a light source in the two or more near-infrared wavelength regions is provided.

The present inventors have made diligent efforts to develop a method for monitoring and/or controlling one or more fermentation broth components in real time, such as microorganisms, proteins, media components, etc., present in the fermentation broth during microbial fermentation for protein production. As a result, each fermentation broth component is The present invention was completed by discovering two or more near-infrared wavelength regions that can be specifically detected/calibrating, and constructing a calibration model that can quantify the electrical signals of each component generated using the two or more near-infrared wavelength regions.

The method may further include the step of controlling the fermentation process in real time in response to the electrical signal of each component indicator generated through the monitoring.

In one embodiment, the monitoring may be performed through a near-infrared spectrometer attached to the fermentation tube in the form of a probe.

In one embodiment, the monitoring may be performed in situ without sampling of the fermentation broth.

The near-infrared spectrometer is attached to the fermentation tube in the form of a probe, so that the change in the component index of the protein fermentation broth can be monitored without sampling. For example, when a genetically modified microorganism strain that is undesirable to leak to the external environment is used, a closed process control (Contained Process Control) that monitors fermentation broth component indicators using a near-infrared spectrometer probe sealed to the fermentation tube as a measurement sensor (Contained) Process control, On-Tank or In-Line Sampling Free Fermentation Process Analytical Control may be used.

The microorganism may include any microorganism as long as it can produce a protein through fermentation from a carbon source. For example, the microorganism may be a microorganism belonging to a genetically modified organism (GMO) including living modified organisms (LMOs). Specifically, the microorganism may be bacteria or yeast. More specifically, the microorganism may be Pichia pastoris .

The protein may include any protein that has been engineered to be produced by a microorganism or to be produced by recombinant technology. In one embodiment, the protein may be Proteinase K.

As a carbon source of microbial fermentation for the production of the protein, glycerol, methanol, or a combination thereof may be used.

The microbial fermentation may be batch, continuous or fed-batch culture. Preferably, the microbial fermentation may be a continuous or fed-batch culture.

Fermentation broth components that need to be monitored in the microbial fermentation may be microorganisms, carbon sources, nitrogen sources, other organic components, inorganic components, and proteins to be produced used for microbial fermentation. The fermentation broth component indicator may be any index capable of determining a change in the concentration of the fermentation broth component in the fermentation broth. For example, it may be the concentration of each component present in the fermentation broth. In the case of a microorganism, it may be an absorbance indicating the growth or concentration of the microorganism.

In one embodiment, the microbial fermentation may be performed by continuous or fed-batch culture of yeast for Proteinase K production. Glycerol and methanol may be used as carbon sources for the microbial fermentation.

In this case, the fermentation broth component indicator may be one or more selected from the group consisting of Proteinase K concentration, glycerol concentration, methanol concentration, and microbial growth (absorbance).

As used herein, the near-infrared wavelength means a wavelength in the near-infrared (NIR) region of 800 to 2500 nm (12,000 to 4,000 cm-1) that exists between visible and mid-infrared rays.

In the method, the provision or monitoring of the calibration model may be performed using a near-infrared spectrometer.

In order to provide or monitor the calibration model, two or more near-infrared wavelength regions may be used in combination for one fermentation broth component indicator. Specifically, it may be desirable to use two near-infrared wavelength regions in combination to provide or monitor a calibration model of one fermentation broth component indicator.

In one embodiment, the two or more wavelength regions include a wavelength region of 9403.5 to 7498.1 cm -1 and a wavelength region of 6101.9 to 5446.2 cm -1 in the case of the Proteinase K concentration; In the case of the glycerol concentration, it includes a wavelength region of 8886.7 to 7745.0 cm -1 and a wavelength region of 5870.4 to 5420.0 cm -1 ; the methanol concentration includes a wavelength range of 8457.7 to 7498.1 cm-1 and a wavelength range of 6101.9 to 5446.2 cm-1; The microbial viability may include a wavelength region of 8535.3 to 7503.4 cm -1 and a wavelength region of 5797.2 to 5439.5 cm -1 .

The calibration model used in the method includes the steps of (a) collecting n samples of the fermentation broth at regular time intervals, (b) calibrating m samples out of the n samples collected using a near-infrared spectrometer. Measuring the raw spectrum for ), (c) performing physicochemical standard analysis on m samples for which the raw spectrum was measured, (d) statistically preprocessing the raw spectrum measured in step (b) Step, (e) Select two or more near-infrared wavelength regions specific for each of one or more fermentation broth component indicators to be quantified from the spectrum pretreated in step (d), and calibrate from the spectrum of the selected region through chemometrics constructing a calibration model, and (f) applying the results obtained by applying the nm samples not used for the raw spectrum measurement to the calibration calibration model developed in step (e), the physicochemical model obtained in step (c) can be obtained by a method comprising the steps of assaying and optimizing compared to standard analytical results.

