Field of the invention

The present invention relates to a method for producing a biomolecule by a cell cultured in vitro.

Background of the invention

A large number of biopharmaceuticals such as antibodies, cytokines and hor mones is produced using biotechnological methods. To do so, a given biopharma- ceutical is expressed in suitable cells that are cultured in large scale cell cultures and the biopharmaceutical is then isolated from the cell culture. Suitable cells can be lower organisms such as bacteria and yeast or cells of cell lines derived from higher order organisms such as mammals, depending on the type of biopharma ceutical to be produced.

To increase the efficiency of the production of biopharmaceuticals and other bio molecules while ensuring maximal product quality, cells, cell lines and production processes are continuously optimized. The optimization of cell lines includes, for example, optimizing cell morphology and metabolism, improving growth, increas ing viability, reducing apoptosis, senescence and autophagy, increasing productiv ity as well as introducing specific changes to post-translational protein modifica tions. To optimize cell lines, various cell line engineering approaches that are generally time- and cost-intensive such as gene knockout of unfavourable genes or gene clusters are applied. Another commonly applied technique is the trans formation of the cells with a vector. In this way, the genome of the cells does not need to be modified. The optimization of such vectors is an area of intense re search.

The optimization of production processes includes, for example, increasing gas sing rate and stirring speed as well as upscaling cell culture volume. To optimize production processes, to date, continuous processes such as perfusion processes are increasingly used. Continuous processes allow higher cell densities compared to non-continuous processes and can thus be performed in smaller cell culture volumes as well as in shorter periods of time. They also provide further ad vantages such as higher product quality, higher product yield and/or cost savings. However, continuous processes also have certain limitations, in particular oxygen deficiency at high cell densities. Oxygen deficiency may negatively affect cellular parameters, cell productivity and/or product quality.

Therefore, new methods and tools for producing biomolecules are needed that overcome the current limitations, in particular that increase the efficiency of bio molecule production in high cell density perfusion processes.

Summary of the invention

The present invention relates to a method for producing a biomolecule by a cell cultured in vitro, the method comprising the steps of:

(a) propagating the cell in a cell culture, wherein the cell is transformed with a vector, wherein the vector comprises

(i) at least one hypoxia-responsive element (HRE) consisting of a nucle ic acid sequence selected from the group consisting of SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3, SEQ ID NO.: 4 and functional variants of any one of these sequences;

(ii) a promoter that is active in the cell, wherein the promoter is operably linked to the HRE;

(iii) a nucleic acid sequence encoding the biomolecule, wherein the nu cleic acid sequence is operably linked to the promoter; and

(b) isolating the biomolecule from the cell culture, wherein the cell is a eukaryotic cell, wherein the cell is not a tumor cell.

Brief description of the figures

Fig. 1 shows the sequences of the hypoxia-responsive element (HRE) of the VEGF-A gene of various species. The conserved binding site of the hypoxia- inducible factor (HIF) is underlined. Sequence alterations between some of the species occur at a few specific positions in the HRE sequence as marked by verti cal boxes. Fig. 2 shows the molecular biological identification of crucial factors for the detec tion of hypoxia by Chinese hamster ovary (CHO) cells. (A) Alignment of the amino acid sequences of hypoxia-inducible factor-1 a (HIF-1a) protein isoforms from hu man (hsa) and hamster (crigri) with highlighted functional domains. (B) PCR on cDNA detecting HIF-1a (Lane 1+5) and HIF-1 b (Lane 2+6) in CHO-K1 (Lane 1-4) and CHO-DG44 (Lane 5-8). Lane 3-4 and 7-8 represent negative controls lacking reverse transcriptase (-RT). (C) PCR on cDNA detecting Von Hippel-Lindau pro tein (VHL) (Lane 1+3) in CHO-K1 (Lane 1-2) and CHO-DG44 (Lane 3-4). Lane 2+4 represent the negative -RT control. (D) Western blot on HIF-1 a (top panel), HIF-1 b (second top panel) and VHL (bottom panel) in CHO-K1 (Lane 1-3) and CHO-DG44 (Lane 4-6) using 30 pg of protein per lane.

Fig. 3 shows a schematic view of fragments of the vector constructs generated by the inventors. (A) Fragment of the vector constructs used for validation studies in CHO cells. 2-9 HREs denotes 2 to 9 copies of the HRE sequence of the human VEGF-A gene or the human EPO gene, mCMV denotes the minimal CMV pro moter and d2GFP denotes the nucleic acid sequence encoding the biomolecule. In this example, the biomolecule is d2GFP, which is a destabilized variant of the green fluorescent protein (GFP). (B) Fragment of the vector construct used for recombinant protein expression studies in CHO cells. 5HRE denotes five copies of the HRE sequence of the human VEGF-A gene, CMV denotes the complete CMV promoter and SEAP denotes the nucleic acid sequence encoding the biomolecule. In this example, the biomolecule is SEAP (secreted embryonic alkaline phospha tase).

Fig. 4 shows the identification of the most promising vector constructs under unde fined hypoxic conditions. (A+B) Flow cytometry analysis of static or shaken culti vated CHO-K1 pools, stably expressing destabilized GFP (d2GFP) under the con trol HREs from VEGF-A (A) or EPO (B) origin. Grey bars stand for shaken, while black bars represent static cultured pools. (C+D) Flow cytometry analysis of static or shaken cultivated CHO-DG44 pools, stably expressing destabilized GFP (d2GFP) under the control of HREs from VEGF (C) or EPO (D) origin. Grey bars stand for shaken, while black bars represent static cultured pools [n = 3 replicates; Mean ± standard deviation (SD)].

Fig. 5 shows the characterization of the response of CHO-DG44 and CHO-K1 cells to defined 02-concentrations. (A) Progression of the controlled oxygen con centration during fermentation in a normoxic batch (solid line) and a stepwise hy poxic cultivation (dotted line). (B) Mean fluorescence of CHO-DG44-Mock (dotted lines) and CHO-DG44-5HRE-VEGF (solid lines) cells cultured in a batch fermenta tion over 156 h under normoxic and stepwise hypoxic conditions. (C) Mean fluo rescence of normoxic and hypoxic cultured CFIO-DG44-Mock (non-striped bars) and CFIO-DG44-5FIRE-VEGF (striped bars) cells after 140 h. (D) Mean fluores cence of CFIO-K1-Mock (dotted lines) and CFIO-K1-5FIRE-VEGF (solid lines) cells cultured in a batch fermentation over 156 h under normoxic and stepwise hypoxic conditions. (E) Mean fluorescence of normoxic and hypoxic cultured CFIO-K1- Mock (non-striped bars) and CFIO-K1-5FIRE-VEGF (striped bars) cells after 140 h. Statistical analysis was conducted using one-way ANOVA with Bonferroni correc tion or students t-test (F) [n = 3 replicates; Mean ± SD]

Fig. 6 shows inducible secreted embryonic alkaline phosphatase (SEAP) expres sion exploiting hypoxia during oxygen deprivation conditions. (A) Specific produc tivity of CHO-DG44-SEAP and CHO-DG44-5HRE-SEAP cells cultured under hy poxic conditions (static cultivated cells) in comparison to normally cultured cells (shaken cultivated cells). (B) Specific productivity of CFIO-DG44-SEAP and CFIO- DG44-5FIRE-SEAP cells cultured in ultra-high cell density compared to normally cultured cells (normal cell density). (C+D) Viable cell density (C) and viability (D) of CFIO-DG44-5FIRE-SEAP cells during fed-batch fermentation. Temp shift = tem perature shift (no oxygen shift) (grey line); Temp+02 shift = temperature and oxy gen shift (black line). (E) SEAP concentration during fed-batch fermentation se creted by CFIO-DG44-5FIRE-SEAP cells with and without hypoxic shift to 1 % O2 (oxygen shift). Temp shift = temperature shift (no oxygen shift) (grey line); Temp+02 shift = temperature and oxygen shift (black line). (F) Specific productivi ty of CFIO-DG44-5FIRE-SEAP cells in average before the temperature and oxygen shift and afterwards. Temp shift = temperature shift (no oxygen shift) (grey bars); Temp+02 shift = temperature and oxygen shift (black bars). Statistical analysis was conducted using one-way ANOVA with Bonferroni correction [n = 3 replicates; Mean ± SD; * = p < 0.05; ** = p < 0.01 ; n.s. = not significant].

Detailed description of the invention

In a first aspect, the invention relates to a method for producing a biomolecule by a cell cultured in vitro, the method comprising the steps of:

(a) propagating the cell in a cell culture, wherein the cell is transformed with a vector, wherein the vector comprises

(i) at least one hypoxia-responsive element (HRE) consisting of a nucle ic acid sequence selected from the group consisting of SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3, SEQ ID NO.: 4 and functional variants of any one of these sequences;

(ii) a promoter that is active in the cell, wherein the promoter is operably linked to the HRE;

(iii) a nucleic acid sequence encoding the biomolecule, wherein the nu cleic acid sequence is operably linked to the promoter; and

(b) isolating the biomolecule from the cell culture, wherein the cell is a eukaryotic cell, wherein the cell is not a tumor cell.

The term “biomolecule” as used herein refers to any compound which is inherently biological in nature and suitable to be produced by a cell and harvested from the cell or from the cell’s culture medium. Preferably, the biomolecule is a biopharma- ceutical, i.e. a pharmaceutical which is inherently biological in nature and manu factured by means of a biotechnological process. The biopharmaceutical may be a therapeutic, prophylactic or diagnostic compound. The biopharmaceutical may be, for example, an antibody, a cytokine, a hormone, an enzyme or a vaccine. The biopharmaceutical may also be a virus or a coding or non-coding RNA molecule.

The cell is a cell that is suitable for producing the biomolecule of interest when the cell is cultured in vitro. The cell can also be referred to as “producer cell”. For bio molecule production, the cell is provided with the nucleic acid sequence encoding the biomolecule and propagated in a cell culture, i.e. grown and proliferated under controlled conditions. The cell culture may be a suspension cell culture or an ad herent cell culture. Preferably, the cell culture is a suspension cell culture. Besides cells that naturally grow in suspension, there are also cell lines that have been modified to be able to survive in suspension cell cultures so they can be grown to higher cell densities than their adherent culture would allow.

The cell is a eukaryotic cell.

The cell is not a tumor cell. It is known in the art that tumor cells differ from non tumor cells in several aspects. For example, tumor cells show an abnormal physi ology.

According to the invention, the nucleic acid sequence encoding the biomolecule is part of the vector with which the cell is transformed.

The term “vector” as used herein refers to a recombinant DNA molecule that is used to introduce genetic material into the cell. The vector may be, for example, an expression vector or a viral vector. The cell is transformed with the vector so that the vector is present in the cell. Methods for transforming the cell with the vec tor are known in the art and include, for example, transfection of the cell. Trans fection is an easy and fast method to transform the cell with the vector.

The vector provides a hypoxia-inducible expression system. By transforming the cell with the vector, the cell is rendered hypoxia-inducible, i.e. it can be referred to as a hypoxia-inducible cell.

