Uncoupling growth and protein production

Abstract

The present invention is in the field of recombinant biotechnology, in particular in the field of protein expression. The invention generally relates to methods of increasing the expression level of a protein of interest of a bacterial host cell in a production process. The invention relates particularly to improving the capacity of a bacterial host cell to express a protein of interest by expressing a phage protein during the production process which inhibits growth of the bacterial host cell. Decoupling growth of the bacterial host cell of manufacturing of the protein of interest during the production process reduces (i) the metabolic burden, (ii) oxygen demand, (iii) metabolic heat development, and (iv) avoids stress response caused by heterologous protein expression and thereby increases the capacity of a host cell to produce the protein of interest. The present invention also relates to uses of the host cell for protein expression, cell culture technology, and more specifically to culturing host cells to produce a protein of interest.

Claims

1. A bacterial host cell which is E. coli and which (i) comprises under the control of an inducible promoter a nucleotide sequence encoding a phage protein which inhibits growth of said bacterial host cell; and (ii) comprises a nucleotide sequence encoding a RNA polymerase which is heterologous for said bacterial host cell, wherein sail RNA polymerase, which is heterologous for said bacterial host cell is bacteriophage T3 RNA polymerase, T7 bacteriophage RNA polymerase, engineered orthogonal TI RNA polymerase, bacteriophage SP6 RNA polymerase or bacteriophage Xp10 RNA polymerase; and (iii) comprises a nucleotide sequence which encodes a protein of interest under the control of a promoter recognized by said RNA polymerase, which is heterologous for said host cell; wherein said nucleotide sequence encoding a phage protein which inhibits growth of said bacterial host cell and sail nucleotide sequence encoding said RNA polymerase, which is heterologous for said bacterial host cell are integrated into the genome of said host cell, wherein said phage protein which inhibits growth of said bacterial host cell is (a) a protein having the amino acid sequence shown in Seq Id No: 1 which inhibits bacterial host cell RNA polymerase; or (b) a protein having an amino acid sequence which has an identity of 80% or more to the full-length of the amino acid sequence shown in Seq Id No: 1 and which inhibits bacterial host cell RNA polymerase.

2. The bacterial host cell of claim 1, wherein said nucleotide sequence encoding a protein of interest is comprised by an extrachromosomal vector.

3. The bacterial host cell of claim 1, wherein said nucleotide sequence encoding said RNA polymerase, which is heterologous for said bacterial host cell, is under the control of an inducible promoter.

4. The bacterial host cell of claim 3, wherein said inducible promoter is regulated by arabinose, IPTG, tryptophane, xylose, rhamnose, phosphate or phage lambda cI protein.

5. The bacterial host cell of claim 1, wherein said host cell has a non-functional arabinose operon.

6. A preparation of a bacterial host cell which is E. coli and which (i) comprises under the control of an inducible promoter a nucleotide sequence encoding a phage protein which inhibits growth of said bacterial host cell wherein said phage protein which inhibits growth of said bacterial host cell is (a) a protein having the amino acid sequence shown in Seq Id No: 1 which inhibits bacterial host cell RNA polymerase; or (b) a protein having an amino acid sequence which has an identity of 80% or more to the full-length of the amino acid sequence shown in Seq Id No: 1 and which inhibits bacterial host cell RNA polymerase; and (ii) comprises a nucleotide sequence encoding an RNA polymerase which is heterologous for said bacterial host cell wherein said RNA polymerase, which is heterologous for said bacterial host cell is bacteriophage T3 RNA polymerase, T7 bacteriophage RNA polymerase, engineered orthogonal T7 RNA polymerase, bacteriophage SP6 RNA polymerase or bacteriophage Xp10 RNA polymerase; and (iii) comprises a nucleotide sequence which encodes a protein of interest under the control of a promoter recognized by said RNA polymerase, which is heterologous for said host cell, wherein said nucleotide sequence encoding a phage protein which inhibits growth of said bacterial host cell and said nucleotide sequence encoding said RNA polymerase, which is heterologous for said bacterial host cell are integrated into the genome of said host cell.

7. A method for the production of a protein of interest, comprising culturing the bacterial host cell of claim 1 under suitable conditions and obtaining said protein of interest.

8. The method of claim 7, wherein said culturing step includes (a) growing the bacterial cells to a density of at least 20 g/L cell dry mass (CDM); (b) inducing expression of the nucleotide sequence encoding a phage protein which inhibits growth of the host cell; (c) feeding bacterial cells with a constant linear feed rate that would allow an initial growth rate of 0.05 h.sup.−1; and (d) further culturing said bacterial cells for at least 12 hours.