Wherein n and m are natural numbers, and n is a natural number greater than m. Preferably, n and m may be natural numbers of 50 or more, 100 or more, 150 or more, 200 or more, 250 or more, 300 or more, 350 or more, 400 or more, 450 or more, 500 or more, but is not limited thereto. It can be any larger natural number to increase the accuracy of the calibration model.

In addition, the difference between n and m may be 10 or more, 50 or more, 100 or more, 150 or more, 200 or more, 250 or more, or 300 or more, but is not limited thereto.

The n and m are large numbers suitable for deriving a calibration model for statistical processing, and can be easily selected and used by a person skilled in the art.

As used herein, "chemometrics" is a generic term for mathematical statistical algorithms that enable qualitative and quantitative analysis by deriving desired information from various analytical chemistry data such as spectra or chromatograms. In chemometrics, multivariate analysis may be mainly used. In contrast to univariate analysis, multivariate analysis is a statistical technique that simultaneously analyzes multiple dependent variables for multiple independent variables. Statistical analysis of chemometrics includes Multiple Linear Regression (MLR), Principal Component Regression (PCL), Partial Least Squares (PLS), Modified Partial Least Squares, MPLS) may be used, and in particular, Partial Least Squares (PLS) may be used.

In this specification, the term “partial least squares method (PLS)” may be used interchangeably with the term “partial least squares regression (PLSR)”.

Through the calibration model, it is possible to monitor in real time the change of each of the indices of the fermentation broth components that are changed in the process of microbial fermentation in real time, and the fermentation process can be automatically controlled in real time.

According to another aspect of the present invention, there is provided a fermentation system for producing a protein by microbial fermentation, (i) using two or more near-infrared wavelength regions specific for each of the fermentation broth component indicators to monitor a change in one or more fermentation broth component indicators. A fermentation system including a control unit for real-time monitoring and control of a protein production process by microbial fermentation is provided.

In one embodiment, the fermentation system may further include (ii) a fermentation unit for microbial fermentation and (iii) a near-infrared spectroscopy unit.

In one embodiment, the control unit may automatically control the fermentation process in real time in response to an electrical signal of each component indicator generated by monitoring changes in the one or more fermentation broth component indicators.

In one embodiment, the fermentation unit may include, but is not limited to, any means, apparatus, etc. known in the art for microbial fermentation, such as a fermenter, a stirrer, a culture solution supply pipe, a gas pipe, etc. necessary for microbial fermentation.

In one embodiment, the near-infrared spectrometer may be mounted in the form of a probe in the fermentation unit. The near-infrared spectrometer is attached to the fermentation tube in the form of a probe, so that the change in the component index of the protein fermentation broth can be monitored in real time without sampling. For example, when a genetically modified microorganism strain that is undesirable to leak to the external environment is used, a closed process control (Contained Process Control) that monitors fermentation broth component indicators using a near-infrared spectrometer probe sealed to the fermentation tube as a measurement sensor (Contained) An on-tank or in-line (in-line) sampling-free fermentation process analysis management system that is process control may be used.

In one embodiment, the fermentation system may be hermetic. For example, the fermentation system may check the fermentation state of the protein production process by microbial fermentation through an on-tank measurement facility.

The microorganism may include any microorganism as long as it can produce a protein through fermentation from a carbon source. For example, the microorganism may be a microorganism belonging to a genetically modified organism (GMO) including living modified organisms (LMOs). Specifically, the microorganism may be bacteria or yeast. More specifically, the microorganism may be Pichia pastoris .

In one embodiment, the fermentation broth component indicator may be one or more selected from the group consisting of Proteinase K concentration, glycerol concentration, methanol concentration, and microbial growth (absorbance). For example, the one or more indicators of fermentation broth components may include a concentration of glycerol, which is a carbon source for microbial growth.

In one embodiment, the two or more wavelength regions include a wavelength region of 9403.5 to 7498.1 cm -1 and a wavelength region of 6101.9 to 5446.2 cm -1 in the case of the Proteinase K concentration; In the case of the glycerol concentration, it includes a wavelength region of 8886.7 to 7745.0 cm -1 and a wavelength region of 5870.4 to 5420.0 cm -1 ; the methanol concentration includes a wavelength range of 8457.7 to 7498.1 cm-1 and a wavelength range of 6101.9 to 5446.2 cm-1; The microbial viability may include a wavelength region of 8535.3 to 7503.4 cm -1 and a wavelength region of 5797.2 to 5439.5 cm -1 .

