A SIMPLE WAY OF ACHIEVING A HIGH CELL CONCENTRATION IN RECOMBINANT Escherichia coli CULTIVATION

Abstract

Abstract - A cultivation strategy based on some previous knowledge of the metabolism of Escherichia coli BL21 (DE3) pLysS containing the troponin C gene cloned into plasmid pET was developed and applied through the use of simple fermentation equipment and a feed-forward control strategy in order to achieve a high cell concentration <FONT FACE="Symbol">¾</font> 92 g l-1 dry cell weight <FONT FACE="Symbol">¾</font> and a high cell productivity <FONT FACE="Symbol">¾</font> 3.7 g l-1 h-1.

Acetic acid; feed-forward; high cell concentration; recombinant Escherichia coli

A SIMPLE WAY OF ACHIEVING A HIGH CELL CONCENTRATION IN RECOMBINANT Escherichia coli CULTIVATION

A.K. Gombert and B.V. Kilikian ^*

Departamento de Engenharia Química da Escola Politécnica da USP - CP 61548 - CEP 05424-970

São Paulo, SP - Brazil - Phone: [phone] - Fax: [phone] - E-mail: [e-mail]

(Received: February 5, 1997; Accepted: May 26, 1997)

ABSTRACT - A cultivation strategy based on some previous knowledge of the metabolism of Escherichia coli BL21 (DE3) pLysS containing the troponin C gene cloned into plasmid pET was developed and applied through the use of simple fermentation equipment and a feed-forward control strategy in order to achieve a high cell concentration ¾ 92 g l^-1 dry cell weight ¾ and a high cell productivity ¾ 3.7 g l^-1 h^-1.

KEYWORDS: Acetic acid, feed-forward, high cell concentration, recombinant Escherichia coli.

INTRODUCTION

Several heterologous proteins are produced by recombinant microorganisms with different purposes, such as human and animal health, and industrial applications. Escherichia coli is one of the most widely employed hosts in such processes. The achievement of high cell concentrations in recombinant Escherichia coli cultivations is necessary to improve the productivity of the heterologous product when the expression of the target gene is controlled by a repressed promoter ¾ such as the lac promoter or one of its derivatives ¾ which allows the cultivation to be separated into two main phases: cell growth and heterologous gene expression. For this purpose several cultivation strategies have been proposed in the literature (Cutayar and Poillon 1989; Konstantinov et al. 1990; Lee and Chang 1990; Riesenberg et al. 1990; Silman 1990; Kleman et al. 1991; Riesenberg et al. 1991; O’Connor et al. 1992; Yamane et al. 1992). Nevertheless, most of them are based on sophisticated control strategies that demand expensive equipment, which in turn may be a constraint in process development.

In this work a simple cultivation strategy ¾ based both on the control of the glucose concentration by means of a feed-forward strategy imposed through an exponential feeding profile determined from mass balance equations and on the maintenance of the dissolved oxygen concentration above a minimum value ¾ is proposed with the aim of attaining a high cell concentration of a recombinant Escherichia coli strain. The heterologous protein used ¾ chicken muscle troponin C ¾ served rather as a model for the investigation of the performance of the expression system, which is based on the control by the lac UV5 promoter, in bioreactor cultivations.

MATERIALS AND METHODS

Bacterial Strain

Escherichia coli BL21 (DE3) pLysS (Studier and Moffatt 1986) was used as a host to insert the plasmid pET (plasmid for expression by T7 RNA polymerase) bearing the chicken muscle troponin C gene (Quaggio et al. 1993). The expression system is inducible by IPTG or lactose (Gombert and Kilikian 1996).

Inoculum

The inoculum was precultivated in two phases. First, transformed E. coli cells were transferred from TYE solid medium (in g l^-1: bacto-tryptone, 10.0; yeast extract, 5.0; NaCl, 8.0; bacteriological agar, 15.0; pH 7.5) to 50 ml of 2xTY medium (in g l^-1: bacto-tryptone, 16.0; yeast extract, 10.0; NaCl, 5.0; pH 7.0) and incubated in 250 ml shake flasks for 10 h at 37 ° C and 200 rev.min^-1. Second, 0.5 ml of this first preculture was used to inoculate 50 ml of a medium (Fass et al. 1989) containing (g l^-1): glucose, 10.0; KH₂PO₄, 13.0; K₂HPO₄, 10.0; (NH₄)₂HPO₄, 3.0; NaH₂PO₄.H₂O, 4.6; MgSO₄.7H₂O, 0.46; yeast extract, 1.0; and 3 ml of a micronutrient solution (Bauer and Shiloach 1974) containing (g l^-1): FeCl₃.H₂O, 27.0; ZnCl₂, 1.3; CoCl₂.6H₂O, 2.0; Na₂MoO₄.2H₂O, 2.0; CaCl₂.2H₂O, 24.6; CuCl₂, 1.0; H₃BO₃, 0.5; Al₂(SO₄)₃.16H₂O, 1.6; MnSO₄.H₂O, 6.8; and 100 ml of concentrated HCl. This second preculture was incubated in 500 ml shake flasks for 6 h at 37 ° C and 200 rev.min^-1 and served as the bioreactor inoculum (inoculum ratio was 10% (v/v)).