9. A method for increasing the yield of a protein of interest, comprising transforming a bacterial host which is E. coli and which comprises: (i) a nucleotide sequence encoding an RNA polymerase being heterologous for said bacterial host cell, wherein said RNA polymerase, which is heterologous for said bacterial host cell is bacteriophage T3 RNA polymerase, T7 bacteriophage RNA polymerase, engineered orthogonal T7 RNA polymerase, bacteriophage SP6 RNA polymerase or bacteriophage XP10 RNA polymerase; and (ii) a nucleotide sequence which encodes said protein of interest, wherein said nucleotide sequence, which encodes said protein of interest is under the control of a promoter which is recognized by said RNA polymerase being heterologous for said bacterial host cell; with a nucleotide sequence encoding a phage protein which inhibits growth of said bacterial host cell, wherein said nucleotide sequence encoding a phage protein which inhibits growth of said bacterial host cell is under the control of an inducible promoter, wherein said nucleotide sequence encoding a phage protein which inhibits growth of said bacterial host cell and sail nucleotide sequence encoding said RNA polymerase being heterologous for said bacterial host cell are integrated into the genome of sail host cell, and wherein said phage protein which inhibits growth of said bacterial host cell is (a) a protein having the amino acid sequence shown in Seq Id No: 1 which inhibits bacterial host cell RNA polymerase; or (b) a protein having an amino acid sequence which has an identity of 80% or more to the full-length of the amino acid sequence shown in Seq Id No: 1 and which inhibits bacterial host cell RNA polymerase.

10. A method for the production of a protein of interest, comprising bringing into contact under suitable conditions the preparation of claim 6 with a nucleotide sequence comprising under the control of a promoter recognized by an RNA polymerase as defined in claim 6 a nucleotide sequence which encodes a protein of interest.

11. The method of claim 9, wherein the protein of interest is toxic for cells, and adversely affects viability, cell growth and/or cell division.

12. The method of claim 9, further comprising modifying the protein of interest and/or formulating the protein of interest into a composition which includes at least one additional component.

13. The method of claim 12, wherein said protein of interest is modified with a label.

14. The bacterial host cell of claim 1, wherein said nucleotide sequence encoding said RNA polymerase which is heterologous for said bacterial host cell is under the control of a constitutive promoter.

Description

FIGURES

(1) FIG. 1: The induced expression of Gp2 inhibits the host cell growth of E. coli strain NEB10-beta in a dose-dependent manner.

(2) The Gp2 encoding sequence under control of an arabinose-inducible promoter was cloned in the low-copy f-plasmid pKLJ12 (Jones and Keasling (1998), Biotechnol Bioengineer 59: 659-665). The plasmid pKLJ12+Gp2 was transformed in E. coli strain NEB10-beta which comprises an araD139 mutation. Consequently, NEB10-beta is not capable of metabolizing the inducer arabinose. Addition of different concentrations of arabinose and thereby expression of Gp2 cause an inhibition of proliferation in a dose-dependent manner.

(3) FIG. 2: Addition of arabinose has no effect on host cell growth.

(4) In order to exclude any effect of the compound arabinose on host cell growth a derivative of plasmid pKLJ12+Gp2 lacking the ribosome binding site of the Gp2 expression cassette has been employed. Consequently, expression of Gp2 cannot be induced and hence no difference in proliferation was observed with or without arabinose.

(5) FIG. 3: The induced expression of Gp2 increases the expression level of the model protein GFP.

(6) E. coli strain HMS174(DE3) TN7::<T7GFP>, comprising the GFP encoding sequence under control of a T7 promoter and T7 RNA polymerase under control of an IPTG-inducible promoter, was transformed with the plasmid pKLJ12+Gp2 which harbors the Gp2 encoding sequence under control of an arabinose-inducible promoter. Three consecutive experiments showed that host cells induced with IPTG and arabinose, expressing T7 RNA polymerase and Gp2, expressed GFP to a higher extent compared to host cells that were induced with IPTG, only, and therefore expressed T7 RNA polymerase but not Gp2.

(7) FIG. 4: Expression of Gp2 from the pKLJ12+Gp2 insert without the vector increases the GFP expression.

(8) E. coli strain HMS174(DE3) TN7::<T7GFP>, comprising the GFP encoding sequence under control of a T7 promoter and T7 RNA polymerase under control of an IPTG-inducible promoter, was transformed with the insert of pKLJ12+Gp2 comprising the Gp2 expression cassette. In two out of three cases the transformation of the Gp2 expression cassette resulted in an increased GFP expression 3 h after IPTG induction.