In one embodiment, the near-infrared spectroscopy unit includes a wavelength region of 9403.5 to 7498.1 cm -1 and a wavelength region of 6101.9 to 5446.2 cm -1 in the case of the Proteinase K concentration during the real-time monitoring; In the case of the glycerol concentration, it includes a wavelength region of 8886.7 to 7745.0 cm -1 and a wavelength region of 5870.4 to 5420.0 cm -1 ; the methanol concentration includes a wavelength range of 8457.7 to 7498.1 cm-1 and a wavelength range of 6101.9 to 5446.2 cm-1; In the case of the microbial growth, it may be set to emit a light source in a wavelength region including a wavelength region of 8535.3 to 7503.4 cm -1 and a wavelength region of 5797.2 to 5439.5 cm -1 .

Referring to Figure 1, the control unit is connected to the fermentation unit by wire or wireless communication, and receives the electrical signal of the fermentation broth component indicator in the fermentation unit obtained through the light source of the wavelength region emitted from the near-infrared spectrometer to be established. It is possible to monitor the index of the fermentation broth converted by the calibration model and control the carbon source supply in real time in response.

For terms related to the fermentation system, microbial fermentation, protein produced, fermentation broth component index, calibration model, and the like, reference may be made to a description of a method for real-time control of the protein production process by the microbial fermentation.

The embodiments according to the present invention described above may be implemented in the form of program instructions that can be executed through various computer components and recorded in a non-transitory computer-readable recording medium. The non-transitory computer-readable recording medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the non-transitory computer-readable recording medium may be specially designed and configured for the present invention, or may be known and available to those skilled in the computer software field. Examples of the non-transitory computer readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, magneto-optical media such as floppy disks ( magneto-optical media), and hardware devices specially configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware device may be configured to operate as one or more software modules for carrying out the processing according to the present invention, and vice versa.

Hereinafter, the present invention will be described in more detail by way of Examples. However, the following examples are only for illustrating the present invention, and the scope of the present invention is not limited thereto.

Now, the process for developing a calibration model of the major control factors applied to the present invention will be described in detail using Proteinase K fermentation as an example.

(P1) Proteinase K Concentration of Proteinase K in fermentation broth, glycerol and methanol concentrations as carbon sources, and acquisition of near-infrared absorption spectra related to cell growth (absorbance)

After taking 300 samples of the fermentation broth of the proteinase K fermentation process that are continuously performed sequentially at regular time intervals (2-hour intervals), 200 samples among them were collected from 12,000 cm-1 to 4,000 cm-1 using a near-infrared spectrometer. The raw spectrum was obtained by scanning in a constant wavelength unit of 2 nm in the wavelength region of . At this time, the passing distance of the light source was 2 mm, the measurement mode was a transmission reflection mode, and the FT-NIR device used a matrix F FT-NIR (FT-NIR Matrix-F Bruker Optics: Germany). Using the standard physicochemical analysis method for the same 200 samples, the concentration of glycerol as a carbon source was determined by high-performance liquid chromatography (HPLC), and the concentration of methanol as another carbon source was determined by gas chromatography (GC), and Proteinase K (Protein) concentration is the Bradford Protein Assay (Kit) method, and the cell viability was measured at 600 nm using a spectrophotometer and used as a standard sample analysis value for the development of a calibration model. The sample of the fermentation broth can be used by sequentially collecting and using the fermentation broth proceeding in a fed batch (Fed Batch) every 2 hours, and 300 samples were collected for broadening the calibration formula and developing a significant calibration model, in the present invention 24 to 30 samples were collected per fermentation cycle, 200 samples per each fermentation broth were used for raw spectrum measurement for the development of a calibration model, and the remaining 100 samples were used as a validation model. ) was used as a test sample. 300 samples were selected as samples evenly distributed throughout the fermentation process.

(P2) Physicochemical standard analysis of samples

About 25 batches (total of 200) in one operation of Proteinase K fermentation (the entire fermentation process in which the proteinase K-producing strain is inoculated into the fermentation medium and then fermented for a certain period of time while adding the carbon source (glycerol and methanol), which is the raw material for growth and the target product) Dog) samples were collected and subjected to physicochemical standard analysis. After taking 30 ml of a sample, 10 ml of the FT-NIR spectrum was immediately measured, and the rest of the sample was refrigerated. After dissolving the refrigerated samples at room temperature, each was analyzed using a conventional standard analyzer. That is, Proteinase K concentration is measured by Bradford Protein Assay (Kit) method, glycerol concentration as carbon source is by high-performance liquid chromatography (LC) method, methanol concentration is by gas chromatography (GC), and cell viability is measured using a spectrometer. Thus, it was measured at 600 nm and used as a standard analysis value for the development of the FT_NIR calibration model.

(P3) Preprocessing of raw spectra

In order to obtain an accurate quantitative calibration model from the raw spectrum of the fermentation broth obtained in (P1) above, it is necessary to remove the spectrum variable according to the change of the instrument condition and the sample, and to remove the noise included in the measurement spectrum. Spectrum correction and pre-processing were performed using BRUKER's OPUS Quantitative Method (Germany) program. This pre-processing is a necessary process to secure the high-quality spectrum required for the development of a robust, accurate and reproducible calibration model by removing the noise of the raw spectrum and matching the baseline.