Culture Conditions

The Bio Flo III system (New Brunswick Scientific) was used for the cultivations. Initial working volume was 1.5 l, temperature was kept constant at 37 ° C, pH was monitored by an Ingold pH electrode and controlled at 7.0 ± 0.2 by the addition of a 25% (w/v) NH₃ solution, agitation was kept between 700 and 950 rev.min^-1 and dissolved oxygen (DO) was measured by a New Brunswick polarographic DO probe and maintained within the range of 30% to 100% saturation by manual enrichment of the inlet air stream with pure oxygen whenever it was necessary. Gas flow rate into the bioreactor was kept between 1.0 and 1.5 l l^-1 min^-1.

Bioreactor cultivations were divided into two phases: first, a batch phase with a medium containing (g l^-1): glucose, 5; KH₂PO₄, 13.0; K₂HPO₄, 10.0; (NH₄)₂HPO₄, 3.0; NaH₂PO₄.H₂O, 4.6; MgSO₄.7H₂O, 2; yeast extract, 1.0; and 3 ml of the same micronutrient solution described above; and second, a fed-batch phase with an 800 g glucose /l and 21.4 g MgSO₄.7H₂O /l feeding solution, which was added to the culture according to equations 2, 3, and 4. During the fed-batch phase, carbon source feeding was performed through a nutrient pump that was controlled by the AFS (Advanced Fermentation Software, New Brunswick Scientific). A simple equation derived from a glucose mass balance was loaded into the software to control the feed rate during exponential growth (Yee and Blanch 1992):

(1)

X₀ is the cell concentration at the beginning of the feeding period and m is the prefixed specific growth rate. Parameters X_o and Y_x/s were obtained from previous experiments.

The equations used (according to Figures 1 and 2) were:

F = 3.346 exp (0.3t) (2)

with prefixed m = 0.3 h^-1;

F = 50.0 (3)

for the linear fed-batch phase;

F = 50.0 exp (0.1t) (4)

with prefixed m = 0.1 h^-1.

Carbenicillin and chloranphenicol were added to the precultures and at the beginning of the bioreactor cultivations (working concentration was 50 mg l^-1 of each antibiotic) to promote selective pressure . When the cell concentration was about 18 g dry cell weight (DCW) /l, an injection of another 3 ml of the micronutrient solution was performed.

Analytical Methods

Samples were periodically taken from the bioreactor cultivations in order to analyze dry cell weight (X), glucose (S) and acetic acid concentrations. During the first 5 h of cultivation, the samples were first vacuum filtered through a 0.22 m m membrane to assess X. During the remainder of the experiment, centrifugation at 9500g for 5 min was used instead of filtration. In both cases, the mass of the pellet was determined after drying at 85 ° C for 6 h (filtered samples) or 32 h (centrifuged samples). The supernatant, which was separated from the pellet, had its glucose and acetic acid content determined.

A Technicon Autoanalyzer II (Technicon Instruments Corporation) was used to perform the glucose-oxidase method (Merck) to assess the glucose content of the samples.

A Waters 600E HPLC (Waters Associates) device equipped with a Waters 410 refractometer at its outlet was used to determine acetic acid concentration. For this purpose, a Shodex SH-1011 column at a working temperature of 60 ° C was used with 0.8 ml min^-1 of 0.01 N H₂SO₄ as the mobile phase.

RESULTS AND DISCUSSION

Some previous knowledge of the metabolism of the strain used was essential to develop the cultivation strategy that led to a high cell concentration in the bioreactor.

First, acetic acid is the main fermentation by-product and it inhibits growth. According to Luli and Strohl (1990), the concentration above which acetic acid inhibits growth of E. coli depends both on the strain and the culture conditions. Preliminary experiments indicated that our strain is inhibited in the presence of as little as 0.9 g/L acetic acid under the working conditions described.