(9) FIG. 5: Amino acid sequences of exemplary, but nevertheless preferred (phage) proteins which inhibit growth of a bacterial host cell.

(10) FIG. 6: Reference fermentation process lacking Gp2 expression.

(11) In the reference fermentation process Gp2 is not expressed and therefore cells continue growing during production of the model protein GFP, as expected. Consequently, total CDM (cell dry mass) and CDM without recombinant protein increase during the entire fermentation process. Induced expression of GFP results in a constant increases of both, specific soluble GFP and total soluble GFP during the entire fermentation process.

(12) FIG. 7: Example fermentation process in which growth and protein production have been uncoupled by Gp2 expression.

(13) In the example fermentation process Gp2 expression causes a growth arrest of the cells. Consequently, CDM without recombinant protein remains constant upon induction of Gp2 expression at the time point 11 h whereas total CDM increases moderately due to the production of recombinant GFP. Both, specific soluble GFP and total soluble GFP increase during the course of the fermentation process despite growth arrest of the cells.

(14) FIG. 8: Expression of Gp2 increases the expression level of GFP and the ratio of GFP to soluble host cell protein (HCP) in the supernatant.

(15) The coomassie stained SDS PAGE gel shows an increase of soluble GFP in the supernatant (S) using the growth decoupled system (E. coli BL21 (DE3) with genome integrated inducible Gp2 protein compared to a standard system (E. coli BL21 (DE3) without genome integrated inducible Gp2 protein). Furthermore, the relative amount of GFP to HCP (excluding Lysozyme) in the supernatant is considerably higher using the growth decoupled system compared to a standard system. Additionally, solubility of GFP is improved by using the growth decoupled system compared to a standard system.

(16) FIG. 9: Reference process scheme. Induction with 0.1 mM IPTG.

(17) FIG. 10: Process scheme of fed-batch cultivations and induction strategy with Arabinose and IPTG.

(18) FIG. 11: SDS page analysis of shake flask cultivations of BL21 (DE3)::TN7(Gp2ΔAra)pET30(HIV1-protease) and BL21(DE3)pET30(HIV1-protease).

(19) Comparison of induced and non-induced samples of reference strain and model strain cultivation. HIV1-protease band is located at 11 kDa.

(20) FIG. 12: Growth and product formation kinetics of strain BL21(DE3)pET30a(HIV1-protease) and BL21(DE3):TN7<GP2ΔAra>pET30a(HIV1-protease).

(21) (A) Reference fermentation process: Induction with 20 μmol IPTG/g CDM at feed 21 h with exponential feed rate of μ=0.10 h.sup.−1; (B) Model fermentation process: Induction with 0.1 M arabinose+20 μmol IPTG/g at feed 11 h where exponential feed (μ=0.20 h.sup.−1) was switched to linear feed;

(22) FIG. 13: SDS page analysis of reference process fermentation with BL21(DE3)pET30(HIV1-protease).

(23) FIG. 14: SDS PAGE analysis: Cultivation of BL21(DE3)::TN7(Gp2ΔAra)pET30(HIV1-protease).

(24) FIG. 15: Comparison of HIV1-protease production yields by BL21(DE3)::TN7(Gp2ΔAra)pET30(HIV1-protease) [model process, green] and BL21(DE3)pET30a(HIV1-protease) [reference process, red].

(25) (A, C, E) Comparison of produced total, soluble and insoluble HIV1-protease; (B, D) Comparision of total CDM and net CDM; (F) Total HIV1-protease produced by both systems.

(26) FIG. 16: Established feed profile for growth decoupled protein expression in HCD bioreactor fed-batch cultivation.

(27) Theoretical trends of growth curve and growth rate, calculation based on a constant glucose yield coefficient throughout the cultivation.

(28) FIG. 17: SDS page analysis of HCD cultivation of. BL21(DE3)::TN7(Gp2ΔAra)pET30(GFPmut3.1).

(29) Induced with 20 μmol IPTG/g CDM and 0.1 M arabinose at feed hour 15 where exponential feed (μ=0.17 h.sup.−1) was switched to linear feed. GFP band is located at 27 kDa.

(30) FIG. 18: Growth and product formation kinetics of HCD cultivation of strain BL21(DE3):TN7<GP2ΔAra>pET30a(GFPmut3.1).

(31) Induction with 0.1 M arabinose+20 μmol IPTG/g at feed 15 h where exp. feed (μ=0.20 h.sup.−1) was switched to linear feed.

(32) FIG. 19: Hourly (A) and total (B) O.sub.2 consumption and CO.sub.2 formation during HCD cultivation of growth decoupled system.

