Glucose-limited high cell density cultivations from small to pilot plant scale using an enzyme-controlled glucose delivery system

The enzyme controlled substrate delivery cultivation technology EnBase(?) Flo allows a fed-batch-like growth in batch cultures. It has been previously shown that this technology can be applied in small cultivation vessels such as micro- and deep well plates and also shake flasks. In these scales high cell densities and improved protein production for Escherichia coli cultures were demonstrated. This current study aims to evaluate the scalability of the controlled glucose release technique to pilot scale bioreactors. Throughout all scales, that is, deep well plates, 3 L bioreactor and 150 L bioreactor cultivations, the growth was very similar and the model protein, a recombinant alcohol dehydrogenase (ADH) was produced with a high yield in soluble form. Moreover, EnBase Flo also was successfully used as a controlled starter culture in high cell density fed-batch cultivations with external glucose feeding. Here the external feeding pump was started after overnight cultivation with EnBase Flo. Final optical densities in these cultivations reached 120 (corresponding to about 40 g L(-1) dry cell weight) and a high expression level of ADH was obtained. The EnBase cultivation technology ensures a controlled initial cultivation under fed-batch mode without the need for a feeding pump. Because of the linear cell growth under glucose limitation it provides optimal and robust starting conditions for traditional external feed-based processes.     

Growth and recombinant protein expression with <Emphasis Type="Italic">Escherichia coli</Emphasis> in different batch cultivation media

Parallel operated milliliter-scale stirred tank bioreactors were applied for recombinant protein expression studies in simple batch experiments without pH titration. An enzymatic glucose release system (EnBase), a complex medium, and the frequently used LB and TB media were compared with regard to growth of Escherichia coli and recombinant protein expression (alcohol dehydrogenase (ADH) from Lactobacillus brevis and formate dehydrogenase (FDH) from Candida boidinii). Dissolved oxygen and pH were recorded online, optical densities were measured at-line, and the activities of ADH and FDH were analyzed offline. Best growth was observed in a complex medium with maximum dry cell weight concentrations of 14 g L⁻¹. EnBase cultivations enabled final dry cell weight concentrations between 6 and 8 g L⁻¹. The pH remained nearly constant in EnBase cultivations due to the continuous glucose release, showing the usefulness of this glucose release system especially for pH-sensitive bioprocesses. Cell-specific enzyme activities varied considerably depending on the different media used. Maximum specific ADH activities were measured with the complex medium, 6 h after induction with IPTG, whereas the highest specific FDH activities were achieved with the EnBase medium at low glucose release profiles 24 h after induction. Hence, depending on the recombinant protein, different medium compositions, times for induction, and times for cell harvest have to be evaluated to achieve efficient expression of recombinant proteins in E. coli. A rapid experimental evaluation can easily be performed with parallel batch operated small-scale stirred tank bioreactors.     

Effect of cultivation conditions on cell-surface display of Flo1 fusion protein using sake yeast

The cell-surface display of the Flo1p anchor system with a flocculation functional domain was examined under various cultivation conditions. As a model system, lipase from Rhizopus oryzae with the pro sequence was genetically fused to the Flo1 short (FS) anchor (FSProROL) and displayed on the sake yeast cell-surface under the control of the SED800 promoter (pSED800). The nutrients and carbon source in the culture media affected the display of the fusion protein FSProROL on the sake yeast cell-surface. The lipase activity in whole cells cultivated in poor media, without peptone and/or yeast extracts, were higher than those cultivated in rich media. In addition, glucose and maltose were effective carbon sources for increasing the lipase activity in whole cells, and the addition of di- or tri-saccharide as the carbon source reduced the release of the lipase activity into the culture supernatants. The initial glucose concentration was found to influence the total lipase activity and it mainly affected the lipase activity in whole cells. Under the optimum condition, sake yeast was found to show high cell density and high lipase activity in short time cultivation.     