The vector comprises at least one hypoxia-responsive element (HRE) consisting of a nucleic acid sequence selected from the group consisting of SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3, SEQ ID NO.: 4 and functional variants of any one of these sequences.

Hypoxia-responsive elements (HREs), also referred to as hypoxia response ele ments, are regulatory enhancer elements that have been identified in the promoter region of numerous oxygen-responsive genes such as the VEGF-A, EPO, PGK1, Ldha, ALDA and GAPDH genes. The HREs of the different genes differ in their nucleotide sequences. The HREs serve as the binding site for hypoxia-inducible factors (HIFs), which are transcription factors that mediate the adaptation of cells to low oxygen levels (hypoxia) by enhancing the transcription of the respective genes. Transcription factors generally enhance the transcription of a gene by bind ing to a specific DNA sequence such as a responsive element.

HIFs are heterodimeric proteins comprising a hypoxia-inducible a subunit with three isoforms (HIF-1a, H IF-2a and H IF-3a) and a constitutively expressed b sub unit with two isoforms (HIF-1 b and H IR-2b). H IF-1 b is also known as aryl hydro carbon receptor nuclear translocator (ARNT). In short, under normal oxygen con ditions (normoxia), i.e. under the presence of sufficient oxygen, the a subunit be comes hydroxylated and ubiquitylated, so that it is degraded by the proteasome. In case of hypoxia, i.e. in case of oxygen deficiency, the a subunit is not hydrox ylated and ubiquitylated and is thus stabilized and binds to the b subunit to form a heterodimer. The heterodimer binds to FIREs in the promoter/enhancer region of target genes, enhancing the transcription of the genes.

The Von Hippel-Lindau protein (VHL protein), also known as Von Hippel-Lindau tumor suppressor, ubiquitylates the a subunit of HIFs and thus marks the a subu nit for degradation. A reduction or loss of VHL protein activity results in an in creased amount of HIF a subunit and thus in an increased transcription level of HRE-regulated genes.

The inventors found that HIF-1 a, HIF-1 b and the VHL protein, which are the key players of the hypoxia sensing pathway, are also present in Chinese hamster ova ry (CHO) cells which are the most commonly used mammalian cells for the pro duction of biomolecules. The inventors also found that crucial protein domains of HIF-1 a such as the DNA binding domain are conserved in the human and hamster HIF-1 a proteins.

The inventors then found that the HRE sequence of the vascular endothelial growth factor A (VEGF-A) gene can be exploited for increasing the production of a biomolecule under hypoxic conditions. To do so, at least one HRE of the VEGF-A gene is operably combined with a suitable promoter and the nucleic acid se quence encoding the biomolecule. This leads to an increased transcription of the nucleic acid sequence encoding the biomolecule under conditions of oxygen defi ciency (i.e. under hypoxic conditions) and thus to an increased production of the biomolecule under hypoxic conditions. The increase of biomolecule production under hypoxic conditions is particularly advantageous for cell cultures with high cell densities, in which low levels of oxygen (hypoxic conditions) are inherently present. HREs of the VEGF-A gene have been identified in various organisms such as in human, mouse, rat, hamster, dog and cattle.

The sequence of SEQ ID NO.: 1 is the HRE sequence of the human VEGF-A gene:

CCACAGTGCATACGTGGGCTCCAACAGGTCCTCTT (SEQ ID NO.: 1).

The VEGF-A gene of Canis lupus familiaris (dog) and the VEGF-A gene of Bos taurus (cattle) are regulated by the same FIRE sequence as the human VEGF-A gene.

The sequence of SEQ ID NO.: 2 is the FIRE sequence of the Mus musculus (house mouse) VEGF-A gene:

ACACAGTGCATACGTGGGTTTCCACAGGTCGTCTC (SEQ ID NO.: 2).

The sequence of SEQ ID NO.: 3 is the FIRE sequence of the Rattus norvegicus (Norway rat) VEGF-A gene:

CCACAGTGCATACGTGGGCTTCCACAGGTCGTCTC (SEQ ID NO.: 3).

The sequence of SEQ ID NO.: 4 is the FIRE sequence of the Cricetulus griseus (Chinese hamster) VEGF-A gene:

CCACAGTGCATACGTGGGCTTCCACAGGTCCTCTT (SEQ ID NO.: 4).

All sequences are indicated in 5’ -> 3’ direction.

A functional variant of any one of the sequences of SEQ ID NO.s: 1-4 is a variant that has one or more alterations in the sequence, wherein the alterations do not alter the ability of the sequence to induce gene expression in response to hypoxia. The alterations are preferably nucleotide exchanges, but may also be nucleotide insertions or nucleotide deletions.

By comparing the FIRE of the erythropoietin (EPO) and the VEGF-A gene in CFIO cells, the inventors found that the FIRE of the VEGF-A gene is superior and pre sents the most potent enhancer for biomolecule production in CFIO cells when the cells are induced by hypoxic conditions. The inventors also found that the use of the HRE sequence of the human phosphoglycerate kinase 1 (PGK1) gene does not result in a hypoxia-inducible expression system in CHO cells.

The vector further comprises a promoter that is active in the cell, wherein the pro moter is operably linked to the HRE. The promoter drives the transcription of the nucleic acid sequence encoding the biomolecule. The promoter can be any pro moter that is active in the cell and is thus primarily selected according to the type of the cell that is used for the production of the biomolecule. When selecting the promoter, further known considerations can be taken into account, such as the type of biomolecule that is to be produced and the level of induction of transcrip tion of the nucleic acid sequence encoding the biomolecule. The promoter can be an endogenous promoter or an exogenous promoter. The promoter can be a natu ral promoter, such as a natural viral promoter, or an engineered (synthetic) pro moter.

Two nucleic acid sequences are “operably linked” when one of the nucleic acid sequences is placed into a functional relationship with the other nucleic acid se quence. The term “operably linked” is thus synonymous to the term “functionally linked”. In most cases, two nucleic acid sequences that are operably linked are contiguous. However, enhancers, for example, do not have to be contiguous to another nucleic acid sequence in order to be operably linked thereto.

The promoter is operably linked to the HRE, i.e. the promoter is placed into a functional relationship with the HRE. In other words, the promoter and the HRE are arranged in the vector in a manner that allows a functional relationship of the promoter with the HRE, so that the promoter is induced when a HIF heterodimer binds to the HRE.

The vector further comprises a nucleic acid sequence encoding the biomolecule, wherein the nucleic acid sequence is operably linked to the promoter. In other words, the nucleic acid sequence encoding the biomolecule and the promoter are arranged in the vector in a manner that allows a functional relationship of the nu cleic acid sequence with the promoter, so that the promoter can drive the tran scription of the nucleic acid sequence. For example, as is known in the art, the promoter needs to be arranged in an appropriate orientation relative to the nucleic acid sequence. The transcription of the nucleic acid sequence encoding the bio molecule leads to the production of the biomolecule by the cell.

The nucleic acid sequence encoding the biomolecule can also be referred to as the gene of interest (Gol).

After the step of propagating the cell in the cell culture, which is the step in which the biomolecule is produced, the biomolecule is isolated from the cell culture. Suitable ways of isolating the biomolecule from the cell culture are known and will be selected according to the type of biomolecule produced. For example, if the biomolecule is a recombinant protein that is secreted into the cell culture medium, i.e. into the medium in which the cell is propagated, the protein can be isolated from the cell culture by separating the medium from the cells (e.g. by centrifuga tion) and by subsequently purifying the protein from the medium (e.g. by chroma tography).

Biomolecule production in cell cultures with high cell densities is limited by oxygen deficiency. Oxygen deficiency naturally occurs in the cell culture when the cell density increases to a level at which oxygen supply becomes insufficient. In the prior art, this limitation could only be improved to a very limited extent due to poor oxygen distribution as well as limited oxygen saturation in the cell cultures. The method of the invention solves this problem by transforming the cells with a vector that leads to an increased production of the biomolecule (i.e. the product) under conditions of oxygen deficiency. The cells show improved productivity, despite higher cell densities and lower levels of available oxygen, which allows using more intensive cell culture processes such as perfusion cell cultures for biomolecule production. The method of the invention thus exploits hypoxia to significantly in crease biomolecule production.

The inventors confirmed that the hypoxia-inducible cells that were generated for the method of the invention did not express the unstable HIF-1a protein under normoxic conditions (40-60 % O2). Low levels of the HIF-1a protein were found during cultivation of the cells at 5 % O2 and higher levels (i.e. a stabilization) of the HIF-1a protein were found at 1 % O2. In accordance with these findings, the amount of biomolecule produced was increased during cultivation of the cells at 5 % O2 compared to their cultivation at 40 % O2 or 60 % O2. The amount of bio molecule produced was further massively increased during cultivation of the cells at 1 % O2 compared to their cultivation at 5 % O2. The amount of biomolecule produced by the hypoxia-inducible cells was nearly 10-fold higher than the amount produced by non-hypoxia-inducible control cells under hypoxic conditions. Taken together, these data confirm the successful establishment of a hypoxia-inducible vector system to increase biomolecule production under conditions of oxygen de ficiency.

The method of the invention results in increased cellular productivity and in creased product yield while product quality is maintained or improved. This in creases the efficiency of the production process and reduces the costs for biomol ecule production. The method of the invention also renders the production pro cess more robust, because the previously negative consequences of decreasing oxygen levels are overcome, even turned around, leading to an improved repro ducibility and an increased predictability of the biomolecule production process.

The method of the invention allows cells to be cultured at high cell densities while improving cellular productivity and maintaining and/or improving product quality under conditions of oxygen deficiency. Cell cultures with high cell densities are, for example, perfusion cell cultures, batch cell cultures, fed-batch cell cultures or N-1 fed-batch cell cultures. Cell cultures with high cell densities save production time and lower the costs for biomolecule production since the volume of the cell culture can be reduced. A reduced cell culture volume leads to a reduction of cell culture medium and feeds needed, to space saving, waste reduction and a lower energy demand for tempering the cell culture at a suitable temperature such as 37 °C.

In addition, the method of the invention allows the reduction of the gassing rate of the cell culture (with compressed air or pure oxygen) or even the elimination of gassing with pure oxygen, which saves oxygen and thereby further reduces the costs for biomolecule production. Pure oxygen gassing also has a certain toxicity for commonly used CHO cells. The method of the invention also allows the reduc tion of the stirring speed, which significantly reduces shear stress acting on the cells. The reduction of the shear stress leads to an improvement in cellular pa rameters such as cell growth, cell viability and cellular productivity. The improve ment in these parameters can simplify the isolation and purification of the biomol ecule produced.

Further, the method of the invention can lead to a reduction of a possible aggrega tion of the biomolecule at interphases, thereby improving product quality and lead ing to a reduction in the costs and workload of downstream processing.