(33) Induction with 0.1 M arabinose+20 μmol IPTG/g CDM at feed 15 h with exponential feed rate of μ=0.17 h.sup.−1; Feed medium supplemented with (NH.sub.4).sub.2SO.sub.4.

(34) FIG. 20: Comparison of GFPmut3.1 production yields between HCD and non-HCD cultivation of BL21(DE3)::TN7(Gp2ΔAra)pET30(GFPmut3.1).

(35) (A, C, E) Comparison of produced total, soluble and insoluble GFPmut3.1; (B, D) Comparison of total CDM and net CDM; (F) Total GFPmut3.1 produced.

EXAMPLES

(36) The following Examples illustrate the invention, but are not to be construed as limiting the scope of the invention.

Example 1: Inhibition of the Host Cells RNA Polymerase Inhibits Growth of the Host Cell

(37) In order to assess the effect of inhibition of the host cells RNA polymerase on proliferation of the host cell the Gp2 encoding sequence under control of an arabinose-inducible promoter was cloned in the low-copy f-plasmid pKLJ12 (Jones and Keasling (1998) Biotechnol Bioeng Vol. 59, Issue 6: 659-665), which constitutes only a small burden to the host cell and is stably maintained. The protein Gp2 is known to inhibit the host cell RNA polymerase by binding to the beta-subunit of the enzyme. The plasmid pKLJ12+Gp2 was transformed in E. coli strain NEB10-beta which comprises an araD139 mutation. Consequently, NEB10-beta is not capable of metabolizing the inducer arabinose. At the time point 0 h several cultures were inoculated and Gp2 expression was induced after 2 h by addition of 1.5%, 0.1% or 0.001% arabinose. Proliferation of the bacteria was measured by determining the OD600 nm value. Addition of different concentrations of arabinose and thereby expression of Gp2 caused an inhibition of proliferation in a dose-dependent manner in comparison to a bacteria culture where arabinose has been omitted (FIG. 1). In order to exclude any effect of the compound arabinose on host cell growth a derivative of plasmid pKLJ12+Gp2 lacking the ribosome binding site of the Gp2 expression cassette has been employed. Consequently, expression of Gp2 cannot be induced and hence no difference in proliferation was observed with or without arabinose (FIG. 2).

Example 2: Integration of the Gp2 Encoding Expression Cassette in the Host Cells Genome

(38) In order to confer a stable expression of the Gp2 protein in the host cell population the Gp2 encoding expression cassette was integrated in the genome of NEB10-beta via homologous recombination at the TN7 locus. The inserted sequence comprised the Gp2 gene under control of an arabinose promoter, a regulator, a terminator, and an ampicillin resistance gene. To this end, a 50 bp overhang was added to the insertion element via PCR. The linear PCR product was concomitantly transformed into the NEB10-beta host cell with a pSIM helper plasmid, which confers heat shock induced expression of proteins mediating the integration of the PCR product in the host cells genome. After successful integration the pSIM plasmid can be withdrawn from the host cell (Sharan et al., 2009, Nat Protoc. 4(2):206-23).

Example 3: Induced Expression of Gp2 in E. coli Increases the Expression of the Model Protein GFP

(39) E. coli strain HMS174(DE3) TN7::<T7GFP>, comprising the GFP encoding sequence under control of a T7 promoter and T7 RNA polymerase under control of an IPTG-inducible promoter, was transformed with the plasmid pKLJ12+Gp2 which harbours the Gp2 encoding sequence under control of an arabinose-inducible promoter. About 2 h after inoculation of the culture IPTG was added to induce the expression of T7 RNA polymerase and thereby GFP. About 3 h after inoculation arabinose was added in order to express Gp2. Three consecutive experiments showed that host cells induced with IPTG and arabinose, expressing T7 RNA polymerase and Gp2, expressed GFP to a higher extent compared to host cells that were induced with IPTG, only, and therefore expressed T7 RNA polymerase but not Gp2. In the two samples lacking IPTG with or without arabinose GFP was slightly expressed due to leakiness of the promoter (FIG. 3).

Example 4: Expression of Gp2 from the pKLJ12+Gp2 Insert without the Vector Increases the GFP Expression

(40) E. coli strain HMS174(DE3) TN7::<T7GFP>, comprising the GFP encoding sequence under control of a T7 promoter and T7 RNA polymerase under control of an IPTG-inducible promoter, was transformed with the insert of pKLJ12+Gp2 comprising the Gp2 expression cassette. In two out of three cases the transformation of the Gp2 expression cassette resulted in an increased GFP expression 3 h after IPTG induction (FIG. 4).