High cell density cultivation and recombinant protein production with <Emphasis Type="Italic">Escherichia coli</Emphasis> in a rocking-motion-type bioreactor

Background

Single-use rocking-motion-type bag bioreactors provide advantages compared to standard stirred tank bioreactors by decreased contamination risks, reduction of cleaning and sterilization time, lower investment costs, and simple and cheaper validation. Currently, they are widely used for cell cultures although their use for small and medium scale production of recombinant proteins with microbial hosts might be very attractive. However, the utilization of rocking- or wave-induced motion-type bioreactors for fast growing aerobic microbes is limited because of their lower oxygen mass transfer rate. A conventional approach to reduce the oxygen demand of a culture is the fed-batch technology. New developments, such as the BIOSTAT^® CultiBag RM system pave the way for applying advanced fed-batch control strategies also in rocking-motion-type bioreactors. Alternatively, internal substrate delivery systems such as EnBase^® Flo provide an opportunity for adopting simple to use fed-batch-type strategies to shaken cultures. Here, we investigate the possibilities which both strategies offer in view of high cell density cultivation of E. coli and recombinant protein production.

Results

Cultivation of E. coli in the BIOSTAT^® CultiBag RM system in a conventional batch mode without control yielded an optical density (OD₆₀₀) of 3 to 4 which is comparable to shake flasks. The culture runs into oxygen limitation. In a glucose limited fed-batch culture with an exponential feed and oxygen pulsing, the culture grew fully aerobically to an OD₆₀₀ of 60 (20 g L^-1 cell dry weight). By the use of an internal controlled glucose delivery system, EnBase^® Flo, OD₆₀₀ of 30 (10 g L^-1 cell dry weight) is obtained without the demand of computer controlled external nutrient supply. EnBase^® Flo also worked well in the CultiBag RM system with a recombinant E. coli RB791 strain expressing a heterologous alcohol dehydrogenase (ADH) to very high levels, indicating that the enzyme based feed supply strategy functions well for recombinant protein production also in a rocking-motion-type bioreactor.

Conclusions

Rocking-motion-type bioreactors may provide an interesting alternative to standard cultivation in bioreactors for cultivation of bacteria and recombinant protein production. The BIOSTAT^® Cultibag RM system with the single-use sensors and advanced control system paves the way for the fed-batch technology also to rocking-motion-type bioreactors. It is possible to reach cell densities which are far above shake flasks and typical for stirred tank reactors with the improved oxygen transfer rate. For more simple applications the EnBase^® Flo method offers an easy and robust solution for rocking-motion-systems which do not have such advanced control possibilities.

     

Enzyme controlled glucose auto-delivery for high cell density cultivations in microplates and shake flasks

Background

Here we describe a novel cultivation method, called EnBase?, or enzyme-based-substrate-delivery, for the growth of microorganisms in millilitre and sub-millilitre scale which yields 5 to 20 times higher cell densities compared to standard methods. The novel method can be directly applied in microwell plates and shake flasks without any requirements for additional sensors or liquid supply systems. EnBase is therefore readily applicable for many high throughput applications, such as DNA production for genome sequencing, optimisation of protein expression, production of proteins for structural genomics, bioprocess development, and screening of enzyme and metagenomic libraries.

Results

High cell densities with EnBase are obtained by applying the concept of glucose-limited fed-batch cultivation which is commonly used in industrial processes. The major difference of the novel method is that no external glucose feed is required, but glucose is released into the growth medium by enzymatic degradation of starch. To cope with the high levels of starch necessary for high cell density cultivation, starch is supplied to the growing culture suspension by continuous diffusion from a storage gel. Our results show that the controlled enzyme-based supply of glucose allows a glucose-limited growth to high cell densities of OD₆₀₀ = 20 to 30 (corresponding to 6 to 9 g l^-1 cell dry weight) without the external feed of additional compounds in shake flasks and 96-well plates. The final cell density can be further increased by addition of extra nitrogen during the cultivation. Production of a heterologous triosphosphate isomerase in E. coli BL21(DE3) resulted in 10 times higher volumetric product yield and a higher ratio of soluble to insoluble product when compared to the conventional production method.

Conclusion

The novel EnBase method is robust and simple-to-apply for high cell density cultivation in shake flasks and microwell plates. The potential of the system is that the microbial growth rate and oxygen consumption can be simply controlled by the amount (and principally also by the activity) of the starch-degrading enzyme. This solves the problems of uncontrolled growth, oxygen limitation, and severe pH drop in shaken cultures. In parallel the method provides the basis for enhanced cell densities. The feasibility of the new method has been shown for 96-well plates and shake flasks and we believe that it can easily be adapted to different microwell and deepwell plate formats and shake flasks. Therefore EnBase will be a helpful tool especially in high throughput applications.     