The method of the invention also allows to establish a novel biphasic production process. Conventionally, during a fed-batch process, a temperature shift is applied to the cell culture to ensure maximum productivity. Standard industrial fed-batch processes apply a shift of temperature from 37 °C to a lower temperature such as 34 °C after the cells have reached the static cultivation phase. In this way, the lon gevity of the production process can be enhanced and the final product titer can be increased as the temperature reduction leads to a favorable redistribution of nutrients for biomolecule production. In addition to the temperature shift or instead of the temperature shift, the method of the invention allows applying an oxygen shift from a normoxic level (e.g. 40 % or 60 % oxygen) to a hypoxic (e.g. 5 % or 1 % oxygen), by which an induction of biomolecule production can be achieved at any cell density, i.e. at any time point of the production process. This requires that the oxygen level in the cell culture can be tightly controlled, which is usually the case when biomolecules are produced. The oxygen shift preferably is applied after the cells have reached the static cultivation phase. The oxygen shift can be ap plied at the same time as the temperature shift.

The inventors found that there is no significant influence of oxygen levels as low as 1 % O2 on cell viability when the oxygen shift to 1 % is performed in parallel with (i.e. at the same time as) the temperature shift to 34 °C after the cells have reached the static cultivation phase.

The inventors further found that after shifting the temperature or after shifting both the temperature and the oxygen level, biomolecule production was induced in both hypoxia-inducible and conventional non-hypoxia-inducible cells, however, to a sig nificant stronger extent in hypoxia-inducible cells which benefit from the reduction of the oxygen level. A nearly 2-fold increase of specific productivity compared to conventional non-hypoxia-inducible cells could be achieved by exploiting hypoxia in this manner.

In a preferred embodiment, the promoter is upstream of the nucleic acid sequence encoding the biomolecule. The distance between the promoter and the nucleic acid sequence encoding the biomolecule can be, for example, 0 to about 100 base pairs. In case the distance is 0 base pairs, the promoter is directly upstream of the nucleic acid sequence encoding the biomolecule.

In a further preferred embodiment, the promoter is directly upstream of the nucleic acid sequence encoding the biomolecule. In other words, the 3’ end of the pro moter sequence is directly adjacent to the 5’ end of the nucleic acid sequence en coding the biomolecule. Alternatively, the promoter may be located further up stream of the nucleic acid sequence encoding the biomolecule.

In a preferred embodiment, the HRE is upstream of the promoter. The distance between the HRE and the promoter can be, for example, 0 to about 100 base pairs. In case the distance is 0 base pairs, the HRE is directly upstream of the promoter.

In a further preferred embodiment, the HRE is directly upstream of the promoter. In other words, the 3’ end of the HRE sequence is directly adjacent to the 5’ end of the promoter sequence. Alternatively, the HRE may be located further upstream of the promoter or downstream of the nucleic acid sequence encoding the biomol ecule.

In a particularly preferred embodiment, the promoter is directly upstream of the nucleic acid sequence encoding the biomolecule and the HRE is directly upstream of the promoter.

In a preferred embodiment, the HRE is present in more than one copy. According ly, in a preferred embodiment, the vector comprises at least two HREs arranged adjacent to one another, wherein the number of HREs preferably is between 2 and 20, further preferred between 2 and 10, further preferred between 2 and 8, most preferred 5. When using more than one copy of the HRE sequence, the in- ventors observed a positive correlation between HRE copy number and response in terms of biomolecule production. Thus, it is preferred to use a vector that com prises at least two HREs. More than 20 copies of the HRE sequence would lead to an unnecessarily large vector. Therefore, it is preferred that the vector comprises between 2 and 20 HREs. In CHO cells, the correlation between HRE copy number and biomolecule production was found to be saturated between 5 and 8 HRE rep etitions. Overall, the inventors found that 5 copies of the HRE gave the best re sults in terms of induction of biomolecule production under conditions of oxygen deficiency.

The term ’’arranged adjacent to one another” refers to a distance of 0 to about 100 base pairs between two adjacent HREs. The distance (spacing) between two ad jacent HREs preferably is between 0 and 50 base pairs, further preferred between 0 and 30 base pairs, further preferred between 0 and 20 base pairs, further pre ferred between 0 and 10 base pairs, most preferred 6 base pairs. In case the dis tance is 0 base pairs, the two HREs are arranged directly adjacent to one another. The HRE copies do not need to be directly adjacent to one another. For example, the inventors used vectors with different HRE copy numbers with a distance of six base pairs between adjacent HREs. A distance of more than 50 base pairs would lead to an unnecessarily large vector. Therefore, it is preferred that the distance between two adjacent HREs is between 0 and 50 base pairs.

The cell may be any cell that is suitable for the biotechnological production of the biomolecule of interest. The cell is a eukaryotic cell.

The cell may be a human cell or a non-human cell. Many human and non-human cell lines that are suitable for the biotechnological production of biomolecules are known. Among the non-human cell-lines, CHO cells and duck embryo cells are of particular importance. CHO cells are the most commonly used mammalian cells for the industrial production of recombinant therapeutic proteins such as antibod ies. Duck embryo cells are particularly useful for the production of virus-based vaccines.

In a preferred embodiment, the cell is a Chinese hamster ovary (CHO) cell, a de rivative of a CHO cell, a SL-29 cell, a duck embryo cell, a Vero cell, a NIH/3T3 cell, a Baby Hamster Kidney 21 (BHK-21) cell, a Sp2/0-Ag14 cell, a BTI-Tn-5B1-4 cell, a hybridoma cell, a Human Embryonic Kidney 293 (HEK293) cell, a derivative of a HEK293 cell, a CAP cell or a CAP-T cell.

In a preferred embodiment, the derivative of the CHO cell is a CHO-K1 cell, a CHO-EBNA cell, a CHO-EBNA GS cell, a CHO-DG44 cell, a CHO-DXB11 cell or a CHO-S cell. CHO-K1 and CHO-DG44 cells are the prevailing derivatives used in industry. CHO-EBNA cells are CHO-K1 cells that are engineered to express the gene encoding Epstein-Barr virus nuclear antigen-1 (EBNA-1). CHO-EBNA GS cells are CHO-K1 cells that are engineered to express the gene encoding EBNA-1 and to express the gene encoding glutamine synthetase (GS).

In a preferred embodiment, the derivative of the HEK293 cell is a HEK293F cell or a HEK293S cell. In contrast to HEK293 cells, HEK293F and HEK293S cells are adapted to grow in suspension. HEK293F cells are known for their fast growth and high transfection efficiency. Further derivatives of the HEK293 cell are, for exam ple, a HEK293T cell (ATCC number: CRL-3216), a HEK293FT cell (a fast growing variant of HEK293T), a HEK293E cell and a HEK293SG cell.

A hybridoma cell is an immortalized cell derived from the fusion of a B lymphoblast with a myeloma fusion partner. The generation of a hybridoma cell starts by im munizing a mouse with a target antigen, thereby eliciting an immune response. The B lymphocytes that are taken from the immunized mouse spleen produce an tibodies to the antigen. Each B lymphocyte is then fused with an immortal myelo ma cell line to produce a hybrid cell line called a hybridoma, The hybridoma cell has both the antibody-producing ability of the B lymphocyte and the longevity and reproductivity of the myeloma, allowing the production of large quantities of mono clonal antibodies. Sp2/0-Ag14 is an example of a hybridoma fusion partner mye loma cell line.

An overview of the preferred cell lines is given in Table 1. Table 1 : Preferred cell lines for the production of biomolecules

The cell preferably is a mammalian cell. In another embodiment, the cell prefera bly is an insect cell.

The promoter preferably is a complete promoter. The complete promoter can also be referred to as full-length promoter. The method of the invention is suitable for using a complete promoter. Complete promoters are generally used when produc ing a biomolecule by a cell cultured in vitro.

In a preferred embodiment, the promoter is selected from the group consisting of cytomegalovirus (CMV) promoter, simian virus 40 early (SV40E) promoter, elon gation factor 1 alpha (EF-1a) promoter, hEF1-HTLV promoter, T7 RNA polymer ase and SP6 RNA polymerase (T7/SP6) promoter, ubiquitin C (UBC) promoter, U6 promoter, ferritin heavy chain (FerFI) promoter, ferritin light chain (FerL) pro moter, phosphoglycerate kinase (PGK) promoter, thyroxine-binding globulin (TBG) promoter, albumin (Alb) promoter, alpha-fetoprotein (AFP) promoter, Autographa californica nuclear polyhedrosis baculovirus (AcNPV) polyhedrin promoter, chick en beta-actin promoter coupled with CMV early enhancer (CAGG) promoter, copia transposon (COPIA) promoter, actin 5C (ACT5C) promoter, simian virus 40 (SV40) promoter, human beta-actin promoter, tetracycline response element (TRE) promoter, Drosophila Gal4 binding sites-containing upstream activation se quence (UAS) promoter, calcium/calmodulin-dependent protein kinase II (CaMKII) promoter, yeast GAL1 promoter, yeast GAL10 promoter, yeast GAL1/GAL10 pro moter, yeast transcription elongation factor 1 alpha (TEF1) promoter, yeast glyceraldehyde 3-phosphate dehydrogenase (GAPDH) promoter, yeast alcohol dehydrogenase I (ADH1) promoter, Cauliflower Mosaic Virus 35S (CaMV 35S) promoter, maize ubiquitin (Ubi) promoter and H1 promoter.

In a further preferred embodiment, the promoter is selected from the group con sisting of cytomegalovirus (CMV) promoter, simian virus 40 early (SV40E) pro moter, elongation factor 1 alpha (EF-1a) promoter, hEF1-HTLV promoter, T7 RNA polymerase and SP6 RNA polymerase (T7/SP6) promoter, ubiquitin C (UBC) promoter, U6 promoter, ferritin heavy chain (FerH) promoter, ferritin light chain (FerL) promoter, phosphoglycerate kinase (PGK) promoter, thyroxine-binding globulin (TBG) promoter, albumin (Alb) promoter and alpha-fetoprotein (AFP) promoter.

The EF-1a promoter, which is also referred to as EF1 promoter, preferably is the human EF-1a (hEF1) promoter.

The hEF1-FITLV promoter is a composite promoter composed of the human elon gation factor 1 alpha (EF-1a) core promoter and the R segment and part of the U5 sequence (R-U5’) of the Fluman T-Cell Leukemia Virus (HTLV) Type 1 Long Ter minal Repeat.

The T7/SP6 promoter is a combination of two promoters, namely the T7 RNA pol ymerase promoter and the SP6 RNA polymerase promoter.

The U6 promoter is an RNA polymerase III promoter that allows expression of small non-coding RNA molecules such as shRNAs or miRNAs with precise 5' and 3' sequences (i.e. no 5' cap or 3' polyA tail).

The AcNPV polyhedrin promoter (polyhedrin promoter) is the polyhedrin promoter of the Autographa californica nuclear polyhedrosis virus. The promoter can be used for high level expression in baculovirus expression systems and derivatives thereof.

The CAGG promoter, which is also referred to as CAG promoter, is a chicken be- ta-actin promoter coupled with the CMV early enhancer element and the splice acceptor of the rabbit beta-globin gene.

The COPIA promoter is the copia transposon promoter that is commonly used in insect cell systems.