Example 5: Description of an Example Fermentation Process in which the Effect of Gp2 Expression on the Yield of the Model Protein GFP was Assessed

(41) Cultivation Mode and Process Analysis

(42) The cells are grown in a 12 L (8 L net volume, 4 L batch volume) computer-controlled bioreactor (MBR; Wetzikon, CH) equipped with standard control units. The pH is maintained at a set-point of 7.0±0.05 by addition of 25% ammonia solution (ACROS Organics), the temperature is set to 37° C.±0.5° C. In order to avoid oxygen limitation, the dissolved oxygen level is stabilized above 30% saturation by stirrer speed and aeration rate control. Fluorescence measurements are performed using a multi-wavelength spectrofluorometer specially designed for online measurements in an industrial environment, the BioView® (DELTA Light & Optics, Lyngby, Denmark). Foaming is suppressed by addition of antifoam suspension (PBG2000) with a concentration of 0.5 ml/l medium. For inoculation, a deep frozen (−80° C.) working cell bank vial, is thawed and 1 ml (optical density OD.sub.600=1) is transferred aseptically to the bioreactor. Feeding is started when the culture, grown to a bacterial dry matter of 22.5 g in 4 L batch medium, entered stationary phase. With start of the feed phase cultivation temperature is reduced to 30° C. The feed medium provided sufficient components to yield another 363 g of bacterial dry matter (4 doublings).

(43) In the reference process (FIG. 6) the (standard) expression system E. coli BL21(DE3)pET30a GFPmut3.1 was used. The growth rate in the feed phase was set to 0.1 h.sup.−1 and 3 doublings past feed start induction of recombinant gene expression was conducted with 20 μmol IPTG per gram CDM by a single pulse directly to the bioreactor.

(44) In the process with the (standard) expression system E. coli BL21(DE3)pET30a GFPmut3.1 containing a genome integrated inducible Gp2 protein (FIG. 7) the growth rate in the feed phase was set to 0.2 h.sup.−1 for 3 doublings via an exponential substrate feed. Afterwards induction with 20 μmol IPTG per gram cell dry mass and 10 mmol arabinose is conducted and the medium feed is switched to a linear feed for another 16 h with an initial growth rate of 0.05 h.sup.−1. The substrate feed is controlled by increasing the pump speed according to the exponential growth algorithm, x=x.sub.0.e.sup.μt, with superimposed feedback control of weight loss in the substrate tank.

(45) Media Composition

(46) The minimal medium used in this study contains 3 g KH.sub.2PO4 and 6 g K.sub.2HPO4*3H.sub.2O per litre. These concentrations provide the required buffer capacity and serve as P and K source as well. The other components are added in relation of gram bacterial dry matter to be produced: sodium citrate (trisodium salt *2H.sub.2O; ACROS organics) 0.25 g, MgSO.sub.4*7H.sub.2O 0.10 g, CaCl.sub.2*2H.sub.2O 0.02 g, trace element solution 50 μl and glucose*H.sub.2O 3 g. To accelerate initial growth of the population, the complex component yeast extract 0.15 g is added to the minimal medium to obtain the batch medium. For the feeding phase 8 L of minimal medium are prepared according to the amount of biological dry matter 363 g to be produced in the feeding phase, whereby P-salts are again added per litre. Trace element solution: prepared in 5 N HCl (g/L): FeSO.sub.4*7H.sub.2O 40.0, MnSO.sub.4*H.sub.2O 10.0, AlCl.sub.3*6H.sub.2O 10.0, CoCl.sub.2 (Fluka) 4.0, ZnSO.sub.4*7H.sub.2O 2.0, Na.sub.2MoO.sub.2*2H.sub.2O 2.0, CuCl.sub.2*2H.sub.2O 1.0, H.sub.3BO.sub.3 0.50.

(47) Offline Analysis

(48) Optical density (OD) is measured at 600 nm. Bacterial dry matter is determined by centrifugation of 10 ml of the cell suspension, re-suspension in distilled water followed by centrifugation, and re-suspension for transfer to a pre-weighed beaker, which is then dried at 105° C. for 24 h and re-weighed. The progress of bacterial growth is determined by calculating the total amount of cell dry mass (total CDM).

(49) The content of recombinant protein GFP is determined by ELISA and electrophoretic protein quantification using densitometric quantification of bands on an SDS-PAGE gel. Soluble recombinant product is quantified via GFP-ELISA, while the recombinant product in the inclusion bodies is determined with SDS-PAGE gel electrophoresis.