Second, glucose and dissolved oxygen (DO) concentrations, besides the specific growth rate, are key factors in acetic acid formation. In order to avoid the inhibitory effects of excess glucose and of a high specific growth rate ¾ a combination of which is called the Crabtree effect ¾ on some enzymes involved in the aerobic dissimilation of glucose which may lead to an increase in fermentation by-product accumulation (Doelle 1975; Doelle 1981), glucose concentration was limited, according to equation 1, in order to promote a specific growth rate of 0.3 h^-1 or below this value during the whole fed-batch period (equations 2, 3, and 4). This values are far below the maximum specific growth rate, which is around 0.8 h^-1, according to preliminary experiments. In order to avoid the effects of oxygen transfer limitation, which might lead to an increase in the fermentative activity of E. coli ¾ also known as the Pasteur effect (Doelle 1975; Doelle 1981) ¾ pure oxygen was supplied to the inlet gas to keep DO concentration between 30% and 100% saturation.

Moreover, the identification of the oxygen transfer capacity as an equipment limitation was also a key factor in process development. Oxygen consumption is a function of the cell concentration and of the specific growth rate. Therefore, a simple way of avoiding this would be to promote cell growth at a low specific growth rate, such as reported by Lee et al. (1989). However, the risk of plasmid loss would become higher and the productivity would drop to very low values. Therefore, a specific growth rate of 0.3 h^-1 was chosen as the best one to promote growth until any of the problems mentioned above arose, i. e., acetic acid accumulation to inhibitory values or insufficient oxygen transfer. Specific growth rates above this value were not chosen due to the significant raise in the specific rate of acetic acid production (data not shown) that occurs in these cases, even with limited glucose concentration. The whole strategy may be represented by the flow rate into the reactor, which is indicated in Figure 1.

The first problem that arose, as a result of which the initial exponential feed was changed to a linear feed (point A), was with the oxygen transfer capacity of the equipment. Nevertheless, in a general way it could also have been acetic acid accumulation had a strain with a higher acetic acid-producing capacity or with a higher sensitivity towards this substance been utilised. The linear flow rate used forced the specific growth rate to drop from 0.3 h^-1 at point A to 0.1 h^-1 at point B (Figure 2) and then maintained it at this value until the end of the cultivation by means of an exponential feed. Specific growth rates below this value were not chosen, as it had been verified that cell lysis occurred in these cases. At point C (around 23 h of cultivation) it was not possible to keep the dissolved oxygen concentration above 30% saturation and the cultivation was interrupted at point D.

In Figure 1 it is possible to verify that the DO concentration is a key factor in acetic acid accumulation, as there is an increase in the concentration of this product due to the drop in DO concentration to values around 10% saturation at point C. From the region representing the initial batch phase in Figure 1, it is possible to verify that the combination of unlimited values of glucose concentration with a high specific growth rate (above 0.4 h^-1) is also a key factor in acetic acid accumulation, as there is an increase in the concentration of this product followed by its consumption when glucose concentration is limited and m decreases.

In Figure 2 it is possible to see that the specific growth rate varied between 0.2 h^-1 and 0.4 h^-1 during the first part of the fed-batch phase, which indicates that the control prefixed by equation 1 should have been improved if a better adjustment of m were desired. However, the control proved to operate well for lower values of m , as may be seen from Figure 2 (third part of the fed-batch phase).

CONCLUSIONS

The present results show that it is possible to achieve a high cell concentration (92 g DCW /l) with a high productivity (3.7 g l^-1 h^-1) in the bioreactor using a simple strategy based on feed-forward control and some knowledge of the metabolism of the E. coli strain used.

Figure 1: Flow rate (¾ ), glucose concentration (¡ ) and acetic acid concentration (n ).

Figure 2: Cell concentration (D) and specific growth rate (¾ ).

The acquisition of such high values of cell concentration and productivity on the basis of some basic knowledge of cell metabolism and conventional equipment is scarce in the literature. Highly automated systems are commonly cited resulting, not rarely, in lower cell productivity values.

Dissolved oxygen, glucose concentration and specific growth rate, as well as oxygen transfer, were identified as key factors in our experiment. The devices used included a bioreactor and a computer-controlled peristaltic pump. The same feed- forward strategy based on the substrate mass balance had been previously applied to shorter runs via manual adjustments of the flow rate through the pump each 30 minutes, which resulted in similar flutuations of m . This indicates that a computer-controlled pump is not essential to perform substrate addition in the way it is described in this paper.

ACKNOWLEDGEMENTS

Support for this research by the Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) is gratefully acknowledged. We also thank Dr F.C. Reinach for making the recombinant strain available.

NOMENCLATURE

DO Dissolved oxygen concentration (%)

F Flow of the feeding solution into the reactor, ml h^-1

S Glucose concentration, g ml-1

S_o Glucose concentration in the feeding solution, g ml-1

t Time, h

V Volume at the beginning of the feeding period, l

XCell concentration, g DCW l

m Specific growth rate, h^-1

Y_x/s Cell yield on glucose, g DCW g glucose^-1

REFERENCES

Bauer, S. and Shiloach, J., Maximal Exponential Growth Rate and Yield of Escherichia coli Obtainable in a Bench-Scale Fermentor. Biotechnology and Bioengineering 17: 933-941 (1974).