In the present invention, by applying the first-order differential method, the second-order differential method, the Norris-Williams derivatives, and the Savitzky-Golay polynomial derivatives, which are generally applied to the transmitted reflection spectrum, the first-order differential method is applied. The raw spectrum was subjected to mathematical pretreatment to remove noise, and by matching the baseline, the utilization of useful information of the spectrum was improved, and a pre-processed spectrum for developing a more accurate and reproducible calibration model was selected and used.

(P4) Calibration calibration model development through preprocessed spectrum region selection and regression analysis

In the case of the fermentation process solution used in the present invention, since it contains many impurities and foreign substances, a specific wavelength region in which the spectrum of the component to be quantified is unique is selected, and a calibration calibration model is developed from the spectrum of the selected region through chemometrics. do. At this time, applicable statistical analysis methods of chemometrics are Multiple Linear Regression (MLR), Principal Component Regression (PCL), Partial Least Squares (PLS), and Modified Partial Least Squares method. There are methods such as (Modified Partial Least Squares, MPLS), and in the present invention, a calibration model was developed using the partial least squares method (PLS). The selection of the optimal calibration model was selected using statistics such as standard error of calibration (SEC) and coefficient of determination in calibration (R2).

(P5) Test and Optimization of Selected Calibration Calibration Model

In order to verify the applicability of the selected calibration model to the new sample to be analyzed, the results obtained by applying the spectrum of 100 new assay samples that are not used as analysis samples to the calibration calibration model are used as physicochemical standard analysis results. It was compared and tested. The applicability test of the calibration calibration model is a calibration adopted using statistics such as SEP (Standard error of prediction), R2 (Coefficient of determination in prediction), Bias (Average difference between reference and NIR values), and SD (Standard deviation). The accuracy and reproducibility were cross-tested by applying unknown samples (100 samples) to the calibration model. When using the final validation model optimized as described above, the proteinase K concentration of the fermentation broth to be analyzed in the present invention, the glycerol and methanol concentrations as carbon sources, and the absorbance as the cell growth rate are simultaneously measured in real time. It can be measured accurately and reproducibly, and the measurement time is completed within seconds, which is incomparably faster than conventional physicochemical analysis methods. It was confirmed that the concentration of various substances can be monitored in real time by scanning the wavelength 500 times within 10 seconds and measuring the average by measuring the spectrum once.

(P6) Control management of time fermentation process through process automation

In addition, by linking the unique electrical signals of each component concentration generated during real-time simultaneous multi-material analysis of the fermentation process by FT-NIR to the automation system, the concentration of glycerol and methanol, the carbon sources, which are supplied to improve productivity, is set in the medium. It is possible to automatically manage the supply of key raw materials for productivity improvement in the culture medium, thereby creating an environment in which microorganisms can produce the target product to the maximum. Therefore, this method can be used for real-time control management of the fermentation process as well as productivity improvement, so it was confirmed that it is an innovative method capable of real-time automatic control of fermentation, which was not possible until now.

Example 1. Development of a method for measuring Proteinase K concentration in Proteinase K fermentation broth using microorganisms

Example 1.1. Installation of FT-NIR and collection of samples

In order to measure and monitor the fermentation status of the fermentation tube in real time online, and convert each signal generated at this time to use it in fermentation process control, the NIR probe is directly installed in the fermentation tube and connected to the existing fermentation process management computer. A Proteinase K fermentation process monitoring and control system was configured so that the measurement signal of NIR could be used for fermentation process control (see FIG. 1).

In order to perform a physicochemical standard analysis while measuring the raw spectrum of the FT-NIR of the continuously proceeding fermentation process, about 30 ml of the sample was taken at intervals of 1 to 2 hours, and a part of the sample (10 ml) was collected over the passage of fermentation. It was used for star spectrum measurement, and the rest was immediately refrigerated and used as a sample for physicochemical standard analysis. Samples were collected during the fermentation culture period of 10 to 15 times, and the number of samples was 300. If necessary, repeated fermentation experiments were performed to collect and use more samples.

Example 1.2. Measurement of raw spectra and analysis of physicochemical standards of samples

Example 1.2.1. Spectral measurement of the sample

About 25 to 30 samples were collected during one operation of the fermentation tube (the whole process of a 130-hour Fed batch culture by inoculating a strain in the fermentation medium), and a part of the sample (10 ml) was FT-NIR. In the entire wavelength region from 12,000 cm -1 to 4,000 cm -1 , the measurement mode of the spectrum with a beam passing distance of 2 mm and a wavelength interval of 2 nm was set to a transmission reflection mode, and raw spectra were measured and collected ( FIG. 2 ). In this case, as the NIR device, FT-NIR Matrix-F Bruker Optics (Germany) was used. In addition, in all the spectra shown in the figure, the x-axis represents a wavenumber (cm-1), and the y-axis represents an absorbance unit.