Semi-defined Medium for in vitro Cultivation of the Fastidious Insect Pathogenic Fungus Cordyceps unilateralis

Cordyceps unilateralis is a fastidious fungal pathogen affecting ants. Up to now, only the complex and expensive Grace’s insect cell culture medium has been used for in vitro cultivation (as blastospores and mycelium) of this fungus. To obtain an inexpensive and less complicated medium, the effects of carbon and nitrogen sources, salt solution and carbon-to-nitrogen (C:N) ratio on the growth of this fungus were examined. Glucose was the most important factor for blastospore formation, and yeast extract could be used as a nitrogen source for blastospore formation and mycelial growth. A suitable C:N ratio (glucose: yeast extract) was 33.3:1. As a result, a new semi-defined medium was achieved, composed of 26.68 g L⁻¹ glucose, 3.3 g L⁻¹ yeast extract and salt solution. This medium supported blastospore formation and mycelial growth of all tested C. unilateralis isolates.     

Obtaining and selection of hexokinases-less strains of <Emphasis Type="Italic">Saccharomyces cerevisiae</Emphasis> for production of ethanol and fructose from sucrose

Saccharomyces cerevisiae hexokinase-less strains were produced to study the production of ethanol and fructose from sucrose. These strains do not have the hexokinases A and B. Twenty-three double-mutant strains were produced, and then, three were selected for presenting a smaller growth in yeast extract–peptone–fructose. In fermentations with a medium containing sucrose (180.3 g L⁻¹) and with cell recycles, simulating industrial conditions, the capacity of these mutant yeasts in inverting sucrose and fermenting only glucose was well characterized. Besides that, we could also see their great tolerance to the stresses of fermentative recycles, where fructose production (until 90 g L⁻¹) and ethanol production (until 42.3 g L⁻¹) occurred in cycles of 12 h, in which hexokinase-less yeasts performed high growth (51.2% of wet biomass) and viability rates (77% of viable cells) after nine consecutive cycles.     

Distribution of the 7S Proteins in Soybean Globulins by Gel Filtration with Sephadex G-200

The level of isocitrate lyase, an enzyme of glyoxylate cycle, in Candida tropicalis was enhanced at the later period of growth when the yeast was cultivated in a semisynthetic glucose medium. On the other hand, such increase in the enzyme activity was not observed in C. lipolytica grown under the same conditions. In the case of C. tropicalis, high concentrations of glucose remaining in the medium permitted the increase in the enzyme activity and the addition of ethanol, one of the major products from glucose, to the glucose medium did not stimulate the enzyme formation, indicating that the enhanced enzyme level in the yeast was not merely attributable to the release from the repression by glucose or to the induction by ethanol. Biotin, one of the growth-stimulating factors for C. tropicalis, affected markedly the level of isocitrate lyase. That is, the supplementation of biotin to the synthetic glucose medium inhibited completely the increase in the enzyme activity, and reversely the absence of biotin stimulated the enzyme formation in the glucose-assimilating cells. Thiamine, another growth-stimulating factor for C. tropicalis, did not show any effect on the level of isocitrate lyase in the yeast. The level of isocitrate lyase in C. lipolytica growing on glucose was not affected by biotin added exogenously.     

Growth and lipid synthesis promotion in mixotrophic Neochloris oleoabundans (Chlorophyta) cultivated with glucose

In the recent years, the studies concerning the cultivation of Neochloris oleoabundans for biofuel purposes have increased, in relation to its capability to accumulate lipids when grown under nutrient starvation. Unfortunately, this cultivation mode does not allow to reach high biomass densities, which are required to improve the feasibility of the process. Increasing knowledge of the microalgal physiology is necessary to obtain new useful information for the improvement of culture performance in the perspective of large-scale cultivation. In this work, the mixotrophic cultivation of N. oleoabundans in a brackish medium added with different glucose concentrations has been tested under shaking, with the aim of stimulating growth alongside lipid accumulation inside cells. Cell morphology, glucose consumption, photosynthetic pigment content and photosynthetic efficiency were also investigated. Among all tested glucose concentrations (0–30 g L^?1), it was observed that 2.5 g L^?1 was the optimal concentration, allowing to obtain the best compromise between glucose supplement, biomass production and lipid accumulation. Growth was highly enhanced in mixotrophic cultures, linked to the release of cells from sporocysts. A unique feature characterising mixotrophy in N. oleoabundans was the promotion of the maximum quantum yield of Photosystem II. Moreover, when mixotrophic cells entered the stationary phase, high lipid accumulation was induced. This study shows that the addition of glucose to N. oleoabundans remarkably increases the production of biomass enriched in lipids and represents an advancement for the cultivation of this microalga for applied purposes.     