(50) Additionally, supernatant and inclusion bodies were analysed using SDS-PAGE gel electrophoresis. The coomassie stained SDS PAGE gel shows an increase of soluble GFP in the supernatant using the growth decoupled system compared to a standard system. Furthermore, the relative amount of GFP to HCP in the supernatant is considerably higher using the growth decoupled system compared to a standard system. Additionally, solubility of GFP is improved by using the growth decoupled system compared to a standard system (BL21(DE3)pET30a GFPmut3.1) which is not growth decoupled (FIG. 8).

Example 6: Production of HIV-1 Protease Using the Growth Decoupled System

(51) To prove the applicability of the developed growth decoupled process, alternative recombinant proteins were required. For that purpose HIV-1 protease was selected as second model protein for verification of the growth decoupled system and the model process, as it is difficult to produce because it is highly toxic for E. coli (Korant and Rizzo, (1991), Biomed Biochim Acta 50: 643-6). Overexpression of this aspartic protease from the human immunodeficiency virus type 1 in E. coli is usually accompanied by toxic effects on the producing cells (Fernandez et al., (2007), Biotechnol Lett 29: 1381-6), possibly linked to its proteolytic activity. Consequently, this protein is generally difficult to express in microbial systems. The retroviral proteins are synthesized as polyprotein precursors and are processed by specific proteases (Volonté et al., (2011), Microb Cell Fact 10: 53). These precursors are Gag and Gag-Pol polypeptides, which are proteolytically processed by HIV-1 protease to mature proteins (Kohl et al., (1988), Proc Natl Acad Sci USA 85: 4686-90).

(52) HIV-1 protease is encoded by HI-virus and thereby plays an important role in the maturation of the virus. It is an appealing target for development of a possible treatment of the acquired immune deficiency syndrome (AIDS). The availability of a system which can express large amounts of HIV-1 protease in bacterial cultivation systems is the ultimate goal of obtaining large quantities of this protein. (Volonté et al., (2011), Microb Cell Fact 10: 53)

(53) Reference Fermentation Process:

(54) The batch phase of the cultivation was performed at a temperature of 37° C. and was inoculated with 1 mL of working cell bank (WCB). Depending on the experiment, following strains, containing two different model proteins, were used for this process scheme: BL21(DE3)pET30(GFPmut3.1) BL21(DE3)pET30(HIV1-protease)

(55) The batch phase was completed after 11 h to 13 h (indicated by a peak in dissolved oxygen) and the feed phase was started immediately afterwards. In the exponential feed phase the temperature was decreased to 30° C. in order to reduce inclusion body formation of the expressed recombinant protein as well as to achieve better 02-solubility. The growth rate (μ) of the fed batch process was kept constant at 0.10 h.sup.−1 by an exponential substrate feed for 4 generations. Feeding was initiated after the cell dry mass (CDM) reached the end of the batch phase with 22.5 g CDM in 4 L batch volume.

(56) Induction with a single pulse of IPTG (20 μmol/g CDM) took place after the 3.sup.rd generation (21 h after feed start) in the feed-phase. The sampling procedure lasted for 1 generation. An overview on the reference process scheme is shown in FIG. 9.

(57) Fermentation Process/Growth Decoupled Production System:

(58) The batch phase of the cultivation was performed at a constant temperature of 37° C. and was inoculated with 1 mL of WCB. The following systems were cultivated with this process scheme: BL21(DE3)::TN7(Gp2ΔAra)pET30(GFPmut3.1) BL21(DE3)::TN7(Gp2ΔAra)pET30(HIV1-protease)

(59) The batch-phase completed after 11 h to 13 h and the feed phase was started immediately afterwards. In the exponential feed phase the temperature was decreased to 30° C. in order to achieve a better solubility of the expressed recombinant protein and to reach a better oxygen transfer rate (OTR). The growth rate was kept constant at μ=0.20 h.sup.−1. The recombinant protein production was induced with a single pulse of 0.1 M arabinose (Gp2) and 20 μmol IPTG/g CDM (gene of interest—GOI) after the 3rd generation (21 h after feed start). During the 4th generation sampling took place, a linear feed profile was applied starting with an initial growth rate of μ=0.050 h.sup.−1 that decreased to μ=0.025 h.sup.−1 in the course of the experiment. The sampling procedure lasted for 1 generation. An overview on the reference process scheme is shown in FIG. 10.

(60) HIV-1 Protease Production in Escherichia coli:

(61) In order to prove broad applicability of the platform process, experiments with alternative recombinant proteins are required. For that purpose HIV-1 protease, a protein difficult to be produced in E. coli, was selected for benchmarking experiments.