The ACT5C promoter, which is also referred to as Ac5 promoter, preferably is the Drosophila ACT5C promoter that can be used for mediating expression in insect cell systems.

The yeast GAL1/GAL10 promoter is also referred to as GAL1/GAL10 bidirectional promoter.

The GAPDH promoter is also referred to as TDH3 or GDS promoter.

The Ubi promoter is also referred to as Ubi-1 promoter. The H1 promoter is an RNA polymerase III promoter that allows expression of small non-coding RNA molecules such as shRNAs or miRNAs. The H1 promoter preferably is the human H1 promoter.

The promoter may also be a minimal promoter, i.e. a promoter that consists of the minimal sequence that can act as an effective promoter. The minimal promoter can be, for example, a minimal CMV promoter.

In a preferred embodiment, the promoter is a strong promoter. Promoters can be classified as strong or weak promoters. The strength of a promoter is the rate of transcription of the nucleic acid sequence controlled by the promoter. In case of a strong promoter, the rate of transcription is high. In case of a weak promoter, the rate of transcription is relatively low. Accordingly, the use of a strong promoter re sults in a higher rate of biomolecule production compared to the use of a weak promoter. As mentioned above, the inventors found that the HRE sequence of the VEGF-A gene can be exploited for increasing the production of a biomolecule un der hypoxic conditions. Surprisingly, the inventors found that this also applies when promoters that already have a strong activity themselves are used. In other words, the high activity of a strong promoter can be even further increased by us ing the HRE sequence of the VEGF-A gene as specified in the method of the in vention.

Examples of strong promoters are the CMV promoter, the SV40E promoter, the EF-1a promoter, the hEF1-HTLV promoter, the T7/SP6 promoter, the Alb promot er, the AFP promoter, the AcNPV polyhedrin promoter and the CAGG promoter.

The promoter preferably is a constitutive promoter, i.e. a promoter that is always active.

Examples of constitutive promoters are the CMV promoter, the SV40E promoter, the EF-1a promoter, the hEF1-HTLV promoter, the T7/SP6 promoter, the UBC promoter, the U6 promoter, the FerH promoter, the FerL promoter, the PGK pro moter, the TBG promoter, the Alb promoter, the AFP promoter, the AcNPV poly hedrin promoter, the CAGG promoter, the COPIA promoter and the ACT5C pro moter. The promoter may alternatively be an inducible promoter, i.e. a promoter that is active only under certain circumstances.

It is particularly preferred that the promoter is the CMV promoter. The term “CMV promoter” as used herein refers to the complete CMV promoter, unless the speci fication specifically refers to the minimal CMV promoter. The CMV promoter is a strong and constitutive promoter.

It is further preferred that the promoter is the CMV promoter and the cell is a CHO cell or a derivative of a CHO cell.

The vector may further comprise one or more regulatory elements other than HRE, in particular one or more enhancers, that are operably linked to the promoter and regulate the transcription of the nucleic acid sequence encoding the biomole cule. Examples of suitable enhancers include hepatitis B virus enhancers (HBVE), tyrosinase (Tyr) enhancer and aldehyde dehydrogenase (AldA) enhancer.

The vector may further comprise a nucleic acid sequence encoding a selection marker that allows selecting the cells that have been stably transformed with the vector. Suitable selection markers depend on the cell used and are known in the art. For example, if the cell is a dihydrofolate reductase (DHFR) deficient cell (e.g. a CHO-DG44 cell), DHFR is a suitable selection marker. In this case, the vector preferably further comprises a nucleic acid sequence encoding DHFR.

In a preferred embodiment, the cell is further transformed with a nucleic acid se quence comprising a region encoding an RNA interference (RNAi) molecule that prevents or reduces expression of Von Hippel-Lindau protein, wherein the RNAi molecule is a small interfering RNA (siRNA), a short hairpin RNA (shRNA) or a microRNA (miRNA), and wherein the region encoding the RNAi molecule is oper ably linked to a promoter that is active in the cell.

As mentioned above, the Von Hippel-Lindau protein (VHL protein) ubiquitylates the a subunit of HIFs and thus marks the a subunit for degradation. Preventing or reducing the expression of the VHL protein results in an increased amount of HIF a subunit and thus in an increased expression level of the HRE-regulated nucleic acid sequence encoding the biomolecule. For this reason, further transforming the cell with the nucleic acid sequence that comprises the region encoding the RNAi molecule that prevents or reduces expression of the VHL protein is of advantage for further increasing the production of the biomolecule.

The following sequences are examples of sequences of siRNA that prevents or reduces the expression of the VHL protein:

UGGUCAAGCCUGAGAACUAUU (SEQ ID NO.: 56)

CCAAAUGUGCGGAAGGACAUU (SEQ ID NO.: 57)

CCGGAAGGCAGCUGGUCAAUU (SEQ ID NO.: 58)

All sequences are indicated in 5’ -> 3’ direction.

RNA interference is a naturally occurring process in which RNA molecules are involved in sequence-specific suppression of gene expression. RNAi molecules suppress (i.e. prevent or reduce) gene expression by mRNA degradation which prevents the translation of the mRNA into the respective protein. The RNAi mole cule is a siRNA, a shRNA or a miRNA, all of which are small non-coding RNA molecules. siRNA molecules are double-stranded RNA molecules of about 19-23 base pair nucleotides in length. shRNA molecules have a hairpin like stem-loop structure since they consist of two complementary RNA sequences of about 19-22 nucleo tides in length that are linked by a short loop of 4-11 nucleotides. Both siRNA and shRNA molecules are highly specific, having only one mRNA target. miRNA molecules are single stranded RNA molecules of about 20-24 nucleotides in length and have multiple mRNA targets.

To enable transcription of the RNAi molecule, the region encoding the RNAi mole cule is operably linked to a promoter that is active in the cell. The promoter drives the transcription of the RNAi molecule. The promoter preferably is comprised by the nucleic acid sequence that comprises the region encoding the RNAi molecule. The promoter is primarily selected according to the type of the cell that is used for the production of the biomolecule and for its suitability regarding the transcription of the RNAi molecule. All promoters mentioned above with respect to the nucleic acid sequence encoding the biomolecule can be used as promoters for the region encoding the RNAi molecule. The promoter preferably is the U6 promoter. In case the RNAi molecule is a miRNA, the CMV promoter is also preferred.

The nucleic acid sequence that comprises the region encoding the RNAi molecule preferably is part of a vector. The vector can be the same vector as the vector comprising the HRE, the promoter and the nucleic acid sequence encoding the biomolecule. Accordingly, in a preferred embodiment, the nucleic acid sequence comprising the region encoding the RNAi molecule is comprised by the vector. In this case, the vector comprises:

(i) at least one HRE consisting of a nucleic acid sequence selected from the group consisting of SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3, SEQ ID NO.: 4 and functional variants of any one of these sequences;

(ii) a first promoter that is active in the cell, wherein the first promoter is operably linked to the HRE;

(iii) a nucleic acid sequence encoding the biomolecule, wherein the nu cleic acid sequence is operably linked to the first promoter;

(iv) a nucleic acid sequence comprising

(a) a second promoter that is active in the cell; and

(b) a region encoding an RNAi molecule that prevents or reduces expression of Von Hippel-Lindau protein, wherein the RNAi molecule is a siRNA, a shRNA or a miRNA, and wherein the region encoding the RNAi molecule is operably linked to the second promoter.

In an alternative embodiment, the nucleic acid sequence that comprises the region encoding the RNAi molecule can be part of a different vector. This would avoid the need to generate an unnecessarily large vector.

In yet another alternative embodiment, the nucleic acid sequence that comprises the region encoding the RNAi molecule is inserted into the cell’s genome. In this way, a modified cell would be provided. This would also avoid the need to gener ate an unnecessarily large vector. In a preferred embodiment, the promoter to which the region encoding the RNAi molecule is operably linked, is operably linked to at least one hypoxia-responsive element (HRE) consisting of a nucleic acid sequence selected from the group consisting of SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3, SEQ ID NO.: 4 and functional variants of any one of these sequences. In this embodiment, the pro duction of the RNAi molecule is regulated in the same manner as the production of the biomolecule, i.e. its production is increased under hypoxic conditions. Ac cordingly, hypoxic conditions lead to (i) an increase of the transcription of the nu cleic acid sequence encoding the biomolecule, and (ii) an increased amount of HIF a subunit by increasing the transcription of the region encoding the RNAi mol ecule that prevents or reduces the expression of the Von Hippel-Lindau protein. The increased amount of HIF a subunit leads to an increased formation of hetero dimers of the HIF a subunit and the HIF b subunit. The heterodimers bind to the HRE(s), thereby further enhancing the transcription of the nucleic acid sequence encoding the biomolecule as well as the transcription of the region encoding the RNAi molecule. Accordingly, in this way, a positive feedback loop is established that has an enhancing effect on biomolecule production under hypoxic conditions.

The positive feedback loop leads to a higher production yield at a given level of oxygen. This also means that the same production yield can be reached at higher levels of oxygen in the cell culture, for example at 10 % oxygen rather than at 1 % oxygen. This is of advantage for cells that show a reduced viability at extremely low oxygen levels such as 1 % oxygen.

The at least one HRE preferably is comprised by the nucleic acid sequence that comprises the region encoding the RNAi molecule. The nucleic acid sequence comprising the region encoding the RNAi molecule thus preferably comprises (a) at least one HRE consisting of a nucleic acid sequence selected from the group consisting of SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3, SEQ ID NO.: 4 and functional variants of any one of these sequences; (b) a promoter that is active in the cell, wherein the promoter is operably linked to the HRE; (c) the region encod ing the RNAi molecule, wherein the region encoding the RNAi molecule is opera bly linked to the promoter. The embodiments described above with respect to the HRE that controls the tran scription of the nucleic acid sequence encoding the biomolecule likewise apply to the HRE that controls the transcription of the region encoding the RNAi molecule.

In a preferred embodiment, the promoter to which the region encoding the RNAi molecule is operably linked is directly upstream of the region encoding the RNAi molecule and the HRE is directly upstream of the promoter.

In another embodiment, the nucleic acid sequence encoding the biomolecule ad ditionally encodes the RNAi molecule (i.e. the nucleic acid sequence encoding the biomolecule additionally comprises the region that encodes the RNAi molecule) and is transcribed into an mRNA transcript that harbours a respective splicing site so that both the biomolecule and the RNAi molecule will be generated from the mRNA transcript by splicing. The splicing site may be, for example, a pre-tRNA splicing site. In this embodiment, one promoter is sufficient to drive the transcrip tion of both the nucleic acid sequence encoding the biomolecule and the region encoding the RNAi molecule. Likewise, one HRE or one stretch of multiple HREs is sufficient to control the transcription of both the nucleic acid sequence encoding the biomolecule and the region encoding the RNAi molecule. In this case, the vec tor comprises:

(i) at least one HRE consisting of a nucleic acid sequence selected from the group consisting of SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3, SEQ ID NO.: 4 and functional variants of any one of these sequences;

(ii) a promoter that is active in the cell, wherein the promoter is operably linked to the HRE;

(iii) a nucleic acid sequence encoding the biomolecule and encoding an RNAi molecule that prevents or reduces expression of Von Hippel- Lindau protein, wherein the RNAi molecule is a siRNA, a shRNA or a miRNA, wherein the nucleic acid sequence is operably linked to the promoter.