Production of chitin deacetylase by Aspergillus flavus in submerged conditions

Chitosan is a biopolymer obtained by deacetylation of chitin and has been proven to have various applications in industry and biomedicine. Deacetylation of chitin using the enzyme chitin deacetylase (CDA) is favorable in comparison to the hazardous chemical method involving strong alkali and high temperature. A fungal strain producing CDA was isolated from environmental samples collected from coastal regions of South Kerala, India. It was identified as Aspergillus flavus by morphological characteristics and ITS DNA analysis. Nutritional requirement for maximum production of CDA under submerged condition was optimized using statistical methods including Plackett–Burman and response surface methodology central composite design. A 5.98-fold enhancement in CDA production was attained in shake flasks when the fermentation process parameters were used at their optimum levels. The highest CDA activity was 57.69 ± 1.68 U under optimized bioprocess conditions that included 30 g L^?1 glucose, 40 g L^?1 yeast extract, 15 g L^?1 peptone, and 7 g L^?1 MgCl₂ at initial media pH of 7 and incubation temperature of 32°C after 48 hr of incubation, while the unoptimized basal medium yielded 9.64 ± 2.04 U.     

Enhancement of asymmetric bioreduction of ethyl 4-chloro acetoacetate by the design of composition of culture medium and reaction conditions

This study investigated the feasibility of improving the stereospecificity of yeast by the adjustment of the culture medium composition and the reaction conditions. The investigation was performed systematically, using an approach that integrates the Taguchi's array method and the steepest ascent method. The reaction yield and the product's ee were the two indexes of reaction performance. A desirability function was applied to combine these two indexes as a single objective function. The removal of peptone and malt extract from the YM medium increased the yeast's stereoselectivity, without reducing the production of biomass. The medium composition and the reaction conditions were then simultaneously optimized. The resulting optimal conditions were 30 g/l glucose for cultivation, 12 g/l yeast extract, a cultivation time of 12 h, 15 g/l glucose for reaction, 150 g/l yeast for reaction, a reaction buffer concentration of 0.2 M and a buffer pH of 8.5. Compared to the one before this study, the product's ee was improved from 82.1 to 92.3%, and the reaction yield was enhanced from 77.3 to 82.3%. Furthermore, the biomass production was increased considerably from 5.94 to 10.14 g DCW/l.     

Heterotrophic production of lutein by selected Chlorella strains

Seven Chlorella strains representing three species obtained from culture collections and research laboratories were screened for their potential of heterotrophic production of lutein on two different media (Basal and Kuhl) containing glucose. While both media supported good growth and lutein formation of the seven strains in darkness, higher biomass concentrations and lutein content were achieved on Basal medium. Chlorella protothecoides CS-41 was chosen from the seven strains for further investigation due to its higher productivities of both biomass and lutein. The maximal biomass concentration and lutein content of C. protothecoides cultivated heterotrophically with 9 g L^-1 glucose in a 3.7-L fermentor were respectively 4.6 g dry cells L^-1 and 4.60 mg lutein g^-1 dry cells on Basal medium, and 4.0 g dry cells L^-1 and 4.36 mg lutein g^-1 dry cells on Kuhl medium. The heterotrophic cultivation process was scaled up successfully to 30 L using a fermentor, in which the Basal medium containing 36 g L^-1 glucose was used; the maximal biomass concentration of 16.4 g dry cells L^-1, specific growth rate of 0.92 d^-1,lutein content of 4.85 mg lutein g^-1 dry cells,growth yield of 0.47 g dry cells g^-1 glucose and lutein yield of 1.93 mg lutein g^-1 glucose were respectively achieved. This revised version was published online in August 2006 with corrections to the Cover Date.     