Example 1.2.2. Physicochemical Standard Analysis of Proteinase K Concentration

After dissolving the samples stored in the refrigerator at room temperature, each concentration was analyzed using the Bardford Protein Assay Kit that analyzes the Proteinase K concentration.

Example 1.3. Development of calibration model for selection and calibration of proteinase K spectral region

The raw spectrum of Proteinase K concentration measured in Example 1.2 and the physicochemical standard analysis result of the sample Proteinase K concentration were processed by the methods mentioned above (P2), (P3) and (P4) to derive a calibration calibration model. . To be more specific, first, the raw spectra of 300 samples are pretreated with the first differentiation to obtain a stable spectrum (see FIG. 3 ), and a spectral region specific to Proteinase K, that is, two wavelength regions as shown in FIG. 4 . was selected to promote the development of a calibration calibration model. The chemometric method used to develop the calibration calibration model was PLS, and the optimal calibration calibration model was selected using statistics such as predicted standard deviation (SEC) and correlation coefficient (R2). The calibration calibration model curve developed in this way, as shown in FIG. 5 , the Y-axis is the FT-NIR analysis value of the Proteinase K concentration in the fermentation broth, and the X-axis is the standard analysis value of the Proteinase K concentration obtained by analysis with the Bardford Protein Assay Kit. Very high level It can be seen that there is a correlation between

Example 1.4. Test of the developed Proteinase K calibration model

To test whether the calibration calibration model developed using the spectra of 200 samples is applicable to new samples not used in the development of the calibration calibration model, the spectra of 100 new samples and physicochemical standard analysis methods Proteinase K concentrations were substituted and comparatively tested, and the results are shown in FIG. 6 (X-axis and Y-axis are the same as those described for FIG. 5) and Table 1.

As can be seen from FIG. 6 and Table 1, it was confirmed that there was no statistically significant difference between the R2 value, the RMSEE value, and the Bias value. In other words, the analysis results measured by FT-NIR and the standard physicochemical analysis results are very similar, so the validation calibration model was successfully developed. Therefore, the development of a real-time continuous monitoring method of Proteinase K concentration in the fermentation broth by FT-NIR that can replace the existing offline analysis method (Bardford Protein Assay Kit method) was completed.

R ² RMSEE RPD Bias Offset Slop Corr. Coeff Calibration 99.26 14.2 11.9 - 1.655 0.993 0.9965 Validation 98.62 17.2 9.61 -0.149 1.990 0.992 0.9946

Example 2. Development of a method for measuring glycerol (carbon source) concentration in Proteinase K fermentation broth using microorganisms

Example 2.1. Installation of FT-NIR and collection of samples

It was carried out in the same manner as in Example 1.1. At this time, the sample used is the same as Example 1.

Example 2.2. Spectral measurement and pretreatment and physicochemical analysis of glycerol concentration

It was carried out in the same manner as in Example 1, and the raw spectrum for glycerol concentration analysis was also used for the same spectrum as the raw spectrum for Proteinase K concentration analysis. The same spectrum as the spectrum for proteinase K concentration measurement was also used for pretreatment of the raw spectrum by the first differentiation. This is the greatest advantage and characteristic of FT-NIR analysis, which can analyze multiple components simultaneously with one spectrum analysis.

Meanwhile, in the present invention, glycerol is a representative sugar component of Proteinase K fermentation broth, and as a result of a preliminary analysis using HPLC, glycerol was defined as the main sugar component. The physicochemical standard analysis of the glycerol concentration was performed using an HPLC analyzer (LC_40A. SHIMADZU. JAPAN).

Example 2.3. Development of calibration model for screening and calibration of glycerol spectral region

The region of the spectrum for measuring the glycerol concentration was selected as two wavelength regions as shown in FIG. 7, and the raw spectrum of the glycerol concentration measured in Examples 2.1 and 2.2 and the physicochemical standard analysis results of the sample glycerol concentration were obtained. The calibration model was derived by processing in the method mentioned in (P2), (P3) and (P4) above (see FIG. 8).

The chemometric method used to develop the calibration calibration model was PLS, and the optimal calibration calibration model was selected using statistics such as predicted standard deviation (SEC) and correlation coefficient (R2). In the calibration calibration model curve developed in this way, as shown in FIG. 9, the Y-axis is the FT-NIR analysis value of the glycerol concentration in the fermentation broth, and the X-axis is the actual physicochemical analysis value, that is, the standard analysis value of the glycerol concentration obtained by HPLC analysis. It can be seen that there is a correlation.

Example 2.4. Test of the developed glycerol calibration calibration model

To test whether it is applicable to the new sample of the calibration model developed using the spectrum of 200 samples (Fig. 9), the spectra of 100 new samples compared to the reference ratio and the glycerol concentration by the standard physicochemical analysis method are substituted , compared and tested, and the results are shown in FIG. 10 (X-axis and Y-axis are the same as those described for FIG. 9) and Table 2.