(62) Before bioreactor cultivations of the reference strain BL21(DE3)pET30a(HIV1-protease) and the model strain BL21(DE3)::TN7(Gp2ΔAra)pET30(HIV1-protease) were carried out, standard shake flask cultivations were performed for verification if the recombinant protein is produced. HIV1-protease band is located at 11 kDa under the lysozyme band. Following linear equation was used for quantification of HIV1-protease:
y=0.0007x R.sup.2=0.9708

(63) According FIG. 11, both strains were able to produce HIV1-protease in insoluble form. Production of the growth decoupled system yielded a concentration of 69 μg/mL, while the reference system produced only 5 μg/mL insoluble HIV1-protease. Consequently BL21(DE3)::TN7(Gp2ΔAra)pET30(HIV1-protease) was able to produce 13 times more HIV1-protease than the reference strain.

(64) As both strains were capable of producing the model protein, lab scale cultivations of both systems were performed. BL21(DE3)pET30a(HIV1-protease) was used as reference system and the fermentation process was performed according to the description above. In parallel the new platform process with the growth decoupled system was performed as described in FIG. 10 with BL21(DE3)::TN7(Gp2ΔAra)pET30(HIV1-protease). A growth rate of p=0.20 h.sup.−1 was applied during the exponential feed phase.

(65) As shown in graph A of FIG. 12, the maximal specific concentration of HIV-1 protease for the standard process was 11.8 mg/g CDM with no soluble expression of the protein. The obtained volumetric yield with 0.3 g/L is also very low. This result confirms the statement that HIV-1 protease belongs to the group of low yield and difficult-to-express proteins (Volonté et al., (2011), Microb Cell Fact 10: 53; Wörsdörfer et al., (2011), Science 331: 589-92). Graph B shows that in the model process, the growth of CDM stopped after induction and there was no decrease in net CDM. At the end of the process the growth decoupled system produced 233.5 g CDM in total. Compared to the reference process, which produced 331.92 g CDM, the model process generated 30% less CDM. After induction of the protein production, the model process also consumed 155.2 g less base compared to the reference system. FIG. 13 shows SDS page analysis of the reference process. Following linear equation was used for quantification of the reference process fermentation:
y=0.0006x R.sup.2=0.9981

(66) According to FIG. 13, the reference system was only capable of producing HIV1-protease in insoluble form (IB). At the beginning of the protein expression phase, BL21(DE3)pET30(HIV1-protease) produced 2 μg insoluble HIV1-protease per mL. At the end of the 4.sup.th generation the reference process yielded 19 μg insoluble HIV1-protease per mL.

(67) FIG. 14 displays SDS page analysis of the growth decoupled process fermentation. Following linear equation was used for quantification of HIV1-protease:
y=0.0009x R.sup.2=0.9611

(68) According to FIG. 14, the growth decoupled system was capable of producing HIV1-protease in soluble (S) and insoluble form (IB). At the beginning of the protein production phase BL21(DE3)::TN7(Gp2ΔAra)pET30(HIV1-protease) expressed 79% soluble and 21% insoluble HIV1-protease. The ratio of soluble protein decreased with prolonged process duration. At the end of the 4.sup.th generation (27 h after feed start) 12% were expressed as soluble and 88% as insoluble HIV1-protease.

(69) A summary of results of these experiments is shown in FIG. 15. As displayed in graph A, the growth decoupled system was capable of producing HIV1-protease with a concentration of 47 mg/g CDM, whereas the reference process produced a total HIV1-protease concentration of 12 mg/g CDM. Thus, BL21(DE3)::TN7(Gp2ΔAra)pET30(HIV1-protease) produced almost four times more HIV1-protease per gram CDM compared to BL21(DE3)pET30a(HIV1-protease). According to graph C, the model process yielded a concentration of 9 mg/g soluble HIV1-protease at the end of the process while the reference strain was not able to produce HIV1-protease in soluble form. As seen in graph F, the model process produced a total mass of 11 g HIV1-protease, whereas the reference process only reached a total output of 4 g. In conclusion the model process produced about three times more HIV1-protease compared to the reference process. Graph D of FIG. 15 shows the calculated net CDM without produced recombinant product (X-P) in gram. X-P remained constant compared to the total CDM production (graph B). After induction of the growth decoupled system 26 g of CDM where built until the end of the process, whereas the reference system produced 122 g CDM during the production phase. In summary, the growth decoupled system produced 282% more HIV1-protease with about 30% less CDM compared to the reference system.