In a preferred embodiment, the cell is further transformed with a nucleic acid se quence encoding a HIF a subunit, wherein the nucleic acid sequence encoding the HIF a subunit is operably linked to a promoter that is active in the cell. The HIF a subunit preferably is H IF-1 a. In this embodiment, the HIF a subunit is overex pressed in the cell. The increased amount of HIF a subunit leads to an increased formation of heterodimers of the HIF a subunit and the HIF b subunit. The hetero dimers bind to the FIRE(s), thereby further enhancing the transcription of the nu cleic acid sequence encoding the biomolecule. The qualitative effect of overex pressing the HIF a subunit is thus comparable to the effect of the RNAi molecule that prevents or reduces the expression of the Von Flippel-Lindau protein in the cell. Overexpressing the HIF a subunit may thus be applied as an alternative to expressing the RNAi molecule that prevents or reduces the expression of the Von Flippel-Lindau protein.

The embodiments described above with respect to the region encoding the RNAi molecule that prevents or reduces the expression of the Von Hippel-Lindau protein or with respect to the nucleic acid sequence that comprises the region encoding the RNAi molecule likewise apply to the nucleic acid sequence encoding the HIF a subunit.

In particular, in a preferred embodiment, the promoter to which the nucleic acid sequence encoding the HIF a subunit is operably linked, is operably linked to at least one hypoxia-responsive element (HRE) consisting of a nucleic acid se quence selected from the group consisting of SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3, SEQ ID NO.: 4 and functional variants of any one of these se quences. In this embodiment, the expression of the HIF a subunit is regulated in the same manner as the production of the biomolecule, i.e. its expression is in creased under hypoxic conditions. Accordingly, hypoxic conditions lead to (i) an increase of the transcription of the nucleic acid sequence encoding the biomole cule, and (ii) an increased amount of HIF a subunit. The increased amount of HIF a subunit leads to an increased formation of heterodimers of the HIF a subunit and the HIF b subunit. The heterodimers bind to the HRE(s), thereby further en hancing the transcription of the nucleic acid sequence encoding the biomolecule as well as the transcription of the nucleic acid sequence encoding the HIF a subu- nit. Accordingly, in this way, a positive feedback loop is established that has an enhancing effect on biomolecule production under hypoxic conditions.

The positive feedback loop leads to a higher production yield at a given level of oxygen. This also means that the same production yield can be reached at higher levels of oxygen in the cell culture.

The embodiments described above with respect to the HRE that controls the tran scription of the nucleic acid sequence encoding the biomolecule likewise apply to the HRE that controls the transcription of the nucleic acid sequence encoding the HIF a subunit.

In a preferred embodiment, the promoter to which the nucleic acid sequence en coding the HIF a subunit is operably linked is directly upstream of the nucleic acid sequence encoding the HIF a subunit and the HRE is directly upstream of the promoter.

Overexpression of the HIF a subunit may also be used in combination with ex pression of the RNAi molecule that prevents or reduces the expression of the Von Hippel-Lindau protein. Accordingly, in a preferred embodiment, the cell is further transformed with both

(i) the nucleic acid sequence comprising the region encoding the RNAi mole cule that prevents or reduces expression of Von Hippel-Lindau protein, wherein the RNAi molecule is a siRNA, a shRNA or a miRNA, wherein the region encoding the RNAi molecule is operably linked to a promoter that is active in the cell, and

(ii) the nucleic acid sequence encoding the HIF a subunit, wherein the nucleic acid sequence encoding the HIF a subunit is operably linked to a promoter that is active in the cell.

In a preferred embodiment, the biomolecule is a recombinant peptide, a recombi nant protein, a recombinant virus, a recombinant virus-like particle or a recombi nant mRNA, wherein the recombinant protein preferably is an antibody. The biomolecule preferably is a biopharmaceutical. Biopharmaceuticals include, for example, recombinant peptides, recombinant proteins, recombinant viruses, recombinant virus-like particles and recombinant mRNAs.

The recombinant protein preferably is an antibody, a derivative of an antibody, a fragment of an antibody, a cytokine, a hormone, a growth factor, a transcription factor, a ribosomal protein, an enzyme or a vaccine.

The recombinant virus preferably is an oncolytic virus or a virus used for gene therapy.

Each of the recombinant virus, the recombinant virus-like particle and the recom binant mRNA may preferably be a vaccine. A virus-like particle (VLP) closely re sembles a virus, but is non-infectious because it comprises no viral genetic mate rial. It can be produced by expressing viral structural proteins, which can then self- assemble into the VLP.

The biomolecule may be an amino acid such as, for example, cysteine, serine or threonine.

The biomolecule may be a vitamin.

The biomolecule may be a coding or non-coding RNA molecule. The non-coding RNA molecule may be, for example, a small non-coding RNA (such as, for exam ple, a siRNA, a shRNA or a miRNA), a long non-coding RNA (IncRNA) or an RNA aptamer. The coding RNA molecule may be, for example, a recombinant mRNA. The RNA molecule may be an RNA-based therapeutic.

The biomolecule may be released from the cell into the cell culture medium, i.e. into the medium in which the cell is propagated. The biomolecule may be released from the cell and isolated from the cell culture as part of an extracellular vesicle (EV) that is released from the cell. An example of such a biomolecule is a recom binant membrane protein. The biomolecule may be released from the cell as car go of an EV that is released from the cell. An example of such an EV is an EV with a coding or non-coding RNA molecule as cargo. The EV may be, for example, a recombinant exosome. The biomolecule may be a single gene product (e.g. a single protein) or the bio molecule may be composed of more than one gene product. In the latter case, the nucleic acid sequence encoding the biomolecule encodes more than one gene product. For example, in case the biomolecule is a virus-like particle, the nucleic acid sequence encoding the biomolecule encodes more than one viral protein. The number of gene products encoded by the nucleic acid sequence encoding the biomolecule is limited only by the size of nucleic acid that can be inserted into the vector.

The biomolecule may also be a molecule that is not intended for a pharmaceutical use. For example, the biomolecule may be a recombinant protein that can be used as an additive or substitute in food and/or animal feed. An example of such a protein is casein.

In a preferred embodiment, the cell culture is a perfusion cell culture, a batch cell culture or a fed-batch cell culture. These types of cell cultures allow high cell den sities, i.e. cell densities of 15x10 ⁶ cells/ml or more. For example, perfusion cell cultures can have cell densities of 15x10 ⁶ cells/ml to 100x10 ⁶ cells/ml. A high cell density cell culture is of advantage for the method of the invention since due to the high cell density, it is typically accompanied by hypoxic conditions, which lead to an increase in the production of the biomolecule. In other words, the method of the invention is particularly beneficial for high cell density cell culture since it ex ploits high cell density-induced hypoxia.

In perfusion cell culture, cells are retained inside the bioreactor or cells are recy cled back to the bioreactor. Fresh cell culture medium is continuously added to the bioreactor and cell-free supernatant comprising the product of interest (i.e. the biomolecule), waste products and depleted cell culture medium is continuously removed from the bioreactor at the same rate. Perfusion cell cultures can have considerably higher cell densities than batch cell cultures.

The inventors performed a mock-perfusion cell culture using the method of the invention by replacing the cell culture medium daily. The inventors surprisingly observed a significantly better performance of the cells in comparison to a normal cell culture of the cells. The increase of the specific productivity of the cells during the mock-perfusion conditions was more than 2-fold compared to the normal cul ture of the cells.

In batch cell culture, all nutrients are provided at the beginning of the cultivation, i.e. no nutrients are added during the subsequent cultivation process.

In fed-batch cell culture, one or more nutrients are supplied to the bioreactor dur ing the cultivation, while the product of interest remains in the bioreactor until the end of the run.

The fed-batch cell culture preferably is an N-1 fed-batch cell culture. In N-1 fed- batch cell culture, a step of intensification of cell growth is performed in an N-1 bioreactor prior to the cell culture in the production bioreactor (N). This facilitates a higher starting cell density in the production bioreactor and shortens the produc tion bioreactor run time.

In a preferred embodiment, the cell culture has a cell density of at least 15x10 ⁶ cells/ml, preferably at least 50x10 ⁶ cells/ml. A cell culture with a cell densi ty of 15x10 ⁶ cells/ml or more, in particular with a cell density of 50x10 ⁶ cells/ml or more, typically shows oxygen deficiency, which leads to an increase in the produc tion of the biomolecule.

In a preferred embodiment, the cell culture is exposed to a hypoxic condition, wherein the hypoxic condition preferably is a level of oxygen from 0.1 % to 15 %, further preferred from 0.1 % to 10 %, further preferred from 1 % to 5 %, most pre ferred 1 %. The level of oxygen can also be referred to as the concentration of oxygen. The hypoxic condition leads to an increase in the production of the bio molecule. The method of the invention allows applying an oxygen shift to the cell culture, namely an oxygen shift from a normoxic condition (e.g. 40 % or 60 % oxy gen) to the hypoxic condition (e.g. 5 % or 1 % oxygen), by which biomolecule pro duction will be increased at any cell density, i.e. at any time point of the production process. This requires that the oxygen level in the cell culture can be tightly con trolled, which is usually the case when biomolecules are produced. The oxygen shift (hypoxic shift) preferably is applied after the cells have reached the static cul tivation phase. The oxygen shift can be applied in combination with the tempera ture shift that is known from conventional production processes. The HRE and the cell can be derived from the same species.

Alternatively, the HRE can be from a different species than the cell. An example for such a cross-species embodiment is the use of the human HRE sequence of SEQ ID NO.: 1 in a CHO cell.

In a preferred embodiment, the HRE consists of a nucleic acid sequence selected from the group consisting of SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3 and SEQ ID NO.: 4.

In a particularly preferred embodiment, the HRE consists of the nucleic acid se quence of SEQ ID NO.: 1.

In a further preferred embodiment, the HRE consists of the nucleic acid sequence of SEQ ID NO.: 1 and the cell is a CHO cell.

The HRE can consists of a nucleic acid sequence that is a functional variant of any one of SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3 and SEQ ID NO.: 4. The functional variant has one or more alterations compared to any one of SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3 or SEQ ID NO.: 4, wherein the alterations do not alter the ability of the sequence to induce gene expression in response to hypoxia. The alterations preferably are nucleotide exchanges. The one or more alterations preferably are not in the conserved HIF binding site, which is the most important sequence element of the HRE sequence. Counting the nu cleotides in in 5’ -> 3’ direction, the conserved binding site of HIF is formed by nu cleotides 11 to 18 of SEQ ID NO.s: 1-4 (i.e. by the nucleotides at positions 11 to 18) (underlined in Figure 1). Accordingly, the one or more alterations preferably concern nucleotides 1 to 10 and/or nucleotides 19 to 35 of SEQ ID NO.s: 1-4.