The citric acid production from raw glycerol by Yarrowia lipolytica yeast and its regulation

The optimal cultivation conditions ensuring the maximal rate of citric acid (CA) biosynthesis by glycerol-grown mutant Yarrowia lipolytica NG40/UV7 were found to be as follows: growth limitation by inorganic nutrients (nitrogen, phosphorus, or sulfur), 28 °C, pH 5.0, dissolved oxygen concentration (pO₂) of 50 % (of air saturation), and pulsed addition of glycerol from 20 to 80 g L^?1 depending on the rate of medium titration. Under optimal conditions of fed-batch cultivation, in the medium with pure glycerol, strain Y. lipolytica NG40/UV7 produced 115 g L^?1 of CA with the mass yield coefficient of 0.64 g g^?1 and isocitric acid (ICA) amounted to 4.6 g L^?1; in the medium with raw glycerol, CA production was 112 g L^?1 with the mass yield coefficient of 0.90 g g^?1 and ICA amounted to 5.3 g L^?1. Based on the activities of enzymes involved in the initial stages of raw glycerol assimilation, the tricarboxylic acid cycle and the glyoxylate cycle, the mechanism of increased CA yield from glycerol-containing substrates in Y. lipolytica yeast was explained.     

Influence of environmental factors on endo-β-1,4-glucanase production by Bacillus HR68, isolated from a Zimbabwean hot spring

The production of endo-β-1,4-glucanase by a Bacillus strain isolated from a hot spring in Zimbabwe was studied in batch culture, chemostat culture, and carbon dioxide-regulated auxostat (CO₂-auxostat). The bacteria produced the enzyme in the presence of excess glucose or sucroso, but not under carbon-limited conditions in a chemostat using mineral medium. There was a specific growth rate dependent linear increase in enzyme production in glucose excess, nitrogen-limited chemostat cultures. A high specific growth rate of 2.2 h^-1 and a high rate of enzyme production of 362 nkat (mg dry mass h)^-1 were attained under nutrient rich conditions in the CO₂-auxostat. The bacteria had the highest specific growth rate and endo-β-1,4-glucanase enzyme production at 50° C. The maximum specific growth rate and the rate of enzyme production increased when yeast extract and tryptone were added in increasing amounts to the mineral medium used for cultivation in separate experiments. Increasing the glucose concentration in the CO₂-auxostat cultures increased the rate of enzyme production but did not affect the specific growth rate.     

Optimization of medium composition for improving biomass production of Lactobacillus plantarum Pi06 using the Taguchi array design and the Box-Behnken method

An immune-enhancing strain, Lactobacillus plantarum Pi06, isolated from a healthy infant was used for biomass production following optimization of the medium in shake-flask culture. Preliminary studies showed that commercial MRS medium and cultivation under static conditions generated higher biomass production than four other tested media with or without a shaking condition. The selected medium composition, consisting of glucose, yeast extract, soy peptone, ammonium citrate, and corn steep liquor, was further optimized using a systematic method that integrated the Taguchi array design and the Box-Behnken method. The response effects of these factors were first investigated using Taguchi design under an L ₁₆ (4⁵) array. The suggested medium composition, derived from Statistica 7.1 using the Taguchi design, was applied to cultivate cells and a biomass of 7.16 g dry cell weight (DCW)/L was obtained. Response surface methodology based on the Box-Behnken method for the three response variables of glucose, yeast extract, and corn steep liquor was then used to further increase the biomass level to 8.94 g DCW/L. The resulting optimum medium consisted of 35 g/L glucose, 35 g/L yeast extract, and 40 mL/L corn steep liquor. Compared with the initial medium, the biomass yield was improved from 4.31 to 8.94 g DCW/L, an enhancement of approximately 107%.     

Synthesis of the extracellular protease byBacillus pumilus

Bacillus pumilus synthesizes an oxtracellular protease during the stationary phase of growth when the intracellular protease level is rather low. The formation of the enzyme is blocked by chloramphenicol. A shift from batch to continuous cultivation is accompanied by a decreased enzyme level in the medium. The original concentration of the enzyme in the medium can be attained only after reverting to batch cultivation. The protease is not synthesized during growth in a mineral medium with glucose. Glutamic acid, arginine and ornithine are among the amino acids which stimulate the enzyme formation. Yeast extract enhances the formation of the protease, the active component of the extract being only the brown pigment with attached Fe³⁺ ions. The B-vitamins and other essential growth factors contained in the yeast extract are without effect on the protease formation. Free ferric ions also induce protease formation, the level being roughly proportional to the concentration of Fe³⁺ in the medium. The synthesis is also enhanced by Mn²⁺. Growth of the culture under oxygen limitation results in the suppression of protease formation.     