Example 7: High-Cell-Density Cultivations (HCDC) Using the Growth Decoupled System

(70) The results from the non-HCD bioreactor cultivations indicated that the growth decoupled expression system should allow variable growth rates before induction which is important for HCD cultivations as a growth rate of μ=0.2 h.sup.−1 is hard to maintain without the risk of oxygen limitation. A too high growth rate would result in suboptimal conditions, especially during the exponential feed phase. The HCD process was planned to reach a CDM concentration of 60 g/L at induction time point. As the performed HCDC should only show the ability of the growth decoupled system to reach comparable specific amounts of protein and higher productivity in a semi-HCDC compared to non-HCDC, GFPmut3.1 was used as only model protein. The HCD fermentation plan is shown in FIG. 16. The batch was performed at a temperature of 37° C. and completed after 16 h. An exponential feeding phase (μ=0.17 h.sup.−1) was started immediately after the batch phase finished, which lasted for about 2 generations. The feed medium was also supplemented with ammonium sulphate to guarantee non-nitrogen-limiting conditions. Afterwards the first linear feed profile was applied which lasted for 1 generation. 15 h after start of the feeding phase the protein production was induced with 0.1 M arabinose+20 μmol IPTG/g CDM. During the production phase a second linear feed profile was applied, which lasted for about 1 generation with a decreased calculated growth rate starting from μ=0.050 h.sup.−1 at the beginning to a μ of 0.020 h.sup.−1 at the end of the process. The protein expression phase lasted for 33 h. The purpose of the resulting low growth rate was to supply just enough glucose to the strain as it needed to express the POI.

(71) FIG. 17 shows SDS page analysis of the HCD cultivation of the growth decoupled system. After 33 h of protein production BL21(DE3)::TN7(Gp2ΔAra)pET30(GFPmut3.1) was capable of producing 63% soluble and 37% insoluble GFP, which is an improvement compared to the non-HCD cultivation of the growth decoupled system process and shows that upscaling to HCDC has no significant impact on the solubility of the expressed protein.

(72) FIG. 18 shows the results of the HCDC of the growth decoupled system. BL21(DE3)::TN7(Gp2ΔAra)pET30(GFPmut3.1) was capable of producing 268.4 mg soluble and 157.6 mg insoluble GFP per g CDM. During the production phase a total amount of 279.5 g GFP has been produced and 176.1 g thereof in soluble form. The process yielded a concentration of 30 g/L GFP which is an increase of 300% compared to the non-HCDC of the growth decoupled system. After induction of protein expression the net CDM stopped and remained more or less constant, which agrees with the results from the non-HCDC of the growth decoupled system.

(73) Analysis of total O.sub.2 consumption and total CO.sub.2 formation during protein production of the HCD model fermentation process showed that the growth decoupled system forms a comparable high amount of total CO.sub.2 as the non-HCDC of the growth decoupled system. As seen in graph A of FIG. 19, after induction of protein production O.sub.2 consumption and CO.sub.2 formation of BL21(DE3)::TN7(Gp2ΔAra)pET30(GFPmut3.1) remained constant, which shows that the HCD process is still metabolically active.

(74) FIG. 20 shows a summary of results between the HCD and the non-HCD cultivation of the growth decoupled system. Graph A shows that non-HCDC of BL21(DE3)::TN7(Gp2ΔAra)pET30(GFPmut3.1) was capable of producing 480.18 mg GFP per g CDM, which is the highest specific concentration of all performed cultivations. The HCD process reached a comparable high specific concentration with 426 mg GFP per g CDM. Furthermore the HCDC process was capable of producing 7% more specific GFP in soluble form. As seen in graph D, during both cultivations the growth of net CDM stopped and decreased after induction of the protein expression. At the time of induction the HCDC process reached a CDM concentration of 57 g/L and generated 216% more gross CDM compared to the non-HCDC process, which reached 22 g/L at the time of induction (graph B). HCDC of the growth decoupled system produced a total amount of 280 g GFP which is an increase of 243% compared to non-HCDC of the system (graph F). In consideration of the total produced CDM, shown in graph B and the produced net CDM, shown in graph F, the HCDC produced 243% more GFP with 300% more net CDM compared to the non-HCDC. Graph D displays the calculated net CDM without produced recombinant protein (X-P) in g. The decrease of net CDM proved that after induction of both system almost exclusively recombinant GFPmut3.1 is produced.

(75) Comparison of the total produced GFP and the produced net CDM between the non-HCD and the HCD process shows that the growth decoupled system shows a linear relationship between the produced GFP and the net CDM even in the up-scaled HCD process. It also reveals that HCD cultivation of the growth decoupled system has high potential for further HCDC fermentation with much higher CDM concentrations. Cultivation with a CDM concentration up to 100 g/L prior induction should yield an enormous amount of GFP.