As can be seen from FIG. 10 and Table 2, it was confirmed that there was no statistically significant difference between the R2 value, the RMSEE value, and the Bias value. In other words, the analysis result measured by FT-NIR and the standard physicochemical analysis result are very similar, so the validation calibration model was successfully developed. there was. In other words, it was confirmed that the change in glycerol concentration in Proteinase K fermentation can be monitored in real time with FT-NIR because the physicochemical standard analysis value and the analysis value by FT_NIR show a very high significant correlation. Therefore, as the monitoring and control of sugar, which has been one of the biggest problems of real-time monitoring and control of changes over time of fermentation, has been solved, real-time monitoring and automatic control process of Proteinase K fermentation has become possible.

R ² RMSEE RPD Bias Offset Slop Corr. Coeff Calibration 99.83 0.734 24.4 - 0.022 0.998 0.9992 Validation 99.73 0.889 19.2 -0.00494 0.039 0.997 0.9986

Example 3. Development of a method for measuring methanol (carbon source) concentration in Proteinase K fermentation broth using microorganisms

Example 3.1. Installation of FT-NIR and collection of samples

It was carried out in the same manner as in Example 1.1. At this time, the sample used is the same as Example 1.

Example 3.2. Spectral measurement and pretreatment and physicochemical analysis of methanol concentration

It was carried out in the same manner as in Example 1, and the same spectrum as the raw spectrum for the analysis of the proteinase K concentration was used for the analysis of the methanol concentration. The same spectrum as the spectrum for proteinase K concentration measurement was also used for pretreatment of the raw spectrum by the first differentiation.

Physicochemical standard analysis of methanol was analyzed using gas chromatography (GC-2010Pluse. SHIMADZU. JAPAN).

Example 3.3. Development of calibration model for screening and calibration of methanol spectral region

The region of the spectrum for measuring the methanol concentration was selected as two wavelength regions as shown in FIG. 11, and the methanol concentration raw spectrum measured in Examples 3.1 and 3.2 and the physicochemical standard analysis results of the sample methanol concentration were obtained. The calibration model was derived by processing in the method mentioned in (P2), (P3) and (P4) above (see FIG. 12).

The chemometric method used to develop the calibration calibration model was PLS, and the optimal calibration calibration model was selected using statistics such as predicted standard deviation (SEC) and correlation coefficient (R2). As shown in FIG. 13, the calibration calibration model curve developed in this way shows that the X-axis is the standard analysis value of the methanol concentration obtained by GC analysis and the Y-axis is the FT-NIR analysis value of the methanol concentration in the fermentation broth. It can be seen that there is a very high level of correlation. can

Example 3.4. Test of the developed methanol calibration calibration model

To test whether the calibration calibration model (Fig. 13) developed using the spectrum of 200 samples is applicable to an unknown new sample, the spectrum of 100 new samples and the methanol concentration by the physicochemical standard analysis method Substitution and comparison tests were performed, and the results are shown in FIG. 14 (X-axis and Y-axis are the same as those described for FIG. 13) and Table 3.

As can be seen from FIG. 14 and Table 3, it was confirmed that there was no statistically significant difference between the R2 value and the RMSEE value. In other words, the analysis results measured by FT-NIR and the results of standard physicochemical analysis were very similar, so the development of a validation calibration model was successful. could In other words, as the physicochemical standard analysis value and the analysis value by FT-NIR show a very high significant correlation, FT-NIR can accurately and reproducibly monitor the temporal change of the methanol concentration in the Proteinase K fermentation broth in real time. Confirmed. Therefore, real-time measurement of methanol concentration, which has been one of the biggest barriers to real-time monitoring and control of changes over time of fermentation, has become possible, making it possible to develop on-line real-time monitoring and automatic control of the Proteinase K fermentation process.

R ² RMSEE RPD Bias Offset Slop Corr. Coeff Calibration 99.37 0.063 12.6 - 0.008 0.994 0.9968 Validation 98.31 0.0984 7.69 -0.0006 0.018 0.987 0.9915

Example 4. Development of a method for measuring the absorbance (OD) of Proteinase K fermentation broth using microorganisms

Example 4.1. Installation of FT-NIR and collection of samples

It was carried out in the same manner as in Example 1.1. At this time, the sample used is the same as Example 1.

Example 4.2. Spectral measurement and pretreatment and physicochemical analysis of absorbance

It was carried out in the same manner as in Example 1, and the same spectrum as the raw spectrum for proteinase K concentration analysis was used for the absorbance analysis. The same spectrum as the spectrum for proteinase K concentration measurement was also used for pretreatment of the raw spectrum by the first differentiation.