Cutayar, J.M. and Poillon, D., High Cell Density Culture of Escherichia coli in a Fed-Batch System with Dissolved Oxygen as Substrate Feed Indicator. Biotechnology Letters 11: 155-160 (1989).

Doelle, H.W., Bacterial Metabolism. 2nd ed. New York: Academic Press (1975).

Doelle, H.W., Basic Metabolic Processes. In: Biotechnology: A Comprehensive Treatise (Rehm H J and Reed G, eds.), pp. 193-201, Verlag Chemie, Weinheim (1981).

Fass, R.; Clem, T.R. and Shiloach, J., Use of a Novel Air Separation System in a Fed-Batch Fermentation Culture of Escherichia coli. Applied and Environmental Microbiology 55: 1305-1307 (1989).

Gombert, A.K. and Kilikian, B.V., Improving the Expression of a Recombinant Gene in Escherichia coli Using Lactose as Inducer. In: Abstracts of the 8th International Congress of Bacteriology and Applied Microbiology Division, pp. 30, Jerusalem: International Union of Microbiological Societies (1996).

Kleman, G.L.; Chalmers, J.J.; Luli, G.W. and Strohl, W.R., A Predictive and Feedback Control Algorithm Maintains a Constant Glucose Concentration in Fed-Batch Fermentations. Applied and Environmental Microbiology 57: 910-917 (1991).

Konstantinov, K.; Kishimoto, M.; Seki, T. and Yoshida, T., A Balanced DO-Stat and Its Application to the Control of Acetic Acid Excretion by Recombinant Escherichia coli. Biotechnology and Bioengineering 36: 750-758 (1990).

Lee, Y.L. and Chang, H.N., High Cell Density Culture of a Recombinant Escherichia coli Producing Penicillin Acylase in a Membrane Cell Recycle Fermentor. Biotechnology and Bioengineering 36: 330-337 (1990).

Lee, J.H.; Choi, Y.H.; Kng, S.K.; Park, H.H. and Kwon, I.B., Production of Human Leukocyte Interferon in Escherichia coli by Control of Growth Rate in Fed-Batch Fermentation. Biotechnology Letters 11: 695-698 (1989).

Luli, G.W. and Strohl, W.R., Comparison of Growth, Acetate Production, and Acetate Inhibition of Escherichia coli Strains in Batch and Fed-Batch Fermentations. Applied and Environmental Microbiology 56: 1004-1011 (1990).

O’Connor, G.M.; Sanchez-Riera, F. and Cooney, C.L., Design and Evaluation of Control Strategies for High Cell Density Fermentations. Biotechnology and Bioengineering 39: 293-304 (1992).

Quaggio, R.B.; Ferro, J.A.; Monteiro, P.B. and Reinach, F.C., Cloning and Expression of Chicken Skeletal Muscle Troponin I in Escherichia coli: the Role of Rate Codons on the Expression Level. Protein Science 2: 1053-1056 (1993).

Riesenberg, D.; Menzel, K.; Schulz, V.; Schumann, K.; Veith, G.; Zuber, G. and Knorre, W., High Cell Density Fermentation of Recombinant Escherichia coli Expressing Human Interferon Alpha 1. Applied Microbiology and Biotechnology 34: 77-82 (1990).

Riesenberg, D.; Schulz, V.; Knorre, W.A.; Pohl, H.D.; Korz, D.; Sanders, E.A.; Ross, A. and Deckwer, W.D., High Cell Density Cultivation of Escherichia coli at Controlled Specific Growth Rate. Journal of Biotechnology 20: 17-27 (1991).

Silman, R.W., Control of Feed Rate to a Fed-Batch Culture Using a Heat-Flux Sensor. Biotechnology Techniques 4: 397-402 (1990).

Studier, F.W. and Moffatt, B.A., Use of Bacteriophage T7 RNA Polymerase to Direct Selective High-Level Expression of Cloned Genes. Journal of Molecular Biology 189: 113-130 (1986).

Yamane, T.; Hibino, W.; Ishihara, K.; Kadotani, Y. and Kominami, M., Fed-Batch Culture Automated by Uses of Continuously Measured Cell Concentration and Culture Volume. Biotechnology and Bioengineering 39: 550-555 (1992).

Yee, L. and Blanch, H.W., Recombinant Protein

Expression in High Cell Density Fed-Batch Cultures of Escherichia coli. Bio/Technology 10: 1550-1556 (1992).

*

To whom correspondence should be addressed

Publication Dates

• Publication in this collection
  09 Oct 1998
• Date of issue
  June 1997

History

• Accepted
  26 Mar 1997
• Received
  05 Feb 1997