In addition to the HIF binding site, the nucleotides at positions 2 to 6 and at posi tions 24 to 28 may be important for the ability of the sequence to induce gene ex pression in response to hypoxia. Accordingly, in a preferred embodiment, the one or more alterations concern nucleotide 1, nucleotides 7 to 10, nucleotides 19 to 23 and/or nucleotides 29 to 35 of SEQ ID NO.s: 1-4.

In a particularly preferred embodiment, the one or more alterations are nucleotide exchanges of nucleotide 1, nucleotide 19, nucleotide 21, nucleotide 23, nucleotide 31 and/or nucleotide 35 of SEQ ID NO.s: 1-4. These nucleotides partially vary be tween SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3 and SEQ ID NO.: 4 (Figure 1). In other words, these nucleotides are not conserved. It can thus be reasonably assumed that an exchange of one or more of these nucleotides will not affect the ability of the sequence to induce gene expression in response to hypoxia.

Examples of functional variants of SEQ ID NO.s: 1-4 according to the particularly preferred embodiment described above are given in the following with nucleotide exchanges being underlined. All sequences are indicated in 5’ -> 3’ direction.

Preferred functional variants of SEQ ID NO.: 1:

ACACAGTGCATACGTGGGCTCCAACAGGTCCTCTT (SEQ ID NO.: 5) CCACAGTGCATACGTGGGTTCCAACAGGTCCTCTT (SEQ ID NO.: 6) CCACAGTGCATACGTGGGCTTCAACAGGTCCTCTT (SEQ ID NO.: 7) CCACAGTGCATACGTGGGCTCCCACAGGTCCTCTT (SEQ ID NO.: 8)

C CAC AGT G CAT ACGTGGGCTC C AAC AG GT C GT CTT (SEQ ID NO.: 9) CCACAGTGCATACGTGGGCTCCAACAGGTCCTCTC (SEQ ID NO.: 10) CCACAGTGCATACGTGGGCTCCAACAGGTCGTCTC (SEQ ID NO.: 11) CCACAGTGCATACGTGGGCTCCCACAGGTCGTCTC (SEQ ID NO.: 12) ACACAGTGCATACGTGGGCTTCCACAGGTCGTCTC (SEQ ID NO.: 13) SEQ ID NO.: 2 SEQ ID NO.: 3 SEQ ID NO.: 4

Preferred functional variants of SEQ ID NO.: 2:

CCACAGTGCATACGTGGGTTTCCACAGGTCGTCTC (SEQ ID NO.: 14) ACACAGTGCATACGTGGGCTTCCACAGGTCGTCTC (SEQ ID NO.: 13) ACACAGTGCATACGTGGGTTCCCACAGGTCGTCTC (SEQ ID NO.: 15) ACACAGTGCATACGTGGGTTTCAACAGGTCGTCTC (SEQ ID NO.: 16) ACACAGTGCATACGTGGGTTTCCACAGGTCCTCTC (SEQ ID NO.: 17) ACACAGTGCATACGTGGGTTTCCACAGGTCGTCTT (SEQ ID NO.: 18) ACACAGTGCATACGTGGGCTCCCACAGGTCGTCTC (SEQ ID NO.: 19) ACACAGTGCATACGTGGGTTCCAACAGGTCGTCTC (SEQ ID NO.: 20) ACACAGTGCATACGTGGGCTCCAACAGGTCGTCTC (SEQ ID NO.: 21) CCACAGTGCATACGTGGGCTCCAACAGGTCGTCTC (SEQ ID NO.: 11) ACACAGTGCATACGTGGGTTTCCACAGGTCCTCTT (SEQ ID NO.: 22) SEQ ID NO.: 1 SEQ ID NO.: 3

SEQ ID NO.: 4

Preferred functional variants of SEQ ID NO.: 3:

ACACAGTGCATACGTGGGCTTCCACAGGTCGTCTC (SEQ ID NO.: 13) CCACAGTGCATACGTGGGTTTCCACAGGTCGTCTC (SEQ ID NO.: 14) CCACAGTGCATACGTGGGCTCCCACAGGTCGTCTC (SEQ ID NO.: 12) CCACAGTGCATACGTGGGCTTCAACAGGTCGTCTC (SEQ ID NO.: 23) CCACAGTGCATACGTGGGCTTCCACAGGTCCTCTC (SEQ ID NO.: 24) CCACAGTGCATACGTGGGCTTCCACAGGTCGTCTT (SEQ ID NO.: 25) CCACAGTGCATACGTGGGCTCCAACAGGTCGTCTC (SEQ ID NO.: 11) ACACAGTGCATACGTGGGCTCCAACAGGTCGTCTC (SEQ ID NO.: 21) ACACAGTGCATACGTGGGCTTCCACAGGTCCTCTT (SEQ ID NO.: 26) SEQ ID NO.: 1

SEQ ID NO.: 2 SEQ ID NO.: 4

Preferred functional variants of SEQ ID NO.: 4:

ACACAGTGCATACGTGGGCTTCCACAGGTCCTCTT (SEQ ID NO.: 26) CCACAGTGCATACGTGGGTTTCCACAGGTCCTCTT (SEQ ID NO.: 27) CCACAGTGCATACGTGGGCTCCCACAGGTCCTCTT (SEQ ID NO.: 8) CCACAGTGCATACGTGGGCTTCAACAGGTCCTCTT (SEQ ID NO.: 7) CCACAGTGCATACGTGGGCTTCCACAGGTCGTCTT (SEQ ID NO.: 25) CCACAGTGCATACGTGGGCTTCCACAGGTCCTCTC (SEQ ID NO.: 24) ACACAGTGCATACGTGGGCTCCAACAGGTCCTCTT (SEQ ID NO.: 5) ACACAGTGCATACGTGGGCTTCCACAGGTCGTCTC (SEQ ID NO.: 13) SEQ ID NO.: 1 SEQ ID NO.: 2

SEQ ID NO.: 3

The preferred functional variants of any one of SEQ ID NO.s: 1-4 are also pre ferred functional variants for the remaining three sequences of SEQ ID NO.s: 1-4. In other words, the preferred functional variants of SEQ ID NO.: 1 are the variants of SEQ ID NO.s: 5-27 and SEQ ID NO.: 2, SEQ ID NO.: 3 and SEQ ID NO.: 4; the preferred functional variants of SEQ ID NO.: 2 are the variants of SEQ ID NO.s: 5- 27 and SEQ ID NO.: 1 , SEQ ID NO.: 3 and SEQ ID NO.: 4; the preferred function al variants of SEQ ID NO.: 3 are the variants of SEQ ID NO.s: 5-27 and SEQ ID NO.: 1, SEQ ID NO.: 2 and SEQ ID NO.: 4; and the preferred functional variants of SEQ ID NO.: 4 are the variants of SEQ ID NO.s: 5-27 and SEQ ID NO.: 1, SEQ ID NO.: 2 and SEQ ID NO.: 3.

Disclosed is a cell for producing a biomolecule in vitro, wherein the cell is trans formed with a vector, wherein the vector comprises (i) at least one hypoxia-responsive element (HRE) consisting of a nucleic acid sequence selected from the group consisting of SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3, SEQ ID NO.: 4 and functional variants of any one of these sequences;

(ii) a promoter that is active in the cell, wherein the promoter is operably linked to the HRE;

(iii) a nucleic acid sequence encoding the biomolecule, wherein the nucleic acid sequence is operably linked to the promoter, wherein the cell is a eukaryotic cell, wherein the cell is not a tumor cell.

The embodiments described above with respect to the cell and the vector used in the method of the invention likewise apply to the cell of the disclosure. In particu lar, the cell may be any cell that is suitable for the biotechnological production of the biomolecule of interest. The cell is a eukaryotic cell. The cell may be a human cell or a non-human cell. Many human and non-human cell lines that are suitable for the biotechnological production of biomolecules are known.

Under hypoxic conditions, the cell of the invention increases its production of the biomolecule by increasing the transcription of the nucleic acid sequence encoding the biomolecule.

The cell is not a tumor cell. The term “tumor cell” as used herein refers to a tumor cell that is transformed with a vector in the context of a gene therapy of a cancer.

In a preferred embodiment, the cell is a Chinese hamster ovary (CHO) cell, a de rivative of a CHO cell, a SL-29 cell, a duck embryo cell, a Vero cell, a NIH/3T3 cell, a BHK-21 cell, a Sp2/0-Ag14 cell, a BTI-Tn-5B1-4 cell, a hybridoma cell, a Human Embryonic Kidney 293 (HEK293) cell, a derivative of a HEK293 cell, a CAP cell or a CAP-T cell, wherein the derivative of the CHO cell preferably is a CHO-K1 cell, a CHO-EBNA cell, a CHO-EBNA GS cell, a CHO-DG44 cell, a CHO- DXB1 1 cell or a CHO-S cell, and wherein the derivative of the HEK293 cell prefer ably is a HEK293F cell or a HEK293S cell. Further aspects of the invention will be apparent to the person skilled in the art by the enclosed description of the examples, in particular the scientific results.

Examples

Materials and methods

Cell culture

CHO (Chinese hamster ovary) cells were cultured in animal component free SFM4CHO medium (GE Flealthcare, Chicago, IL, USA), supplemented with 4 mM L-Glutamine (Lonza, Basel, Switzerland) and 10 g/L glucose (Roth, Karlsruhe, Germany). Cultivation was performed at 140 rpm (25 mm orbit) with 5 % CO2 and 85 % humidity at 37 °C. Cells have been passaged every 3-4 days to a viable cell density (VCD) of 0.5 x 10 ⁶ cells/mL. VCD and viability have been determined via trypan blue exclusion using CEDEX XS (Roche Diagnostics, Mannheim, Germa ny). Transfection of CFIO cells was performed using 15 pg vector using the NEON transfection kit (Thermo Fisher, Waltham, MA, USA). Cell lines stably expressing variants of the pEF-myc-cyto-mCMV-d2GFP (Addgene, Watertown, MA, USA) vector have been selected using 500 pg/mL G418 (Genaxxon, Ulm, Germany). To evaluate the response of CFIO cells on undefined hypoxic cultivation conditions, CFIO cells have been inoculated with a density of 0.5 x 10 ⁶ cells/mL and cultivated static with 5 % CO2 and 85 % humidity at 37 °C. For cultivation at defined O2 con centrations, CFIO cells were inoculated in 1 L at a VCD of 0.3 x 10 ⁶ cells/mL using a 2 L stirred tank bioreactor (Sartorius, Gottingen, Germany). The bioreactors were controlling pFH at 7.15, temperature at 37 °C, stirring speed at 100 rpm and O2. The O2 set point was adjusted every 24 h.