Physicochemical standard analysis of absorbance was performed at 600 nm using a spectrometer (LIBAR S80W BIOCHROM USA), and the obtained value was used as a standard analysis value.

Example 4.3. Development of calibration model for screening and calibration of absorbance spectral region

The spectrum region for measuring the absorbance was selected as two wavelength regions as shown in FIG. 15, and the raw absorbance spectrum measured in Examples 4.1 and 4.2 and the physicochemical standard analysis result of the sample absorbance were analyzed above (P2 ), (P3), and (P4) were treated in the same way to derive a calibration model (see Fig. 16)

The chemometric method used to develop the calibration calibration model was PLS, and the optimal calibration calibration model was selected using statistics such as predicted standard deviation (SEC) and correlation coefficient (R2). As shown in FIG. 17, the calibration calibration model curve developed in this way, the Y-axis is the FT-NIR analysis value of the absorbance of the fermentation broth, and the X-axis is the actual physicochemical analysis value, that is, the standard analysis value obtained with a spectrometer at 600 nm. It can be seen that the absorbance measurements are also highly correlated.

Example 4.4. Test of the developed absorbance calibration calibration model

To test whether the calibration model developed using the spectra of 200 samples (FIG. 17) is applicable to an unknown new sample, the spectra of 100 new samples compared to the standard and absorbance by standard physicochemical analysis methods are substituted , compared and tested, and the results are shown in FIG. 18 (X-axis and Y-axis are the same as those described for FIG. 17) and Table 4.

As can be seen from FIG. 18 and Table 4, it was confirmed that there was no statistically significant difference between the R2 value, the RMSEE value, and the Bias value. In other words, the absorbance (OD) analysis value measured by FT-NIR and the results of standard physicochemical analysis were very similar, so the validation calibration model was successfully developed. It was confirmed that the cell growth rate could be measured accurately and reproducibly. Therefore, real-time measurement of cell growth, one of the biggest problems of real-time monitoring and control of changes over time of fermentation, has become possible, making it possible to develop on-line real-time monitoring and automatic control of the Proteinase K fermentation process.

R ² RMSEE RPD Bias Offset Slop Corr. Coeff Calibration 99.26 14.2 11.9 - 1.655 0.993 0.9965 Validation 98.62 17.2 9.61 -0.149 1.990 0.992 0.9946

Example 5. Monitoring and control results of Proteinase K fermentation process using FT-NIR

Calibration formula that can measure the major factors of fermentation developed through Examples 1 to 4, that is, Proteinase K (protein) concentration, glycerol and methanol concentrations as carbon sources, and absorbance (OD _600nm ) indicating cell growth. By applying the model-equipped FT-NIR system to the actual fermentation process, a plurality of fermentation components were continuously measured and monitored at the same time, and the electrical signal generated at this time was used to test whether the Proteinase K fermentation process could be controlled in real time. The results are shown in FIG. 19 .

Specifically, the spectrum measured in real time through the probe installed in the fermentation tube is calculated through a calibration model and shows the Proteinase K concentration in the Proteinase K fermentation broth by FT-NIR, the concentration of glycerol and methanol as carbon sources, and the cell growth rate. Not only is the absorbance (OD _600nm ) analyzed and monitored accurately and reproducibly in real time, but also the signal generated during monitoring is converted into an electrical signal to link the existing fermentation process control program, PLC (Programmable Logic Controller), to control the fermentation process in real time. could be controlled automatically.

In addition, as shown in FIG. 19, after taking the sample offline, the standard physicochemical analysis result and the FT-NIR measurement value, that is, the proteinase K concentration of the target product, the concentration of glycerol and methanol as carbon sources, and the absorbance ( OD _600nm ) showed very good analysis values with little difference from the standard analysis values, so it was confirmed that real-time monitoring of the Proteinase K fermentation process by FT-NIR and real-time automatic control of the fermentation process using the signal proceeded very well. .

Claims (21)