Efficient oleaginous yeasts for lipid production from lignocellulosic sugars and effects of lignocellulose degradation compounds on growth and lipid production

Microbial lipid production using lignocellulosic biomass is considered an alternative for biodiesel production. In this study, 418 yeast strains were screened to find efficient oleaginous yeasts which accumulated large quantities of lipid when cultivated in lignocellulosic sugars. Preliminary screening by Nile red staining revealed that 142 strains contained many or large lipid bodies. These strains were selected for quantitative analysis of lipid accumulation by shaking flask cultivation in nitrogen-limited medium II containing 70 g/L glucose or xylose or mixture of glucose and xylose in a ratio of 2:1. Rhodosporidium fluviale DMKU-SP314 produced the highest lipid concentration of 7.9 g/L when cultivated in the mixture of glucose and xylose after 9 days of cultivation, which was 55.0% of dry biomass (14.3 g/L). The main composition of fatty acids were oleic acid (40.2%), palmitic acid (25.2%), linoleic acid (17.9%) and stearic acid (11.1%). Moreover, the strain DMKU-SP314 could grow and produce lipid in a medium containing predominantly lignocellulose degradation products, namely, acetic acid, formic acid, furfural, 5-hydroxymethylfurfural (5-HMF) and vanillin, with however, some inhibitory effects. This strain showed high tolerance to acetic acid, 5-HMF and vanillin. Therefore, R. fluviale DMKU-SP314 is a promising strain for lipid production from lignocellulosic hydrolysate.     

Trehalose phosphorylase from Pichia fermentans and its role in the metabolism of trehalose

During a screening for novel microbial trehalose phosphorylase three Pichia strains were identified as producers of this particular enzyme that have not yet been described. To our knowledge, this is the first time that this enzyme activity has been shown in yeasts. Pichia fermentans formed trehalose phosphorylase when cultivated on a growth medium containing easily metabolizable sugers such as glucose. Addition of NaCl (0.4 M) to the medium increased the synthesis of the enzyme significantly. Production of trehalose phosphorylase was found to be growth-associated with a maximum of activity formed at the transition of the exponential to the stationary phase of growth. Trehalose phosphorylase catalyzes the phosphorolytic cleavage of trehalose, yielding glucose 1-phosphate (glucose-1-P) and glucose as products. In vitro the enzyme readily catalyzes the reverse reaction, the synthesis of trehalose from glucose and glucose-1-P. For this reaction, the enzyme of P. fermentans was found to utilize -glucose-1-P preferentially. A partially purified enzyme preparation showed a pH optimum of 6.3 for the synthesis of trehalose. The enzyme was found to be rather unstable; it was easily inactivated by dilution unless Ca²⁺ or Mn²⁺ were added. This instability is presumably caused by dissociation of the enzyme. In contrast to other yeasts, P. fermentans rapidly degraded intracellularly accumulated trehalose when the carbon source in the medium was depleted. Trehalose phosphorylase seems to be a key enzyme in the degradative pathway of trehalose in P. fermentans. Additional enzymes in this catabolic pathway of trehalose include phosphoglucomutase, glucose-6-phosphate dehydrogenase, and gluconolactonase.This contribution is part of the Ph.D. thesis of Ingrid Schick     

Yeast concentration estimation and prediction with static and dynamic neural network models in batch cultures

The second fermentation is one of the most important steps in Champagne production. For this purpose, yeasts are grown on a wine based medium to adapt their metabolism to ethanol. Several models built with various static and dynamic neural network configurations were investigated. The main objective was to achieve real-time estimation and prediction of yeast concentration during growth. The model selected, based on recurrent neural networks, was first order with respect to the yeast concentration and to the volume of CO₂ released. Temperature and pH were included as model parameters as well. Yeast concentration during growth could thus be estimated with an error lower than 3% (±1.7×10⁶ yeasts/ml). From the measurement of initial yeast population and temperature, it was possible to predict the final yeast concentration (after 21 hours of growth) from the beginning of the growth, with about ±3×10⁶ yeasts/ml accuracy. So a predictive control strategy of this process could be investigated.