Fed-batch fermentation

Fed-batch fermentation (fed-batch cell culture) was performed in a 2 L Biostat benchtop bioreactor (Sartorius Stedim Biotech GmbFI, Goettingen, Germany). 3x10 ⁵ cells/mL were inoculated in 1 L SFM4CFIO medium (GE Flealthcare, Chica go, IL, USA), supplemented with 4 mM L-Glutamine (Lonza, Basel, Switzerland) and 10 g/L glucose (Roth, Karlsruhe, Germany). Fermentation was performed at 37 °C, pFH 7.15, stirring speed at 100 rpm and an initial O2 concentration of 40 %. During the process, the glucose concentration was maintained at 5 g/L and L- Glutamine at 2 mM. The main feed (1 L) consisted of 11.6 g HyClone Cell Boost 6 solved in 330 ml_ Millipore water (Cytiva, Marlborough, MA, USA), 20 g/L glucose and 660 mL SFM4CHO. The glutamine feed consisted of 100 mL 200 mM L- glutamine and 100 mL SFM4CHO. Viable cell density and viability was determined daily via trypan blue exclusion and after reaching the static cultivation phase, tem perature was reduced to 34 °C and O2 concentration in one fermenter to 1 %. Every day a sample was taken to determine SEAP concentration.

Molecular biology

The starting plasmid 5HRE/GFP was a gift from Martin Brown & Thomas Foster (Addgene plasmid #46926) and has been described in Vordermark et al. 2001. Constructs were cloned using the following HRE sequences of the human vascu lar endothelial growth factor A (VEGF-A) and erythropoietin (EPO) genes (se quences are indicated in 5’ -> 3’ direction):

HRE sequence of the human VEGF-A gene:

CCACAGTGCATACGTGGGCTCCAACAGGTCCTCTT (SEQ ID NO.: 1).

HRE sequence of the human EPO gene:

GCCCTACGTGCTGTCTCACACAGCCTGTCTGAC (SEQ ID NO.: 28)

The oligonucleotides used are listed in Table 2 (Fw, forward; Rev, reverse; VEGF, VEGF-A). For cloning, oligonucleotides have been annealed by slowly reducing temperature from 95 °C to 25 °C, phosphorylated using T4 PNK (New England Biolabs, Ipswich, MA, USA) and ligated by T4 ligase (New England Biolabs, Ips wich, MA, USA). Randomly ligated double stranded oligonucleotides have been cloned into the pEF-myc-cyto-mCMV-d2GFP after removing the 5HREs by digest ing with Xhol and Bglll (both New England Biolabs, Ipswich, MA, USA).

Table 2: Oligonucleotide sequences for molecular biology

To generate the vector for SEAP production, a DNA fragment comprising “5HRE- CMV-SEAP” was generated via gene synthesis and cloned into a suitable vector. 5HRE denotes five copies of the HRE sequence of the human VEGF-A gene, CMV denotes the complete CMV promoter and SEAP denotes the nucleic acid sequence encoding SEAP. Subsequently, the dihydrofolate reductase (DHFR) cassette was excised from the pOptiVEC-TOPO vector (Thermo Fisher Scientific, Darmstadt, Germany) and cloned into the “5HRE-CMV-SEAP”-comprising vector for selection purposes. The DHFR cassette was necessary for stable SEAP ex pression in CFIO-DG44 (DFIFR deficient) cells.

RNA isolation

RNA isolation was performed using the miRNeasy Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol with 5 x 10 ⁶ cells. Purity and concentra tion of the isolated RNA was assessed using a Nanodrop 1000 spectrophotometer by absorbance at 260 nm (Thermo Fisher Scientific, Darmstadt, Germany).

PCR and RT-PCR

After RNA isolation, cDNA was written using the Transcriptor High Fidelity cDNA Synthesis Kit (Roche Diagnostics, Mannheim, Germany) according to the manu facture’s protocol. The transcription of HIF-1 a, HIF-1 b and VFIL was demonstrated performing PCR on cDNA level with the following primers (sequences are indicat ed in 5’ -> 3’ direction; Fw denotes the forward primer; Rev denotes the reverse primer):

HIF-10 Fw TCCAGTTGCGCTCCTTTGAT (SEQ ID NO.: 43) HIF-10 Rev AT C C ATT GATT G C C C C AG C A (SEQ ID NO.: 44) HIF-1 b Fw TTGTCTCGTGTGAGACTGGC (SEQ ID NO.: 45) HIF-1 b Rev C C AAT G G C C ACT AG G C AG AA (SEQ ID NO.: 46) VHL Fw GGTCATCTTCTGCAACCGCA (SEQ ID NO.: 47) VHL Rev AAGCTGGAATTCAGGACCACT (SEQ ID NO.: 48)

Amplified DNA was separated via gel electrophoresis and detected using Fusion FX imager (Vilber Lourmat, Eberhardszell, Germany). Quantitative real-time PCR on d2GFP was performed using the following primers (sequences are indicated in 5’ -> 3’ direction; Fw denotes the forward primer; Rev denotes the reverse primer): d2GFP Fw CTACCCCGACCACATGAAGC (SEQ ID NO.: 49) d2GFP Rev AAGTCGATGCCCTTCAGCTC (SEQ ID NO.: 50) with GAPDH as loading control:

GAPDH Fw G ACT CTAC C CAT G G C AAGTT C A (SEQ ID NO.: 51)

GAPDH Rev TCGCTCCTGGAAGATGGTGATG (SEQ ID NO.: 52)

The measurement with SybrGreen Mastermix (Genaxxon, Ulm, Germany) was recorded using a Lightcycler480 (Roche Diagnostics, Mannheim, Germany).

Flow cytometry analysis

GFP fluorescence of hypoxic and normoxic cultivated CHO cells was measured using MACSQuant Analyzer 10 without further staining (Miltenyi Biotec, Bergisch Gladbach, Germany). Resulting data was analyzed via the MACSQuantify soft ware.

Immunoblotting

Cells were lysed in RIPA buffer (1 % v/v NonidetTM P40, 0.5 % w/v sodium deox- ycholate, ethylenediaminetetraacetic acid 1mM, all agents Carl Roth, Karlsruhe, Germany). Protein concentration was assed via BCA-assay. The sodium dodecyl sulfate polyacrylamide gel electrophoresis was performed using TruPAGE Precast 4-20 % gradient gels (Sigma-Aldich, St. Louis, MO, USA) with 30 pg protein per lane. Blotting of proteins was done on polyvinylidene fluoride membrane for 75 min with 60 mA and immunoblotting with the following antibodies was per formed over night at 4 °C: anti-HIF1 -alpha antibody (Bio-Techne, Minneapolis, MN, USA, #NB100-479SS, 1:1000); anti-VHL antibody (Santa Cruz Biotechnolo gy, Dallas, TX, USA, #sc-135657 HRP, 1:200); anti-ARNT antibody (anti-HIF-1 b antibody) (Santa Cruz Biotechnology, Dallas, TX, USA, #sc-17811, 1:200) and anti-p-actin antibody (Sigma-Aldrich, St. Lousis, MO, USA, #A5441-2ML, 1:1000). The secondary antibodies (anti-mouse IgG-HRP (Sigma Aldrich, St. Louis, MO, USA, #A4416; 1:5000); anti-rabbit IgG-HRP F(ab’)2 (Jackson ImmunoResearch, Cambridge, United Kingdom, #711-036-152, 1:50000)) were incubated for 1 h at room temperature. Detection was performed using Immobilon® Western chemi luminescent HRP substrate (Millipore Corporation, Billerica, MA, USA) and Fusion FX imager (Vilber Lourmat, Eberhartszell, Germany). SEAP assay

Secreted embryonic alkaline phosphatase (SEAP) is a reporter widely used to study promoter activity or gene expression. To determine SEAP concentration in the SEAP reporter gene assay, chemiluminescent (Roche, Mannheim, Germany) was used according to the manufacturer’s advice. Samples were heat inactivated at 65 °C for 30 min and have been diluted 1:50 to 1:300 depending on the SEAP concentration to fit to the linear area of the standard curve. Afterwards, 50 pL heat inactivated sample have been pipetted into a black non-binding flat bottom 96-well plate (Greiner Bio One, Frickenhausen, Germany) and incubated for 5 min with inactivation buffer and freshly prepared substrate was added. After 10 min incuba tion while gently shaking, luminescence was measured with 1 s integration time via SpectraMacs (Molecular Devices, San Jose, CA, USA).

Statistical analysis

Statistical differences between data was determined using GraphPad Prism 6 software. ONE-WAY-ANOVA test with Bonferroni correction was conducted to calculate statistical differences of the mean between groups. All values of groups are shown as mean ± standard deviation (SD).

Results

Before the scientific results are described in detail, it is referred to Figure 1 , which shows the sequences of the FIRE of the VEGF-A gene of various species in 5’ -> 3’ direction. The conserved binding site of H IF is underlined. It can be seen that the binding site is identically present in all sequences shown. It can further be seen that the sequences are highly conserved. The sequences differ between the species, if at all, only by slight modifications at a few specific positions that have been marked by vertical boxes for illustration purposes. The positions are nucleo tide 1, nucleotide 19, nucleotide 21 , nucleotide 23, nucleotide 31 and/or nucleotide 35.

The experimental results will now be described in detail. Presence and conservation of the hypoxia-responsive pathway in CHO cells

Besides hypoxia response elements (HREs) of hamster origin, HREs from other species such as human or mouse could serve as building blocks for the construc tion of hypoxia-responsive vectors to selectively enhance protein expression under hypoxic culture conditions in Chinese hamster ovary (CHO) cells. Among others, suitable elements to induce expression in CHO cells under hypoxic conditions could include HREs from the erythropoietin (EPO) and vascular endothelial growth factor A (VEGF-A) gene. To verify the potential binding of hamster proteins to hu man HREs in CHO cells, protein sequences of the key binding protein HIF-1a of human and hamster were aligned and analyzed for conservation of crucial do mains (Figure 2A). For human HIF-1a three isoforms and for the hamster HIF-1a two isoforms displayed over 77 % conservation. Notably, critical areas like the DNA-binding domain, the H IF-1 b heterodimerization domain and the nuclear local ization signal were 100 % similar across all isoforms (Figure 2A). Especially the identical DNA-binding domain may allow the hamster HIF-1a to bind to human HREs introduced into CHO cells.