As a method for real-time control of the protein production process by microbial fermentation,
(i) providing a calibration model obtained using two or more near-infrared wavelength regions specific for each of the one or more fermentation broth component indicators, and
(ii) monitoring each change in the fermentation broth component index with a light source in the two or more near-infrared wavelength regions,
The calibration model is
(a) collecting n samples of the fermentation broth at regular time intervals;
(b) measuring the raw spectrum for calibration using a near-infrared spectrometer for m samples among the collected n samples;
(c) performing physicochemical standard analysis on m samples for which the raw spectrum has been measured;
(d) statistically preprocessing the raw spectrum measured in step (b);
(e) Select two or more near-infrared wavelength regions specific for each of one or more fermentation broth component indicators to be quantified from the spectrum pretreated in step (d), and a calibration calibration model from the spectrum of the selected region through chemometrics building a step, and
(f) comparing the results obtained by applying the nm samples not used for raw spectrum measurement to the calibration calibration model developed in step (e) with the physicochemical standard analysis results obtained in step (c), and is obtained by a method comprising the step of optimizing, wherein n and m are natural numbers greater than or equal to 50, and n is a natural number greater than m.
method. The method of claim 1,
The method further comprises the step of controlling the fermentation process in real time in response to the electrical signal of each component indicator generated through the monitoring
method. The method of claim 1,
The monitoring is performed through a near-infrared spectrometer attached to the fermentation tube in the form of a probe.
method. The method of claim 1,
The monitoring can be performed without sampling the fermentation broth.
method. The method of claim 1,
The protein is Proteinase K
method. The method of claim 1,
The microorganism is yeast
method. The method of claim 1,
The microorganism is Pichia pastoris ( Pichia pastoris )
method. The method of claim 1,
The fermentation broth component indicator is at least one selected from the group consisting of Proteinase K concentration, glycerol concentration, methanol concentration, and microbial growth (absorbance).
method. 9. The method of claim 8,
the two or more wavelength ranges include a wavelength range of 9403.5 to 7498.1 cm-1 and a wavelength range of 6101.9 to 5446.2 cm-1 in the case of the Proteinase K concentration; In the case of the glycerol concentration, it includes a wavelength region of 8886.7 to 7745.0 cm -1 and a wavelength region of 5870.4 to 5420.0 cm -1 ; the methanol concentration includes a wavelength range of 8457.7 to 7498.1 cm-1 and a wavelength range of 6101.9 to 5446.2 cm-1; In the case of the microbial viability, including a wavelength region of 8535.3 to 7503.4 cm-1 and a wavelength region of 5797.2 to 5439.5 cm-1
method. delete A fermentation system for the production of protein by microbial fermentation, comprising:
(i) a control for real-time monitoring and control of a protein production process by microbial fermentation, including a calibration model obtained using two or more near-infrared wavelength regions specific for each of the fermentation broth component indicators to monitor changes in one or more fermentation broth component indicators includes wealth,
The calibration model is
(a) collecting n samples of the fermentation broth at regular time intervals;
(b) measuring the raw spectrum for calibration using a near-infrared spectrometer for m samples among the collected n samples;
(c) performing physicochemical standard analysis on m samples for which the raw spectrum has been measured;
(d) statistically preprocessing the raw spectrum measured in step (b);
(e) Select two or more near-infrared wavelength regions specific for each of one or more fermentation broth component indicators to be quantified from the spectrum pretreated in step (d), and a calibration calibration model from the spectrum of the selected region through chemometrics building a step, and
(f) comparing the results obtained by applying the nm samples not used for raw spectrum measurement to the calibration calibration model developed in step (e) with the physicochemical standard analysis results obtained in step (c), and is obtained by a method comprising the step of optimizing, wherein n and m are natural numbers greater than or equal to 50, and n is a natural number greater than m.
fermentation system. 12. The method of claim 11,
(ii) a fermentation unit for microbial fermentation; and
(iii) further comprising a near-infrared spectroscopy unit
fermentation system. 13. The method of claim 12,
The near-infrared spectroscopy unit is mounted in the form of a probe in the fermentation unit.
fermentation system. 12. The method of claim 11,
The control unit automatically controls the fermentation process in real time in response to the electrical signal of each component indicator generated by simultaneously monitoring changes in the one or more fermentation broth component indicators in real time.
fermentation system. 12. The method of claim 11,
The fermentation broth component indicator is at least one selected from the group consisting of Proteinase K concentration, glycerol concentration, methanol concentration, and microbial growth (absorbance).
fermentation system. 12. The method of claim 11,
The microorganism is a LMO (Living modified organisms) strain
fermentation system. 12. The method of claim 11,
The microorganism is a yeast Pichia pastoris ( Pichia pastoris )
fermentation system. 13. The method of claim 12,
The fermentation system is closed
fermentation system. 13. The method of claim 12,
The fermentation system can check the fermentation state of the protein production process by microbial fermentation through an on-tank measurement facility.
fermentation system. 12. The method of claim 11,
The one or more fermentation broth component indicators include a glycerol concentration that is a carbon source for microbial growth
fermentation system. 16. The method of claim 15,
the two or more wavelength ranges include a wavelength range of 9403.5 to 7498.1 cm-1 and a wavelength range of 6101.9 to 5446.2 cm-1 in the case of the Proteinase K concentration; In the case of the glycerol concentration, it includes a wavelength region of 8886.7 to 7745.0 cm -1 and a wavelength region of 5870.4 to 5420.0 cm -1 ; the methanol concentration includes a wavelength range of 8457.7 to 7498.1 cm-1 and a wavelength range of 6101.9 to 5446.2 cm-1; In the case of the microbial viability, including a wavelength region of 8535.3 to 7503.4 cm-1 and a wavelength region of 5797.2 to 5439.5 cm-1
fermentation system.

Images