For the production of biopharmaceuticals, mainly CHO-K1 and CHO-DG44 cells are industrially used. To verify the suitability of these cell lines to introduce exoge nous HREs, the expression of the essential hypoxia-related genes HIF-1a, H IF-1 b and VHL was examined on RNA and protein level. For both cell lines HIF-1a and H IF-1 b expression was detected using RT-PCR analysis (Figure 2B) as well as VHL, the antagonist to HIF-1a (Figure 2C). Finally, the protein expression of HIF- 1a, H IF-1 b and VHL was analysed via Western Blot (Figure 2D). While both VHL and H IF-1 b were detected in both CHO-K1 and CHO-DG44 cultured under normoxic conditions, the instable HIF-1a could only be detected in CHO-DG44 cell lysates isolated from hypoxic cultivated cell cultures at around 100 kDa (Figure 2D). In summary, these data indicate the presence and conservation of key play ers of the hypoxia sensing pathway in CHO production cells which might therefore allow for cross species cell engineering. Vector development to enable hypoxia-inducible protein expression

To create and validate a hypoxia-inducible expression system, several vectors were designed carrying a minimal CMV (mCMV) promotor flanked upstream by HRE sequences and controlling the expression of the unstable d2GFP. HRE se- quences were derived from hypoxia sensitive human genes EPO and VEGF-A and cloned as oligonucleotides with varying repetitions or without FIREs to provide for a mock vector. A fragment of the vectors with HRE sequences is schematically shown in Figure 3A, in which 2-9 HREs denotes 2 to 9 copies of the HRE se quence of the human VEGF-A gene or the human EPO gene, mCMV denotes the minimal CMV promoter and d2GFP denotes the nucleic acid sequence encoding d2GFP. The HRE repetitions have a distance of six base pairs between adjacent HRE repeats. For all generated vectors (Table 3), stable cell pools for CHO-K1 and CHO-DG44 cell lines were subsequently generated. Table 3: Generated vector constructs for the analysis of HRE functionality Induction of GFP expression under oxygen deprivation conditions

To verify the functionality of the resulting cell lines expressing GFP under the con trol of hypoxia-inducible elements, cell lines were initially cultivated in small scale batches under normal shaken and oxygen deprivation static conditions. Induced GFP expression was measured daily using flow cytometry. CFIO-K1 cells express ing the mock vector showed no induction of GFP expression with a comparable minimal mean-fluorescence of around 0.3 - 0.5 under shaken and static cultivation between 24 h to 78 h (Figure 4A; in Figure 4, grey bars stand for shaken, while black bars represent static cultured pools). Similarly, CFIO-K1 cells stably express ing the 2FIRE-VEGF, 5FIRE-VEGF or 8FIRE-VEGF construct and cultivated under shaken normoxic conditions did not induce GFP expression displaying as well a GFP-mean fluorescence of around 0.3 - 0.5 (Figure 4A). In contrast, static cul tured CFIO-K1 cells showed a strongly increased GFP expression with mean fluo rescence of 0.7 (2HRE-VEGF), 1.0 (5HRE-VEGF) and 2.0 (8HRE-VEGF) over time (Figure 4A). Comparable or even slightly pronounced induced expression results were obtained with CFIO-DG44 cells, where mock expressing or shaken cultured cells showed no induced GFP expression with values between 0.3-0.6, whereas a strong increase in fluorescence intensity could be detected with stati cally cultivated 2HRE-VEGF (1.4), 5HRE-VEGF (3.0) and 8HRE-VEGF (2.4) CHO- DG44 cells (Figure 4C). When using CFIO-K1 cells expressing hypoxia inducible constructs with 3, 4 or 9 FIREs derived from the EPO gene, only a slight increase in GFP expression was observed for static (0.4 - 0.7) compared to mock or shak en (0.2 - 0.5) cultured cells (Figure 4B). CFIO-DG44 cells on the other hand showed a remarkable induction of GFP expression (1.5 - 2.2) containing EPO- HRE constructs at static, hypoxia imitating conditions in comparison to the shaken cultured control cells (0.3 - 0.6) (Figure 4D). These data indicated that the VEGF- HRE constructs performed significantly better in both cell lines than the EPO-FIRE counterparts. In addition, CFIO-DG44 cells responded stronger to static and hy poxic cultivation with both VEGF- and EPO-FIRE constructs than CFIO-K1 cells. When using repetitive FIRE sequences, a correlation between FIRE number and expression response was observed, which was saturated between 5-8 FIRE repe- titions. Therefore, the 5HRE-VEGF construct was chosen for further analysis due to its strong induction of protein expression under oxygen limitation.

Scale-up of batch fermentation under defined oxygen concentrations

After identifying the most hypoxia-responsive vector regarding type and number of HRE elements, the performance of the 5HRE-VEGF vector construct was further characterized in a scale up batch fermentation (batch cell culture) under defined oxygen concentrations. Both stable mock and 5FIRE-VEGF expressing CFIO-K1 and CFIO-DG44 cell pools were cultivated as batch in a 2L-fermenter under normoxic conditions (60 % oxygen) for 156 hours. In addition, all cell lines were cultured in a batch fermentation where oxygen concentrations were switched to hypoxia starting after 12 hours and continuing in a stepwise manner (Figure 5A). For all fermentations, potential hypoxia-induced GFP fluorescence was recorded at regular intervals by flow cytometry. CFIO-DG44 Mock and 5FIRE-VEGF cells displayed identical mean fluorescence over the first 84 h of cultivation (Figure 5B). Flowever, after a reduction to 5 % O2 the CFIO-DG44-5FIRE-VEGF cells showed an induction of the GFP signal to about 0.6, which increased additionally signifi cantly to 1.5 mean fluorescence at an oxygen concentration of 1 %. In contrast, the control cells showed continuously decreasing values from about 0.4 after 12 h to 0.15 after 156 h throughout the fermentation (Figure 5B). When comparing the mean fluorescence of normoxic and hypoxic CFIO-DG44 Mock and 5FIRE-VEGF cells, a significant nearly 10-fold induction of the GFP signal could be detected for CFIO-DG445FIRE-VEGF cells in comparison to the control cell lines (Figure 5C).

For CFIO-K1 mock and CFIO-K1-5FIRE-VEGF cells, the cultivation under both normoxic and hypoxic conditions showed for the first 84 h hours of batch fermen tation cultured a comparable pattern of the fluorescence signal of about 0.5 de creasing slightly but continuously to a value of about 0.25 (Figure 5D). After a de crease to 5 % oxygen, CFIO-K1-5FIRE-VEGF cultured cells revealed a slight in crease in mean fluorescence to 0.3, which further increased to 0.5 after a reduc tion to 1 % oxygen. Control cells decreased to about 0.15 at 5 % oxygen and 0.1 at 1 % oxygen (Figure 5E). When comparing the mean fluorescence of all cultured CFIO-K1 cell lines after 140 h a significant stronger GFP expression could be ob- served in hypoxic cultured CHO-K1-5HRE-VEGF cell lines compared to Mock or normoxic 5HRE-VEGF control cell lines (Figure 5E). Interestingly, the hypoxic cul tivation using oxygen concentrations of 5 % or 1 % had no significant negative effect on growth or viability of CFIO-K1 cells (data not shown).

Taken together, these data point towards the successful establishment of a hy poxia-inducible vector system by using the 5FIRE-VEGF constructs to drive protein expression under oxygen limitation.

The observed differences in the extent of induction of protein expression under hypoxic conditions between CFIO-DG44 and CFIO-K1 cells may be due to the fact that the basal expression of the H IF-1 a protein is increased in CFIO-DG44 cells compared to CFIO-K1 cells (Figure 2D). In addition, the inventors found that the stabilization of the H IF-1 a protein under hypoxic conditions is more effective in CFIO-DG44 cells than in CFIO-K1 cells (data not shown). This may explain the more effective induction of protein expression under hypoxic conditions observed in CFIO-DG44 cells compared to CFIO-K1 cells.

To further increase the induction of protein expression under hypoxic conditions, the cells may be further transformed so that (i) the cells overexpress H IF-1 a, or (ii) the expression of the VFIL protein is prevented or reduced. These measures may be particularly interesting for CFIO-K1 cells in order to compensate for their H IF-1 a deficiency.

Increased recombinant protein expression exploiting hypoxia-responsive induction

To demonstrate the potential of FIREs to improve recombinant protein expression, a stable secreted embryonic alkaline phosphatase (SEAP) expressing CFIO-DG44 cell pool was generated with and without the introduction of 5 HREs originated from the VEGF-A gene upstream of the CMV promoter. A fragment of the SEAP expression vector having the 5 HREs is schematically shown in Figure 3B, in which 5HRE denotes five copies of the HRE sequence of the human VEGF-A gene, CMV denotes the complete CMV promoter and SEAP denotes the nucleic acid sequence encoding SEAP. SEAP and 5HRE-SEAP cells have been cultured under normal and static oxygen deprivation conditions. After 72 h SEAP titer was determined in all culture conditions and specific productivity calculated for compar- ison. Thereby, CHO-DG44-SEAP (SEAP) cells lost around 80 % of their specific productivity during hypoxia deprivation conditions in comparison to normally cul tured cells (Figure 6A). In contrast, CHO-DG44-5HRE-SEAP (5HRE-SEAP) cells maintained over 50 % specific productivity during static cultivation and performed significantly better than cells lacking the HREs (Figure 6A). Additionally, the pro duction capacity of SEAP and 5FIRE-SEAP cells during very high cell densities was identified. In this context, both cell lines were inoculated with a normal (0.5 x 10 ⁶ cells/mL) and ultra-high (50 x 10 ⁶ cells/mL) cell density and a mock- perfusion batch was performed by replacing the medium daily. After 3 media re placements the specific productivities of SEAP and 5FIRE-SEAP cells within 24 h were calculated revealing a slight decrease in specific productivity of ultra-high inoculated SEAP cells to around 80 % when compared to the normally inoculated control (Figure 6B). Strikingly, 5FIRE-SEAP cells even increased their specific productivity during mock-perfusion conditions about 2.6-fold when compared to normally cultured 5FIRE-SEAP cells leading to a significantly better performance during the mock-perfusion of cells carrying 5FIREs of the gene VEGF-A (Figure 6B).

To finally validate the potential of FIREs to improve the production capacity of CFIO cells, a fed-batch fermentation of CFIO-DG44-5HRE-SEAP cells was per formed in a 2 L benchtop bioreactor. Cells were cultured under common condi tions (37 °C, 40 % O2) until reaching the static growth phase and then both a tem perature shift to 34 °C alone and the combination of temperature shift and oxygen shift to 1 % was conducted. Although the same cell pool was cultured in both fer menters, the viable cell density (VCD) of one fermenter was slightly lower before the condition shift. However, after reducing temperature or temperature and oxy gen concentration, VCD remained stable for both cultures (Figure 6C). Interesting ly, a nearly identical viability for both culture conditions of >90 % was observed over 12 days, demonstrating the availability of CHO cells to be cultured under strong oxygen deprivation conditions (Figure 6D). By analyzing the SEAP titer of both fermenters a comparable SEAP concentration of around 13 pg/mL was moni tored (Figure 6E) with a similar specific productivities of approximately 1.7 pg/cell/day over the first 6 days of cultivation (Figure 6F). Notably, after shifting temperature or temperature and oxygen concentration SEAP expression was in duced in both fermenters, however, to a significant stronger extent in CHO cells benefiting from having adjusted the oxygen concentration to 1 % (Figures 6E and 6F). In this study, the hypoxia-inducible system led to a SEAP titer of around 460 pg/mL (Figure 6E) and a nearly 2-fold higher specific productivity than non- hypoxia-induced cells by reaching approximately 20 pg/cell/day SEAP expression (Figure 6F).

References

Vordermark D, Shibata T, Brown JM. Green Fluorescent Protein is a Suitable Reporter of Tumor Hypoxia Despite an Oxygen Requirement for Chromophore Formation. Neoplasia. 2001 ;3: 527-534.